Vol. 254, No. 2, Issue of January 25, pp. 356-366, 1979 Printed m U.S.A.

Catalytic Role of the Metal Ion of Carboxypeptidase A in Ester Hydrolysis*

(Received for publication, May 15, 1978)

Marvin W. Makinen,$ Lawrence C. Kuo,@ John J. Dymowski, and Shams Jaffer

From the Department of Biophysics and Theoretical Biology, Cummings Life Science Center, The University of Chicago, Chicago, Illinois 60637

The mechanism of action of bovine pancreatic car- boxypeptidase A, (peptidyl-L-amino acid hydrolase; EC 3.4.12.2) has been investigated by application of cryoen- zymologic methods. Kinetic studies of the hydrolysis of the specific ester substrate 0-( trans-p-chlorocinna- moyl)-L-P-phenyllactate have been carried out with both the native and the Co’+-substituted enzyme in the 25 to -45°C temperature range. In the -25 to -45°C temperature range with enzyme in excess, a biphasic reaction is observed for substrate hydrolysis charac- terized by rate constants for the fast (kf) and the slow (k8) processes. In Arrhenius plots, /z~ extrapolates to kcat at 25°C for both enzymes in aqueous solution, indicat- ing that the same catalytic rate-limiting step is ob- served. The slow process is analyzed for both metal enzymes, as previously reported (Makinen, M. W., Ya- mamura, K., and Kaiser, E. T. (1976) Proc. Nutl. Acud. Sci. U. S. A. 73, 3882-3886), to involve the deacylation of a mixed anhydride acyl-enzyme intermediate. Near -60°C the acyl-enzyme intermediate of both metal en- zymes can be stabilized for spectral characterization. The pH and temperature dependence of k, reveals a catalytic ionizing group with a metal ion-dependent shift in pK, and an enthalpy of ionization of 7.2 kcal/mol for the native enzyme and 6.2 kcal/mol for the Coz+ enzyme. These parameters identify the ioniz- ing catalytic group as the metal-bound water molecule. Extrapolation of the pKc, data to 25°C indicates that this ionization coincides with that observed in the acidic limb of the pH profile of log( k,,t/K,,+,p,)) for sub- strate hydrolysis under steady state conditions. The results indicate that in the esterolytic reaction of car- boxypeptidase A deacylation of the mixed anhydride intermediate is catalyzed by a metal-bound hydroxide group.

The mechanism of action of carboxypeptidase A (peptidyl- L-amino acid hydrolase; EC 3.4.12.2) has been conjectural despite a wide variety of thorough kinetic, structural, and chemical studies (l-4). The roles of catalytic residues of the enzyme are not unambiguously defined for both peptide and ester hydrolysis, and kinetic studies have not revealed the nature and number of intermediates subsequent to substrate binding. Structural requirements for substrate specificity and

* This work was supported by Public Health Service Grant GM 21900 of the National Institutes of Health. 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.

$ Established Investigator of the American,Heart Association. § Predoctoral trainee of the National Institutes of Health (GM 780

and GM 07183).

rapid hydrolysis are a free carboxyl group and an aromatic or branched aliphatic side chain of the COOH-terminal residue. The binding relationships of substrates to carboxypeptidase A, determined by x-ray crystallographic methods, indicate that the terminal, free carboxyl group is salt-linked to arginine 145 and the aromatic side chain is enclosed in a deep hydro- phobic pocket of the enzyme (1,2). Once this binding mode is assumed, the only residues close enough to the substrate to enter into catalysis are the side chains of glutamate 270 and tyrosine 248, the tetrahedrally coordinated Zr?’ ion, and the metal ion-bound water molecule. Atomic model building stud- ies, however, do not differentiate between the possible roles of glutamate 270 in direct nucleophilic attack on the carbonyl carbon of the substrate or in hydrolysis by a general base- catalyzed mechanism. Suggestions have been made that the binding of ester and peptide substrates may involve partially overlapping sites on the enzyme (5) and that their hydrolysis may occur by different mechanisms (6). While L-phenylala- nine and L-P-phenyllactate, frequently employed as the COOH-terminal residue in synthetic peptide and ester sub- strates, bind in similar configurations (7), x-ray studies have been carried out only for peptide substrates. No esterolytic substrate has been sufficiently sluggish for data collection by x-ray diffraction methods. In addition, difference Fourier stud- ies (1,2,7) indicate that the structure of the enzyme in binding sluggish peptide substrates in the crystal is not compatible with hydrolysis catalyzed by a metal-hydroxide species, as demonstrated for small molecule metal ion complexes (8-10).

An approach developed in recent years to detect and char- acterize intermediates of enzyme reactions is the application of organic-aqueous cosolvent mixtures under conditions of subzero temperatures (11-13). By this method, reaction inter- mediates under nonturnover conditions can be accumulated in approximate stoichiometric concentrations in the 0 to -100°C range. We have demonstrated in a preliminary study (14) that temporal resolution of the reaction catalyzed by carboxypeptidase A can be achieved in the subzero tempera- ture range and that hydrolysis of the specific ester substrate 0-t trans-p-chlorocinnamoyl)-L-P-phenyllactate proceeds via formation of the covalent acyl-enzyme intermediate illus- trated schematically in Structure I.

p-------&~45 Glu-270 C-O-C-R

lo 0

STRUCTURE I

This initial study has served as an incisive approach in determining new details of the mechanism of carboxypepti-

356

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Mechanism of Carboxypeptidase A 357

dase A since discrete intermediates of the enzyme reaction had not been previously detected. In this communication, we demonstrate that the steps of the esterolytic reaction involving the chemical transformation of the substrate into products are identical in both cosolvent mixtures and in aqueous me- dium. In contrast to the interpretations of structural and chemical studies, our results indicate that the active species in ester hydrolysis catalyzed by carboxypeptidase A is a metal- hydroxide complex.

EXPERIMENTAL PROCEDURES

Materials-The specific ester substrate ClCPL’ was synthesized as described by others (15, 16). The tightly bound inhibitor L-benzylsuc- cinate (17) was a gift from Professor E. T. Kaiser of the Department of Chemistry at The University of Chicago.

The crystalline a-isozyme of bovine pancreatic carboxypeptidase A obtained from Sigma was used as previously described (14). Prepa- ration of CoCPA was carried out as described by Latt and Vallee (19) with use of deionized distilled water. The extent of metal substitution was evaluated by atomic absorption methods. Identical results in kinetic studies were obtained for all preparations of CoCPA with 88 to 98% metal replacement.

All chemicals were of analytical reagent grade. CoClp was obtained by treatment of Specpure Co bar (Johnson Matthey Chemicals, Ltd.) with Ultrex HCl (Baker). Deuterium oxide (99% D20) was obtained from Stohler Isotopic Chemicals.

Methods-Initial velocity data were collected spectrophotometri- tally at 310 nm, specific for the hydrolysis of ClCPL. For ZnCPA, the reaction mixture was supplemented (19) with 10m4 M Zn*+ below pH 6, and for CoCPA, with 10m4 M Co’+ below pH 7.

For studies with cryosolvent mixtures, solutions were prepared as prescribed by Douzou and co-workers (11,18). The pan of cryosolvent mixtures was confirmed spectrophotometrically with indicator dyes. For thermostating the spectrophotometer cuvette in kinetic and spectral studies, a cryostat assembly developed in this laboratory for use in a Cary 15 spectrophotometer was employed as described previously (20). The preparation of enzyme solutions in ternary solvent mixtures in the subzero temperature range was carried out below -2O’C by injection of small aliquots of MeOH into an EC/Hz0 mixture, coordinated with temperature regulation (21) to prevent precipitation of the enzyme. The addition of MeOH is associated with a marked decrease in the solubility of dissolved gases in the EG/HzO mixture, observed by the appearance of small gas bubbles. Therefore, MeOH addition is best carried out only in the -25 to -2O’C range where the escape of gas bubbles proceeds readily.

