Eosinophil Peroxidase Nitrates Protein Tyrosyl Residues

Eosinophil peroxidase (EPO) has been implicated in promoting oxidative tissue injury in conditions ranging from asthma and other allergic inflammatory disorders to cancer and parasitic/helminthic infections. Studies thus far on this unique peroxidase have primarily focused on its unusual substrate preference for bromide (Br−) and the pseudohalide thiocyanate (SCN−) forming potent hypohalous acids as cytotoxic oxidants. However, the ability of EPO to generate reactive nitrogen species has not yet been reported. We now demonstrate that EPO readily uses nitrite (NO2 −), a major end-product of nitric oxide (⋅NO) metabolism, as substrate to generate a reactive intermediate that nitrates protein tyrosyl residues in high yield. EPO-catalyzed nitration of tyrosine occurred more readily than bromination at neutral pH, plasma levels of halides, and pathophysiologically relevant concentrations of NO2 −. Furthermore, EPO was significantly more effective than MPO at promoting tyrosine nitration in the presence of plasma levels of halides. Whereas recent studies suggest that MPO can also promote protein nitration through indirect oxidation of NO2 − with HOCl, we found no evidence that EPO can indirectly mediate protein nitration by a similar reaction between HOBr and NO2 −. EPO-dependent nitration of tyrosine was modulated over a physiologically relevant range of SCN− concentrations and was accompanied by formation of tyrosyl radical addition products (e.g. o,o′-dityrosine, pulcherosine, trityrosine). The potential role of specific antioxidants and nucleophilic scavengers on yields of tyrosine nitration and bromination by EPO are examined. Thus, EPO may contribute to nitrotyrosine formation in inflammatory conditions characterized by recruitment and activation of eosinophils.

Eosinophil peroxidase (EPO) has been implicated in promoting oxidative tissue injury in conditions ranging from asthma and other allergic inflammatory disorders to cancer and parasitic/helminthic infections. Studies thus far on this unique peroxidase have primarily focused on its unusual substrate preference for bromide (Br ؊ ) and the pseudohalide thiocyanate (SCN ؊ ) forming potent hypohalous acids as cytotoxic oxidants. However, the ability of EPO to generate reactive nitrogen species has not yet been reported. We now demonstrate that EPO readily uses nitrite (NO 2 ؊ ), a major end-product of nitric oxide ( ⅐ NO) metabolism, as substrate to generate a reactive intermediate that nitrates protein tyrosyl residues in high yield. EPO-catalyzed nitration of tyrosine occurred more readily than bromination at neutral pH, plasma levels of halides, and pathophysiologically relevant concentrations of NO 2 ؊ . Furthermore, EPO was significantly more effective than MPO at promoting tyrosine nitration in the presence of plasma levels of halides. Whereas recent studies suggest that MPO can also promote protein nitration through indirect oxidation of NO 2 ؊ with HOCl, we found no evidence that EPO can indirectly mediate protein nitration by a similar reaction between HOBr and NO 2 ؊ . EPO-dependent nitration of tyrosine was modulated over a physiologically relevant range of SCN ؊ concentrations and was accompanied by formation of tyrosyl radical addition products (e.g. o,o-dityrosine, pulcherosine, trityrosine). The potential role of specific antioxidants and nucleophilic scavengers on yields of tyrosine nitration and bromination by EPO are examined. Thus, EPO may contribute to nitrotyrosine formation in inflammatory conditions characterized by recruitment and activation of eosinophils.
Eosinophils play a central role in host defenses against a variety of cancers and both parasitic and helminthic infections (1)(2)(3). Increased levels of circulating and tissue eosinophils are also implicated in promoting cellular injury during asthma and other allergic inflammatory disorders (1)(2)(3)(4)(5)(6). Eosinophils are thought to mediate many of their cytotoxic and tissue-destroying effects through their exceptional ability to generate oxidizing species (1)(2)(3)(4)(7)(8)(9). Indeed, the respiratory burst of eosino-phils generates several times as much superoxide (O 2 . ) and hydrogen peroxide (H 2 O 2 ) as a corresponding number of neutrophils (9,10). Despite their known proclivity for producing reactive oxygen species and a wealth of clinical evidence linking eosinophils to host defenses and inflammatory tissue injury, structural identification of specific oxidation products formed by these leukocytes is lacking. Eosinophil peroxidase (EPO), 1 an abundant heme protein secreted from activated eosinophils, plays a central role in oxidant production by eosinophils (1-4, 8, 11, 12). EPO amplifies the oxidizing potential of H 2 O 2 produced during the respiratory burst by using it as a co-substrate to generate cytotoxic oxidants. Studies thus far have focused on the unusual substrate preference of EPO for physiological levels of bromide (Br Ϫ ) (8,(13)(14)(15)(16)(17) and the pseudohalide thiocyanate (SCN Ϫ ) (18 -20), even in the presence of a vast molar excess of chloride (Cl Ϫ ), as is seen in vivo. Studies with proteins incubated in the presence of radioactive Br Ϫ or SCN Ϫ salts and either activated eosinophils or the EPO-H 2 O 2 system have demonstrated covalent incorporation of radiolabel into target proteins (13,14). However, structural identification of the oxidation products formed on proteins following exposure to EPO-generated oxidants is essentially unexplored. We recently identified 3-bromotyrosine and 3,5-dibromotyrosine as major products of protein oxidation by the EPO-H 2 O 2 -Br Ϫ system (21). The role of brominating oxidants in promoting tissue injury in eosinophilic inflammatory disorders has not yet been established.
Another potential pathway for oxidative tissue damage by eosinophils may involve formation of nitrating intermediates by EPO. Immunohistochemical studies using antibodies raised against nitrotyrosine, a global marker for protein damage by reactive nitrogen species, intensely stain eosinophils present in the inflamed lung tissues of asthmatics (22). Klebanoff (23) demonstrated that the antimicrobicidal properties of eosinophil peroxidase-H 2 O 2 systems are enhanced in the presence of nitrite (NO 2 Ϫ ), the autoxidation product of ⅐ NO. Moreover, recent studies have shown that myeloperoxidase (MPO), a neutrophiland monocyte-specific peroxidase, can use NO 2 Ϫ and H 2 O 2 as substrates to generate a reactive intermediate capable of nitrating phenolic compounds and protein tyrosyl residues (25-28). However, MPO and EPO are distinct gene products with divergent physical properties. Although both peroxidases contain heme prosthetic groups, structural studies have established that MPO contains a six-coordinate, high spin chlorin, while EPO possesses a high spin, six-coordinate protoporphyrin (24). Moreover, although human MPO is tetrameric and comprised of two heavy and two light chains, EPO is dimeric, having only one heavy and one light chain (7,24). Whether these structural differences underlie the distinct halide and substrate specificities observed between MPO and EPO remains unknown. Furthermore, although increased levels of NO 2 Ϫ and eosinophils have been reported in a variety of conditions, the ability of EPO to generate nitrating intermediates has not yet been explored.
