Improving the activity and enantioselectivity of PvEH1, a Phaseolus vulgaris epoxide hydrolase, for o-methylphenyl glycidyl ether by multiple site-directed mutagenesis on the basis of rational design

Article abstract of DOI:10.1016/j.mcat.2019.110517

Substrate spectrum assay exhibited that PvEH1, which is an epoxide hydrolase from P. vulgaris, had the highest specific activity and enantiomeric ratio (E) for racemic o-methylphenyl glycidyl ether (rac-1) among tested aryl glycidyl ethers (1–5). To produce (R)-1 via kinetic resolution of rac-1 efficiently, the catalytic properties of PvEH1 were further improved on the basis of rational design. Firstly, the seven single-site variants of PvEH1-encoding gene (pveh1) were PCR-amplified as designed, and expressed in E. coli BL21(DE3). Among all expressed single-site mutants, PvEH1^L105I and PvEH1^V106I had the highest specific activities of 17.6 and 16.4 U/mg protein, respectively, while PvEH1^L196D had an enhanced E value of 9.2. Secondly, to combine their respective merits, one triple-site variant, pveh1^{L105I/V106I/L196D}, was also amplified, and expressed. The specific activity, E value, and catalytic efficiency of PvEH1^{L105I/V106I/L196D} were 23.1 U/mg, 10.9, and 6.65 mM^?1 s^?1, respectively, which were 2.0-, 1.8- and 2.4-fold higher than those of wild-type PvEH1. The source of PvEH1^{L105I/V106I/L196D} with enhanced E value for rac-1 was preliminarily analyzed by molecular docking simulation. Finally, the scale-up kinetic resolution of 100 mM rac-1 was conducted using 5 mg wet cells/mL E. coli/pveh1^{L105I/V106I/L196D} at 25 °C for 1.5 h, producing (R)-1 with 95.0% ee_s, 32.1% yield and 3.52 g/L/h space-time yield.

Full text of DOI:10.1016/j.mcat.2019.110517

MolecularCatalysis476(2019)110517
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Improving the activity and enantioselectivity of PvEH1, a Phaseolus vulgaris
epoxide hydrolase, for o-methylphenyl glycidyl ether by multiple site-
directed mutagenesis on the basis of rational design
Chuang Lia,1, Ting-Ting Kanb,1, Die Huc, Ting-Ting Wangb, Yong-Jun Sub, Chen Zhangb,
Jian-Qing Chengc, Min-Chen Wuc,⁎
a Key Laboratory of Carbohydrate Chemistry and Biotechnology, Ministry of Education, School of Biotechnology, Jiangnan University, Wuxi, 214122, China
b School of Pharmaceutical Science, Jiangnan University, Wuxi, 214122, China
c Wuxi School of Medicine, Jiangnan University, Wuxi, 214122, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
PvEH1
Catalytic property
Site-directed mutagenesis
Rational design
Kinetic resolution
Substrate spectrum assay exhibited that PvEH1, which is an epoxide hydrolase from P. vulgaris, had the highest
specific activity and enantiomeric ratio (E) for racemic o-methylphenyl glycidyl ether (rac-1) among tested aryl
glycidyl ethers (1–5). To produce (R)-1 via kinetic resolution of rac-1 efficiently, the catalytic properties of
PvEH1 were further improved on the basis of rational design. Firstly, the seven single-site variants of PvEH1-
encoding gene (pveh1) were PCR-amplified as designed, and expressed in E. coli BL21(DE3). Among all expressed
single-site mutants, PvEH1L105I and PvEH1V106I had the highest specific activities of 17.6 and 16.4 U/mg protein,
respectively, while PvEH1L196D had an enhanced E value of 9.2. Secondly, to combine their respective merits,
one triple-site variant, pveh1L105I/V106I/L196D, was also amplified, and expressed. The specific activity, E value,
and catalytic efficiency of PvEH1L105I/V106I/L196D were 23.1 U/mg, 10.9, and 6.65 mM−1 s−1, respectively, which
were 2.0-, 1.8- and 2.4-fold higher than those of wild-type PvEH1. The source of PvEH1L105I/V106I/L196D with
enhanced E value for rac-1 was preliminarily analyzed by molecular docking simulation. Finally, the scale-up
kinetic resolution of 100 mM rac-1 was conducted using 5 mg wet cells/mL E. coli/pveh1L105I/V106I/L196D at 25 °C
for 1.5 h, producing (R)-1 with 95.0% ees, 32.1% yield and 3.52 g/L/h space-time yield.
