Effect of low organic solvents concentration on the stability and catalytic activity of HSL-like carboxylesterases: Analysis from psychrophiles to (hyper)thermophiles

Article abstract of DOI:10.1016/j.molcatb.2012.06.002

A comparative study aiming at analysing the catalytic activity and stability of four related enzymes belonging to the HSL family in the presence of several water-miscible organic solvents was performed. The studied enzymes were four carboxylesterases: Psy

Full text of DOI:10.1016/j.molcatb.2012.06.002

Journal of Molecular Catalysis B: Enzymatic 82 (2012) 46–52
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Effect of low organic solvents concentration on the stability and catalytic
activity of HSL-like carboxylesterases: Analysis from psychrophiles to
(hyper)thermophiles
Luigi Mandrich∗, Concetta De Santi, Donatella de Pascale, Giuseppe Manco
Institute of Protein Biochemistry, CNR, Via Pietro Castellino, 111, I-80131 Naples, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A comparative study aiming at analysing the catalytic activity and stability of four related enzymes
belonging to the HSL family in the presence of several water-miscible organic solvents was performed.
The studied enzymes were four carboxylesterases: PsyHSL (psychrophilic), Aes (mesophilic), EST2 (ther-
mophilic) and AFEST (hyperthermophilic). The effect of the solvents on catalytic activity was measured
by adding solvents up to 20% (v/v) to the assay mixture. For all enzymes at low solvent concentrations
(4–5%) the catalytic activity (U/mg) increased up to 3–4-fold; only diethyl ether exerted a negative effect
when compared to the reference value without solvent. EST2 and AFEST retained more than 100% of
activity for concentrations of solvent >10% (v/v). Similar results were obtained for the stability analysis;
in fact these two enzymes appeared more stable than the mesophilic and psychrophilic counterparts.
The kinetic parameters of PsyHSL, EST2 and AFEST in all the tested solvents showed an increase in
kcat but a negative effect on KM values. In the case of PsyHSL the KM increase was very low resulting
in a substantial increase of the specific constant (S); in particular, the addition of dimethyl sulfoxide
increased the kcat and the s values about 4- and 3-fold, respectively. We tested the ability of the enzymes
to synthesize esters with the aim to discriminate if the decrease of activity in the presence of organic
solvents was due to protein denaturation or an inversion of the hydrolytic reaction toward synthesis.
While for PsyHSL and Aes the synthesis was very low and completely abolished in presence of solvents,
instead for EST2 and AFEST the esters synthesis increased up to 10-fold in the presence of solvents,
although the hydrolytic activity dropped about 300-fold.
Received 11 January 2012
Received in revised form 17 May 2012
Accepted 2 June 2012
Available online 12 June 2012
Keywords:
Carboxylesterases
Enzyme kinetics
Organic solvent effect
Thermal stability
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
solvent may alter these mechanisms and lead to improvement or
worsening of the enzyme performance. In general, the activity and
The properties of an enzyme in organic solvents may differ dra-
matically with respect to those in aqueous environment [1]. It is
accepted that water is essential for the maintenance of native con-
formation but it is also involved in several mechanisms of enzyme
inactivation [2]. Irreversible thermo-inactivation of enzymes, in
aqueous solution at 90–100 ◦C depends on covalent modifications
at level of: cysteine residues (thiol-catalysed disulfide interchange,
oxidation of –SH groups); asparagines and/or glutamine residues
(deamidation); and hydrolysis of peptide bonds nearby aspartic
acid residues [3,4]. In most of these thermo-inactivation processes
water is one of the reactants as well as in heat-induced protein
denaturation where aggregation of enzymes and incorrect struc-
ture formation are observed [3,4]. The presence of an organic
stability of enzymes depend on the nature of the solvent (water-
miscible or immiscible, protic or aprotic, etc.) and the water content
of the enzyme; biocatalysts which show increased activity and/or
stability in organic solvents are considered attractive for biotech-
nological applications [5].
Normally, an enzyme shows enhanced thermostability when
dispersed in pure water-immiscible organic solvents. This has been
observed for porcine pancreatic lipase and lipase from Candida
cylindracea ribonuclease, chymotripsin and several other enzymes
[5,6]. Enzymes in dry organic solvents show a significant character-
istic that is higher conformational rigidity, that prevents them to
unfold and this could explain their increased thermostability com-
pared to that in aqueous solution [7]. In organic solvents enzyme
activity changes principally for the absence or drastic reduction of
water which assure the required flexibility for catalysis.
Proteins in mixtures of water and water-miscible organic sol-
vents (>20% of organic solvents) are known to be generally less
stable and less active than in water only, even though some
∗
Corresponding author. Tel.: +39 081 6132581; fax: +39 081 6132277.
E-mail address: l.mandrich@ibp.cnr.it (L. Mandrich).
1381-1177/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcatb.2012.06.002

L. Mandrich et al. / Journal of Molecular Catalysis B: Enzymatic 82 (2012) 46–52
47
exception exist depending on the solvent or the protein [8,39,41].
2.2. Esterase activity
Protein folding is a balance between non-covalent, hydrophobic
and electrostatic interactions, and they are influenced by the sur-
rounding conditions (temperature, ionic strength of the solvent)
as well as type of enzyme. In the presence of organic solvents the
hydrophobic interactions are disrupted and proteins aggregate or
unfold [6,9]. Studies of the behaviour of enzymes of biotechnologi-
cal interest in the presence of organic solvents are useful for further
applications.
