Lack of Tyrosine Nitration by Peroxynitrite Generated at Physiological pH*

Nitration of tyrosine residues of proteins has been suggested as a marker of peroxynitrite-mediated tissue injury in inflammatory conditions. The nitration reaction has been extensively studied in vitro by bolus addition of authentic peroxynitrite, an experimental approach hardly reflecting in vivo situations in which the occurrence of peroxynitrite is thought to result from continuous generation of ⋅NO and O⨪2 at physiological pH. In the present study, we measured the nitration of free tyrosine by ⋅NO and O⨪2 generated at well defined rates from the donor compound (Z)-1-[N-[3-aminopropyl]-N-[4-(3-aminopropylammonio)butyl]-amino]-diazen-1-ium-1,2-diolate] (spermine NONOate) and the xanthine oxidase reaction, respectively. The results were compared with the established nitration reaction triggered by authentic peroxynitrite. Bolus addition of peroxynitrite (1 mm) to tyrosine (1 mm) at pH 7.4 yielded 36.77 ± 1.67 μm 3-nitrotyrosine, corresponding to a recovery of about 4%. However, peroxynitrite formed from ⋅NO and O⨪2, which were generated at equal rates (∼5 μm × min−1) from 1 mmspermine NONOate, 28 milliunits/ml xanthine oxidase, and 1 mm hypoxanthine was much less efficient (0.67 ± 0.01 μm; ∼0.07% of total product flow). At O⨪2fluxes exceeding the ⋅NO release rates, 3-nitrotyrosine formation was below the detection limit of the high performance liquid chromatography method (<0.06 μm). Nitration was most efficient (∼0.3%) with the ⋅NO donor alone, i.e.without concomitant generation of O⨪2. Nitration by ⋅NO had a pH optimum of 8.2, increased progressively with increasing tyrosine concentrations (0.1–2 mm), and was not enhanced by NaHCO3 (up to 20 mm), indicating that it was mediated by ⋅NO2 rather than peroxynitrite. Our results argue against peroxynitrite produced from ⋅NO and O⨪2 as a mediator of tyrosine nitration in vivo.

Nitration of tyrosine residues of proteins has been suggested as a marker of peroxynitrite-mediated tissue injury in inflammatory conditions. The nitration reaction has been extensively studied in vitro by bolus addition of authentic peroxynitrite, an experimental approach hardly reflecting in vivo situations in which the occurrence of peroxynitrite is thought to result from continuous generation of ⅐ NO and O 2 . at physiological pH. In the present study, we measured the nitration of free tyrosine by ⅐ NO and O 2 . generated at well defined Nitric oxide ( ⅐ NO) 1 is a cellular messenger regulating numerous biological processes, including relaxation of blood vessels and neurotransmitter release in the brain, but overproduction of ⅐ NO appears to contribute essentially to tissue injury in inflammatory and ischemic conditions (1). The molecular mech-anisms underlying the cytotoxicity of ⅐ NO are not well understood. The potent oxidant peroxynitrite, which is formed in a rapid reaction from ⅐ NO and O 2 . , is thought to be a key mediator of ⅐ NO toxicity in atherosclerosis, congestive heart failure, glutamate excitotoxicity, and other disease states involving inflammatory oxidative stress (2). Formation of peroxynitrite from ⅐ NO and O 2 . occurs at nearly diffusion-controlled rates (4.3-6.7 ϫ 10 9 M Ϫ1 s Ϫ1 ) (3,4). Therefore, ⅐ NO out-competes the reaction of O 2 . with superoxide dismutase at steady-state concentrations that are likely to occur in vivo (5). Peroxynitrite is stable at alkaline pH but has a half-life of less than 1 s at pH 7.4 (pK a ϭ 6.8) (6). Depending on the pH, the corresponding peroxynitrous acid either rearranges to NO 3 Ϫ or decomposes to NO 2 Ϫ and O 2 (7). Peroxynitrite has been shown to react with virtually all classes of biomolecules (8). The reaction with phenolic compounds, including free and protein-bound tyrosine, results in the formation of nitrated, hydroxylated, and dimeric products (9 -12). The nitration of tyrosine, yielding mainly 3-nitrotyrosine, is markedly enhanced by CO 2 , which reacts with peroxynitrite anion at physiological pH to form the potent nitrating species ONO 2 CO 2
Peroxynitrite-triggered tyrosine nitration has been extensively studied in vitro by bolus addition of alkaline solutions of peroxynitrite to tyrosine-containing samples. However, this experimental approach does not reflect the in vivo situation in which peroxynitrite is thought to be formed by the rapid reaction of ⅐ NO with O 2 . at physiological pH. Intriguingly, we 2 and others (12) found that the nitrating potential of the sydnonimine-based nitrovasodilator SIN-1, which releases ⅐ NO and O 2 . at the same time (25), is much lower than that of authentic peroxynitrite. However, the decomposition of SIN-1 is highly pH-dependent, its decomposition pathways may be more complicated than hitherto assumed, and co-products of SIN-1 metabolism may interfere with tyrosine nitration, making this drug an inappropriate tool to investigate the nitrating potential of ⅐ NO/O 2 . . In the present study, we have addressed this issue with ⅐ NO and O 2 . generated simultaneously by a system that has been used previously to study hydroxylation and Snitrosation reactions by ⅐ NO/O 2 . (26,27).
