Nitric Oxide is a Physiological Substrate for Mammalian Peroxidases

We now show that NO serves as a substrate for multiple members of the mammalian peroxidase superfamily under physiological conditions. Myeloperoxidase (MPO), eosinophil peroxidase and lactoperoxidase all catalytically consumed NO in the presence of the co-substrate hydrogen peroxide (H 2 O 2 ) at neutral pH. Near identical rates of NO consumption by the peroxidases were observed in the presence versus absence of plasma levels of Cl - (100 mM). Although rates of NO consumption in buffer were accelerated in the presence of a superoxide-generating system, subsequent addition of catalytic levels of a model peroxidase, MPO, to NO-containing solutions resulted in the rapid acceleration of NO consumption. The interaction between NO and Compounds I and II of MPO were further investigated during steady-state catalysis by stopped-flow kinetics. NO dramatically influenced the build-up, duration and decay of steady state levels of Compound II, the rate limiting intermediate in the classic peroxidase cycle, in both the presence and absence of Cl - . Collectively, these results suggest that peroxidases may function as a catalytic sink for NO at sites of inflammation, influencing its bioavailability. They also support the potential existence of a complex and interdependent relationship between NO levels and the modulation of steady-state catalysis by peroxidases in vivo .


INTRODUCTION
Nitric oxide (NO, nitrogen monoxide) plays essential bioregulatory roles in a wide range of processes critical to normal functions in the cardiovascular, nervous, and immune systems (1,2).
Under pathological conditions, such as during inflammation and vascular disease, rates of NO consumption become excessive and impaired response to endothelium-derived relaxing factor or NO are observed (1). Accordingly, factors which influence rates of NO removal following its synthesis by nitric oxide synthases (NOS) are of significant interest. The autoxidation of NO in aqueous solutions is slow at physiological concentrations of O 2 and NO (3). In the vascular compartment, a major pathway for NO removal is through near diffusion-controlled interaction with erythrocyte oxyhemoglobin yielding ferric (met)hemoglobin and nitrate (NO 3 -15-lipoxygenase (14), an enzyme implicated in atherogenesis (15,16). The role of these pathways in modulating NO-dependent signaling in vivo remains to be determined.
The ability of NO to react with hemoproteins at nearly diffusion-controlled rates, promoting activation of guanylate cyclase and possibly inhibition of many heme and non-heme proteins by interacting with their metal centers is well known (17)(18)(19)(20)(21)(22)(23). Likewise, a variety of studies have documented the ability of NO to bind to the heme moiety of peroxidases (24)(25)(26)(27). Both spectroscopic and rapid kinetics measurements were recently used to demonstrate that NO rapidly binds to both ferric [Fe(III)] and ferrous [Fe(II)] forms of myeloperoxidase (MPO) (28), a hemoprotein which is present in abundance in neutrophils, monocytes and certain sub-populations of tissue macrophages, such as in atherosclerotic lesions (29,30). Although human neutrophils isolated from peripheral blood do not normally contain inducible NOS, neutrophils within human buffy coat preparations pre-treated with cytokines are reported to express inducible NOS (31). Moreover, immunohistochemical studies demonstrate that MPO and inducible NOS in cytokine-treated human neutrophils are both co-localized and secreted from the primary granules of activated leukocytes (31). Finally, numerous cell types generate NO at sites of inflammation. Hence, MPO typically performs its functions in environments where NO is formed. MPO uses H 2 O 2 and a variety of co-substrates to generate reactive oxidants and diffusible radical species (32)(33)(34)(35)(36)(37). Under physiological conditions, a major co-substrate is Clyielding hypochlorous acid (HOCl), a potent chlorinating oxidant with microbicidal and viricidal properties (38). The reactive species formed are thought to play a key role in the ability of MPO to promote destruction of invading parasites and pathogens during the host response (39)(40)(41). However, MPOgenerated oxidants are also linked to tissue oxidation in cardiovascular disease and other inflammatory disorders (42)(43)(44).  fell gradually to the origin as NO was depleted by autoxidation (Fig. 1A). Addition of H 2 O 2 to the reaction mixture had no significant effect on the rate of NO decay (Fig. 1A), similar to prior reports (24). Subsequent addition of either MPO, EPO or LPO to the reaction mixture caused a rapid decay in the level of free NO ( Fig. 1 To examine the potential physiological significance of these observations, we next determined the effect of additional substrates on peroxidase-catalyzed consumption of NO. MPO was initially used as a model peroxidase because of its abundance at sites of leukocyte recruitment and activation during inflammation (30), and its well-known use of the abundant halide Clas substrate (32,58).
