Characterization and structure basis of Pseudomonas alcaligenes lipase's enantiopreference towards d,l-menthyl propionate

Article abstract of DOI:10.1016/j.molcatb.2014.01.020

In this >work, a lipase from Pseudomonas alcaligenes CGMCC4405 (PaL) was cloned and expressed. It was very attractive that the recombinant PaL exhibited excellent enantioselectivity (E > 200) in the resolution of racemic d,l-menthyl propionate to produce l-menthol. The structure basis of enantiopreference is a fundamental scientific problem which needs to be resolved. In our research, molecular dynamic simulation (MD) research was employed to research the different binding modes of d and l-menthyl propionate. The results showed that when bound with slow-reacting enantiomer (d-menthyl propionate), the steric requirements of the large substituent (isopropyl) of the d-menthyl propionate force a rotation of the imidazole ring of catalytic residue His271 and further pushed the active site His271 away from its proper orientation. Moreover, the average distance between alcohol oxygen (O _alc) and H_NEε of catalytic His271 increased to 3.7 A, which was too far to form an essential hydrogen bond and further prevented efficient catalysis of slow enantiomer. This correlation of the distance between alcohol oxygen (O_alc) and H _NEε of catalytic His271 and the enantioselectivity was also confirmed by the result of site-directed mutagenesis.

Full text of DOI:10.1016/j.molcatb.2014.01.020

Journal of Molecular Catalysis B: Enzymatic 102 (2014) 81–87
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Characterization and structure basis of Pseudomonas alcaligenes
lipase’s enantiopreference towards d,l-menthyl propionate
∗
∗
Hui Chen, Jianping Wu , Lirong Yang , Gang Xu
Institute of bioengineering, Department of Chemical and Biological Engineering, Zhejiang University, Hangzhou 310027, China
a r t i c l e i n f o
a b s t r a c t
Article history:
In this >work, a lipase from Pseudomonas alcaligenes CGMCC4405 (PaL) was cloned and expressed. It was
very attractive that the recombinant PaL exhibited excellent enantioselectivity (E > 200) in the resolution
of racemic d,l-menthyl propionate to produce l-menthol. The structure basis of enantiopreference is a
fundamental scientific problem which needs to be resolved. In our research, molecular dynamic simula-
tion (MD) research was employed to research the different binding modes of d and l-menthyl propionate.
The results showed that when bound with slow-reacting enantiomer (d-menthyl propionate), the steric
requirements of the large substituent (isopropyl) of the d-menthyl propionate force a rotation of the
imidazole ring of catalytic residue His271 and further pushed the active site His271 away from its proper
orientation. Moreover, the average distance between alcohol oxygen (Oalc) and HNε of catalytic His271
Received 1 November 2013
Received in revised form 3 January 2014
Accepted 23 January 2014
Available online 6 February 2014
Keywords:
PaL
Enantiopreference
l-menthol
Molecular dynamics simulation
Site-directed mutagenesis
increased to 3.7
A˚ , which was too far to form an essential hydrogen bond and further prevented efficient
catalysis of slow enantiomer. This correlation of the distance between alcohol oxygen (Oalc) and HN␧ of
catalytic His271 and the enantioselectivity was also confirmed by the result of site-directed mutagenesis.
©
2014 Elsevier B.V. All rights reserved.
1
. Introduction
catalytic region [8,9]. The empirical rule helps researchers pre-
dict favored enantiomers in the lipase-catalyzed new reactions.
However, it does not provide detailed description of structure
basis of enantiomers recognition. Since then, this subject has been
widely studied. Hæffner et al. [10] proposed a calculated model
based on the energy of the interaction between the specific residue
or water molecule and the transition state of substrate to pre-
dict the fast-reacting enantiomer qualitatively. The research of
Raza et al. [11] focused on the potential energies of a number
of subsets within the model systems and identified the specific
regions which based on atoms that make up the core structure
elements of the transition state are important for the predic-
tion of the enantioselectivity by simulation method. The entropic
component of the free energy of activation is as important as
the enthalpic components for the enantioselectivity [12,13]. The
research of Ottosson et al. revealed the relation between the
substrate-related component of activation and the spatial free-
dom of the substrate in transition state by MD simulation. Some
other new approaches besides MD simulation such as substrate
imprinting reaction method [14], substrate-imprinted docking [15]
or QM/MM [16] were also employed to investigate the recognition
mechanism of enantiomers and achieved good effects.
