Relative Chlorinating, Nitrating, and Oxidizing Capabilities of Neutrophils Determined with Phagocytosable Probes*

The capabilities of stimulated neutrophils to initiate intraphagosomal and extracellular chlorination, nitration, and other oxidative reactions has been evaluated using a fluorescent particle and soluble phenolic compounds as target molecules. Neutrophils activated by the soluble stimulus, phorbol myristate acetate, both chlorinated fluorescein that was covalently attached to polyacrylamide microspheres and initiated tyrosine dimerization. When nitrite ion was present at millimolar concentration levels in the medium, nitration of the phenolic rings also occurred; the relative extent of nitration increased as the nitrite concentration was increased. Myeloperoxidase (MPO) also catalyzed nitration and chlorination of fluorescein and the fluorescein-conjugated particles in cell-free solutions; the relative nitration yields increased with increasing [NO2 −]/[Cl−] ratios. Nitration did not involve intermediary formation of nitrating agents derived from reaction between MPO-generated HOCl and NO2 − because this reaction also occurred in chloride-free media and direct addition of HOCl to solutions containing NO2 − and fluorescein gave only chlorinated products. In marked contrast to these extracellular reactions, intraphagosomal nitration of the fluorescein-conjugated particles could not be detected (even at [NO2 −] as high as 0.1m), whereas chlorination of the probe was extensive. These data indicate that intraphagosomal aromatic nitration in neutrophils is negligible, although extracellular nitration of phenolic compounds by secreted MPO could occur at physiological concentration levels of NO2 −.

trating phenolic rings (13); and HOCl can apparently react with the immediate product of the respiratory "burst," superoxide anion (O 2 . ) (5,6), to generate hydroxyl radical ( ⅐ OH), viz., (14). Neutrophils activated with soluble stimuli to secrete MPO catalyzed extracellular aromatic nitration when the medium contained NO 2 Ϫ at concentration levels approximating those found in biological fluids (10), establishing that physiologically derived NO 2 Ϫ could be an alternative MPO substrate. Extracellularly generated ⅐ OH is probably too short lived to function as an effective bactericidal agent in biological environments (15). Nonetheless, conversion of a substantial fraction to the longer lived (and bactericidal) carbonate ( ⅐ CO 3 Ϫ ) radical should occur in the CO 2 -rich environment of respiring phagosomes via the reaction: ⅐ OH ϩ HCO 3 Ϫ 3 H 2 O ϩ ⅐ CO 3 Ϫ (16). Thus, each of the reactions involving NO 2 Ϫ and O 2 Ϫ are potentially important contributors to host defense mechanisms against pathogenic organisms.
These reactions might also be major contributors to the pathogenesis of diseases associated with oxidative stress. Neutrophil activation at sites of infection is generally accompanied by some degranulation at the plasma membrane, releasing MPO into the extracellular environment, which could then participate in oxidative reactions leading to damage of surrounding tissues. For example, extracellular MPO has been shown to accumulate within human atherosclerotic lesions (17) which also contain elevated levels of chlorotyrosines (18). Since MPO and stimulated neutrophils are capable of chlorinating peptide tyrosyl groups (19,20), it follows that the reactions leading to protein tyrosyl chlorination in the lesions probably involve MPO-generated HOCl. Similarly, high levels of nitrotyrosine are often found in damaged tissues of individuals with respiratory or neurological diseases (21). Although these findings have often been attributed to involvement of peroxynitrous acid (ONO 2 H), which could be formed by radical coupling of O 2 . with nitric oxide ( ⅐ NO) (22), alternative possibilities in environments containing accumulated neutrophils are MPOcatalyzed nitration (10) or reaction of MPO-generated HOCl with NO 2 Ϫ to form nitrating intermediates (13). In this study, we have utilized a particulate chemical probe that acts as a bacterial mimic (23) to determine relative contributions of the various possible MPO-mediated reactions by stimulated neutrophils. The probe contains a fluorescein derivative as a reporter group, which gives easily distinguishable products when reacted with chlorinating (23), nitrating (23), or other putative neutrophil-generated oxidizing agents (16). When opsonized with serum-derived proteins, the probe stimulates neutrophil respiration and is simultaneously phagocytosed; unopsonized particles, however, elicit no cellular response (23). Opsonization therefore provides a means to control the reaction locus, either within the phagosome or in extracellular environments. Results described herein will show that the reactivity patterns differ in the two environments, with chlorination predominating within the phagosome, but being accompanied by appreciable nitration when the reactions are extracellular.

