Myeloperoxidase scavenges peroxynitrite: A novel anti-inflammatory action of the heme enzyme.

Peroxynitrite, a potent pro-inflammatory and cytotoxic species, interacts with a variety of heme containing proteins. We addressed the question whether (i) the interaction of myeloperoxidase (MPO, an enzyme generating hypochlorous acid from hydrogen peroxide and chloride ions) with peroxynitrite affects the clearance of peroxynitrite, and (ii) if peroxynitrite could modulate the chlorinating activity of MPO. Our results show that this interaction promotes the decomposition of the highly reactive pro-inflammatory oxidant, whereby MPO Compound II (but not Compound I) is formed. The efficiency of MPO to remove peroxynitrite was enhanced by L-tyrosine, nitrite and (-)-epicatechin, substances known to reduce Compound II with high reaction rate. Next, peroxynitrite (added as reagent) diminished the chlorinating activity of MPO in the presence of hydrogen peroxide. Alternatively, SIN-1, a peroxynitrite donor, reduced hypochlorous acid formation by MPO, as measured by aminophenyl fluorescein oxidation (time kinetics) and taurine chloramine formation (end point measurement). At inflammatory loci, scavenging of peroxynitrite by MPO may overcome the uncontrolled peroxynitrite decomposition and formation of reactive species, which lead to cell/tissue damage.

Introduction

Peroxynitrite, a reactive oxidant with a short half-life (approx. 10–20 ms at physiological pH [1,2]), is generated by the diffusion-controlled reaction of nitric oxide (•NO) and superoxide anion (O2•−) with a rate constant of 1 × 1010 M−1 s−1 at pH 7.0 [3,4]. •NO is a relatively stable radical, while O2•− has much shorter half-life and undergoes spontaneous dismutation with a second-order rate constant of 2 × 105 M−1 s−1 at pH 7.4 [5].

Peroxynitrite is a potent pro-inflammatory and cytotoxic molecule [6-8], which exerts a variety of inflammatory effects such as inhibition of antioxidants [9] and ion channels [10,11], lipid peroxidation [12,13], thiol oxidation and tyrosine nitration of proteins [14] as well as modulation of cyclooxygenase activity [15,16]. Under inflammatory conditions, the excessive production of •NO and O2•− favors elevated levels of peroxynitrite that trigger dysregulation of cellular signaling pathways, activation of inflammatory stress and cell/organ injury in further consequence [17-19]. Peroxynitrite interacts with a variety of redox-active heme proteins including hemoglobin [20], myoglobin [20], catalase [21,22], lactoperoxidase [23], cytochrome P450 [24], nitric oxide synthase [25], cyclooxygenase-2 [26], and myeloperoxidase (MPO)1 [27,28], respectively.

In activated neutrophils, dimeric MPO contributes to the inactivation and killing of phagocytosed microorganisms [29,30]. Under inflammatory conditions, cationic MPO protein becomes attached to negatively charged proteins and membrane epitopes [30-33]. Most importantly, in the presence of hydrogen peroxide (H2O2), MPO catalyzes the oxidation of small molecules including tyrosine, tryptophan, nitrite, •NO, sulfhydryls, phenol and indole derivatives, xenobiotics and others [34-39]. As reported for the other peroxidases, the heme moiety of MPO undergoes characteristic conversions from the ferric state to the oxo-ferryl states: Compound I (which has an additional porphyryl radical function) and Compound II [40]. A unique property of MPO is its ability to oxidize chloride ions (Cl−) by the abstraction of two electrons to form hypochlorous acid (HOCl); during this reaction Compound I is reduced to the ferric state [41,42].

At inflammatory loci, both, the release of MPO from leukocytes as well as the formation of peroxynitrite might coincide. Peroxynitrite is known to interact with ferric MPO as well as with Compound I, whereby in both cases Compound II is formed [27]. Reports about the functional effects of peroxynitrite on MPO are scarce. Peroxynitrite diminishes the consumption of H2O2 by MPO and causes a heme depletion that could be prevented by Cl− [28]. Any effects of peroxynitrite on the chlorinating activity of MPO remain unknown. The same holds for factors modulating the clearance of peroxynitrite by MPO.

Therefore, we aimed to investigate the effects of MPO on peroxynitrite decomposition in order to determine conditions favouring enhanced removal of peroxynitrite. Furthermore, we addressed the question how peroxynitrite modulates the chlorinating activity of MPO.

