Inhibition effects in the hydrolysis reactions of esters and peptides catalyzed by carboxypeptidase A: An example of cooperative binding effects with a monomeric enzyme

Article abstract of DOI:10.1006/bioo.2000.1161

N-benzoyl-L-phenylalanyl-L-phenylalanine is an excellent peptide substrate for carboxypeptidase A; at 30°C and pH 7.5, K_m is 2.6 × 10^-5 M while κ_cat is 177 s^-1 κ_cat/K_m = 6.8 × 10⁶ M^-1 s^-1). Indole-3-acetic acid is a noncompetitive or mixed inhibitor towards the peptide and toward hippuryl-L-phenylalanine; plots of E/V vs [Inhibitor] are linear. N-Benzoyl-L-phenylalanine is a competitive inhibitor of peptide hydrolysis, and plots of E/V vs [Inhibitor] are again linear. One molecule of inhibitor binds per active site, and these inhibitors bind in different sites. At constant peptide substrate concentration and a series of constant concentrations of indole-3-acetic acid, plots of E/V vs the concentration of N-benzoyl-L-phenylalanine are linear and intersect behind the E/V axis and above the [Inhibitor] axis. This shows that both inhibitors can bind simultaneously and that binding of one facilitates the binding of the other (β = 0.18). Employing the ester substrate hippuryl-DL,βphenyllactate, the same type of behavior is observed in the reverse sense; N-benzoyl-L-phenylalanine is a linear noncompetitive inhibitor and indole-3-acetic acid is a linear competitive inhibitor. Again the two inhibitor plot is linear and intersects above the [Inhibitor] axis (β = 0.12). Previous X-ray crystallographic studies have indicated that indole-3-acetic acid binds in the hydrophobic pocket of the S₁ site, while N-benzoyl-L-phenylalanine binds in the S₁-S₂ site. The product complex for hydrolysis of N-benzoyl-L-phenylalanyl-L-Phenylalanine (phenylalanine + N-benzoyl-L-phenylalanine) occupies both of these sites. However, the present work shows that the peptide substrate does not bind to the enzyme at pH 7.5 so as to be competitive with indole-3-acetic acid. The binding sites may be formed via conformational changes induced or stabilized by substrate and product binding.

Full text of DOI:10.1006/bioo.2000.1161

Bioorganic Chemistry 28, 226–241 (2000)
doi:10.1006/bioo.2000.1161, available online at http://www.idealibrary.com on
Inhibition Effects in the Hydrolysis Reactions of Esters
and Peptides Catalyzed by Carboxypeptidase A: An Example
of Cooperative Binding Effects with a Monomeric Enzyme
Abdulkader Rahmo and Thomas H. Fife
Department of Biochemistry, University of Southern California, Los Angeles, California
Received September 1, 1999
N-benzoyl-L-phenylalanyl-L-phenylalanine is an excellent peptide substrate for carboxy-
−
5
−1
peptidase A; at 30°C and pH 7.5, K
1
m
is 2.6 × 10 M while kcat is 177 s (kcat/K
m
= 6.8 ×
6
−1 −1
0 M s ). Indole-3-acetic acid is a noncompetitive or mixed inhibitor towards the pep-
tide and toward hippuryl-L-phenylalanine; plots of E/V vs [Inhibitor] are linear. N-Benzoyl-
L-phenylalanine is a competitive inhibitor of peptide hydrolysis, and plots of E/V vs
[
Inhibitor] are again linear. One molecule of inhibitor binds per active site, and these
inhibitors bind in different sites. At constant peptide substrate concentration and a series of
constant concentrations of indole-3-acetic acid, plots of E/V vs the concentration of N-
benzoyl-L-phenylalanine are linear and intersect behind the E/V axis and above the [Inhibitor]
axis. This shows that both inhibitors can bind simultaneously and that binding of one facil-
itates the binding of the other (β = 0.18). Employing the ester substrate hippuryl-DL,β-
phenyllactate, the same type of behavior is observed in the reverse sense; N-benzoyl-L-
phenylalanine is a linear noncompetitive inhibitor and indole-3-acetic acid is a linear compet-
itive inhibitor. Again the two inhibitor plot is linear and intersects above the [Inhibitor] axis
(
β = 0.12). Previous X-ray crystallographic studies have indicated that indole-3-acetic acid
binds in the hydrophobic pocket of the S
′ site, while N-benzoyl-L-phenylalanine binds in the
–S site. The product complex for hydrolysis of N-benzoyl-L-phenylalanyl-L-phenylala-
1
S
1
2
nine (phenylalanine + N-benzoyl-L-phenylalanine) occupies both of these sites. However, the
present work shows that the peptide substrate does not bind to the enzyme at pH 7.5 so as to
be competitive with indole-3-acetic acid. The binding sites may be formed via conformation-
al changes induced or stabilized by substrate and product binding.
© 2000 Academic Press
INTRODUCTION
1
Carboxypeptidase A (peptidyl L-amino acid hydrolase (EC 3.4.17.1) (CPA) is a
Zn(II) metalloenzyme that catalyzes the hydrolysis of C-terminal peptides and O-acyl
derivatives of β-phenyllactic acid and mandelic acid (1–3). The enzyme is composed of
a single chain of 307 amino acids (3). X-ray crystallographic analysis at 2 Å resolution
showed that the poor peptide substrate glycyltyrosine is complexed to Zn(II) through the
1
Abbreviations used: CPA, bovine pancreatic carboxypeptidase A; BPP, N-benzoyl-L-phenylalanyl-L-
phenylalanine; BP, N-benzoyl-L-phenylalanine; IAA, indole-3-acetic acid; HPL, hippuryl-DL, β-phenyllactic
acid; HPA, hippuryl-L-phenylalanine; PPA, N-(phenoxycarbonyl)-L-phenylalanine; Tris, tris-hydroxy-
methylaminomethane.
2
26
0
045-2068/00 $35.00
Copyright © 2000 by Academic Press
All rights of reproduction in any form reserved.