Initial velocity data were evaluated with use of the computer program ENZKIN provided by Professor J. L. Westley of the De- partment of Biochemistry at The University of Chicago. The steady state kinetic parameters &at and K,(,,,, were calculated with this algorithm by an iterative nonlinear least squares fit to a rectangular hyperbolic function. Rate constants for the separated kinetic data for ZnCPA studies with enzyme in excess collected under subzero tem- perature conditions were calculated with use of a modified version of the nonlinear least squares algorithm of Marquardt (22), and, for CoCPA studies, by nonlinear regression analysis of the separated rate data with programmable calculators or by graphical analysis.

RESULTS

Ester Hydrolysis under Steady State Conditions

Influence of Cryosoluents-The influence of EG and MeOH on the hydrolysis of ClCPL has been determined on the basis of initial velocity data. The results are compared in Fig. 1 and in Table I. MeOH exhibits noncompetitive inhibition of

’ The abbreviations used are: ClCPL, O-( trans-p-chlorocin- namoyl)-L-P-phenyllactate; CoCPA, Co*+-substituted carboxypepti- dase A,: E,, Arrhenius activation energy; EG, ethylene glycol; (Eo), initial enzyme concentration; MeOH, methanol; (SO), initial substrate concentration; ZnCPA, native carboxypeptidase A,; AHi,,, enthalpy of ionization. The composition of cryosolvent mixtures is designated in percent(%) on a volume basis. Hydrogen ion activity in cryosolvent systems is measured as the pan (11, 18). This is used interchangeably in the text with pH.

I ZnCPA I cocpA /

I/% x 10-a (WI I/So X IO-’ CM-‘)

FIG. 1. Comparison of MeOH inhibition of ZnCPA and CoCPA in the hydrolysis of ClCPL. Reaction mixtures contained 0.25 M sodium chloride buffered to pH 7.5 with 0.01 M sodium cacodylate at 23°C. The MeOH concentration ranged from 0.12 to 7.4 M and is indicated in percent (%) of total volume (v/v). (.I&) = 2.1 X 1O-8 M, and (SO) ranged from 3.0 to 20.0 x 10m5 M. The dissociation inhibitor constants (Kr) are estimated to be -0.6 M and -2.2 M for ZnCPA and CoCPA, respectively. The lines in both plots are drawn according to values of K m(app) and V,,, calculated from the nonlinear least squares fit of initial velocity data to a rectangular hyperbolic function.

ZnCPA but competitive inhibition of CoCPA. The inhibitory influence of EG on both ZnCPA and CoCPA is noncompeti- tive, in contrast to that of MeOH. For CoCPA in binary mixtures containing more than 30% EG, the value of K,,,(appJ is increased so that sufficiently high substrate concentrations cannot be achieved for accurate estimation of the kinetic parameters. In ternary solvent mixtures, the overall inhibitory influence of both cryosolvents appears to remain noncom- petitive. In contrast to ZnCPA, steady state kinetic parame- ters for CoCPA in mixed solvents are complicated by substrate inhibition that becomes more pronounced with decreasing temperature.

Influence of Temperature-The variation of kcat with tem- perature is compared in Fig. 2 for aqueous and cryosolvent mixtures. The Arrhenius plots yield values of E, of 16.5 f 0.8 kcal/mol for ZnCPA and 14.6 + 0.9 kcal/mol for CoCPA in aqueous solution. With binary solvents containing MeOH, a change in the temperature dependence of kcat is observed only for ZnCPA. For CoCPA the data for kcai overlap with those observed in aqueous medium. This behavior is expected since MeOH acts noncompetitively toward ZnCPA but competi- tively toward CoCPA. On the other hand, the temperature dependence of kcat is altered for both ZnCPA and CoCPA in EG-containing solvent mixtures. The change in the tempera- ture dependence of k,,, indicates that above 0°C the enzyme reaction does not follow the same rate law under conditions of noncompetitive inhibition by mixed solvents as in aqueous solution. This interpretation is further supported by data in Table II. A deuterium isotope effect is observed for both ZnCPA and CoCPA in aqueous solution identical with that reported for the hydrolysis of 0( trans-cinnamoyl)-L-P-phen- yllactate by the y-isozyme (23). This effect is not observed with noncompetitive inhibition by cryosolvents at room tem- perature.

Ester Hydrolysis under Conditions of Enzyme in Excess

Temporal Resolution of the Enzyme Reaction in the Sub- zero Temperature Range-With enzyme in excess, the hy- drolysis of ClCPL in the subzero temperature range is char- acterized by a biphasic decrease in absorbance. In Fig. 3, the

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

358 Mechanism of Carboxypeptidase A

TABLE I

Comparison of kinetic parameters for hydrolysis of ClCPL in cryosolvent mixtures by &CPA and CoCPA

Solvent mixture”

Z&PA C&PA

kc.,, K”hd Solvent mixture” k cat K rnimml

s-l lo4 M s-l 104 M

0% EG 59.01 + 3.54 1.01 + 0.31 0% EG 116.0 * 4.0 0.68 f 0.29 50% EG’ 3.12 + 0.76 0.98 -+ 0.53 5% EG 122.5 + 16.8 0.91 + 0.22

10% EG 96.5 + 28.1 0.72 f 0.32 30% EG* 53.6 + 2.3 1.07 + 0.08

EG/MeOH/HzO (25:25:50) 9.29 k 3.84 2.64 + 1.67 EG/MeOH/HzO (25:25:50) 28.5 + 1.5 2.66 + 0.20 EG/MeOH/HzO (40:20:40) 2.41 + 0.48 1.21 + 0.40

n Solvent conditions: 24”C, 0.25 M NaCl, buffered to pH 7.5 with 0.01 M sodium cacodylate. * Dissociation inhibitor constants (Kz) of EG are estimated as -0.6 M and -2.9 M for ZnCPA and CoCPA, respectively.

IOC

p IC

z 2

I

0.1

16.5

9.6

I 1 I I I I

3.4 3.6 3.6

B

I

:-i:-i;

I

“Cd%

‘\\,

14.6

II 2

IO 2

f I / I I I 1 3.4 36 38

I, , I I I/T x,10’ I I I I

l 30 *I3 0 -IS l 30 *I3 0 -15

“C

TABLE II

Comparison of steady state kinetic parameters for hydrolysis of CKPL in H20- and D20-containing solvents ZnCPA C&PA

FIG. 2. Comparison of the tempera- ture dependence of kcat for the hydrolysis of ClCPL in aqueous and cryosolvent mixtures. A, kcat for ZnCPA; and B, kcat for CoCPA. The following solvent sys- tems were employed: A, A, HzO; 0, +, EG/MeOH/HzO (25:25:50); q , W, EG/HzO (50:50). Ionic strength, buffer, (Eo), and range of (SO) are as in Fig. 1. The Arrhenius activation energy is indi- cated in kcal/mol for each solvent sys- tem calculated from the least squares slope of the data. Steady state kinetic parameters could not be estimated for the reaction of CoCPA in the EG/H20 (50:50) mixture. The temperature de- pendence of VO/(EO) in this cosolvent mixture was determined under condi- tions of (S,)/(Eo) =600:1.

