Regulation of neutrophil superoxide by antichymotrypsin-chymotrypsin complexes.

The ability of neutrophils to generate free radicals is a crucial component of host defense (Babior, B. M. (1978) N. Engl. J. Med. 298, 659-668, 721-725. Neutrophil oxidants, however, can cause significant host tissue destruction (Weiss, S. J. (1989) N. Engl. J. Med. 320, 365-376), and the regulation of free radical production is not well understood. We have previously shown that recombinant antichymotrypsin (rACT), a serine protease inhibitor, inhibits superoxide production in intact neutrophils (Kilpatrick, L., Johnson, J. L., Nickbarg, E. B., Wang, Z., Clifford, T. F., Banach, M., Cooperman, B. S., Douglas, S. D., and Rubin, H. (1991) J. Immunol. 146, 2388-2393). Using a cell-free NADPH oxidase preparation, we now demonstrate that rACT alone has no effect on superoxide production and that antichymotrypsin-chymotrypsin (rACT.CT) complexes are required to inhibit superoxide, suggesting that neutrophil chymotrypsin-like proteases produce conformational changes in ACT, allowing it to become active in regulating superoxide production. Additionally, we have identified NADPH oxidase itself as the target for rACT.CT and have demonstrated that rACT.CT interferes specifically with activation of the NADPH oxidase without changing the Km for NADPH or the rate constant describing the rate-limiting step in activation. These observations suggest an important role for antichymotrypsin in the regulation of NADPH-oxidase activation, which is a prerequisite for neutrophil superoxide production, and predict possible therapeutic uses for rACT in conditions where unregulated neutrophil-free radical production has been implicated in the mechanism of tissue destruction.

The ability of neutrophils to generate free radicals is a crucial component of host defense ( (1991) J. Immunol. 146, 2388-2393). Using a cellfree NADPH oxidase preparation, we now demonstrate that rACT alone has no effect on superoxide production and that antichymotrypsin-chymotrypsin (rACT-CT) complexes are required to inhibit superoxide, suggesting that neutrophil chymotrypsin-like proteases produce conformational changes in ACT, allowing it to become active in regulating superoxide production. Additionally, we have identified NADPH oxidase itself as the target for rACT*CT and have demonstrated that rACT*CT interferes specifically with activation of the NADPH oxidase without changing the K,,, for NADPH or the rate constant describing the rate-limiting step in activation. These observations suggest an important role for antichymotrypsin in the regulation of NADPHoxidase activation, which is a prerequisite for neutrophil superoxide production, and predict possible therapeutic uses for rACT in conditions where unregulated neutrophil-free radical production has been implicated in the mechanism of tissue destruction.
During the acute phase response to infection and tissue injury, stimulated neutrophils undergo a respiratory burst, producing superoxide radicals via NADPH oxidase. Upon stimulation, the normally dormant NADPH oxidase is assembled from cytosolic and membrane-bound components. Superoxide free radicals, as well as H202, OH., and HOC1 formed in subsequent reactions are important in the defense against microrganisms (1-2) but can also cause significant damage to * This work was supported by H & Q Life Sciences, the Garchik Family Fund, and the United States Department of the Navy. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. host tissue and have been implicated in disorders such as respiratory distress syndrome, myocardial reperfusion damage and rheumatoid arthritis (3-5). Naturally occurring antioxidants such as superoxide dismutase and vitamin E play a role once free radicals have been produced, but little is known about the control of the actual production of free radicals during inflammation.
In addition to the generation of free radicals, stimulated neutrophils release powerful proteases such as elastase and cathepsin G, a chymotrypsin-like enzyme. The serine proteinase inhibitors antichymotrypsin and cul-proteinase inhibitor increase dramatically during the acute phase response (7) and are considered to play an important role in mediating the inflammatory response by neutralizing the effects of cathepsin G and elastase respectively. Additional biological functions have been attributed specifically to the complex of the serpin with its target enzyme, and include chemoattractant activity (8) and recognition by membrane receptors (9). Recently, we have shown that in addition to its well known role in inactivating cathepsin G, antichymotrypsin can inhibit the formation of superoxide by stimulated neutrophils (6).