2
27
COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
carbonyl oxygen (3,4). The γ-carboxyl group of Glu-270 is also located in close prox-
imity. On this basis, nucleophilic and general base mechanisms involving Glu-270 were
postulated for peptide hydrolysis, and it was assumed that similar or identical mecha-
nisms would occur with ester substrates. Christianson and Lipscomb (5) have suggest-
ed that the carbonyl complexation of glycyltyrosine may be anomalous and that the pep-
tide carbonyl is normally polarized by hydrogen bonding from Arg-127.
Christianson and Lipscomb (5,6) determined that N-benzoyl-L-phenylalanine (BP)
binds in the S –S subsite of CPA with the carboxyl group complexed with the metal
1
2
ion. In the same complex L-phenylalanine binds in the hydrophobic region of the S1′
subsite with the carboxylate anion electrostatically linked to Arg-145. They consid-
ered that I represented the productive complex for the synthesis of N-benzoyl-L-
phenylalanyl-L-phenylalanine (BPP). If that is the case, then by the principle of micro-
scopic reversibility I should also resemble the productive complex for the hydrolysis
of BPP since the transition states for the two reactions must be identical.
There is strong evidence that peptides and esters bind initially to CPA in different
sites (7,8). Indole-3-acetic acid (IAA) is a competitive inhibitor toward ester sub-
strates but is noncompetitive toward peptides (7). The carbamate ester N-(phenoxy-
carbonyl)-L-phenylalanine is a linear competitive inhibitor toward the peptide hip-
puryl-L-phenylalanine but is a linear noncompetitive inhibitor toward the ester
hippuryl-L,β-phenyllactate (8). The carbamate ester is also an observable substrate for
−6
the enzyme at high enzyme concentration (10 M), and K = K . The carbamate is
m
i
therefore binding in the catalytic site for peptides.

2
28
RAHMO AND FIFE
The above findings do not rule out similar mechanisms for peptides and esters (9).
Furthermore, the catalytic site could be formed by a conformational change induced
or stabilized by substrate binding. Galdes et al. (10) and Geoghagen et al. (11) have
found that at least two intermediates occur before the rate-determining step (ES s
1
ES ) in the CPA-catalyzed reactions of both peptides and esters. The rate-limiting step
2
in peptide hydrolysis occurs before the formation of an enzyme-product complex (12).
It is clear that insight into the enzyme mechanism is dependent on a deeper under-
standing of the sites in which peptides and esters bind and of the interrelationships
between these sites. We have in the present work studied the CPA catalyzed hydroly-
sis of the large peptide substrate N-benzoyl-L-phenylalanyl-L-phenylalanine (BPP) in
the presence and absence of the inhibitors IAA and BP. X-ray crystallographic studies
have indicated that these inhibitors bind in the S ′ and S –S sites, respectively (6,13).
1
1
2

COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
229
We have also employed these inhibitors together in the reactions of hippuryl-DL,β-
phenyllactic acid (HPL), a typical ester substrate for CPA that has been extensively
studied in previous investigations (3,7,8).
Multiple inhibition analysis can allow the determination of whether binding of the
inhibitors to the enzyme is mutually exclusive or whether both can bind simultane-
ously (14). If both can bind simultaneously, then it can be determined whether the
binding of one hinders or facilitates the binding of the other. As a consequence, con-
siderable information can be obtained pertaining to the nature of the binding sites and
2
their possible interactions. We have found that IAA and BP can bind simultaneously
and that their binding is synergistic, even though CPA is a monomeric enzyme.
EXPERIMENTAL SECTION
N-Benzoyl-L-phenylalanyl-L-phenylalanine (BPP) was prepared by benzoylation of
the dipeptide L-Phe-L-Phe (Sigma) by employing the method of Auld and Vallee (16).
L-Phe-L-Phe (1 g) was added to 10 ml of water and 1 ml of 10 N NaOH. To this solu-
tion was added at 0°C an equivalent amount of benzoyl chloride (Aldrich) in small
portions with vigorous stirring over a period of 10 min. The solution was acidified
with 6 N HCl. The white precipitate was washed with water. After recrystallization
from water-ethanol and vacuum dessication, the crystalline compound melted at
1
85–186°C. Anal. Calcd for C H N O : C, 72.10; H, 5.81; N, 6.73. Found: C, 72.19;
25
24
2
4
H, 5.79; N, 6.56.
Indole-3-acetic acid was purchased from Aldrich. After recrystallization from ethyl
acetate-hexane it melted at 166–167°C. N-Benzoyl-L-phenylalanine, hippuryl-L-
phenylalanine, and hippuryl-DL,β-phenyllactate sodium salt were purchased from
Sigma. Ninhydrin reagent solution (ninhydrin, 20 g/l; hydrindantin, 3 g/l; DMSO,
7
5%; lithium acetate, 1 M) was purchased from Sigma and stored at 4°C under nitro-
gen. This assay solution must be stirred gently before its application to the amino acid
solutions. All buffer components were reagent grade materials, and deionized water
was used throughout.
Carboxypeptidase A from Sigma was dialyzed in 0.05 M Tris-HCl buffer (µ =
2
+
−4
0
.5 M with NaCl, [Zn ] = 10 M), pH 7.5, at 5°C for 36 h. The buffer solution was
changed after 18 h. After dialysis the clear enzyme solution was centrifuged at
5,000 rpm for 15 min at 5°C. The supernatant was filtered through Millipore
Millex” filters and stored at 5°C. The protein concentration was determined by
1
“
5
−1
−1
employing the extinction coefficient at 222.5 nm (⑀ = 5.27 × 10 M cm ) and 278 nm
⑀ = 6.42 × 10 M cm ) (17).
4
−1
−1
(
Kinetic methods. The hydrolysis of BPP cannot be followed spectrophotometrically
due to insufficient absorbance changes. Therefore, the hydrolysis reactions were followed
using a ninhydrin assay that detects the release of the primary amino group of phenylala-
nine (18). In the initial rate measurements the reaction conditions were chosen so that less
than 10% of the substrate was hydrolyzed. Absorbance measurements were carried out
using a Zeiss Model PM2DL spectrophotometer. The background absorbance due to the
2
An example is the multiple inhibitor analysis that was carried out in reactions of D-glyceraldehyde-3-
phosphate dehydrogenase (15).

