Chlorination of Taurine by Myeloperoxidase KINETIC EVIDENCE FOR AN ENZYME-BOUND INTERMEDIATE*

The chlorination of taurine by the myeloperoxidase- H2O2-Cl- system was investigated under steady state conditions. By systematically varying the pH and the concentrations of H202, C1-, and taurine such that chloride inhibition and the unwanted formation of inactive compound I1 intermediate are minimized, rate data were found to fit a mechanism involving an enzyme- bound chlorinating intermediate. The mechanism we propose is as follows. are: lo6 M - ~ s-l. The rate constant for compound I formation (kl) is of the same order of magnitude as the value (1.8 10’ s-l) obtained using transient state techniques in a previous study by our group. The value of k3 is 2 orders of magnitude greater than the nonenzymatic reaction between HOCl and taurine at the same The results of this study indicate that the chlorination reaction mediated by the myeloperoxidase system in vivo may involve an enzyme intermediate species rather than free HOCl. Not only does this mechanism offer

I formation (kl) is of the same order of magnitude as the value (1.8 x 10' s-l) obtained using transient state techniques in a previous study by our group. The value of k3 is 2 orders of magnitude greater than the nonenzymatic reaction between HOCl and taurine at the same pH.
The results of this study indicate that the chlorination reaction mediated by the myeloperoxidase system in vivo may involve an enzyme intermediate species rather than free HOCl. Not only does this mechanism offer the advantage of substrate specificity but also of speed compared to the non-enzymatic reaction. This mechanism can also explain how the indiscriminate oxidation reactions by HOCl are prevented in the leukocyte. The fast formation of taurine monochloramine, a relatively nontoxic and stable compound compared to HOC1, is consistent with the proposed role of taurine in the neutrophil, that of protecting certain targets including myeloperoxidase from the attack by potent chlorinated oxidants.
Myeloperoxidase (donor: Hz02 oxidoreductase, EC 1.11.1.7 ) is present in high concentrations in the granules of polymorphonuclear leukocytes and monocytes (1). It is released during Research Council of Canada. The costs of publication of this article were * This work was supported by the Natural Sciences and Engineering defrayed in part by the payment of page charges. This article must U.S.C. Section 1734 solely to indicate this fact. therefore be hereby marked "aduertisement" in accordance with 18 $ On study leave from the University of the Philippines at Los Baiios, Laguna, Philippines.
To whom correspondence should be addressed. Tel.:  phagocytosis and catalyzes the oxidation of halide ions in the presence of Hz02 to yield products that oxidize and halogenate microbial components (2). In uiuo, chloride, with concentrations of as high as 0.11 M in the leukocytes (31, is the dominant electron donor, and it has often been assumed that hypochlorous acid (HOC11 is produced as the oxidation product (4-6).
HOCl is known as a potent oxidizing agent. It can directly oxidize a variety of biological molecules including carbohydrates, nucleic acids, peptide linkages, amino acids, and lipids (7-9). Bacteria exposed to HOCl are killed within milliseconds after the addition of the oxidant (10). In vitro studies show that low concentrations (10-20 PM) of HOCl rapidly cause oxidation of plasma membrane sulfhydryls, while slightly higher doses (>50 w) lead to cell lysis (10). While serving a protective role, HOCl and other oxidants derived from the MPO1-HzO2-C1system can also cause deleterious effects to host tissues (11). For example, there have been reports implicating these oxidants to adult respiratory stress syndrome (12) and pulmonary inflammatory injury (13). HOCl reacts with nitrogen-containing compounds such as ammonium ions, amines, and amino acids to form derivatives containing the N-Cl bond. These N-C1 derivatives retain the two oxidizing equivalents of HOCl and therefore can act as as oxidizing agents as well.
Taurine, 2-aminoethanesulfonic acid, is one of the most abundant free amino acids in mammalian tissues. It has been implicated in various functions in the body including as a neurotransmitter, as protector of the photoreceptor of the retina, anti-convulsant agent, anti-oxidant, and alleviator of congestive heart failure. It is now widely used as an additive in infant formula, pet foods, tonics, health agents, and eyedrops (14)(15)(16)(17)(18)(19).
While ubiquitous in animal tissues, the concentration of taurine is particularly high in cells that possess considerable potential for producing oxidants (20)(21)(22). Intracellular taurine concentrations of 26 and 22 mM were reported for human leukocytes (23) and neutrophils (24), respectively. There is an intuitive feeling that a substance present in such high concentration must play a significant role. Taurine appears to act as a trap for HOCl produced by the MPO-HzO2-C1-system of leukocytes forming the long-lived oxidant taurine monochloramine, TauNHC1, which is much less reactive and less toxic than HOCl (25). In fact, taurine and other amines account for 90% of the N-C1 derivatives that accumulate in stimulated neutrophils (24).
