the of vol. no. the of printed m u. a. inhibition of the serine … · 2017-08-15 · inhibition of...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1984 by The American Society of Biological Chemists, Inc

Vol. 259, No. 24, Issue of December 25, pp. 15106-15114,1984 Printed m U. S. A.

Inhibition of the Serine Proteases Leukocyte Elastase, Pancreatic Elastase, Cathepsin G, and Chymotrypsin by Peptide Boronic Acids*

(Received for publication, May 21, 1984)

Wilmington, Delaware 19898

Three a-aminoboronic acid-containing analogs of good peptide substrates for serine proteases were syn- thesized, MeO-Suc-Ala-Ala-Pro-boro-Phe-OH, MeO- SUC - Ala - Ala - Pro - boro - Ala - OH, and M e 0 - SUC - Ala - Ala-Pro-boro-Val-OH. They were effective inhibitors of chymotrypsin, cathepsin G, and both leukocyte and pancreatic elastase at nanomolar concentrations (0.10-20 nM). Except for cathepsin G, inhibition was not simply competitive, but showed kinetic properties corresponding to the mechanism for slow-binding inhibition, i.e. E + I + EI + E P , where EI and EI* are enzyme-inhibitor complexes and EI* is more stable than EI. This type of inhibition has not been observed previously for synthetic inhibitors or serine proteases and in this study it was observed only for peptide boronic acids which satisfy the primary specificity requirements of the protease.

Leukocyte elastase and cathepsin G are serine proteases associated with the granular fraction of polymorphonuclear leukocytes which hydrolyze multiple bonds of protein substrates. Particular emphasis has been placed on the study of leukocyte elastase due to its strong implication in diseases such as emphysema and septicemia. Its role in emphysema is in the aberrant proteolysis of lung elastin (Mittman (1972; Turino et al., 1974; Hance and Crystal, 1975) while in septicemia, elevated plasma levels of elastase associated with in- fection suggest that it is involved in nonspecific degradation of plasma proteins (Jochum et al., 1983). The role of cathepsin G in disease processes is less clear, but its ability to degrade proteoglycans and elastin suggest that it may also be involved in emphysema (Reilly and Travis, 1980). Cathepsin G also liberates the hypertensive peptide angiotensin I1 from angiotensin I (Reilly et al., 1982) and from angiotensinogen (Win- troub et al., 1984); however, the physiological significance of these reactions is unclear. Therefore, specific inhibitors of leukocyte elastase and cathepsin G may be useful therapeu- tically in the control of emphysema and septicemia as well as in determining the physiological significance of other reactions catalyzed by cathepsin G and elastase.

A number of synthetic inhibitors of leukocyte and pancreatic elastase and of cathepsin G and chymotrypsin are known. These inhibitors include peptide chloromethyl ke- tones (Powers et d. , 1977), azapeptides (Powers and Carroll, 1975), trifluoroacetyl peptides (Renaud et al., 1983), arylsul- fonyl fluorides (Yoshimura et al., 1982), and heterocyclic

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Charles A. Kettner and Ashok B. Shenvi From the Central Research and Deuehment DeDartment, E. I. Du Pont de Nemours and Company, Experimental Station,

agents (Zimmerman et al., 1980; Teshima et al., 1982). K, values for the more effective inhibitors obtained in these studies range from 0.10 to 0.010 FM (see Teshima et al., 1982, for comparisons).

The approach we have taken in a search for more selective and effective inhibitors of the leukocyte proteases pancreatic elastase and chymotrypsin has been to prepare peptide boronic acids. This group of compounds takes advantage of binding in the PI-PS sites (using the nomenclature of Schech- ter and Berger, 1967) and of the binding of the boronic acid which can potentially act as a “transition state analog.” Se- lection of the residues in the P,-P6 binding sites of the peptide boronic acids was based on the amino acid sequence of the better substrates of these enzymes from the extensive specificity studies of Zimmerman and Ashe (1977), Nakajima et al., (1979), and McRae et al. (1980). The proposed mechanism of binding of peptide boronic acids is shown in Fig. 1.

The serine proteases chymotrypsin and subtilisin have been shown to be inhibited by aromatic boronic acids (Koehler and Lienhard, 1971; Lindquist and Terry, 1974). The crystal structure of subtilisin complexed to either phenylethane boronic acid or phenyl boronic acid has been determined (Matthews et al., 1975). In the crystal structure, the trigonal boronic acid exists as a transition state-like, tetrahedral adduct with the active site serine of the protease. One -OH of boron occupies the oxyanion hole and the other -OH corresponds to the expected position of the departing NH- group for normal hydrolysis.

Matteson et al. (1981) have further demonstrated the po- tential of boronic acids as serine protease inhibitors by the preparation of the boronic acid analog of N-acetyl-phenylal-

Subrtroto Hydrolysis

B peptide-NH-CH-C

I \ OH

NH’

Boronic Acid Binding

Oe I

I I R O

+ peptide-NH-CH-C-NH-

QH peptide “NH-$H”$ + pept ide -NH-~H-B~OH

/ I

A ‘OH OH

FIG. 1. Scheme for the mechanism of binding for peptide boronic acids.

15106

Inhibition of Serine Proteases 15107

anine, (R)-acetamido-2-phenylethane boronic acid.’ They have shown that this compound is effective in the micromolar concentration range and is several orders of magnitude more effective than the aromatic boronic acids in the inhibition of chymotrypsin. In addition, the synthetic route developed by these workers was instrumental for us in the synthesis of other a-amino boronic acids used in the preparation of peptide boronic acids for the present study.