Solutions-All solutions were prepared fresh each day. Water was ultrafiltered type I (resistance, Ն18 M⍀ ϫ cm Ϫ1 ) from a Barnstead NanoPure apparatus. Spermine NONOate was prepared as a 10-fold stock solution in 10 mM NaOH. SIN-1 was dissolved to 10 mM at pH 5.0. DHR was dissolved to 10 mM in acetonitrile and kept in the dark until use. ⅐ NO solutions were prepared as described (28). Alkaline solutions of peroxynitrite were prepared from acidified NO 2 Ϫ and H 2 O 2 as described (7). Stock solutions were diluted with H 2 O to 10 mM (pH ϳ12.8) immediately before the experiments and added to 50 mM K 2 HPO 4 /KH 2 PO 4 buffer, pH 7.4, to obtain final peroxynitrite concentrations of 1 mM. Changes of buffer pH were Ͻ0.1 unit. oxyHb was prepared as described (29).
Tyrosine Nitration-Unless indicated otherwise, hypoxanthine (1 mM) and tyrosine (1 mM) were incubated at ambient temperature for 12 h in 50 mM K 2 HPO 4 /KH 2 PO 4 buffer, pH 7.4, in the presence of various concentrations of XO and spermine NONOate or SIN-1 (final concentration, 1 mM). Authentic 3-nitrotyrosine was stable under these conditions (recovery, 97.2 Ϯ 3.44% (n ϭ 4) after 12 h of incubation). In some experiments, reaction mixtures were incubated for 1 h with a solution of authentic ⅐ NO (final concentration, ϳ1 mM) instead of the ⅐ NO donor. To study tyrosine nitration by authentic peroxynitrite, alkaline stock solutions of peroxynitrite (final concentration, 1 mM) were added dropwise under vigorous vortexing to tyrosine (1 mM) in 50 mM K 2 HPO 4 /KH 2 PO 4 buffer, pH 7.4, followed by incubation for 1 h. The combined presence of hypoxanthine (1 mM), XO (28 milliunits/ml), and spermine NONOate (1 mM) had no effect on tyrosine nitration triggered by authentic peroxynitrite. The following buffers were used to study the pH dependence of tyrosine nitration: pH 6.0 -8.0, 0.2 M K 2 HPO 4 / KH 2 PO 4 ; pH 8.2-9.0, 0.2 M Tris/HCl; pH Ն 10, solutions of NaOH. The effect of CO 2 was studied by incubation in the presence of up to 20 mM NaHCO 3 .
Determination of 3-Nitrotyrosine-HPLC analysis of 3-nitrotyrosine was performed on a C 18 reversed phase column with 50 mM KH 2 PO 4 / H 3 PO 4 buffer (pH 3) containing 10% (v/v) methanol at 1 ml/min and detection at 274 nm as described (32). Calibration curves were recorded daily with authentic 3-nitrotyrosine (0.06 -5 M for experiments with ⅐ NO/O 2 . and ⅐ NO, and 10 -100 M for experiments with authentic peroxynitrite).
Oxidation of DHR-DHR oxidation was measured at 501 nm with a Shimadzu 160A spectrophotometer at ambient temperature in a total volume of 0.2 ml of a 50 mM K 2 HPO 4 /KH 2 PO 4 buffer, pH 7.4, as described (33,34). The amount of oxidized DHR was calculated using an extinction coefficient of 78.78 mM Ϫ1 cm Ϫ1 . With authentic peroxynitrite, the recovery of DHR oxidation was 37.4 Ϯ 1.43% (n ϭ 4). Spermine NONOate (1 mM) or hypoxanthine/XO (28 milliunits/ml) alone led to DHR oxidation rates of Ͻ0.06 and Ͻ0.3 M ϫ min Ϫ1 , respectively.  (Table I).