Although HOCl is usually assumed to be the chlorinating agent in the MPO-catalyzed oxidation of chloride, one cannot exclude the possibility that an enzyme-bound chlorinating intermediate is formed as has been claimed for chloroperoxidase (26)(27)(28). advantage of this chromophoric product to monitor the chlorination rate of the MPO-H202-C1-system under steady state conditions. This three-substrate reaction was analyzed in terms of the two most plausible mechanisms: one that involves free HOCl as chlorinating agent and another that involves an enzyme-bound chlorinating intermediate. Kinetic parameters were extracted from secondary plots of steady state data and compared with rate constants obtained using transient state techniques and non-enzymatic methods.

MATERIALS AND METHODS
Bovine spleen MPO was isolated and purified using a combination of modified procedures (29-31). Complete details are available (32). The enzyme preparation used in this study exhibited a Reinheitszahl (A43d A,,,) of 0.83. The MPO concentration was determined spectrophotometrically using of 178 m-l cm" (33). Hydrogen peroxide (-30% solution, BDH Chemicals) concentration was determined after appropriate dilution using the horseradish peroxidase assay (34) and was confirmed by absorbance measurements a t 240 nm where E~~~~ is 39.4 M" cm-' (35). Taurine (Sigma), KC1 (Aldrich), and the chemicals used for the phosphate buffers (Fisher) were used without further purification. Aqueous solutions were prepared using deionized water obtained from the Milli Q system (Millipore Corp.), and concentrations were determined by weight. A stock solution of HOCl was prepared by bubbling Cl, through 0.1 M NaOH (Aldrich). The C12 was prepared by dripping concentrated HC1 (Anachem) on reagentgrade MnO, (Matheson, Coleman and Bell) (36). This stock solution was stored in the cold protected from light. Its concentration was determined by measuring the absorbance at 295 nm where eOCl ~ is 0.35 mM-' cm" (37).
Routine spectral measurements were made on a Beckman DU-650 spectrophotometer. Stopped flow measurements were performed using the Photal RA-601 rapid reaction analyzer equipped with a 1-cm obser- Steady state experiments were performed using the stopped flow mode of the Photal instrument. The chlorinating activity of MPO was determined by following the formation of TauNHCl at 252 nm (eTauNHCI 429 M-' cm") (38). One reservoir contained H,O, in water; the other, MPO, KC1 and taurine in buffer. The solutions were buffered at the desired pH to a final ionic strength of 0.1 M. pH measurements were made using a Fisher Accumet model 25 digital pH meter.
The initial rate was determined from the slope of the first portion of the traces obtained after mixing all the components in about 4 ms. Usually 6 8 traces were recorded, and the mean values were used in the plots. A nonlinear regression data analysis program (Enzfitter) was used for curve fittings and calculations. The initial rate ( u ) and substrate concentration [SI were fitted to a typical steady state equation, where [E],, is the total MPO concentration and kc,, and KM are typical Michaelis-Menten parameters. The kc,, parameters were further fitted to a hyperbolic equation as a function of chloride. Steady state rates were also measured at various taurine concentrations.
The nonenzymatic chlorination of taurine by HOCl was studied in the stopped flow apparatus at different pH levels by monitoring the increase in absorbance a t 252 nm as TauNHCl forms. HOCl solutions in one reservoir (final concentration 0.15 m) were prepared immediately before the experiment by diluting the stock solution in buffer of ionic strength 0.1 M. A large excess of taurine (final concentrations of 5-25 m) was placed in the other reservoir. The stopped flow traces were single exponentials from which pseudo-first order rate constants were determined by a nonlinear least squares fit of seven or more traces. The mean kobs values were plotted against taurine concentration, and the apparent second order rate constant of the reaction was calculated from the slopes using linear regression analysis. areducing substrate of MPO compounds I and 11. Rapid scans of the reaction between native enzyme and H202 showed no difference in the absence or presence of taurine. Neither was the rate of formation of compound I1 from compound I affected. Moreover, a wavelength scan indicated that taurine does not hasten the return of compound I1 to native enzyme. These results indicate that taurine reduces neither compound I nor compound 11.