The biological roles and potential targets of mammalian peroxidases are defined by their unique and often nonoverlapping substrate preferences (29,30). We therefore sought to determine whether the EPO-H 2 O 2 system of eosinophils could promote protein oxidative damage through formation of reactive nitrogen species. We now show that isolated EPO readily uses NO 2 Ϫ as substrate to generate a reactive intermediate that nitrates protein tyrosyl residues in high yield. We demonstrate that EPO is more efficient than MPO at nitrating tyrosine and that EPO-dependent protein nitration is a preferred activity of the enzyme. Finally, we provide evidence that EPO also generates tyrosyl radical and that tyrosine nitration by EPO may involve a tyrosyl radical intermediate. Isolation and Characterization of EPO and MPO-Porcine EPO was isolated according to the method of Jorg (31) employing guaiacol oxidation as the assay (32). Purity of EPO preparations was assured before use by demonstrating a RZ of Ͼ0.9 (A 415 /A 280 ), SDS-polyacrylamide gel electrophoresis analysis with Coomassie Blue staining, and in-gel tetramethylbenzidine peroxidase staining to confirm no contaminating MPO activity (33). MPO was initially purified from detergent extracts of human leukocytes by sequential lectin affinity and gel filtration chromatography as described (34). Trace levels of contaminating EPO were then removed by passage over a CM-52 ion exchange column (20). Purity of isolated MPO was established by demonstrating a RZ of 0.87 (A 430 /A 280 ), SDS-polyacrylamide gel electrophoresis analysis with Coomassie Blue staining, and in-gel tetramethylbenzidine peroxidase staining to confirm no contaminating EPO activity (33). Enzyme concentrations were determined spectrophotometrically utilizing extinction coefficients of 89,000 and 112,000 M Ϫ1 cm Ϫ1 /heme of MPO (35) and EPO (36,37), respectively. The concentration of the MPO dimer was calculated as half of the indicated concentration of hemelike chromophore.

Materials-Organic
General Procedures-SDS-polyacrylamide gel electrophoresis analysis was performed as described by Laemmli (38). Protein content was measured by the Markwell-modified Lowry assay with bovine serum albumin (BSA) as the standard (39). HOBr free of Br Ϫ and bromate was prepared from liquid bromine the day of use as described (40). HOBr was quantified spectrophotometrically (⑀ 331 ϭ 315 M Ϫ1 cm Ϫ1 ) (41) as its conjugate base, hypobromite (OBr Ϫ ), immediately prior to use. The concentration of reagent H 2 O 2 (⑀ 240 ϭ 39.4 M Ϫ1 cm Ϫ1 ) (42) was determined spectrophotometrically. Production of H 2 O 2 by the glucose/glucose oxidase system was determined by oxidation of Fe(II) and formation of a Fe(III)-thiocyanate complex (25). Western blot analysis of nitrated proteins was determined as described using immunopurified rabbit polyclonal antibody against nitrotyrosine (Upstate Biotechnology, Inc., Lake Placid, NY) (43). The specificity of the primary antibody was confirmed by blocking immunoreactivity in coincubations with either 10 mM nitrotyrosine or 1 mM of the tripeptide Gly-nitro-Tyr-Ala as described (43). o,oЈ-Dityrosine, pulcherosine, trityrosine, and isodityrosine were prepared and quantified as described previously (44). 3-Bromotyrosine and 3,5-dibromotyrosine standards were synthesized from L-tyrosine and isolated by preparative HPLC as recently described (21). [ 13 C 6 ]Ring-labeled analogs of 3-bromotyrosine and 3,5-dibromotyrosine were similarly prepared using L-[ 13 C 6 ]tyrosine as starting material. UV-visible studies on authentic and EPO-generated 3-nitrotyrosine were performed on a Bio Spectrophotometer (Perkin-Elmer).

5,5Ј-Dithiobis(2-nitrobenzoic acid) (DTNB) Formation by EPO-5-
Thio-2-nitrobenzoic acid (TNB) was prepared fresh by reduction of 1 mM DTNB in 100 ml of sodium phosphate buffer (50 mM, pH 7.0) with 4 l of 2-mercaptoethanol (46). EPO (14.2 nM) was then incubated at 37°C in sodium phosphate buffer (50 mM, pH 7.0) with TNB (53 M) in the absence or presence of the indicated concentrations of Br Ϫ , SCN Ϫ , NO 2 Ϫ , or Cl Ϫ . Oxidation of TNB to DTNB was initiated by the addition of H 2 O 2 (30 M) and followed spectrophotometrically at 412 nm (⑀ 412 ϭ 27,200 M Ϫ1 cm Ϫ1 ) (46). Oxidation of Free L-Tyrosine and Protein-bound Tyrosyl Residues-Unless otherwise specified, reactions were initiated by the addition of oxidant (H 2 O 2, HOBr or glucose to the glucose/glucose oxidase system) and performed in sodium phosphate buffer (20 mM, pH 7.0) at 37°C for 60 min under the conditions described in the figure and table legends. Reactions were stopped by the addition of phenol (1% final) and catalase (30 nM; for glucose/glucose oxidase studies) to the reaction mixture. The pH dependence of 3-nitrotyrosine formation was assayed in phosphate buffer (50 mM final) composed of mixtures of phosphoric acid and monobasic and dibasic sodium phosphate. The pH of each reaction mixture was determined at the end of the incubation and did not change by more than 0.2 pH units over the course of the reaction. Experiments utilizing the glucose/glucose oxidase system for H 2 O 2 generation were performed in the presence of 100 g/ml glucose and 20 ng/ml glucose oxidase (grade II; Roche Molecular Biochemicals) at 37°C for the indicated times. Preliminary studies demonstrated that under these conditions, a constant flux of H 2 O 2 (0.18 M/min) was generated.
