Bicarbonate enhances the hydroxylation, nitration, and peroxidation reactions catalyzed by copper, zinc superoxide dismutase. Intermediacy of carbonate anion radical.

The effect of bicarbonate anion (HCO(3)(-)) on the peroxidase activity of copper, zinc superoxide dismutase (SOD1) was investigated using three structurally different probes: 5, 5'-dimethyl-1-pyrroline N-oxide (DMPO), tyrosine, and 2, 2'-azino-bis-[3-ethylbenzothiazoline]-6-sulfonic acid (ABTS). Results indicate that HCO(3)(-) enhanced SOD/H(2)O(2)-dependent (i) hydroxylation of DMPO to DMPO-OH as measured by electron spin resonance, (ii) oxidation and nitration of tyrosine to dityrosine, nitrotyrosine, and nitrodityrosine as measured by high pressure liquid chromatography, and (iii) oxidation of ABTS to the ABTS cation radical as measured by UV-visible spectroscopy. Using oxygen-17-labeled water, it was determined that the oxygen atom present in the DMPO-OH adduct originated from H(2)O and not from H(2)O(2). This result proves that neither free hydroxyl radical nor enzyme-bound hydroxyl radical was involved in the hydroxylation of DMPO. We postulate that HCO(3)(-) enhances SOD1 peroxidase activity via formation of a putative carbonate radical anion. This new and different perspective on HCO(3)(-)-mediated oxidative reactions of SOD1 may help us understand the free radical mechanism of SOD1 and related mutants linked to amyotrophic lateral sclerosis.

Stadtman, Yim, and co-workers (6) originally proposed that SOD1 reacts with H 2 O 2 to form hydroxyl radicals as evidenced by increased hydroxylation of nitrone traps and peroxidation of ABTS chromophore. Subsequently, it was suggested that familial amyotrophic lateral sclerosis-associated SOD1 mutants increased formation of hydroxyl radical upon reaction with H 2 O 2 (7,8). HCO 3 Ϫ stimulated the peroxidase activity of the extracellular SOD activity, which was attributed to an increased formation of hydroxyl radicals (9). Hydroxyl radicals formed from the reaction between SOD1 and H 2 O 2 were suggested to be responsible for the increased cytotoxicity of SOD1 (10).
In the present work, we show that HCO 3 Ϫ enhances (i) the hydroxylation of nitrone spin trap, DMPO, (ii) oxidation and nitration of tyrosine, and (iii) oxidation of the peroxidase probe ABTS to the ABTS radical cation in the presence of SOD1 and H 2 O 2 . We propose that the carbonate anion radical (CO 3 . ) formed from oxidation of HCO 3 Ϫ at the active site of SOD1 by the copper, zinc SOD-bound hydroxyl radical (SOD1-Cu 2ϩ -⅐ OH) is responsible for hydroxylation of DMPO, oxidation and nitration of tyrosine, and oxidation of ABTS. The proposed mechanism offers a new and different perspective on the peroxidative reactions catalyzed by the enzyme SOD1.
EXPERIMENTAL PROCEDURES SOD1 (bovine) was purchased from Roche Molecular Biochemicals. Tyrosine, hydrogen peroxide, sodium bicarbonate, sodium nitrite, and 3-nitrotyrosine were purchased from Sigma. Bio-gel P-2 was obtained from Bio-Rad. DMPO was obtained from Sigma and double distilled to remove the paramagnetic impurity. 17 O-Labeled water (45%) was obtained from ICON Isotopes (Summit, NJ).
Synthesis of 3,3Ј-Dityrosine-Dityrosine was synthesized according to the published literature (11). Briefly, 500 ml of tyrosine (5 mM) in borate buffer (0.1 M, pH 9.1) was mixed with 10 ml of horseradish peroxidase (1 mg/ml) and 1.42 ml of H 2 O 2 (3%) at 37°C for 3 h. The reaction mixture was then mixed with 175 l of ␤-mercaptoethanol and lyophilized to dryness. The lyophilized powder was dissolved in distilled water, and dityrosine was separated from other by-products through a series of chromatography; DEAE-cellulose, Bio-Gel P-2 at neutral pH, and Bio-Gel P-2 at low pH (Ͻ3). Fractions containing dityrosine were pooled, lyophilized, and stored at Ϫ20°C. Dityrosine was identified by comparing its spectral properties with those reported in the literature (11). The purity of dityrosine was verified by both UV and fluorescence HPLC.
