Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Proteinase 3 (Pr3) is an abundant serine protease stored in the azurophilic granules of neutrophils together with human leukocyte elastase (HLE) and cathepsin G (1). This 29-kD glycoprotein, alternatively named myeloblastin, AGP7, and p29b (2), is the main target recognized by the antineutrophil cytoplasmic antibodies from patients with Wegener's granulomatosis (5). The amino acid sequence and crystal structure of Pr3 are highly homologous with those of HLE (6). Like HLE, Pr3 is able to degrade a number of extracellular matrix proteins: elastin, type IV collagen, fibronectin, vitronectin, and laminin (7). Intratracheal instillation of Pr3 induced emphysemalike injury in the lungs of hamsters (8), suggesting that Pr3 may play a role in neutrophil-mediated tissue destruction in inflammatory pulmonary disorders. Whether an imbalance between Pr3 and its endogenous inhibitors is involved in the pathogenesis of emphysema and other diseases remains to be clarified.

Several endogenous proteins were reported to inhibit Pr3, including ₁-proteinase inhibitor (₁-PI), ₂-macroglobulin, the 6-kD inhibitory domain of elafin, and monocyte/neutrophil elastase inhibitor (7). Elafin was originally isolated from the scales of patients with psoriasis (12) and is also present in lung secretions at concentrations estimated around 10⁶ M (13). Elafin inhibits HLE, Pr3, and porcine pancreatic elastase (10, 12). The inhibition of HLE by elafin has already been evaluated in detail (14). The inhibition of Pr3 by elafin was previously only studied with an insoluble substrate, elastin (10). In this report, kinetics of the inhibition by elafin of Pr3 in solution and in a form associated with phospholipid bilayer vesicles have been studied. The results have implications for estimating the contribution of elafin to the regulation of proteolytic activity of Pr3 in the lungs.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Purified Pr3, HLE, and ₁-PI were purchased from Athens Research and Technology (Athens, GA). Chemically synthesized 6-kD elafin was from Peptides International (Louisville, KY). The active site concentration of HLE was determined by titration with N-benzyloxycarbonyl-Ala-Ala-Pro-azaAla-ONp (Enzyme System Products, Livermore, CA). The active site concentrations of ₁-PI and elafin were measured with titrated HLE, assuming that one mole of ₁-PI or elafin completely inactivates one mole of HLE. The active site concentration of Pr3 was measured with ₁-PI, also assuming a one-to-one inhibition. Synthetic peptide substrates were from the following sources: MeO-Suc-Lys(pic)-Ala-Pro-Val-pNA and Boc-Ala-Pro-Nva-SBzl(Cl), Bachem (Philadelphia, PA); Boc-Ala-ONp and MeO-Suc-Ala-Ala-Pro-Val-pNA, Sigma (St. Louis, MO); and pyroglutamyl-Pro-Val-pNA (S-2484), Chromogenix (Franklin, OH). Dimyristoyl phosphatidylcholine (DMPC) and sodium dimyristoyl phosphatidylglycerol (DMPG) were purchased from Sigma and used without further purification.

Preparation of Phospholipid-Bound Pr3

Single bilayer liposomes composed of DMPC and DMPG in a molar ratio of 1:1 were prepared by a solvent injection method (15). DMPC and DMPG were dissolved in methanol at a concentration of 12 µmol/ml each. A fluorescent tracer, monomyristoyl phosphatidylcholine with an acyl chain containing a BODIPY fluorophore, -BODIPY 581/591 C₅-HPC from Molecular Probes (Eugene, OR), was added to the solution at a molar ratio of tracer: phospholipids, 1:800. The lipid solution, 1.5 ml, was warmed to 37°C and rapidly (~ 1 ml/s) injected via a 22-gauge needle into 18.5 ml of Dulbecco's modified phosphate-buffered saline (DPBS) containing 0.02% (wt/vol) NaN₃, pre-warmed to 37°C, and rapidly stirred. The resulting suspension was dialyzed against the same buffer at room temperature to remove methanol. Changes of phospholipid concentration during the operation were monitored by the absorption of -BODIPY 581/591 C₅-HPC at 582 nm. In order to allow Pr3 to associate with the phospholipid vesicles, Pr3 and the liposome preparation were mixed in DPBS to final concentrations of 4 and 500 µM, respectively (throughout this report, the concentration of liposomes is expressed as the total concentration of phospholipids). The mixture was repeatedly warmed and cooled between 37°C and 4°C for four cycles (16). The active site concentration of this preparation, which we hereafter refer to as phospholipid-bound Pr3, was assayed with ₁-PI. The preparation was stored at 4°C until used within 1 wk.