1. Introduction
epoxidation and kinetic resolution, have been developed to prepare
enantiopure epoxides, in which the expensive chiral ligands and ha-
Chirality is a new emerging challenge for pharmaceutical develop-
ment given that (R)- and (S)-enantiomers of a racemic drug generally
exhibit different or even antagonistic pharmacological activities [1]. In
view of the so-called ‘chiral switch’ from racemates to single en-
antiomers advocated by the US Food and Drug Administration (FDA)
since the early 1990s, an increasing demand for enantiopure inter-
mediates to synthesize chiral compounds has been observed [2]. En-
antiopure epoxides, which are highly value-added and versatile syn-
thons, have been diffusely applied in pharmaceutical, agrochemical and
fine chemical industries [3]. For example, (R)-1 is an essential chiral
building block for a series of benzazepinone derivatives, which can be
clinically used as β-adrenoreceptor antagonists to treat hypertension
and angina pectoris [4].
zardous heavy metals are required [5,6]. Considering the ever-in-
creasing environmental awareness, biocatalysis by using whole resting
cells or enzymes — an environment-friendly process with high stereo-
selectivity and little or no byproducts, is an attractive supplement or
alternative to chemocatalysis [7]. For example, the kinetic resolution of
epoxides by epoxide hydrolases (EHs; EC 3.3.2.-) provided a promising
strategy to produce chiral epoxides because EHs are ubiquitous in
nature, cofactor-independent, and capable of operating in organic/
aqueous biphasic system [8]. Single epoxide enantiomers can be re-
tained from their corresponding racemates with a theoretical yield of
50% via EH-catalyzed kinetic resolution [9]. However, the in-
dustrialized application of these EHs was restricted by the low activity
and enantioselectivity (i.e., E value) [10]. Therefore, excavating novel
EHs with superior catalytic properties or modifying some local
Several chemical-based approaches, such as Jacobsen’s asymmetric
⁎ Corresponding author.
E-mail address: biowmc@126.com (M.-C. Wu).
1 C. Li and T.-T. Kan, the two first authors, contributed equally to this work.
https://doi.org/10.1016/j.mcat.2019.110517
Received 5 March 2019; Received in revised form 15 July 2019; Accepted 15 July 2019
2468-8231/©2019ElsevierB.V.Allrightsreserved.

C. Li, et al.
MolecularCatalysis476(2019)110517
and -pveh1L105I/V106I, and one transformant expressing PvEH1, desig-
nated E. coli/pveh1, were constructed, and preserved in our laboratory.
Rac-1–5 were purchased from TCI (Shanghai, China).
2.2. Expression and purification of PvEH1 or its mutant
A single colony of E. coli transformant, such as E. coli/pveh1 or
/pveh1L105I/V106I/L196D, was aerobically inoculated into LB medium
containing 100 μg/mL kanamycin sulfate, and cultured at 37 °C for
12–14 h as the seed culture. Then, the same fresh medium was in-
oculated with 2% (v/v) seed culture, and cultured until the OD600 value
reached 0.6–0.8. The expression of PvEH1 or its mutant was induced by
0.2 mM IPTG at 20 °C for 10 h. The induced E. coli transformant cells
were harvested by centrifugation, and resuspended in 50 mM
NaH2PO4–Na2HPO4 buffer (pH 7.0). E. coli transformant harboring pET-
28a(+), designated E. coli/pET-28a, was used as a negative control. The
recombinantly expressed PvEH1 or its mutant having a 6×His tag at its
N-terminus was purified by affinity chromatography using a nickel-ni-
trilotriacetic acid (Ni-NTA) column (Tiandz, Beijing, China). Enzyme
purification and protein assay were performed as previously described
[15].
Fig. 1. Substrate spectrum assay of PvEH1 for five rac-aryl glycidyl ethers (rac-
1–5). 1, o-methylphenyl glycidyl ether; 2, m-methylphenyl glycidyl ether; 3, p-
methylphenyl glycidyl ether; 4, phenyl glycidyl ether; 5, benzyl glycidyl ether.
structures of existing EHs by protein engineering is necessary [11].