In the present study, we explore the mechanism of inhibition of superoxide formation by examining the effect of ACT and the products of the complexation of ACT with chymotrypsin on the activation and turnover of NADPH oxidase in a cell-free system. To study the effect of antichymotrypsinchymotrypsin complexes on the activation of the respiratory burst oxidase, we employed a model proposed by Babior et al. lanylalanylprolylphenylalanine p-nitroanilide, and sodium arachidonate were obtained from Sigma. Lymphoprep was obtained from Accurate Chemical and Scientific Corp. (Westbury, NY). All other reagants were of reagant grade and commercially available.
Antichymotrypsin-Chymotrypsin Complexes-Recombinant human antichymotrypsin was complexed with bovine chymotrypsin by titration to a 1:l stoichiometry based on hydrolysis of the specific chymotrypsin substrate succinylalanylalanylprolylphenylalanine pnitroanilide measured by change in absorbance at 410 nM. Complex formation was verified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (7).
Isolation of Neutrophik-Neutrophils were isolated from healthy donors by dextran sedimentation followed by centrifugation over Lymphoprep and hypotonic lysis of red blood cells as previously described (12). Neutrophils were >98% viable by trypan blue exclusion criteria. The neutrophil pellet was suspended in 5 mM PIPES, pH 7.3, 1.75 mM MgC12, 0.5 mM ATP, and 0.34 M sucrose.
Isolation of Membrane and Cytosolic Fractions-Membrane and cytosolic fractions were isolated as previously described (13). Briefly, neutrophils were disrupted with three 15 s bursts of 30% power of a Fisher model 300 sonicator. This was followed by centrifugation at 1200 rpm for 10 min to remove unbroken cells and cellular debris. The supernatant was layered over 40% sucrose and centrifuged at 150,000 X g in a Beckman swing-out rotor (SW 55) for 30 min. The top layer containing cytosolic factors was separated from the sucrose layer containing membranes and stored at -70 C. The pellet was discarded.
In preincubation experiments, antichymotrypsin-chymotrypsin complexes were added 10 min prior to activation. In some experiments, antichymotrypsin-chymotrypsin complexes were added after activation by arachidonate, in order to examine the effect of complexes after activation already occurred. 50 pg/ml superoxide dismutase was added to the reference cuvette. The final volume was 750 pl. NADPHdependent superoxide formation was determined at room temperature as the superoxide-inhibitable reduction of cytochrome c measured continuously for 15-30 min at 550 nM. The change in absorbance was converted to nanomoles of Oz, as previously described (12). In each cell-free preparation, 1 mg of particulate protein represents approximately 4 X 1 0 ' cell equivalents. No superoxide production was seen when either membrane or cytosolic fractions were omitted.
Determination of K,,, for NADPH, VE, and k-K, for NADPH was determined by Lineweaver-Burk analysis. Values for k and VE, were determined by computer-fitting curves of ASb0 uersus time to the equation: A ( t ) = VE, l)/k] + t) + CY (lo), where A(t) is the A660 measured at time t, Vis the rate constant for superoxide production by active NADPH oxidase at saturating NADPH concentration, Et is proportional to the total amount of NADPH oxidase in the reaction mixture, and CY is a figure that enables the computer to adjust for uncertainty in A6s0 at time t = 0. Computer fitting was achieved using a general nonlinear function minimization routine with VEt, k, and a as adjustable variables (lo). In all cases, an excellent fit of the equation to the curves was achieved.

Effect of rACT and rACT. CT on Superoxide Production-
T h e addition of up to 12 PM rACT to the cell-free assay had no inhibitory effect irrespective of when it was added. We postulated that in the intact neutrophil, ACT can complex with chymotrypsin-like enzymes released during stimulation, and the complexes might be responsible for inhibition of superoxide production. Such chymotrypsin-like enzymes are essentially absent from our cell-free preparations, which contain negligible amounts of elastase or CT activity (<0.4 and t0.002 prnollpg of membrane or cytosolic protein, respectively). The addition of 2-12 PM rACT. CT complex inhibited superoxide production in a concentration-dependent manner when added prior to activiation by arachidonate (Fig. 1, A   and B ) .
Effect of rACT. CT Added after Activation of NADPH Oxidase-To separate the effect of rACT on activation uersus turnover of NADPH oxidase, rACT.CT was added after activation had occurred. Complete activation of the NADPH oxidase was identified as the linear rate of superoxide production following a lag during which superoxide formation gradually accelerates. When rACT.CT complex was added after activation had already occurred (turnover phase), no effect on superoxide formation was seen (Fig. 1C).