EXPERIMENTAL PROCEDURES

Bovine a-chymotrypsin, 49 units mg”, was obtained from Wor- thington. Stock solutions were prepared in 1.0 mM HCl and were stored for no longer than three days at 4 “C. Cathepsin G , was a generous gift from Dr. James Travis, University of Georgia, and was stored frozen in 0.10 M pyridine acetate buffer, pH 5.5, and 0.40 M NaC1. Porcine pancreatic elastase and human sputum leukocyte elastase were obtained from Elastin Products, Pacific, MO. Human leukocyte elastase was also prepared by the procedure of Baugh and Travis (1976). Solutions of lyophilized pancreatic elastase were prepared in 0.05 M sodium acetate buffer, pH 5.5, containing 0.40 M NaCl, and were then diluted to 50% glycerol (w/w) and stored at -20 “C. Stock solutions of leukocyte elastase (1.0 mg ml-I) were prepared in 0.05 M sodium acetate buffer, pH 5.5, containing 0.40 M NaCl and stored at -20 “C.

Chymotrypsin and cathepsin G were assayed with the substrate Suc-Ala-Ala-Pro-Phe-p-nitroanilide reported by Nakajima et al. (1979). More sensitive assays of chymotrypsin used Suc-Ala-Ala-Pro- Phe-7-amidino-4-methylcoumarin (Enzyme Systems Products). Pan- creatic elastase was assayed with Suc-Ala-Ala-Ala-p-nitroanilide (Sigma) as a substrate (Bieth et aL., 1974). Leukocyte elastase was assayed using MeO-Suc-Ala-Ala-Pro-Val-p-nitroanilide (Nakajima et al., 1979). Assay solutions contained 1% dimethyl sulfoxide for chymotrypsin and leukocyte elastase and 5% dimethyl sulfoxide for cathepsin G.

The hydrolysis of p-nitroanilide substrates was monitored at 405 nm on a Cary 219 spectrophotometer. Hydrolysis of 7-amidino-4- methylcoumarin substrates was monitored by excitation at 380 nm and measuring emission at 460 nm on a Perkin-Elmer 650-40 fluo- rometer.

The concentrations of pancreatic elastase, leukocyte elastase, and chymotrypsin were estimated by the titrimetric procedure described by Morrison (1969) using the most effective peptide boronic acid inhibitor. See data in Fig. 6 for an example. The concentration of cathepsin G was based on weight.

All kinetic constants were determined in 0.10 M sodium phosphate buffer, pH 7.5, containing 0.50 M NaCl at 25 “C. K, values for peptide boronic acids were determined from double reciprocal plots of velocity uersus substrate concentration by the method of Linweaver and Burk. Data were fitted to the best straight lines by the least-squares method. In all cases the inhibitor concentration was at least 5-fold greater than the enzyme concentration. Five or six concentrations of substrate were chosen so that values of reciprocal concentrations were evenly proportioned. The substrate concentration ranges and kinetic constants for substrate hydrolysis are given in Appendix I (Table I).z

For reactions in which progressive, time-dependent inhibition was

(R)-Acetamido-2-phenylethane boronic acid is an analog of N- acetyl-phenylalanine in which the -COOH group is replaced by -B(OH)*. We have abbreviated this compound and other a-amino boronic acids by the prefix “boro” and the name of the corresponding amino acid. All amino acidi are in the L configuration unless specified. The designation of L for the configuration of a-amino boronic acids is that of naturally occurring L-amino acids. Other abbreviations used are: SUC, succinyl, MeO-Suc, methoxysuccinyl; Hepes, N-2-hydroxy- ethylpiperazine-N’-2-ethanesulfonic acid; MS, mass spectrum; Me, methyl; Bz, benzyl; i-Pr, iso-propyl; Boc, t-butyloxycarbonyl.

Portions of this paper (including Appendices 1-111, Tables 1-111, a scheme, and additional references) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 84M-1514. cite the authors. and include a check-or money order for $6.00 per set of photocopies: Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press. wn.

observed, the reactions were initiated by the addition of enzyme and the initial and steady-state velocities were measured. Reaction velocities were measured after 30 and 60 min to establish whether a steady state had been reached. In all cases, no significant difference was observed between double reciprocal plots used to calculate K, values after 30 and 60 min, and data at 30 min were used to calculate Ki. This technique was generally applicable except for chymotrypsin in which a small correction for substrate depletion was necessary. Values of k’ , the apparent first-order rate constant for the transition from the initial velocity (ui) to the steady-state velocity Cur), were determined by the method of Cha (1975) using Equation 1. The spectrophotometer tracings of the reaction progress were extrapolated to the y axis (time = 0) after a steady state was reached and the value of the y intercept, K, was determined. For determining k ’ , the uf was measured at either 30 or 60 min depending on the inhibitor and enzyme. This method was routinely used for determining values of k’ reported in Appendix I (Table 11). The fit of the data to Equation 2 (Cha, 1975) was also determined using the MLAB computer pro- gram on the DEC-10 computing system. In Equation 2, P is the concentration of product a t time t , Po is P at t = 0.

k’ = K-’ ( ~ i - u/) (1)

P = PO + (ui - uf ) /k ’ + uft + (u, - u;)e-k’t/k’ (2)

The synthesis of amino boronic acids is described in Appendix I1 and the preparation of peptide boronic acids is described in Appendix 111. In all cases, nuclear magnetic spectra and elemental analyses were consistent with structure.

RESULTS

Synthesis and Properties of Peptide Boronic Acids-The structures of the a-amino boronic acid peptides and corresponding diol esters are given in Fig. 2. Compounds were prepared by coupling MeO-Suc-Ala-Ala-Pro-OH to pinacol esters of racemic a-amino boronic acids. The pinacol esters were converted to diethanolamine esters which were hydrolyzed in the presence of aqueous acid to yield the unprotected peptide boronic acid. In two cases, MeO-Suc-Ala-Ala-Pro- boro-Phe-OH and MeO-Suc-Ala-Ala-Pro-boro-Ala-OH, crystalline diethanolamine derivatives were obtained which al- lowed separation of the diastereomers. Assignment of the configuration of the boronic acids was based on the reactivity of crystallization fractions with chymotrypsin and pancreatic elastase assuming that the most active isomer is the “L” form. Previously, this has been shown for the “D” and L forms of Ac-Boro-Phe-OH (Matteson et al., 1981).

