Studies on the Chlorinating Activity of Myeloperoxidase*

Two methods were utilized to demonstrate the peroxidation of chloride ion to a free species (HOC1 or Cl& by myeloperoxidase. The peroxidase caused the volatilization of radioactivity from solutions containing hydrogen peroxide and [*Y!l]NaCl, and catalyzed the formation of HOC1 when solutions containing these components were passed through a Millipore filter to which the peroxidase was adsorbed. In this flow system, 90 rg of canine myeloperoxidase generated 80 flM HOC1 in the presence of 200 pM H1O, at a rate corresponding to a turnover of 190 min-‘. Under these conditions, o-tolidine, whose oxidation can be coupled to Cl- peroxidation in free solution, did not accelerate turnover. In contrast to chloroperoxidase and horseradish peroxidase. myeloperoxidase does not utilize chlorite for chlorination reactions. This oxidant inactivates the enzyme. At low pH, chloride ion suppresses the oxidation of myeloperoxidase (to the stable compound II) by both hydrogen peroxide and hypochlorite. Acceptor chlorination is therefore not a rate-controlling reaction in the myeloperoxidase mechanism, and the potential of the functional peroxidase couple is higher than the HOCl/Cl-couple under chlorinating conditions. The product-forming step may be a reverse of compound I formation at the expense of HOCl, rather than the chlorination of Cl-by

From the Papanicolaou Cancer Research Institute, Miami, Florida 33123 Two methods were utilized to demonstrate the peroxidation of chloride ion to a free species (HOC1 or Cl& by myeloperoxidase. The peroxidase caused the volatilization of radioactivity from solutions containing hydrogen peroxide and [*Y!l]NaCl, and catalyzed the formation of HOC1 when solutions containing these components were passed through a Millipore filter to which the peroxidase was adsorbed. In this flow system, 90 rg of canine myeloperoxidase generated 80 flM HOC1 in the presence of 200 pM H1O, at a rate corresponding to a turnover of 190 min-'. Under these conditions, o-tolidine, whose oxidation can be coupled to Cl-peroxidation in free solution, did not accelerate turnover. In contrast to chloroperoxidase and horseradish peroxidase. myeloperoxidase does not utilize chlorite for chlorination reactions. This oxidant inactivates the enzyme. At low pH, chloride ion suppresses the oxidation of myeloperoxidase (to the stable compound II) by both hydrogen peroxide and hypochlorite. Acceptor chlorination is therefore not a rate-controlling reaction in the myeloperoxidase mechanism, and the potential of the functional peroxidase couple is higher than the HOCl/Cl-couple under chlorinating conditions. The product-forming step may be a reverse of compound I formation at the expense of HOCl, rather than the chlorination of Cl-by a chloroperoxidase-like chlorinating intermediate.