Oxidation and Nitration of Tyrosine by SOD1 Peroxidase Activity-Tyrosine (1 mM) was incubated at 37°C with SOD1 (1 mg/ml), H 2 O 2 (1 mM), and NaHCO 3 (25 mM) in sodium phosphate buffer (0.1 M, pH 7.4) containing DTPA (100 M). After incubating for 30 min, samples were centrifuged (10,000 rpm) in an ultrafilter (molecular weight cut-off, 10,000) at 1°C for 5 min to remove SOD1 and subsequently used for HPLC analysis. All reagents were prepared in double-distilled and deionized water. The pH of the incubation mixture remained at 7.40 Ϯ 0.02 before and after the addition of NaHCO 3 . SOD1 activity was measured using the ferricytochrome c reduction assay as described previously (12).
HPLC Analysis-Nitrotyrosine, nitrodityrosine, and dityrosine were * This work was supported by National Institutes of Health Grants RR01008 and HL63119 and by a grant from the Amyothophic Lateral Sclerosis Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡  1 The abbreviations used are: SOD1, copper, zinc superoxide dismutase; DTPA, diethylenetriaminepentaacetic acid; DMPO, 5,5Ј-dimethyl-1-pyrroline N-oxide; DMPO-OH, 5,5Ј-dimethyl-1-pyrroline Noxide-hydroxyl adduct; ABTS, 2,2Ј-azino-bis-[3-ethylbenzothiazoline]-6sulfonic acid; ESR, electron spin resonance; HPLC, high pressure liquid chromatography. separated on an HPLC system equipped with fluorescence and UV detectors. The mobile phase was methanol:phosphate buffer (50 mM, pH 3.0) (4:96) for 30 min. The stationary phase was a C 18 reverse phase column (Partisil ODS-3 250 ϫ 4.6 mm, Alltech). UV detection at 280 nm was used to monitor nitrotyrosine, and fluorescence detection at 284 nm (excitation) and 410 nm (emission) was used to monitor dityrosine and higher oxidation products of tyrosine (13,14). The authentic standards, tyrosine, dityrosine, 3-nitrotyrosine, and nitrodityrosine were eluted at 7. 5, 9.5, 17.5  C, same as in B but without NaHCO 3 . Note that HCO 3 Ϫ is absolutely essential for dityrosine formation. Inset, dityrosine formation as a function of HCO 3 Ϫ concentration. Tyrosine (1 mM) was incubated with SOD1 (1 mg/ml), H 2 O 2 (1 mM) and various amounts of NaHCO 3 at 37°C for 4 h and analyzed by HPLC (n ϭ 3 Ϯ S.D.).   1E). However, the signal intensity of the DMPO-OH adduct (␣ N ϭ ␣ H ϭ 14.9 G) increased with the addition of HCO 3 Ϫ in these incubation mixtures (Fig. 1, A-D). The pH level was measured immediately before and after the addition of 25 mM HCO 3 Ϫ to ensure that the HCO 3 Ϫ -mediated enhancement was not due to a change in the pH of the reaction mixture. The addition of Me 2 SO, a commonly used hydroxyl radical scavenger, had no effect on the ESR signal intensity of DMPO-OH (data not shown). In contrast, when Me 2 SO was added to a mixture containing H 2 O 2 (1%) and DMPO followed by irradiation with UV light, the spectrum caused by the DMPO-OH adduct was replaced by that of the DMPO-methyl radical adduct (␣ N ϭ 16.4 G, ␣ H ϭ 23.4 G) (data not shown). From these results, we conclude that free hydroxyl radicals are not responsible for the hydroxylation of DMPO by SOD1/H 2 O 2 /HCO 3 Ϫ . However, unequivocal proof that H 2 O 2 (the precursor of hydroxyl radical) was not responsible for DMPO-OH formation came from studies using 17 O-labeled water experiments.