Protein Hydrolysis-Preliminary studies confirmed that removal of NO 2 Ϫ prior to acidification of proteins was essential to avoid artifactual nitration of proteins. Proteins oxidized in vitro were therefore prepared for analysis by first precipitating and desalting them in a single-phase extraction mixture composed of H 2 O/methanol/H 2 O-saturated diethyl ether (1:3:7, v/v/v) as described (47). Proteins were typically hydrolyzed by incubating the desalted protein pellet with 6 N HCl (0.5 ml) supplemented with 1% phenol for 24 h (47). Prior to initiating hydrolysis, acid mixtures were degassed under vacuum and then sealed under a blanket of argon. When chlorotyrosine and/or bromotyrosine determinations were performed, formation of trace levels of halogenated tyrosine analogs during acid hydrolysis was avoided by hydrolysis in 4 N methane sulfonic acid (0.5 ml) supplemented with 1% phenol for 24 h at 100°C under a blanket of argon (21). Control experiments utilizing proteins supplemented with L-[ 2 H 4 ]tyrosine as internal standard added prior to acid hydrolysis and then analyzed by GC/MS (48)  Reverse Phase HPLC Quantification of Tyrosine Oxidation Products-Quantification of oxidation products from free L-tyrosine (3-nitrotyrosine, 3-bromotyrosine, 3,5-dibromotyrosine, 3-chlorotyrosine, and o,oЈ-dityrosine) was performed on a Beckman Gold HPLC system equipped with diode array and fluorescence detectors. Separations were performed on a C18 column (Beckman Ultrasphere, 5 m, 4.6 ϫ 250 mm) equilibrated with solvent A (0.1% trifluoroacetic acid, pH 2.5). L-Tyrosine and its oxidation products were eluted at a flow rate of 1 ml/min with a linear gradient generated with solvent B (0.1% trifluoroacetic acid in methanol, pH 2.5) as follows: 0% solvent B for 5 min, 0 -100% solvent B over 30 min, 100% solvent B for 10 min. Products were quantified employing standard curves constructed with authentic synthetic standards. N ␣ -acetyl-L-tyrosine-derived oxidation products were isolated, and their structures were confirmed by on-line electrospray ionization mass spectrometric analysis as described below. Routine quantification of N ␣ -acetyl-L-tyrosine oxidation products was determined by reverse phase HPLC with UV detection (A 280 ) as above, utilizing external calibration curves constructed with known amounts of their nonacetylated counterparts (and assuming identical extinction coefficients). The retention times of N ␣ -acetyl-L-tyrosine analogs of tyrosine oxidation products were determined using authentic standards isolated at the time of on-line liquid chromatography/mass spectrometric analysis (LC/MS).
3-Nitrotyrosine, 3-bromotyrosine, 3,5-dibromotyrosine, o,oЈ-dityrosine, and 3-chlorotyrosine in protein hydrolysates were routinely quantified by reverse phase HPLC with electrochemical (coulometric) detection on an ESA (Cambridge, MA) CoulArray HPLC instrument equipped with UV detector and electrochemical cells (six channels) arranged in series and set to increasing specified potentials: channel 1 (300 mV); channel 2 (350 mV); channel 3 (480 mV); channel 4 (650 mV); channel 5 (800 mV); channel 6 (880 mV). Chromatographic separation of amino acids was performed on a a Projel TSK ODS-80 TM column (5 m, 4.6 ϫ 250 mm) equilibrated with mobile phase A (15 mM lithium phosphate, 3 mg/liter lithium dodecyl sulfate, pH 3.2). Products were eluted at a flow rate of 1 ml/min with a nonlinear gradient generated with mobile phase B (50% methanol, 15 mM lithium phosphate, 3 mg/liter lithium dodecyl sulfate, pH 3.2) as follows: isocratic elution at 0% mobile phase B for 10 min, 0 -15% mobile phase B over 10 min, isocratic elution at 15% mobile phase B for 10 min, 15-20% mobile phase B over 10 min, isocratic elution at 20% mobile phase B for 10 min, 20 -100% mobile phase B over 20 min, isocratic elution at 100% mobile phase B for 10 min. Peak identity was established by demonstrating the appropriate retention time, redox potential, and ratio of integrated currents in adjacent channels and by the method of standard additions for each analyte. L-Tyrosine, 3-nitrotyrosine, 3-bromotyrosine, 3,5-dibromotyrosine, o,oЈ-dityrosine, and 3-chlorotyrosine standards (0.5-50 pmol each on column) were also dissolved together and used to generate external calibration curves. The experimental error in data obtained by this assay is less than Ϯ5%. Control experiments comparing quantitation of the tyrosine analogues by stable isotope dilution GC/MS (45) versus reverse phase with electrochemical detection demonstrated comparable results.
Mass Spectrometry-LC/MS was performed using electrospray ionization (ESI) and detection with a Quatro II triple quadrupole mass spectrometer (Micromass, Inc.) interfaced with an HP 1100 high performance liquid chromatograph (Hewlett Packard, Wilmington, DE) equipped with diode array detector. L-Tyrosine oxidation products were resolved on an Ultrasphere C18 column (Beckman; 5 m, 4.6 ϫ 250 mm) at a flow rate of 1 ml/min and a linear gradient between H 2 O (with 0.3% formic acid) and methanol (with 0.3% formic acid) over 30 min. Column eluent was split (970 l/min to diode array, 30 l/min to mass detector) and analyzed in the positive ion mode with a cone potential of 30 eV. GC/MS analyses of HPLC-isolated L-tyrosine oxidation products were performed following derivatization to their n-propyl perheptafluorylbutyryl or n-propyl perpentafluorylproprionyl derivatives (49). 3-Ni-trotyrosine was also analyzed by GC/MS analysis as its n-propyl perheptafluorylbutyryl derivative following reduction to 3-aminotyrosine (50). Negative ion chemical ionization GC/MS studies were performed utilizing a Perkin Elmer (Norwalk, CT) TurboMass spectrometer equipped with chemical ionization probe.