Materials and methods

Materials

Peroxynitrite solutions (Cayman-81565, Hamburg, Germany) were prepared to minimize nitrite formation [43]. Stock solutions were prepared in 0.01 M NaOH to avoid protonation and homolytic decay of peroxynitrous acid. The concentration of peroxynitrite solutions was determined by measuring absorbance at 302 nm (ε302 = 1705 M−1 cm−1) [44]. For experiments, final pH values were measured (using pH meter) and adjusted to 9.0 by adding required volume of peroxynitrite stock solution. Purified human MPO was purchased from Planta Natural Products (Vienna, Austria). For denaturation, MPO was heated to 95 °C for 15 min. H2O2 (Sigma–Aldrich-216763, Munich, Germany) was diluted in cold water and the concentration was determined using ε240 = 43.6 M−1 cm−1 [45]. H2O2 solutions were used within 2 h after dilution.

5-Amino-3-(4-morpholinyl)-1,2,3-oxadiazolium chloride (SIN-1, Cayman-82220), aminophenyl fluorescein (APF, Cayman-10157), potassium cyanide (KCN, Sigma–Aldrich-60178), 4-aminobenzoic acid hydrazide (4-ABAH, Sigma–Aldrich-A41909), sodium nitrite (NaNO2, Sigma–Aldrich-71759), taurine (Sigma–Aldrich-T0625), 5,5′-dithiobis(2-nitrobenzoic acid) (TNB, Sigma–Aldrich-D8130), l-tyrosine (Sigma–Aldrich-93829), (−)-epicatechin (Sigma–Aldrich-68097), superoxide dismutase (SOD, Sigma–Aldrich-S7571) and catalase (Sigma–Aldrich-C40) were diluted in 10 mM phosphate buffered saline (PBS, Sigma–Aldrich-P4417, pH 7.4).

Kinetic study of peroxynitrite decomposition

Peroxynitrite decomposition was followed by measuring the absorbance at 302 nm every min at 37 °C for indicated time periods using a Varian Cary 50 UV–vis spectrophotometer (Mulgrave, Australia). The effects of MPO alone (either in its native or denatured form) or in the presence of KCN, 4-ABAH, NaNO2, (−)-epicatechin or l-tyrosine on the decomposition of peroxynitrite were examined. Absorbance values from at least three experiments were recorded and plotted as a function of time. Curves were fitted mono-exponentially to determine the corresponding decomposition rate constants (kobs). The kobs values were averaged and plotted against the proton ([H+]) or MPO concentration to determine corresponding catalytic rate constant (kcat).

Kinetic studies of the APF oxidation by MPO, SIN-1 or peroxynitrite

APF was added to MPO alone or in the presence of SIN-1/peroxynitrite at 37 °C and the formation of fluorescein from APF was measured every 30 s for indicated time periods. Two min after starting the experiments, H2O2 was added using an injection device followed by measurement of fluorescence intensities for another 30 min.

Furthermore, the effects of SIN-1/peroxynitrite on APF oxidation were assessed by addition of APF to peroxynitrite or SIN-1 (alone or supplemented with SOD or catalase) at 37 °C followed by recording fluorescence intensities at indicated time periods.

All fluorescence measurements were performed in 96-well plates (a final volume of 250 μl/well) using a fluorescence microplate reader Tecan Infinite 200 PRO (Männedorf, Switzerland). Fluorescence intensity was monitored at 488/522 nm (excitation/emission wavelengths) [46]. For normalization, fluorescence intensity of APF alone was used.

Detection of taurine chloramine

The chlorination of taurine by the MPO–H2O2–Cl− system was followed by measuring formation of taurine chloramine using TNB [47]. Briefly, taurine (10 mM) was pre-incubated with MPO (10 nM in PBS, pH 7.4) in the absence or presence of SIN-1 (10–50 μM in PBS, pH 7.4, kept at 37 °C for 2 h) for 1 min at 37 °C. Then, 25–150 μM H2O2 was added to induce the chlorinating activity of MPO. After 5 min the reaction was stopped by addition of 1000 U/ml catalase. After adding 60 μM TNB, formation of taurine chloramine was estimated at 412 nm using ε = 14,100 M−1 cm−1 [47]. Absorbance values were normalized to added H2O2 concentrations to calculate the efficiency of MPO to form HOCl.

Statistics

All values are represented as mean ± SEM and n represents the number of experiments. Statistical significances were tested by Student’s t-test or one-way ANOVA with adequate post hoc tests (Tukey or Dunnett), using IBM SPSS 20 software. All p values ⩽0.05 were considered statistically significant and all tests were 2-sided.