Solvent"

k Kn,a,,, (kcaJHzO /(kcdu2o

k Lm,, (kcrAHzO eat cat /(kc.,)u20

s-1 lo4 M s-1 lo4 M

Hz0 73.2 + 5.3 0.98 f 0.14 1.9 115.8 + 3.9 0.66 + 0.04 Dd 38.5 f 5.2 0.74 f 0.17 63.9 -I 6.7 0.61 f 0.10 1.8

EG/MeOH/H20 (25:25:50) 9.3 + 3.7 2.64 + 1.67 0.9

30.3 f 1.5 2.83 + 0.21 EG/MeOH/DzO (25:25:50) 10.9 +- 3.2 5.71 & 0.29 39.9 + 3.8 8.77 f 0.99' 0.8

” All solvents contained 0.25 M sodium chloride, buffered to pH 7.5 or pD 7.9 with 0.01 M sodium cacodylate at 25°C. b Determined from Lineweaver-Burk plot of initial velocity data.

small deviation from the linear portion of pseudo-first order plots shows two kinetically discrete processes. The biphasic reaction is characterized by rate constants for the fast (& and slow (h,) processes which correspond to formation and break- down of a covalent acyl-enzyme intermediate (14). For CoCPA, the deviation from the linear portion of the plot is observed most readily only at lower pH, as shown in Fig. 4. Here the deviation from the linear portion of the slow process is too small for accurate estimation of hf above -40°C. Below -45”C, on the other hand, the high viscosity of the cryosolvent mixtures results in long mixing times so that the fast process is not observed.

There is an additional difference in the nature of the de-

crease in absorbance for the ZnCPA and CoCPA reactions. For the ZnCPA reaction, the linear portion of the slow process characteristically extrapolates to the value of (A* - Am) pre- dicted by (So). For the CoCPA reaction, the linear portion of the slow process extrapolates to values of (A,, - A,) that are greater by a value of Ac of -4000 M-' cm-’ than that predicted by (So). The higher extrapolated value of (A0 - A,) is of interpretative significance and is discussed later.

The temperature dependence of the kinetic data in the -25 to -45°C range is compared in Fig. 5. For both ZnCPA and CoCPA, h, extrapolates to Scat measured in aqueous solution at 25°C. Similarly, as previously noted (14), kf extrapolates to the value of kcat at 25°C determined for ZnCPA in the identical

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Mechanism of Carboxypeptidase A 359

cosolvent mixture. Comparable results are obtained for the three cryosolvent mixtures EG/H20 (50:50), EG/MeOH/HzO (25:25:50), and EG/MeOH/HtO (40:20:40) with both ZnCPA and CoCPA.

IC

0:

8 7 01 a

0.0:

00 I 1 I 1 / /

200 400 600

Time bed FIG. 3. First order plot of the decrease in absorbance at 310 nm

during hydrolysis of ClCPL by ZnCPA at -4O”C, illustrating graphi- cally two kinetically discrete processes. The enzyme solution was prepared in the ternary mixture containing EG/MeOH/H20 (40:20: 40) buffered to pH 7.5 at -40°C with 0.01 M sodium cacodylate at an ionic strength of 0.26 M sodium chloride. The enzyme concentration was 1.2 x 10e4 M, and the initial substrate concentration was 4.8 X 10W5 M. 0, observed absorbance decrease; 0, separated fast process with rate constant Iz,. - - -, indicates the separated slow step with rate constant k,.

For conditions of enzyme in excess, k, is equivalent to the observed pseudo-first order rate constant kObs, where kobs = kcatWo)/Kn(app) + (Eo)); and may be, therefore, influenced by kcat and Kw+) if saturating levels of enzyme are not achieved. In connection with our preliminary study (14), we have esti- mated the limiting value of k, under conditions of very high ZnCPA concentrations. At ZnCPA concentrations of -8 X

lo-* M, the limiting value of k, at -40°C was -9 X 10m3 s-l. In Fig. 5, this coincides precisely with the value obtained by extrapolation to -40°C of the least squares-determined slope of the kcat data in aqueous solution. The experimental condi- tions employed thus permit reasonable estimates of the ob- served first order rate constants, and the results indicate that the enzyme reaction is characterized by the same rate-limiting step in the subzero temperature range as in aqueous solution.

pH Dependence of Hydrolysis of CLCPL in Cryosolvent Mixtures--The pH dependence of k, in cryosolvent mixtures is illustrated in Fig. 6. The data provide evidence for ionization of a group associated with the catalytic rate-limiting step. The pH dependent change in k, for CoCPA parallels that for ZnCPA but is shifted to a lower pH range. The pH dependence of k, has been plotted according to Equation 1

k, I (k,him Ka ’ K,+ (H’)

(1)

where (ks)lim represents the graphically estimated limiting value of k, at high pH, and K, the ionization constant. The temperature dependence of the pK, values yields a AHi,” of 7.2 k 0.1 kcal/mol for ZnCPA and 6.2 + 0.1 kcal/mol for CoCPA for the ionizing group.

In Tables III and IV, a comparison is made of the pH dependencies of the activation parameters of the slow process. Possible interpretations of the large positive entropy of acti- vation have been discussed previously (14). By deriving the relationship for d (l&,)/d(l/T) on the basis of Equation 1, the observed enthalpy of activation can be separated into a pH independent AH+ term associated with (ks)li, and the AH,,” associated with the pK, of the ionizing group. An estimate of the AHi,, can be made thus from the maximum difference

oL 2& 4kJ 6&J :,' ! I 2Ao 4th &o time (sac)

FIG. 4. First order plot of the de- crease in absorbance at 310 nm during hydrolysis of ClCPL by CoCPA at -40°C. The enzyme solution was pre- pared as in Fig. 3 but buffered to pH 5.0 at -40°C with 0.01 M sodium acetate. A, the total enzyme concentration was 1.2 x 10e4 M and the initial substrate con- centration was 0.40 X 1O-4 M. & the total enzyme concentration was 0.60 x 10e4 M and the initial substrate concentration was 0.20 X 10e4 M.

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

360 Mechanism of Carboxypeptidase A

+20 0 -20 -40 +20 0 -20 -40 *c

n

! FIG. 6. Comparison of the pH dependence of k, in hydrolysis of

CICPL by carboxypeptidase A, in the subzero temperature range. A, k, data for ZnCPA in EG/MeOH/&O (25:25:50); B, k, data for CoCPA in EG/MeOH/HzO (40:20:40). Cryosolvent mixtures were buffered with either 0.01 M sodium acetate, 0.01 M sodium cacodylate, or 0.01 M Tris-HCl according to DOUZOU and co-workers (18) at an ionic strength of 0.26 M sodium chloride. Data correspond to the following temperatures: 0, -25’C; A, A, -30°C; II, W, -35°C; 0, l , -40°C. The data are visually fit to the solid line drawn for a theoretical curve corresponding to Equation 1. The horizontal line on the right-hand side of each graph indicates the graphically esti- mated asymptotic limiting value of K,. The estimated pK, values indicated by arrows in order of decreasing temperature are 7.25, 7.30, 7.52, and 7.65 for Z&PA and 6.08, 6.20, and 6.33 for CoCPA.

observed in the values of AH+ at high and low pH. This treatment yields values of 9.4 and 6.6 kcal/mol for ZnCPA and CoCPA, in good agreement with those derived from the temperature dependence of the pK,. The similar behavior of the two metal enzymes in substrate hydrolysis in aqueous solutions and in cryosolvent mixtures indicates that identical catalytic processes obtain for both Z&PA and CoCPA.

Metal-substituted derivatives of carboxypeptidase A, are less stable in the acidic pH range, requiring addition of excess metal ion for maximum activity (19, 24). Addition of 10e4 M

FIG. 5. Comparison of the tempera- ture dependence of rate constants for the hydrolysis of ClCPL. A, rate constants for hydrolysis by ZnCPA; B, rate con- stants for hydrolysis by CoCPA. Solu- tions were buffered to pH 7.5 with 0.01 M sodium cacodylate and contained 0.26 M sodium chloride. Data for & in aqueous medium are provided from Fig. 2. Data for h, and & are obtained from experiments as in Figs. 3 and 4. The solvent mixtures are: A, A, H20; 0, EG/MeOH/HzO (25:25:50); 0, 0, EG/MeOH/HzO (40:20:40). Error bars indicate the standard error where this is larger than the diameter of the symbol used. The lines for extrapolation of sub- zero temperature data to the higher tem- perature range are drawn according to the least squares slope of Ir, or kf data.