The time course of the esterase-catalysed hydrolysis of pNP-
esters (C5 and C6 acyl chain length) was followed by monitoring
of p-nitrophenol production at 405 nm, in 1-cm path-length cells
with a Cary 100 spectrophotometer (Varian, Australia). Initial rates
were calculated by linear least-squares analysis of time courses
comprising less than 10% of the total substrate turnover. Assays
were performed at the optimal temperature of each enzymes
(35 ◦C for PsyHSL; 65 ◦C for Aes; 70 ◦C for EST2; 80 ◦C for AFEST)
[10,12,13,29], in a mixture of 40 mM Tris–HCl pH 8.0 for PsyHSL
and 40 mM buffer phosphate pH 7.1 for the other enzymes, 4%
acetonitrile at the indicated pH (3% for PsyHSL), containing pNP-
esters (100 ␮M). Stock solutions of pNP-esters, C5 and C6, were
prepared by dissolving substrates in pure acetonitrile [10,12,13].
Samples of identical composition as the assay mixture, omitting
the enzyme, provided suitable blanks. Assays were done in dupli-
cate and results were means of two independent experiments.
One unit of enzymatic activity was defined as the amount of
the protein releasing 1 ␮M of p-nitrophenoxide/min from pNP-
pentanoate or -hexanoate at indicated temperature. The absorption
coefficient used for p-nitrophenoxide was 19,000 M−1 cm−1 at
35 ◦C and pH 8.0 or 20,000 M−1 cm−1 form 65 to 80 ◦C at
pH 7.1.
In this paper, we studied the effect of the increasing concen-
tration of several water-miscible organic solvents on stability,
catalytic activity, kinetic parameters and ester synthesis of four
carboxylesterases, which cover a range of catalysis temperature
from 5 to 85 ◦C: the psychrophilic PsyHSL from Psychrobac-
ter sp. TA144 [10], the mesophilic Aes from Escherichia coli
[11], the thermophilic EST2 from Alicyclobacillus acidocaldar-
ius [12] and the hyperthermophilic AFEST from Archaeoglobus
fulgidus [13]. These enzymes belong to the same group, the HSL
family [14–17], from the classification by ESTHER DATABASE
(http://bioweb.ensam.inra.fr/ESTHER/), that corresponds to the
family IV of the classification by Arpigny and Jaeger [18]. Previ-
ously only kinetic parameters of EST2 in acetonitrile and alcohols
such as methanol, ethanol and 2-propanol have been reported [25].
Since these enzymes were related, we studied not only their dif-
ferent behaviour with several solvents but also differences due
to their altered thermal adaptation. The effect of organic sol-
vents on the activity and stability of enzymes is not universal
and seems to depend on the properties of the solvent and/or
the considered enzyme. The organic solvents were selected for
their different physico-chemical characteristics: polar aprotic (ace-
tonitrile, dimethyl formamide, dimethyl sulfoxide), polar protic
(methanol), polar poly-protic (ethylene glycol) and non polar
(diethyl ether). In Table 1 are summarized the main features of
these solvents.
In modern biotechnology, the application of microorganisms
and enzymes to the production of chemicals, biopolymers, mate-
rials and fuels offers great opportunities for the industries. In
the case of lipases/esterases their substrates and/or products are
generally not miscible in water and the industrial biotransfor-
mation processes run under harsh conditions, such as high and
low temperature, extremes of pH and high pressure, presence
of salt, organic solvents and metals [24]. For these reasons, the
organic solvent tolerance of enzymes is a relevant property, which
show potential in biotechnological applications. In the case of the
enzyme belonging to the HSL family some data about the organic
solvent tolerance were reported [10,25]. Herein we reported a
complete analysis on the effect of the six organic solvents on
PsyHSL, Aes, EST2 and AFEST enzymes. Moreover we checked
the effect of organic solvents on a reaction of biotechnologi-
cal interest, namely the modification of EST2 enantioselectivity
towards the substrate (RS)-p-nitrophenyl-2-chloropropionate. This
substrate is used to produce (S)-2-chloropropionic acid, an inter-
mediate in the synthesis of optically pure phenoxypropionic acid
herbicides [26,27].
2.3. Effect on enzymatic activity by organic solvents
Enzyme activity was evaluated in the optimal conditions of
temperature and pH of each enzymes using the better reported sub-
strate (100 ␮M final concentration), that are pNP-pentanoate for
PsyHSL [10] and pNP-hexanoate for Aes, EST2 and AFEST [12,13,35].
Stock solutions of each substrate were prepared in all the solvents
tested; the activity was measured using an increasing concentra-
tion of the solvents up to 20% (v/v) in the assay mixture. Results
were reported as relative activity with respect to the value mea-
sured without solvents.
2.4. Stability to solvents
The stability to solvents of the enzymes was studied at the opti-
mal temperature and pH of each enzyme incubated in a mixture
containing 2.5, 5.0 and 10.0% (v/v) of each organic solvent. Pure
enzymes (0.2 mg/ml) were incubated in sealed glass tubes at the
appropriates temperatures. Aliquots were withdrawn at 15, 30, 60,
120, 240, and 360 min (in some case up to 24 h) and assayed at the
optimal temperature of each enzymes, in the standard conditions
described above.