Kinetic Analysis for the Inhibition of Pr3 in Various Media

Kinetics of the inhibition of Pr3 by elafin were elucidated by the progress curve method under pseudo-first-order conditions, i.e., the initial concentration of inhibitor was at least ten times greater than that of the enzyme, [I]₀ > 10 [E]₀. The assays were performed in DPBS, Ph 7.2, ionic strength 0.15, containing 0.1% (wt/ vol) Triton X-100 and 10% (vol/vol) dimethyl sulfoxide (DMSO) at 25°C with Boc-Ala-Pro-Nva-SBzl(Cl) as the substrate (17). For a typical assay, 10 µl of Pr3 were added to 990 µl of a pre-equilibrated solution containing elafin, the substrate, and 5,5'-dithio- bis-2-nitrobenzoic acid (DTNB), to initiate the reaction. The reaction of the p-chlorothiobenzyl group released from the substrate by Pr3 with DTNB was monitored at 410 nm ( _410nm = 13,600 M¹ cm¹) for 25 min. The data collected were digitized and fit to an integrated rate equation, equation 1 (18), by the nonlinear regression data analysis programs, Enzffitter (Elsevier, Cambridge, UK) or SigmaPlot 4.0 (SPSS, Chicago, IL).

(1)

where [P]₀ and [P] are the product concentrations at time zero and time t, respectively, and v₀ and v_s are the substrate hydrolysis velocities at time zero and at steady state, respectively. K is an apparent first-order rate constant for the decline of v₀ to v_s. By analyzing the relationship between K and inhibitor concentration, the reaction mechanism and reaction rate constants were determined.

The reaction rate constants were also determined by the progress curve method under pseudo-second-order conditions in which [I]₀ > [E]₀ but [I]₀ < 5[E]₀. In these sets of analyses, MeO-Suc-Lys(pic)- Ala-Pro-Val-pNA (19) and Boc-Ala-ONp (7) were used as the substrates. With the p-nitroanilide substrate, the medium used was DPBS containing 10% DMSO, with or without 0.1% Triton X-100, and amidolysis was monitored at 405 nm ( _405nm = 9950 M¹ cm¹) for 10 min. With the ester substrate, the medium used was DPBS containing 1% (vol/vol) methanol, and esterolysis was monitored at 347 nm ( _347nm = 5,500 M¹ cm¹) for 10 min. All of the measurements were performed in cuvets made of polystyrene, except for the esterolysis measurements, which were performed in cuvets made of methacrylate (Fisher, Springfield, NJ). The data collected were fit to an integrated rate equation, equation 2 (20),

(2)

where  is the autohydrolysis rate of the substrate. The term t in the equation is of importance only for the substrate, Boc-Ala-ONp, because its autohydrolysis is not negligible.  is a variable for fitting the data, as defined by equation 3, 

(3)

The fitting procedure yielded a set of curve parameters, v₀, v_s, K, and [P]₀, from which the association rate constant K_a and the dissociation rate constant K_d were calculated according to equation 4 and 5, respectively (20).

(4)

(5)

where [S] is the concentration of substrate, and K_m is the Michaelis constant.