The diethanolamine ester of MeO-Suc-Ala-Ala-Pro-boro- Val-OH was not crystalline and therefore the separation of diastereomers was not possible by fractional crystallization. The all-L form of MeO-Suc-Ala-Ala-Pro-boro-Val-OH in

OH p 3 A’-NH-CH-B, R’-NH-CH-

/ /‘?-CH3

A OH R \wC“CH, dH3

Free Boronic Acid Pi nacol Ester

Diethonolamine Ester Pinonediol Ester FIG. 2. Structures of peptide boronic acids and correspond-

ing esters. R’ is MeO-Suc-Ala-Ala-Pro-; R is benzyl for MeO-Suc- Ala-Ala-Pro-boro-Phe-OH, R is methyl for MeO-Suc-Ala-Ala-Pro- boro-Ala-OH, and R is isopropyl for MeO-Suc-Ala-Ala-Pro-boro-Val- A T 7

15108 Inhibition of Serine Proteases

which the boronic acid was in the form of its pinanediol ester was prepared from H-L-boro-Val-pinanediol which was obtained by a stereospecific synthetic route similar to that described by Matteson et al. (1981).

The stability of MeO-Suc-Ala-Ala-Pro-DL-boro-Val-OH and its pinacol ester have been studied fairly extensively. The pinacol ester is approximately 100-fold less active in the initial inhibition of leukocyte elastase than the free acid. However, after incubation in 0.10 M phosphate buffer, pH 7.5, for less than an hour, its inhibitory activity is identical to the free boronic acid. The free boronic acid is stable under these conditions up to 24 h at room temperature. The inhibitory activity of the diethanolamine esters of MeO-Suc-Ala-Ala- Pro-boro-Phe-OH and MeO-Suc-Ala-Ala-Pro-boro-Ala-OH and the corresponding free boronic acids were identical even without preincubation.

Results obtained with the pinanediol ester indicate that this ester is also hydrolyzed to yield the free boronic acid. Comparison of the effectiveness of MeO-Suc-Ala-Ala-Pro-L- boro-Val-pinanediol after incubation in phosphate buffer and of MeO-Suc-Ala-Ala-Pro-DL-boro-Val-OH in the inhibition of leukocyte elastase were consistent with the hydrolysis of the ester to the free boronic acid. Since similar results were obtained with the other enzymes we have studied, the free boronic acid of MeO-Suc-Ala-Ala-Pro-boro-Val-pinanediol was not characterized, but instead the pinanediol ester was preincubated in 0.10 M sodium phosphate buffer, pH 7.5, for an hour prior to assays in the subsequent studies. The other peptide boronic acids were chemically characterized as the free boronic acids unless specifically stated.

Mechanism of Inhibition by Peptide Boronic Acids-The peptide boronic acids corresponding to the better substrates for chymotrypsin, pancreatic elastase, and leukocyte elastase inhibit these enzymes in a progressive, time-dependent manner. We have examined in some detail the mechanism of the inhibition of chymotrypsin by MeO-Suc-Ala-Ala-Pro-boro- Phe-OH, the most effective inhibitor of this enzyme. As shown in Fig. 3, rather than a linear increase in product with time, a decrease in reaction rate is observed in enzyme-initiated assays until a steady state is obtained (middle curve). When

0.16 1 1 I I I I I

-

0.0 20.0 30.0 40.0 50.0 60.0 Time (min)

FIG. 3. Effect of MeO-Suc-Ala-Ala-Pro-boro-Phe-OH (4 on hydrolysis of substrate ( S ) by chymotrypsin (E) . Hydrolysis of substrate Suc-Ala-Ala-Pro-Phe-p-nitroanilide was monitored by measuring the change in absorbance at 405 nm with time. In all cases, the final concentrations of S, E, and Z were 20.0 pM, 0.76 nM, and 10.0 nM, respectively. Z = 0 is the reaction in the absence of inhibitor. For ( S + T) + E, the reaction was initiated by adding E to a solution of S and I . ( E + Z) + S is the reaction in which E and Z were preincubated for 5 min and diluted 20-fold into a solution of substrate.

the inhibitor is preincubated with enzyme and the reaction is initiated by adding substrate, a progressive increase in reaction rate is observed until the same steady-state velocity as enzyme-initiated assays is obtained. The two progress curves were adequately defined by Equation 2 for time-dependent inhibition. Similar values of k ’ , the rate constant for the first- order transition from initial velocity to steady-state velocity, were obtained regardless of the method of initiating the reaction.

The initial and final velocities of chymotrypsin in the presence of 5.0 nM MeO-Suc-Ala-Ala-Pro-boro-Phe-OH were analyzed by the method of Lineweaver and Burk using data similar to those shown in Fig. 3. As shown in Fig. 4, double reciprocal plots of initial and final velocities intercept the y axis close to the intercept obtained in the absence of inhibitor indicating that inhibition is competitive. K; values calculated for the initial and final velocities are reported in Table I.

The results in Figs. 3 and 4 are consistent with the mechanism, Equation 3, for “reversible, slow-binding inhibition” as defined by Williams and Morrison (1979).

[K;(initial)/Ki(final)] - 1 = k3/kl (4)

k’ = k4 + k3 [(Z/K;(initial)/l + Z/K;(initial) + S/K,)] (5)

Equation 4 defines K;(initial), which is k2/kl , and Ki(final) in terms of rate constants in Equation 3. Equation 5 gives the relationship between k’ and the rate constants, k3 and k4 of Equation 3, Ki(initial), inhibitor concentration (I), substrate concentration (S), and the Michaelis constant (K,) (Schloss et al., 1980).