2.5. Kinetic measurements
Initial velocity versus substrate concentration data were fitted
to the Lineweaver–Burk transformation of the Michaelis–Menten
equation, by weighted linear least-squares analysis with a per-
sonal computer and the GRAFIT program 3.0 [30]; pNP-pentanoate
ranged from 1 to 200 ␮M and pNP-hexanoate from 1 to 150 ␮M.
Assays were done in duplicate or triplicate and results for kinetic
data were mean of two independent experiments. Kinetic param-
eters were calculated for each organic solvents (3%, v/v for PsyHSL
and 4%, v/v for the other enzymes), dissolving pNP-pentanoate or
-hexanoate in each pure organic solvent, and also without solvent,
dissolving pNP-pentanoate or -hexanoate in water at their max-
imum of solubility. Measurements were done at the optimal pH
and temperature of each enzymes.
2. Experimental
2.1. Purification of proteins
Enzymes used in the experiments were expressed and purified
as previously reported: the psychrophilic Psy HSL from Psychrobac-
ter sp. TA144, see reference De Santi et al. [10]; the mesophilic Aes
from E. coli [11,29]; the thermophilic EST2 from A. acidocaldarius,
see reference Manco et al. [12]; and the hyperthermophilic AFEST
from A. fulgidus, see reference Manco et al. [13].

48
L. Mandrich et al. / Journal of Molecular Catalysis B: Enzymatic 82 (2012) 46–52
Table 1
Main characteristics of the organic solvents used in this study.
Structure
Characteristics
Hydrophobicity, log Pa
Denaturation capacity (DC)b
64.3
Polarity index, Ic
5.8
Dielectric constant, εd
−0.3
−1.0
37
63.3
60.3
6.4
7.2
38
47
−1.3
−0.76
−1.8
30.5
18.7
5.1
5.4
33
38
−0.8
50.2
6.7
4.3
a
The log P value were taken from, or calculated on the basis of Ref. [19].
The denaturing capacities of the solvents were taken from Ref. [20].
The polarity indexes of solvents were taken from Ref. [21].
The dielectric constant value were taken from Refs. [22,23].
b
c
d
2.6. In vitro ester synthesis
expressed as log P, that is empirically calculated log10 of the coeffi-
cient for solvent partitioning P between 1-octanol and water [19],
log P could be used as a criterion to correlate organic solvents and
their denaturating strength only for solvents of the same class such
as alcohols [20]; (b) denaturation capacity, that is an index of ther-
modynamic stability of a protein obtained by taking into account
the solvent concentration at the protein is half inactivated.
The assays were carried out in buffered solutions with or
without solvents as previously described [31]. Briefly, 50 ␮M
p-nitrophenol and 500 ␮M hexanoic acid were added to a buffered
solution containing or not the solvent, and after the addition of
the enzymes the activity was recorded at 405 nm by following
the decrease of absorbance due to ester synthesis. Each enzyme
was tested at its optimal temperature and pH in the presence of
5 and 10% (v/v) solvent. Results are the mean of two independent
experiments.
It has been experimentally verified that exists a linear rela-
tionships between DC of a solvent and half-inactivation of several
proteins. Thus DC represents a good index of the denaturing
strength of organic solvents [20]; (c) polarity index, that is a
measure of the enzymes denaturation degree by the solvent, in
other words solvents having polarity index of 5.8 and above if
are used as cosolvents cause reversible conformational changes
in the enzyme molecule up to a concentration of 50% (v/v). Sol-
vents having polarity index of 5.0 and below cause irreversible
conformational changes of the enzymes [21,22]; (d) dielectric con-
stant, that is a rough measure of a solvents polarity. Solvents with
a dielectric constant of less 15 are considered to be non polar [36].
The polar aprotic solvents, defined as solvents do not contain-
ing an –OH or –NH bond, were acetonitrile, dimethyl formamide
and dimethyl sulfoxide; they have different values of hydropho-
bicity, polarity index and dielectric constant [19,21–23]. Methanol
was used as polar protic solvent, defined as containing an –OH or
–NH bond, and ethylene glycol as polar polyprotic; they differ in
hydrophobicity and denaturation capacity values [19,20]; diethyl
ether was used as apolar solvent. Polar aprotic solvents tend to
have large dipole moment and solvate positively charged species
via their negative dipole. For all enzymes, we measured the activ-
ity as function of the solvent percentage, up to 20% by the volume,
and at different temperatures as follows: (i) at the reported optimal
temperature of each enzyme, (ii) at the living temperature of the
bacterium from which the enzymes have been isolated, (iii) at room
2.7. Enantioselectivity measurements
The assays were carried out in buffered solutions with or
without 4% (v/v) solvent (as previously described [26]). The
substrates were prepared dissolving (R)- or (S)-p-nitrophenyl-
2-chloropropionate in pure organic solvents. Assays were done
at 25 ◦C and in 50 mM sodium citrate buffer at pH 5.0. The
kinetic determinations were performed only for EST2 wild type
enzyme and for EST2 L212P mutant in acetonitrile [26]. Initial
velocity versus substrate concentration data were fitted to the
Lineweaver–Burk transformation of the Michaelis–Menten equa-
tion, by weighted linear least-squares analysis with a personal
computer and the GRAFIT program 3.0 [30]. Assays were done in
duplicate or triplicate and results for kinetic data were mean of two
independent experiments.
