Esterase-Like Activity of Serum Albumin: Characterization of Its Structural Chemistry Using p-Nitrophenyl Esters as Substrates

Article abstract of DOI:10.1023/B:PHAM.0000016241.84630.06

Purpose. To elucidate the catalytic mechanism of the esterase-like activity of serum albumin (SA). the reactivity of SA from six species was investigated using p-nitrophenyl esters as model substrates. Methods. The effect of pH and the energetic and therm

Full text of DOI:10.1023/B:PHAM.0000016241.84630.06

Pharmaceutical Research, Vol. 21, No. 2, February 2004 (© 2004)
Research Paper
tions, including the maintenance of blood osmolarity, acting
as an antioxidant, and serving as a solubilizing agent and
carrier for many endogenous and exogenous compounds (1).
It is also well-known that it is a major binding protein for the
transport of a large number of drugs (1).
Esterase-Like Activity of Serum
Albumin: Characterization of Its
Structural Chemistry Using
p-Nitrophenyl Esters as Substrates
Studies of SA have also revealed the important role of
this protein as a catalyst for the hydrolysis of various com-
pounds, such as esters, amides, and phosphates (2,3). Ikeda et
al. reported that the most prominent catalytic, esterase-like,
active sites of human serum albumin (HSA) are closely re-
lated to its drug binding sites, because various drugs inhibit
this activity (4). Thus, the active site of HSA with respect to
p-nitrophenyl acetate (PNPA) is thought to be the same as
the binding site for several benzodiazepines (site II), and the
enzymatic active site with respect to nitroaspirins is in close
proximity to the warfarin binding site (site I) (4). However, in
spite of the extensive literature on these reactive sites, their
properties and their differences among species are not known
in detail.
1
1
1
Yuji Sakurai, Shen-Feng Ma, Hiroshi Watanabe,
2
2
Noriyuki Yamaotsu, Shuichi Hirono,
Yukihisa Kurono, Ulrich Kragh-Hansen, and
Masaki Otagiri
3
4
1,5
Received September 21, 2003; accepted October 10, 2003
Purpose. To elucidate the catalytic mechanism of the esterase-like
activity of serum albumin (SA), the reactivity of SA from six species
was investigated using p-nitrophenyl esters as model substrates.
Methods. The effect of pH and the energetic and thermodynamic
Typically, the esterase activity (e.g., carboxylesterase ac-
profiles of SA were determined for all species for p-nitrophenyl ac- tivity) is more pronounced in plasma from primates than in
etate (PNPA). Then, kinetic and thermodynamic studies using a se- other animals such as dogs, rabbits, snakes, and fish (5,6). The
ries of p- and o-nitrophenyl esters with different side chains and esterase-like activity of SA has been reported to be highest in
human SA (HSA) were carried out. The influence of deuterium oxide
human and is completely absent in SA from horse (Order
was also evaluated. Finally, the information gained was used to con-
Perissodactyla) (1,7). The primary structure of several species
struct a computer model of the structural chemistry of the reaction.
of albumin is now known (1), and mammalian types show
Results. The pH profiles suggest that the nucleophilic character of the
amino acid sequence identities of about 70–80%. Therefore, it
catalytic residue (Tyr-411 in the case of HSA) is essential for activity.
is conceivable that they have comparable three-dimensional
structures. In fact, X-ray diffraction studies by Ho et al.
showed that the crystal structure of horse albumin is similar to
This kcat-dependent activity was found to increase with a decrease in
the activation free energy change (⌬G). Hence, the magnitude of ⌬G,
which is dependent on activation entropy change (⌬S), as calculated
from the thermodynamic analysis, can be regarded as an indicator of the three-dimensional structures of HSA (8,9). Studies involv-
hydrolytic activity. It indicates that p-nitrophenyl propionate (PNPP) ing site-directed mutagenesis have shown that Arg-410 and,
is the best substrate by evaluating the reactions of nitrophenyl esters especially, Tyr-411 are important for the esterase-like activity
with HSA. The findings here indicate that deuterium oxide has no
significant effect on the rate of hydrolysis of PNPA by HSA.
of HSA (10). However, many albumin species, including
horse albumin, also contain arginine and tyrosine in the same
Conclusions. The results are consistent with a scenario in which HSA
or corresponding positions, but have different enzymatic ac-
becomes acylated due to a nucleophilic attack by Tyr-411 on the
tivities (1). Therefore, other, and presently unknown, factors
substrate and then is deacylated by general acid or base catalysis with
must be important as well.
the participation of water.
In the current study, we report on a detailed examination
of the esterase-like activity of albumin and the mechanism of
the catalytic reaction. Species differences were first examined.
KEY WORDS: esterase-like activity; p-nitrophenyl esters; serum al-
bumin; species difference; structure–activity relationship.