Kinetic constants k3 and k4 were determined for the reaction of MeO-Suc-Ala-Ala-Pro-boro-Phe-OH with chymotrypsin by two different methods and are compared in Table I. First, in method 1, k4 was determined experimentally in a manner similar to that described by Schloss et al. (1980). The boronic acid was incubated in a 5-fold molar excess with the enzyme and the mixture was diluted by a large factor in the assay

0.8 -

0.6 -

w \ -

0.4 t / -

INITIAL VELOCITY 0

I .I=O ~ ~ ~ ‘ ~ ~ ~ ~ 1 ’ ~ ~ ~ ” 10.0 15.0 20.0

1/S mM“

FIG. 4. Lineweaver-Burk plots for the inhibition of chymotrypsin by MeO-Suc-Ala-Ala-Pro-boro-Phe-OH. Reactions were initiated by adding chymotrypsin to a solution of substrate (S), Suc-Ala-Ala-Pro-Phe-p-nitroanilide, and inhibitor. The final concentration of inhibitor was 5.0 nM and of enzyme was 0.76 nM. Initial velocities were measured immediately after the addition of enzyme, and final velocities were measured after 30 min. Velocities (2,000 X change in absorbance at 405 nm/min) are compared with reactions run in the absence of inhibitor, Z = 0.


TABLE I Kinetic constants for substrate hydrolysis and inhibitor, MeO-Su-

Ala-Ala-Pro-boro-Phe-OH, binding by chymotrypsin in H z 0 and Dzo Kinetic constants were determined on 0.10 M sodium phosphate

buffer, pH 7.5, and pD 7.5, in 0.50 M NaCl at 25 "C. The pD was determined with a glass electrode using the relationship pD = pH + 0.40.

Hz0 DzO K, (PMY 24.7 f 4.4 18.2 -t 1.9 kat (s-')O 26.0 f 4.2 9.4 k 0.4 K,(initial) (nM)b 3.4 f 1.1 4.75 rt 0.01 K,(final) (nM)* 0.16 f 0.03 0.147 +- 0.017

kdkr 20.2 12.2 31.3 13.1 k3 (s") 0.020 0.012 0.034 0.014 k4 (s") 0.0010 0.00098 0.0011 0.0011

Method 1' Method 2d Method 1' Method 2d

K,,, and Lt values were determined by the method of Lineweaver and Burk for the hydrolysis of MeO-Suc-Ala-Ala-Pro-Phe-p-nitroanilide. In both HZ0 and DzO, the concentration range of substrate was 0.050-0.50 nM and the concentrations of enzyme were 0.78 and 2.4 nM, respectively.

* K, values were determined by the method of Lineweaver and Burk for MeO-Suc-Ala-Ala-Pro-boro-Phe-OH (see "Experimental Proce- dures"). Constants measured in Hz0 are reported in Table 11. K; values determined in D20 were measured using the conditions in Footnote a and an inhibitor concentration of 10.0 nM.

Rate constants were determined for the mechanism in Equation 3. In method 1, k4 was determined directly, as illustrated in Fig. 5. For HzO, the reported value is the average of 4 determination (S.D. k 0.00011 s-'). For D20, reaction conditions were identical to those used in H20. The reported value is the average of 5 determinations (S.D. k 0.00015 s"). Using the relationship in Equation 4 ((Ki(initial)/K,(final)) - 1 = k3/k,) k3/k4 was calculated and k3 was calculated from k d k 4 and k4.

method 2, rate constants were determined from the graphic solution of Equation 5, k' = k, + k, [(Z/Ki(initial)/l + [/&(initial) + S/K,,,)] by plotting the observed values of k' (Appendix I, Tables I1 and Tables 111) uersus the term in the parentheses of Equation 5. The plot was linear over a substrate range of 0.10-0.50 mM for reactions run both in H20 and D20.

solution. The increase in enzymatic activity could be monitored by using a sensitive fluorogenic substrate (Fig. 5 ) . Under these conditions, the term in parentheses of Equation 5 is small, and the observed k' is a good approximation of k4. The ratio, k3/kq, was calculated from the values of K,(initial) and Ki(final) using Equation 4.

In method 2, k3 and k4 were determined from the experimental values of k' (Appendix I, Table 11) and the graphic solution of Equation 5 . A comparison of rate constants obtained by methods 1 and 2 indicates that both gave consistent values of k4 and that there is some variation in k,. We feel method 1 is more accurate since method 2 is dependent on measuring rate constants at high substrate concentrations over a relatively small concentration range.

The deuterium isotope effects on the hydrolysis of Suc-Ala- Ala-Pro-Phe-p-nitroanilide and on the binding reactions of MeO-Suc-Ala-Ala-Pro-boro-Phe-OH were determined (Table I). For substrate hydrolysis, values of K, were similar in both H 2 0 and D20 while a large isotope effect, approximately 3- fold, was observed for kcat. Comparison of the isotope effect on inhibitor binding demonstrates only small differences in Ki(initial) and Ki(final).

Values of k3 and k4 determined by the graphic solution of Equation 5 (method 2) were very similar. In method 1, again, almost identical values of k4 were obtained but a larger deviation, approximately 80%, was observed in k,. Although our methods of analysis are not sensitive to small changes in rate, an isotope effect of the magnitude observed in kc, for substrate hydrolysis was not observed.