To investigate the origin of the oxygen atom in the DMPO-OH adduct, spin trapping experiments were performed in buffers prepared with 17  The Effect of Tyrosine on Hydroxylation of DMPO-Spin trapping data suggest that an oxidant formed from the oxidation of HCO 3 Ϫ by SOD1 and H 2 O 2 is responsible for the hydroxylation of DMPO. We hypothesized that this oxidant is the carbonate radical anion (CO 3 . ) (17). It is well known that CO 3 .
reacts rapidly with phenolic substrates such as tyrosine (18). Thus, we wished to determine whether tyrosine can compete with DMPO in its ability to react with this putative oxidant.
The effect of tyrosine on DMPO-OH formation is shown in Fig.  3 (B-E). A dose-dependent inhibition in DMPO-OH formation was observed with increasing addition of tyrosine to incubations containing SOD1, H 2 O 2 , DMPO, and HCO 3 Ϫ in phosphate buffer. The DMPO-OH signal intensity was almost completely inhibited in the presence of tyrosine (5 mM) (Fig. 3E). This result suggests that tyrosine is able to effectively compete with DMPO for the reaction with HCO 3 Ϫ -derived oxidant. Bicarbonate Enhances Oxidation and Nitration Mediated by the Peroxidase Activity of SOD1- Fig. 4 shows the HPLC profile for dityrosine formation in incubations containing SOD1, tyrosine, H 2 O 2 , and DTPA in the presence and absence of HCO 3 Ϫ . In the absence of HCO 3 Ϫ , no dityrosine was detected (Fig. 4C). In the presence of added HCO 3 Ϫ , dityrosine was detected (Fig. 4B), the level of which increased with increasing HCO 3 Ϫ (Fig. 4, inset). In addition to dityrosine, two minor products (peaks I and II) (Fig. 4B) were identified. Based on the literature report (14), peaks I and II were tentatively attributed to higher oxidation products of tyrosine (e.g. tri-and tetratyrosine).
Next we investigated the effect of HCO 3 Ϫ on SOD1/H 2 O 2 / nitrite-mediated nitration of tyrosine. In the absence of HCO 3 Ϫ , addition of SOD1 to the incubation mixture containing nitrite, H 2 O 2 , and DTPA in phosphate buffer yielded nitrotyrosine, which was measured 2 min after starting the reaction (Fig. 5, bottom trace) (19). In the presence of HCO 3 Ϫ (25 mM), there was a significant increase in nitrotyrosine formation (Fig. 5, middle trace). Fig. 5 (inset) shows the formation of nitrotyrosine as a function of HCO 3 Ϫ concentration. Fig. 6 shows that nitrotyrosine is further oxidized with time in this incubation mixture (Fig. 6A). Nitrotyrosine levels decreased with time, along with a concomitant increase in nitrodityrosine formation (the satellite peak detected at 20 min). After 60 min, a significant fraction of nitrotyrosine was converted to nitrodityrosine.
We assigned the peak appearing at 20 min to nitrodityrosine (Fig. 6) based on the following results. Authentic nitrodityrosine prepared by the addition of peroxynitrite to dityrosine (Fig. 6B, bottom trace) gave a peak that eluted at 20 min, as did incubations containing SOD1, H 2 O 2 , nitrite, and dityrosine (Fig. 6B, top trace). Nitrodityrosine can therefore be used as a unique diagnostic marker product of both oxidation and nitration of tyrosine.
The addition of DMPO (5 mM) markedly inhibited HCO 3 Ϫ -dependent dityrosine and nitrotyrosine formation (data not shown). This result further confirms that both DMPO and tyrosine react with the same oxidant derived from HCO 3 Ϫ .

Bicarbonate Enhances the Oxidation of ABTS to ABTS Cation Radical in the Presence of SOD1 and H 2 O 2 -The oxidation of ABTS to ABTS .
ϩ is conveniently monitored at 415 nm, and this optical change has been used to assay the peroxidase activity of heme proteins (20). As reported previously (5), we confirmed that bicarbonate is needed for peroxidation of ABTS to ABTS . ϩ in the presence of SOD1. Fig. 7A shows the optical changes occurring during SOD1/H 2 O 2 /HCO 3 Ϫ -mediated oxidation of ABTS. The increase in the absorbance of ABTS .