This involved a determination of pH profiles and the collec-
tion of thermodynamic data on the esterase activity of SA
INTRODUCTION
from human, bovine, dog, rabbit, rat, and horse using PNPA
as a substrate. To characterize the structural chemistry of the
active site, hydrolytic reactions of a series of p-nitrophenyl
and o-nitrophenyl esters with HSA were also investigated.
Finally, the optimum structure of the substrate–HSA complex
As the most abundant soluble protein in the body (about
% in serum) and the most prominent protein in plasma,
serum albumin (SA) is responsible for a multiplicity of func-
4
1
Graduate School of Pharmaceutical Sciences, Kumamoto Univer- was constructed using the X-ray structure of the HSA:my-
sity, Kumamoto 862-0973, Japan.
The School of Pharmaceutical Sciences, Kitasato University, Tokyo
ristate (MYR):tri-iodobenzoic acid (TIB) complex.
MATERIALS AND METHODS
Materials
2
3
4
5
108-8641, Japan.
Faculty of Pharmaceutical Sciences, Nagoya City University,
Nagoya 467-8603, Japan.
Department of Medical Biochemistry, University of Aarhus,
Aarhus C DK-8000, Denmark.
To whom correspondence should be addressed. (e-mail:
otagirim@gpo.kumamoto-u.ac.jp)
ABBREVIATIONS: Arg, arginine; HSA, human serum albumin;
MD, molecular dynamics; MYR, myristate; PNPA, p-nitrophenyl ac-
HSA was a gift from the Chemo-Sera-Therapeutic Re-
search Institute (Kumamoto, Japan), and dog, rabbit, bovine,
horse, and rat serum albumin samples were purchased from
the Sigma Chemical Co. (St. Louis, MO, USA). All of the
etate; PNPE, p-nitrophenyl esters; PNPP, p-nitrophenyl propionate; albumin samples were defatted according to Chen’s method
SA, serum albumin; TIB, tri-iodobenzoic acid; Tyr, tyrosine. before use (11).
285
0724-8741/04/0200-0285/0 © 2004 Plenum Publishing Corporation

2
86
Sakurai et al.
PNPA, p-nitrophenol, o-nitrophenyl acetate, and o- can be calculated from the intercept and slope of a double-
nitrophenol were purchased from Nacalai Tesque (Kyoto, reciprocal plot using Eq. (2):
Japan), and other p-nitrophenyl esters were ordered from
Sigma Chemical Co. Deuterium oxide (D > 99.75%) was ob-
tained from Merck Co. (Darmstadt, Germany). All other
chemicals were of analytical grade.
1
KS
1
=
+
(2)
kobs kcat[albumin] kcat
Determination of pKa
Assuming that the hydrolytic reaction catalyzed by albu-
min is dependent on the ionic form of Tyr-411, the rate of the
Procedures for Kinetic Runs
reaction should be both the ionization constant (K ) and pH-
dependent, as described in Eq. (3):
a
The buffer systems used were as follows: pH 6.0–8.5, 1/15
M phosphate; pH 9.0–10.5, 1/20 M borate. The ionic strength
of each buffer was not fixed, because the addition of NaCl to
the buffer decreased the reaction rates in a complicated man-
ner (12). The reaction medium contained 0.5% (v/v) acetoni-
trile. The temperature was 25°C except for the thermody-
namic analysis, which was done in a range of 15∼35°C.
Hydrolysis of substrate (5 ␮M) by various species of al-
bumin (at least a 5-fold excess concentration over the sub-
strate) was carried out in such a way that complications by
any multiple reactive sites of albumin were avoided. It is said
that the substrates preferentially reacted with the primary
reactive site of albumin under these conditions (12,13). The
reactions were followed with a stopped-flow spectrophotom-
kcat
k
1
cat
ͩ
ͪ
=
ͩ ͪ
K
a
×
(3)
+
KS
K
S
H
K + ͓H ͔
a
Equation (3) has the same form as the Michaelis–Menten
equation, where kcat/K is the value at completely ionized Tyr
S
+
form, and (kcat/K ) is the measured value at a given [H ]. K
can be calculated as follows, where [H ] is the hydrogen ion
concentration.
S H
a
+
kcat
KS
k
k
1
cat
cat
+
ͩ
ͪ ͩ ͪ ͩ ͪ
=
−
[H ] ×
(4)
K
K
Ka
H
S
S
H
The pK values were obtained using least squares
a
eter (Otsuka Electronics Co. Ltd., Osaka, Japan) at 400 nm method.
by monitoring the appearance of the corresponding phenol.
Under these conditions, pseudo-first-order rate constants Thermodynamic Analysis
could be obtained.
The free-energy change for an enzymatic reaction has
generally been discussed according to the literature (14).
Each free-energy parameter was calculated using Eqs. (5),
Measurement of Reaction Parameters
(6), and (7), where ⌬GS is the free-energy change for the
initial reaction of the enzyme and substrate, ⌬G the activation
free energy for the rate-determining step, ⌬GT the free-
energy difference for the reaction, R the gas constant, T the
absolute temperature, kB the Boltzmann constant, and h the
Planck constant.