Kinetic constants for the inhibition of HLE by elafin were also assayed with the progress curve method under pseudo-second-order conditions. The substrate used was MeO-Suc-Ala-Ala-Pro-pNA and the medium was DPBS containing 10% DMSO, with or without 0.1% Triton X-100.

Kinetic Analysis for the Inhibition of Phospholipid-Bound Pr3

The inhibition of phospholipid-bound Pr3 by elafin was studied by the progress curve method under pseudo-second-order conditions. Into 970 µl of DPBS at 25°C, 10 µl of 60 mM Boc-Ala-ONp in methanol and 10 µl of elafin were sequentially added and mixed well. To initiate the reaction, 10 µl of phospholipid-bound Pr3 were added and esterolysis was monitored at 347 nm for 10 min. Controls were run with the same concentrations of substrate (600 µM) and liposomes (5 µM) but without Pr3 to determine the autohydrolysis rate of the substrate. The data recorded were fit to equation 2 and analyzed as described previously.

Additional details of the kinetic analyses and other experiments not described in this section are presented subsequently in the text, table, and the legends for figures. All of the results obtained are expressed as mean ± standard deviation, based on three or more measurements.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Mode of the Inhibition

The digitized data for hydrolysis of the substrate Boc-Ala-Pro-Nva-SBzl(Cl) by Pr3 in the absence and presence of elafin under pseudo-first-order conditions, [I]₀ > 10[E]₀, are shown in Figure 1. In the presence of elafin, the substrate hydrolysis velocity exponentially approached a steady state over a time course of minutes, indicating that elafin is a slow-binding inhibitor of Pr3. To identify the mode of inhibition, the steady-state velocities were measured from the digitized data and plotted according to the methods of Dixon (21) and Cornish-Bowden (22) (Figure 2). The Dixon plot demonstrates that the inhibition is neither noncompetitive nor uncompetitive. The Cornish-Bowden plot further rules out mixed-type inhibition, confirming that the mode of inhibition is simple competitive. The linearity of the Dixon plot indicates that a single elafin molecule binds a single Pr3 molecule; if one Pr3 molecule bound more than one elafin molecule, the Dixon plot would be curved upwards (23).

View larger version (18K): [in this window] [in a new window]   Figure 1.   Progress curves for inhibition of Pr3 by elafin. The reactions were performed at 25°C in DPBS containing 0.1% Triton X-100 and 10% DMSO with Boc-Ala-Pro-Nva-SBzl(Cl) (250 µM) as the substrate and detection of the free thiol product with DTNB (250 µM). The concentration of Pr3 was 0.5 nM. The concentrations of elafin were 0 (straight line), 30 (upper curve), and 45 nM (lower curve). Points, digitized data; curves, theoretical curves computed by fitting the digitized data to equation 1.

View larger version (16K): [in this window] [in a new window]   Figure 2.   Dixon (A) and Cornish-Bowden plots (B) of the inhibition of Pr3 by elafin. The steady-state velocities of substrate hydrolysis, vs, were measured from the digitized data as shown in Figure 1. The concentrations of the substrate Boc-Ala-Pro-Nva-SBzl(Cl) were 150 (open circles) and 250 µM (solid circles).

A Minimum Mechanism for the Inhibitory Reaction

The continuous curves in Figure 1 were computed by fitting the digitized data to equation 1. Cha (18) has shown that equation 1 may be used to explore four different reaction mechanisms. Stojan (24) extended the applicability of equation 1 to eight mechanisms, including three of Cha's mechanisms. Since the mode of inhibition has already been demonstrated to be simple competitive (Figure 2), mechanisms in which an inhibitor-enzyme-substrate triple complex is involved can be ruled out. We examine only the competitive mechanisms in schemes 1 to 3 (for purposes of simplification, the contribution of the reaction with substrate in each of the schemes has been omitted).