The laboratory-directed evolution of an Aspergillus niger EH (AnEH)
was conducted to improve its E value for rac-4. After one round of error-
prone PCR, an AnEH mutant (IS002B1), whose E value increased to
10.8 from 4.6, was selected [12]. In another example, an Agrobacterium
radiobacter EH (ArEH) was subjected to iterative saturation mutagen-
esis. Through screening, the best mutant, that is, T247 K/I108 L/D131S,
was obtained. Its catalytic efficiency (kcat/Km) and E value for rac-epi-
chlorohydrin were 4.5- and 2.1-fold higher than those of wild-type
ArEH, respectively [10]. The directed structural modification of EHs on
the basis of rational design also provides additional insight into the
molecular mechanism [13].
2.3. EH activity assay
The activities of PvEH1 and its mutants for rac-1 were determined as
previously described [9], with slight modification. In detail, a total of
475 μL EH solution or cell suspension, which was suitably diluted with
50 mM phosphate buffer (pH 7.0), was preincubated at 25 °C for
10 min. The reaction was initiated by adding 25 μL 200 mM (at a final
concentration of 10 mM) methanol-dissolved rac-1, incubated for
10 min, and terminated by adding 2 mL methanol. The sample was
analyzed by HPLC, using an e2695 apparatus (Waters, Milford, MA)
equipped with an XBridge C18 column. The mobile phase, H2O/me-
thanol (3:7, v/v), was used at a flow rate of 0.8 mL/min, and monitored
using a 2489 UV–Vis detector at 220 nm. One unit (U) of EH activity
was defined as the amount of purified EH or whole resting cells hy-
drolyzing 1 μmol rac-1 per minute under the given assay conditions.
Correspondingly, the activities of PvEH1 for rac-2–5 were measured by
replacing rac-1 with each of them.
Previously, pveh1 (GenBank: KR604729) was cloned from P. vulgaris
total mRNA, and expressed in E. coli BL21(DE3). The enantioconvergent
hydrolysis of styrene epoxides by PvEH1 or its best mutant was studied
[14]. In the present work, PvEH1 had the highest specific activity and E
value for rac-1 among tested five common rac-aryl glycidyl ethers,
which were abbreviated to rac-1–5 in this work (Fig. 1 and Table 1). To
improve the catalytic properties of PvEH1 for rac-1 further, a total of
eight variant genes were constructed by two-stage whole-plasmid PCR
on the basis of rational design. The specific activities and E values of the
purified PvEH1 and its mutants were measured, and compared. The
2.4. EH enantioselectivity assay
kinetic parameters of the best mutant, that is, PvEH1L105I/V106I/L196D
,
were also measured. The source of its enhanced E value was analyzed
by molecular docking (MD) simulation. The kinetic resolution of rac-1
at elevated concentration was also performed using whole resting cells
EH enantioselectivity, which is quantitatively described by E value,
was used to evaluate the preferential hydrolytic degree of one en-
antiomer over its antipode [9]. The E value of PvEH1 or its mutant was
measured as follows: 1.8 mL suitably diluted purified EH solution and
200 μL 200 mM (at a final concentration of 20 mM) rac-1, 2, 3, 4 or 5
were mixed, and then incubated at 25 °C. Aliquots of 100 μL reaction
sample were drawn out at different time points, extracted with 400 μL
ethyl acetate, and analyzed by HPLC equipped with a Chiralcel OD-H
column (Daicel, Osaka, Japan) under the same analytical conditions as
stated above, except for isopropanol/n-hexane (2:8, v/v) as a mobile
phase. The absolute configurations of single enantiomeric 1–5 were
judged by comparing their retention times with those reported pre-
viously [16]. The conversion ratio (c) of substrate was calculated based
on its depleted concentration. The ees of retained epoxide enantiomer
was obtained using the following equation: ees = [(R – S)/(R + S)] ×
100%, where R and S were the concentrations of (R)- and (S)-aryl
glycidyl ether. On the basis of these parameters stated above, the E
value was calculated from an equation: E = ln [(1 – c) × (1 – ees)]/ln
[(1 – c) × (1 + ees)] [17,18].
of E. coli/pveh1L105I/V106I/L196D
.
2. Experimental
2.1. Strains, plasmids, and chemicals
E. coli BL21(DE3) (Novagen, Madison, WI, USA) was used for EH
expression, while PrimeSTAR HS DNA polymerase and Dpn I en-
donuclease (TaKaRa, Dalian, China) were for the site-directed muta-
genesis of PvEH1. Two recombinant plasmids, namely, pET-28a-pveh1
Table 1
Substrate spectrum assay of the purified PvEH1 for rac-1–5.