ϩ is detected at 415 nm (Fig. 7A, inset). A time-dependent increase in the absorbance at 415 nm was monitored as a function of HCO 3 Ϫ concentration (Fig. 7B). In the absence of HCO 3 Ϫ , no increase in the absorbance because of ABTS . ϩ was noticed.

DISCUSSION
Bicarbonate-mediated Hydroxylation of DMPO-Most studies on SOD1 and SOD1 familial amyotrophic lateral sclerosis mutant-induced hydroxylation of DMPO were carried out in bicarbonate buffers (6 -8, 16). To our knowledge, Sato et al. (21) were the first to report the finding that SOD1/H 2 O 2 /HCO 3 Ϫ caused a dramatic increase in the hydroxylation of DMPO. The role of bicarbonate anion in these studies was largely ignored. It was suggested that DMPO was oxidized at the active site of SOD1 by a copper-bound hydroxyl radical (i.e. SOD1-Cu 2ϩ -⅐ OH) to DMPO-OH (21). As shown in this and other investigations (5,21,22), HCO 3 Ϫ is absolutely necessary for SOD1/H 2 O 2induced hydroxylation of DMPO. It was proposed that CO 3 .
formed from a one-electron oxidation of HCO 3 Ϫ at the active site of SOD1 could diffuse out of the active site and cause oxidation of substrates in solution (17,22,23). Based on these results, we put forth the following mechanisms for hydroxylation of DMPO (Scheme 1). The proposed mechanisms are consistent with those proposed previously for the hydroxylation of a related nitrone trap by the sulfate radical anion (24). In the present study, experiments using 17 O-labeled H 2 O unambiguously demonstrated that nearly all of the DMPO-OH originated from the addition of water to the DMPO radical intermediate. Although this estimate is significantly higher than that reported previously (16), the reasons for the observed differences, however, are not understood.
As shown in Fig. 2, the incorporation of the 17 O atom into the DMPO-OH adduct gives rise to the additional lines. The incorporation of an 17 O atom into DMPO-OH increases the number of ESR lines from 4 to 15 because of 17 O coupling. Fig. 8 16 OH is obtained from the computer simulation (Fig. 2).
To verify the electron transfer mechanism, the nitrone-derived cation radical must be directly detected. However, the DMPO cation radical is too unstable to be detected by direct ESR (25). In this regard, it is important to note that only the azulenyl nitrone possesses the necessary redox and structural characteristics to form a persistent cation radical (26,27).
Bicarbonate-mediated Oxidation and Nitration of Tyrosine  Fig. 4, HCO 3 Ϫ is required for the peroxidatic oxidation of tyrosine by SOD1. Previously, we reported that CO 3 . formed at the active site of SOD1 was able to oxidize ␣-tocopherol to ␣-tocopheryl quinone (17). CO 3 . reacts rapidly with tyrosine (k ϭ 10 7 to 10 8 M Ϫ1 s Ϫ1 ) to form the tyrosyl radical that subsequently dimerizes to form dityrosine (Scheme 2). Dityrosine can be oxidized by CO 3 . to form the dityrosyl radical, which will react with the tyrosyl radical to form trityrosine. Similarly, one can visualize the formation of other higher oxidative marker products. Based on the literature report (14), we attribute the minor HPLC peaks eluted at longer retention times (Fig. 4) to the tri-and tetratyrosines. Bicarbonate stimulated nitrotyrosine formation in incubations containing nitrite anion (NO 2 Ϫ ), SOD1, and H 2 O 2 . We attribute this to a fast electron transfer reaction between CO 3 . and NO 2 Ϫ (k ϭ 4 ϫ 10 6 M Ϫ1 s Ϫ1 ) that results in the formation of the nitrogen dioxide radical (NO 2 ⅐ ), a potent nitrating agent (28,29). NO 2 ⅐ can abstract the phenolic hydrogen atom from tyrosine to form the corresponding tyrosyl radical, which then recombines with NO 2 ⅐ to form the nitrotyrosine. Our results indicated that nitrotyrosine decayed and a new peak appeared with time (Fig. 6A). This new peak was assigned to nitrodityrosine, the formation of which can be explained as follows: CO 3 . ϩ . However, the addition of HCO 3 Ϫ to this system induced the oxidation of ABTS to ABTS . ϩ (Fig. 7). Similar results have previously been reported (5). The addition of Me 2 SO (200 mM) did not have any effect on the formation of ABTS .