EXPERIMENTAL PROCEDURES
Materials-Human neutrophils were isolated, counted, and stored as described previously (23). Myeloperoxidase, purified from bovine spleens (24), had an A 430 /A 280 ϭ 0.85 and a specific activity of 284 guaicol units/mg of protein (25); enzyme concentrations were determined spectrophotometrically using ⑀ 430 ϭ 91 mM Ϫ1 cm Ϫ1 /heme (26). Commercial HOCl was purified by vacuum distillation; HOCl concentrations were determined spectrophotometrically using ⑀ 236 ϭ 100 M Ϫ1 cm Ϫ1 (27). Stock peroxynitrite solutions were prepared by reacting potassium nitrite with hydrogen peroxide (28). 3,3Ј-Dityrosine was prepared by horseradish peroxidase-catalyzed oxidation of tyrosine by hydrogen peroxide (H 2 O 2 ) (29). Fluorescein-conjugated 1-m diameter polyacrylamide microspheres containing cystamine linker groups were synthesized as described previously (23). The particles used in this study contained 3.5 ϫ 10 7 fluorescein groups/bead and ϳ8 ϫ 10 8 unreacted carboxyl end-groups/bead. Other chemicals and biochemicals were the highest grade commercially available and were used as supplied. All solutions were prepared from deionized water that had been further purified by passage through a laboratory reverse osmosis/ion exchange unit.
Enzymatic Reactions-Peroxidase-catalyzed nitration reactions were generally carried out at 37°C in 25 mM sodium phosphate, pH 7.4, containing 105 M fluorescein, 120 M tyrosine, or 5 mg/ml bovine serum albumin and varying amounts of H 2 O 2 , NO 2 Ϫ , and the peroxidase (MPO, horseradish peroxidase, or lactoperoxidase). For some reactions involving MPO, 0.10 M NaCl was also added to the reaction medium. Reactions were terminated after 30 min either by adding 20 g/ml catalase or by cooling on ice; these methods gave equivalent results. For studies with fluorescein-conjugated beads, 1.5 ϫ 10 9 beads/ml ([fluorescein] ϭ 80 M) were incubated for 30 min at 37°C in the same buffer with 70 nM MPO, 100 M H 2 O 2 , 0.10 M Cl Ϫ , and varying amounts of nitrite ion.
Reactions with Neutrophils-For most experiments with neutrophils, 2 ϫ 10 7 cells/ml were incubated in human serum with 3 ϫ 10 8 fluorescein-conjugated beads/ml. During the incubation period, the reaction mixtures were gently rotated (23). Reactions were terminated after 30 min at 37°C by cell homogenization. Comparative studies were made utilizing preopsonized beads to stimulate neutrophils suspended in phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 (PBS)); rates and extents of phagocytosis under these conditions are nearly identical to phagocytosis in serum (23). Extracellular reactions involving neutrophil-secreted MPO were investigated by using unopsonized beads in PBS containing 30 ng/ml phorbol myristate acetate (PMA) as a soluble stimulant. Following the reaction, the fluorescein was solubilized for chemical analysis by adding ϳ0.25 M dithiothreitol to cleave the cystamine disulfide bond and centrifuging to remove cell debris (25). Experiments in which extracellular reactions involving tyrosine and intracellular reactions of fluorescein were simultaneously measured were conducted in PBS containing 1 mM tyrosine; immediately following incubation of the fluoresceinated particles with neutrophils, the mixture was centrifuged at 90,000 ϫ g, and the tyrosine-containing supernatant was analyzed by HPLC. The pellet was subsequently homogenized, and the fluorescein was recovered for analysis by adding dithiothreitol.