Results

Influence of pH and MPO on peroxynitrite decomposition

As peroxynitrite is highly unstable at physiological pH [1], we aimed to reveal the decomposition kinetics of this anion under basic conditions. For these experiments 200 μM peroxynitrite was used to get detectable absorbance values. Fig. 1A shows pH-dependent self-decomposition of peroxynitrite. At pH 9.0, about 37% peroxynitrite decomposed during 60 min, while at lower pH (8.0–8.5) faster decomposition was observed. However, at physiological pH peroxynitrite decomposed within a few seconds.

Fig. 1

Influence of pH and MPO on peroxynitrite decomposition. Peroxynitrite solutions (200 μM) were either kept at indicated pH values (A) or incubated with indicated MPO concentrations at pH 9.0 (C). Absorbance was measured at 302 nm every min (up to 60 min) at 37 °C. (A) Absorbance measurements were stopped when zero levels were reached. One representative kinetic curve (A and C) out of three independent experiments is shown. Curves were fitted mono-exponentially to determine the corresponding apparent rate constants (kobs) and statistical analysis is shown in (B and D). Values are expressed as mean ± SEM (n = 3) (B and D). ∗p ⩽ 0.05 vs. 0 nM MPO (D).

Next, curves from Fig. 1A were fitted mono-exponentially and kobs were calculated. Fig. 1B shows an inverse relation between kobs and pH. At pH 9.0, kobs was 0.01 ± 0.001 min−1, which increased with decreasing pH to 0.376 ± 0.0156 min−1 at pH 7.4. However, no linear correlation between [H+] and kobs values was found (Supplemental Fig. 1A).

Due to the high stability of peroxynitrite at pH 9.0, this condition was chosen to investigate the effects of MPO on peroxynitrite decay. MPO accelerated the decomposition of peroxynitrite in a concentration-dependent manner (Fig. 1C). In all experiments the concentration of Cl− was 0.14 M. An increase of MPO concentration (up to 40 nM) was paralleled by increased kobs values ranging from 0.013 ± 0.0021 min−1 to 0.025 ± 0.0017 min−1 at pH 9.0 (Fig. 1D). Data analysis revealed a linear correlation between MPO concentration and kobs resulting a kcat value of 6.25 ± 0.01 × 103 M−1 s−1 for rate-determining step of MPO-derived peroxynitrite decomposition (Supplemental Fig. 1B).

MPO-mediated peroxynitrite decomposition is independent of Compound I formation

As MPO decomposes peroxynitrite (Fig. 1C), we tried to reveal whether MPO activity is the prerequisite for this phenomenon. Fig. 2A and B shows that denatured MPO did not affect peroxynitrite decomposition kinetics. Then, we tested which redox states of MPO are involved in the enzyme-mediated peroxynitrite decomposition. In the presence of KCN (known to inhibit the enzyme activity by forming a low-spin adduct with ferric MPO [48,49]), kobs values decreased from 0.025 ± 0.0017 min−1 to 0.012 ± 0.003 min−1 (Fig. 2B). In contrast, 4-ABAH (an inhibitor of MPO that interacts with Compound I [50]) was ineffective to modulate MPO-mediated peroxynitrite decay (Fig. 2A and B). From these findings, it is obvious that active MPO (ferric state) is the prerequisite for peroxynitrite decomposition; the reaction is independent of Compound I formation.

Fig. 2

MPO-mediated peroxynitrite decomposition is independent of Compound I formation. (A) Peroxynitrite solutions (200 μM, pH 9.0) were incubated with either MPO (40 nM), denatured MPO (40 nM, denatured at 95 °C for 15 min) or MPO (40 nM) incubated with either KCN (40 nM) or 4-ABAH (1 mM). Absorbance was measured at 302 nm every min (up to 60 min) at 37 °C. One representative kinetic curve (A) out of three independent experiments is shown. Curves were fitted mono-exponentially to determine the corresponding apparent rate constants (kobs) and statistical analysis is shown in (B). Values are expressed as mean ± SEM (n = 3) (B). ∗p ⩽ 0.05 vs. control and #p ⩽ 0.05 vs. MPO (B).

Decomposition of peroxynitrite is accelerated by substances mediating reduction of Compound II to the ferric state

Stopped-flow data addressing the interaction of ferric MPO and peroxynitrite (at pH 7.0) revealed the formation of Compound II with a rate of 6.8 × 106 M−1 s−1. This is followed by a reverse reaction with an apparent rate of 440 M−1 s−1, whereby the reduction of Compound II was not mediated by peroxynitrite but a decay product [27]. Therefore, we aimed to analyze whether an accelerated conversion of Compound II to the native enzyme state promotes enzyme-mediated peroxynitrite decomposition.

