High dissociation rate constant of ferrous-dioxy complex linked to the catalase-like activity in lactoperoxidase.

Heme reduction of ferric lactoperoxidase (LPO) into its ferrous form initially leads to the accumulation of the unstable form of LPO-Fe(II), which spontaneously converts to a more stable species, the two of which can be identified by Soret peaks at 440 and 434 nm, respectively. Our data demonstrate that both LPO-Fe(II) species are capable of binding O(2) at a similar rate to generate the ferrous-dioxy complex. Its formation with respect to O(2) was first order and monophasic and with rate constants of k(on) = 3.8 x 10(4) m(-1) s(-1) and k(off) = 11.2 s(-1). The dissociation rate constant for the formation of LPO-Fe(II)-O(2) is relatively high, in contrast to hemoprotein model compounds. This high dissociation rate can be attributed to a combination of effects that include the positive trans effect of the proximal ligand, the heme pocket environment, and the geometry of the Fe-O(2) linkage. Our results have also shown that the decay of the LPO-Fe(II)-O(2) complex occurs by two sequential O(2)-independent steps. The first step involves formation of a short-lived intermediate that can be characterized by its Soret absorption peak at 416 nm and may be attributed to the weakening of the Fe(II)-O(2) linkage with a rate constant of 0.5 s(-1). The second step is spontaneous conversion of this intermediate to generate the native enzyme and presumably superoxide as end products with a rate constant of 0.03 s(-1). A comprehensive kinetic model that links LPO-Fe(II)-O(2) complex formation to the LPO catalase-like activity, combined with the classic catalytic cycle, is presented here.

Hydrogen peroxide (H 2 O 2 ) is freely diffusible through biological membranes, and its overproduction or inhalation is extremely destructive to cells and tissues (1,2). Superoxide (O 2 . ) is mainly formed through the univalent reduction of molecular oxygen by the membrane-bound NADPH oxidase in phagocytic cells and endothelial cells (3). In most non-phagocytic cells, however, the main source of O 2 . is from the mitochondria (3).
Most of the O 2 . generated in vivo is utilized primarily to produce H 2 O 2 nonenzymatically or by superoxide dismutase (4). Nitric oxide synthase and several oxidase enzymes, such as monoamine and amino acid oxidase, can also directly produce H 2 O 2 (5,6). Because H 2 O 2 has both beneficial and harmful functions, its effect in a given biological setting requires precise control to regulate its action at tissues throughout the body. Therefore, understanding the circumstances that influence the rate of H 2 O 2 synthesis and degradation is of critical importance for optimal biological function. Catalase and cytoplasmic glutathione peroxidase are the major H 2 O 2 scavengers and protect cells from the toxic effects of H 2 O 2 by catalyzing its decomposition into molecular oxygen and water (4,7,8). Recently, lactoperoxidase (LPO) 1 has been identified as the major macromolecular consumer of H 2 O 2 in animal airway secretions (9,10). At this site, catalase displays very low levels of expression that are insufficient to protect the epithelium against pathogens (11,13). In addition, LPO and other members of the mammalian peroxidase superfamily (e.g. eosinophil peroxidase and myeloperoxidase) are all implicated in NO scavenging and protein nitration, as well as production of cytotoxic substances that protect against microbial, fungal, and bacterial infections (14 -22). LPO is a hemoprotein enzyme that typically uses H 2 O 2 with a combination of halides or pseudohalides to generate the corresponding hypohalous acid as a final end product (23)(24)(25)(26)(27). LPO also utilizes H 2 O 2 in combination with various organic and inorganic substrates to generate free radicals and other reactive components as primary end products that ultimately lead to nitration or halogenation of tyrosine residues or nitrosation of thiol residues in proteins (12-17, 28 -32). Organic and inorganic compounds may also serve as substrates enhancing the catalytic cycle of mammalian peroxidases by accelerating Compound II (LPO-Fe(IV)-O) formation and subsequent decay to the ground state, which is thought to be the rate-limiting step in the classic peroxidase cycle (12-17, 28 -32). The enzyme has been identified as an antimicrobial agent in milk, saliva, and tears and is produced by goblet cells and submucosal glands that form airway mucus secretions (33,34). A variety of evidence suggests that the peroxidase heme prosthetic group is involved in a wide range of important processes through its catalytic reactions that include binding, transport, and activation of oxygen, as well as deactivation of reactive oxygen compounds, oxidation/reduction reactions, and electron transport (17,19,35). In addition, the heme prosthetic group of LPO accommodates a large variety of molecules as ligands of the iron cation (17). Binding of these ligands to the peroxidase heme iron causes enzyme inhibition (17).