Product Analyses-Products formed in the various reactions were identified by HPLC on a Gilson 305/306 system equipped with a 5-m Microsorb C-18 column and UV-visible detector. For reactions involving tyrosine, components were separated by isocratic elution with 20 mM phosphate, pH 3.0, containing 8% methanol, and were detected at 276 nm; for reactions involving fluorescein or the fluorescein-conjugated particles, the eluant was 20 mM phosphate, pH 7.4, containing 32% methanol, and detection was at 498 nm. In all cases, peak identities were confirmed by co-chromatography with authentic samples and by optical absorption spectroscopy. Product yields were determined from the peak areas by comparison to reference standards, which were prepared by reacting fluorescein or the fluorescein-conjugated beads with HOCl, MPO-H 2 O 2 -Cl Ϫ , or ONO 2 H and isolating the products by HPLC (23). Identities of the purified standards were confirmed by pneumatically assisted electrospray (Ionspray) mass spectrometry (23). Concentrations of chlorinated standards prepared from free fluorescein were determined spectrophotometrically at pH 9.0 using ⑀ 500 ϭ 7.5 ϫ 10 4 M Ϫ1 cm Ϫ1 for monochlorofluorescein and ⑀ 508 ϭ 8.0 ϫ 10 4 M Ϫ1 cm Ϫ1 for dichlorofluorescein (23). Two optically distinguishable mononitrofluorescein isomers were obtained, which gave band maxima at 490 nm for the more rapidly eluting isomer and 500 nm for the other; for both compounds, ⑀ max ϭ 7.6 ϫ 10 4 M Ϫ1 cm Ϫ1 . Concentrations of reference standards for the chlorinated and nitrated fluorescein derivatives isolated from the beads were determined assuming ⑀ values identical to the other prepared chlorofluorescein and nitrofluorescein standards.

MPO-catalyzed Reactions Involving Fluorescein-Incubation of MPO with fluorescein in solutions containing H 2 O 2 , Cl Ϫ , and NO 2
Ϫ resulted in formation of three major products ( Fig. 1), one of which had previously been identified by mass spectrometry as a monochlorofluorescein (23). The other two peaks co-eluted with nitrofluoresceins formed by reacting fluorescein with ONO 2 H. The latter were identified as isomeric mononitrofluoresceins by Ionspray mass spectrometry, for which each product gave a major ion peak at m/z ϭ 378.0 atomic mass units Ϫ , but not Cl Ϫ , and was nearly completely blocked by addition of the MPO inhibitor, azide ion (Table I and Fig. 1). The nitrofluorescein yields increased with increasing NO 2 Ϫ concentrations up to about 1 mM (Fig. 2a), then decreased slowly at higher concentrations; at the maximal yield, ϳ30% of the initially added H 2 O 2 was converted to nitrofluoresceins. The isomer distribution was unaffected by the NO 2 Ϫ concentration. However, the chlorofluorescein yield rapidly declined, reaching a level of ϳ4% of the added H 2 O 2 at the nitrofluorescein maximum and becoming negligible at higher NO 2 Ϫ concen- trations (Fig. 2a). In contrast, at very low NO 2 Ϫ (Յ60 M) the chlorofluorescein yield was enhanced (Fig. 2a, inset) consistent with earlier observations (10). Direct reaction between HOCl and fluorescein was also inhibited by NO 2 Ϫ , although in this case nitrofluorescein yields were less than 1% of the added oxidant under all reaction conditions investigated (Fig. 2b).