2
30
RAHMO AND FIFE
−9
enzyme was negligible at the enzyme concentration employed (1.4–6 × 10 M). The reac-
tion measurements were carried out in duplicate. Control duplicates that did not include
CPA gave the background absorbance not due to the enzyme. Under the assay conditions
the inhibitors BP and IAA do not react with ninhydrin to a significant extent. At pH 7.5,
Tris–HCl buffer was employed, which contributes slightly to the background absorbance.
An absorbance range of 0.2–1.0 was generally utilized. The reaction temperature was
−4
maintained constant at 30 ± 0.1°C. The reaction was stopped using a 1 M citrate, 10 M
EDTA, pH 5.2, solution. The ninhydrin assay has been employed previously to monitor
CPA catalyzed peptide hydrolysis in Tris buffer (16).
The CPA catalyzed hydrolysis of the ester substrate hippuryl-DL, β-phenyllactic
acid (HPL) was monitored spectrophotometrically by following the absorbance changes
at 254 nm with a Beckman DU-7500 spectrophotometer. Initial rates were determined.
The presence of the D-isomer has been shown not to affect the kinetics of the CPA cat-
alyzed reaction (17). The concentrations employed were corrected to reflect that of the
L-isomer. The pH measurements were made with a Radiometer PHM-22 pH meter.
The peptidase activity of the enzyme stock solution was routinely checked with HPA
and the esterase activity with HPL.
The values of the kinetic parameters were obtained from computer analysis of plots
of V/E vs [S ], E/V vs 1/[S ], and E/V vs [Inhibitor]. The solid lines in the plots pre-
0
0
sented for illustration represent the best fit. The conventional nomenclature (14) is
employed to describe the inhibition experiments, i.e., competitive, noncompetitive,
and mixed. Noncompetitive inhibition is strictly an effect on k . If both k and the
cat
cat
apparent K are affected, then the inhibition is mixed. Noncompetitive inhibition has
m
K (slope) = K (intercept). In the inhibition experiments in which the substrate con-
i
i
centration was varied, at least two constant concentrations of inhibitor were employed.
RESULTS
The CPA catalyzed hydrolysis of N-benzoyl-L-phenylalanyl-L-phenylalanine (BPP) at
−
1
pH 7.5 and 30°C (µ = 0.5 M with NaCl) is characterized by k = 177 s , and k /K =
6
cat
cat
m
6
−1 −1
.8 × 10 M s . The accuracy of rate measurements with the ninhydrin assay was
less at very low substrate concentrations (less than K ) than with higher substrate con-
m
centrations. The small apparent K of BPP was therefore measured using a suitable
m
competitive inhibitor so that relatively high concentrations of substrate could be con-
veniently employed. N-Benzoyl-L-phenylalanine (BP), a product of the hydrolysis of
BPP, was found to be a competitive inhibitor of the CPA catalyzed hydrolysis of BPP.
The plots of E/V vs 1/[S ] at constant concentrations of BP shown in Fig. 1 intersect on
0
−1
the E/V axis and give the k value of 177 ± 2 s . The plots of E/V vs [BP] (Fig. 2) are
cat
also linear. Correlation coefficients were greater than 0.99 for the plots of Figs. 1 and
2
. Dixon plots of E/V vs [I] intersect at [I] = –K when the inhibition is linear compet-
i
itive. A horizontal line from the point of intersection will then touch the E/V axis at
/k (14). The intersection of the plots of Fig. 2 for the competitive inhibition by BP at
1
cat
high substrate concentrations occurs close to the E/V axis at 1/k because of the required
cat
−
5
scale of the [BP] axis. The average K from Figs. 1 and 2 is 2.6 × 10 M. A K of 9.2 ×
m
i
−5
−5
1
0 M was found for BP. This K value compares well with the value of 8.8 × 10 M
i
previously reported in other reactions (19). The slopes of Fig 2. give K /K = 3.6.
i
m

COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
231
FIG. 1. Plots of E/V vs 1/[S
.002 and 0.004 M BP at 30°C, pH 7.5, and µ = 0.5 M NaCl.
] for the CPA (1.4 × 10−9 M) catalyzed hydrolysis of BPP in the presence of
0
0
FIG. 2. Plots of E/V vs [BP] for the CPA (1.5 × 10−9 M) catalyzed hydrolysis of 1.3 × 10−3 and 6.4 × 10−4
BPP at 30°C, pH 7.5, and µ = 0.5 M NaCl.
M

2
32
RAHMO AND FIFE
Indole-3-acetic acid is a noncompetitive or mixed inhibitor of the CPA catalyzed
hydrolysis of BPP, and again plots of E/V vs [IAA] are linear as shown in Fig. 3.
Varying the constant substrate concentration produced nearly identical lines, as is
required for noncompetitive inhibition at high [S ]/K ratios (see equations III-28 and
0
m
−4
III-29 in (14)). The value of K is 1.5 × 10 M.
i
Indole-3-acetic acid is a fully noncompetitive inhibitor toward hippuryl-L-
phenylalanine (HPA). The plots of E/V vs 1/[S ] at 30°C and pH 7.5 in that case inter-
0
sect on the 1/[S ] axis, and the values of K (slope) and K (intercept) are closely sim-
0
i
i
−
4
−4
ilar, 2.1 × 10 M and 2.3 × 10 M, respectively (α = 1.1). The plots of E/V vs [IAA]
−4
are linear. From the point of intersection on the [IAA] axis, a K of 2.2 × 10 M was
i
calculated for IAA inhibition of the CPA catalyzed hydrolysis of HPA.
Figure 4 is a plot of E/V vs the concentration of BP at a constant concentration of
BPP and a series of constant concentrations of IAA. Equation [1] is applicable to the
case of linear competitive and noncompetitive (or mixed) inhibitors, where I is the
competitive inhibitor and X is noncompetitive. The scheme of Eq. [1] leads to Eq. [2].
I
kcat
EI
E
+
+
S
S
ES
E
+
P
Ki
Ks
+
X
+
X
+
X
[1]
βKx
Kx
αKx
Kl
Kl
Kl
EIX
EX
ESX
αKs
FIG. 3. Plot of E/V vs [IAA] for the CPA (1.5 × 10−9 M) catalyzed hydrolysis of 6.4 × 10−4 M BPP at
0°C, pH 7.5, and µ = 0.5 M NaCl.
3

COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
233
−5
−4
FIG. 4. Plots of E/V vs [BP] at a series of constant concentrations of IAA (0, 6 × 10 , 1 × 10 , and 2 ×
0−4 M) in the CPA (1.5 × 10 M) catalyzed hydrolysis of 6 × 10 M BPP 30°C, pH 7.5, and µ = 0.5 M NaCl.
–9
–4
1
V
Vmax
[ ]
S
=


[
I
]
[
X
]
[
I
] [
X
]
x



[ ]
X
αKx
[2]
K 1 +
+
+
+
[
S
]
1 +
s
Ki Kx βK K 
i
The equation for the Dixon plot of 1/V vs [I ] is given in Eq. [3],
1
V
Ks
VmaxKi


[
X
βKx 
]

1 
[
X
αK 
]
x

K 
[
X
]
 

s
=
1 +
[
I
]
+
1 +
+
1 +
[3]
[4]


[
S
]
Vmax

[
S
]

K 

x
while that of 1/V vs [X] is given in Eq. [4] (14).
1
V
1

αKs 
S
]
[
I
]
 

1
Vmax

Ks 
[
I
]
 

=
1 +
1 +
[
X
]
+
1 +
1 +


αVmaxKx
[

βKi 
[
S
]

Ki 




The plots of Fig. 4 are linear with correlation coefficients of 0.98 or greater, and the slope
increases as the concentration of IAA increases. Therefore, both inhibitors bind to the
enzyme simultaneously (14). The plots intersect behind the E/V axis and above the [BP]
axis, which shows that the binding of one inhibitor facilitates binding of the other. A lin-
ear plot of the slopes of Fig. 4 vs the concentration of IAA gives −βK
X
as the intercept on
−4
the IAA axis. Employing K
intercepts of Fig. 4 vs [IAA] allows the calculation of α; αK
X
= 1.5 × 10 M, β is calculated to be 0.18. A linear plot of the
−4
x
= 1.75 × 10 M.

2
34
RAHMO AND FIFE
The ester substrate hippuryl-DL,β-phenyllactate (HPL), has k and K values of 750
cat
m
−5
−1
−4
s and 1.0 × 10 M, respectively, at 30°C and pH 7.5. A K of 9 × 10 M was pre-
m
viously reported (8). Indole-3-acetic acid is a competitive inhibitor and N-benzoyl-L-
phenylalanine is noncompetitive. Plots of E/V vs 1/[S ] intersect on the E/V and 1/[S ]
0
0
axis, respectively, in the presence of constant concentrations of the inhibitor, as illus-
trated for BP in Fig. 5. In the noncompetitive inhibition by BP the values of K (inter-
i
cept) and K (slope) do not differ greatly (α = 0.9). The plots of E/V vs [Inhibitor] are
i
linear and intersect on the BP axis. In contrast, the linear plots of E/V vs IAA (not
shown) intersect behind the E/V axis and above the IAA axis. A line from the inter-
section point to the E/V axis is close to 1/k . The linear plots of E/V vs 1/[S ] in the
cat
0
−
1
presence of IAA intersect precisely on the E/V axis at 1/k (k = 752 s ). The values
cat
cat
−4
−4
of K determined from the Dixon plots are 2.2 × 10 M for IAA and 1.0 × 10 M for
i
BP. Previously reported values of K for IAA as a competitive inhibitor of ester
i
−4
hydrolysis and as a noncompetitive inhibitor of peptide hydrolysis are 1.6 × 10 M
−4
and 1.7 × 10 M, respectively (7). The linear plots in Fig. 6 of E/V vs the concentra-
tion of BP at a constant concentration of HPL and a series of concentrations of IAA,
intersect behind the E/V axis and above the [BP] axis (β = 0.12). Correlation coeffi-
cients for the straight lines in Fig. 6 were again greater than 0.99.
DISCUSSION
N-Benzoyl-L-phenylalanyl-L-phenylalanine (BPP) is an excellent peptide substrate
−5
for CPA. The K value at pH 7.5 of 2.6 × 10 M is the smallest that has yet been mea-
m
FIG. 5. Plots of E/V vs 1/[S
0
] for the CPA (2.5 × 10−9 M) catalyzed hydrolysis of HPL in the presence of BP;
4
1
× 10− M (᭹); 5 × 10−5 M ( ); and in the absence of inhibitor (᭪) at 30°C, pH 7.5, and µ = 0.5 M NaCl.

COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
235
−4
−4
FIG. 6. Plots of E/V vs [BP] at a series of constant concentrations of IAA (0, 1 × 10 , 1.5 × 10 , and 2 ×
0−4 M) in the CPA (2.5 × 10−9 M) catalyzed hydrolysis of 1 × 10 M HPL at 30°C, pH 7.5, and µ = 0.5 M
−4
1
NaCl.
sured for benzoylated peptide substrates of the enzyme. It is quite unlikely that the
−1
small K is influenced by nonproductive binding since k (177 s ), which would be
m
cat
reduced by nonproductive binding, is enhanced over that of usual dipeptide substrates;
−
1
hippuryl-L-phenylalanine (HPA) has k = 100 s at pH 7.5 and 30°C (8). The value
cat
6
−1 −1
of k /K for BPP (6.8 × 10 M s ) is again the largest yet observed for a benzoyl-
cat
m
ated peptide substrate of the enzyme. Only some dansylated peptides provide larger
values (20). It is probable that these favorable values are produced by the multiplici-
ty of contacts that the substrate can make with the enzyme. In comparison, HPA (N-
−3
5
−1 −1
benzoylglycyl-L-phenylalanine) has a K of 10 M and k /K = 10 M s at pH 7.5.
m
cat
m
An “edge to face” interaction of a benzyl side chain with an aromatic enzyme residue
(
Tyr-198) may contribute to the specificity effects for binding in the S subsite (5,21).
1
Removal of a phenylmethylene group from water to a hydrocarbon environment will
proceed with a ∆G° of −3.6 kcal/mol (22). Consequently, the smaller K for BPP in
m
comparison with HPA can be explained on the basis of the substitution of a phenylalanyl
residue for glycyl. Note that the precision of binding (assuming that K = K , as is the
m
s
case with other peptide substrates (16,23,24)), manifests itself in a relatively favorable
k /K . The latter constant is the second-order rate constant for reaction of the free
cat
m
enzyme with the substrate and is not affected by any nonproductive binding effects.
The discovery of enzyme substrates that give large rate constants can be mechanis-
tically important. In addition to the general base and nucleophilic mechanisms pro-
posed by Lipscomb (3,4), more recent suggestions have involved Glu-270 general

2
36
RAHMO AND FIFE
base abstraction of a proton from a metal ion bound water molecule (II) (5), and the
kinetically equivalent attack of metal ion bound hydroxide ion (III) assisted by Glu-
2
70 general acid catalysis (25).
The latter mechanism is attractive because it should be capable of generating large rate
constants and large rate enhancements (25,26), whereas mechanism II has not been
−
shown to be chemically feasible. Metal ion promoted OH catalysis has been consid-
ered to be a general mechanism for hydrolytic metalloenzymes (27). Kinetically equiv-
alent mechanisms may be distinguished if one of the mechanisms demands a rate con-
8
9
−1 −1
stant that is greater than that for a diffusion controlled reaction (10 –10 M s in an
enzymatic reaction). If the apparent pK values of 6.3 and 9 are assigned to the car-
a
boxyl group of Glu-270 and a metal ion bound water molecule, as suggested
(
1,2,28,29), then the concentration of active site II is 500-fold larger than that of active
3
site III at any pH. Therefore, from the experimental rate constant for BPP at pH 7.5
6
−1 −1
(7 × 10 M s ), the true rate constant for mechanism III is calculated to be 3.5 ×
9
−1 −1
1
0 M s . This value is slightly greater than that usually considered for a diffusion-
controlled enzyme reaction, but does not allow mechanism III to be ruled out.
However, it is clear that the pK of zinc ion bound water cannot be much greater than
a
9
for mechanism III to be possible. On the other hand, a water pK much less than 9,
a
e.g., 6.3, would preclude the need for general base catalysis by Glu-270 (Mechanism
II) because of the high concentration of the ionized species. A mechanism involving
−
attack of metal ion bound OH without involvment of Glu-270 in the rate-determining
step would be consistent with the pH dependence of k in peptide hydrolysis (23) if
cat
4
the water pK is as low as 6.3.
a
3
The ratio of the concentrations of the monoanionic active sites in II and III is given by the ratio of the
dissociation constants for ionization of metal ion bound water and ionization of the γ-carboxyl group of
Glu-270, assuming that these dissociations do not influence each other. Thus:
−
II
K
[COO ][M − H O]
COOH
2
=
II
−
K
MII
–H O [COOH][M − OH
]
2
4
II
The value of the dissociation constant of the Zn −H
2
O of CPA has not been determined unambiguous-
of the aquo complex of Zn(II) is 8.8.
ly; pK
a
values ranging from near 6 to 9 have been suggested. The pK
a

COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
237
Inhibition of CPA. Linear inhibition is observed in the CPA catalyzed hydrolysis of
both peptide and ester substrates. The competitive inhibition by the carbamate ester N-
phenoxycarbonyl-L-phenylalanine (PPA) in the CPA catalyzed hydrolysis of the peptide
hippuryl-L-phenylalanine, gives rise to linear E/V vs [PPA] plots (8). The same type of
behavior is observed with the competitive inhibitor BP when the peptide BPP is the sub-
strate. Indole-3-acetate is a noncompetitive or mixed inhibitor towards peptides, and plots
of E/V vs [IAA] are again linear. The reverse is the case when the substrate is the ester
HPL, IAA is then a linear competitive inhibitor, whereas PPA and BP are linear noncom-
petitive. The linear plots of E/V vs [I] for the CPA catalyzed hydrolysis of constant con-
centrations of BPP, HPA, or HPL, show that one molecule of inhibitor (PPA, BP, or IAA)
binds per active site at pH 7.5.
Linear plots of E/V vs 1/[S ] intersecting on the E/V axis in the presence of an inhibitor
0
indicate competitive inhibition but do not establish that the substrate and inhibitor bind
in the same site on the enzyme. Partially competitive inhibition, in which an EIS com-
plex can break down to products will also produce such plots. Partially competitive inhi-
bition can be identified by hyperbolic plots of 1/V vs [I] at a constant substrate concen-
tration. To establish that substrate and inhibitor are mutually exclusive, it is necessary
for plots of 1/V vs [I] to be linear. Since the plots of E/V vs [I] are linear in the CPA
catalyzed reactions of BPP, HPA, and HPL, an EIS complex is not formed in the com-
petitive cases, and there is direct competition for binding to the active site.
As only one molecule of inhibitor binds per active site at pH 7.5, and the K values
i
are nearly constant regardless of whether the inhibition is competitive or noncompet-
itive (7,8), it might be reasonably inferred that the binding of each inhibitor is to a
single favored site. If it is considered that the inhibitors are binding in the sites indi-
cated by the X-ray analysis, then the linear competitive inhibition of peptide hydroly-
sis by BP in conjunction with the observed binding of BP in the S –S subsite in the
1
2
X-ray crystallographic studies of Christianson and Lipscomb (5,6), indicates that pep-
tide substrates bind in that site. In view of its structural similarity to BP, PPA proba-
bly also binds to the same site (8). L-Phenylalanine binds in the S′ site with the phenyl
1
group in the hydrophobic pocket, as in structure I (3,5). Structure I, the suggested
complex for the synthesis of BPP, would then resemble the productive complex for
the forward hydrolysis reaction. However, the lack of apparent competitive inhibition
of IAA toward BPP and HPA at pH 7.5 poses an interpretive problem.
The X-ray structure of the IAA–CPA complex shows that IAA binds in the
hydrophobic pocket of the S′ subsite with the carboxylate group salt-linked to Arg-
1
1
45 in the same manner as phenylalanine (13). No other binding sites for IAA were
observed at the concentrations employed (0.01 M). The competitive inhibition shown
by IAA toward esters was considered by Auld and Holmquist to indicate the same
binding site for those substrates and IAA (7). A different binding site was indicated
for peptides toward which IAA was noncompetitive. Lipscomb suggested that peptides
initially bind in a remote site and that the rate-determining step is migration into the
S1′ site, i.e., there are different initial binding sites and rate-determining steps for esters
and peptides but the same catalytic site (9). The difficulty of binding peptides in S1′
was ascribed to a twisting effect necessary to achieve binding. However, the large
k /K for BPP and dansylated peptides (10,20) implies that the prevalent ES species
cat
m
is the one in which the substrate is bound efficiently to the catalytic site. Also, rate-