Substrate
Enantiopreference
Specific activity
E value
(U/mg protein)
Rac-1
Rac-2
Rac-3
Rac-4
Rac-5
(S)-1
(S)-2
(S)-3
(S)-4
(S)-5
11.7
2.4
2.0
4.3
1.8
0.4
0.1
0.1
0.2
0.1
6.1
3.0
2.7
3.4
2.3
2.5. Homology modeling of PvEH1 or its best mutant
The three-dimensional (3-D) structure of PvEH1 or its best mutant
was homologically modeled using the MODELLER 9.11 program
2

C. Li, et al.
MolecularCatalysis476(2019)110517
(http://salilab.org/modeller/) by selecting the known crystal structure
of a Vigna radiata EH (VrEH1, PDB: 5XMD) at 2.0 Å resolution as a
template, which shared 87.4% primary structure identity with PvEH1.
Then, the 3-D structure was subjected to molecular mechanics optimi-
zation by using the CHARMM27 force field in GROMACS 4.5 package
(http://www.gromacs.org/). The outputted result with the best geo-
metry quality was validated using the SAVES program (http://services.
mbi.ucla.edu/SAVES/). Meanwhile, the 3-D structures of (S)- and (R)-1
were handled using a ChemBio3D Ultra 12.0 software (http://www.
cambridgesoft.com/).
hydrolytic reactions of rac-1 at concentrations ranging from 50 to
120 mM were carried out in a 1 mL 50 mM phosphate buffer (pH 7.0)
system, using 5 mg wet cells/mL E. coli/pveh1L105I/V106I/L196D at 25 °C
for 4 h to ensure the maximum allowable concentration of rac-1.
Subsequently, the scale-up kinetic resolution, in a 100 mL phosphate
buffer system containing 5 mg wet cells/mL and rac-1 at the maximum
allowable concentration, was performed at 25 °C. During the hydrolytic
process, aliquots of 100 μL reaction sample were periodically drawn
out, extracted with 400 μL ethyl acetate, and then analyzed by chiral
HPLC.
2.6. Rational design of PvEH1 for its site-directed mutagenesis
3. Results and discussion
The mutual action between the 3-D structures of PvEH1 and (S)-1
was predicted by MD simulation by using the AutoDock vina program
(http://autodock.scripps.edu/) to locate the most appropriate binding
site and steric orientation, that is, a binding state with the lowest en-
ergy. The 3-D conformation of a docked complex PvEH1-(S)-1 was
optimized using the GROMACS 4.5 package, and then visualized using a
PyMol software (http://pymol.org/) to identify the amino acid residues
in proximity to (S)-1 within 8 Å. Other plant EHs sharing more than
55% primary structure identity with PvEH1 were searched by BLAST
server in the NCBI website (http://blast.ncbi.nlm.nih.gov/). Among
them, four plant EHs with superior activities and/or E values for rac-1
were selected to conduct the multiple sequence alignment with PvEH1
by using the ClustalW2 program (http://www.ebi.ac.uk/Tools/msa/
clustalw2/). According to the above computer-aided analysis, the spe-
cific residues in PvEH1 were selected to be separately substituted with
the corresponding and frequently emerging residues (the identity is
equal to 75% or 100%) among other four EHs.
3.1. Substrate spectrum assay of PvEH1
The kinetic resolution reactions of several common rac-aryl glycidyl
ethers by various EHs, such as PvEH2, VrEH3 and TpEH1, have been
investigated [9,15,17]. In our previous studies, only the en-
antioconvergent hydrolysis of rac-styrene epoxides, such as styrene
oxide, by PvEH1 or its mutant was investigated [14]. Herein, to develop
the industrialized application of PvEH1 in the kinetic resolution of rac-
aryl glycidyl ethers, its specific activities and E values for rac-1–5 were
tested, that is, the substrate spectrum assay. As shown in Table 1, the
purified PvEH1 enantiopreferentially hydrolyzed the (S)-enantiomers of
all tested rac-1–5, and possessed the highest specific activity of 11.7 U/
mg protein and E value of 6.1 for rac-1, suggesting that the different
aryl groups in aryl glycidyl ethers significantly influenced the activity
and E value of PvEH1. However, the catalytic properties of PvEH1,
especially the E value, were still unsatisfactory for efficiently preparing
(R)-1 via the kinetic resolution of rac-1. Therefore, substituting several
specific residues in PvEH1 by single or multiple site-directed muta-
genesis on the basis of computer-aided design is necessary for im-
proving its activity and E value.