The reactions of the substrates with albumin are assumed
to proceed through the pathway shown in Fig. 1 (4). In this
chart, PNPE is the substrate and EST–Alb denotes the Mi-
chaelis–Menten type complex between EST and albumin.
Acyl-Alb denotes albumin acylated by the substrate, and KS
represents the dissociation constant of the complex. The rate
⌬
G = −RTln͑1
ր
K )
(5)
(6)
(7)
S
S
constants of EST–Alb are represented by kcat and k , respec-
0
⌬
G = RT
h͒-lnkcat
͕ ͖
ln͑k T
ր
B
tively. k is the rate constant for the dissociation of Acyl-Alb.
The pseudo-first-order rate constant for the release of
phenol, kobs, can be represented as follows:
⌬
G = ⌬G − ⌬GS
T
The thermodynamic investigation of the albumin-
catalyzed reaction was performed in the temperature range
k K + kcat[albumin]
0
S
kobs
=
(1) 15∼35°C. The activation energy (E ) was obtained from an
a
K + [albumin]
S
Arrhenius plot of the rate constant vs. temperature following
Eq. (8).
Here, [albumin] is the concentration of albumin. Because
the value of kcat is much larger than that of k0 (2000- to
d ln kcat Ea
=
(8)
6
000-fold), so k can be ignored (4). The K and k values
d ͑1
ր
T͒
R
0
S
cat
Fig. 1. Spontaneous and albumin (Alb) catalyzed hydrolysis of p-nitrophenyl esters (EST).

Esterase-Like Activity of SA Using p-Nitrophenyl Esters
287
The activation enthalpy change (⌬H) and the activation nors in protein atoms surrounding site II were defined as
entropy change (⌬S) were calculated using Eqs. (9) and (10), UNITY queries. However, this was not done for hydrogen
respectively.
bond acceptors because there are no donors in PNPP. The
PNPP molecule should form a hydrogen bond to Arg-410,
which is important for esterase-like activity (10), and to at
least one of the other residues. However, Tyr-411 was not
included in the UNITY queries because it is a catalytic resi-
due. A hydrogen bond tolerance was set within 0.5 Å. We
searched for conformers that satisfied the above conditions
using a UNITY 3D Search and obtained only one hit. In this
search, the HSA template and each PNPP conformer were
kept rigid. Finally, the docked HSA–PNPP complex was mini-
mized in a same manner as above. AM1-Wang-Ford charges
⌬
H = E − RT
(9)
a
⌬
S = ͑⌬H − ⌬G͒
ր
T
(10)
Conformational Analysis of PNPP
The conformational analysis of PNPP was performed us-
ing the CAMDAS 2.1␤2 program (15). Ten molecular dynam-
ics (MD) calculations were simultaneously performed using
different initial structures. Each of the MD calculations was
carried out for 600 ps with an integral time step of 1 fs. The
lengths of the covalent bonds were fixed. The temperature of
the system was maintained at 900 K in order to enhance the
sampling efficiency. The Tripos force field (16) was used to
evaluate the potential energy surface of the molecule. The
electrostatic and hydrogen bonding term of the potential were
(25,26) were generated as charges for the PNPP using CS
MOPAC (27).
RESULTS AND DISCUSSION
ignored in order to avoid intramolecular hydrogen bonding. Species Differences in Esterase-Like Activity
The angle and torsional terms of the potential were reduced
by a factor of 0.8 in order to increase the flexibility of the
molecule. Conformers were sampled at 100 step intervals,
thus producing 6000 conformations for each MD calculation.
pH Profiles
The pH profiles for kcat, K , and k /K for the hydro-
S
cat
S
A total of 60,000 conformations were preclustered with an
root mean square deviation (rmsd) threshold of 0.2 Å. Super-
position and rmsd calculations were performed for all of the
heavy atoms. After sampling, the reclustering of the sampled
conformers was performed with an rmsd threshold of 0.2 Å.
Before the clustering, each conformer was minimized until
the root mean square of the potential energy gradients was
lysis of PNPA by the different types of SA were constructed.
The kcat values are markedly dependent on pH in all species
examined (data not shown) and show that the susceptibility of
the active sites to nucleophilic attack increases with pH. In
contrast, the K values decrease with increasing pH (data not
S
shown). It has been reported that pH-dependent conforma-
tional changes (N–B transition) occur in albumin when going
from neutral to slightly alkaline pH (28). Therefore, alter-
ations in nucleophilic attack in active sites and in the affinity
to PNPA could be due, totally or partly, to changes in the
tertiary structure of albumin, which accompany the pH-
−
1
−1
below 0.005 kcal mol
obtained.