2
38
RAHMO AND FIFE
determining binding of peptides in S′ is not in accord with structure-reactivity con-
1
siderations (8). Thus, the problem of the observed noncompetitive inhibition by IAA
in peptide hydrolysis remains in apparent conflict with the X-ray results.
The pH dependence of pK for IAA noncompetitive inhibition is a gentle sigmoid
i
(30), similar to that of phenylacetate (31). L-phenylalanine also gives a sigmoidal log
K vs pH plot (32). In contrast, the log K vs pH profile for the inhibition toward pep-
i
i
tides of L-benzylsuccinate, which may bind to the metal ion, shows a large pH depen-
dence that is nearly linear from pH 6–10 (33). These results are consistent with the
view that IAA in its noncompetitive mode does not bind to the metal ion.
Carboxylic acids can exhibit different types of inhibition effects (34). The partially
competitive inhibition of HPL hydrolysis by cyclohexyl acetate was interpreted in
terms of an EI complex (35). At high concentrations (greater than 0.01 M), a second
2
binding site can be detected for carboxylate inhibitors of CPA, and the inhibition
toward peptides changes from noncompetitive to mixed or competitive at high pH (> 8)
(31,36). In view of the linear inhibition, binding of IAA in the two sites is mutually
exclusive at pH 7.5. Furthermore, a competitive component is not detected when the
peptide HPA is the substrate at pH 7.5; the IAA inhibition is linear noncompetitive,
and α = 1.1.
Linear mixed inhibition can result from the binding of an inhibitor at two different
but mutually exclusive sites (37). A scheme involving mixed competitive and non-
competitive inhibition, in which there are two exclusive sites for binding of inhibitor,
can give results characteristic of noncompetitive inhibition if coincidentally Ki
(
slope) = K (intercept) (37). However, two preexisting sites for the binding of inhibitor
i
is not consistent with the X-ray results for IAA (13) unless binding in the second site
is very poor.
A scheme involving mixed competitive and uncompetitive inhibition with exclusive
5
binding of inhibitor in two different sites is shown in Eq. [5]. This scheme will also
appear to be pure noncompetitive
KS
kcat
E
+
S
ES
I
E
+
P
γKi
I
Ki
[5]
Kl
Kl
IE
ESI
inhibition if γ = 1. In Eq. [5] the second binding site for inhibitor is created by the binding
of substrate. A scheme of that type can explain the observed noncompetitive inhibition by
IAA in peptide hydrolysis without invoking different binding sites for esters and peptides
or conflicting with the X-ray data. The difficulty in the scheme stems from the requirement
that γ = 1.
L-Phenylalanine might, of course, move into the S′ site after the C–N bond break-
1
ing step. Such an alternative would imply that the X-ray crystal structure depicted in
5
Uncompetitive inhibition has been reported in CPA catalyzed ester hydrolysis reactions (34).

COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
239
I does not represent the initially formed product complex in water. Failure of the C-
terminal phenylalanine residue of BPP to bind in S′ would then imply a rigid active
1
site. Conformationally inflexible ES and EP complexes are not, however, consistent
with the present double inhibitor experiments.
The double inhibitor plots of E/V vs the concentration of one inhibitor at a series of
fixed concentrations of the other inhibitor (Figs. 4 and 6), are linear and intersect,
which shows that both inhibitors can bind simultaneously (14). The value of β influ-
ences the slope of the plot and determines whether the lines intersect. Intersection
indicates the presence of the ternary complex EIX. Moreover, β denotes the effect of
the binding of one inhibitor on the binding of the other (see Eqs. [3] and [4]). A β of
∞
indicates that binding of the inhibitors is mutually exclusive, and the lines are
parallel. β = 1 indicates that binding of one inhibitor has no effect on the binding of
the other (Dixon plots intersect on the [inhibitor] axis), while β < 1 indicates that bind-
ing of one inhibitor promotes the ease of binding of the other. That the plots inter-
sect above the [inhibitor] axis indicates that binding is synergistic, i.e., binding of
one inhibitor favors binding of the other. The value of the inhibitor interaction factor
β was found to be ∼ 0.2 when either the peptide BPP or the ester HPL is the substrate.
Thus, the shape of the plots and the values of β are the same with these structural-
ly different peptide and ester substrates with which different assay methods were
employed. There are at least two distinct but not independent sites for the binding
of the two inhibitors. A conformational change is occurring in CPA catalyzed reac-
tions that is either induced or stabilized by inhibitor binding, and these conformational
effects are communicated between sites.
When HPL is the ester substrate and BP is the noncompetitive inhibitor, the values
of K (intercept) and K (slope) are similar although not identical (α = 0.9). If the K for
i
i
i
IAA acting as a competitive inhibitor vs. HPL is taken as a fixed value, then α in the
IAA inhibited reactions of BPP (K (noncompetitive)/K (competitive)) is 0.68. Thus,
i
i
synergistic effects occur predominantly within the EIX complex. A large substrate
may make sufficient contacts with the enzyme so that the optimum conformation for
binding is achieved with minimum mutual binding effects. Clearly, cooperative bind-
ing effects are highly dependent on the structure of the ligands.
Conformational changes of CPA have been noted previously (3,5), in particular, the
movement of Tyr-248 and the associated peptide chain, when substrates bind to the
enzyme. Such conformational changes may affect peptides and esters differently.
Tyrosyl acetylation reduces kcat for peptides to 3–6% of the kcat for the unmodified
enzyme (16,38), but the ester HPL is still hydrolyzed efficiently. Both kcat and K for
m
HPL hydrolysis are increased by acetylation. Likewise, conformational changes involv-
ing Arg-127 occur that could be mechanistically important with peptide substrates
(5,39). The extended conformation of Arg-127 can bind to the substrate carbonyl oxy-
gen and thereby exert a polarizing effect. Thus, conformational changes can be crucial
in the formation of the catalytic site and are very likely dependent on the ligands.
Cooperative binding of ligands has been frequently observed with enzymes composed
of subunits but is rare in reactions of monomeric enzymes (see (40) and references there-
in for an example). The classical equilibrium explanations for cooperativity (41,42)
cannot be easily applied to monomeric enzymes. The positive or negative cooperativity
effects that often occur with enzymes composed of subunits can be considered to arise