Incubation of fluorescein-conjugated beads with the MPO-H 2 O 2 -Cl Ϫ -NO 2 Ϫ system gave qualitatively the same behavior as incubation of free fluorescein, with chlorination yields decreasing and nitration yields increasing as the NO 2 Ϫ concentration was increased (Fig. 3). Quantitative differences were: (i) the maximal nitration yield was considerably less, with only ϳ9% of the added H 2 O 2 being converted to nitrofluoresceins under optimal conditions, and (ii) the maximal yield was attained at much higher [NO 2 Ϫ ], i.e. Ն 10 mM. In contrast, the optimal chlorofluorescein yield was the same as for free fluorescein within experimental uncertainty, i.e. 52-67% (cf. Figs. 2 and 3 with [NO 2 Ϫ ] ϭ 0). Enzymatic nitration of the conjugated fluorescein was partially inhibited by preopsonization of the beads in 25% serum and completely inhibited when unopsonized beads were reacted in an assay medium comprising 100% serum. Specifically, the nitrofluorescein yield obtained following incubation of 1.5 ϫ 10 9 opsonized beads/ml containing 100 M fluorescein with 70 nM MPO, 100 M H 2 O 2 , and 3-10 mM NO 2 Ϫ in PBS gave only ϳ50% of the nitration product obtained using unopsonized beads. Reaction of unopsonized beads in serum with [NO 2 Ϫ ] as high as 100 mM gave no nitrofluorescein product.

Reactions of Neutrophils with Fluorescein-conjugated
Beads-Human neutrophils were incubated with the fluores-ceinated beads in 100% human serum at a bead:neutrophil ratio of ϳ15:1. Under these conditions, 80 -90% of the beads and nearly all of the MPO are entrapped within sealed phagosomes (23). Addition of NO 2 Ϫ to the medium had no effect upon chlorination yields until its concentration exceeded ϳ1 mM, at which point the chlorination yield declined progressively with increasing NO 2 Ϫ , becoming negligible at [NO 2 Ϫ ] Ӎ 100 mM (Fig.  4). Very similar results were obtained using serum-opsonized beads in PBS, i.e. addition of up to 1 mM NO 2 Ϫ had no effect upon the chlorination yields, which in this case were ϳ8 nmol/ 10 7 cells. Thus, in marked contrast to the MPO-catalyzed reactions of the fluorescein-conjugated beads in solution (Fig. 3), nitrofluorescein was not formed under any experimental conditions in the intraphagosomal reactions of stimulated neutrophils (Fig. 4). However, the neutrophils were capable of catalyzing aromatic nitration reactions, at least in the extracellular medium. Stimulation of neutrophils with PMA in media containing tyrosine and NO 2 Ϫ caused formation of both nitrotyrosine and dityrosine; the relative yield of nitrotyrosine increased with increasing NO 2 Ϫ concentration and only dityrosine was formed when NO 2 Ϫ was deleted (Table II). Similarly, unopsonized fluorescein-conjugated beads reacted extracellularly (23) with PMA-stimulated neutrophils to give both chloro-and nitrofluorescein products (Table II). In this case, the relative yields of the chlorinated product decreased and the nitrated products increased with increasing [NO 2 Ϫ ] in a manner similar to the cell-free reaction of the beads with MPO-H 2 O 2 -Cl Ϫ -NO 2 Ϫ (Fig. 3), although the overall nitration yields were less. For these reactions, the rate of O 2 consumption by the PMA-stimulated neutrophils (ϳ20 nmol/10 7 neutrophils/min) was approximately the same as previously measured for cells stimulated with opsonized beads in PBS (ϳ16 nmol/10 7 neutrophils/ min) and unopsonized beads in 100% serum (ϳ30 nmol/10 7 neutrophils/min) at a 15:1 bead:neutrophil ratio (23). When the buffer contained both opsonized beads (bead:neutrophil ϭ 44:1) and tyrosine, both chlorofluorescein and dityrosine products were obtained. Although only ϳ40% of the beads were phagocytosed, previous studies have established that nearly all of the chlorination is intracellular under these conditions (23). The reduced yield of dityrosine (2.7 nmol/10 7 neutrophils) compared with analogous reactions using PMA-stimulated cells in the absence of the beads (18 nmol/10 7 neutrophils) (Table II) is consistent