Å . Finally, 723 conformers were
Preparation of HSA Template
dependent N–B transition. The kcat/K values increase with
raising pH (data not shown). This difference in activity ap-
S
To obtain the HSA template structure for docking, we
used the crystal structure of the HSA:myristate (MYR):tri-
iodobenzoic acid (TIB) complex (PDB ID code: 1bke) (17).
Due to the lack of N-terminal residues from the 1st to the 3rd
and of the C-terminal residue of the 585th in the PDB struc-
ture, we treated the 4th residue as the N-terminus and the
pears to be mainly due to differences in kcat
.
The results of the pH experiments indicate that nucleo-
philicity in the active sites is very important in catalysis activ-
ity. The pK value for the hydrolysis of PNPA with various
a
albumins is calculated, and the result for human is shown in
Fig. 2, as a typical example. It is lowest in human albumin, in
which, however, the highest enzymatic activity was found.
5
84th residue as the C-terminus. The initial positions of the
other missing atoms in the PDB structure were generated
using the SYBYL 6.6.2 program package (18). The Tripos
force field was used. The AMBER 1991 charges (19) were
used as the charges for HSA. Gasteiger-Hückel (20–23)
charges were used as the charges for MYR and for TIB. Only
the missing atoms were minimized using SYBYL to below
Horse albumin has the highest pK value for hydrolysis but
a
has the lowest enzymatic activity (Table I). These results in-
dicate that hydrolytic activity is dependent on the pK and on
a
the nucleophilicity of the most important catalytic residue,
Tyr-411, in the various albumins. According to X-ray crystal-
lographic studies of HSA, the oxygen of the hydroxyl group
of Tyr-411 is adjacent to the nitrogen of the guanidine group
of Arg-410 (10). Therefore, the species differences in ester-
ase-like activity could be affected by structural changes after
substrate binding.
−1
−1
0
.05 kcal mol
Å
for the root mean square of the gradients.
The cut-off distance for nonbonded interactions was 8 Å.
Finally, the MYR and TIB molecules on HSA were deleted.
Docking of PNPP with HSA
The docking calculation of each PNPP conformation
On the other hand, Tyr-411 and Arg-410, which are es-
with HSA was performed using the UNITY 4.1.2 program sential for hydrolytic activity, are perfectly preserved in all
package (24). A MOLCAD fast Connolly surface was created species examined to date, and the amino acid residues in the
for the protein atoms surrounding site II of HSA (within 8 Å immediate vicinity of Tyr-411 residue are highly conserved
from MYR-1003 and MYR-1004). We defined a UNITY sur- among species. Thus, the significant decrease in reactivity in
face volume constraint expanded by a tolerance of 1.0 Å from rat and horse albumin must reflect differences in the micro-
the MOLCAD surface. The PNPP molecule should be the environment in the active sites rather than a deficiency of
inside of the surface volume constraint. Hydrogen bond do- basic activity. The SA with a lower pK can cause an easier
a

2
88
Sakurai et al.
Table II. Free Energy and Other Thermodynamic Parameters for the
Hydrolysis of PNPA by Various Albumins*
⌬
GT
⌬GS
⌬G
⌬H
T⌬S
Species (kcal/mol) (kcal/mol) (kcal/mol) (kcal/mol) (kcal/mol)
Human
Bovine
Rabbit
Dog
Rat
Horse
13.9
14.5
14.6
14.9
15.5
15.7
−5.0
−4.6
−4.6
−4.8
−5.1
−5.1
18.9
19.1
19.2
19.7
20.6
20.8
15.8
15.3
15.0
15.3
14.8
14.9
−3.2
−3.9
−4.2
−4.4
−5.7
−5.9
PNPA, P-nitrophenyl acetate.
*
Reaction conditions: 1/15 M phosphate buffer (pH 7.4) at 25°C.
Average of values from two experiments, which coincided with each
other within 6%.
the substrate (the ester portion) for the hydrolysis, shows a
smaller difference in entropy between the transition state and
the ground state. This is the reason why hydrolysis by HSA
proceeds more readily than by the other albumins. Otherwise,
the entropic difference can be also explained by the differ-
Fig. 2. Plots of [log kcat/K ] against pH for the hydrolytic reaction of
PNPA with HSA. Reaction conditions: 1/15 M phosphate buffer at
S H
25°C. Plots represent mean ± SD (n ס 3).
dissociation of Tyr-411 residue, which can be predicted to ences in the effects on water structure in active site(s).
have a higher reactivity in consequence.