TABLE III $3 dependence of activation parameters associated with k, in

hydrolysis of CZCPL by .&CPA

PH" Eab AHi AS+

kcal/mol kcal/mol e.u.

5.5 28.7 f 4.4 28.1 34.4 6.0 26.2 + 0.3 25.6 30.0 6.5 26.3 + 0.4 25.7 35.5 7.0 24.7 + 0.5 24.1 30.7 7.5 20.2 + 0.3 19.5 14.8 8.0 21.9 + 0.5 21.3 22.7 8.5 19.3 c 0.3 18.7 11.6 9.5 18.2 + 0.3 17.6 6.gd

B Solvent conditions are those in Fig. 6. * E, and standard errors are calculated from the least squares slope

of Arrhenius plots of k, data. Identical errors are associated with each corresponding value of the enthalpy of activation.

‘Estimated from the calculated least squares slopes of plots of Tln( k,/T) uersus T to minimize statistical correlation with estimates of E,. The estimated standard error is -*3 e.u. for each entry.

d Since here activation parameters may be influenced by an over- lapping pK at high pH (16), as suggested by the markedly lower AS& the value of AH+ has not been included in the estimate of the AH,,,.

TABLE IV $7 dependence of activation parameters associated with k, in the

hydrolysis of ClCPL by CoCPA

PH" Lb AH+ A+'

kcal/mol kc&no1 e.u.

5.0 27.0 26.4 29.9 6.0 21.0 20.4 16.0 6.5 21.6 21.0 19.2 7.0 21.0 20.4 17.7 7.5 20.7 20.1 16.4 8.0 20.4 19.8 13.5 8.5 20.4 19.8 14.9

n Solvent conditions as in Fig. 6. * Determined as described in Table III. The largest estimate of the

standard error for E, throughout the pH range is +O.l kcal/mol.

metal ion to cosolvent mixtures for low temperature studies caused a slow rate of background hydrolysis of ClCPL not related to the active site, detected by the addition of excess L-benzylsuccinate prior to the addition of substrate. Subtrac- tion of this background rate gave results comparable to those observed without added metal ion. This background hydro- lytic activity is possibly related to extra metal ion-binding

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Mechanism of Carboxypeptidase A 361

A0.D 0.1 t

I I I I I , J 240 260 260 300 320

Wavelength (nm)

FIG. 7. Comparison of the ultraviolet difference absorption spec- trum of equimolar mixtures of carboxypeptidase A, and ClCPL at -60°C versus free enzyme and substrate at ambient temperature. The ternary solvent EG/MeOH/HzO (40:20:40) was buffered to pH 7.5 with 0.01 M sodium cacodylate and contained 0.26 M sodium chloride. The total enzyme concentration was 1.1 x lo-” M with an equimolar quantity of ClCPL. Further details of the procedure for recording the difference spectrum have been provided in Ref. 14. -, corresponds to the spectrum for ZnCPA; - - -, corresponds to the spectrum for CoCPA. Extinction coefficients at 310 nm estimated for both covalent intermediates employing the extinction coefficient of 12,000 M-’ cm-’ for the substrate (15, 16) are -22,000 and 19,000 M-’ cm-’ for ZnCPA and CoCPA, respectively, assuming complete conversion of the enzyme to the covalent intermediate.

sites on the enzyme (25) observed at the high enzyme concen- trations employed at subzero temperatures. We conclude that the low temperature conditions retard metal ion loss from the enzyme in contrast to the behavior of the enzyme at room temperature. Moreover, in Fig. 6, the metal ion-dependent shift in pK, is opposite to the pH dependent loss of metal ion from the enzyme in aqueous solution (19, 24).

Spectral Characterization of the Acyl-enzyme Intermediate

The uitraviolet difference spectrum of carboxypeptidase A, at subzero temperatures uersus carboxypeptidase A,, at am- bient room temperature remains invariant to at least -60°C. Addition of ClCPL causes a change in the difference spectrum attributable to accumulation of a covalent intermediate of the reaction (14). The difference spectra observed under these conditions for ZnCPA and CoCPA are illustrated in Fig. 7. The spectrum is not altered when L-benzylsuccinate is added after substrate, and addition of the inhibitor prior to addition of substrate prevents formation of the spectrally identified species. Over the pH range 6.0 to 9.0, the difference spectrum is invariant for both ZnCPA and CoCPA and cannot be accounted for by the difference spectrum of products uersus the substrate. Furthermore, under conditions of enzyme in excess near -4O”C, the slow phase of the biphasic decrease in absorption due to hydrolysis of ClCPL is not retarded by addition of L-benzylsuccinate. These observations indicate that a complex of the substrate with the enzyme is formed which must have a dissociation equilibrium constant less than the Kl - lo-’ M of L-benzylsuccinate (17). Noncovalent inter- actions of the substrate with carboxypeptidase A, are unlikely to account for these observations. The spectrally identified species for ZnCPA and CoCPA, therefore, are best interpreted as resulting from accumulation of a covalent intermediate of the enzyme reaction formed prior to the rate-limiting step governed by k,. We have previously outlined the evidence indicating that the covalent intermediate is formed by acyla-

tion of the y-carboxylate oxygen of glutamate 270, yielding the highly reactive mixed anhydride acyl-enzyme intermediate illustrated in Structure I (14). The chemical bonding structure of the intermediate has been confirmed by nucleophilic trap- ping studies with the use of hydroxylamine.’

As indicated above, extrapolation of the linear portion of the slow process of kinetic data for CoCPA, as in Fig. 4, yields a value of (A0 - A,) consistently higher than that predicted by the initial substrate concentration. This is equivalent to approximately 80% of the extinction coefficient of the covalent intermediate of CoCPA at 310 nm. Similarly, for ZnCPA an extinction coefficient of the covalent intermediate of approx- imately 70% of that predicted by the difference spectrum could be estimated from kinetic data (14). We conclude that in kinetic studies with CoCPA at subzero temperatures, the initial fast process has occurred largely within the mixing time and only the breakdown of the intermediate species identified in Fig. 7 is observed.

DISCUSSION

The Reaction Scheme of Carboxypeptidase A in Ester Hydrolysis-Data presented in Figs. 3-5 and 7 indicate that two kinetically discrete reactions with observed rate constants kf and k, are correlated with formation and breakdown of a covalent acyl-enzyme intermediate in the subzero temperature range. The basis for the analysis of the hydrolysis of ClCPL catalyzed by carboxypeptidase A, as two consecutive first order reactions has been outlined (14). The analysis leads to a straightforward mathematical relationship between the time-dependent change in absorbance, the observed rate con- stants, and the concentrations of the substrate, intermediate, and products of the reaction, as given in Equation 2

A - A, = (SO) E.W - +, + (k, EEY - hs EPJ

k, - kf 1 exp(-kft)

(2)

or Equation 3

A - A, = a exp(-k/t) + b exp(-k,t). (3)

In Equation 2, (S&ST represents the initial absorbance at to, and the various E values refer to the molar extinction coeffi- cients of the chemical species of the reaction3 (defined below). Kinetic data such as those in Figs. 3 and 4 can be described by Equation 3 only under the condition that b > 0. This condition assigns the rate-limiting step governed by k, to deacylation of the covalent acyl-enzyme intermediate (14).