The presence of two different conformational states of the reduced form of LPO (LPO-Fe(II) state) has been reported in studies using a variety of spectroscopic techniques (36 -38). However, functional differences between the two forms have not been reported until recently (39). For example, previous resonance Raman spectroscopy studies demonstrated that the two LPO-Fe(II) species display no significant differences in the character of the Fe-N(histidine) bond, and both share the same state configuration (40). However, our recent diatomic binding studies have shown the existence of two spectroscopically and kinetically distinguishable LPO-Fe(II) species at equilibrium, one with the partially open heme pocket and one with the heme pocket relatively closed (39). We have also demonstrated that the presence of pathophysiologically relevant levels of peroxidases and H 2 O 2 serves as a catalytic sink for NO and reversibly inhibits NO-mediated bronchodilation of preconstricted tracheal rings (14 -16). Therefore, LPO and other members of mammalian peroxidase superfamily display multiple functions in airway diseases.
Characterization of the LPO-Fe(II)-O 2 complex, known as Compound III, generated by the addition of a slight excess of H 2 O 2 relative to LPO native enzyme has been accomplished recently by employing techniques such as optical absorption and resonance Raman spectroscopy (37,(41)(42)(43). However, what role the LPO-Fe(II)-O 2 complex plays during LPO catalysis is still unclear. In this study we examine the potential physiological relevance of the LPO-Fe(II)-O 2 interaction. We utilize direct rapid kinetic methods to characterize O 2 interactions with both forms of LPO-Fe(II). A modified comprehensive kinetic model is presented that describes LPO-Fe(II)-O 2 accumulation, decay, and interaction with H 2 O 2 , as well as involvement in the catalase-like activity during steady state catalysis.

EXPERIMENTAL PROCEDURES
Materials-O 2 gas was purchased from Matheson Gas Products, Inc., and used without further purification. All other reagents and materials were of the highest purity grades available and obtained either from Sigma or from Aldrich (Milwaukee, WI). Bovine LPO was obtained from Worthington Bio-Chemistry Corp. (Lakewood, NJ) and used without further purification. Purity was confirmed by demonstrating a RZ of Ͼ0.75 (A 415 /A 280 ), as well as by SDS-PAGE analysis. LPO concentration was determined spectrophotometrically by utilizing an extinction coefficient of 112,000 M Ϫ1 cm Ϫ1 at 412 nm (44).
Optical Spectroscopy and Rapid Kinetic Measurements-Optical spectra were recorded on a Cary 100 UV-visible spectrophotometer at 25°C. Anaerobic spectra were recorded using septum-sealed quartz cuvettes that were attached through quick-fit joints to an all-glass vacuum train system. LPO samples were made anaerobic by several cycles of evacuation and equilibrated with catalyst-deoxygenated N 2 . Separate buffer solutions were evacuated, gassed with N 2 , and anaerobically transferred either to the stopped-flow instrument or to anaerobic cuvettes using gas-tight syringes. Cuvettes were maintained under N 2 positive pressure during spectral measurements.
All kinetic measurements were performed with a temperature-controlled stopped-flow apparatus (Hi-Tech Scientific, model SF-61) equipped for anaerobic work. Addition of a slight excess of dithionite to an anaerobic solution of LPO-Fe(III) initially caused rapid buildup of an unstable LPO-Fe(II) intermediate that can be characterized by its Soret absorption peak at 444 nm. This oxygenated intermediate converts slowly (t1 ⁄2 ϭ ϳ30 min at 10°C) to a more stable intermediate that displays a Soret absorption peak at 432 nm. The reaction for O 2 binding to both the stable and the unstable forms of LPO-Fe(II) was monitored at wavelengths determined based on the spectral changes that occur upon O 2 binding to the LPO-Fe(II). Experiments were carried out at 10°C and initiated by rapidly mixing equal volumes of the unstable form of LPO-Fe(II) (1.45 M) (obtained immediately after reduction) with buffer solution supplemented with increasing concentrations of O 2 . In parallel, a similar study was repeated when a solution of the stable form of LPO-Fe(II) was rapidly mixed with a buffer solution supplemented with increasing concentrations of O 2 . To determine the apparent rate constants for generation of the LPO-Fe(II)-O 2 complex, the time course of absorbance changes was fit to single (Y ϭ 1 Ϫ e Ϫkt ) or double exponential (Y ϭ Ae Ϫk1t ϩ Be Ϫk2t ϩ C) functions using a nonlinear least squares method provided by the instrument manufacturer. Signal-to-noise ratios were improved by averaging 7-10 individual traces for each experiment.