3ao "t 8 s 25.0 - K 5 2ao .- e z

-

6 - 15.0 - -

-

10.0 l / k 20.0 30.0 40.0 50.0 60.0 Time tmin)

FIG. 5. Experimental determination of k4. MeO-Suc-Ala-Ala- Pro-boro-Phe-OH, 10.0 nM, was incubated with chymotrypsin, 2.2 nM, in assay buffer. After incubation for 5 min, the solution was diluted 0.180 to 1.00 ml and a 10-pl aliquot was diluted into 0.590 ml of 0.50 mM Suc-Ala-Ala-Pro-Phe-7-amido-4-methylcoumarin to yield a final concentration of 30 p~ for the inhibitor and 6.6 PM for the enzyme. The rate of increase in fluorescence was compared with controls run under identical conditions in the absence of inhibitor (I = 0). The fluorometric tracing after a steady state was reached was extrapolated to the x axis which is l/k, (Cha, 1975). Repetitive measurements (n = 4) gave a value of 16.5 f 1.7 min for the intercept and a value of 0.0010 & 0.0001 s-' for k,.

100.0 I I I I I

0.0 I I 1 I I I \ I 0.0 100.0 200.0 3oao 400.0 500.0 m.0

Concentrotion of I InM)

FIG. 6. Stoichiometry of the binding of MeO-Suc-Ala-Ala- Pro-boro-Phe-OH (I) to chymotrypsin. Chymotrypsin (15 pg/ml) was incubated with varying concentrations of I as its diethanolamine ester in 0.300 ml of 0.10 M Hepes buffer, pH 7.5, containing 0.50 M NaCl and 10% dimethylsulfoxide for 5 min. Aliquots (8.0 p l ) were removed and assayed in the above buffer, 2.00 ml, containing 20.0 PM Suc-Ala-Ala-Pro-Phe-p-nitroanilide. Initiak velocities were compared with the control (Z = 0), 0.0261 rnin", and per cent activities were calculated. Concentrations of I are for the 0.300-ml incubation.

Finally, the stoichiometry of binding of MeO-Suc-Ala-Ala- Pro-boro-Phe-OH to chymotrypsin was determined as shown in Fig. 6. The protease was incubated with less than stoichi- ometric amounts of inhibitor, and per cent activity was determined from controls run in the absence of inhibitor. A linear relationship was obtained between per cent activity and inhibitor concentration over most of the range of inhibitor concentrations. The concentration of chymotrypsin, based on the dry weight of protein and molecular weight of 25,100 is 597 nM. The deviation we have observed is well within experimental error and leaves little doubt that MeO-Suc-Ala-Ala- Pro-boro-Phe-OH forms a 1:l complex with chymotrypsin.


Specificity of Peptide Boronic Acids-Peptide boronic acids were prepared which correspond to the structure of the pre- ferred sequence of substrates for chymotrypsin, cathepsin G, pancreatic elastase, and leukocyte elastase by varying the residue in the P, site. Chymotrypsin and cathepsin G prefer an aromatic residue in the PI site while pancreatic elastase and leukocyte elastase prefer Ala and Val, respectively. Smaller differences exist in the specificity of this group of enzymes for residues in the p2-P~ sites and MeO-Suc-Ala- Ala-Pro- is one of the best sequences for all the proteases.

The kinetic constants for the reactivities of the proteases with peptide boronic acids are shown in Table 11. Ki values shown are average values reported with standard deviations for repetitive measurements. Inhibitor concentrations used in measuring K j values are also given in Table I1 while substrates, substrate concentration ranges, protease concentrations, and kinetic constants obtained for substrate hydrolysis are given in Appendix I (Table I). All kinetic constants for substrate hydrolysis were consistent with those reported in the litera- ture.

In all cases, inhibition was competitive and both final and initial Ki values are reported for all reactions of inhibitors in which slow-binding inhibition was observed. The possibility that other compounds in Table I1 exhibit time-dependent inhibition upon prolonged incubation with the enzyme exist, but it was not observed during the 3-4-min assay period over the concentration ranges of substrates and inhibitor we have used. Therefore, any transformation into a complex more stable than that reflected by the Kt for initial complex would have to be a considerably slower process than observed with the time-dependent inhibitors we report.

For chymotrypsin, MeO-Suc-Ala-Ala-Pro-boro-Phe-OH exhibited slow-binding inhibition with an initial Ki of 3.4 nM

and a final or steady-state K, of 0.16 nM. All other compounds appeared to be simple competitive inhibitors of this enzyme. The affinity of MeO-Suc-Ala-Ala-Pro-boro-Phe-OH for chymotrypsin is at least 50-fold greater than the corresponding D isomer and is orders of magnitude greater than the corresponding boro-Val-OH and boro-Ala-OH peptides. Compari- son of Ac-boro-Phe-OH and MeO-Suc-Ala-Ala-Pro-boro- Phe-OH illustrates the extent to which the peptide portion contributes to overall binding affinity. The kinetic constant for Ac-boro-Phe-OH in Table I1 is that of Matteson et al. (1981) who report that it is a simple competitive inhibitor.

Comparison of the affinity of cathepsin G and chymotrypsin for the peptide boronic acids indicates that binding has a similar dependence on the nature of the residue in the PI site. However, in contrast to chymotrypsin and the other serine proteases in Table 11, no time-dependent inhibition was observed for cathepsin G. Time-dependent inhibition could not be detected for even the most effective inhibitor, MeO-Suc- Ala-Ala-Pro-boro-Phe-OH, after a 30-min incubation with the enzyme.

For pancreatic elastase, both MeO-Suc-Ala-Ala-Pro-boro- Val-OH and MeO-Suc-Ala-Ala-Pro-boro-Ala-OH are similar in effectiveness, exhibiting steady-state K j values in the 0.2- 0.3 nM range. Differences between these two inhibitors are reflected in a %fold lower Ki for initial binding of the boro- Ala-OH peptide. For leukocyte elastase, MeO-Suc-Ala-Ala- Pro-boro-Val-OH is the only inhibitor which exhibited slow- binding inhibition and it is several orders of magnitude more effective than the corresponding boro-Ala-OH peptide. The boro-Phe-OH peptide was considerably less active for both elastolytic enzymes. Individual kinetic constants, k3 and k d , for the slow-binding inhibitors of the elastolytic enzymes are given in Appendix I (Table 111). It should be noted that there

TABLE I1 Inhibition of serine proteoses

K; values were determined at pH 7.5 and 25 "C by the method of Lineweaver and Burk. Values are reported with S.D. for repetitive measurements. Initial Ki values were determined from initial velocity measurements. For inhibitors which exhibited slow-binding inhibition according to the mechanism in Equation 3, E + Z 9 EZ EZ*, the steady-state K; values, K;(final), were determined 30 min after initiating the reaction.