Peroxynitrite Formation-We
With the photometric assay, the release rates of ⅐ NO and O 2 .
could not be measured for longer than the first 1-3 min. We had to consider the possibility that the release rates fell off after this time and that consequently the initial rates overestimated the total peroxynitrite production. However, we found a progressive increase in tyrosine nitration over 10 -12 h of incubation (Fig. 1). Kinetic simulations of spermine NONOate decomposition indicated that after 10 h, i.e. the time point at  (Fig. 2). At a fixed rate of 1.9 Ϯ 0.09 M Tyrosine Dependence-The finding that tyrosine nitration due to spermine NONOate alone was more efficient than in the presence of hypoxanthine/XO implied that the nitration was triggered by ⅐ NO or a product of ⅐ NO autoxidation rather than by peroxynitrite. We next characterized the tyrosine dependence of the nitration reaction due to spermine NONOate compared with that caused by authentic peroxynitrite. In the presence of 1 mM spermine NONOate, formation of 3-nitrotyrosine increased progressively with increasing tyrosine concentration up to 2 mM (Fig. 3A), whereas the reaction triggered by authentic peroxynitrite reached a maximum at 0.3 mM tyrosine (39.17 Ϯ 3.48 M 3-nitrotyrosine) (Fig. 3B) and then declined at higher concentrations (20.53 Ϯ 1.61 M 3-nitrotyrosine at 2 mM).
pH Dependence-Another characteristic of tyrosine nitration by authentic peroxynitrite is its pH dependence (10,12,35). We therefore studied the pH dependence of ⅐ NO-mediated tyrosine nitration for comparison. To avoid complications arising from the known pH sensitivity of ⅐ NO release from spermine NONOate (36), we used an aqueous solution of ⅐ NO gas as the ⅐ NO source. The pH dependence of tyrosine nitration by this solution showed a sharp maximum at pH 8.2; in contrast, tyrosine nitration triggered by authentic peroxynitrite was most effi-cient at pH 7.0, in agreement with the previous studies ( Fig. 4) (2,37) but can cause nitration in the presence of additional oxidants generating tyrosyl radicals. The second order reaction of ⅐ NO with tyrosyl radicals (k Ͼ 10 9 M Ϫ1 s Ϫ1 ) (38) leads to formation of C-nitroso and/or O-nitrosotyrosine products that can be converted to 3-nitrotyrosine in a two-electron oxidation reaction (39,40). It is unlikely that this mechanism accounts for tyrosine nitration under our experimental conditions, i.e. in the absence of additional oxidants required for tyrosyl radical formation and oxidative rearrangement of the intermediate(s). The recovery of 3-nitrotyrosine was considerably higher when a high initial concentration of ⅐ NO was applied as a bolus as compared with the continuous release of ⅐ NO from spermine NONOate giving low steady state concentrations of free ⅐ NO. This observation hints at an essential involvement of ⅐ NO autoxidation, a reaction that follows second order kinetics with respect to ⅐ NO and, therefore, occurs at significant rates only at relatively high ⅐ NO concentrations (41). The initial product of ⅐ NO autoxidation, ⅐ NO 2 , was shown to nitrate free tyrosine and tyrosine residues of proteins in aqueous solution (12,42,43). The pH dependence of tyrosine nitration caused by ⅐ NO (Fig. 4) resembles that obtained by others with authentic ⅐ NO 2 (42), suggesting that the nitration of tyrosine observed under our experimental conditions was mediated by ⅐ NO 2 formed in the course of ⅐ NO autoxidation.
The most interesting finding of this study was the apparent lack of significant tyrosine nitration by peroxynitrite generated from ⅐ NO and O 2 . at physiological pH, even though alkaline solutions of peroxynitrite efficiently nitrated tyrosine under identical conditions. These results suggest that the peroxynitrite formed from ⅐ NO and O 2 . at physiological pH differs from the species present in alkaline solutions. The efficiency of tyrosine nitration triggered by alkaline solutions of peroxynitrite is markedly enhanced by CO 2 , which reacts with peroxynitrite anion to give the potent nitrating species ONO 2 CO 2 Ϫ (13, 15, 35, 44 -46). We observed that NaHCO 3 led to an about 2-fold increase in tyrosine nitration by alkaline peroxynitrite, a result that agrees well with previous data from other laboratories (14,35,45,46). However, Berlett et al. (47) have recently suggested that formation of ONO 2 CO 2 Ϫ is obligatory for peroxynitritemediated nitration (47). The apparently CO 2 -independent reaction observed in the absence of added bicarbonate would accordingly be due to contaminating bicarbonate present in appreciable concentrations (0.1-0.2 mM) in buffer solutions (35).