The antimicrobial
activity of myeloperoxidase has been related to the chloride peroxidase activity of the enzyme (l-3). Agner first brought attention to the oxidative potential of myeloperoxidase in the presence of Cl-and proposed that the formation of free HOC1 could account for the observed chemical transformations (2). Subsequently Zgliczynski and coworkers have described some of the properties of the myeloperoxidase-H,O&-system and have shown that the decarboxylation of amino acids proceeds via the formation of amino acid chloramines (4,5). In no case, however, has the formation of free HOC1 been demonstrated. In this paper we demonstrate that myeloperoxidase catalyzes the peroxidation of chloride ion to free HOCl, and that the mechanism of chlorination is distinct from that of chloroperoxidase to the extent that a product-forming step is not rate-controlling.
The rate of HOC1 formation can account for the rate of acceptor chlorination.
Since the couple involved in chloride peroxidation must be of higher potential than the HOCl/Cl-couple, HOC1 formation may proceed via a reverse of peroxidase compound I formation at the expense of HOCl. were collected into tubes containing 1.0 ml of ortho tolidine reagent. The time per fraction was checked using a stopwatch, and was found not to vary by more than 5% from run to run. The initial fraction collected was generally ignored, since a surge of effluent could not be avoided.
The concentration of HOC1 in each fraction was calculated from the absorbance at 445 nm and the volume of each fraction (18 dps = 0.32 ml).
Trapping of seCI,-The solution containing myeloperoxidase and chloride ion at pH values between 3.5 and 6.0, the odor of Cl, can be detected within a few seconds. This simple test is negative when applied to horseradish peroxidase and chlorite under conditions where this system catalyzes chlorination.
This observation suggested that a quantitative difference exists between the myeloperoxidase-H,O&l-and horseradish peroxidase-ClO*-(Cl-) systems in their capacities to generate free Cl*. Fig. 1 shows the generation of Cl, by myeloperoxidase as measured by the volatilization of %-from solution ("Experimental Procedures").
The formation of product is dependent upon myeloperoxidase and HIO1. This experimental setup is incapable of measuring the true activity of the system, because the peroxidase is inactivated within a few seconds due to the accumulation of HOC1 in solution, and because it depends on the rate of loss of Cl, from solution in competition with other processes. destruction of HOC1 by reaction with HnOI may be significant under these circumstances. These problems can be partially alleviated by the use of a flow system in which the substrates are pumped over immobilized myeloperoxidase, and the effluent, containing both substrates and products, is rapidly analyzed for HOCl. The rapid analysis of HOC1 prevents a possible reaction with H,OI from continuing after the enzymatic reaction has occurred. A second advantage is that it avoids the exposure of the enzyme to destructive levels of HOCl. Fig. 2 shows the generation of HOC1 (or Cl,) by approximately 96 pg of myeloperoxidase using the flow system. The initial concentration of HOC1 (bars) is 80 PM, in contrast to substrate (H,O,) concentration of 200 PM. By quenching alternate tubes with 0.5 mM taurine (4) in 100 mM Na,HPO,, pH 8.6, and subsequently testing for peroxide, it could be shown that excess peroxide is present in the effluent in the initial stages of the reaction. This, together with the fact that apparent activity declines from the second fraction, indicate that the formation of HOC1 is limited by enzymatic turnover, rather than by substrate concentration.
When myeloperoxidase was added to reaction mixture identical in composition with that present in the tubes following quenching, no color formation was observed. Thus, elution of myeloperoxidase from the membrane does not contribute to the observed yield of product. The odor of Cl, can be detected in the effluent from the system.
The measured activity of myeloperoxidase in this system is low, being only ~100 min-' under the conditions, compared to apparent rates of chlorination and chloride-dependent oxidation of 2006 min-' in free solution. When ortho-tolidine was included in the substrate buffer, however, the same rates of turnover were obtained as in its absence (Fig. 3). In free solution, o-tolidine competes effectively with other chloride acceptors, and under identical conditions is oxidized at the same rate as monochlorodimedone is chlorinated. Thus an apparent chloride acceptor, or HOCl-reactive reagent, does not increase the rate of myeloperoxidase turnover. The low activity observed with Millipore-bound myeloperoxidase is probably *  laboratory has shown that like chloroperoxidase, myeloperoxidase catalyzes a single chlorination of monochlorodimedone to yield dichlorodimedone in high yield. The essential differences between myeloperoxidase and horseradish peroxidase in their respective capacities to utilize chlorite and peroxide in the chlorination of monochlorodimedone is shown in Fig. 5. The data on horseradish peroxidase has been previously published by Hager and co-workers (6) and is included here to underline the differences and to serve as a control. Myeloperoxidase exhibits very low apparent activity at the expense of chlorite (cf. HzO,) and undergoes inactivation, whereas horseradish peroxidase exhibits the converse pattern. Experiments with '?-have shown that at the level of myeloperoxidase utilized in this assay, no chlorine incorporation into dichlorodimedone can be detected. Hence the low extent of loss of ultraviolet absorbance seen in the figure, which does not exceed the chloride-independent rate of destruction (Fig. 4) is not due to chlorination.
Studies on the mechanism of inactivation of myeloperoxidase by ClOz-using substrate enzyme levels have revealed that the heme moiety is rapidly oxidized to an oxyheme (12). horseradish peroxidase and chloroperoxidase have been shown to undergo oxidation at the expense of HrOr and HOC1 (6). However, an enzymatic couple capable of generating HOC1 from Cl-should exhibit resistance to oxidation by the HOCl/ Cl-couple. At pH 8.6 both H,Oz and HOCl, in 10x excess over heme, generate myeloperoxidase compound II to similar extents (not shown). At pH 4.5, the extent of formation of compound II at the expense of peroxide is slightly increased, but decreased at the expense of HOCl. When the same determination is made in the presence of 50 mM Cl-, the difference spectra obtained are not significantly different from the prerecorded base-lines (Fig. 6). Under these conditions, peroxide is decomposed with the formation of HOCl, as shown above. Following the addition of HOCl, the solution remains positive for an oxidizing agent (o-tolidine test) for several minutes. DISCUSSION Unlike the protoheme peroxidase chloroperoxidase, myeloperoxidase (a) can continuously peroxidase Cl-to HOCl, (5) does not require organic chloride acceptors for turnover or does not undergo rate acceleration in the presence of the latter, (c) does not utilize chlorite for chlorination reactions, and (d) is not oxidized by HOC1 to a stable oxidation state (under chlorinating conditions). There are two ways in which nonacceleration of turnover can be rationalized.
The first of these is simply that organic acceptors do not interact with a chlorinating form of myeloperoxidase, either because such a form is inaccessible, or because it does not exist as an intermediate.
Under this alternative, a free species (HOC1 or Cl,) would be an obligatory intermediate in acceptor chlorination.
The second alternative is that a rate-determining step precedes the decomposition of a chlorinating intermediate.
If, for example, our experimental setup were measuring a 95% rate-determining step, then the direct interaction of organic acceptors with a chlorinating intermediate could occur (at any rate), but turnover would be accelerated 5% at most. The data rule out a chlorinating intermediate whose decomposition is rate-determining. A previous report from this laboratory has ruled out myeloperoxidase compound II as an intermediate in the catalytic cycle (13). and recent studies have confirmed the existence of an unstable primary peroxide compound (compound I). Two possible mechanisms for chloride peroxidation suggest themselves. The first of these is that generally accepted for halide peroxidations, in which a halogenating intermediate reacts with a second halide ion, to form free halogen (14). The second alternative is that oxygen transfer from compound I to bound hydrogen chloride (H+, Cl-) occurs. This scheme requires a reaction similar to the reverse of compound I formation at the expense of peroxides (15), peroxybenzoic acids (16,17), and HOC1 itself. The latter reaction has been demonstrated in horseradish peroxidase (6) and presumably occurs in myeloperoxidase itself with the Cl-free form (Fig. 6).
Preliminary experiments utilizing the flow system have shown that the effluent is capable of killing Escherichia coli. It is therefore probable that the antibacterial activity of myeloperoxidase described by Agner (1) and Klebanoff (3) and the cytotoxic effect on normal and tumor cells (18) is due to the generation of HOCl. The involvement of singlet Or in killing (19) cannot yet be ascertained, but our data appear to indicate that the reaction between HOC1 and HrOr does not proceed at a significant rate compared with the rate of chloride peroxidation.