The data presented in Fig. 5 demonstrate that the reaction catalyzed by ZnCPA and CoCPA is characterized by the same rate-limiting step in aqueous solution as in cryosolvent mix- tures at subzero temperatures. Thus, the subzero temperature data permit deduction of the nature of the rate-limiting step of the enzyme-catalyzed reaction in aqueous solution. For the enzyme reaction, hydrolysis of ClCPL must involve first non- covalent binding to form the Michaelis complex prior to formation of the acyl-enzyme intermediate. The reaction in aqueous solution can be described on this basis by the minimal scheme in Equation 4.

Ed ES5 EY%EEPP,+Pz k-1 +P1

(4) K

E + P, =” EPI

’ E. T. Kaiser, personal communication (1978). 3 In our preliminary communication (14), a transcription error has

appeared in the expression for the coefficient of exp(-k,t) and is hereby corrected.

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

362 Mechanism of Carboxypeptidase A

where EY is the covalent intermediate of the reaction, KP is the dissociation constant of the enzyme complexed with the fist product of hydrolysis, L-P-phenyllactate, and kS governs the rate-limiting step and, therefore, is to be associated with h, in cryosolvent mixtures at subzero temperatures.

The results in Fig. 2 and in Table II indicate an alteration in the rate law for the enzyme reaction in the presence of noncompetitive cryosolvent inhibitors. To account for the overall noncompetitive inhibition of the cryosolvent mixtures employed, we modify the reaction scheme in Equation 4. The appropriate series of consecutive reactions is illustrated in Scheme I, and the corresponding rate equation is provided under “Appendix.”

In Fig. 5, data for kf for the reaction catalyzed by ZnCPA in the subzero temperature range extrapolate to the value of &at observed at 25°C in the same cryosolvent mixture. We have previously suggested (14) that the fast process represents rate- limiting acylation above --10°C both in cryosolvent mixtures and in aqueous solution, but that at lower temperatures deacylation becomes rate-determining. This suggestion must be modified in view of the more complete data presented here which define the interaction of cryosolvents with the enzyme during substrate hydrolysis. In aqueous solution, deacylation governed by K3 is rate-limiting. The rate law describing the enzyme reaction in cryosolvent mixtures in the 25-0°C range is altered from that in aqueous solution and contains a com- posite rate constant which includes the equilibrium governed by Ka’. Under conditions of noncompetitive inhibition, en- zyme with bound cryosolvent represented as EQ can further bind substrate to form the noncovalent complex ESQ. This complex is in equilibrium with the Michaelis complex ES, and further catalytic steps of the reaction proceed only along the pathway normally observed in the absence of noncompetitive inhibition. The change in the rate law simply arises as a result of binding of noncompetitive cryosolvent inhibitor, altering the availability of ES for further reaction.

A comparable reaction scheme must also obtain in the subzero temperature range with respect to the interaction of cryosolvent inhibitors with the enzyme. However, to account for the pH dependence of h,, an equilibrium involving ioniza- tion of a residue on the mixed anhydride intermediate prior to rate-limiting deacylation must be included in the reaction scheme. The corresponding sequence of reaction is illustrated in Scheme II, and the appropriate rate equation is again provided under “Appendix.”

Evaluation of the interaction of cryosolvents with carboxy- peptidase A, during the hydrolysis of ClCPL is comparable to the analysis (26) of the effect of inhibitors on the kinetic parameters of an enzyme reaction that proceeds via an acyl- enzyme intermediate. In the presence of cryosolvent inhibi- tors, the enzyme reaction is altered only by formation of dead- end complexes. In the case of the hydrolysis of ClCPL by

EQ’ IF

II KOS KO

EQ w E

a k,(S) S

ESQ KO

/ ES

EY

SCHEME I

EHQ’

it Ko, Ko

EHQ -

o k,(S)

“0 EHSC’ -

EHY

SCHEME II

carboxypeptidase A,, complexing of the Michaelis complex with cryosolvents only alters its availability in the reaction pathway. Therefore, the important result of this analysis is that the catalytic steps of the reaction of ClCPL with carbox- ypeptidase A, responsible for the chemical transformation of substrate to products are identical in aqueous solution and in cryosolvent mixtures.

Identification of the Ionizing Group in the Acyl-enzyme Intermediate-The pH dependence of k, provides evidence for ionization of a group associated with deacylation of the covalent acyl-enzyme intermediate in the subzero temperature range. Douzou and co-workers (27, 28) have previously ob- served that the enthalpies of ionization of amino acid residues in cryosolvent mixtures are dependent upon solvent compo- sition and that the compensation temperature of an ionizing residue remains the only identifiable criterion for assignment. Therefore, the parallel behavior of the pH dependence of k, and the AHi,, for both ZnCPA and CoCPA independent of solvent composition indicates that the environment of the ionizing group must be very nearly identical for both enzymes. This behavior could occur only for a group buried in the interior region of the acyl-enzyme intermediate inaccessible to bulk solvent.

The metal ion-dependent shift in pK, indicates that this group must be in the immediate environment of the metal ion. The decrease in pK, upon substitution of the open shell Co’+ cation for the closed shell Zn2+ cation is consistent with the change expected for ionization of a ligand directly coordinated to the metal. The amino acid side chains of histidine 196, histidine 69, and glutamate 72 that coordinate the metal ion are sterically inaccessible for direct participation in the cata- lytic process (1, 2). Also, the pK, of ionization of the pyrrole hydrogen of a metal-coordinated imidazole group is too high (29) to account for the pH dependence of k,. On this basis, the only residue that could be responsible for the pH dependence of k, is the metal-coordinated water molecule. The estimates for AHion of 6.2 + 0.1 and 7.2 + 0.1 kcal/mol are in good agreement with values of 6.0 to 8.3 kcal/mol observed for ionization of the water ligand in model five-coordinate Co’+ and Zn2+ complexes (30). In contrast, a AHi,” of 0 to 1 kcal/mol would be expected for the catalytic amino acid side chain of glutamate 270 (31). We also note that an estimate of the compensation temperature of the ionizing group based on the combined data in Tables III and IV yields a value of -345 K. This is consistent only with that expected for ionization of a cationic acid (27, 28). The only cationic acid group in the active site region that can be directly involved in the catalytic process is the solvent molecule coordinated to the divalent metal ion with an overall charge of +l.

The temperature dependence of the pK, for ZnCPA in Fig. 6 with an estimated AHi,” of 7.2 f 0.1 kcal/mol extrapolates to a pK, of 6.1 at 25°C for the catalytic residue. Corresponding considerations place the pK, for CoCPA at 4.9. These pK,

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Mechanism of Carboxypeptidase A 363

values for the ionization of a metal-coordinated water mole- cule are low in comparison to that of -10 generally associated with hexaaquo complexes of Zn2+ and Co’+ ions (32). There are, however, several factors suggesting that they are reason- ably expected for the acyl-enzyme intermediate. In complexes of transition metal ions, the redistribution of bonding capacity upon a decrease in coordination number results in stronger metal-ligand bonding interactions. In fact, the infrared v(OH)stretching frequency of hydrates of transition metals, particularly of Zn2+ complexes of low coordination number, indicate enhanced O-H bond polarization and, therefore, an increased ease of proton ionization (33). Furthermore, the electrostatic interactions between the positively charged metal cation and the bound water molecule polarizing the O-H bond will be increased in the acyl-enzyme intermediate. Loss of the negative charge associated with the carboxylate of glutamate 270 in the active site cavity upon formation of the covalent acyl-enzyme intermediate will enhance the electro- static field of the metal cation on the bound water molecule. Conversion of the enzyme active site upon substrate binding from a water-ftied to a hydrophobic region (1,2) is associated with a change in the local environment of the metal ion to one with a low dielectric constant. The decrease in the dielec- tric constant will aid the ionization of the metal-bound water molecule through the electrostatic energy to be gained (34) by removing a proton from the vicinity of the cation. More- over, stronger metal-ligand bonding interactions resulting from ligand-field stabilization effects (35), as reflected through infrared stretching data (36, 37), should lower the pK, of the Co’+ complex in comparison to that of the Zn2+ complex. In support of these arguments, we note that the pK, of -8.0 and 8.7 observed for the ionization of a water molecule in model five-coordinate complexes of Co2+ and Zn2+ (30) are substan- tially lower than the pK, values of 9.8 and 10.2 for their corresponding hexaaquo complexes (32). We, therefore, assign the ionization of the group responsible for the pH dependence of K, in carboxypeptidase A, to the metal-coordinated water molecule.