(s1)

(s2)

(s3)

In scheme 1, the enzyme E and inhibitor I directly form a reversible complex EI. In scheme 2, E and I first rapidly form a "loose" complex EI, which slowly converts to a tighter, reversible complex EI*. In scheme 3, E maintains a slow equilibrium with an isomer E', but only the isomer can bind the inhibitor (25). The three mechanisms may be distinguished by the relationship between k obtained from the progress curve fits versus inhibitor concentration [I], as expressed by equations 6, 7, and 8, respectively (18, 25).

(6)

(7)

(8)

The plot of K versus [I] is a straight line only for the mechanism of scheme 1. Figure 3 shows that for the inhibition of Pr3 by elafin, the plot of K versus [I] maintains good linearity. This result suggests that the inhibition of Pr3 by elafin is best described by the one-step mechanism in scheme 1. We will discuss this mechanism later and will show that this mechanism might be only valid for low concentrations of elafin. From the plot of K versus [I], the association rate constant K_a and the dissociation rate constant K_d were measured to be 6.0 × 10⁵ M¹ s¹ and 1.6 × 10³ s¹, respectively (Table 1).

View larger version (13K): [in this window] [in a new window]   Figure 3.   Plot of K versus [I] for the inhibition of Pr3 by elafin. The apparent first-order rate constant K was computed by fitting the substrate hydrolysis data to equation 1 as described in Figure 1.

Reversibility of the Inhibition

The previous analysis is based on an assumption that the reaction between Pr3 and elafin is fully reversible. The kinetic behavior shown by the Dixon and Cornish-Bowden plots in Figure 2 and the quality of fit of the digitized data to equation 1 in Figure 1 support this hypothesis. However, because the reactions were monitored only for a short period (25 min), it cannot be excluded that a small fraction of the Pr3-elafin complex might not be dissociable and that during the reaction, a portion of elafin molecules might be inactivated due to cleavage by Pr3. If these events occurred over time, the reaction cannot be considered to be completely reversible. To examine the extent of true reversibility, the following experiments were performed. Elafin and Pr3 in a molar ratio of 1:1.5 were incubated at 25°C. A control solution with elafin at the same concentration but without Pr3 was incubated at 4°C. After an extended incubation for 24 h, inhibitory activities of both solutions were measured against a single HLE preparation. The experiments used the observation that in the presence of 0.1% Triton X-100 and 10% DMSO, the affinity of HLE for elafin is at least 19 times higher than that of Pr3 for elafin (see the results in subsequent paragraphs and Table 1), so that HLE can almost completely replace Pr3 from a fully dissociable Pr3-elafin complex. However, if a fraction of the Pr3-elafin complex was not dissociable, or if a portion of elafin was inactivated during the 24-h incubation, more HLE activity would be inhibited by the control solution. HLE activity in the experiments was measured with S-2484 as the substrate, against which Pr3 has no detectable activity. The results from one of two such experiments are illustrated in Figure 4. The figure shows that the extents of inhibition of HLE by elafin in the two solutions are nearly identical, regardless of whether the inhibitor had been pre-incubated with Pr3. The results demonstrate that the Pr3-elafin complex is fully dissociable and that the inhibitor released from the complex is fully active.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Inhibition of HLE by elafin and a pre-incubated mixture of elafin and Pr3. Elafin (0.2 µM) and Pr3 (0.3 µM) were incubated at 25°C for 24 h. A control solution of 0.2 µM elafin was stored at 4°C, also for 24 h. The inhibitory activities of both solutions against HLE (18 nM) in a total volume of 1 ml were measured with the HLE-specific substrate S-2484 (562 µM). The reaction medium used was DPBS containing 0.1% Triton X-100. Open circles, elafin solution; solid circles, elafin/Pr3 mixture. The straight line is the best linear fit to the data points obtained with the solution of elafin alone.