Remarkably, the rates of NO consumption mediated by MPO in the presence versus absence of plasma levels of Clwere virtually indistinguishable (Fig. 2). These results are consistent with the fact that MPO is far from saturated at plasma levels of Cl - (52). NO consumption by MPO was to a large extent prevented by pre-incubation of the enzyme solution with sodium azide, a peroxidase inhibitor (59) was best fit to a single exponential function, giving an apparent pseudo first-order rate constant of 3.2 s -1 . The subsequent decrease in absorbance at 455 nm observed was also fit to a single exponential function with a rate constant of 0.008 s -1 and was attributed to the decay of compound II. Together, these results indicate that the buildup of MPO Compound II in the absence of NO is rapid, monophasic, and occurs with a much faster rate than its decay.
The addition of NO to reaction mixtures results in dramatic effects on the rates of MPO Compound II build-up, duration and decay, as assessed by stopped-flow spectroscopy (Fig. 6). NO was readily used as a one esubstrate by Compound I, as indicated by the rapid buildup of MPO Compound II (Fig. 5B). The rate of Compound II accumulation was enhanced nearly 20 fold in the presence of NO and increased in a concentration-dependent and saturable manner (Fig. 7A) (Fig. 2) and in the presence of a cell-free O 2 •--generating system both in the absence and presence of Cl - (Fig.   3). However, this process can be partially or completely blocked by pre-incubation of the enzyme sample with classic peroxidase inhibitors (Fig. 2) (Figs 8 and 9).
The ability of NO to influence Compound II rates of formation and decay strongly supports the notion that NO undergoes a one rather than a two eoxidation transition following interaction with  (Fig. 6,7) and presence ( Fig. 8,9)  NO also may serve as a ligand for MPO-Fe(III) leading to inhibition of peroxidase activity and formation of a MPO-Fe(III)-NO complex (Fig. 10) (28). Examination of the NO concentrationdependence for the rate of Compound II formation revealed saturable kinetics at levels of NO > 2.5 µM in both the absence (Fig. 7A) and presence (Fig. 9A) (Fig. 7) and presence (Fig. 9) (49). Despite these differences in experimental design, several major distinctions between NO interactions with MPO and HRP are apparent: 1) NO dramatically accelerates the rate of MPO Compound II formation in the nanomolar to low micromolar (< 2.5 µM) range (Fig. 7), whereas significantly higher levels of NO are required with HRP (49); 2) NO reduces HRP Compound II faster than HRP Compound I (49) -the opposite was observed with MPO (Figs. 7,9); and 3) the rate constant for the reduction of HRP Compound II by NO appears to be much faster than the rate of NO-dependent reduction of MPO Compound II. Regardless of these differences, NO appears to serve as a general substrate for both plant and animal peroxidases.
An interesting feature of MPO -NO interactions is its parallel behavior to that observed during peroxidase interactions with another physiological diatomic ligand, O 2 •-(45,48). Both serve as reductants for Compound II and lead to enhanced overall peroxidase activity due to acceleration of this rate limiting step in the peroxidase cycle. Both O 2 •and NO also serve as ligands for MPO-Fe(III) and generate inactive complexes, Compound III and MPO-Fe(III)-NO, respectively (Fig. 10). Formation of each is a reversible process and addition of H 2 O 2 to each result in spectral changes consistent with formation of Compound II. Thus, both Compound III and MPO-Fe(III)-NO may still promote peroxidation reactions. This contrasts with the mechanism for inactivating MPO following reduction of the ground state ferric to ferrous form (Fig. 10). Here, heme reduction appears to be accompanied by collapse of the heme pocket, as defined by any conformational alteration, however subtle, that limits access of substrate to the distal heme center (28). An example would be the binding of a 6th axial ligand from an amino acid residue on the opposing side/wall of the heme pocket. The slow rate of NO binding to MPO-Fe(II) observed (28) is consistent with ligand replacement rather than direct binding of NO to the MPO heme iron. Changes in heme pocket geometry upon ligand binding (e. g. movement of a number of amino acid residues and a rearrangement of active site water molecules) have been described for cytochrome c peroxidase (26). Moreover, slower rates of NO binding to the Fe(III) forms of a number of heme proteins have been attributed to ligand replacement (21,67).
Collapse in the heme pocket geometry upon heme reduction has also been reported for other heme proteins (68).
Another remarkable feature of the present studies is the demonstration that peroxidases catalytically consume NO under a variety of conditions that mimic those found in biological systems.
For example, NO consumption rates were not inhibited by addition of physiologically relevant amounts of alternative substrates, such as plasma levels of Cland the model peroxidase MPO (Fig. 2).
Studies with MPO examining H 2 O 2 consumption rates in the presence of plasma levels of Cland additional alternative substrates (e.g. thiocyanate) have revealed that MPO is far from saturated at plasma levels of Cl - (52). It is also remarkable that peroxidases like MPO effectively act as catalysts for NO consumption, even in the presence of a O 2 •--generating system (Fig. 3).