RESULTS
Taurine had insignificant absorption at 252 nm, and it did not affect the weak absorbance due to catalytic amounts of MPO. However, when C1-and HzOz were added to the system, there was an increase in the absorbance at 252 nm, which remained stable after several minutes (Fig. 1). This can be attributed t o the formation of TauNHCl. The formation of dichlorinated product (TauNC12) proceeded subsequently. TauNClz had a weaker absorbance at 252 nm ( E 190 M -~ cm-l) (38) but could nevertheless cause a n interference in the rate measurements. To avoid this, it was necessary to measure the rate of TauNHCl formation from the initial portion of the traces (Fig. 2) before this product accumulated.
The chlorination rate was measured as a function of H202 at different pH levels, and the results are shown in Fig. 3. Above pH 5.2, the rates at the higher H2O2 concentrations were so fast that we could only see the latter stages of the reaction. Inhibition by HzOz becomes significant at about pH 6.6 and higher (data not shown).
The next set of steady state experiments were performed by varying the concentrations of Cl-and Hz02 systematically. We used lower C1-concentrations to minimize inhibition due to competitive binding of chloride to MPO. Moreover, the concentration of HzOz relative to the enzyme was reduced in order to prevent the formation of compound 11, which is inactive in the chlorination reaction. The dependence of the chlorination rates on the concentrations of C1-and H2O2 a t fmed concentration of taurine are shown in Fig. 4. The hyperbolic plots yield two parameters, kcat and K M from Equation 2.
From inspection of Fig. 4, it is apparent that both kcat and K M increase with increasing chloride concentration. Secondary plots of kcat and K M against the chloride concentration were both nonlinear (Figs. 5 and 6). The data in Fig. 5 were fitted to a hyperbolic equation, Nonenzymatic chlorination of taurine was carried out from pH 2.9 to 5.3. The addition of HOCl to a solution of taurine caused the rapid formation of a stable peak at 252 nm indicative of the formation of TauNHCl. Pseudo-first order rate constants from exponential traces were plotted against taurine concentration. The plots intersected very close to the origin and were linear over the concentrations studied (data not shown). The slopes gave values of the apparent second order rate constant, which were then plotted as a function of pH (Fig. 8). An interpolation from Fig. 8  been presumed to be the chlorinating species in the MPOcatalyzed reaction, the formation of an enzyme-bound chlorinating intermediate cannot be ruled out. In this paper we considered the two most plausible mechanisms for MPO-catalyzed chlorination using taurine as substrate. These mechanisms may be represented as follows. Both mechanisms involve the reversible formation of compound I, for which evidence has been presented recently.' The reverse rate constant ( k l ) becomes insignificant when large excess of H202 is used over the enzyme concentration as was the case for the steady state experiments in this work.
The two mechanisms also include the reversible binding of C1-to native enzyme. This binding has been documented in several reports (31, 39-43). In chlorination reactions, chloride not only acts as a substrate for MPO but also behaves as a competitive inhibitor of HzOz. Chloride binding to MPO is pHdependent. When the pH increases, the affinity of C1-for the enzyme decreases. In contrast, the formation of the inactive intermediate compound I1 increases with increasing pH (44). Thus, HzOz also inhibits the chlorination reaction. It is now evident that there is no fmed optimum pH for MPO-catalyzed chlorination. It depends on the relative concentrations of C1and Hz02 (45). Various combinations of pH, C1-, taurine, and Hz02 concentrations were tried so that the conditions were obtained in which the chlorination reaction was the rate-determining step and that inhibition by C1-and Hz02 was at least lication. minimized. Thus, using selected ranges for these three substrates, we were able to measure accurately rates of formation of TauNHC1. The catalytic amount of MPO used in this study did not interfere with the absorbance of TauNHCl (Fig. 1). Moreover, the linear portion of the traces on a short time scale (Fig. 2) allowed us to measure the first chlorination step before enough TauNHCl accumulates to be further chlorinated to TauNCl'.
We then carried out the chlorination reaction at several pH levels to determine which pH would be most suitable (Fig. 3). Our choice of pH 4.7 for subsequent experiments was based on the following reasons. (a) At pH above 5.6, the rate of enzyme inactivation increases (46, 47) and HzOz inhibition becomes important (39) leading to a decrease in chlorinating activity. ( b ) At pH below 4, the inhibitory binding between C1-and native enzyme becomes stronger (31). (c) The pH under physiological conditions in the phagosome has been reported t o fall in the range from 4.5 to 5.0 (48). Moreover, we obtained traces with good signal-to-noise ratio for the runs performed at this pH. Thus, subsequent measurements were conducted at pH 4.7.