Control Experiments-The results of several experiments indicate that the observed inhibition is due to a specific effect' of the rACT. CT complex on NADPH-dependent oxidase activity. First, rACT.CT complexes, formed by titration of CT to a 1:1 stoichiometry with rACT, were shown to have no measurable CT activity (based on the hydrolysis of the specific CT substrate N-succinylalanylalanylprolylphenylalanine pnitroanilide) before or after addition to the cell-free assay.
Although CT enhances superoxide production in the intact neutrophil (15), we observe that CT alone can inactivate the NADPH oxidase in the cell-free system. The assay was sufficiently sensitive to detect C T levels below those inhibiting superoxide production. Second, a 30-kDa cutoff filtrate of rACT. CT (obtained using a Centricon 30 microconcentrator) had no effect on superoxide production, excluding lower molecular weight species, including CT, as the agent responsible for inhibition. Third, the degree of inhibition by rACT-CT was not decreased on addition of excess arachidonic acid (Fig.  l D ) , ruling out simple sequestration of the activator as the mechanism of inhibition. Additionally, our assays typically contain 10 FM GTPyS, which causes a 2-fold increase in superoxide production (16). However, the percent of NADPHoxidase inhibition resulting from the addition of rACT. CT is the same in the presence or absence of GTPyS (data not shown).
Effect of rACT on K , for NADPH, VE, and k-The K , for NADPH was unaffected by the presence of rACT. VEt and k, the first-order rate constant for the rate-limiting step of the activation reaction were determined in the presence and absence of rACT (Table I). It is clear that the inhibitory effect of rACT. CT is principally on lowering VEL, rather than on k or on the K , for NADPH.
Preincubation Experiments-Preincubation of rACT. CT with the cytosolic fraction yielded significantly less inhibition of superoxide generation than that following preincubation of rACT.CT with membrane component alone or membrane together with cytosol (Table II), suggesting that the target for rACT. CT is on the membrane. DISCUSSION T h e results presented above support three important conclusions. First, an ACT-protease complex, rather than ACT alone, is responsible for inhibition of superoxide production.
We have shown this directly for the cell-free NADPH oxidase system. By extension, we suggest that in the intact neutrophil, ACT can complex with CT-like enzymes released during stimulation, and it is these complexes that are responsible for inhibition of superoxide production. Second, inhibition of superoxide production in the cell-free system indicates that the target for rACT.CT is the NADPH-oxidase itself, and  ' NO significant difference in k in the presence or absence of rACT. CT).
CT. In the intact neutrophil, superoxide generation could be inhibited by ACT even after activation by met-Leu-Phe (6). In the cell-free system, however, rACT. CT had no effect when added after full activation of the NADPH-oxidase by arachidonic acid (Fig. IC). These results are analogous to those seen with N-ethylmaleimide, which inhibited continued production of superoxide in the whole neutrophil but not in the preactivated cell-free system (17). The N-ethylmaleimide results have been interpreted as indicating the existence of a n intrinsic, rapid deactivation step and continuously replenished pool of active oxidase that are present in intact neutrophils but not in the cell-free system (17). Consequently, effects measured on intact neutrophils cannot distinguish between activation and turnover of the NADPH-oxidase. The observation that rACT. CT does not abort superoxide production in the preactivated cell-free system confirms that the complex is not a free radical scavenger and demonstrates that rACT. CT specifically inhibits activation of the NADPH-oxidase and not its turnover.
The studies reported here, along with earlier work on the intact neutrophil (6) indicate a potentially important biological role for ACT in regulation of the inflammatory response.
ACT normally circulates at approximately 4 PM, but shortly after an inflammatory stimulus the concentration can increase by a factor of 5 (7). During inflammation, there is a disruption of the dynamic equilibrium between proteases and preincubation with cytosol + membrane.
duction. Curves represent typical tracings of results repeated at least 10 times. Arrows indicate time of rACT.CT addition. Arachidonic acid was added at zero time in both experiments. D, dependence of the maximal rate of superoxide production on arachidonic acid concentration in the presence (open circles, n = 3) and absence (closed circles, n = 4) of 8 p~ rACT.CT (f standard error).