Lipases (glycerol ester hydrolases, EC 3.1.1.3) mainly catalyzes
the hydrolysis of triglycerides and several other substrates con-
taining ester bonds [1–3]. They can also perform other kinds of
bioconversion, such as esterification, transesterification, alcoholy-
sis, acidolysis and aminolysis [2]. Currently, enantiomerically pure
compounds are of rapidly increasing importance to the chemical
industrial as they are widely used in pharmaceuticals, agrochem-
icals, flavors and fragnences [4–6]. Biocatalysis based on the
enzymatic resolution of racemic mixtures remains an important
method of choice [7]. In particular, lipases are among the most
employed biocatalysts in organic synthesis to catalyze the kinetic
resolution of a wide range of substrates and yield optically pure
compounds, as they are stable, do not require cofactors, have a
broad substrate spectrum and frequently exhibit high activities as
well as chemo-, regio- or stereoselectivity [5,7].
For the enantiomers recognition of lipase towards secondary
alcohols, there is an empirical “Kazlauskas” rule in which enan-
tiomers recognition relies on the match between the size of
substituents at the stereo center and the size of pocket in the
The former researches [17,18] indicated that the geometric
parameter d(HN␧–Oalc) was an essential influencing factor for the
enantiomers discrimination. The research of Ema [19] showed the
enantioselectivity resulted from the suppression mechanism work-
ing on the slower-reacting enantiomer in the transition state. The
∗
Corresponding authors at: Zhejiang University, Institute of bioengineering,
Department of Chemical and Biological Engineering, Hangzhou, Zhejiang 310027,
China. Tel.: +86 571 87952363; fax: +86 571 87952363.
E-mail addresses: wjp@zju.edu.cn (J. Wu), lryang@zju.edu.cn (L. Yang).
1
381-1177/$ – see front matter © 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcatb.2014.01.020

8
2
H. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 81–87
Scheme 1. Resolution of racemic d,l-menthyl propionate to produce l-menthol catalyzed by PaL.
◦
unfavourable enantiomer cannot pass over the transition state at
all because of very high activation energy. Although some attempts
have been performed, the uniform conclusion is still not been
achieved. The structure basis of enantiopreference is problem
which still needs further research. Consequently, the method of
directed evolution is often employed to improve lipase enantios-
elecvitity in most cases [20–23] due to it does not require the
knowledge of structure basis of enantiopreference [24].
at OD600 = 0.8 for 10 h at 18 C. After induction, 100 mL volumes
cells were collected by centrifugation and resuspened by 20 mL
Gly-NaOH buffer (50 mM pH = 9.0), then disrupted by ultrasound.
FF
Affinity chromatograph of HisTrapTM column was employed to
purify PaL. The concentration of imidazole in binding buffer and
elution buffer is 50 mM and 250 mM, respectively.
2
.3. PaL molecular weight determination and identification by
Recently, we obtained a novel extracellular lipase (PaL) from
Pseudomonas alcaligenes CGMCC4405 which was isolated from soil
samples. PaL gene (pal) had been cloned by the short gun cloning
method and expressed in heterologous host Escherichia coli BL21
peptide mass fingerprinting
The purified PaL was fractionated by means of SDS-PAGE.
The protein band between 40 kD and 60 kD was excised from
the Coomassie brilliant blue-stained SDS–PAGE gel. The PaL was
recovered and sent for matrix-assisted laser desorption ionization
time-of-flight (MALDI–TOF) to determine the precise molecular
weight.
(
DE3) [25]. PaL exhibited excellent enantiopreference in selective
hydrolyzing racemic d,l-menthyl propionate to produce l-menthol
(
(1R,2S,5R)-(−)-2-isopropyl-5-methylcyclohexanol), which was a
kind of important fine chemicals [26] (Scheme 1).