3. Results and discussion
3.1. Activity and stability
In Table 1 are reported the six solvents used and their main
characteristics; it has been considered their: (a) hydrophobicity,

L. Mandrich et al. / Journal of Molecular Catalysis B: Enzymatic 82 (2012) 46–52
49
Table 2
was inversely related to the activity increasing. In the case of AFEST
at 25 and 50 ◦C and up to 20% of acetonitrile we recorded enzyme
activation with respect to the sample without solvent set as 100%.
Instead EST2 appears to be more sensitive to high concentrations
of acetonitrile at all temperatures tested.
Enzymes half-live. Acetonitrile stability was measured at three different solvent
concentrations: 2.5, 5.0 and 10.0% (v/v). The stability was followed for different times
of incubation at the optimal temperature and pH of each enzyme; pNP-pentanoate
(PsyHSL) or pNP-hexanoate (Aes, EST2, AFEST) were used as substrates. The stability
was reported as half life-time t1/2 (min).
As far as the enzyme stability in acetonitrile is concerned, it
turned out that PsyHSL had low stability at all solvent concentra-
tions tested, with a half-life of few minutes (Table 2), while the
other enzymes seemed to be more stable, with a half-life of about
800 min for Aes at 2.5% (v/v) acetonitrile. A decreasing stability was
recorded at higher solvent concentration for all the enzymes tested
(Table 2).
Enzymes
Acetonitrile concentration (v/v)
2.5%
3 min
5.0%
10.0%
PsyHSL
Aes
EST2
2 min
250 min
100 min
95 min
1 min
170 min
90 min
60 min
800 min
330 min
105 min
AFEST
temperature (25 ◦C) and (iv) at 50 ◦C. The stability of enzymes to
solvents was measured at three different concentrations: 2.5, 5.0
and 10.0% (v/v), and at the optimal temperature and pH of each
enzyme. The kinetic parameters were calculated in presence of 4%
of each solvent and the rate of ester synthesis of each enzyme in all
solvents were calculated aiming at verify if the decrease of activ-
ity at high concentration of solvent was due to denaturation or to
some change of the equilibrium reaction.
3.3. Dimethyl sulfoxide
Concerning the addition of dimethyl sulfoxide, the greatest acti-
vation was obtained for PsyHSL (Supplemental data, Fig. S1A) and
its behaviour was very similar to that of EST2; AFEST (Supplemen-
tal data, Fig. S1C and D), showed a relevant activity also at high
solvent concentration, while for Aes was recorded a fast decrease
of activity at concentrations over 6% (Supplemental data, Fig. S1B).
As for the enzyme stability in dimethyl sulfoxide, the behaviour
was very similar to that obtained for acetonitrile at all concentra-
tions analysed; PsyHSL was less stable than the others enzymes and
Aes seemed to be less stable with respect to acetonitrile, showing
half-lives of about 600 and 150 min in 2.5 or 10%, DMSO, respec-
tively. EST2 seemed to be more stable than in acetonitrile showing
half-lives of 500 and 90 min in 2.5% and 10% DMSO, respectively
(Supplemental data, Table S1).
3.2. Acetonitrile
Fig. 1 shows the results obtained for the enzyme activity as
function of acetonitrile concentration and temperatures. In panel
A (Fig. 1) is shown the profile obtained for PsyHSL: we observed
a maximum increase of activity in presence of solvent at about
3% (v/v) at 35 ◦C; lower increments were recorded at 25 and 4 ◦C.
At a concentration of acetonitrile ranging between 6 and 8%, the
activity dropped down below the value measured without the sol-
vent. For other enzymes (Aes panel B, EST2 panel C, AFEST panel
D, Fig. 1), the maximum increase of activity was observed at 25 ◦C
and in presence of about 6% of solvent; the temperature increasing
3.4. Dimethyl formamide
The analysis of activity in DMF shows that the maximum
enhancement was obtained in the range of 2–6% of solvent for all
Fig. 1. Enzyme activity vs acetonitrile concentration (%, v/v). Data obtained for: (A) PsyHSL; (B) Aes; (C) EST2; (D) AFEST. For each enzyme the assays were done at the optimal
pH, and using pNP-hexanoate as substrate.

50
L. Mandrich et al. / Journal of Molecular Catalysis B: Enzymatic 82 (2012) 46–52
Table 3
Ester synthesis. We followed spectrophotometrically the synthesis of p-NP-hexanoate, starting from p-nitrophenol and hexanoic acid, using a reaction mixture containing 5
or 10% of each solvents. Results obtained at 10% of solvent concentration are reported in parentheses. Results are mean of two different experiments. One unit of synthetic
activity was defined as the amount of protein that synthesized 1 ␮mol of pNP-hexanoate/min.
Solvents
Enzymes activity (U/mg)
PsyHSL
Aes
EST2
AFEST
None
Acetonitrile
0.0051 0.0002
n.d.
0.0045 0.0005 (0.0028 0.0002)
0.0042 0.0005 (n.d.)
n.d.
n.d.
n.d.