Solution Preparation-The molecular oxygen concentration was determined using an Apollo 4000 device (World Precision Instruments, Sarasota, FL) equipped with an O 2 -selective electrode. The various solutions were made by mixing different volumes of O 2 -saturated buffer with anaerobic buffer solutions. The O 2 -saturated buffer was made by bubbling O 2 gas through the solution for 1 h in a septum-tipped flask. Utilizing the O 2 -selective electrode, the estimated ratio of the current (nA) measurement of the O 2 sensor for O 2 -saturated and air-equilibrated buffers is ϳ4.8. Using an estimated concentration of air-equilibrated buffer (21% O 2 ) of ϳ250 M (45), we calculated that the O 2 concentration of O 2 -saturated buffer is ϳ1.2 mM. These values, with the corresponding correction for temperature, were used to calculate the final O 2 concentrations of all the various O 2 solutions used in this study. The current for 0% oxygen concentration of the O 2 sensor was adjusted prior to each measurement by adding several mg of dithionite, a strong reducing agent, per 20 ml of buffer. All the oxygen solutions were then immediately transferred to the drive syringes of the stopped-flow apparatus that were surrounded by a thermostated anaerobic water jacket to maintain a constant temperature of 10°C.

Initial Rapid Spectroscopic Characterization of O 2 Binding
to Ferrous LPO-To determine the role of the LPO-Fe(II)-O 2 complex in catalytic activity, as well as to further our understanding of the potential role of LPO in catalase-like function, the direct reaction between LPO-Fe(II) and O 2 was carried out using rapid kinetic measurements. Addition of a slight molar excess of dithionite to LPO-Fe(III) has been shown previously to cause immediate LPO heme iron reduction, as judged by a shift in the Soret absorption peak from 412 to 444 nm, and the appearance of additional absorbance peaks in the visible range at 561 and 595 nm. This intermediate is unstable and converted over a short period of time to a more stable form of LPO-Fe(II) that can be characterized based on the Soret absorbance spectrum centered at 434 nm (39). The absorption spectrum was obtained as a function of time by rapidly mixing a solution of the stable and unstable forms of LPO-Fe(II) with an equal volume of air-saturated buffer (ϳ250 M O 2 ) as shown in Fig. 1. In both cases, the formation of the LPO-Fe(II)-O 2 intermediate was essentially the same and completed within 0.5 s after mixing. This intermediate has an absorption Soret peak centered at 426 nm and additional visible peaks at 549 and 588 nm (Fig. 1, A and B, respectively). In both cases, the spectral changes as a function of time indicate that the reactions involve a simple one-step mechanism in which both generate a six-coordinate ferrous-dioxy complex, Compound III. The subsequent change in the absorption spectra was a result of the decay of this intermediate to ground state, which occurs through the generation of a transient intermediate whose spectrum was characterized by absorbance peaks at 416, 548, and 586 nm ( Fig. 2A). This spectrum differs from that of either LPO Compound I or LPO Compound II, whose Soret absorption peaks centered at 410 and 430 nm, respectively, but is similar to that of oxyhemoglobin and nitric oxide synthase ferrousdioxy complex (46 -49), indicating that this is another oxygenated LPO intermediate. The spectral time line in Fig. 1 is much different from that found in Fig. 2. Fig. 1A shows spectra collected at 3. of LPO were identical and obtained by a fit of the data to one exponential function equation. The second order rate constants for the reaction of LPO-Fe(II) with O 2 were obtained by a fit of the pseudo-order rates, determined as a function of the oxygen concentration, to one exponential function. The plot of the pseudo-first order rate constant for the initial Fe(II)-O 2 complex formation was linear with a positive intercept, indicating that the reaction is reversible and proceeds through a simple one-step mechanism with k on of 3.8 ϫ 10 4 M Ϫ1 s Ϫ1 and k off of 11.2 s Ϫ1 (Fig. 3). The decay of the LPO-Fe(II)-O 2 complex was monitored by following the absorbance changes at 460 nm. These decreases in absorbance were successfully fitted to a double exponential function with observed rates of 0.5 and 0.03 s Ϫ1 . The subsequent conversion of this intermediate into the kinetically and spectroscopically distinct intermediate occurred via mechanisms independent of O 2 concentration (Fig. 4). DISCUSSION The present studies support the notion that the formation of the LPO-Fe(II)-O 2 complex during steady state catalysis is a fundamental feature of the kinetic reactions of LPO. Its formation and decay operate to accelerate the removal of the unwanted H 2 O 2 from the LPO milieu. This is specifically important in the human airway where catalase, a major H 2 O 2 scavenger, exists in very low levels (50), thus leaving LPO to serve as the major H 2 O 2 scavenger at this site.