Effects of Additives on the Reaction Rates

All the assays with substrate Boc-Ala-Pro-Nva-SBzl(Cl) were carried out in DPBS containing 10% DMSO and 0.1% Triton X-100. High concentrations of DMSO are required for solubilizing the substrate, whereas the detergent is added to stabilize the enzyme. Pr3 rapidly loses activity when handled in glass vessels (19). In polystyrene or methacrylate cuvets, the rate at which Pr3 loses activity is decreased but is still measurable in the absence of Triton X-100. It is assumed that Pr3 tends to adhere to the surface of vessels, leading to the loss of activity, and that Triton X-100 blocks the adhesion. Boc-Ala-Pro-Nva-SBzl(Cl) is a sensitive substrate for Pr3 (K_cat/K_m = 6.5 × 10⁵ M¹ s¹). In kinetic assays with this substrate, Pr3 concentrations used were on the level of 10¹⁰ M. Further loss of small amounts of Pr3 would introduce large experimental errors, and consequently, all kinetic assays employing the thiobenzyl ester substrate were carried out in the presence of Triton X-100. To examine the effects of Triton X-100 and DMSO on the kinetics of inhibition of Pr3 by elafin, two other substrates, which allowed the kinetic assays to be performed at higher concentrations of Pr3 and lower concentrations of organic solvents, were employed. For both substrates, the assays were run under pseudo-second-order conditions, [I]₀ > [E]₀, but [I]₀ < 5[E]₀, with an alternative progress curve method. The integrated rate equation used was equation 2, which was derived from the one-step mechanism in scheme 1 (20). We first confirmed that the two progress curve methods produced the same results in the same medium, DPBS containing 10% DMSO and 0.1% Triton X-100. With the substrate MeO-Suc-Lys(pic)- Ala-Pro-Val-pNA under pseudo-second-order conditions, K_a and K_d were determined to be 7.7 × 10⁵ M¹ s¹ and 1.5 × 10³ s¹, as compared with 6.0 × 10⁵ M¹ s¹ and 1.6 × 10³ s¹, respectively, obtained with the substrate Boc-Ala-Pro-Nva-SBzl(Cl) under pseudo-first-order conditions (Table 1). The differences between the two sets of results are all within the range of experimental error. We next measured the kinetic constants in DPBS, which contained only 10% DMSO. In the absence of Triton X-100, K_a was determined to be 2.7 × 10⁶ M¹ s¹, threefold greater than that in the presence of Triton X-100. However, K_d was also increased, by a factor of approximately two, making the inhibition constant, K_i, almost unchanged (Table 1). The critical micellar concentration of Triton X-100 is 0.24 mM (26), or 0.015% (wt/vol). At the Triton X-100 concentration used in this study, 0.1%, most of the detergent molecules should aggregate to form micelles. Interaction between the micelles and Pr3 might be responsible for the lower association rate of Pr3 and elafin. The detailed mechanism for the effect of Triton on the inhibition kinetics, however, remains unknown. Finally, we measured the kinetic constants with the substrate Boc-Ala-ONp in medium that contained neither Triton X-100 nor DMSO, but 1% methanol. The association rate constant was further increased to 4 × 10⁶ M¹ s¹, whereas the dissociation rate constant was the same as that in the presence of Triton X-100. These data clearly demonstrate that both Triton X-100 and DMSO significantly affect the reaction between Pr3 and elafin, especially the association rate.

Previously we reported the reaction kinetics for HLE and elafin in buffers other than DPBS but also containing Triton X-100 and DMSO (14). For the purpose of comparison, we redetermined the reaction rate constants in DPBS containing 10% DMSO, with and without 0.1% Triton. The results are summarized in Table 1. In DPBS containing 10% DMSO and 0.1% Triton X-100, the association of HLE and elafin is nearly tenfold faster than that of Pr3 and elafin. Furthermore, HLE and elafin form a complex that is more stable than that formed from Pr3 and elafin. Removal of Triton X-100 from the medium has minor effects on both association and dissociation of HLE with elafin (Table 1). However, in DPBS containing 1% methanol with Boc-Ala-ONp as the substrate, the inhibition of HLE by elafin progressed extremely rapidly. Because the slopes of the progress curves rapidly declined to their steady-state values, we were unable to obtain sufficient data from the pre-steady-state portion of these progress curves to fit to equation 2 for reliable determination of the curve parameters. Only a lower limit for the association rate constant of 2 × 10⁷ M¹ s¹ can be estimated.