0.012 0.001
0.0063 0.0005 (n.d.)
0.015 0.001 (0.0074 0.0005)
0.010 0.001 (n.d.)
n.d.
n.d.
n.d.
2.4 0.1
6.9 0.2 (4.3 0.1)
n.d.
3.3 0.1 (n.d.)
30.0 5.0 (3.1 0.1)
n.d.
n.d.
1.10 0.05
2.8 0.01 (2.1 0.01)
n.d.
2.5 0.02 (n.d.)
9.5 0.1 (4.0 0.2)
n.d.
Dimethyl sulfoxyde
Dimethyl formamide
Methanol
Ethylene glycol
Diethyl ether
n.d.
enzymes; for additional increments of solvent the activity dropped
down below the value measured in absence of solvent (Supplemen-
tal data, Fig. S2A–D).
3.7. Diethyl ether
In Fig. S5 (Supplemental data) it is reported the effect of diethyl
ether on the enzyme activities. In all the conditions tested, Aes,
EST2 and AFEST proved to be less active than without solvent, while
PsyHSL enzymatic activity increased at 25 and 37 ◦C up to 5% of
solvent concentration (Supplemental data, Fig. S5A). Interestingly
diethyl ether affects the stability of the same three enzymes in the
same way as the solvent which shows the lower half-lives, with
respect to the other solvents analysed, and at all the used per-
centages (2.5, 5.0 and 10.0%); in contrast the psychrophilic protein,
PsyHSL appeared to be more stable in diethyl ether with respect
to the other solvents tested here a result similar to the effect of
methanol (Supplemental data, Table S5).
By analysing all data regarding the enzyme stability in DMF,
we could hypothesize a stronger denaturing action of this solvent
most likely, due to the increased polarity with respect to the two
solvents used before. This could destabilize charges present on
the protein surface. Accordingly, PsyHSL showed half lives of few
minutes, and even EST2 and AFEST were demonstrated to be less
stable with respect to other solvents analysed (half lives of about 60
and 20 min at 2.5 and 10% DMF, respectively; Supplemental data,
Table S2).
How may we explain these results? An hypothesis could be that
being the structural core of (hyper)thermophilic enzymes much
more stable with a overall less flexible structure due to a large
number of ionic and hydrophobic interactions [32,33], an apolar
solvent reaching the inner core of the protein might interfere with
the protein hydrophobic interactions leading to protein unfolding.
On the other hand, psychrophilic enzymes are characterized by
a high structural flexibility, due to a lower number of hydrophobic
interactions [34], and as a consequence these proteins might be
less sensitive to the presence of apolar solvents (diethyl ether in
this case).
3.5. Methanol
The analysis of activity in methanol shows a behaviour sim-
ilar to other solvents, with an increase of activity in the range
3–5% of methanol, which reached about 4-fold for PsyHSL (Sup-
plemental data, Fig. S3). EST2 activation at 70 ◦C decreased
rapidly above 10% methanol (Supplemental data, Fig. S3C). This
result was confirmed also in terms of stability; in fact EST2
was demonstrated to be less stable in methanol with respect
to the other solvents analysed, whereas Aes showed an excel-
lent stability in this solvent (Supplemental data, Table S3).
Since methanol is a strong protein denaturant, because it dis-
rupt the hydrophobic interaction [38], we could speculate a
general effect due to the high polarity of methanol; further-
more, for EST2 and AFEST, the denaturation effect was enhanced
by the high temperatures used for the incubation, e.g. 70 and
80 ◦C.
3.8. Kinetic parameters
The kinetic parameters were calculated at 4% final concentra-
tion of each solvent for Aes, EST2 and AFEST and 3% for PsyHSL.
The results are summarized in Table S6 (Supplemental data). Fig. 2
shows the kcat and s = kcat/KM values as ratio of the value obtained
in water and solvents used for each enzyme. A careful reading of
the kinetic parameters pointed out an increase of enzymatic activ-
ity for all the tested enzymes, only exception being PsyHSL and
Aes, for which a slight reduction was observed in diethyl ether
(Fig. 2A). PsyHSL showed the strongest increases of kcat (around
4-fold) with respect to the other enzymes, in dimethyl sulfoxide.
Concerning Aes, we observed a large increase in kcat (up to 3-fold) in
DMF, ethylene glycol and methanol. Slighter increasing, up to 1.5-
fold, were recorded for EST2 and AFEST (Fig. 2A); finally, for all the
enzymes analysed, the addition of solvents increased the KM val-
ues (Supplemental data, Table S6). In terms of specificity constants
s (kcat/KM) the variations recorded correspond to a weak increase
for Aes, a decrease for EST2 and AFEST and 2–3 fold increase for
PsyHSL (Fig. 2B).
3.6. Ethylene glycol
PsyHSL activity was dramatically affected by the presence of
ethylene glycol; in fact, we observed a decrease of activity with
respect to the control, in all tested conditions (Supplemental data,
Fig. S4A). Concerning the addition of ethylene glycol to the other
three enzymes, we observed that activity remained above the value
measured without solvent for Aes and EST2 up to 12.5% of ethylene
glycol (Fig. S4B); for AFEST this trend was maintained almost up to
20% of solvent (Supplemental data, Fig. S4B–D).