One of the most remarkable findings of the present study is the low affinity of LPO-Fe(II) toward O 2 and the instability of this complex in contrast to related hemoprotein model compounds (46 -49, 51-54). peak at 416 nm prior to its decay to the ground state. These observations are consistent with structural data of hemoprotein model compounds, in which changes in the position of the heme iron lead to a fundamental alteration in their UV-visible spectra (51)(52)(53)(54).
A modified working kinetic model that incorporates our current results is shown in Fig. 5. In this model LPO reacts rapidly and reversibly with H 2 O 2 generating a ferryl -cation radical, Compound I (reaction 1). LPO Compound I is capable of oxidizing co-substrates such as SCN Ϫ through a single 2e Ϫ transition generating LPO-Fe(III) and the corresponding hypohalous acid (reaction 2) (23)(24)(25)(26)(27). Compound I may also oxidize an array of substrates through two sequential one-e Ϫ steps forming a long-lived intermediate, Compound II, and LPO-Fe(III) (reactions 3 and 4), respectively (28 -31, 55). O 2 . interacts with the LPO-Fe(III) heme iron to generate LPO-Fe(II)-O 2 (reaction 5). Formation of Fe(II)-O 2 complexes through this route is reversible, is relatively fast, and occurs via a one-or two-step mechanism (56,57). The presence of superoxide dismutase completely inhibited Compound III formation, but the presence of catalase had no significant effect on this process (57). In the presence of a slight excess of H 2 O 2 , Compound II is readily converted to Compound III (reaction 6). LPO heme reduction causes immediate buildup of a transient intermediate that displays a relatively open, unrestricted pocket (reaction 7), which converts with time to one that is constrained and able to bind small ligands at low rates (reaction 8) (39,58,59). Despite the remarkable alteration in the LPO-Fe(II) heme pocket, both LPO-Fe(II) forms display similar affinity toward O 2 and generate low spin six-coordinated ferrous-dioxy complex that can be characterized by the Soret absorbance peak at 424 nm and two resolved absorbance peaks centered at 549 and 588 nm (reactions 9 and 10) (42,60). The O 2 dissociation constant of LPO-Fe(II)-O 2 binding is significantly high compared with other hemoproteins (reactions 11 and 12) (47,51,54). Three major factors can account for the high dissociation rate in LPO-Fe(II)-O 2 including the positive trans effect contributed by the proximal ligand, the heme pocket environment, and the geometry of Fe-O 2 linkage. Indeed, previous resonance Raman spectroscopy studies demonstrated that the (Fe-O 2 ) frequency for LPO was considerably lower than those reported for related and relevant hemoprotein model compounds (37,61,62). Collapse or narrowing in the heme pocket geometry and/or the orientation of the O 2 bond may increase the interaction between the OϭO and the positively charged L-arginine that is localized above the heme iron on the distal side of LPO. Therefore, the electron density on the antibonding * orbital of the OϭO bond is pulled up by the influence of a positively charged residue localized on the distal side above the heme moiety (62,63). Weakening the Fe-O 2 linkage might allow the ligand trans to O 2 to pull the iron out of the plane away from the O 2 . This course of action is associated with the generation of a new oxygenated LPO species that can be identified by its Soret absorbance peak centered at 416 nm (reaction 13) similar to oxyhemoglobin and nitric oxide synthases (46,(47)(48)(49). The later intermediate is unstable, and its formation and autooxidation to ground state, which involves electron transfer to O 2 as a primary step to form superoxide (reaction 14), occurs in an oxygen-independent fashion. Our rapid kinetic measurements demonstrated that the rate for O 2 dissociation is relatively high and leads to the formation of LPO-Fe(II) (reactions 11 and 12). LPO-Fe(II) is not a dead end product as judged by its ability to utilize H 2 O 2 in the production of Compound II (reaction 15).