Next, nitrite (200 μM), (−)-epicatechin (5 μM) or l-tyrosine (40 μM) was added to the mixture of MPO and peroxynitrite. Nitrite reacts with Compound II with the rate constant of 5.5 × 102 M−1 s−1 at pH 7.0 [36]. The bimolecular rate constant of (−)-epicatechin is 4.5 × 106 M−1 s−1 at pH 7.0, which is by far the highest reaction rate constant for Compound II reduction [51]. l-Tyrosine reduces Compound II with a rate of 1.6 × 104 M−1 s−1 at pH 7.0 [34]. Our results show that none of these substances (when tested alone) altered the decomposition kinetics (Fig. 3A, C and E) or rate constants (Fig. 3B, D and F) of peroxynitrite decay. Moreover, all three substances enhanced MPO-mediated peroxynitrite clearance (Fig. 3A, C and E). Data analysis of peroxynitrite decay revealed an increase of kobs values from 0.025 ± 0.0017 min−1 (MPO alone) to 0.059 ± 0.0072 min−1 (MPO + nitrite, Fig. 3B), 0.051 ± 0.0044 min−1 (MPO + (−)-epicatechin, Fig. 3D) or 0.032 ± 0.003 min−1 (MPO + l-tyrosine, Fig. 3F). Comparable results were obtained when different concentrations of these reducing agents were tested (data not shown). These data suggest that the promotion of the recovery of the ferric state by endogenous (e.g. nitrite, l-tyrosine) or exogenous ((−)-epicatechin) substances accelerates the removal of peroxynitrite.

Fig. 3

Decomposition of peroxynitrite is accelerated by substances mediating reduction of Compound II of MPO, at pH 9.0, MPO (40 nM in PBS, pH 7.4) was incubated with NaNO2 (200 μM, A), (−)-epicatechin (epi, 5 μM, C) or l-tyrosine (Tyr, 40 μM, E) followed by addition of peroxynitrite (200 μM). Absorbance was measured at 302 nm every min (up to 60 min) at 37 °C. One representative kinetic curve (A, C and E) out of three independent experiments is shown. Curves were fitted mono-exponentially to determine the corresponding apparent rate constants (kobs) and statistical analysis is shown in (B, D and F). Values are expressed as mean ± SEM (n = 3) (B, D and F). ∗p ⩽ 0.05 vs. control and #p ⩽ 0.05 vs. MPO (B, D and F).

SIN-1-induced oxidation of APF is a measure of peroxynitrite

In order to investigate potential effects of peroxynitrite on the chlorinating activity of MPO, selected experiments were performed using the peroxynitrite generator, SIN-1. Peroxynitrite formation was followed using the non-fluorescent dye, APF, which is oxidized to fluorescein upon interaction with peroxynitrite [52]. Fig. 4A demonstrates that the formation of peroxynitrite from SIN-1 was paralleled by an increase in APF oxidation at both, the physiological and basic pH. Thereby, a lag phase in the fluorescence increase was observed at pH 7.4 and 8.0, respectively. This is due to the fact that at first SIN-1 produces O2•− and •NO, two radicals that further react to form peroxynitrite. The most rapid increase in fluorescence intensity was observed at pH 9.0. As fluorescein is equally stable at pH values higher than 7.0 [53], the most rapid increase in APF oxidation at pH 9.0 is most likely due to the high stability of peroxynitrite. To sum up, peroxynitrite formation can be followed at pH 7.4 using SIN-1 and APF.

Fig. 4

SIN-1-induced oxidation of APF is a measure of peroxynitrite. SIN-1 solutions (20 μM at indicated pH [A] or indicated concentrations in PBS, pH 7.4 [B]) were added to 5 μM APF. SIN-1 solutions (20 μM in PBS, pH 7.4) were kept at 37 °C for indicated time periods (C) followed by addition of APF and measurement of APF fluorescence. Superoxide dismutase (SOD, 2000 U/ml) or catalase (1000 U/ml) was added to SIN-1 solutions (20 μM in PBS, pH 7.4) followed by APF addition and measurement of APF fluorescence (D). Fluorescence intensities were measured every 30 s (up to 30 min [A], 120 min [B and D] or 5 min [C]) at 488/522 nm (excitation/emission wavelength) at 37 °C. One representative kinetic curve (A–D) out of three independent experiments is shown.