In terms of protein stability, in the presence of ethylene gly-
col, EST2 and Aes demonstrated to be very stable, with a half life
of 240 and 360 min, respectively at 10% of solvent, whereas AFEST
showed a moderate stability (Supplemental data, Table S4). Com-
paring the results obtained in ethylene glycol with that obtained
with methanol, it turned out that the less destabilizing effect we
observed in ethylene glycol could be ascribed to the lower “denat-
uration capacity” (Table 1) of the ethylene glycol with respect to
methanol.
3.9. Ester synthesis
With the aim to understand if the decrease in enzymatic activ-
ity, observed after incubation in organic solvents, was due to a

L. Mandrich et al. / Journal of Molecular Catalysis B: Enzymatic 82 (2012) 46–52
51
Table 4
The parameters were calculated using (R)- or (S)-p-nitrophenyl-2-chloropropionate as substrate, in a mixture containing sodium citrate buffer 50 mM pH 5.0 and 10% of each
solvents at 25 ◦C. Stock solution of the substrates dissolved in each solvent were prepared. The kinetic determinations were performed only for EST2.
Solvent
(R)-p-Nitrophenyl-2-chloropropionate
(S)-p-Nitrophenyl-2-chloropropionate
Ratio Ss/Rs
kcat (s−1
)
KM (mM)
kcat/KM (s−1 mM−1
)
kcat (s−1
)
KM (mM)
kcat/KM (s−1 mM−1
)
–
37.5 1.5
53.5 2.0
39.0 2.0
48.0 1.0
43.2 1.0
43.7 2.3
0.100 0.014
0.170 0.020
0.105 0.015
0.230 0.025
0.190 0.013
0.190 0.030
375 60
314 50
371 50
208 27
227 18
230 40
52.6 3.4
56.6 2.5
48.8 2.4
36.7 1.6
35.5 1.0
41.5 1.8
0.110 0.030
0.076 0.013
0.124 0.024
0.125 0.025
0.130 0.015
0.115 0.020
480 100
745 80
394 50
293 75
273 35
360 80
1.28 0.27
2.37 0.39
1.06 0.06
1.41 0.39
1.20 0.20
1.57 0.55
Acetonitrile
Dimethyl sulfoxide
Dimethyl formamide
Methanol
Ethylene glycol
denaturation process or to an inversion of the hydrolytic reaction,
the rates of ester synthesis were calculated.
Our results (Table 3) show an excellent rate of ester synthesis
for EST2 and AFEST in 5% methanol, while the increase of methanol
percentage (10%) corresponds to a drastic decrease of synthetic
reaction. Other esterases tested under the same conditions demon-
strated a very poor catalytic rate of synthesis, suggesting that the
presence of organic solvent led to the protein unfolding.
3.10. Enantioselectivity
To translate the information gathered on the solvents stabil-
ity into a reaction/process of biotechnological interest, we used
as a model reaction the EST2 hydrolysis of (RS)-p-nitrophenyl-2-
chloropropionate to obtain (S)-2-chloropropionic acid, a compound
used in the synthesis of optically pure phenoxypropionic acid her-
bicides [27]. Previously, it has been reported that EST2 is able to
hydrolyse this compound, showing a slightly preference for the S
form [26]; in the same work by a random mutagenic approach, con-
sisting in error prone PCR of the EST2 gene, expression of the genes
and screening for altered activities against the R or the S form of
the substrate we isolated a mutated version of EST2, L212P that
displayed about 6-fold increase of specificity for the S form [26]. In
Table 4 are reported the kinetic parameters measured for EST2 in
the different solvents, using the (R) or (S) forms of the substrate. In
the presence of acetonitrile we measured the maximum difference
in affinity for the S form, respect to the parameters measured with-
out solvent showing a ratio between the specific constants Ss/Rs in
acetonitrile of 2.37 (without solvent the ratio was 1.28, Table 4).
In presence of acetonitrile the parameters measured on the R form
of the substrate were the same obtained without solvent and the
difference was due to a decrease of KM value, corresponding an
increase of affinity for the S form. Regarding DMF an inversion of
ratio of activities was observed favouring the R form. Whereas using
DMSO a decrease of ratio Ss/Rs was measured respect to the value
without solvent (Table 4), due to a decrease of specific constant for
the S form, remaining unchanged for the R form. Finally we tested
the behaviour of the mutant L212P in acetonitrile, obtaining a slight
increase of the ratio Ss/Sr with respect the value measured without
solvents, from 6.6 to 7.4.
4. Conclusions
The study of behaviour of enzymes in different conditions such
as temperature, pH, presence of salts and/or organic solvents may
be a preparatory phase to understand the potential of enzymes in
biotechnological application. By now, many enzymes have been
described soluble in pure organic solvents, more stable there than
in water and show different catalytic properties, generally reduced
activity but altered specificity [39–41]. Systems made of homo-
geneous aqueous–organic solvent mixtures are used in catalysis
when the substrates are not completely soluble in water, and they
may alter the synthesis/hydrolysis equilibrium and the enantios-
electivity [42]. In this light, the nature of organic solvent used
causes changes in the properties of the reaction medium which
can determine different conformational flexibility, deformation of
the enzyme [43], and different solubility and polarization of the
substrates.