with other data indicating only limited secretion of MPO under phagocytosing conditions (23). Addition of NO 2 Ϫ to the medium gave rise to nitrotyrosine formation, enhancement of dityrosine yields and, at [NO 2 Ϫ ] Ն 1 mM, partial inhibition of   (Table II). These patterns of reactivity are qualitatively the same as were observed when the intracellular and extracellular reactions were studied separately. Enzyme-catalyzed Tyrosine Nitration-The extent to which phenolic nitration might be a general attribute of peroxidase-catalyzed reactions was briefly explored. Tyrosine was chosen as the substrate because nitrotyrosine formation is thought to be a specific marker for participation of reactive nitrogen species in in vivo oxidative damage (21). Incubation of 5 mg/ml bovine serum albumin with 20 nM MPO, 200 M H 2 O 2 , 0.10 M Cl Ϫ , and 5 mM NO 2 Ϫ gave spectrophotometric evidence of extensive nitration of its tyrosyl residues, i.e. upon reaction, a strong peak appeared at ϳ360 nm, which further intensified and shifted to ϳ430 nm upon increasing the medium alkalinity to pH 11 (data not shown). These spectral changes are very similar to changes measured for nitrotyrosine in PBS, where max ϭ 356 nm at pH 5 and max ϭ 432 nm at pH 11. Incubation of 120 M tyrosine with 30 nM MPO, 100 M H 2 O 2 , and 5 mM NO 2 Ϫ gave ϳ32% conversion to nitrotyrosine. Nitrotyrosine yields increased with increasing [NO 2 Ϫ ] over the experimentally investigated range of 0 -10 mM. Dityrosine also formed when the concentration of tyrosine was increased, and was accompanied by a corresponding reduction in the nitrotyrosine yield. Comparable amounts of nitrotyrosine were formed when the MPO was replaced with 500 nM horseradish peroxidase, but replacement with 260 nM lactoperoxidase gave a product yield that was about 10-fold lower than the other peroxidases. No products were formed when 300 nM catalase was the catalyst.
Fluorescein Recovery from Neutrophil Phagosomes-The amount of dye that was recovered from 17:1 bead:neutrophil suspensions in 100% serum is plotted in Fig. 5 as a function of incubation time. In these reactions, mono-, di-, and trichlorofluoresceins are formed (23); the chlorination yield, plotted on the same time scale, is the total yield of chlorinated products determined by HPLC. Previous studies have established that phagocytosis of the beads, respiratory uptake of O 2 , and bead  Ϫ ; set B: 6 ϫ 10 6 neutrophils/ml stimulated with 30 ng/ml PMA in PBS containing 3 ϫ 10 9 unopsonized beads/ml (ϳ150 M fluorescein) and varying amounts of NO 2 Ϫ ; set C: 8 ϫ 10 6 neutrophils/ml stimulated with 4 ϫ 10 8 opsonized beads (ϳ21 M fluorescein) in PBS containing 1.0 mM tyrosine and varying amounts of NO 2 Ϫ . Fluorescein nitration and chlorination yields were calculated as in Table I chlorination are approximately parallel events (23); these relationships hold for both opsonized beads in PBS and unopsonized beads in 100% serum as stimuli. 2 Furthermore, respiratory activity was not influenced by NO 2 Ϫ ion, i.e. at 15/1 unopsonized beads/neutrophil in 100% serum, the initial rates of O 2 uptake were 31 and 33 nmols/10 7 neutrophils/min in the presence of 1 mM and 10 mM NO 2 Ϫ , respectively, compared with ϳ30 nmols/10 7 neutrophils/min when NO 2 Ϫ was absent (23). Chlorination therefore tracks both phagocytosis and respiratory activity, which are essentially complete after 15-20 min. The data at zero time indicate that ϳ95% of the fluorescein bound to the beads can be recovered prior to phagocytosis; the data obtained at 10 and 30 min in the presence of 1 mM N 3 Ϫ (triangles, Fig. 5), for which phagocytosis and respiration are normal (23), indicate that the same amount of dye can be recovered from the neutrophil phagosomes when MPO activity is inhibited. However, in neutrophils containing active MPO, phagocytosis was accompanied by partial loss of recoverable fluorescein, e.g. at completion of the respiratory burst, only ϳ70% of the initially added dye was recovered (Fig. 5).