Kinetics of Inhibition of Phospholipid-Bound Pr3

A portion of the total neutrophil content of Pr3 has been found to be located on the surface of activated neutrophils and a subset of nonactivated neutrophils (27, 28). Pr3 has also been shown to associate with high affinity to phospholipid bilayer vesicles (16). Using a liposome preparation of DMPC/DMPG, we have studied the effects of membrane binding on the inhibition of Pr3 by elafin. This liposome preparation was selected because its phase transition temperature is not high (23.8°C) (16), so that Pr3 can be incorporated into the membrane under mild conditions. Assays using inhibition by ₁-PI show that after the thermal cycling (37°C/4°C) treatment (see MATERIALS AND METHODS), the apparent concentration of Pr3 active sites underwent no detectible diminution (data not shown). To examine if Pr3 is bound by the liposome membranes, and how tight the binding is, we employed a series of experiments based on the observation that binding by phospholipids markedly raises the esterolytic activity of Pr3. Ten microliters of Pr3, before and after thermal cycling with liposomes, were added into 990 µl of DPBS with 600 µM Boc-Ala-ONp, and esterolysis was monitored at 347 nm for 10 min. Results reported in Figure 5 show that thermal cycling with a liposomal suspension did indeed raise the esterolytic activity of Pr3, indicating that the enzyme had become associated with the liposomal membranes. The progress curve for lipid-bound Pr3 maintained good linearity, indicating that the enzyme remained bound to the phospholipids over the time course of the observations. If a fraction of the protease-lipid complex had dissociated to release free Pr3 within the observation period, the progress trace would have displayed a distinct downward curvature. It might be argued that at the moment when phospholipid-bound Pr3 and DPBS were mixed, the protease-lipid complex could have partially dissociated upon dilution to establish a new rapid equilibrium between free Pr3, free liposomes, and the complex. This possibility has been ruled out by the results from a control experiment also reported in Figure 5. When the medium contained the same concentration of liposomes that were prevented from association with Pr3 by eliminating the thermal cycling step, the rate of esterolysis by the free Pr3 in the liposome suspension was the same as that in the absence of liposomes. The observation that dilution does not induce dissociation of the protease-lipid complex is critical to our kinetic analysis for the inhibition of phospholipid-bound Pr3 by elafin because the method includes a dilution step.

View larger version (14K): [in this window] [in a new window]   Figure 5.   Esterolysis by phospholipid-bound Pr3 and free Pr3. The reactions were performed at 25°C in DPBS containing 1% methanol with Boc-Ala-ONp (600 µM) as the substrate. Top straight line, esterolysis by 5 nM Pr3 bound to phospholipid liposomes after thermal cycling; the total liposome concentration in the reaction solution was 5 µM. Lower straight line, esterolysis by 5 nM free Pr3; no liposomes were added to the reaction solution. Discontinuous points, digitized data from esterolysis by 5 nM free Pr3 in the presence of 5 µM liposomes that had not been exposed to thermal cycling with the enzyme.

The digitized data for hydrolysis of the substrate Boc-Ala-ONp by phospholipid-bound Pr3 in the absence and presence of elafin are shown in Figure 6. The figure also includes the data for inhibition of free Pr3 by elafin. Comparison of the two sets of data clearly demonstrates that association with phospholipid bilayers significantly retards the rate of inhibition of Pr3 by elafin. The continuous curve in the middle of Figure 6 was obtained by fitting the relevant digitized data to equation 2. From the progress curve parameters obtained by the curve fitting procedure described previously, the association and dissociation rate constants for phospholipid-bound Pr3 and elafin were calculated to be 3.3 × 10⁵ M¹ s¹ and 9.1 × 10⁴ s¹, respectively. In comparison to the rate constants determined for free Pr3 and elafin (Table 1), binding with phospholipid bilayers deceases the association rate of Pr3 and elafin by a factor more than one order of magnitude, whereas the phospholipid-bound Pr3 and elafin form a protease-inhibitor complex that is more stable than that formed by free Pr3 and elafin.