3-Nitrotyrosine and Tyrosyl Radical Addition Products Are Formed during L-Tyrosine Oxidation by the Eosinophil Peroxidase-H 2 O 2 -NO 2
Ϫ System-The addition of purified EPO and H 2 O 2 to reaction buffer (20 mM sodium phosphate, 100 M diethylenetriaminepentaacetic acid, pH 7.0) supplemented with L-tyrosine generated several new species with distinct retention times on reverse phase HPLC analysis with diode array detection (Fig. 1, upper panel, peaks a-c). The ultraviolet absorption maximum of each peak was nearly indistinguishable from L-tyrosine ( max 275-285 nm). LC/MS analyses were consistent with formation of the tyrosyl radical addition products o,oЈ-dityrosine and isodityrosine (peak a; m/z 361 (M ϩ H) ϩ ), pulcherosine (peak b; m/z 540 (M ϩ H) ϩ ), and trityrosine (peak c; m/z 540 (M ϩ H) ϩ ) (44). Likewise, HPLC analysis with UV and fluorescence detection demonstrated co-migration with authentic synthetic standards prepared as described under ؊ system. Top, the major L-tyrosine oxidation product generated by the EPO-H 2 O 2 -NO 2 Ϫ system (peak I, Fig. 1, bottom) was isolated by reverse phase HPLC as described under "Experimental Procedures." The isolated product was dried, resuspended at the indicated pH, and then analyzed by UV-visible spectroscopy (200 -600 nm). Bottom, the major L-tyrosine oxidation product generated by the EPO-H 2 O 2 -NO 2 Ϫ system (peak I in Fig. 1, bottom) was analyzed by LC/MS as described under "Experimental Procedures." The positive ion ESI mass spectrum is shown. The structure and predicted m/z of protonated 3-nitrotyrosine is also shown (inset). "Experimental Procedures." Following the addition of NO 2 Ϫ , the reaction mixture developed an intense yellow color. Reverse phase HPLC analysis demonstrated the formation of an additional major (peak I) and minor (peak d) product (Fig. 1, bottom panel) as well as enhanced production of the tyrosyl radical addition products (peaks a-c). The UV-visible spectra of the new products (peaks I and d) both illustrated similar absorbance maxima in the visible wavelength range ( max 350 -355 nm), consistent with nitration of the aromatic amino acid (51). Use of heat-inactivated EPO or omission of either L-tyrosine or H 2 O 2 from the reaction mixture both resulted in no detectable production of the products. The reaction products formed (peak I and peaks a-d) were stable to treatment with acid (4 N HCl at 100°C for 24 h), incubation at 37°C with H 2 O 2 , and the addition of a molar excess of either reducing agents or nucleophilic scavengers (e.g. NaCNBH 3 , methionine, 2-mercaptoethanol, taurine, ammonium acetate).
To characterize the structures of the L-tyrosine oxidation products generated by the EPO-H 2 O 2 -NO 2 Ϫ system, a variety of analytical techniques were employed. The pH dependence of the UV-visible absorbance spectrum of HPLC-isolated peak I is illustrated in Fig. 2 (upper panel) and is identical to that observed with authentic 3-nitrotyrosine. To confirm the structure of the major L-tyrosine oxidation product (peak I) as 3-nitrotyrosine, LC/MS analysis was performed. The retention time and positive ion mass spectrum of peak I was identical to that of authentic 3-nitrotyrosine and demonstrated a m/z of 227 (M ϩ H) ϩ (Fig. 2, lower panel). LC/MS analysis of the minor products generated during oxidation of L-tyrosine by the complete EPO-H 2 O 2 -NO 2 Ϫ system was consistent with the prior structural assignments of peaks a-c and suggested that peak d was composed of a nitrated analog of o,oЈ-dityrosine (e.g. 3-nitro-o,oЈ-dityrosine, m/z 406 (M ϩ H) ϩ ). These results, combined with the chemical stability and fluorescence spectra (data not shown) of the compounds formed, suggest that the major compounds generated during oxidation of free L-tyrosine by EPO and H 2 O 2 are tyrosyl radical addition products. In contrast, in the presence of EPO, H 2 O 2 , and NO 2 Ϫ , 3-nitrotyrosine is the major product formed.
Characterization of Reaction Requirements for EPO-dependent Formation of 3-Nitrotyrosine-The reaction requirements for EPO-catalyzed nitration of L-tyrosine yielding 3-nitrotyrosine are illustrated in Fig. 3. At pH 7.0 and 100 M L-tyrosine and NO 2 Ϫ , 3-nitrotyrosine synthesis by EPO (following the bolus addition of 100 M H 2 O 2 ) was linear over the first 15 min and reached a plateau within 30 min (Fig. 3A). At higher L-tyrosine concentrations, the yield of 3-nitrotyrosine decreased, consistent with the competing bimolecular mechanism of o,oЈ-dityrosine formation. EPO-catalyzed nitration was optimal between pH 6 and 6.5, where approximately 0.35 mol of product was formed for each mole of oxidant consumed (Fig.  3B). A dose-dependent increase in the synthesis of 3-nitrotyrosine was observed over a (patho)physiologically relevant range of NO 2 Ϫ concentrations (Fig. 3C). The H 2 O 2 concentration dependence for 3-nitrotyrosine production by EPO yielded a similarly shaped plot and overall yield (Fig. 3D). At high concentrations of H 2 O 2 , however, the total amount of 3-nitrotyrosine formed decreased, consistent with either substrate inhibition or an interaction between H 2 O 2 and an intermediate in 3-nitrotyrosine formation (data not shown).

3-Nitrotyrosine Is a Preferred Product of Eosinophil Peroxidase at Physiologically Relevant Concentrations of L-Tyrosine, Halides, and NO 2
Ϫ -Thiocyanate (SCN Ϫ ) and bromide (Br Ϫ ) are reported to be preferred substrates for EPO (13)(14)(15)(16)18). To estimate the relative substrate preferences of isolated EPO for NO 2 Ϫ versus other reported substrates for the enzyme, we ex- amined the initial rate of DTNB formation in reaction mixtures containing EPO, TNB, H 2 O 2 , and various concentrations of anionic substrates (SCN Ϫ , Br Ϫ , NO 2 Ϫ , Cl Ϫ ) at neutral pH. The relative rates of TNB oxidation observed by EPO were as follows: SCN Ϫ Ͼ Br Ϫ Ͼ NO 2 Ϫ Ͼ Ͼ Cl Ϫ (Fig. 4). Although EPO does not conform to Michaelis-Menten kinetics (e.g. high concentrations of H 2 O 2 inhibit tyrosine nitration), a measure of the relative ability of EPO to discriminate in favor of a particular substrate in the presence of a mixture of competing substrates can be calculated as an apparent specificity constant (k x Ϫ) for each substrate (33). Because high levels of H 2 O 2 can inhibit oxidant production, we estimated the kinetic parameters under conditions that minimize inhibition and mimic physiological conditions (e.g. neutral pH and low concentrations of H 2 O 2 ) ( Table I). SCN Ϫ was the preferred substrate for EPO, demonstrating an approximately 2.7-fold higher apparent specificity constant than Br Ϫ , the next best substrate examined. EPO preferred Br Ϫ approximately 4-fold more than NO 2 Ϫ (Table I). Implicit in this type of kinetic analysis is the assumption that the reactive nitrogen species formed by the EPO-H 2 O 2 -NO 2 Ϫ system oxidize TNB at a rate and stoichiometry comparable with that of other hypohalous acids formed by the enzyme.