Structure–Activity Relationships
Energy and Thermodynamic Profiles
Significant variations in hydrolytic activity were found
Using the hydrolysis parameters, we calculated the en- for different substrates, even if they contained the same func-
ergy changes for the different albumins at pH 7.4. ⌬G , con- tional group. Structure–activity studies with enzymes have
T
sidered to be an indicator of the energy difference, is depen- made it relatively easy to define the structure and the physi-
dent on ⌬G and can be calculated from kcat (Table II). As cochemical features of the active sites by measuring the ef-
described above, the hydrolysis reaction of albumin is kcat fects of changes in the structure of the substrate on its inter-
dependent. This observation suggests that the esterase activ- action with the enzyme (14).
ity is due to the energy difference between the substrate–
protein complex in the transition state (ES*) and in the PNPA, the reactions of p- and o-nitrophenyl esters with HSA
ground state (ES).
were investigated kinetically and thermodynamically. The ki-
The free-energy diagram for the hydrolysis of human and netic parameters for the hydrolytic reactions of nitrophenyl
In order to characterize the active site(s) of HSA toward
horse albumins toward PNPA is given in Fig. 3. The different esters with HSA are listed in Table III. The values of K
esterase-like activity of various albumins is due to differences
in ⌬G, namely, the extent of stabilization of the transition
state.
S
In order to investigate the characteristic feature of hy-
drolysis with respect to PNPA in greater detail, thermody-
namic studies on this catalytic reaction were performed. As
given in Table II, ⌬G, mentioned above, which reflects the
activity difference, is due to an entropy change (⌬S). Hence,
the entropic difference between the ground state and the
transition state is especially significant for hydrolytic reac-
tions of albumin. The active site of HSA to which the sub-
strate binds, having a perfect orientation to the binding site of
Table I. pK Values for the Catalytic Group in Various Albumins
a
Species
pKa
Human
Bovine
Rabbit
Dog
Rat
Horse
9.2
9.3
9.4
9.4
9.5
9.7
Fig. 3. Comparison of the free-energy profiles of the hydrolytic re-
action of PNPA with albumin from human and horse. ⌬G , free-
T
Average of values from 2–3 experiments, which coincided with each energy differences; ⌬G , free-energy change for the initial reaction of
S
other within 2%.
albumin and PNPA; ⌬G, activation free energy.

Esterase-Like Activity of SA Using p-Nitrophenyl Esters
Table III. Kinetic Parameters for the Hydrolysis of p- and o-Nitrophenyl Esters by HSA*
289
Number of C
in side chain
k
(10
K
k
(M
/K
S
s )
cat
3
S
cat
−1 −1
−
−1
−2
Substrate
s
)
(10 mM)
p-nitrophenyl ester
PNPA
PNPP
Butyrate
Valerate
Capronate
Caprylate
Caprinate
C2
C3
C4
C5
C6
C8
C10
C12
C5
86.8 ± 4.8
166.2 ± 12.2
19.3 ± 0.8
2.8 ± 0.1
1.2 ± 0.1
1.1 ± 0.2
0.6 ± 0.0
0.3 ± 0.0
0.4 ± 0.2
21.7 ± 0.3
15.5 ± 0.2
7.5 ± 0.1
4.5 ± 0.1
2.3 ± 0.1
0.9 ± 0.0
1.0 ± 0.0
1.3 ± 0.0
11.3 ± 0.3
403.4 ± 35.6
1079.3 ± 135.8
261.0 ± 23.9
63.3 ± 6.2
55.0 ± 19.0
133.7 ± 43.6
61.0 ± 14.4
25.7 ± 6.8
Laurate
Trimethyl acetate
o-nitrophenyl ester
Acetate
3.1 ± 1.0
C2
3.3 ± 0.6
10.7 ± 0.0
30.6 ± 4.6
HSA, human serum albumin; PNPA, p-nitrophenyl acetate; PNPP, p-nitrophenyl propionate.
*
Reaction conditions: 1/15 M phosphate buffer (pH 7.4) at 25°C. The concentration of substrate is 5 mM
and that of HSA is 25–150 mM. Each value is shown as mean ± SD. (n ס 4–5).
decreased with increasing length of the side chain. Up to transition state of capronate–HSA intermediate (ES*) has the
p-nitrophenyl valerate (a five-carbon side chain) a linear de- highest energy of all the p-nitrophenyl esters. Stabilization of
crease in KS values against the increase in the number of the transition state is an important aspect of the hydrolytic
carbons in the side chain was found, showing that the hydro- reaction, and ⌬G is an indicator of the esterase-like activity of
phobic side chain of the substrate influences the hydrolytic albumin. As a marker of chemical conversion, kcat is more
reaction. However, the decrease in kcat/KS and kcat values, significant for hydrolytic activity than K , which can be
S
depending on the incremental length of the side chain, was thought of as an index of affinity for the substrate.