This conclusion is not in agreement with assignments of ionizing residues associated with the productive binding of substrates to carboxypeptidase A, from steady state kinetic studies. The pH profile of log kcat/Km(appj indicates that ioni- zations of two groups with pK, values of approximately 6.4 and 9.3 are involved in the productive binding of ClCPL (15, 16). Comparable pK, values are observed in the hydrolysis of peptide substrates (24, 38). The ionization with pK, > 9 has been attributed to a tyrosine residue (16, 24, 38) or the ionization of the metal-coordinated water molecule (3, 16,39). The only basis for the latter interpretation appears to be the pK, of -10.2 observed for the hexaaquo-coordinated Zn’+ complex (32). The estimate of AHi,, of -5 kcal/mol for this ionization (38) favors the tyrosine assignment but does not adequately differentiate between the two possible interpreta- tions. We note, however, that this ionization is not sensitive to metal ion substitution in the hydrolysis of peptide and ester substrates (24).

The pK, of -6.4 on the acidic side of the pH profile of log kcat/Km(app) has been assigned (l-3, 16, 24, 39) to glutamate 270 on the basis that the hydrophobic environment of the active site cavity would tend to raise the pK, of a carboxylic acid group. However, the estimate of AHi,” of -6.5 kcal/mol obtained from kinetic studies of peptide hydrolysis (38) clearly rules against the assignment of a carboxylic acid group for this ionization. It is noteworthy that for peptide hydrolysis the pK, detected on the acidic side of the pH profile of log kcat/Km(app) is sensitive to metal ion substitution and is shifted from 6.2 in ZnCPA to -5.0 in CoCPA (24). As indicated above,

pK, values of -6.1 for ZnCPA and -4.9 for CoCPA at 25°C are predicted for the ionizing group from low temperature data. The metal ion-dependent shift in pK, associated with the rate-limiting process in the subzero temperature range together with the AHi,” serve as firm diagnostic criteria for assignment of the ionization to the metal-coordinated water molecule. We, therefore, suggest that the pK, observed on the acidic side of the pH profile of log kcat/Km(appI of carboxypep- tidase A, in substrate hydrolysis under steady state conditions is due to the ionization of the metal-bound water molecule in the catalytic rate-determining step.

The Mechanism of Action of Carboxypeptidase A in Ester Hydrolysis-Results of atomic model building studies suggest that in productive binding the carbonyl oxygen of the suscep- tible peptide bond of the substrate molecule is coordinated to the metal ion as the fourth ligand by displacing the water molecule (1, 2). This configuration is generally considered as characteristic of the mechanism of action of carboxypeptidase A, for both peptide and ester hydrolysis (l-3, 40). The pH dependence of k, indicating ionization of the metal-bound water molecule cannot be rationalized by simple displacement of the solvent molecule by the substrate, and an alternative explanation of the chemical processes responsible for deacyl- ation of the acyl-enzyme intermediate must be made. To this end we have illustrated in Fig. 8 a schematic outline of the chemical pathway for the hydrolysis of the specific ester substrate ClCPL catalyzed by carboxypeptidase A,. The scheme differs from other formulations (l-3, 6, 40, 42) of the mechanism of action of carboxypeptidase A, with respect to the catalytic role of the metal ion and water molecule, based previously only on the results of atomic model building stud- ies. However, the structural basis for specific binding of the terminal carboxyl group of the substrate by arginine 145 and the aromatic side chain in the hydrophobic cavity established by x-ray diffraction studies (1, 2, 7) remains unaltered.

The suggested pathway for the hydrolysis of the ester substrate in Fig. 8 is consistent with (i) the minimal reaction scheme in Equation 4; (ii) the temperature and metal ion dependence of the pK, in Equation 1; and (iii) the evidence for a covalent acyl-enzyme mixed anhydride intermediate (14). Nucleophilic attack on the carbonyl carbon of the sub- strate to form a mixed anhydride intermediate most likely proceeds via formation and collapse of a transient tetrahedral intermediate. Deacylation of the mixed anhydride intermedi- ate with a metal-bound hydroxide group in the active site may then occur by two possible pathways, as suggested by model studies of the metal ion-catalyzed hydrolysis of esters and peptides (8-10). Direct nucleophilic attack by the metal- bound hydroxide group on the carbonyl carbon atom of the substrate may occur with solvolysis of the resultant complex of the metal ion. Alternatively, deacylation may occur via indirect general base catalysis by an intervening water mole- cule hydrogen bonded to the metal-bound hydroxide group. Our data do not distinguish between the two possibilities. Direct nucleophilic attack of the metal-bound hydroxide group is the kinetically preferred pathway in model systems (9). While the deuterium isotope effect on kcat supports the role of glutamate 270 as a nucleophile (23, 43), the influence of deuterium on k,,, as shown in Table II also remains com- patible with the general base catalyzed process of an interven- ing water molecule. For either pathway cleavage of the C-O bond must occur between the substrate carbonyl carbon atom and the oxygen of the y-carboxyl group of glutamate 270, for studies (44) of the incorporation of ‘*O into active site residues of carboxypeptidase A, show that the oxygens of the carboxyl group of glutamate 270 are not exchangeable. Model building studies indicate that binding of the acyl moiety of the sub-

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

364 Mechanism of Carboxypeptidase A

Glu-270 Michaelis Complex

FIG. 8. Schematic illustration of the pathway for the hydrolysis of the specific ester substrate ClCPL catalyzed by carboxypeptidase A, showing the structural relationships of the catalytic residues of the enzyme to the substrate molecule. See text for discussion. Since formation of a metal-hydroxide complex is required for deacylation of the mixed anhydride intermediate, we have indicated for purposes of simplification that the metal-bound water molecule is also retained during formation of the preceding intermediates of the reaction. Our

strate is likely to prevent dissociation of the first product of hydrolysis L-/?-phenyllactate (1,2). We expect, therefore, that this product remains bound in the interior region of the active

site cavity until deacylation of the mixed anhydride interme- diate is complete, consistent (14) with the large positive en- tropy of activation for deacylation.

These results account for the hydrolysis of the specific ester substrate ClCPL by carboxypeptidase A,. There are no defin- itive studies to indicate that the catalytic roles of glutamate 270 and of the metal-bound water molecule in the active site are different in peptide hydrolysis. Recently, Breslow and Wernick (6) have suggested that formation of an acyl-enzyme mixed anhydride intermediate may be characteristic only for the hydrolysis of ClCPL since isotopic exchange data do not indicate directly formation of a covalent intermediate in pep- tide hydrolysis. These investigators, however, point out that their isotopic exchange data would be consistent with forma- tion of a mixed anhydride intermediate if a binding site for water on the enzyme is retained. Since a metal-bound water molecule in the acyl-enzyme intermediate of ester hydrolysis is demonstrated through our studies and anhydride formation

I;; Ip: y2

0

R-&@

HO-CH-CO;

data do not define the role of the metal-bound water molecule during formation of noncovalently bound enzyme-substrate complexes. The metal ion requirement for the binding of ester substrates (41) may be compatible also with outer sphere coordination although displace- ment of the metal-bound water molecule by the carbonyl oxygen atom of the substrate is generally assumed (l-3, 6, 39, 40,42) on the basis of model building studies (1, 2). For purposes of simplification, the reaction scheme is not balanced with respect to protons.

is sterically compatible with peptide hydrolysis (1,2), a metal- bound water molecule could also account for the apparent inconsistencies (6) of the isotope exchange studies. Moreover, difference Fourier studies of sluggishly reactive peptides in- dicate that not all of the solvent molecules in the active site cleft are displaced upon substrate binding f 1, 2, 7). The similarity in the pH profiles of log hcat/I&app) for peptide and ester hydrolysis (16, 24, 38, 45) and the estimate of AHi,” of -6.5 kcal/mol for the pK, of 6.2 in peptide hydrolysis (38), closely overlapping with that estimated in this study for the hydrolysis of CICPL, indicate that a metal-hydroxide species is also formed in peptide hydrolysis.