As an alternative means of assessing the potential significance of EPO-dependent tyrosine nitration, we performed product analyses of the major L-tyrosine oxidation products formed by the enzyme system (Table I). When plasma levels of Ltyrosine and halides were incubated with isolated EPO, H 2 O 2 , and concentrations of NO 2 Ϫ as might be observed in inflammatory tissues or fluids (tyrosine ϭ Br Ϫ ϭ NO 2 Ϫ ϭ 100 M, Cl Ϫ ϭ 100 mM; Refs. 25 and 52-56), 3-nitrotyrosine was formed in high yield (Table II). In fact, 3-nitrotyrosine was the major product formed and was present at 2-3-fold greater levels than either 3-bromotyrosine or o,oЈ-dityrosine (Table II). Despite that vast molar excess of Cl Ϫ , no detectable formation of 3-chlorotyrosine was observed under all of the conditions examined, consistent with the low apparent specificity constant observed (Tables I and II). The highest yields of 3-nitrotyrosine production were observed during incubations in the absence of Br Ϫ , consistent with the unusual substrate preference of EPO for this halide.

3-Nitrotyrosine Formation by Eosinophil Peroxidase Occurs in the Presence of Nucleophilic Scavengers and Is Inhibited by
Ascorbate and Urate-Reactive oxidants and diffusible radical species generated by peroxidases in vivo will encounter a variety of nucleophilic scavengers and antioxidant compounds. To characterize the nature of the reactive halogenating and nitrating intermediates generated by EPO, we incubated the enzyme with H 2 O 2 , Br Ϫ , and NO 2 Ϫ under a variety of conditions and then quantified 3-bromotyrosine and 3-nitrotyrosine production. Again, when L-tyrosine was incubated with EPO, H 2 O 2 (100 M), and equivalent amounts of Br Ϫ and NO 2 Ϫ (100 M each), 3-nitrotyrosine was the major oxidation product formed and was produced in 2-3-fold higher yield than 3-bromotyrosine (Fig. 5). The addition of methionine, a potent scavenger     of reactive halogenating species, completely ablated 3-bromotyrosine production. In contrast, the addition of the thiol ethercontaining amino acid to the reaction mixture had no significant effect on 3-nitrotyrosine production (Fig. 5). Likewise, the addition of reduced thiol compounds such as glutathione totally inhibited tyrosine bromination but only partially blocked aromatic nitration by the peroxidase. The addition of either ascorbic acid or uric acid potently inhibited both 3-bromotyrosine and 3-nitrotyrosine production. The addition of primary aminecontaining species (e.g. N ␣ -acetyl lysine) to the reaction mixture did not attenuate either 3-bromotyrosine or 3-nitrotyrosine production. These results are consistent with our recent report that N-monobromamines serve as brominating intermediates for tyrosine (21). They also suggest that the nitrating intermediate formed by the EPO-H 2 O 2 -NO 2 Ϫ system is not scavenged by amine groups. Finally, these results suggest that aromatic nitration reactions may be mediated by EPO in inflammatory fluids and tissues where NO 2 Ϫ levels are increased and antioxidants such as ascorbate and urate may be depleted.
Eosinophil Peroxidase Nitrates Protein Tyrosyl Residues in High Yield-The ability of EPO to nitrate protein tyrosyl residues was first examined by incubating a target protein (e.g. BSA) with the complete EPO-H 2 O 2 -NO 2 Ϫ system and then performing SDS-polyacrylamide gel electrophoresis analysis with Western blot analysis using affinity-purified polyclonal antibodies specific for 3-nitrotyrosine (43). A NO 2 Ϫ -dependent increase in the intensity of staining was observed following exposure of the protein to the enzymatic system (Fig. 6). To confirm that 3-nitrotyrosine was produced in proteins exposed to the EPO-H 2 O 2 -NO 2 Ϫ system, protein hydrolysates were directly analyzed for 3-nitrotyrosine production by GC/MS in the negative-ion chemical ionization mode employing synthetic 3-nitro-[ 13 C 6 ]tyrosine as an internal standard, as described under "Experimental Procedures." Ions with the appropriate mass-to-charge ratio (m/z 464 (M) . and 444 (M Ϫ HF) . ) and identical retention time to that observed with their corresponding stable isotope-labeled fragment ions were readily observed for the n-propyl-perheptafluorylbutyryl derivative of 3-nitrotyrosine. Moreover, reduction of the products prior to analysis by GC/MS demonstrated ions with identical retention time and appropriate m/z for that of the n-propyl perheptafluorylbutyryl derivative of 3-aminotyrosine (m/z 806 (M Ϫ HF) . , m/z 762 (M Ϫ HF Ϫ CO 2 ) . and m/z 628 (M Ϫ heptafluorylbutyryl) . ). Collectively, these results unambiguously confirm that EPO catalyzes nitration of protein tyrosyl residues in the presence of the co-substrates, H 2 O 2 and NO 2 Ϫ . To characterize the significance of EPO-dependent protein nitration, the content of protein-bound 3-nitrotyrosine produced by the enzyme was determined by reverse phase HPLC with electrochemical detection as described under "Experimental Procedures." BSA was incubated with the complete EPO-H 2 O 2 -NO 2 Ϫ system, while the concentrations of either NO 2 Ϫ (Fig. 7, upper panel) or H 2 O 2 (Fig. 7, lower panel) were varied. Nitration of protein tyrosyl residues occurred readily in a concentration-dependent manner for both co-substrates. The overall yield of EPO-catalyzed nitration of protein tyrosyl residues as a function of each substrate is shown in Fig. 8. The highest yields (10 -20%) of protein nitration were noted when NO 2 Ϫ concentrations were high and H 2 O 2 concentrations were low. These results suggest that protein nitration would be highest under conditions where H 2 O 2 production occurs at a low rate, similar to the conditions that are likely to be present in vivo.