In this study, the structure basis of enantiopreference of PaL in
the resolution of d,l-menthyl propionate was the research focus. On
the basis of structure characterization of PaL, molecular dynamics
For peptide mass fingerprinting, the PaL was dissolved in
7
5 ␮L of denaturing buffer (6 M guanidine hydrochloride, 0.5 M
◦
Tris hydrochloride, 2 mM EDTA, pH8.0) and incubated at 37 C for
(
MD) simulation was employed to get insights into the structural
3
1
0 min. The denatured protein solution was diluted 1:1(v/v) with
00 mM ammonium hydrogencarbonate (pH = 7.8) and 20 ng of
differences of Pal-l–menthyl propionate and Pal-d-menthyl pro-
pionate complexes. The essential role of the geometric parameter
d(HN␧–Oalc) in the enantiomers discrimination and the influencing
factors which caused the changes of the geometric parameter were
also investigated and identified. In order to verify the results of MD
simulation, site-directed mutagenesis guided by MD simulation
was employ to regulate the enantioselectivity (E value).
◦
trypsin were added. Digestion was performed at 37 C overnight.
Finally, the sample was dried by means of a vacuum concentrator
centrifuge and dissolved in 20 ␮L of acetonitrile −0.1% TFA (1:9, v/v)
for subsequent mass spectrometric analysis. ␣-Cyano-4-hydroxy
cinnamic acid was used as matrix and ions were extracted into
the linear TOF using an extraction potential of 20 kV in high-mass
detection mode [28,29].
2
. Materials and methods
Peptide mass fingerprinting data was also employed to search
in the NCBI nonredundant protein databases using Mascot soft-
ware from Matrix Science (http://www.matrix-science.com). One
missed cleavage of trypsin per peptide, an accuracy level of
100 ppm and a mass tolerance of 1 Da were allowed.
2.1. Strains, vectors and chemicals
P. alcalignens CGMCC4405 was isolated from soil. E.coli BL21
(
DE3) and vector pET30 (+) (Invitrogen) were used for pal expres-
sion. The DNA purification kit, restriction endonucleases, T DNA
2.4. Circular dichroism (CD) measurements
4
TM
ligase, and PrimerSTAR
DNA polymerase were purchased from
CD spectra were recorded on a JASCO® J-715 spectropolarime-
ter. The far-UV spectra was taken at a protein concentration of
0.2 mg/mL in Gly-NaOH buffer (50 mM, pH = 9.0) using a 1-mm
path length cell. All spectra were collected from 190 nm to 240 nm
for far-UV spectra. The background noise was corrected against
Gly–NaOH buffer (50 mM, pH = 9.0) blank. The spectra of pro-
tein samples were subtracted by blanks. Molecular ellipticity was
TaKaRa (Japan). The racemic d,l-menthol was purchased from
Sigma (St. Louis, MO). The corresponding racemic d,l-menthyl pro-
pionate was synthesized by our own lab following the method
described by Croteau [27]. All other chemicals were of analytical
grade and commercially available.
2.2. Cloning, expression and purification of PaL
determined as [Â] = [100 × (MRW) × Â /(c × l)], where Â
is the
ꢀ
obs
obs
observed ellipticity in degrees at a given wavelength, c is the pro-
tein concentration in mg/mL and l is the length of the light path in
cm. Noise in the data was smoothed using JASCO® J-715 software,
including the fast Fourier transform noise reduction routine that
allows enhancement of most noisy spectra without distorting their
peak shapes [30].
According to the DNA sequence of PaL (Genbank: JN674069) two
ꢀ
ꢀ
oligonucleotide primers: forward F, 5 - CCG GAA TTC GTT TCT TAC
ꢀ
ATG TTG C TC AGC A-3 ; reversed R, 5 -CGC CTC GAG TGC ATT CAT
ꢀ
TCA GAA ACT GCC A-3 were designed and added the restriction site
of EcoR I and Xhol I, respectively. The amplified 1.5 kb fragments
and the plasmids pET30a (+) were digested by EcoR I and Xhol I and
then ligated by T DNA ligase. Then, the recombinant pET30-pal was
4
transformed into E.coli BL21 (DE3) component cells.