In regards to the biological relevance of our findings, patients with asthma and chronic obstructive pulmonary disease display higher levels of H 2 O 2 in their breath condensates compared with normal subjects (9,64,65). LPO and other mammalian peroxidases are also all present at high levels in the airways of asthmatic subjects. Although LPO functions to maintain sterility of the airway in the setting of constant exposure to inhaled debris and potential pathogens, it may also serve to scavenge the excess H 2 O 2 and protect the airway epithelium from H 2 O 2 toxicity. Related studies by Huwiler et al. (41) have demonstrated that oxygen is released by LPO in the presence of a slight excess of peroxidase. Kohler et al. (66) have also shown that the stoichiometry of O 2 release to the consumption of H 2 O 2 was 1:2, typical of catalase activity. It is important to note that this ratio is applied when high or low H 2 O 2 concentrations have been used (41). Several kinetic mechanisms have been proposed to provide a mechanistic explanation for these experimental findings and to explain the buildup of the LPO ferrous-dioxy complex during steady state catalysis (43,66). Our modified working kinetic model shown in Fig. 5 indicates the formation of an LPO-Fe-O 2 complex intermediate in the catalytic mechanism of the enzyme. As such, it might represent an alternative pathway, whose biological function is to accelerate the removal of H 2 O 2 from the LPO milieu by having catalase-like activity.
We have shown that the high off-rate observed for the O 2 complex with LPO may be a key feature that drives decomposition of the enzyme-O 2 complex and promotes generation of ligand-free LPO-Fe(II) (k off ϭ 11 s Ϫ1 ). In a similar action, at higher levels of H 2 O 2 , LPO-Fe(II) can bind H 2 O 2 generating Compound II, thereby closing the catalase-like cycle (18). The removal of LPO-Fe(II) from the equilibrium mixture of Fe-O 2 , Fe(II), and O 2 causes the reaction to shift from the bound to unbound form enhancing the overall rates of catalysis. Alternatively, Compound III may decay to ground state (k on ϭ 0.5 s Ϫ1 ) through another oxygenated intermediate that can be characterized by its absorption peak at 416 nm and again engage in the LPO catalytic cycle. The value is ϳ20-fold smaller than the rate constant for the dissociation rate constant of LPO-Fe(II)-O 2 . In an air-saturated solution, formation of the LPO-Fe(II)-O 2 complex occurred at rates that were 40 -100ϫ faster than complex decay. Under such circumstances, practically all the LPO-Fe(II) sample exists in its Fe(II)-O 2 form prior to decay. Therefore, accumulation of the LPO-Fe(II)-O 2 during steady state catalysis of LPO requires continuous production of LPO-Fe(II)-O 2 complex, which can be achieved in the presence of a high H 2 O 2 concentration. Under these circumstances, the degree of LPO-Fe(II)-O 2 accumulation and stability depends in part on the H 2 O 2 concentration used and the rate of the conversion of Compound II to Compound III, FIG. 4. Subsequent decay of the LPO-O 2 complex consists of two independent steps. Spectral changes were monitored at 460 nm and best fit to two exponential functions. A plot of each of the observed decay rates of LPO-Fe(II)-O 2 , when the stable (E) and unstable (q) forms of LPO-Fe(II) were used, is shown as a function of O 2 concentration. The two lines present in the figure indicate that the complex decay under this condition was biphasic, and each step follows an irreversible O 2 -independent mechanism. Experimental conditions were as described in Fig. 3.