Fig. 2. Graphical representation of kinetic parameters reported in Table S6
(Supplemental data). Parameters were reported as ratio with respect to the
value obtained in water. (A) kcat solvent/kcat H2O values are reported; (B)
ssolvent/sH
values are reported. To better understand the real difference
O
in activi2ty among the enzymes used, we report the value measured in
H2O: PsyHSL kcat = 19.0 1.3 s−1
,
s = 190 30 s−1 mM−1
;
Aes kcat = 40.3 0.6 s−1
,
s = 260 10 s−1 mM−1
;
EST2 kcat = 3800 80 s−1
,
s = 475 55 (×103) s−1 mM−1
;
AFEST kcat = 1530 5 s−1, s = 352 23 (×103) s−1 mM−1
.

52
L. Mandrich et al. / Journal of Molecular Catalysis B: Enzymatic 82 (2012) 46–52
The ester hydrolysis reaction is a nucleophilic substitution
be taken into account when planning to study specific biotransfor-
mation reactions.
which involves a carbocation intermediate, anything that can facil-
itate this will speed up the catalytic reaction. The addition of
solvents to the reaction can stabilize ionic intermediates in the
case of aprotic solvents; and to solvate the leaving group in the
case of protic solvents, among these there are water and alco-
hols, which will also act as nucleophiles [37]. In this work, we
observed that ethylene glycol (polyprotic) is less destabilizing than
methanol for all proteins; in fact, the half-lives were higher with
respect to methanol (Tables S3 and S4), also the activity in the
presence of increasing percentage of solvent (v/v) was higher in
ethylene glycol than in methanol (Figs. S3 and S4). In the case of
aprotic solvents no direct correlation was observed, for instance,
while Aes showed an high half-life in both DMSO and DMF, EST2
half-life was similar to Aes in DMSO, but 10 fold less in DMF. In
terms of kinetic parameters (Table S6) the aprotic solvents gave
increased specific constants for PsyHSL and Aes due to essentially
an increase in kcat. In the case of EST2 and AFEST an increase of kcat
and KM values was observed, with consequent decrease of specific
constants.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.molcatb.2012.06.002.
References
[1] A.M. Klibanov, Trends Biochem. Sci. 14 (1989) 141–144.
[2] A.M. Klibanov, Adv. Appl. Microbiol. 29 (1983) 1–28.
[3] T.J. Ahern, A.M. Klibanov, Science 228 (1985) 1280–1284.
[4] S.E. Zale, A.M. Klibanov, Biochemistry 25 (1986) 5432–5444.
[5] G. Carrea, S. Riva, Angew. Chem. Int. Ed. Engl. 39 (2000) 2226–2254.
[6] A. Zacks, A.M. Klibanov, Science 224 (1984) 1249–1251.
[7] D.B. Volkin, A. Staubli, R. Langer, A.M. Klibanov, Biothechnol. Bioeng. 37 (1991)
843–853.
[8] K. Xu, K. Griebenow, A.M. Klibanov, Biotechnol. Bioeng. 56 (1997) 485–491.
[9] L.M. Simon, M. Kotorman, G. Garab, I. Laczko, Biochem. Biophys. Res. Commun.
280 (2001) 1367–1371.
[10] C. De Santi, M.L. Tutino, L. Mandrich, M. Giuliani, E. Parrilli, P. Del Vecchio, D.
de Pascale, Biochimie 92 (2010) 949–957.
The activity and stability of the tested enzymes was altered by all
solvents tested; from a kinetic point of view, dimethyl formamide,
dimethyl sulfoxide, methanol and ethylene glycol enhance the
activity (kcat) of all enzymes; parallelly it was observed a gen-
eral and slight increase in the KM values suggesting a reduction
of the substrate affinity. These effects are not unexpected because
organic solvents might change local or global polarity of the enzyme
and therefore they could stabilize the surface binding site for sub-
strate or they might remove or change the distribution of water
molecules in the active site that in turn could have an effect
on kcat and KM. At substrate level, the solvents might give some
conformational differences or have a mere inhibitory effect [28].
For the same reasons the organic solvents might have a different
effect on the enantiomer substrate (RS)-p-nitrophenyl chloropro-
pionate because the R or the S form could interact better with
EST2 enzyme and as consequence give a different rate of hydrol-
ysis or a different affinity. In the model case analysed here, the
mutation L212P concerns an exposed residue in the active site
likely to be involved in slight modifications [26], which in turn
help the S form of substrate to have a better interaction with
the enzyme ultimately leading to an increase of activity. These
alterations could be easily mimicked by the so called “medium
engineering” as reported here (Table 4). If we assume that for
L212P mutant the enzyme conformation is optimized for the
S form, the presence of organic solvents could have a nega-
tive or slight positive effect on the catalysis because the organic
solvent could interfere with the optimal active site conforma-
tion.
In this light, we could explain why we obtained only a slight
increase in affinity of S form in presence of acetonitrile for the
mutant L212P and the inversion of ratio of activity when dimethyl
formamide was used.