Protease Inhibitor (R-MeO-Suc-Ala-Ala-Pro-)

K; Concentration of Number of

Initial Final inhibitof determinations"

Chymotrypsin Ac-Boro-Phe-OHb R-Boro-Phe-OH' R-D-Boro-Phe-OH R-Boro-Val-OH R-Boro-Ala-OH

Cathepsin G R-Boro-Phe-OH R-D-Boro-Phe-OH R-Boro-Val-OH R-Boro-Ala-OH

Pancreatic elastased R-Boro-Phe-OH R-Boro-Val-OH R-Boro-Ala-OH

Leukocyte elastasedP R-Boro-Phe-OH R-Boro-Val-OH R-Boro-Ala-OH

nM 2,100

3.4 f 1.1 11.0 f 2.3

1,200 f 50 9,100 r 900

21.0 * 0.2 500 f 80

74,000 f 14,000 113,000 f 2,000

270 f 50 30 f 13

18.2 f 2.5 350 f 80 15 f 6 79 f 25

nM

0.16 f 0.03 5.0 100,50 5,000, 10,000 50,000 25,000 20 400 100,000 150,000 200

0.25 f 0.02 10 0.32 & 0.04 10

0.57 f 0.12 40 2,500

100

4 3 2 2 2 2 2 2 2 2 2 2 3 3

"The concentrations of inhibitor and number of determinations are reported. The enzyme concentrations, substrates, substrate concentration ranges, and values of K,,, and katare reported in Appendix I (Table I).

The K; for Ac-boro-Phe-OH was reported by Matteson et al. (1981). The rate constants, kS and k4, for the EZ - EZ* transition are reported in Table I.

0.00030 s-' and for the boro-Ala-OH peptide were 0.018 and 0.00041 s-'; for leukocyte elastase and the Boro-Val- For pancreatic elastase, the respective rate constants, ka and kr, for the boro-Val-OH peptide were 0.026 and

OH peptide they were 0.0092 and 0.00038 s-'. See Appendix I, Tables I1 and 111. 'The reported kinetic constants were determined with elastase isolated from sputum. Similar values were

obtained for elastase isolated by the procedure of Baugh and Travis (1976) for MeO-Suc-Ala-Ala-Pro-boro-Val- OH, i.e. K,(initial) 15 f 6 nM and K;(final) 0.38 f 0.08 nM.


is not much more than a %fold difference in values of k3 or in values of k4 for the elastases and chymotrypsin (Table I). Considerable refinement of kinetic measurements would be required to ascertain the effect that structure has on the small differences in rate constants.

DISCUSSION

Peptide boronic acids are a unique group of effective protease inhibitors which exhibit reversible, slow-binding inhibition. The property of slow-binding inhibition, however, is limited to those inhibitors which satisfy the primary specificity requirements for the protease and in this study were analogs of the more effective substrates. Slow-binding inhibition has been observed for inhibitors of a number of enzymes (Rich and Sun, 1980; Baici and Gyger-Marazzi, 1982; Morri- son, 1982; Schloss and Cleland, 1982), but it has not been reported for a synthetic, serine protease inhibitor. Very re- cently, slow-binding inhibition was reported for the serine protease neutrophil elastase by eglin, a 70-residue polypeptide inhibitor isolated from leech (Baici and Seemuller, 1984). In many cases, slow-binding inhibition is associated with inhibitors which act as transition-state analogs (Stark and Bartlett, 1983) or more generally, as reaction-intermediate analogs (Schloss and Cleland, 1982).

There is little doubt the peptide boronic acids, which are the more effective protease inhibitors, form a transition state- like complex with the active site serine of the protease. The affinity of the peptide boronic acid inhibitor for chymotrypsin exceeds that of the corresponding substrate by at least 4 orders of magnitude and other peptide analogs by at least several orders of magnitude. For example, chymostatin, a bacterial peptide terminating in the aldehyde analog of phenylalanine, has an ICGo of 2.5 X M for chymotrypsin (Umezawa and Aoyagi, 1977) while the final Kt for MeO-Suc- Ala-Ala-Pro-boro-Phe-OH is 1.6 X 10"' M. It would be diffi- cult to explain the high affinity of the peptide boronic acids without the contribution of the tetrahedral complex between the active site serine and the boronic acid moiety. Further- more, a tetrahedral complex has been observed in the crystal- lographic study of the bacterial serine protease subtilisin and phenylethane boronic acid (Matthews et al., 1975). The affinity of phenylethane boronic acid for subtilisin is considerably lower than even the less effective peptide boronic acids examined in this study.

The progressive, time-dependent inhibition we have observed is not readily explained on a molecular basis. One can speculate that initial binding is the formation of a ground- state complex with the enzyme and that the slower kinetic stage of the reaction involves the formation of a transition state-like tetrahedral complex of the boronic acid and the active-site serine. An analogy can be drawn between this mechanism and the mechanism for substrate hydrolysis lead- ing to the formation of an acyl-enzyme (Fig. 1). Formation of the acyl enzyme is the rate-limiting step in the hydrolysis of most amide substrates, and a significant isotope effect is observed on kcat due to the associated proton transfer (Stein, 1983). Nakatani et al. (1975) have shown that binding of phenyl boronic acid to subtilisin occurs by a similar two-step process. The first is the formation of a complex between the enzyme and the trigonal boronic acid and the second is the formation of a tetrahedral adduct with the active site serine. Like substrate hydrolysis, a deuterium isotope effect was observed for the latter step. It should be noted that the transition these workers have observed is 5 orders of magnitude faster than those we have observed and would not be considered slow-binding inhibition (Williams and Morrison,

1979). Therefore, comparable deuterium isotope effects on k3

for peptide boronic acid binding and on kcat for substrate hydrolysis would provide evidence for this mechanism. This, however, was not observed. Even though our measurements of the rate constant for the formation of the more stable complex, k3, are error prone, it is clear that the 3-fold isotope effect for kcat is not observed for k3.