2.7. Construction of E. coli transformants harboring variant genes
The single or triple site-directed mutagenesis of pveh1 was con-
ducted by two-stage whole-plasmid PCR as reported previously [19]. In
detail, using pET-28a-pveh1 as the template, the first-round PCR was
performed with a forward primer, such as L105I-F and a reverse one (T7
terminator primer) (Table S1) under the following conditions: a dena-
turation at 95 °C for 4 min, followed by 30 cycles of at 98 °C for 10 s,
56 °C for 10 s, and 72 °C for 50 s. Then, the second-round PCR was
carried out using the first-round PCR’s product as a megaprimer: 30
cycles of at 98 °C for 10 s, 53 °C for 10 s and 72 °C for 6 min. The PCR-
3.2. Specific residues selection in PvEH1
To improve the enantioselectivity of PvEH1 for rac-1, that is, the
enantiopreference for (S)-1 efficiently, a total of 51 amino acid residues
in proximity to (S)-1 within 8 Å were first confirmed using a visualized
PyMol software on the basis of 3-D conformation of a docked complex
PvEH1-(S)-1. Then, according to the result of the multiple sequence
alignment of PvEH1 with other four selected plant-derived EHs, 32
absolutely conserved residue sites in PvEH1 were eliminated from 51
ones (Fig. 2A). After further consideration of the remaining 19 non-
conserved residue sites, 12 ones can be excluded: the highest residue
identities among other four plant EHs were only 50% at six sites (such
as Leu129 and Thr178), meanwhile, at the six other sites, the highest
identity residues were the same as those of PvEH1, such as Val138 and
Ala140 (Fig. 2B). Consequently, the remaining seven specific residues,
Ile76, Leu105, Val106, Met175, Leu196, His267, and Thr272, were selected to
be separately substituted with the corresponding and frequently
emerging residues, Cys, Ile, Ile, Ile, Asp, Tyr, and Met, among the four
other EHs (the identity is equal to 75% or 100%). The 3-D structural
analysis of PvEH1 displayed that Met175 and Leu196 are located in the
cap domain of PvEH1, whereas Ile76, Leu105, Val106, His267, and Thr272
are in the α/β domain (Fig. 2C). The mutations of specific residues in
these two domains and near the substrate-binding pocket of EHs con-
siderably influenced their catalytic properties [20–22].
amplified target recombinant plasmids, such as pET-28a-pveh1L105I
,
were digested by Dpn I to decompose the methylated template, and
transformed into E. coli BL21(DE3), respectively, thereby constructing
seven E. coli transformants harboring single-site variant genes, such as
E. coli/pveh1L105I. Analogously, one recombinant plasmid, pET-28a-
pveh1L105I/V106I/L196D
,
was amplified from pET-28a-pveh1L105I/V106I
using a pair of primers, L196D-F and T7 terminator primer, and used for
the construction of the corresponding E. coli transformant, that is, E.
coli/pveh1L105I/V106I/L196D
.
2.8. Kinetic parameter assay of purified PvEH1L105I/V106I/L196D
The initial hydrolytic rate (μmol/min/mg protein) of rac-1 by
PvEH1L105I/V106I/L196D was measured under the EH activity assay con-
ditions, except for the rac-1 concentrations from 2.0 to 20.0 mM. Km
and Vmax values were calculated by non-linear regression analysis using
an Origin 9.0 software (http://www.orignlab.com/). The turnover
number (kcat) of EH was deduced from its apparent molecular weight
and Vmax, and its catalytic efficiency was defined as the ratio of kcat to
Km.
3.3. Screening of E. coli transformants harboring single-site variants
The recombinant plasmids harboring single-site variants of pveh1
were amplified by whole-plasmid PCR, followed by transforming them
into E. coli BL21(DE3), respectively, thereby constructing seven corre-
sponding E. coli transformants expressing PvEH1 mutants, namely, E.