DISCUSSION
General Considerations-Neutrophils are thought to possess multiple microbicidal mechanisms, both oxidative (1-4) and nonoxidative (30,31). Recent studies by Winterbourn and associates using selective inhibitors of the neutrophil NADPH oxidase and MPO in conjunction with superoxide dismutaseconjugated Staphylococcus aureus provide evidence that the major pathway for killing of this organism is oxidative and involves MPO-catalyzed reactions (32), and Klebanoff (1) has summarized indirect arguments supporting the primacy of MPOdependent mechanisms in other phagocytic reactions. As discussed in the Introduction, the actual toxins might include HOCl formed by MPO-catalyzed oxidation of Cl Ϫ , secondary oxidants formed by reaction of HOCl with, e.g. endogenous amines (33), NO 2 Ϫ (13), or O 2 . (11), or other oxidants formed by direct reaction of other substrates with MPO (3,10). Use of fluorescein as a chemical probe offers an unusual opportunity to evaluate the relative chlorinating, nitrating, and other oxidative capabilities of stimulated neutrophils because these reactions give rise to distinct reaction products (3). When attached to 1-m polyacrylamide spheres via cystamine spacer groups, the fluorescein can be used to monitor the dynamics of intraphagosomal reactions and can be recovered in near-quantitative yield for subsequent chemical analyses (23).
Intraphagosomal Chlorination Versus Nitration-The most remarkable finding of these studies is that intraphagosomal nitration of fluorescein in neutrophils simply does not occur, even when the reaction medium contains 100 mM NO 2 Ϫ (Fig. 4). The reason for preferential chlorination of the probe within the phagosome is not apparent. It is not attributable to the intrinsic reactivity of fluorescein, which undergoes highly selective MPO-catalyzed ring nitration. For example, the amount of nitrofluorescein formed in a solution containing 1 mM NO 2 Ϫ and 100 mM Cl Ϫ was 5-6-fold greater than the amount of chlorofluorescein formed (Figs. 1 and 2a). The selectivity for fluorescein nitration over chlorination decreased slightly when the dye was attached to polyacrylamide beads, but nitration was the predominant product when [NO 2 Ϫ ] Ն 1 mM (Fig. 3). Likewise, NO 2 Ϫ is not excluded from the phagosome since intraphagosomal MPO-catalyzed chlorination was inhibited at high NO 2 Ϫ concentrations (Fig. 4). The apparent inhibition constant (K i ) at pH 7.4 was ϳ5 mM, based upon the amount of NO 2 Ϫ required to inhibit chlorination by 50%. This value is very similar to the ligand dissociation constant (K d ) for NO 2 Ϫ binding to the peroxidase heme extrapolated from EPR data, i.e. K d Ϸ2 mM (pH 7.4) (34) and from spectrophotometric determinations at other pH values, i.e. K d ϭ 0.6 mM (pH 6.0); 55 mM (pH 8.5) (35). Axial heme ligation blocks reaction with H 2 O 2 , preventing formation of the reactive ferryl -cation, compound I (7); consequently, one expects that K d տ K i . The correspondence between K d and the measured intraphagosomal K i strongly suggests that the effective intraphagosomal NO 2 Ϫ concentrations are approximately the same as in bulk solution. Inhibition of intraphagosomal MPO reactions by N 3 Ϫ , an anion with solution properties similar to NO 2 Ϫ , is well established (36). Finally, the possibility that the cellular environment presents additional target molecules that react preferentially with MPO-generated ⅐ NO 2 or other reactive nitrogen species is not in quantitative accord with the experimental results. Our data indicate that the medium does contain alternative reactants, i.e. fluorescein nitration was ϳ50% inhibited when the beads were opsonized and completely inhibited when the reactions were carried out in serum; fluorescein chlorination by HOCl is also completely inhibited in serum. 2 If intraphagosomal fluorescein nitration were blocked by intracellular scavengers, one would expect a sharp decline in chlorination yields similar to that observed in solution at [NO 2 Ϫ ] Ն 0.5 mM (Figs. 2 and 3); this occurs because enzymatic reaction of NO 2 Ϫ reduces the HOCl yield. However, no significant decrease in chlorination yield was observed in reactions with neutrophils until the NO 2 Ϫ concentration reached levels that inhibited the enzyme (Fig. 4). Consequently, the data suggest that NO 2 Ϫ is not competitive with Cl Ϫ as an MPO substrate within the neutrophil phagosome.