To more fully examine the potential physiological significance of protein nitration by EPO, we quantified the levels of formation of multiple tyrosyl residue oxidation products (3nitrotyrosine, 3-bromotyrosine, 3-chlorotyrosine, and o,oЈ-dityrosine) in proteins incubated with isolated EPO, a hydrogen in sodium phosphate buffer (20 mM, pH 7.0) at 37°C for 60 min. After incubation, 1.2 g of protein was loaded on 10% SDS-polyacrylamide gels for electrophoresis, and proteins were transferred and immunostained using a rabbit polyclonal antibody against 3-nitrotyrosine (top) or stained with Coomassie Blue (bottom). peroxide-generating system (glucose/glucose oxidase), and concentrations of halides and NO 2 Ϫ as might be observed in inflammatory fluids (i.e. 100 mM Cl Ϫ , 100 M Br Ϫ , and 100 M NO 2 Ϫ ). Again, EPO nitrated protein tyrosyl residues to a greater extent than that of all other tyrosine oxidation products examined (Table III). Under the conditions employed, only minimal amounts of o,oЈ-dityrosine and no detectable levels of 3-chlorotyrosine were formed (Table III). EPO-catalyzed bromination of protein tyrosyl residues occurred, albeit at a level approximately one-third that of tyrosine nitration. Generation of the tyrosine oxidation products demonstrated an absolute requirement for EPO and the H 2 O 2 -generating system. Collectively, these results demonstrate that EPO uses NO 2 Ϫ as a preferred substrate to form a reactive nitrogen intermediate(s) even in the presence of physiological levels of halides.
The pseudohalide SCN Ϫ is a preferred substrate of EPO (18 -20). We were therefore interested in examining the poten-tial influence of SCN Ϫ levels on EPO-catalyzed nitration. Extracellular fluid and tissue levels of this unusual anion have not been reported; however, plasma levels of SCN Ϫ vary considerably (normal values are Ͻ69 M) depending upon diet and smoking habits in normal individuals (57). In a separate set of experiments, either free L-tyrosine (Fig. 9, left) or BSA (Fig. 9, right) was incubated with NO 2 Ϫ (100 M) and plasma levels of halides (100 mM Cl Ϫ and 100 M Br Ϫ ) in the absence and presence of varying amounts of SCN Ϫ , and the content of 3-nitrotyrosine was determined. EPO-mediated nitration was attenuated by the addition of SCN Ϫ over the reported range of plasma levels of SCN Ϫ (Fig. 9). In the presence of equivalent levels of NO 2 Ϫ and SCN Ϫ (100 M each), 3-nitrotyrosine formation was significantly inhibited; however, nitration of both free and protein-bound tyrosine residues was still detectable.
Eosinophil Peroxidase Is More Effective Than Myeloperoxidase at Generating Nitrotyrosine in the Presence of NO 2 Ϫ , H 2 O 2 , and Plasma Levels of Halides-Since MPO-dependent formation of nitrotyrosine may also contribute to protein nitration at sites of inflammation (25-28, 58, 59), we sought to compare the ability of isolated MPO and EPO to nitrate target proteins. BSA was incubated with equal concentrations of each peroxidase in the presence of plasma levels of halides (100 mM Cl Ϫ and 100 M Br Ϫ ) and an H 2 O 2 -generating system (glucose/ glucose oxidase) to more closely mimic the low steady flux of peroxide that might be formed in vivo. Kinetic analyses comparing the ability of each peroxidase to use NO 2 Ϫ as substrate for protein nitration demonstrated that EPO formed at least 4-fold more 3-nitrotyrosine on target proteins than MPO at all times examined (Fig. 10, left). Moreover, regardless of the concentration of NO 2 Ϫ examined (0 -100 M), EPO was significantly more effective than MPO at nitrating protein tyrosyl residues (Fig. 10, right). Thus, at sites of eosinophilic inflammatory disorders, EPO may contribute to 3-nitrotyrosine formation.

The Mechanism of Eosinophil Peroxidase-dependent Production of 3-Nitrotyrosine Involves Formation of a Tyrosyl Radical
Intermediate-To examine the mechanism of protein tyrosyl residue nitration by the EPO-H 2 O 2 -NO 2 Ϫ system, we initially performed a variety of experiments utilizing N ␣ -acetyl-L-tyrosine, a low molecular weight surrogate for a protein-bound tyrosyl residue. In the absence of NO 2 Ϫ , incubation of N ␣ -acetyl-L-tyrosine with EPO and H 2 O 2 resulted in formation of the N ␣ -acetyl dityrosine analog (Table IV). The addition of Br Ϫ to the reaction mixtures resulted in production of the N ␣ -acetyl-L-tyrosine analogs of 3-bromotyrosine and 3,5-dibromotyrosine but no detectable production of the N ␣ -acetyl-L-tyrosine analog of the tyrosyl radical-addition product o,oЈ-dityrosine. In contrast, the addition of NO 2 Ϫ to N ␣ -acetyl-L-tyrosine and the EPO-H 2 O 2 system formed predominantly the N ␣ -acetyl-L-tyrosine analog of 3-nitrotyrosine and modest levels of the N ␣ -acetyl-Ltyrosine analog of o,oЈ-dityrosine (Table IV). The addition of both NO 2 Ϫ and Br Ϫ to N ␣ -acetyl-L-tyrosine and the EPO-H 2 O 2 system formed all of the expected tyrosine oxidation products (N ␣ -acetyl-L-tyrosine analogs of 3-nitrotyrosine, o,oЈ-dityrosine, 3-bromotyrosine, and 3,5-dibromotyrosine). Thus, in every instance where tyrosine nitration occurred, concurrent formation of tyrosyl radical addition products was observed. In contrast, during tyrosine bromination (in the absence of NO 2 Ϫ ), no formation of tyrosyl radical addition products (i.e. N ␣ -acetyl-Ltyrosine analog of o,oЈ-dityrosine) could be detected. These results are consistent with 3-nitrotyrosine production through an addition reaction between an intermediate tyrosyl radical and reactive nitrogen species such as ⅐ NO 2 , the one electron oxidation product of NO 2 Ϫ .

Co-incubation of HOCl and NO 2
Ϫ has been reported to form a nitrating and chlorinating intermediate, presumably NO 2 Cl (58). Moreover, we observed modest increases in the levels of free and protein-bound nitrotyrosine formed during incubation with the EPO-H 2 O 2 system in the presence of NO 2 Ϫ and Cl Ϫ compared with NO 2 Ϫ alone (Tables II and III). However, the mechanism for the increase does not appear to involve reaction of HOCl with NO 2 Ϫ , since comparable increases in tyrosine nitration were observed in the presence of Cl Ϫ during incubations that contained excess methionine, a potent scavenger of HOCl and other halogenating intermediates (data not shown).