The initial rate of chlorination of taurine was dependent upon both HzOz and C1-and showed saturation behavior (Fig.  4). After fitting the experimental data to Equation 1, parameters kc,, and K M were obtained. Both of these parameters showed nonlinear dependence with C1- (Figs. 5 and 6). Of these parameters kc,, has the more straightforward meaning; it is the maximum turnover number for fixed concentrations of chloride and taurine. kcat values were fitted to a rectangular hyperbola and a second set of parameters, kcl-and Kc,-, were obtained.
We have derived the rate equations for Mechanisms I and I1 (see "Appendix"). Mechanism I predicts a linear dependence of kcat and K M on chloride concentration, whereas Mechanism I1 predicts hyperbolic behavior. Thus, our data clearly support Mechanism 11, which invokes the formation of an enzymebound chlorinating intermediate. We can extract kinetic parameters from secondary plots, particularly the plot of kcat uersus C1-concentration (Fig.  5 ) . Using Equation 20A (see "Appendix"), kc,-= k, [taurine] we determined k3 to be (3.6 5 0.3) x lo5 M" s-l. Substituting this value to Equation 21A (see "Appendix"), we calculated k 2 to be (2.8 2 1.2) x lo6 s-l.
Another criterion that can distinguish between Mechanisms I and I1 is the dependence of the initial rate of chlorination on taurine concentration. Mechanism I predicts no dependence, while Mechanism I1 does. Fig. 7 demonstrates a hyperbolic dependence of rate on taurine concentration. A curve fit of the data yields parameters Krcat and K M . The value of kl can be calculated from k',at (Equation 25A under "Appendix") using the value of k 2 previously determined. The value of K for the dissociation of the MPO-chloride complex was taken from an average of previously reported values: 1.2 mM (43), 0.8 mM (interpolated from Fig. 9  Thus, we were able to satisfy three criteria supporting the existence of a MPO-chlorinating intermediate . Such an enzyme-bound species has similarly been proposed for chloroperoxidase (26-28). We were also able to estimate rate constants for the different steps in the MPO mechanism. We show from the magnitudes of the rate constants that the formation of TauNHCl is indeed the rate-controlling step under the steady state conditions employed in this study. We also obtained a rate constant for the reaction between compound I and C1-to form the intermediate that cannot be determined using transient state kinetic techniques due to the instability of compound I.
In the presence of excess taurine, the non-enzymatic reaction between taurine and HOCl shows pseudo-first order kinetic behavior. Apparent second order rate constants were obtained from the slopes of the plot of kObs uersus taurine concentration at various pH levels. An interpolation from the pH rate profile gives a value of 3.5 x lo3 M -~ s-l for the non-enzymatic reaction of taurine with HOCl at pH 4.7. This result indicates that a 100-fold increase in the rate of chlorination is obtained in the enzymatic reaction. Thus, a MPO-bound chlorinating intermediate offers not only the advantage of substrate specificity but also that of speed.
There have been numerous studies on MPO-catalyzed chlorination reactions (5,6,8,39,43,45,49), but it appears that the evaluation of steady state parameters has been complicated by peroxidatic side reactions (43). For example, monochlorodimedon, which is routinely used to measure chlorinating activity of peroxidases, was found to compete with C1-for compound I and promote the accumulation of compound I1 (48). These complications were circumvented when taurine was used as the substrate because it did not exhibit reactivity with either compound I or 11.
The results of our study are physiologically relevant. Taurine concentration is higher than that of other free amino acids in the leukocyte (23). In fact, taurine accounts for a large part of the amines that are available for reaction with chlorinating reagents (25). Chloramines were found to accumulate in the extracellular medium when stimulated leukocytes are incubated in vitro (50). Apparently, taurine acts as a trap for chlorinated oxidants produced by the MPO-H202-C1-system. If HOCl were produced and allowed to diffuse freely, this potent oxidant could cause extensive damage to cellular components (51,52). For example, it has been reported that when the amine concentration is low, HOCl reacts with and inactivates MPO (53). HOCl also reacts with H202 (54), and this results in the loss of oxidizing equivalents. Taurine has recently been found to be effective in protecting biomembranes against oxidant in-jury (55,56). The formation of an enzyme-bound species that chlorinates taurine is consistent with taurine's protective role against oxidant damage, which could otherwise be indiscriminately brought about by free HOC1. While it has been demonstrated that chloramines like TauNHCl can inactivate enzymes by attacking sulfhydryl groups, lyse erythrocytes, and kill bacteria (571, TauNHCl is much less reactive and less toxic than HOCl (25). Therefore, taurine may function as moderator of neutrophil cytotoxicity.

APPENDIX
The derivation of the steady state equations for Mechanisms I and I1 are shown. The abbreviations are defined in Footnote 1.   K , = k',,/k,