2.5. Gas chromatograph GC analysis
The transformants grew overnight in Luria–Bertani (LB)
media containing 50 ␮g/mL Kanamycin at 37 C. Cultured cells
◦
◦
The reaction was realized in 35 C on a thermostatic rotator
induced by 0.5 mM isopropyl-␤-D1-thiogalactopyranoside (IPTG)
(Eppendorf, Germany) with the speed 200 rpm. The substrate and

H. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 81–87
83
product were extracted from the reaction mixture with hexane,
dried with Na SO and then analysis by GC. The samples from
the hydrolysis reaction were injected in to a Agilent 6890 GC
ensemble from 0 to 300 K over 100 ps. All MD productions were run
using Sander module. The simulations were carried out for a total
of 8 ns at constant temperature (300 K) and pressure (1 bar) condi-
tions. Trajectories analysis was carried out using the Ptraj module
of Amber11 package. The root mean of square of deviation (RMSD)
was calculated for the lipase protein backbone atoms using least
square fit. The average distance between alcohol oxygen (Oalc) and
HN␧ of catalytic His271 (d(HN␧–Oalc)), the dihedral angle of His271
imidazole ring was calculated by using the coordinates of last 5 ns
trajectories.
2
4
instrument (Agilent, Inc, USA) equipped with CP-cyclodextrin ␤-
2
,3,6-M-19,FS 50X.25.(.25) column (50 m × 0.25 mm, 0.25 ␮m film
thickness) and a flame-ionization detector for analysis. N was used
2
as carrier gas. The column, injection and detector temperature is
◦ ◦ ◦
1
30 C 260 C and 250 C, respectively. n-dodecane was used as
internal standard.
Enantiomeric excess (eep) of product was calculated as defined
blow (Eq.(1)), where [l-menthol] and [d-menthol] were the con-
centration of l-menthol and d-menthol, respectively.
2.9. Site-directed mutagenesis
eep = [
L-menthol] − [D-menthol]
(1)
The QuickChangeTM site-directed mutagenesis method (Strata-
[
L-menthol] + [D-menthol]
gene, La Jolla, CA) was used to introduce point mutations into the
PaL sequence under the recommended condition in the manual of
E value was calculated as defined Eq. (2), where c value was the
conversion ratio.
TM
PrimerSTAR . The following primers and their complementary
ln[1 − c(1 + eep)]
reverse counterparts contained the desired mutations were used
to construct the variant enzymes. Mutated plasmids were verified
by DNA sequencing.
E =
(2)
ln[1 − c(1 − eep)]
2
.6. Model building
ꢀ
The site-directed mutagenesis at position S156: 5 GCC-
ꢀ
GTGGCCGCGCTCXXXGGCTGGGCCGATCTG-3 with XXX being ACA
A 3D model of WT PaL was constructed using the APM module
for S156 T, CTT for S180L, CAT for S156H.
of the Sybyl-X1.1 program (Tripos, Saint Louis, USA). This mod-
ule employs the fold recognition method, which is based on the
limitation of the folding type of the protein [31]. In this process, a
library of known 3D structures was searched to determine the fold
that gave the best alignment with the sequence of interest. In our
approach, the backbone of the structurally conserved regions was
constructed according to the template and the remaining parts of
backbone were constructed with Loop Search command in Sybyl.
Side chains were then added. Finally, all hydrogen atoms were
added and the charges were calculated. Several cycles of energy
minimization were carried out using the Tripos force field. The final
model validated using the PROCHECK program [32].
3
. Results and discussion
3.1. Molecular weight determination and peptide mass
fingerprinting of purifid PaL
PaL was purified by using HisTrapTMFF column (GE Healthcare)
affinity chromatography. The purified PaL showed a single band
with molecular mass about 50 to 60 kD in SDS-PAGE. To deter-
mine the precise molecular weight, MALDI-TOF mass spectra was
employed. The result showed that the precise molecular weight
of PaL was 58094.3 Da. In order to further determine whether the
amino acid sequence of PaL matched the one predicted from DNA
sequencing, MALDI-TOF mass of tryptic digest PaL was character-
ized to identify the recombinant protein. By comparing the mass
of identified peptide to the hypothetical tryptic peptides for the
proteins in non-redundant NCBI database using MASCOT search
engine, PaL was obviously identified with MOWSE score of 186.
The identified 13 peptide fragments matched of the deduced amino
acid sequence of PaL accounted for 55% peptide mass fingerprint
sequence coverage (Fig. 1). The result indicated that the amino acid
sequence deduced from DNA sequencing was reliable.