2
40
RAHMO AND FIFE
so that the enzyme can respond to drastic increases or decreases in the substrate con-
centration. However, with a monomeric enzyme such as CPA, intersubsite cooperativi-
ty is very likely a consequence of the large binding site required for the binding of pep-
tides and the necessity of high velocity with different substrates. A conformational
change of a flexible active site induced or stabilized by the binding of substrate would
allow the formation of the catalytic site in the most advantageous manner with a variety
6
of peptide substrates. With a flexible active site a conformational change need not be
energetically prohibitive. Small molecules that can bind simultaneously in different
regions of the active site might then bind cooperatively.
ACKNOWLEDGMENT
This work was supported by a grant from the National Science Foundation.
REFERENCES
1. Kaiser, E. T., and Kaiser, B. L. (1972) Acc. Chem. Res. 5, 219.
2. Hartsuck, J. A., and Lipscomb, W. N. (1971) The Enzymes, 3rd ed., 3, 1.
3. Ludwig, M. L., and Lipscomb, W. N. (1973) in Inorganic Biochemistry (Eichhorn, G. L., Ed.), Vol. 1,
p. 438, American Elsevier, New York.
4
5
6
7
8
9
. Lipscomb, W. N. (1970) Acc. Chem. Res. 3, 81.
. Christianson, D. W., and Lipscomb, W. N. (1989) Acc. Chem. Res. 22, 62.
. Christianson, D. W., and Lipscomb, W. N. (1987) J. Am. Chem. Soc. 109, 5536.
. Auld, D. S., and Holmquist, B. (1974) Biochemistry 13, 4355.
. King, S. W., Lum, V. R., and Fife, T. H. (1987) Biochemistry 26, 2294.
. Lipscomb, W. N. (1980) Proc. Natl. Acad. Sci. USA 77, 3875.
1
1
0. Galdes, A., Auld, D. S., and Vallee, B. L. (1983) Biochemistry 22, 1888.
1. Geoghegen, K. F., Galdes, A., Hanson, G., Holmquist, B., Auld, D. S., and Vallee, B. L. (1986) Biochemistry
2
5, 4669.
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2. Galdes, A., Auld, D. S., and Vallee, B. L. (1986) Biochemistry 25, 646.
3. Rees, D. C., and Lipscomb, W. N. (1981) Proc. Natl. Acad. Sci. USA 78, 5455.
4. Segel, I. H. (1975) in Enzyme Kinetics, Wiley Interscience, New York.
5. DeMark, B. R., Benjamin, B. M., and Fife, T. H. (1983) Arch. Biochem. Biophys. 223, 360.
6. Auld, D. S., and Vallee, B. L. (1970) Biochemistry 9, 602.
7. Whitaker, J. R, Menger, F., and Bender, M. L. (1966) Biochemistry 5, 386.
8. Moore, S. (1968) J. Biol. Chem. 243, 6281.
9. Galardy, R. E., and Kortylewicz, Z. P. (1984) Biochemistry 23, 2083.
0. Williams, A. C., and Auld, D. S. (1986) Biochemistry 25, 94.
1. Burley, S. K., and Petsko, G. A. (1986) J. Am. Chem. Soc. 108, 7995.
2. Fersht, A. (1977) Enzyme Structure and Mechanism, p. 232, W. H. Freeman, San Francisco.
3. Auld, D. S., and Vallee, B. L. (1970) Biochemistry 9, 4352.
4. Auld, D. S., and Vallee, B. L. (1971) Biochemistry 10, 2892.
5. Fife, T. H. (1991) in Perspectives in Bioinorganic Chemistry, (Hay, R. W., Ed.), Vol. 1, pp. 43–93, JAI Press,
London.
2
2
6. Fife, T. H., and Bembi, R. (1993) J. Am. Chem. Soc. 115, 11358.
7. Wilkins, P. C., and Wilkins, R. G. (1997) in Inorganic Chemistry in Biology, p. 57, Oxford Univ. Press,
New York.
2
2
3
8. Auld, D. S., Bertini, I., Donaire, A., Messori, L., and Moratal, J. M. (1992) Biochemistry 31, 3840.
9. Zhang, K., and Auld, D. S. (1993) Biochemistry 32, 13844.
0. Auld, D. S., Larson, K., and Vallee, B. L. (1986) in Zinc Enzymes (Bertini, I., Luchinat, C., Maret, W., and
Zeppezauer, M., Eds.), pp. 133–154, Birkhauser, Boston.
6
CPA does, of course, have an absolute specificity requirement in the free C-terminal carboxyl group,
but numerous structurally different peptides and esters will serve as substrates.

COOPERATIVE BINDING EFFECTS WITH A MONOMERIC ENZYME
241
3
3
3
3
3
3
1. Auld, D. S., Latt, S. A., and Vallee, B. L. (1972) Biochemistry 11, 4994.
2. Ref 30, p. 150.
3. Byers, L. D., and Wolfenden, R. (1973) Biochemistry 12, 2070.
4. Bunting, J. W., and Myers, C. D. (1975) Can. J. Chem. 53, 1984, 1993.
5. Bunting, J. W., and Myers, C. D. (1974) Biochim. Biophys. Acta 341, 222.
6. Bertini, I., Monnanni, R., Pellacani, G. C., Sola, M., Vallee, B. L., and Auld, D. S. (1988) J. Inorg. Biochem.
2, 13.
3
3
3
3
4
4
4
7. Ibid (14), pp. 176–178.
8. Davies, R. C., Riordan, J. F., Auld, D. S., and Vallee, B. L. (1968) Biochemistry 7, 1090.
9. Kim, H., and Lipscomb, W. N. (1991) Biochemistry 30, 8171.
0. Cardenas, M. L., Rabajlle, E., and Niemeyer, H. (1984) Eur. J. Biochem. 145, 163.
1. Monod, J., Wyman, J., and Changeux, J. P. (1965) J. Mol. Biol. 12, 88.
2. Koshland, D. E., Jr., Nemethy, G., and Filmer, D. (1966) Biochemistry 5, 365.