To determine if EPO might indirectly promote protein nitration by forming a nitrating intermediate by oxidation of NO 2 Ϫ with HOBr, we performed the following experiment. N ␣ -Acetyl-Ltyrosine was incubated with HOBr in the presence and absence of NO 2 Ϫ , and the extent of tyrosine nitration was determined. No detectable formation of the 3-nitrotyrosine analog was observed, even in the presence of 1 mM concentrations of both HOBr and NO 2 Ϫ (Table IV). Similar results were observed in reactions performed under acidic (pH 4) conditions (data not shown). Collectively, these results suggest that the nitrating intermediate formed by EPO arises from direct oxidation of NO 2 Ϫ and not by secondary oxidation of NO 2 Ϫ by HOCl, HOBr, or some other halogenating agent.
In a final series of experiments, we sought to further explore Ϫ BSA (0.4 mg/ml) was incubated with eosinophil peroxidase (57 nM), glucose (100 g/ml), glucose oxidase (20 ng/ml), and the indicated anionic substrates (Br Ϫ ϭ NO 2 Ϫ ϭ 100 M, Cl Ϫ ϭ 100 mM) in sodium phosphate buffer (20 mM, pH 7.0) at 37°C for 4 h. Following reaction, protein was desalted and subjected to acid hydrolysis, and the content of 3-nitrotyrosine, 3-bromotyrosine, o,oЈ-dityrosine, and 3-chlorotyrosine was determined by reverse phase HPLC with electrochemical detection as described under "Experimental Procedures." Results are the mean Ϯ S.D. of three independent experiments. The limit of detection for these analyses is Ͻ0.05 mmol/mol tyrosine for each analyte examined.  the potential role of a tyrosyl radical intermediate in 3-nitrotyrosine formation by monitoring the extent of aromatic nitration using tyrosine analogs that do not form tyrosyl radicals. Because of its acidic character, the phenoxyl hydrogen of Ltyrosine is the preferred site of hydrogen atom abstraction from tyrosine; consequently, O-methyl-L-tyrosine is resistant to formation of tyrosyl radical (60). If 3-nitrotyrosine production by the EPO-H 2 O 2 -NO 2 Ϫ system instead occurred by an electrophilic addition reaction (e.g. through a NO 2 ϩ intermediate), use of the O-methyl-L-tyrosine analog should not significantly effect the yield of 3-nitrotyrosine formation. Incubation of the complete EPO-H 2 O 2 -NO 2 Ϫ system with L-tyrosine readily formed 3-nitrotyrosine; in contrast, no nitrated products were formed from O-methyl-L-tyrosine at all concentrations of NO 2 Ϫ examined, as determined by reverse phase HPLC with on-line UV and ESI/MS (Fig. 11). Collectively, these results strongly support the hypothesis that the EPO-H 2 O 2 -NO 2 Ϫ system nitrates tyrosyl residues through a tyrosyl radical intermediate. DISCUSSION Eosinophils play an essential role in tissue surveillance and host defense mechanisms (1-3). These cells have evolved enzymatic mechanisms to inflict oxidative damage upon invading parasites, pathogens, and cancer cells. The reactive species they form, however, also have the potential to harm host tissues and contribute to inflammatory tissue injury. Microbicidal and cytotoxic oxidants generated by the EPO-H 2 O 2 system are thought to participate in promoting many of these functions (2)(3)(4)(5)(6)(7)(8). However, despite the numerous links between EPO, oxidant production, and tissue injury in eosinophilic inflammatory disorders, structural identification of oxidation products formed by the action of EPO on target proteins is lacking.    The present studies identify aromatic nitration reactions as a new category of oxidation reactions mediated by EPO. EPO readily used NO 2 Ϫ , a major end product of ⅐ NO metabolism, as substrate to generate a reactive intermediate that nitrates free and protein-bound tyrosyl residues in high yield. Multiple independent analytical methods confirmed the structure of the major tyrosine oxidation product formed by the EPO-H 2 O 2 -NO 2 Ϫ system as 3-nitrotyrosine: (i) studies employing LC/MS with electrospray ionization detection demonstrated that the retention time and mass spectrum of the oxidation product were identical to that of authentic 3-nitrotyrosine; (ii) multiple distinct derivatives of the major tyrosine oxidation product formed by the EPO-H 2 O 2 -NO 2 Ϫ system demonstrated identical GC retention times and negative ion electron impact mass spectra to that of similarly derivatized authentic 3-nitrotyrosine and co-migrated with synthetically prepared 3-nitro-[ 13 C 6 ]tyrosine (both in native form and following reduction of both to 3-aminotyrosine); (iii) the UV-visible absorbance spectra of the tyrosine oxidation product was identical to that of authentic 3-nitrotyrosine at acidic, neutral, and basic pH; (iv) the retention time and electrochemical properties (i.e. hydrodynamic voltamogram) of the EPO-generated product were identical to those of authentic 3-nitrotyrosine when analyzed by reverse phase HPLC with electrochemical (coulometric array) detection; and finally (v) antibodies raised against 3-nitrotyrosine specifically recognize protein oxidized by the EPO-H 2 O 2 -NO 2 Ϫ system. Collectively, these results unambiguously identify 3-nitrotyrosine as a product formed by action of the EPO-H 2 O 2 -NO 2 Ϫ system on free and protein-bound tyrosine residues.
The present results also provide the first evidence for tyrosyl radical production by EPO. EPO-dependent formation of this diffusible radical species is strongly supported by the structural identification of multiple EPO-generated phenolic addition products of tyrosyl radical (e.g. o,oЈ-dityrosine, isodityrosine, pulcherosine, and trityrosine). Moreover, the mechanism of aromatic nitration reactions mediated by the EPO-H 2 O 2 -NO 2 Ϫ system is consistent with formation of a tyrosyl radical intermediate. Thus, tyrosyl radical-dependent oxidative damage of biological targets represents another potential pathway that EPO may contribute to inflammatory tissue injury.