Considering both the relative high effectiveness of peptide boronic acids over other peptide analogs and the absence of a definitive deuterium isotope effect on 123, we feel that all compounds probably form a tetrahedral complex with the active site serine and the slow binding phase of the inhibitor- enzyme reaction is due to a conformational change of the enzyme to optimize binding similar to that expected for binding of substrate in the transition state. Conformational changes have been observed in the crystal structure of the complex between Streptomyces griseus protease A and a tetra- peptide aldehyde inhibitor which also forms a transition state- like complex with the active-site serine (Brayer et al., 1979). However, additional studies are required to provide definitive proof of the origin of slow-binding inhibition observed for the peptide boronic acids.

Some of the structural requirements for boronic acids to exhibit slow-binding inhibition of chymotrypsin have been determined. First, binding in the primary specificity site (P,) alone, as is the case for the inhibition of chymotrypsin by Ac- boro-Phe-OH (Matteson et al., 1981)) will not give slow- binding inhibition. For peptide boronic acids, the specificity requirement of the enzyme in the PI site must be met for progressive inhibition since analogs of MeO-Suc-Ala-Ala-Pro- boro-Phe-OH containing Boro-Ala-OH and Boro-Val-OH are simple competitive inhibitors. The stereochemistry of the PI boroamino acid is also important since MeO-Suc-Ala-Ala- Pro-D-boro-Phe-OH behaves as a simple competitive inhibitor. Although this inhibitor is almost two orders of magnitude less effective than the corresponding all-L isomer, it is still a very effective inhibitor.

All peptide boronic acids which correspond to the sequence of the better substrates for individual proteases were by far the most effective inhibitors. However, inhibitor effectiveness was not entirely predicted from the substrate specificity of the protease. Zimmerman and Ashe (1977), Nakajima et al. (19791, and McRae et al. (1980) have studied the specificity of serine proteases by varying the PI residue of N-acyl peptide- p-nitroanilide and amide substrates containing P,Ala, P3Ala, and P2Pro. These studies have shown that substrates containing a Phe in the P1 site are selective substrates for chymotrypsin and cathepsin G, and those with Ala and Val in the P, sites are selective substrates for leukocyte and pancreatic elastase. The chymotrypsin substrates are not hydrolyzed by the elastases, and the elastase substrates are not hydrolyzed by chymotrypsin and cathepsin G. Comparing the two elastolytic enzymes, leukocyte elastase exhibits an approximate 10-fold preference for the Val substrate, and pancreatic elastase has a similar preference for the Ala substrate. MeO-Suc- Ala-Ala-Pro-boro-Phe-OH was 4 orders of magnitude more effective in the inhibition of chymotrypsin and cathepsin G than reagents with boro-Val and boro-Ala in the PI site. MeO- Suc-Ala-Ala-Pro-boro-Val-OH is the most effective inhibitor for leukocyte elastase and the boro-Ala and boro-Phe peptides were 100 and 500-fold less effective. For pancreatic elastase, MeO-Suc-Ala-Ala-Pro-boro-Ala-OH and MeO-Suc-Ala-Ala- Pro-boro-Val-OH are similar in effectiveness and are 3 orders of magnitude more effective than the corresponding Borophe peptides. It should be noted that the Borophe peptide is still an effective inhibitor of both elastases and this clearly would


not have been predicted from the enzyme’s substrate specific- ities.

The high affinity of peptide boronic acids (Ki = 0.2-20 nM) and their selectivity for particular proteolytic enzymes should make them valuable for study of the physiological roles of cathepsin G and leukocyte elastase. In addition, they may be useful for the treatment of diseases in which aberrant proteolysis has been implicated such as in emphysema and septicemia (Hardie et al., 1984):

Acknowledgments-We are grateful to Jim Travis, University of Georgia, for the gift of cathepsin G , to Donna Reed and Karl Lonberg- Holm for providing purified leukocyte elastase, and to John Schloss who was extremely helpful in the interpretation of kinetic data. The excellent technical assistance of L. J. Mersinger and J. H. McCartney is acknowledged. Finally, we would like to thank John Schloss and Karl Lonberg-Holm for reviewing this manuscript and for their helpful suggestions.

REFERENCES

Baici, A., and Gyger-Marazzi, M. (1982) Eur. J. Biochem. 129, 33-

Baici, A., and Seemiiller, U. (1984) Biochem. J. 218, 829-833 Baugh, R. J., and Travis, J. (1976) Biochemistry 15,836-841 Bieth, J., Spiess, B., and Wermuth, C. G. (1974) Biochem. Med. 11,

Brayer, G. D., Delbaere, L. T. J., James, M. N. G., Bauer, C.-A., and Thompson, R. C. (1979) Proc. Natl. Acad. Sci. U. S. A. 76,96-I00

Cha, S. (1975) Biochem. Pharmacol. 24,2177-2185 Hance, A. J., and Crystal, R. G. (1975) Am. Reu. Respir. Dis. 112,

Jochum, M., Duswald, K.-H., Neumann, S., White, J., Fritz, H., and Seemiiller, U. (1983) in Proteinase Inhibitors (Katunuma, N., Ume- zawa, H., and Holzer, H., eds) pp. 85-95, Springer-Verlag, New York

Koehler, K. A., and Lienhard, G. E. (1971) Biochemistry 10 , 2477- 2483

Lindquist, R. N., and Terry, G. (1974) Arch. Biochem. Biophys. 160, 135-144

Matteson, D. S., Sadhu, K. M., and Lienhard, G. E. (1981) J. Am. Chem. SOC. 103,5241-5242

41

350-357

657-711

Hardie, R. D., Soskel, N. T., Watanabe, S., Shenvi, A. B., Punt, J. A., and Kettner, C. (1984) Abstract American Thoracic Society Meeting, May 20-23, Miami, FL.