2.7. Covalent docking of substrate
The tetrahedral intermediates (TIs) of d and l-menthyl pro-
pionate were generated using Sybyl-X 1.1. Starting from the
initial protein and substrate structure, the conformation space
accessible by the substrate covalently bound to the catalytic ser-
ine residue (S133) in a tetrahedral form was determined using
the covalent docking program Flex X. All parameters were set
to the standard value. The resulting conformations of the two
enantiomers were checked for the existing of five key compo-
nent H-bonds. These H-bonds were characteristic for TIs, and
included O␦2(Asp243)· · ·H␦1(His271), H␧(His271)· · ·O␥(Ser133),
3.2. CD measurements
−
H␧(His271)· · ·O(alcohol), H-bonds between oxyanion (O ) of sub-
The far-UV CD spectrum of PaL was presented in Fig. 2. The spec-
strate and hydrogen atoms connected to the backbone nitrogen of
Tyr134 and indolyl nitrogen of Trp62. If the binding mode contained
all the five hydrogen bonds, it was considered as the productive
binding mode.
trum was typical for ␣/␤ fold hydrolases, which comprise mainly
-helix and ␤-sheet protein. It contains 26.8% ␣-helix, 34.2% ␤-
sheet, 14.2% turn and 24.7% random coil.
␣
3
.3. Selective hydrolysis of racemic d,l-menthyl propionate to
2.8. MD simulation
produce l-menthol
The productive binding modes were subjected to MD simu-
Chiral GC analysis of sample at different conversion ratio
revealed the amount of d-menthyl propionate remained apparently
constant during the course of reaction. PaL had essentially only
cleaved the l-menthyl propionate. As shown in Fig. 3, the enan-
tiomeric excess of l-menthyl propionate was more than 99%, when
the conversion ratio achieved 49.5%. Accordingly, the E value was
calculated to be more than 200 for the PaL. This indicated that PaL
was a ideal biocatalyst for the resolution of racemic d,l-menthyl
propionate to produce l-menthol selectively.
lation. All MD calculations were performed using the Amber11
software package with the ff99SB forcefield. In the simulation sys-
tem, a TIP3P water box was added using the LEaP module integrated
+
in the Amber11 package. Seven Na cations were also added to
neutralize the molecular system. Electrostatic interaction was cal-
culated using the Particle-Mesh Ewald method with a non-bonded
cutoff of 12 A˚ . The system was minimized by 500 steps of steep-
est descent followed by 2000 steps of conjugate gradient. After the
minimization, the system was gradually heated in the canonical

8
4
H. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 81–87
Fig. 1. MALDI-TOF spectra of tryptic digest PaL. The score, amino acid sequence coverage and the number of identified peptides are shown. Matched peptide sequences are
shown in red. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
3
.4. Modelling of PaL
about 3% fell into generously allowed and disallowed region of
the Ramachandran’s plot. Analyzing the PaL and aligned with tem-
plate, the catalysis residues were determined to be Ser133, Asp243
and His271, and the residues constructed oxyanion hole was Trp62
and Tyr134 (Fig. 4B). In order to verify the accuracy of the lipase
model, residues of catalysis and oxyanion hole were mutated to
Ala, respectively [34]. The results indicated that the replacement
of the five important active residues with Ala caused the loss of
hydrolysis activity (Table S1, Table S2). These results demonstrated
the five residues acted as active triads and the model of PaL was
reliable.
Our target was to obtain a three-dimensional (3D) model of
PaL in order to shed some light on the structure basis of excellent
enantioselectivity. A cocaine esterase (PDB entry: 1ju3, sequence
similarity 35.6%) obtained from Rhodococcus sp. strain MB1 [33]
was employed as modelling template by APM module of Sybyl-X
1
.1 to construct 3D structure of PaL (Fig. 4A). Z-Score of the model
is 69.1, which indicated the model is certain. (A certain model
requires Z-Score > 6.0). The predicted model was then checked for
main chain conformation using PROCHECK software implemented
in SWISS-MODEL online. The geometry analysis showed that ini-
tially closed to 97% of the ꢁ–ꢂ plots were in core region, and
Fig. 2. The far-UV CD spectra for the PaL.
Fig. 3. The relationship between conversion ratio and eep of l-menthol.

H. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 81–87
85
Fig. 4. The PaL lipase model. (A) Overall representation of PaL. (B) The active residues of PaL.