At physiological pH, SIN-1 (5–80 μM) generated peroxynitrite in a concentration-dependent manner (Fig. 4B). However, it is unclear whether the observed plateau in fluorescence intensity is due to the lack of either peroxynitrite generation or APF-derived fluorescein. To clarify, time-dependent generation of peroxynitrite was studied. SIN-1 solutions (20 μM) were kept up to 4 h at 37 °C followed by APF fluorescence measurement for another 5 min. Fig. 4C shows that peroxynitrite generation at pH 7.4 reached a maximum between 2 and 3 h, while at longer incubation periods fluorescence intensities decreased. From these observations it is evident that a 2–3 h incubation period is a prerequisite for SIN-1 to generate maximum peroxynitrite levels.

Basically, APF does not react with O2•− or •NO [52], which are generated by SIN-1. To confirm that the observed fluorescence intensity is a measure of peroxynitrite alone, catalase or SOD were implemented. Catalase (a peroxynitrite scavenger [22]) inhibited SIN-1-mediated APF oxidation (approx. 85%, Fig. 4D), while SOD (a O2•− scavenger) was less effective (approx. 22%, Fig. 4D). The inefficiency of SOD could be due to the fact that peroxynitrite is formed from O2•− and •NO at nearly diffusion-controlled rate (1 × 1010 M−1 s−1 at pH 7.0), which is higher compared to the dismutation of O2•− by SOD (1–2 × 109 M−1 s−1 at pH 7.0) [3,4]. Since the reaction rate is dependent also on the reactant concentration, SOD inhibited peroxynitrite formation in a concentration-dependent manner (data not shown); thus data of the high SOD concentrations (2000 U/ml) used is shown in Fig. 4D. These rate constants as well as actual reactant concentrations contribute to the inefficiency of SOD to inhibit the SIN-1-mediated APF fluorescence.

SIN-1/peroxynitrite inhibit the chlorinating activity of MPO leading to accumulation of Compound II

Next, we were interested how peroxynitrite modulates the chlorinating activity of MPO. During these experiments Cl− concentration was 0.14 M. As APF is highly reactive to HOCl (but not to H2O2) [46,52], this dye was used to follow MPO activity. Fig. 5A demonstrates HOCl formation via the MPO–H2O2–Cl− system at pH 7.4. In case of 25 μM H2O2, we observed the highest fluorescence intensity (at 2 min), which decreased with increasing H2O2 concentrations. These data are indicative for the competition of H2O2 with Cl− for Compound I. The second order rate constant for both substrates with Compound I are 2.5 × 104 M−1 s−1 (for Cl− [54]) and 4.4 × 104 M−1 s−1 (for H2O2 [55]) at pH 7.0. With increasing H2O2 levels, this competition favors the accumulation of Compound II that is unable to oxidize Cl−. Moreover, after reaching a maximum (at 2 min, 25–50 μM H2O2, Fig. 5A), fluorescence intensities decreased; this is most probably due to the self-quenching of fluorescein and the formation of chlorinated fluorescein products [53].

Fig. 5

SIN-1/peroxynitrite inhibit the chlorinating activity of MPO leading to accumulation of MPO Compound II. MPO (10 nM in PBS, pH 7.4) was added to indicated H2O2 concentrations (A) followed by measurement of APF fluorescence (5 μM APF). Effects of SIN-1 (20 μM in pH 7.4, kept at 37 °C for 2 h, B and D) or peroxynitrite (ONOO−, 10 nM, pH 9.0, C and E) on the MPO (10 nM)-H2O2 (25 μM)-chloride system (B and C) or the interaction of MPO with SIN-1/peroxynitrite (D and E) were analyzed by fluorescence measurements. Fluorescence intensities were measured every 30 s (up to 30 min) at 488/522 nm (excitation/emission wavelength) at 37 °C. To compare different curves and to avoid congestion, all curves were marked by symbols every 2 min (A–C) or 4 min (D and E). In all experiments, initially MPO was incubated with APF and SIN-1/peroxynitrite followed by H2O2 addition using an injector (2 min after fluorescence measurements started). One representative kinetic curve (A–E) out of three independent experiments is shown. The vertical arrow (between dotted horizontal lines) in B and C represents the inhibition of MPO activity at 2 min.