The highest protein stability of EST2 and AFEST in all tested
solvents, is in good agreement with previous knowledge about
properties of (hyper)thermophilic enzymes; accordingly PsyHSL
seemed to be less stable in presence of solvents, in agreement
with its lower thermal stability; in other words these two prop-
erties: thermal and solvent stability are strictly related suggesting
a common denaturation pathway in which molecular flexibility is
a fundamental determinant of enzyme denaturation, and worth to
[11] R. Peist, A. Koch, P. Bolek, S. Sewitz, T. Kolbus, W. Boos, J. Bacteriol. 179 (1997)
7679–7686.
[12] G. Manco, E. Adinolfi, F.M. Pisani, G. Ottolina, G. Carrea, M. Rossi, Biochem. J.
332 (1998) 203–212.
[13] G. Manco, E. Giosuè, S. D’Auria, P. Herman, G. Carrea, M. Rossi, Arch. Biochem.
Biophys. 373 (2000) 182–192.
[14] E. Krejci, N. Duval, A. Chatonnet, P. Vincens, J. Maussoliè, Proc. Natl. Acad. Sci.
U.S.A. 88 (1991) 6647–6651.
[15] D. Langin, H. Laurell, L. Stenson-Holst, P. Belfrage, C. Holm, Proc. Natl. Acad. Sci.
U.S.A. 90 (1993) 4897–4901.
[16] H. Hemilä, T.T. Koivula, I. Palva, Biochem. Biophys. Acta 1210 (1994) 249–253.
[17] T. Hotelier, L. Renault, X. Cousin, V. Negre, P. Marchot, A. Chatonnet, Nucleic
Acids Res. 32 (database issue) (2004) D145–D147.
[18] J.L. Arpigny, K.E. Jaeger, Biochem. J. 343 (1999) 177–183.
[19] R.F. Rekker, The Hydrophobic Fragmental Constant, Pharmacochemistry
Library, vol. 1, Elsevier, New York, 1977.
[20] Y.L. Khmelnitsky, V.V. Mozhaev, A.B. Belova, M.V. Sergeeva, K. Martinek, Eur. J.
Biochem. 198 (1991) 31–41.
[21] M.N. Gupta, R. Batra, R. Tyagi, A. Sharma, Biotechnol. Prog. 13 (1997) 284–288.
[22] C. Reichardt, Angew. Chem. Int. Ed. 18 (1979) 98–110.
[23] J.A. Riddick, W.B. Bunger, T.K. Sakano, Organic Solvents. Physical Properties and
Methods of Purification, 4th ed., Wiley, New York, 1986.
[24] K. Egorova, G. Antranikian, Review 8 (2005) 649–655.
[25] G. Manco, L. Mandrich, M. Rossi, J. Biol. Chem. 276 (2001) 37482–37490.
[26] G. Manco, G. Carrea, E. Giosuè, G. Ottolina, G. Adamo, M. Rossi, Extremophiles
6 (2002) 325–331.
[27] A. Kurata, M. Fujita, A.M. Mowafy, H. Kamachi, T. Kurihara, N. Esaki, J. Biosci.
Bioeng. 105 (2008) 429–431.
[28] G. Zandonella, L. Haalck, F. Spener, K. Faber, F. Faltauf, A. Hermetter, Eur. J.
Biochem. 231 (1995) 50–55.
[29] L. Mandrich, E. Caputo, B.M. Martin, M. Rossi, G. Manco, J. Biol. Chem. 277 (2002)
48241–48247.
[30] R.J. Leatherbarrow, Staines, Erithacus Software, UK, 1992.
[31] L. Mandrich, G. Manco, M. Rossi, E. Floris, T. Jansen-van den Bosch, G. Smit, J.A.
Wouters, Appl. Environ. Microbiol. 72 (2006) 3191–3197.
[32] P.J. Haney, J.H. Buldak, C.I. Reich, C.R. Wouse, G.J. Olsen, Proc. Natl. Acad. Sci.
U.S.A. 96 (1999) 3578–3583.
[33] C. Veille, G.J. Zeikus, Microbiol. Mol. Biol. Rev. 65 (2001) 1–43.
[34] C. Gerday, M. Aittaleb, J.L. Arpigny, E. Baise, J.P. Chessa, G. Garsoux, I. Petrescu,
G. Feller, Biochim. Biophys. Acta 1342 (1997) 119–131.
[35] T. Farias, L. Mandrich, M. Rossi, G. Manco, Protein Pept. Lett. 14 (2007) 165–169.
[36] T.H. Lowery, K.S. Richardson, Mechanism and Theory in Organic Chemistry, 3rd
ed., Harper Collins Publisher, New York, 1987.
[37] K.S. Peters, Chem. Rev. 107 (2007) 859–873.
[38] Q. Shao, Y. Fan, L. Yang, Y. Qin Gao, J. Chem. Phys. 136 (2012) 115101–115109.
[39] S. Torres, G.R. Castro, Food Technol. Biotechnol. 42 (2004) 271–277.
[40] D.M. Ruiz, R.E. De Castro, J. Ind. Microbiol. Biotechnol. 34 (2007) 111–115.
[41] L. Costas, V.E. Bosio, A. Pandey, G.R. Castro, Appl. Biochem. Biotechnol. 151
(2008) 578–586.
[42] G.R. Castro, T. Knubovets, Crit. Rev. Biotechnol. 23 (2003) 195–231.
[43] K. Watanabe, S. Ueji, Biotechnol. Lett. 22 (2000) 599–603.