3
.5. Molecular dynamic (MD) simulation of Pal-substrate
enantioselecitvity directly. Some comparison of different sub-
strates through kinetic studies had demonstrated that differences
in the conversion of enantiomers do not result from enhanced
reactivity of fast-reacting enantiomers but from reduced reac-
complexes
Quantum-chemical method had identified the catalytically
important distance d(HN␧–Oalc) as a crucial to catalytic activity
35,36]. The hydrogen-bonding network, above all the dis-
tance d(HN␧–Oalc) for both enantiomers, should be relevant to
tivity of slow-reacting enantiomers [37]. So, the
E value
[
correlated with the d(HN␧–Oalc) of slow-reacting enantiomer
[38].
Fig. 5. MD simulation results. (A) The binding modes of l-menthyl propionate and d-menthyl propionate. (B) The d(HN␧–Oalc) of PaL binding with l-menthyl propionate
black) and d-menthyl propionate (red). (C) The dihedral angel of PaL binding with l-menthyl propionate (black) and d-menthyl propionate (red). (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)
(

8
6
H. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 81–87
Table 1
The relationship between d(HN␧–Oalc) of PaL-d-menthyl propionate complex and E
value of PaL.
PaL variant
d(HNε–Oalc) (Å)
E value
WT
3.7
3.2
2.9
2.6
>200
>200
168
52
S156T
S156L
S156H
3.6. Verifying the role of d(HN␧–Oalc) in E value by site-directed
mutagenesis
In practical research, the S156 was identified as a sensi-
tive spot for the enantioselectivity of PaL. Firstly, the geometric
parameter d(HN␧–Oalc) of three variant PaLs complexed wth slow-
reacting enantiomer was calculated by MD simulation. Then, the
corresponding variant PaLs were constructed by site-directed
mutagenesis and the E values of mutant PaLs were also detected.
The result verified the model obtained above and confirmed the
important role of d(HN␧–Oalc) in E value. As shown in Table 1, the
d(HN␧–Oalc) of slow-reacting enantiomer has obvious correlation
with the E value. For the WT and variant S156T, d(HN␧–Oalc) of
PaL-d-menthyl propionate complexes were 3.7 A˚ and 3.2 A˚ , which
was too far to form a productive hydrogen bond [40] and the E
values of them were very high (>200). Conversely, for the variant
S156L and S156H, d(HN␧–Oalc) of PaL-d-menthyl propionate com-
plex decreased to 2.9 A˚ and 2.5 A˚ , respectively, which were short
Fig. 6. Calculated B-factor of WT PaL residues from MD simulation. For binding with
l-menthyl propionate (black), for binding with d-menthyl propionate (red). (For
interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)
The models of PaL-l-menthyl propionate and PaL-d-menthyl
propionate complex were subjected to 8 ns MD simulation under
constant temperature and pressure conditions which mentioned
above. The distance d(HN␧–Oalc) was calculated (Fig. 5B). For fast-
reacting enantiomer (l-menthyl propionate), the d(HN␧–Oalc) was
about 2.2 A˚ . A short distance d(HN␧–Oalc) for fast-reacting enan-
tiomer should promote a fast conversion of l-menthyl propionate.
Otherwise, Distance d(HN␧–Oalc) for slow-reacting enantiomer (d-
menthyl propionate) had an average value around 3.7 A˚ , which was
too far to form a productive hydrogen bond between HN␧ of cat-
alytic histidine and alcohol oxygen (Oalc) of substrate (Fig. 5A) and
prevented the efficient catalysis of slow-reacting enantiomer. In the
research of Schelz et al. [18], the enantioselectivity of lipase could
be predicted by the distance d(HN␧–Oalc) of slow-reacting enan-
tiomer. The formation and strength of the hydrogen bond between
HN␧ of catalytic histidine and alcohol oxygen (Oalc) of substrate was
also the essential factor for the enantioselectivity.
B-factor can provide information about local flexibility. Higher
B-factor indicates higher structural flexibility [39]. To investi-
gate changes in backbone flexibility, MD simulation was used
to calculate the B-factors of PaL bound with each of l-menthyl
propionate and d-neomenthyl propionate. As shown in Fig. 6,
the binding of d-menthyl propionate lead to obvious variation
of flexibility around the catalytic residue site His271 (residues
enough to form a productive hydrogen bond. The E values of the
two variant PaL also decreased to 158 and 52, respectively. This
result verified the crucial role of d(HN␧–Oalc) of PaL-slow-reacting
enantiomer in E value of PaL and confirmed the structure basis of
enantioselectivity obtained from the simulation of WT PaL.