by SOD1 and H 2 O 2 -As shown in
ϩ (data not shown). Thus, free hydroxyl radicals are not responsible for oxidation of ABTS to ABTS . ϩ in this system. We propose that ABTS ϩ ⅐ is formed from the oxidation of ABTS by CO 3 . . CO 3 . is a selective yet potent oxidant that can diffuse out from the active site of SOD1 and oxidizes ABTS to ABTS . ϩ by an electron transfer mechanism (22). This proposal also explains how a large molecule like ABTS, which is unlikely to reach the active site of SOD1, could still be oxidized by the peroxidase activity of SOD1 in the presence of added HCO 3 Ϫ .

A Common Mechanism for Oxidation and Hydroxylation of Peroxidase Substrates by SOD1/H 2 O 2 /HCO 3
Ϫ -The x-ray crystal structure of SOD1 indicates that access to the active site of copper via a narrow channel is restricted for large molecules (30) (Scheme 3). However, a relatively small anion such as HCO 3 Ϫ could reach the active site of SOD1 where it can be oxidized to CO 3 . by the enzyme-bound hydroxyl radical (22).
CO 3 . is a diffusible oxidant that could leave the active site and cause the oxidation of various compounds outside the active site. The one-electron oxidation potential for the HCO 3 Ϫ /CO 3 .
couple is ϩ1.59 V (31), which makes oxidation of HCO 3 Ϫ by SOD1-Cu 2ϩ -⅐ OH to CO 3 . thermodynamically feasible. HCO 3 Ϫ can thus effectively export oxidation from the sterically hindered active site to large molecules in bulk solution (22). This model was originally proposed by Hodgson and Fridovich (32,33). They proposed that the "copper-bound hydroxyl radical," which is generated in the reaction between SOD1 and H 2 O 2 , oxidizes several small molecular weight anion ligands such as azide, formate, and nitrite anions (32,33). The oxidizing potential of the putative SOD1-Cu 2ϩ -⅐ OH species was suggested to be similar to that of "free" hydroxyl radical, which is considerably higher than the potentials associated with conventional SCHEME 2. Bicarbonate-mediated oxidation/nitration of tyrosine. This scheme describes the major reaction pathways leading to the formation of dityrosine, nitrotyrosine, nitrodityrosine, and other higher oxidation products of tyrosine formed during SOD1/H 2 O 2 /HCO 3 Ϫ -induced oxidation of tyrosine in the presence and absence of nitrite. SCHEME 3. Oxidation of bicarbonate to the carbonate anion radical by the copper-bound hydroxyl radical at the active site of SOD1. The CO 3 . radical, a potent oxidant formed at the active site of SOD1, is able to diffuse out of the active site and oxidizes several structurally different molecules including DMPO, ABTS, tyrosine, and nitrite anion in the bulk solution.
peroxidases (17,19). The earlier studies were performed in bicarbonate buffers, but the possibility of oxidation of HCO 3 Ϫ to CO 3 . radical was not realized at that time.
The present data and the published reports (17,22,23) strongly implicate a role for CO 3 . in the HCO 3 Ϫ -dependent peroxidative mechanism of SOD1. This new perspective clearly rules out a mechanism involving direct oxidation/hydroxylation of DMPO, ABTS, or tyrosine by either free or bound hydroxyl radical (7,8,16,34,35). HCO 3 Ϫ , however, facilitates oxidation/ hydroxylation of substrates through formation of the CO 3 . intermediate. Thus, irrespective of the structure of peroxidase substrates, SOD1/H 2 O 2 /HCO 3 Ϫ induces their oxidation, hydroxylation, or nitration via a common oxidizing intermediate (Scheme 3).