View larger version (15K): [in this window] [in a new window]   Figure 6.   Progress curves of the inhibition of phospholipid-bound Pr3 and free Pr3 by elafin. The reactions were performed at 25°C in DPBS containing 1% methanol with Boc-Ala-ONp (600 µM) as the substrate. Upper curve (straight line), esterolysis by 40 nM phospholipid-bound Pr3 in the absence of elafin. Middle curve, esterolysis by 40 nM phospholipid-bound Pr3 in the presence of 128 nM elafin. Bottom curve, esterolysis by 40 nM free Pr3 in the presence of 128 nM elafin.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The kinetic data obtained for the inhibition of Pr3 by elafin are most consistent with a reaction mechanism in which a single elafin molecule binds a single Pr3 molecule to form a fully reversible complex without an intermediate, as formulated by scheme 1. This mechanism is the same as that we previously elucidated for HLE and elafin (14). We have postulated that elafin and HLE could first form a loose complex as described by the mechanism in scheme 2, but that within the limits of our experimental method, the presence of this putative intermediate could not be detected kinetically due to the low concentrations of elafin we have employed (14). This same caveat may also apply to Pr3 and elafin. When (1 + [S]/K_m) >> [I]/K_i, equation 7 simplifies to equation 6, and the two-step mechanism may be approximated by the one-step mechanism with K_a  K₁/K_i. In our experiments, [S] = 250 µM, K_m = 115 µM, and the highest elafin concentration used was 50 nM. As long as K_i  0.16 µM, the condition of (1 + [S]/K_m) >> [I]/K_i is fulfilled, and the progress curve method can no longer distinguish between the one-step and two-step mechanisms. Whether this postulated intermediate is present in the reaction pathway of elafin and its cognate proteases remains to be experimentally clarified. However, allowing for the possibility that the one-step mechanism is only an approximation that is valid at low concentrations of elafin, the fit of the data to this mechanism permitted us to employ the integrated rate equation, equation 2, and to calculate K_i by using the simple relationship, K_i = K_d/K_a, reasonably.