4. Conclusion
All the results of this work revealed that the geometric parame-
ter d(HN␧–Oalc) of slow-reacting enantiomer is a crucial influencing
factor of enantiopreference. Upon binding of d-menthyl propio-
nate, the steric requirements of the large substituent of the alcohol
form a rotation of the imidazole ring of His271, disrupting the
hydrogen bond to the alcohol oxygen. The loss of this hydrogen
bond is likely the cause of the differences in reactivity of the two
enantiomers. This result should be a breach for people to research
and understand the structure basis of enantiopreference.
2
65–280). Compared with Pal-l-menthyl propionate complex, the
B-factors of the regions surrounding the catalytic residue His271 of
PaL-d-menthyl propionate complex increased obviously. B-factor
simulation showed the particularly high flexibility in this region. It
also revealed that the binding of d-menthyl propionate introduced
disturbance effect to the region around His271 and make His271 be
in an unstable state. Moreover, the floating rate the d(HN␧–Oalc) of
the PaL-d-menthyl propionate complex was larger than that of PaL-
l-menthyl propionate complex. This result corroborated the distur-
bance effect introduced by the binding of d-menthyl propionate.
As shown in Fig. 5C, when PaL bound with d-menthyl propio-
Acknowledgements
This work was supported by National Basic Research Program
of China (973 Program, No. 2011CB710800), Hi-Tech Research and
Development Program of China (863 Program, 2011AA02A209),
National Natural Science Foundation of China (No. 20936002), Sci-
ence and Technology Planning Project of Zhejiang Province (No.
2
010C31127)
Appendix A. Supplementary data
◦
◦
nate, the imidazole ring of His271 rotated about 30 (from −75
◦
to 115 ) compared with the binding of l-menthyl propionate. This
rotation enlarged the d(HN␧–Oalc) and further broken the hydrogen
bond between alcohol oxygen (Oalc) of d-menthyl propionate and
HNε of catalytic His271. Analyzing the conformation, the isopropyl
substituent of d-menthyl propionate pointed toward the imidazole
ring of His271. Upon binding of the slow reacting enantiomer, the
steric requirements of the large substituent of the d-menthyl pro-
pionate force a rotation of the imidazole ring of His271, and further
disrupting the hydrogen bond to the alcohol oxygen.
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcatb.
2
014.01.020.
References
[
[
1] U.T. Bornscheuer, FEMS Microbiol. Rev. 26 (2002) 73–81.
2] H. Ogino, S. Hiroshima, S. Hirose, M. Yasuda, K. Ishimi, H. Shikawa, Mol. Gen.
Genomics 271 (2004) 189–196.

H. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 81–87
87
[
[
[
[
[
3] J. Krzeslak, G. Gerritse, R. van Merkerk, R.H. Cool, W.J. Quax, Appl. Environ.
Microbiol. 74 (2008) 1402–1411.
4] K. Liebeton, A. Zonta, K. Schimossek, M. Nardini, D. Lang, B.W. Dijkstra, M.T.
Reets, K-E. Jaeger, Chem. Biol. 7 (2000) 709–718.
5] D. Guieysse, J. Cortés, S. Puech-Guenot, S. Barbe, V. Lafaquiére, P. Monsan, T.
Siméon, I. André, M. Remaud-Siméon, ChemBioChem 9 (2008) 1308–1317.
6] J.S. Carey, D. Laffan, C. Thomson, M.T. Williams, Org. Biomol. Chem. 4 (2006)
[22] S.A. Funke, A. Eipper, M.T. Reetz, N. Otte, W. Thiel, G.van. Pouderoyen,
B.W. Dijkstra, K.-E. Jaeger, T. Eggert, Biocatal. Biotransform. 21 (2003)
67–73.
[23] K. Engström, J. Nyhlén, A.G. Sandström, J-E. Bäckvall, J. Am. Chem. Soc. 132
(2010) 7038–7042.
[24] U.T. Bornscheuer, Microbiol. Rev. 26 (2006) 73–81.
[25] M.M. Sun, H. Chen, J.P. Wu, L.R. Yang, Microbiol. Chin. 39 (2012) 1580–1588.
[26] G.W. Zheng, H.L. Yu, J.D. Zhang, J.H. Xu, Adv. Synth. Catal. 351 (2009) 405–414.