not a decrease but a pronounced increment in the case of
The magnitude of the activation free energy (⌬G) is re-
PNPP (a three-carbon side chain). This effect was one of lated to the magnitude of the activation entropy change (⌬S)
affinity to albumin. Corresponding to the hydrolytic differ- as seen from the thermodynamic analysis (Table IV). Hence,
ences of the various albumins, the kcat-dependent hydrolytic because of the entropic difference between the enzyme–
reaction catalyzed by albumin is due to the directing proper- substrate complex states (ES) and the transition state (ES*),
ties of binding to the substrate rather than affinity to albumin. PNPP has the minimum value of the related nitrophenyl ester
It is conceivable that PNPP has a structure that is more suit- derivatives tested and, therefore, has the highest activity in
able to the active expression of albumin, compared with other this study. That is, PNPP has an optimal structure and a more
p-nitrophenyl esters. However, among structural isomers with perfect orientation for binding to albumin than the other es-
the same molecular weight, such as PNPA and o-nitrophenyl ters during the hydrolytic reaction.
acetate or p-nitrophenyl valerate and p-nitrophenyl trimeth-
Considering the findings of hydrolytic differences be-
ylacetate, substantial differences in activities were observed, tween various species, it can be concluded that a ⌬S depen-
which are believed to be related to steric exclusion caused by dence of the transition state stabilization is significant for
the different branches and sites of branching in the side chain. albumin hydrolysis and that the binding of the substrate to
The findings relative to the reactions of p-nitrophenyl albumin is a contributing factor that determines activity.
esters with HSA are consistent with the above experimental
studies on species differences using PNPA, suggesting that
hydrolytic activity is dependent on the activation free energy
Structural Mechanism of Esterase-Like Activity
(
⌬G) (Table IV). Differences in the free-energy diagrams for
The catalytic reaction of serine-proteases, a primary hy-
the hydrolysis of PNPP vis-à-vis capronate can be illustrated drolase, can be divided into two processes. Initially, the en-
as following the same pathway as shown in Fig. 3, where the zyme and substrate associate to form a noncovalent enzyme–
Table IV. Free Energy and Other Thermodynamic Parameters for the Hydrolysis of Esters by HSA*
Number of C
in side chain
⌬GT
(kcal/mol)
⌬GS
(kcal/mol)
⌬G
(kcal/mol)
⌬H
(kcal/mol)
T⌬S
(kcal/mol)
Substrate
PNPA
PNPP
Butyrate
Valerate
Capronate
C2
C3
C4
C5
C6
13.9
13.3
14.2
15.0
15.1
−5.0
−5.2
−5.6
−5.9
−6.3
18.9
18.5
19.8
21.0
21.5
15.8
16.3
14.3
14.7
12.8
−3.1
−2.2
−5.5
−6.2
−8.7
HSA, human serum albumin; PNPA, p-nitrophenyl acetate; PNPP, p-nitrophenyl propionate.
*
Reaction conditions: 1/15 M phosphate buffer (pH 7.4) at 25°C. Average of values from two experi-
ments, which coincided with each other within 4%.

2
90
Sakurai et al.
Fig. 4. Binding model for PNPP in site II of HSA (A) and the X-ray crystallographic structure of the HSA–myristate (MYR)
complex in site II of HSA (B). The illustration was made with the RasMol program.
substrate complex, which is held together by interactions be- during the first step of the hydrolytic reaction, generating
tween the hydroxyl group of the serine residue and the p-nitrophenol. Because no significant deuterium oxide iso-
substrate (referred to as a catalytic triad). This complex rep- tope effect could be detected, this implies that a water mol-
resents the first tetrahedral intermediate. Water is then in- ecule is not involved in the hydrolysis catalyzed by HSA.
volved in the formation of the enzyme–product complex via a More specifically, the first step of the hydrolysis does not
second tetrahedral intermediate with the acyl-serine. This is proceed via general acid or base catalysis, but is predomi-
followed by transacylation from serine to water. In addition nantly due to the nucleophilic attack of Tyr-411 residue.
to this catalytic triad mechanism, a region of the enzyme,
The same proton exchange technique has been used to
termed the oxyanion hole, has been reported to exist and investigate the catalytic mechanism of many serine proteases
which promotes catalysis by providing transition-state stabi- (30). In these cases, the reactivity was dependent on the mole
lization (29).
fraction of deuterium present, suggesting that the isotope ef-
The catalytic reaction between HSA and PNPA can be fect observed for these reactions originated from the active-
expressed quantitatively using the Michaelis–Menten equa- site serine hydroxyl group (i.e., the reactions involved the
tion. Means et al. reported that, after combining rapidly and formation of catalytic triads). The current results are in con-
reversibly into an active albumin–substrate intermediate, p- trast to these findings and imply that a catalytic triad is not
nitrophenol is generated and rapid acetylation occurs at the involved in the hydrolysis of PNPA by HSA.
reactive residue, Tyr-411 (2). Ohta et al. observed that the
On the other hand, using N-trans-cinnamoylimidazoles
second reaction, deacetylation of acetyl-albumin, is consider- containing amide bonds as model substrates, during the sec-
ably slower than the first step (13). Except for these findings, ond step of hydrolysis, the deacylation rate of the cinnamoyl–
little is known concerning the esterase-like activity of HSA. albumin was found to be about 3- to 4-fold smaller in deute-
For example, there have been no reports concerning the pos- rium oxide than in water (13). This deuterium effect indicates
sible presence of a catalytic triad or an oxyanion hole, as has that a water molecule plays a role in the deacylation reaction.
been reported for serine-proteases.