Triton X-100, which serves as a stabilizer for Pr3, and DMSO, which facilitates solubilization of the rather hydrophobic substrates, are frequently incorporated in media for enzymatic analysis of Pr3. The present study shows that both reagents reduce the association rate of Pr3 and elafin. The rate constants measured in DPBS containing 1% methanol are probably closer to their true values in vitro because the medium contained the least additives. This set of kinetic constants has been selected for considering the functional implications of our results. The efficiency of inhibition of a protease by an inhibitor may be quantified by a parameter suggested by Bieth (29), the delay time of inhibition, d(t) = 5/K_a[I]_vivo, where K_a is the association rate constant in vitro and [I]_vivo is the inhibitor concentration in vivo. In simple terms, d(t) is the time required by the inhibitor to nearly completely inactivate the protease. In bronchoalveolar lavage samples collected from normal individuals, the average molar ratio of 6-kD elafin/ ₁-PI was reported to be 0.73 (13). Because the ₁-PI concentration in lung secretions has been reported to range from 2 to 6 µM (30), it may be assumed that the elafin concentration in the same fluids may range from 1.5 to 4.4 µM. Given a value of K_a = 4.0 × 10⁶ M¹ s¹ (Table 1), d(t) would have a calculated value between 0.3 and 0.8 s for inhibition by elafin. A value of d(t) of less than 1 s implies that elafin may inhibit Pr3 efficiently. However, the efficiency of Pr3 inhibition by elafin is much lower than that by the other inhibitor of Pr3, ₁-PI, which is also present in lung secretions. First, the concentration of ₁-PI is higher than that of 6-kD elafin. Second, the association rate constant of ₁-PI and Pr3 is one order of magnitude larger than that of elafin and Pr3 (for the association rate constant for ₁-PI and Pr3, see References 7 and 9). Third, ₁-PI behaves as an irreversible inhibitor, whereas the Pr3-elafin complex is dissociable, with a half-life time of 6.8 min (t_1/2 = 0.693/K_d). Thus, the inhibition of Pr3 by elafin could be transient under certain conditions. For these reasons, if Pr3 were released by activated neutrophils into the extracellular milieu in the normal lungs, most of the protease would eventually be bound by ₁-PI. A small fraction of Pr3 may initially bind elafin but would progressively be transferred to ₁-PI. We therefore deduce that elafin may participate in the regulation of the proteolytic activity of Pr3, but its contribution is limited. It is interesting to compare the situation of 6-kD elafin with that of secretory leukoprotease inhibitor (SLPI), another small protein in the lungs that inhibits HLE and cathepsin G, but not Pr3. Similar to elafin, SLPI is a reversible inhibitor; the association rate of SLPI with HLE is much slower than that of ₁-PI and HLE (31). However, SLPI plays a major role in regulating the proteolytic activity of HLE in upper airways, where the local concentration of SLPI is higher than that of ₁-PI (32). The possibility that 6-kD elafin is also unevenly distributed in the lungs merits further investigation. A further complication in interpreting the physiologic significance of these results arises from the possible contribution of full-length elafin to the overall antiproteinase defenses, especially in the upper airways. In full-length elafin, the 6-kD antiproteinase domain is linked to an amino-terminal domain that serves as a substrate for transglutaminases and may anchor the antiproteinase covalently to the extracellular matrix (33). The kinetics of inhibition by this full-length inhibitor have not been investigated.

Pr3 is unique among the neutrophil serine proteases in that a substantial amount of the protease can be detected on the surface of the plasma membranes of a subset of resting neutrophils (27). The membrane association of Pr3 seems to be genetically determined and has been postulated as a risk factor for vasculitis and rheumatoid arthritis (34). Antineutrophil cytoplasmic antibodies recognize and associate with the membrane-bound protease, inducing the activation of neutrophils (35). Witko-Sarsat and colleagues (28) have studied the binding properties of Pr3 with the neutrophil membrane and concluded that binding was unlikely to involve a ligand/receptor mechanism. Binding through charge-charge interactions was also considered to be unlikely because high salt concentrations or extremes of pH did not release the protease from the membrane. These authors suggested that some form of covalent bond might be involved in the binding. This study shows that Pr3 binds model membranes formed by DMPC/DMPG liposomes tightly. In the experiments reported in Figure 5, the phospholipids were diluted by a factor of 100, and Pr3 was diluted by a factor of 800. However, no dissociation of the protease-lipid complex could be detected. We incorporated Pr3 into the lipid vesicles by a thermal cycling procedure that repeatedly drives the transition of the bilayers between a gel phase and a liquid crystalline phase. This treatment should not result in the formation of any covalent bonds. Binding to the model membranes increases the catalytic activity of Pr3 toward the substrate Boc-Ala-ONp; it also significantly decreases the association and dissociation rates of Pr3 with elafin. A change in dissociation rate constant is an indication that the protease-inhibitor complex formed is still associated with the bilayer vesicles. Otherwise, the complex would be expected to dissociate at a rate that is approximately equivalent to that of the complex formed from free Pr3 and elafin. Preliminary experiments ongoing in this laboratory have found that binding to the same model membranes also delays the association of Pr3 and ₁-PI. We may hypothesize that plasma membrane-bound Pr3 on activated or resting neutrophils performs some physiologically relevant functions under conditions that disfavor rapid inactivation of the membrane-bound protease by its endogenous inhibitors. However, further consideration of the physiologic significance of these kinetic results must await accumulation of more data.