Next, SIN-1 (diluted in PBS for 2 h) or peroxynitrite solutions were added to MPO prior to the addition of H2O2 (25 μM) (Fig. 5B and C). The initial increase of fluorescence was lower in samples containing SIN-1 or peroxynitrite. Our data show that MPO-mediated formation of HOCl was impaired (approx. 65% at 2 min) by either SIN-1 (20 μM, pH 7.4, Fig. 5B) or peroxynitrite (10 nM, pH 9.0, Fig. 5C). In line with previous data [56], the chlorinating activity of MPO was lower (approx. 30% at 2 min) at pH 9.0 (Fig. 5C) compared to pH 7.0 (Fig. 5B). In contrast to diluted SIN-1 (2 h, Fig. 5B), freshly prepared SIN-1 (Supplemental Fig. II) did not alter the chlorinating activity of MPO. Thus, peroxynitrite-mediated Compound II formation (and not O2•−-mediated Compound III formation) is the primary cause for SIN-1-induced inhibition of MPO activity.

Fig. 5D and E shows SIN-1/peroxynitrite-induced APF oxidation, which is decreased (approx. 15% or 30%, respectively) in the presence of MPO. In the absence of SIN-1/peroxynitrite, MPO did not induce any APF oxidation. These observations imply that MPO scavenges peroxynitrite not only at pH 9.0 (Figs. 1C and 5E) but also at physiological pH (Fig. 5D). As APF gets oxidized by HOCl (Fig. 5A) as well as to a lower extent by SIN-1/peroxynitrite (Fig. 5D and E), fluorescence intensities generated by MPO + SIN-1/peroxynitrite were subtracted from those generated by the MPO–H2O2–Cl− system in the presence of SIN-1/peroxynitrite; only subtracted values are shown in Fig. 5B and C.

In selected experiments, the production of HOCl by the MPO–H2O2–Cl− system was assessed using taurine as a HOCl scavenger. The formation of taurine chloramine was inversely related to H2O2 concentrations (Fig. 6A). The chlorinating activity of MPO accounted in between 91% (25 μM H2O2) and 6% (150 μM H2O2). Next, taurine chloramine measurements were performed in the presence of a peroxynitrite donor. SIN-1 impaired the chlorinating activity of MPO (in the presence of 25 μM H2O2) in a concentration-dependent manner (Fig. 6B). Using increasing SIN-1 concentrations the inhibition of taurine chloramine formation ranged in between 9% (10 μM) and 55% (50 μM) (Fig. 6B). Control experiments revealed that SIN-1 did not affect the detection of taurine chloramines by the TNB assay (data not shown). The findings from Fig. 6A and B parallel data presented in Fig. 5A and B.

Fig. 6

SIN-1 inhibits the efficacy of MPO to generate HOCl and in tern taurine chloramine. MPO (10 nM in PBS, pH 7.4) was incubated for 1 min with (A) 10 mM taurine or (B) the combination of 10 mM taurine with SIN-1 (0–50 μM in PBS, pH 7.4, kept at 37 °C for 2 h). Then (A) indicated concentration of H2O2 or (B) 25 μM H2O2 was added and the samples were incubated for 5 min at 37 °C. The reaction was terminated by addition of 1000 U/ml catalase. Afterwards 60 μM TNB was added followed by absorbance measurement at 412 nm to estimate HOCl-mediated taurine chloramine formation. Absorbance values were normalized to H2O2 concentrations to calculate the efficiency of MPO to form HOCl. Values are expressed as mean ± SEM (n = 3) (A and B). ∗p ⩽ 0.05 vs. 0 nM SIN-1.

Discussion

Peroxynitrite is known to induce the oxidation of ferric MPO to Compound II followed by a much slower back conversion to the ferric enzyme [27,28]. The reverse reaction is not mediated by peroxynitrite, but by a decay product [27]. These results were obtained by time-resolved stopped-flow measurements on the basis of absorbance changes of the MPO forms (Fig. 7A). In order to clarify the effects of MPO on peroxynitrite clearance, we examined the reaction between MPO and peroxynitrite at 302 nm, the absorbance maximum for peroxynitrite.