Assuming that formation of ONO 2 CO 2 Ϫ is indeed essential for peroxynitrite-triggered tyrosine nitration as suggested by Ber-  We can only speculate about the nature of the two postulated peroxynitrite species. The most obvious assumption is the involvement of the cis-and trans-rotamers. Interestingly, the cis-and trans-isomers exhibit clearly different pK a values of 6.8 and 8.0 (48,49), respectively, and data obtained by Raman spectroscopy indicate that peroxynitrite is present exclusively in the cis-conformation in alkaline solution (48). Based on this information, we propose the scheme shown in Fig. 6 as a hypothetical explanation of our data. It is postulated that trans-peroxynitrite is the initial product of the reaction between ⅐ NO and O 2 . at pH 7.4.
Because CO 2 reacts with peroxynitrite anion but not with peroxynitrous acid (13), protonation of trans-peroxynitrite at pH 7.4 (pK a ϭ 8.0) is expected to prevent formation of the nitrating ONO 2 CO 2 Ϫ adduct. In contrast, ϳ80% of the cis-isomer (pK a ϭ 6.8) exist as anion at physiological pH allowing for the reaction with CO 2 and consequent tyrosine nitration.
What are the alternative possibilities to explain our observations? Peroxynitrite in statu nascendi might represent a novel as yet unrecognized form of this molecule, but there are no indications that such a species exists. Alternatively, peroxynitrite generated from ⅐ NO and O 2 . at physiological pH could react with CO 2 to yield an unreactive form of ONO 2 CO 2 Ϫ . Formation of unreactive ONO 2 CO 2 Ϫ , kinetically distinguishable from the reactive form, has been reported (50), but we cannot prove or disprove formation of such a species under our experimental conditions. Finally, the obligatory role of ONO 2 CO 2 Ϫ in peroxynitrite-triggered tyrosine nitration is not generally accepted. 3 Thus, it cannot be excluded that peroxynitrous acid acts as nitrating species even in the complete absence of CO 2 . In this case, the explanation of our data would require the additional assumption that the cis-but not the trans-form of peroxynitrous acid nitrates tyrosine.
Our data demonstrate that the simultaneous generation of ⅐ NO and O 2 . does not cause tyrosine nitration under physiological conditions. Together with the unequivocal demonstration of highly elevated nitrotyrosine levels in several human disease states (21,24,51,52), these surprising results raise the question, again, of what nitrates tyrosine in vivo (53). ⅐ NO 2 formed by autoxidation of ⅐ NO may have acted as nitrating species in our experiments with authentic ⅐ NO and spermine NONOate, but it is unclear whether ⅐ NO autoxidation could account for the extensive nitration of tyrosine residues observed in tissues. At micromolar concentrations of ⅐ NO, the third-order autoxidation reaction is certainly not fast enough to generate sufficient ⅐ NO 2 (42). However, recent data indicate that the reaction of ⅐ NO with O 2 is accelerated about 300-fold in biological membranes, rendering hydrophobic compartments of cells important sites for the formation of ⅐ NO-derived reactive species (54). Thus, it is conceivable that high concentrations of ⅐ NO 2 are formed in such compartments and cause tyrosine nitration of adjacent membrane proteins under certain conditions. Several studies hint at peroxynitrite-independent pathways of tyrosine nitration in tissues. Myeloperoxidase, which is secreted by phagocytic neutrophils in inflammatory conditions, could be a key player in these events. It generates tyrosyl radicals (55) that can be trapped by ⅐ NO, leading to formation of 3-nitrotyrosine in the presence of additional oxidants, e.g. H 2 O 2 (40). In the presence of H 2 O 2 , myeloperoxidase may also cause nitration by oxidation of NO 2 Ϫ to ⅐ NO 2 (56). Finally, myeloperoxidase catalyzes formation of hypochlorous acid (HOCl), which reacts nonenzymatically with NO 2 Ϫ to form the potent nitrating agent nitryl chloride (NO 2 Cl) (57,58). Thus, tyrosine nitration may become significant at sites where ⅐ NO synthase is activated together with oxidative pathways generating tyrosyl radicals and H 2 O 2 or other oxidants. The presence of myeloperoxidase appears to be important but would not be obligatory if other enzymatic pathways acting in a similar manner were activated together with ⅐ NO synthase in tissue inflammation.
In conclusion, our results show that peroxynitrite generated from ⅐ NO and O 2 . at physiological pH does not nitrate tyrosine.
This forces us to question the relevance of previous reports on protein nitration by bolus addition of alkaline peroxynitrite. It should also strongly stimulate new efforts to discover the true mechanism responsible for tyrosine nitration in inflammatory and infectious disease.