The primary advantage of the application of cryoenzymo- logic methods (11-13) to the study of enzyme reactions derives from the potential to accumulate, in stoichiometric accounts, and to stabilize, for long periods of time, an intermediate of the reaction for physical characterization. The results of this investigation thus provide firm support for the conclusions that hydrolysis of substrates by carboxypeptidase A, involves at least one covalent enzyme-substrate intermediate formed after the Michaelis complex and, most importantly, that hy-

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Mechanism of Carboxypeptidase A 365

drolysia proceeds via a metal-hydroxide species. These results stand in direct contradistinction to the assessment of the catalytic configuration of substrates bound to carboxypepti- dase A, on the basis of x-ray diffraction studies (1, 2, 7). To date, three-dimensional structural evidence in support of pro- posed mechanisms of enzyme action has been obtained on the basis of x-ray diffraction studies of crystalline complexes of enzymes with inhibitors or with sluggishly reactive substrates. The nonproductive spatial relationships between the enzyme and pseudo-substrates in such cases are responsible for the long term stability of the complexes. X-ray evidence in support of proposed mechanisms is for this reason of an indirect nature. In the specific case of carboxypeptidase A,, difference Fourier syntheses of the sluggishly reactive peptide substrates Gly-L-Tyr (1, 2) and Phe-Gly-Phe-Gly (7) indicate that a metal-bound hydroxide group is not sterically compatible with coordination of the carbonyl oxygen of the susceptible peptide bond. Since ester hydrolysis catalyzed by carboxypeptidase A, proceeds via a metal-hydroxide species, we conclude that interaction of the metal ion coordinating directly the carbonyl oxygen cannot be a catalytically productive configuration in ester hydrolysis.

Cryosolvents as Inhibitors in Low Temperature Studies-In the application of cryoenzymologic methods, it is necessary to demonstrate that the catalytic pathway in cry- osolvent mixtures remains unchanged from that in aqueous medium. This requirement has been implicit since the intro- duction of cryoenzymology (46). However, except for the obvious case of alcoholic cryosolvents competing for water in the enzyme reaction, the mechanistic implications of the influence of cryosolvents by complexing to intermediates of enzyme reactions have not been assessed quantitatively. Thus, the results of these studies outlining the differential action of two cryosolvents as inhibitors of carboxypeptidase A, under- line emphatically the importance of evaluating solvent effects.

A variety of hydrolytic enzymes that have been investigated in the subzero temperature range exhibit minimal reaction schemes comparable to that in Equation 4 with at least one covalent intermediate (11-13). In numerous cases the cryosol- vent was noted to act as an inhibitor of the enzyme reaction (12). Therefore, the analysis presented here suggests that equilibria involving formation of complexes of catalytic inter- mediates with cryosolvent molecules also may obtain for those enzymes. Our observations on the influence of cryosolvents on carboxypeptidase A,, therefore, introduce a note of caution into the application of cryosolvent methods. For instance, heating-cooling cycles are often applied to effect transforma- tion of intermediates of the enzyme reaction (11). Such tem- perature changes will influence equilibria governing the for- mation of dead-end complexes or may result in formation of complexes with bound cryosolvent that enter into parallel catalytic reaction pathways. The assessment that spectral changes correspond to formation of true productive interme- diates of the natural catalytic pathway consequently rests on the rigorous application of kinetic criteria. The importance of application of such criteria in cryoenzymologic studies cannot be overemphasized.

APPENDIX

The Rate Eauation for Scheme I-For Scheme I the rate equation has the form ’

(Ed

-[

(k-1 + kz)(l + QKa)(l + Q’KB,) UO k&z

k3

(Al)

where Q and Q’ represent the concentration of noncompetitive and competitive cryosolvent inhibitor molecules, k’s represent microscopic rate constants, and K’s the association equilib- rium constants. The rate equation is derived for the condition Q and Q’ >> ( Eo) and independent binding of inhibitors Q and Q’ to the enzyme. The catalytic rate constant for Scheme I is then given by the relation

1 -= kLt

(1 + QKQ’) + 1 kz kz

L42)

Comparison of Equation A2 to the catalytic rate constant for the reaction in aqueous solution (Equation 4)

for which kCat = kS under the condition kZ > kB, indicates that binding of the Michaelis complex ES by a noncompetitive cryosolvent could alter the rate law for the reaction in the 25-O% temperature range.

Since the pH dependence of k’,,, has not been evaluated, ionizations associated with the rate constants kz and k3 have been omitted for purposes of simplification. For Scheme I, a pH dependence of k3 is implicit from the discussion for the derivation of the rate equation for Scheme II given below.

The Rate Equation for Scheme II-For Scheme II the rate equation is given by Equation A3

(Ed

-[

(k-1 + kz)(l + QK&(l + Q’&) 1

UO k,kz l- (S)

where the appropriate rate constants and association equilib- rium constants are indicated respectively by k’s and K’s. The catalytic rate constant is then given by the relation

1 (1 + Q&3 -= k&t

k2 + [“;~“];+&. (A4)

Further simplification of Equation A4 is made by evaluating the relative magnitudes of the component terms. In Fig. 6, the dependence of k, on pH indicates that k-H >> k3. Therefore, the predominant term in Equation A4 is the part containing l/k3. Then

645)

The relationship of Equation A5 to Equation 1 is immediately recognizable. Scheme II thus accounts for the kinetic behavior of the reaction of ClCPL with carboxypeptidase A, and the pH dependence of k, in the subzero temperature range.

On the other hand, binding of noncompetitive cryosolvent inhibitor by EHY and EY to form the dead-end complexes EHYQ and EYQ (governed by the association equilibrium constants RbY and KQ’ leads to the following relation for the catalytic rate constant

k” = kg KG Cat (1 + QKay) K

n + (H+ Cl+ Q&*)

+l + Q&‘) 1 The noncompetitive behavior of cryosolvents and the pH dependence of k, were identical in all cryosolvent mixtures studied, indicating that differential binding affinities of cry- osolvents to EHY and EY are unlikely, and, therefore, KQy = KQY. Under these conditions, the term within the brackets in the above equation containing K, becomes equivalent to that in Equation A5. Furthermore, under conditions of enzyme

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

366 Mechanism of Carboxypeptidase A

in excess, the values of k,, estimated as observed pseudo-first order rate constants, are independent of solvent composition, as shown in Fig. 5. These data, thus, indicate that the denom- inator term (1 + QKoY) = 1 and that at most only very weak binding of cryosolvent inhibitor by EHY and EY can obtain. We conclude, therefore, that Scheme II is the simplest mini- mum scheme which outlines the sequence of reactions com- patible with the kinetic data.

The analysis of Laidler and Bunting (26) suggests that binding of a competitive inhibitor could also occur with EY forming a complex which collapses to products by a reaction parallel to that governed by hy in Scheme II. This condition implies methanolysis in the case of CoCPA. Transesterifica- tion would result in little change in the absorbance properties of the p-chlorocinnamoyl moiety at 310 nm from those of ClCPL. Since comparable changes in absorbance at 310 nm are observed for substrate hydrolysis catalyzed by both ZnCPA and CoCPA in cryosolvent mixtures, we conclude that methanolysis is not catalyzed by CoCPA. Moreover, meth- anolysis has not been detected for ZnCPA (47).