Fig. 7

Scheme for the interaction between peroxynitrite and MPO. (A) The effects of peroxynitrite on MPO. Reactions mediated by peroxynitrite are highlighted with bold black arrows. Gray arrows indicate all other reactions. The protonated form of peroxynitrite, peroxynitrous acid, oxidizes ferric MPO to Compound II (reaction 1), which in turn is reduced by NO2− or other substrates (here collectively given as AH) to the ferric enzyme form. This cycle between ferric MPO and Compound II (reactions 1 and 2) mediates the decomposition of peroxynitrite. In the presence of H2O2, ferric MPO is oxidized to Compound I (reaction 3). Peroxynitrite reduces Compound I to Compound II (reaction 4). In this way, peroxynitrite affects other reactions of Compound I such as the oxidation of chloride (reaction 5) or the one-electron oxidation of other substrates (reaction 6). (B) The effects of Compound II-resolving substances on MPO-mediated peroxynitrite decomposition. Peroxynitrite decomposition by MPO (reaction 1) is accelerated by substances (e.g. (−)-epicatechin, tyrosine, NO2−), which reduce Compound II to the ferric enzyme form (reaction 2).

The pKa value of peroxynitrite is 6.8 [57]. While peroxynitrous acid decomposes homolytically to nitrogen dioxide and hydroxyl radical, the deprotonated form decays in the presence of carbon dioxide to nitrogen dioxide and carbon trioxide anion radicals [57]. Thus, the decomposition of peroxynitrite is strongly pH-dependent. To minimize the rate of self-destruction of peroxynitrite, we incubated peroxynitrite and ferric MPO at pH 9.0. Under these conditions, ferric MPO scavenges and inactivates peroxynitrite in a concentration-dependent manner. Both, the denaturated MPO and the MPO-cyanide complex resulting from binding of hydrocyanic acid to ferric MPO [48,49] were unable to affect the decomposition of peroxynitrite. Thus, the heme structure in ferric MPO should be intact and freely accessible for peroxynitrite.

Compound II, but not Compound I, is involved in MPO-mediated decomposition of peroxynitrite. This is evident by the inability of 4-ABAH to affect peroxynitrite clearance. To act as an inhibitor of the halogenating capacity of MPO, 4-ABAH should first be oxidized by Compound I to the 4-ABAH radical [50]. Furthermore, selected substances known to reduce Compound II markedly enhanced the decomposition of peroxynitrite in a concentration-dependent manner. For our experimental approach, (−)-epicatechin, l-tyrosine, and nitrite were chosen; these substances did not affect the self-destruction of peroxynitrite.

Thus, during the inactivation of peroxynitrite by MPO, the enzyme cycles permanently between ferric form and Compound II, whereby the reverse reaction to ferric MPO is the rate-determining step. Intriguingly, (−)-epicatechin is effective in enhancing MPO-mediated peroxynitrite removal at low micromolar concentrations. This flavonoid has by far the highest rate constant with Compound II [51]. It is likely that other related flavonoids exhibit similar properties towards Compound II. In other words, at inflammatory sites MPO dampens nitrosative stress by removal of peroxynitrite. This process is enhanced by suitable exogenous catalysts such as (−)-epicatechin and related flavonoids. These substances are known to exert anti-inflammatory and health-promoting effects [58-60].

In the presence of H2O2, the short-lived MPO intermediate Compound I is formed that can oxidize Cl− and other substrates [34-39,41,42]. Peroxynitrite decreased the chlorinating activity of MPO. This was shown by applying either the peroxynitrite generator SIN-1 or addition of peroxynitrite. There are two aspects necessary to consider in interpretation of these results. Both H2O2 and peroxynitrite are known to react with ferric MPO, whereby H2O2 has a 2–3-fold higher reaction rate constant than peroxynitrite at pH 7.0 [27,55]. Although the reaction of ferric MPO with H2O2 will dominate, some competitive interference of this enzyme form with peroxynitrite diminishes the consumption of H2O2 [28]. Moreover, SIN-1 produces O2•− that may transform ferric MPO into Compound III [61].

Alternatively, there is further competition of peroxynitrite and its decay products with Cl− for Compound I. A bimolecular rate constant of 7.6 × 106 M−1 s−1 was determined for the reaction between peroxynitrite and Compound I at pH 7.0, where Compound I is reduced to Compound II [27]. Nitrite resulting from the decomposition of peroxynitrite also reacts with Compound I at a rate of 2 × 106 M−1 s−1 at pH 7.0 [36]. Thus, due to addition of SIN-1 several new products will be formed that may compete efficiently with Cl− for Compound I. It is likely that freshly dissolved SIN-1 does not contain sufficient amounts of peroxynitrite or its decay products (see Supplemental Fig. II).