Therefore, general acid or base catalysis rather than nucleo-
In order to elucidate the detailed catalytic mechanism of philic catalysis predominates in the deacylation of cinnamoyl–
the esterase-like activity of HSA, the influence of deuterium albumin. The deacetylation of acetyl-albumin, in the second
oxide on the reactivity was investigated. Subsequently, a com- step of the hydrolysis of PNPA, appears to proceed according
plex of PNPP, showing the highest activity among the p- to the same mechanism as N-trans-cinnamoylimidazoles.
nitrophenyl esters tested, and HSA was constructed using
computer modeling techniques.
In summary, the hydrolysis of PNPA by HSA could take
place as follows: HSA is initially acetylated as the result of a
nucleophilic attack of Tyr-411 on the substrate, and the
deacetylation of acetyl-albumin then proceeds via general
acid or base catalysis involving a water molecule.
Isotope Effect of Deuterium Oxide
The involvement of a water molecule and proton transfer
can be evaluated by measurement of the deuterium oxide
isotope effect (14). The influence of this molecule on the
hydrolysis of PNPA by HSA was investigated to track the
behavior of the proton (data not shown). The result can be
Three-Dimensional Structure of the Modeled
HSA–PNPP Complex
In the model of HSA–PNPP complex, the binding of
regarded as a reflection of the acetylation reaction that occurs PNPP was similar to that of MYR-1004 in the X-ray structure

Esterase-Like Activity of SA Using p-Nitrophenyl Esters
291
Fig. 5. Simulated mechanism for the hydrolysis of PNPP by HSA.
(16) (Fig. 4). The p-nitrophenyl moiety of PNPP is bound to REFERENCES
the hydrophobic pocket of MYR-1004 in the X-ray structure.
In addition, the p-nitrophenyl moiety is associated with the
aromatic residues of Tyr-411 and Phe-488. The hydrogen
bond between the hydroxyl group of Ser-489 and the carboxyl
oxygen atom of PNPP stabilizes the complex. The ethyl moi-
ety of PNPP is bound to a part of the hydrophobic pocket for
MYR-1003 in the X-ray structure. The distance between the
phenoxyl oxygen atom of Tyr-411 and the carbonyl carbon
atom of PNPP was found to be about 3.8 Å and might enable
a nucleophilic reaction. The side chain of Arg-410 forms a
hydrogen bond with the carbonyl oxygen atom of PNPP, and
this hydrogen bond could function as an oxyanion hole.
Moreover, The distance between the phenoxyl oxygen atom
of Tyr-411 and the carbonyl oxygen atom of PNPP was about
1
2
. T. Peter Jr. All About Albumin, Biochemistry, Genetics, and
Medical Applications, Academic Press, San Diego, 1996.
. G. E. Means and M. L. Bender. Acetylation of human serum
albumin by p-nitrophenyl acetate. Biochemistry 14:4989–4994
(
1975).
3
. M. A. Sogorb, A. Monroy, and E. Vilanova. Chicken serum al-
bumin hydrolyzes dichlorophenyl phosphoramidates by a mecha-
nism based on transient phosphorylation. Chem. Res. Toxicol.
11:1441–1446 (1998).
4. K. Ikeda and Y. Kurono. Enzymatic activity and drug binding
activity of human serum albumin. Yakugaku Zasshi 106:841–855
(
1986).
5
6
. C. Y. Quon, K. Mai, G. Patil, and H. F. Stampfli. Species differ-
ences in the stereoselective hydrolysis of esmolol by blood ester-
ases. Drug Metab. Dispos. 16:425–428 (1988).
. T. Satoh, M. Hosokawa, R. Atsumi, W. Suzuki, H. Hakusui, and
E. Nagai. Metabolic activation of CPT-11, 7-ethyl-10-[4-(1-
piperidino)-1-piperidino] carbonyloxycamptothecin, a novel an-
titumor agent, by carboxylesterase. Biol. Pharm. Bull. 17:662–664
(1994).
2
.8 Å, thus favoring a hydrogen bond formation. This hydro-
gen bond could account for isotope effect observed for the
acetylation of HSA (Fig. 5). This model is consistent with the
experimental results; that is, Tyr-411 as the catalytic residue
7. A. Awad-Elkarim and G. E. Means. The reactivity of p-
nitrophenyl acetate with serum albumin. Comp. Biochem. Phys-
iol. 91B:267–272 (1988).
(
2,10) and the importance of Arg-410 (10), as mentioned
above.