One of the remarkable features of protein nitration by the EPO-H 2 O 2 -NO 2 Ϫ system is the overall high yield of 3-nitrotyrosine formation. In the presence of plasma levels of halides and levels of NO 2 Ϫ that approximate those found in inflammatory tissues and fluids, 3-nitrotyrosine was a preferred product formed with both free and protein-bound tyrosine residues as targets (e.g. Tables II and III). Moreover, tyrosine nitration was optimal in a physiological pH range (6.0 -7.0) and accounted for between 10 and 20% of the H 2 O 2 consumed during oxidation of proteins by the EPO-H 2 O 2 -NO 2 Ϫ system at neutral pH. Finally, common nucleophilic scavengers such as primary amines, thiols, and thiol ethers only modestly attenuated nitrotyrosine formation; however, physiological levels of glutathione, ascorbate, and SCN Ϫ are anticipated to have a major impact on the extent of nitrotyrosine formation by EPO. Taken together, these results suggest that EPO-dependent nitration of protein tyrosyl residues is likely to occur in vivo, particularly at sites where eosinophilic infiltration and ⅐ NO (and hence NO 2 Ϫ ) production are enhanced and ascorbate, glutathione, and SCN Ϫ levels are limiting. Indeed, recent immunohistochemical studies using antibodies raised against nitrotyrosine demonstrated intense staining over eosinophils and throughout eosinophilrich tissues of individuals with severe asthma (22), a condition where eosinophils are implicated in promoting inflammatory tissue injury and enhanced oxidant stress and ⅐ NO production occur (2)(3)(4)(5)(6)(61)(62)(63)(64). Although the authors interpreted the 3-nitrotyrosine immunostaining as evidence of peroxynitrous acid production (22), the current results suggest that an alternative source may be the EPO-H 2 O 2 -NO 2 Ϫ system of eosinophils. It is interesting to note that the extent of 3-nitrotyrosine formation by EPO is modulated by physiologically relevant levels of SCN Ϫ . Thiocyanate ion is produced following the enzymatic hydrolysis of certain plant glycosides. Thus, eating certain vegetables (e.g. cabbage) results in a transient increase in the concentration of thiocyanate in blood and urine (57). Normal plasma levels of SCN Ϫ are under 70 M (57), but higher levels may be observed following excessive tobacco smoking or prolonged intravenous administration of sodium nitroprusside, a potent vasodilator used in critically ill patients (57,65). Eosinophils perform their biological functions in a large variety of tissues and fluids whose concentration of halides, NO 2 Ϫ , and SCN Ϫ may differ considerably from plasma. Indeed, our kinetic studies (Table I) suggest that at plasma concentrations of substrates (SCN Ϫ , Br Ϫ , NO 2 Ϫ , and Cl Ϫ ), EPO is far from saturated. Thus, the relative contribution of EPO to 3-nitrotyrosine formation in vivo may be tissue-specific, even at sites of inflammation characterized by eosinophil recruitment and activation.
The formation of reactive nitrogen species by direct peroxidasedependent oxidation of NO 2 Ϫ , or through oxidation of NO 2 Ϫ by oxidants such as HOCl, has recently been reported (25-28, 58, 59). Indeed, the ability of MPO secreted from activated neutrophils to catalyze aromatic nitration reactions has suggested that nitrotyrosine formation at sites of inflammation may arise from pathways independent of peroxynitrous acid generation (26). However, whether EPO could catalyze similar reactions given the unusual substrate preferences, distinct prosthetic heme group, and unique physical properties of this highly cationic protein was unclear. The results of the present studies demonstrate that aromatic nitration reactions are indeed a preferred activity of the enzyme. Moreover, the nitrating capacity of EPO was shown to exceed that of the related leukocyte-derived peroxidase, MPO, by at least 4-fold at physiological concentrations of halides and every concentration of NO 2 Ϫ examined. Finally, it should be noted that O 2 . and H 2 O 2 production by activated eosinophils is severalfold greater than that of an equal number of comparably stimulated neutrophils (9,10) and that the total content of EPO in eosinophils is reported to be 2-4-fold higher than that of MPO in neutrophils (e.g. 30 -33 pmol of MPO per 1 ϫ 10 6 neutrophils versus 70 -135 pmol of EPO per 1 ϫ 10 6 eosinophils) (16,66,67). Thus, EPOdependent formation of reactive nitrogen intermediates is likely to occur in vivo, particularly during inflammatory disorders characterized by the presence of eosinophils.
Another interesting distinction between EPO-and MPO-dependent nitration reactions arises from their differing halide specificity. Although MPO generates chlorinating oxidants at physiological concentrations of halides (46, 68 -70), and HOCl is reported to react with NO 2 Ϫ to form a nitrating intermediate (58), we found no evidence for a parallel reaction between NO 2 Ϫ and HOBr, the primary halogenating intermediate formed by EPO at plasma levels of halides. Moreover, although EPO-dependent nitration of tyrosine was optimal in the presence of both NO 2 Ϫ and plasma levels of Cl Ϫ (e.g. see Tables II and III), the addition of methionine, a potent scavenger of halogenating intermediates (46,71,72), failed to attenuate nitrotyrosine formation by EPO. The mechanism of the modest but reproducible increase in both free and protein-bound 3-nitrotyrosine formation observed in the presence versus absence of Cl Ϫ is unclear. However, the inability of methionine to attenuate tyrosine nitration by EPO in the presence of either Br Ϫ or Cl Ϫ suggests that indirect oxidation of NO 2 Ϫ by an EPO-generated hypohalous acid does not contribute significantly to tyrosine nitration.
One interesting question is the chemical nature of the reactive nitrogen species formed by EPO-catalyzed oxidation of NO 2 Ϫ . Peroxidases catalyze both one-and two-electron oxidation reactions, and both the one-electron (i.e. nitrogen dioxide) and two-electron (e.g. peroxynitrous acid) oxidation products of NO 2 Ϫ have been suggested as potential intermediates formed by MPO (25, 28). Both intermediates can form nitrotyrosine, as well as abstract a phenoxyl hydrogen atom to form tyrosyl radical intermediates. Future studies aimed at elaborating the oxidant(s) formed and the mechanisms of free and proteinbound tyrosine nitration mediated by peroxidase-dependent oxidation of NO 2 Ϫ are warranted. Eosinophils are typically rare leukocytes with a limited circulating life span (t1 ⁄2 ϳ 2 h) before they take residence in tissues (1). However, they do play a prominent role in certain forms of inflammation (e.g. allergic), where their numbers in tissues can greatly exceed those of other leukocytes (1)(2)(3)(4)(5)(6). Indeed, some of the most devastating inflammatory disorders in humans are associated with intense eosinophilic infiltration (1-6, 73, 74). Identifying the mechanisms and products of EPO-dependent oxidative damage is a critical step toward development of targeted interventions designed to interrupt oxidative tissue injury in eosinophilic inflammatory disorders.