To derive appropriate expressions for the rate equations of Schemes I and II, we have considered only irreversibility of the enzyme reaction in transformation of the Michaelis com- plex EHS to EHY. Cleland (40) has pointed out that the reaction catalyzed by carboxypeptidase A, (Equation 4), in which PI is released by the enzyme only after deacylation, should be regarded as a sequential hydrolysis by water rather than as a double displacement mechanism. It can be shown that the requirement of irreversibility in formation of the acyl- enzyme intermediate leads to identical rate equations for Schemes I and II when considered either as a double displace- ment or as a sequential mechanism.

Acknowledgments-We wish to acknowledge helpful discussions with Professors J. L. Westley, E. T. Kaiser, and F. J. Kezdy. We thank Dr. A. K. Churg for assistance in the use of the Marquardt algorithm, Miss J. M. Quinn for atomic absorption measurements, and Dr. J. Ito for the use of an atomic absorption spectrometer. Estimates of the limiting value of k, were carried out with Dr. K. Yamamura.

REFERENCES 36.

1. Hartsuck, J. A., and Lipscomb, W. N. (1971) in The Enzymes (Boyer, P. D., ed), Vol. 3, pp. l-56, Academic Press, New York

2. Quiocho, F. A., and Lipscomb, W. N. (1971) Ado Protein Chem. 25, 1-78

37.

3. Kaiser, E. T., and Kaiser, B. L. (1972) Accts. Chem. Res. 5, 219-224

38. 39.

4. Makinen, M. W. (1975) in Techniques and Topics in Bioinor- ganic Chemistry (McAuliffe, C. A., ed) pp. l-106, Macmillan, London

5. Vallee, B. L., Riordan, J. F., Bethune, J. L., Coombs, T. L., Auld, D. S., and Sokolovsky, M. (1968) Biochemistry 7,3547-3556

6. Breslow, R., and Wernick, D. L. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 1303-1307

40. 41. 42. 43.

44. 45.

46.

7. Quiocho, F. A., Bethge, P. H., Lipscomb, W. N., Studebaker, T. F.. Brown. R. D.. and Koenie. S. H. (1972) Cold Spring Harbor Symp. Quhnt. Biol. 36,561-$67 - -

8. Buckingham, D. A., and Engelhardt, L. M. (1975) J. Am. Chem. sot. 97,5915-5917

9. Buckingham, D. A. (1977) in Biological Aspects of Inorganic

10.

11. 12.

13. 14.

15. 16. 17.

18.

19. 20.

21. 22. 23.

24. 25. 26.

27. 28.

29.

30. 31.

32.

33.

34. 35.

47.

Chemistry (Dolphin, D., ed) pp. 141-196, John Wiley and Sons, Inc., New York

Wells, M. A., and Bruice, T. C. (1977) J. Am. Chem. Sot. 99, 5341-5356

Douzou, P. (1977) Cryobiochemistry, Academic Press, New York Makinen, M. W., and Fink, A. L. (1977) Annu. Rev. Biophys.

Bioeng. 6, 301-343 Fink, A. L. (1977) Accts. Chem. Res. 10, 233-239 Makinen, M. W., Yamamura, K., and Kaiser, E. T. (1976) Proc.

Natl. Acad. Sci. U. S. A. 73, 3882-3886 Suh. J. (1975) Ph.D. thesis. The Universitv of Chicano Sub; J., and Kaiser, E. T. (1976) J. Am. &em. Sot. 9%, 1940-1947 Byers, L. D., and Wolfenden, R. (1973) Biochemistry 12,

2070-2078 Douzou, P., Hui Bon Hoa, G., Maurel, P., and Travers, F. (1976)

in Handbook of Biochemistry and Molecular Biology (Fasman, G. D., ed) 3rd Ed, pp. 520-540, Chemical Rubber Publishing Co., Cleveland, Ohio

Latt, S. A., and Vallee, B. L. (1971) Biochemistry 10, 4263-4270 Churg, A. K., Gibson, G., and Makinen, M. W. (1978) Rev. Sci.

Instrum. 49, 212-214 Balny, C., and Becker, L. (1971) J. Chim. Phys. 1008-1009 Marquardt, D. M. (1963) J. Sot. Indust. Appl. Math. 11,431-441 Kaiser, B. L., and Kaiser, E. T. (1969) Proc. N&l. Acad. Sci. U.

S. A. 64,36-41 Auld, D. S., and VaIIee, B. L. (1970) Biochemistry 9,4352-4359 Folk, J. E., and Gladner, J. A. (1960) J. Biol. Chem. 235,60-63 Laidler, K. J., and Bunting, P. S. (1973) The Chemical Kinetics

of Enzyme Action, Chapt. 3, pp. 89-110, Clarendon Press, Oxford

Maurel, P., and Douzou, P. (1975) J. Biol. Chem. 250,2678-2680 Maurel, P., Hui Bon Hoa, G., and Douzou, P. (1975) J. Biol.

Chem. 250, 1376-1382 Pitner, T. P., Wilson, E. W., Jr., and Martin, R. B. (1972) Inorg.

Chem. 11, 738-742 Wooley, P. (1975) Nature 258, 677-682 Izatt, R. M., and Christensen, J. J. (1970) in Handbook of Bio-

chemistry (Sober, H. A., ed) pp. J-58-J-173, Chemical Rubber Publishing Co., Cleveland, Ohio

Sillen. L. G.. and Martell. A. E. (1964) Stability Constants of Metal-Zon’Complexes, pp. 54-63, The Chemical Society, Lon- don

Zundel, G. (1969) Hydration and Intermolecular Interaction, pp. 69-115, Academic Press, New York

Schwarzenbach, G. (1970) Pure Appl. Chem. 24, 307-334 George, P., and McClure, D. S. (1959) Prog. Inorg. Chem. 1,

381-463 Nakamoto, K. (1968) in Spectroscopy and Structure of Metal

Chelate Compounds (Nakamoto, K., and McCarthy, P. J., eds) pp. 268-275, John Wiley and Sons, Inc., New York

Haigh, J. M., Hancock, R. D., Hulett, L. G., and Thornton, D. A. (1969) J. Mol. Struct. 4, 369-375

Auld, D. S., and VaIIee, B. L. (1971) Biochemistry 10,2892-2897 Kaiser, E. T., Chan, T. W., and Suh, J. (1974) in Metal-Protein

Interactions (Friedman, M., ed) pp. 59-80, Plenum Press, New York

Cleland, W. W. (1977) Adv. Enzymol. 39, 273-387 Auld, D. S., and Holmquist, B. (1974) Biochemistry 13,4355-4361 Mock, W. L. (1975) Bioorg. Chem. 4, 270-278 Fife, T. H., and Squillacote, V. L. (1977) J. Am. Chem. Sot. 99,

3762-3769 Nau, H., and Riordan, J. F. (1975) Biochemistry 14, 5285-5294 Bunting, J. W., Murphy, J., Myers, C. D., and Cross, G. G. (1974)

Can. J. Chem. 52,2648-2659 Douzou, P., Sireix, R., and Travers, F. (1970) Proc. Natl. Acad.

Sci. U. S. A. 66, 787-792 Hall, P. L., and Kaiser, E. T. (1967) Biochem. Biophys. Res.

Commun. 29,205-210

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from

M W Makinen, L C Kuo, J J Dymowski and S JafferCatalytic role of the metal ion of carboxypeptidase A in ester hydrolysis.

1979, 254:356-366.J. Biol. Chem. 

  http://www.jbc.org/content/254/2/356Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/254/2/356.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on October 14, 2018

http://ww

w.jbc.org/

Dow

nloaded from