Matthews, D. A., Alden, R. A., Birktoft, J . J., Freer, S. T., and Kraut,

McRae, B., Nakajima, K., Travis, J., and Powers, J. C. (1980) Bio-

Mittman, C. (ed) (1972) Pulmonary Emphysema and Proteolysis, pp.

Morrison, J. F. (1969) Biochim. Biophys. Acta 185, 269-286 Morrison, J. F. (1982) Trends Biochem. Sci. 7, 102-105 Nakajima, K., Powers, J. C., Ashe, B. M., and Zimmerman, M. (1979)

Nakatani, H., Uehara, Y., and Hiromi, K. (1975) J. Biochem. 78 ,

Powers, J. C., and Carroll, D. L. (1975) Biochem. Biophys. Res.

Powers, J. C., Gupton, B. F., Harley, A. D., Nishino, N., and Whitley,

Reilly, C. F., and Travis, J. (1980) Biochim. Biophys. Acta 621 , 147-

Reilly, C. F., Tewksbury, D. A., Schechter, N. M., and Travis, J.

Renaud, A., Lestienne, P., Hughes, D. L., Bieth, J. G., Dimicoli, J.-

Rich, D. H., and Sun, E. T. (1980) Biochem. Pharmacol. 29 , 2205-

J. (1975) J. Biol. Chem. 2 5 0 , 7120-7126

chemistry 19,3973-3978

1-537, Academic Press, New York

J. Biol. Chem. 254 , 4027-4032

611-616

Commun. 67,639-644

R. J. (1977) Biochim. Biophys. Acta 485 , 156-166

157

(1982) J. Biol. Chem. 257,8619-8622

L. (1983) J. Biol. Chem. 258,8312-8316

2212 Schechter, I., and Berger, A. (1967) Biochern. Biophys. Res. Commun.

27,157-162 Schloss, J. V., and Cleland, W. W. (1982) Biochemistry 2 1 , 4420-

Schloss, J. V., Porter, D. J . T., Bright, H. J., and Cleland, W. W.

Stark, G. R., and Bartlett, P. A. (1983) Pharrnacol. Ther. 23,45-78 Stein, R. L. (1983) J. Am. Chem. SOC. 105, 5111-5116 Teshima, T., Griffin, J. C., and Powers, J. C. (1982) J. Biol. Chem.

Turino, G. M., Rodriquez, J. R., Greenbaum, L. M., and Mandl, I. (1974) Am. J. Med. 6 7 , 493-503

Umezawa, H., and Aoyagi, T. (1977) in Proteinases in Mammalian Cells and Tissues (Barrett, A. J., ed) pp. 637-662, North-Holland Publishing Co., Amsterdam

Williams, J. W., and Morrison, J. F. (1979) Methods Enzymol. 63,

Wintroub, B. U., Klickstein, L. B., Dzau, V. J., and Watt, K . W. K.

4427

(1980) Biochemistry 19 , 2358-2362

267,5085-5091

437-467

(1984) Biochemistry 23, 227-232 Yoshimura, T., Barker, L. N., and Powers, J. C. (1982) J. Biol. Chem.

257.5077-5084 Zimmerman, M., and Ashe, B. M. (1977) Biochim. Biophys. Acta 480,

Zimmerman, M., Morman, H., Mulvey, D., Jones, H., Frankshun, R.,

Additional references are found on p. 15114.

241-245

and Ashe, B. M. (1980) J. Biol. Chem. 255,9848-9851


0.0017~.00013 0 . 0 0 2 8 ~ . 0 0 0 l 3

0.0039stO.WO48 0 . W 3 6 e . w O I

- O.W192+-0.0002 0.003bs+a.00032 0.00432~.00060 0.0062~.00213

0.0 nn, 2) o .oo lo4~ .wol l

o.oos1~o.ooas2 O.O0172~.WOl5 0.00238+O.WWl

O.O03S~.OOOil

O.OOl38+0.OOW5

O . W 2 8 ~ . O W 2 6

0.004l7+-0.0001 0.003l~O.OOQ24

.o nn. 2)

o.ool8%.ow22

3 nn. 2 )

dp" 11. spthe.i. d o-hinobo-c Acid.

SYNTHETIC SCHEME FOR a-AMINOBORONIC ACIDS

Ma he

c . R = B z ; )=>$ d.R=Me ; )=>::


17.60; B . 3 . 2 4 .

Pound: C. 51.541; 8, l.562;2$,3?3?bh; a d 8 . 2.llX. mal. Old. for C II W 0 8: C, 51.661; H, 7.501. N. 13.701; and 8 . 2.111.

~ ReQBReNCES

C a s t i t l o . H. 53-64. J.. Nakajlm. K.. Zime-. If.. m d P a e r a . J.C. (1979) Anal. B i w h e m . 99.

Dale, J. A , , 111. 0. I.. and nosher . H. S. (1969) J. 0%. Chem.. 34. 2543-2549.

b y . R. and )(Btreeon. D.S. (1980) Tetrahedron Lett. 21, 449-450.

Washburn. 1. II.: Levens. e., N b r i g h l . C. Q . , and B i l l i g , P. A. (1963) Organic Synthes is . C e l l . Val y. 68-72.

the of vol. no. the of printed m u. a. inhibition of the serine … · 2017-08-15 · inhibition of...

Documents