8
. X. M. He and D. C. Carter. Atomic structure and chemistry of
human serum albumin. Nature 358:209–215 (1992).
. J. X. Ho, E. W. Holowachuk, E. J. Norton, P. D. Twigg, and
D. C. Carter. X-ray and primary structure of horse serum albu-
min (Equus caballus) at 0.27-nm resolution. Eur. J. Biochem.
215:205–212 (1993).
9
ACKNOWLEDGMENTS
This work was supported in part by Grant-in-Aid for
scientific research from the Ministry of Education, Science, 10. H. Watanabe, S. Tanase, K. Nakajou, T. Maruyama, U. Kragh-
and Culture of Japan (11694298 for M.O.) and was also sup-
ported in part by Grant-in-Aid for scientific research, encour-
agement of young scientists (B) (13771414 for N.Y.) from
Japan Society for the Promotion of Science.
Hansen, and M. Otagiri. Role of Arg-410 and Tyr-411 in human
serum albumin for ligand binding and esterase-like activity. Bio-
chem. J. 349:813–819 (2000).
1
1. R. F. Chen. Removal of fatty acids from serum albumin by char-
coal treatment. J. Biol. Chem. 242:173–181 (1967).

2
92
Sakurai et al.
1
2. Y. Kurono, T. Maki, T. Yotsuyanagi, and K. Ikeda. Esterase-like 21. M. Marsili and J. Gasteiger. Croat. Chem. Acta 53:601–614
activity of human serum albumin: structure-activity relationships (1980).
for the reactions with phenyl acetates and p-nitrophenyl esters. 22. J. Gasteiger and M. Marsili. Prediction of proton magnetic reso-
Chem. Pharm. Bull. 27:2781–2786 (1979).
nance shifts: the dependence on hydrogen charges obtained by
iterative partial equalization of orbital electronegativity. Organ.
Magn. Reson. 15:353–360 (1981).
1
3. N. Ohta, Y. Kurono, and K. Ikeda. Esterase-like activity of hu-
man serum albumin II: reaction with N-trans-cinnamoylimid-
azoles. J. Pharm. Sci. 72:385–388 (1983).
4. A. Fersht. Structure and Mechanism in Protein Science, Freeman,
New York, 1998.
5. H. Tsujishita and S. Hirono. CAMDAS: an automated confor-
mational analysis system using molecular dynamics. Conforma-
tional analyzer with molecular dynamics and sampling. J. Com-
put. Aided Mol. Design 11:305–315 (1997).
6. M. Clark, R. D. Cramer III, and N. Van Opdenbosch. Validation
of the general purpose Tripos 5.2 force field. J. Comp. Chem.
23. W. P. Purcell and J. A. Singer. A brief review and table of
semiempirical parameters used in the Hüeckel molecular orbital
method. J. Chem. Eng. Data 12:235–246 (1967).
1
1
2
2
4. UNITY 4.1.2, Tripos Inc., St. Louis, MO, 2000.
5. G. P. Ford and B. Wang. New approach to the rapid semiempiri-
cal calculation of molecular electrostatic potentials based on the
AM1 wave function: Comparison with Ab Initio HF/6-31G* re-
sults. J. Comp. Chem. 14:1101–1111 (1993).
6. B. Wang and G. P. Ford. Atomic charges derived from a fast and
accurate method for electrostatic potentials based on modified
AM1 calculations. J. Comp. Chem. 15:200–207 (1994).
7 CS ChemOffice Ultra 2000 Enhanced 5.5, CambridgeSoft, Cam-
bridge, MA, 2000.
1
1
2
10:982–1012 (1989).
7. S. Curry, H. Mandelkow, P. Brick, and N. Franks. Crystal struc-
ture of human serum albumin complexed with fatty acid reveals
an asymmetric distribution of binding sites. Nature Struct. Biol.
2
2
8. M. Dockal, D. C. Carter, and F. Ruker. Conformational transi-
tions of the three recombinant domains of human serum albumin
depending on pH. J. Biol. Chem. 275:3042–3050 (2000).
5:827–835 (1998).
1
1
8. SYBYL 6.6.2, Tripos Inc., St. Louis, MO, 2000.
9. S. J. Weiner, P. A. Kollman, D. T. Nguyen, and D. A. Case. An 29. C. Branden and J. Tooze. Introduction to Protein Structure, Gar-
all atom force field for simulations of proteins and nucleic acids.
land Publishing, Inc., New York, 1991.
J. Comp. Chem. 7:230–252 (1986).
30. R. L. Stein. Catalysis by human leukocyte elastase: substrate
structural dependence of rate-limiting protolytic catalysis and op-
eration of the charge relay system. J. Am. Chem. Soc. 105:5111–
5116 (1983).
2
0. J. Gasteiger and M. Marsili. Iterative partial equalization of or-
bital electronegativity—a rapid access to atomic charges. Tetra-
hedron 36:3219–3228 (1980).