[27] R. Croteau, C.L. Hooper, Plant Physiol. 61 (1978) 737–742.
[28] Z.M. Zehl, L. Ivana, A. Marija, R. Andreas, K.P. Biserka, J. Mass Spectrom. 39
(2004) 1474–1483.
[29] C. Chang, X. Liu, K. Chen, Adv. Biosci. Biotechnol. 2 (2011) 123–127.
[30] K. Boren, P. Andersson, M. Larsson, U. Carlsson, Biochim. Biophys. Acta 1430
(1999) 111–118.
2
337–2347.
7] S. Tomi c´ , B. Berto sˇ a, B. Koji c´ -Prodi c´ , I. Kolosvary, Tetrahedron: Asymmetry 15
2004) 1163–1172.
(
[
[
8] S. Tanja, D.S. Rolf, P. Jurgen, J. Mol. Model. 7 (2001) 265–270.
9] R.J. Kazlauskas, A.N.E. Weissfloch, A.T. Rappaport, L.A. Cuccia, J. Org. Chem. 56
(
1991) 2656–2665.
[
[
[
[
[
10] F. Hæffner, T. Norin, K. Hult, Biophys. J. 74 (1998) 1251–1262.
11] S. Raza, L. Fransson, K. Hult, Protein Sci. 10 (2001) 329–338.
12] J. Ottosson, L. Fransson, K. Hult, Protein Sci. 11 (2002) 1462–1471.
13] R. Kazlauskas, Biocatal. Biotransfrom. 4 (2000) 81–88.
[31] A. Imberty, V. Piller, F. Piller, C. Breton, Protein Eng. 10 (1997) 1353–1356.
[32] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, J. Appl. Crystallogr.
16 (1993) 283–291.
14] L. Chaput, Z. Marton, P. Pineau, L. Domon, M. Graber, V. Tran, J. Mol. Catal. B:
Enzym. 84 (2012) 55–61.
[33] N.A. Larsen, J.M. Turner, J. Stevens, S.J. Rosser, A. Nasran, R.A. Lerner, N.C. Bruce,
I.A. Wilson, Nat. Struct. Biol. 9 (2002) 17–21.
[
[
[
15] P.B. Juhl, P. Trodler, S. Tyagi, J. Pleiss, BMC Struct. Biol. 9 (2009) 1–17.
16] M.P. Furshicheva, A. Warshel, ChemBioChem 13 (2012) 215–223.
17] M. Cygler, P. Grochulski, R.J. Kazlauskas, J.D. Schrag, F. Bouthillier, B. Rubin, A.N.
Serreqi, A.K. Gupta, J. Am. Chem. Soc. 116 (1994) 3180–3186.
[34] A. Tramontano, Methods 14 (1998) 293–300.
[35] C. Hu, T. Brinck, K. Hult, Int. J. Quantum Chem. 69 (1998) 89–103.
[36] P. Monecke, R. Friedemann, S. Naumann, R. Csuk, J. Mol. Model. 4 (1998)
395–404.
[
[
18] T. Schulz, J. Pleiss, R.D. Schmid, Protein Sci. 9 (2000) 1053–1062.
19] T. Ema, K. Yamaguchi, Y. Wakasa, A. Yabe, R. Okada, M. Fukumoto, F. Yano, T.
Korenage, M. Utaka, T. Sakai, J. Mol. Catal. B: Enzym. 22 (2003) 181–192.
20] K-E. Jaeger, T. Eggert, Curr. Opin. Biotechnol. 15 (2004) 305–313.
21] M.T. Reetz, M. Puls, J.D. Carballeira, A. Vogel, K-E. Jaeger, T. Eggert, W. Thiel, M.
Bocola, N. Otte, ChemBioChem 8 (2007) 106–112.
[37] T. Ema, J. Kobayashi, S. Maeno, T. Sakai, M. Utaka, Bull. Chem. Soc. Jpn. 71 (1998)
443–453.
[38] T. Schulz, R.D. Schmid, J. Pleiss, J. Mol. Model. 7 (2001) 265–270.
[39] A. Schlessinger, B. Rost, Proteins 61 (2005) 115–126.
[40] W.V. Tuomi, R.J. Kazlauskas, J. Org. Chem. 64 (1999) 2638–2647.
[
[