Experiments utilizing both opsonized fluorescein-conjugated beads and tyrosine (Table II)  and Table II), although extracellular nitration of tyrosine became appreciable at NO 2 Ϫ concentrations as low as 500 M (Table II). The results are consistent with a prior report that PMA-stimulated neutrophils are capable of nitrating 4-hydroxyphenylacetic acid (10); our data further establish that sufficient MPO is secreted from actively phagocytosing neutrophils to catalyze extracellular nitration of tyrosine. Thus, these MPO-catalyzed reactions may contribute to oxidative injury in these and related diseases where peroxidase-containing leukocytes are involved, as previously discussed (10).
HOCl-mediated Nitration-Chlorination of soluble fluorescein by HOCl was inhibited by NO 2 Ϫ without formation of appreciable nitrofluorescein (Fig. 2b). This inhibition was only partial, even at NO 2 Ϫ concentrations as high as 10 mM, and appeared to plateau at ϳ20% yield. Very similar behavior has been reported for competitive nitration and chlorination of 4-hydroxyphenylacetic acid by HOCl in NO 2 Ϫ -containing media, although in these reactions ring nitration was more extensive (13 Ϫ ] տ 2 mM for substantial NO 2 Cl formation; the data ( Fig. 2b) are consistent with these constraints. Assuming the validity of this mechanism, the data indicate that NO 2 Cl and any reactive species formed from it are less efficient chlorinating agents than HOCl. More generally, one notes that the rate constant for the HOCl-NO 2 Ϫ reaction under physiological conditions is relatively low. Since biological fluids contain numerous reductants that are more highly reactive toward HOCl (4,27,33), it is unlikely that this reaction contributes appreciably to intracellular oxidations. The data also provide further evidence that MPO-catalyzed nitration reactions do not proceed via intermediary formation of HOCl (cf. Fig. 2, a and b), but probably via radical coupling reactions involving ⅐ NO 2 (8,28).
Chlorination Versus Other Oxidations-Approximately 95% of the fluorescein originally bound to the beads could be recovered by cleavage of the linker disulfide group; studies using neutrophils whose MPO was inhibited with N 3 Ϫ indicated that entrapment within neutrophil phagosomes did not alter the recovery yields (Fig. 5). These data also established that fluorescein does not undergo MPO-independent reactions with products of the respiratory burst. Exposure of fluorescein to ⅐ OH or ⅐ CO 3 Ϫ radicals causes chromophoric bleaching without accumulation of isolable intermediates (16). If these radicals were formed by reaction of MPO-generated HOCl with O 2 Ϫ (14), we might expect the yields of recoverable fluorescein to decrease. The data indicate a loss of about 30% of the fluorescein occurs in MPO-dependent reactions in normal neutrophils (Fig.  5), which is potentially attributable to these reactions. However, chlorination of the dye within the phagosome is extensive, and reaction of HOCl with the trichlorinated derivative also destroys the chromophore (23). Since the relative contribution of this reaction to bleaching cannot be ascertained from these experi-ments, the 30% loss of chromophore must be considered an upper limit to the extent of participation of ⅐ OH and ⅐ CO 3 Ϫ radicals. Conclusions-Fluorescein exhibits a selectivity for nitration over chlorination in MPO-catalyzed reactions that is very similar to tyrosine and other phenolic compounds. Nonetheless, MPO-catalyzed chlorination is strongly favored over aromatic nitration within the neutrophil phagosome. The selectivity is different in extracellular media, however, where MPO-catalyzed phenolic nitration could possibly compete with chlorination. The data also suggest that ⅐ OH and ⅐ CO 3 Ϫ radicals might be important secondary oxidants that are generated within phagosomes from reaction between HOCl and O 2 Ϫ . No evidence was found for nitration by reactive nitrogen species; kinetic arguments suggest that the HOCl-NO 2 Ϫ reaction will be relatively unimportant in physiological environments.