In the present study, we implemented the non-fluorescent dye, APF, which is converted to fluorescein by both peroxynitrite and HOCl [52], while other oxidants such as O2•−, •NO and H2O2 do not oxidize APF [52]. As long as peroxynitrite or HOCl is present alone, the use of APF for the detection of these oxidants is uncomplicated. In some experiments (Fig. 5B–E), both peroxynitrite and HOCl were present concomitantly. We concluded that peroxynitrite dampens the chlorinating activity of MPO and favors the reduction of Compound I to Compound II. APF is highly oxidized by HOCl and to a much lower extent by peroxynitrite (when both reactive species are present); at neutral pH HOCl/OCl− is much more stable than peroxynitrite, and HOCl has much higher affinity towards APF compared to peroxynitrite [52]. Moreover, the actual concentration of peroxynitrite in SIN-1 solutions accounts only 1–3% of added SIN-1 [62]. This is also observed when APF was used to measure peroxynitrite generated by SIN-1. In these experiments maximum fluorescence intensity was observed while using 10 μM APF and 0.5 mM SIN-1 [63]. In contrast, there is nearly an equimolar reaction between HOCl and APF [46].

As an alternative approach, taurine was used to determine HOCl production [47]. In line with APF data, following taurine chloramine formation we show that MPO activity is dependent on H2O2 concentration. Moreover, SIN-1 impairs the efficacy of MPO to generate HOCl. However, APF has an advantage over taurine to monitor time kinetics of the chlorinating activity of MPO. Taurine is commonly used to follow the end point production of HOCl after a fixed time interval. Taurine is a weak scavenger of peroxynitrite [64]. Furthermore, catalase acts a scavenger of peroxynitrite in addition to H2O2. Thus, the use of this scavenger during our experimental conditions removes both substances [21,22]. In fact, control experiments showed no TNB oxidation while using SIN-1 alone. In addition to 25 μM H2O2 (Fig. 6B), SIN-1 inhibited taurine chloramine formation in a concentration-dependent manner also at higher H2O2 concentrations (data not shown).

Importantly, the interaction of peroxynitrite with MPO leads, both in the absence or presence of H2O2, to the formation of Compound II, which is unable to oxidize Cl− [54]. Only a few number of exogenous (flavonoids such as (−)-epicatechin [51]) and endogenous substances (tyrosine [34], serotonin [65], ascorbate [66], and urate [67]) are known to reduce Compound II with high reaction rates. Thus, at inflammatory loci, following interaction of MPO with peroxynitrite, Compound II will accumulate when antioxidants get consumed. Concomitantly, the anti-inflammatory activity of MPO is disturbed by peroxynitrite.

Other heme proteins [20-26] and iron-porphyrin complexes [68,69] also contribute to peroxynitrite scavenging. During the interaction of peroxynitrite with ferric porphyrin complexes, oxo-ferryl intermediates were observed [68], which correspond to Compound II. Catalysts of peroxynitrite decomposition possess an anti-inflammatory activity [70-72].

Several adverse effects have been reported for peroxynitrite. These effects might be potentiated under conditions of antioxidant depletion and reduced removal of peroxynitrite at inflammatory loci. Peroxynitrite can permeate through cellular membranes and can adversely affect surrounding cells [1,2,73]. The reduction of neutrophil migration during sepsis and the exocytosis of its vacuoles are influenced by peroxynitrite [74,75]. In vivo production rate of peroxynitrite by macrophages was estimated approx. 50–100 μM/min [73]. Moreover, lipopolysaccharide-stimulated neutrophils and monocytes (where MPO accounts up to 5% and 1%, respectively, of total cell protein content [39]) can produce and release significant amounts of peroxynitrite [76].

Conclusions

Our data demonstrate that the heme enzyme MPO is a potent endogenous scavenger for peroxynitrite. During this interaction Compound II is formed. The efficacy of MPO to decompose peroxynitrite is enhanced by substances known to reduce Compound II with sufficient reaction rates (Fig. 7B).

MPO is a complex enzyme with beneficial as well as harmful effects [30]. Scavenging of peroxynitrite by MPO in vivo could act as a protective mechanism to avoid damage under oxidative/nitrosative stress. This may limit the formation of highly reactive species such as hydroxyl radical or carbon trioxide anion radicals formed during decomposition of peroxynitrite and peroxynitrous acid [57]. Moreover, excess formation of peroxynitrite in combination with depletion of antioxidants and inactivation of heme proteins (Fig. 7A) can lead to a pathophysiological scenario, where cells/tissues may be damaged by reactive oxygen/nitrogen species. Therefore, it is of importance to get insight into specific protective mechanisms in order to optimize therapeutic strategies to overcome inflammatory conditions.