J. Debord et al. / Biochimie xxx (2013) 1e6
5
Table 2
the two compounds is in agreement with the very close values of
Comparison with some literature values. The activation enthalpy
D
Hz and the
their octanolewater partition coefficients: log P ¼ 1.49 for phenyl
acetate and 1.46 for phenol, as estimated by Leo’s method [20].
However, the thermodynamics of enzyme binding differs widely
from a simple octanolewater transfer. For phenol, this transfer is
Arrhenius activation energy Ea are in kJ molꢁ1. Q10 is the factor by which kcat in-
creases when the temperature is raised by 10 K, from 298 K to 308 K. kunc is the rate
constant of the uncatalyzed reaction. ARE: human serum arylesterase; SA: serum
albumin.
slightly exothermic (
D
Hꢀ z ꢁ8 kJ molꢁ1) with an important rise in
Catalyst
Substrate
D
Hz
Ea
Q10 kcat/kunc
Ref.
entropy (
D
Sꢀ z þ18 J molꢁ1
K
ꢁ1) [21]. On the other hand, the
None
ARE (purified)
ARE (whole serum) PhOAc
PhSAc
SA (various species) pNO2ePhOAc 61.9e66.1
PhOAc
PhOAc
77.4
40
26
(a)
(b)
(c)
(c)
enzyme binding is about 4 times more exothermic with an
important loss in entropy. The increased exothermicity can be
attributed to the calcium complexation and to the formation of
hydrogen bonds. The entropy loss can be attributed to the inclusion
of the ligand into a highly organized network of intermolecular
bonds involving calcium and several amino acids (Fig. 4).
37.4
1.69 2.5 ꢄ 108
1.42
45.5 1.81
1000e2000 (d)
(a) Ref. [4].
(b) This work; kunc value from Ref. [4].
(c) Refs. [23,24].
It has been suggested that the combination of an unfavourable
entropy change compensated by a favourable enthalpic term is
characteristic of an ‘induced fit’ process [22]. But this phenomenon
is alꢀso accompanied by an important change in heat capacity
(d) Ref. [5].
our results is difficult since the assay methods differ widely.
However, the higher activation energy of thiophenyl acetate
supports the conclusion that the formation of the tetrahedral
intermediate is the rate-limiting step: if the decomposition
of the intermediate was limiting, one could expect a lower
activation energy for the thioester, since thiophenol is a
better leaving group (lower pKa) than phenol.
(D K
Cp > 4 kJ molꢁ1 ꢁ1) due to the ‘ordering of flexible ligand
binding sites’. In our case, the linearity of the plots in Fig. 2 indicates
a negligible heat capacity change. Indeed, an attempt to fit an
ꢀ
improved version of eq. (5), incorporating a
D
Cp term, resulted in
statistically non-significant parameters. This result suggests that
both the substrate and reaction product bind to a pre-existing
conformer of the enzyme, in agreement with the conclusions of
Ben-David et al. [2]. According to these authors, each class of ary-
lesterase substrates (carboxylic esters, phosphate esters,
lactones.) interacts with a specific conformer of the enzyme.
Lorentz et al. [23,24] reported Km values at 25 ꢀC and 30 ꢀC for the
hydrolysis of phenyl- and thiophenyl acetate by whole human serum.
(3) Sakurai et al. [5] studied the hydrolysis of 4-nitrophenyl ac-
etate by serum albumin, a non-specific catalyst. Later, Lock-
ridge et al. [26] showed that the hydrolysis of nitrophenyl
esters by albumin is due mainly to the acetylation of 82
amino acids, only one of them (Tyr-411) being deacetylated,
but at a very low rate (k3 z 0.0002 minꢁ1). So, it is likely that
the activation energies reported by Sakurai et al. refer to the
global acetylation step. With arylesterase, the kcat for 4-
nitrophenyl acetate is only about 2% of its value for phenyl
acetate [11]. This would increase the activation energy by
about 10 kJ molꢁ1, a value insufficient to explain the observed
difference in the activation enthalpies. So, arylesterase is
much more efficient than serum albumin, which seems to be
due to the presence of a calcium ion.
From their results we estimated
D
respectively. Although the precision of these estimates is poor (since
there are only two temperatures), and the enzyme is not purified, the
values for phenyl acetate compare well with our results.
D
Hꢀ z ꢁ40 and ꢁ16 kJ molꢁ1
,
Sꢀ z ꢁ77 and ꢁ0.6 J molꢁ1 Kꢁ1 for phenyl- and thiophenyl acetate,
Cléry-Barraud et al. [18] studied the influence of hydrostatic
pressure upon the hydrolysis of several substrates catalysed by
arylesterase, with or without HPBP. The results allowed to deter-
mine the volume variation
D
Vꢀ for the binding step. For phenyl
Cléry-Barraud et al. [18] found an activation volume
acetate, in the presence of HPBP, a slight decrease in volume
D
Vz z þ3.3 mL molꢁ1 for the hydrolysis of phenyl acetate by the
(
D
Vꢀ z ꢁ2.7 mL molꢁ1) was noticed for pressures up to 150 MPa.
arylesterase/HPBP complex. An increase in activation volume has
been attributed to bond cleavage and charge neutralization, while a
decrease has been attributed to bond formation and ionization [27].
It is likely that the breakage of the carbonyl double bond and the
subsequent neutralization of the oxygen charge by the calcium ion
contribute to the observed increase in activation volume.
This seems to be the consequence of the many intermolecular
bonds which occur during the binding of the substrate.
5.4. Catalysis by arylesterase
The rate-limiting step for the hydrolysis of phenyl acetate, cat-
alysed by arylesterase, is the nucleophilic attack of the ester carbon
by water. The activation enthalpy of 37.4 kJ molꢁ1 corresponds to an
5.5. Practical applications
Arrhenius activation energy Ea z 40 kJ molꢁ1 (since Ea ¼
D
Hz þ RT
The present work may lead to applications in the following
fields:
and the mean temperature in our assay is 298 K). This also corre-
sponds to a Q10 of 1.69 (Q10 is the factor by which kcat increases
when the temperature is raised by 10 K, from 298 K to 308 K).
Table 2 compares our findings with some literature values. The
following comments apply:
(1) Arylesterase assays: for the reasons explained in Section 4.2, a
continuous assay at low substrate concentration may be more
useful than an initial rate assay, for instance if a progressive
inhibitor must be studied in the presence of substrate.
(2) Determination of thermodynamic parameters: equation (7)
provides a useful alternative to the classical Arrhenius plots.
(1) The activation energy for arylesterase is about half the value
for the uncatalysed reaction. So, the enzyme is very efficient
in its function of stabilizing the transition state. It has been
suggested that this stabilization is due to the ‘electrostatic
preorganization of the enzyme active site’ [25]. Obviously,
the calcium ion contributes an important part in this process.
Also, the existence of a preorganized site is in agreement
with our conclusions regarding the binding step.
(3) Ligand optimization: the knowledge of both DH and DS, and
of course the knowledge of the reaction mechanism, helps
designing specific ligands for an enzyme [28].
References
(2) Lorentz et al. [23,24] studied the hydrolysis of phenyl- and
thiophenyl acetate by whole human serum. Comparison with
Please cite this article in press as: J. Debord, et al., Temperature dependence of binding and catalysis for human serum arylesterase/paraoxonase,