2.9. Kinetic resolution of rac-1 at elevated concentration
Using the ees and yield of (R)-1 as the criteria, the asymmetric
coli/pveh1I76C
,
/pveh1L105I
,
/pveh1V106I
,
/pveh1M175I
,
/pveh1L196D
,
3

C. Li, et al.
MolecularCatalysis476(2019)110517
Fig. 2. Rational design of PvEH1 for its site-directed mutagenesis. (A) Multiple sequence alignment of PvEH1 with other four selected plant-derived EHs. PvEH1
(GenBank: AKJ75509, in this work); PvEH2 (ASS33914, 80.1% identity with PvEH1); PvEH3 (ATG22745, 71.3%); VrEH3 (AKJ75505, 71.7%); NbEH1 (ACE82566,
66.4%). Among the 51 residues in proximity to (S)-1 within 8 Å, 32 absolutely conserved residue sites were marked with asterisks, while 19 non-conserved sites were
with inverted triangles. (B) The selection of seven specific residues for the site-directed mutagenesis of PvEH1. (C) The topological diagram of PvEH1. The selected
residue sites were marked with solid circles, while the identified sites of a catalytic triad (Asp101-Asp264-His299) and two proton donors (Tyr150 and Tyr234) with stars.
E. coli/pET-28a (as a negative control) under the same expression
conditions (data not shown). Subsequently, the seven single-site mu-
tants of PvEH1 were purified to homogeneity, and their specific activ-
ities and E values for rac-1 were measured (Table 2). As a result,
PvEH1L105I and PvEH1V106I had the highest specific activities of 17.6
and 16.4 U/mg, which were about 1.5- and 1.4-fold higher than that
(11.7 U/mg) of wild-type PvEH1. Meanwhile, the E value of PvEH1L196D
increased to 9.2 from 6.1 of PvEH1.
Table 2
Specific activities and E values of the purified PvEH1 and its mutants for rac-1.
Enzyme
Specific activity (U/mg protein)
E value
PvEH1
11.7
6.2
0.4
0.3
6.1
2.3
7.7
7.5
5.9
9.2
4.8
5.0
10.9
PvEH1I76C
PvEH1L105I
PvEH1V106I
PvEH1M175I
PvEH1L196D
PvEH1H267Y
PvEH1T272M
PvEH1L105I/V106I/L196D
17.6
16.4
10.3
10.8
9.0
0.6
0.5
0.4
0.5
0.5
0.2
0.7
3.4. Obtaining one triple-site mutant of PvEH1
6.5
23.1
A synergistic effect on the improvement in catalytic properties via
combinatorial site-directed mutagenesis may be present [23]. There-
fore, on the basis of above experimental results of the single site-di-
/pveh1H267Y, and /pveh1T272M. Sodium dodecyl sulfate-polyacrylamide
gel electrophoresis (SDS-PAGE) analysis displayed that the seven single-
site mutants of PvEH1 were successfully expressed in their respective E.
coli transformants, such as PvEH1L105I in E. coli/pveh1L105I, at 20 °C for
10 h by 0.2 mM IPTG, just the same as PvEH1 expressed in E. coli/pveh1
(as a positive control). However, no target protein band was detected in
rected mutagenesis, the three specific residues in PvEH1, Leu105, Val106
,
and Leu196, were subjected to a combinatorial triple site-directed mu-
tagenesis to superimpose the superior catalytic properties of
PvEH1L105I, PvEH1V106I, and PvEH1L196D for the kinetic resolution of
rac-1. A recombinant plasmid connecting a triple-site variant, pET-28a-
4

C. Li, et al.
MolecularCatalysis476(2019)110517
37,893 Da) that were identified by ProtParam (http://web.expasy.org/
protparam/).
As expected, PvEH1L105I/V106I/L196D simultaneously had the highest
specific activity of 23.1 U/mg and E value of 10.9 among all the tested
single- and triple-site mutants (Table 2). Its specific activity was about
2.0-fold that of PvEH1, while the EH activity (736.9 U/g wet cell) of E.
coli/pveh1L105I/V106I/L196D was 4.7 folds higher than that of E. coli/
pveh1. The EH activity assays of purified enzyme and whole resting
cells, as well as SDS-PAGE analysis (Fig. 3, lanes 1 and 2) suggested that
the increased soluble expression level of EH in E. coli/pveh1L105I/V106I/
L196D
contributed more than half to its evidently enhanced EH activity.
It has been also reported that the expression levels of target enzymes in
E. coli were elevated via site-directed mutagenesis [11,24].
3.5. Kinetic parameter of best mutant PvEH1L105I/V106I/L196D
Fig. 3. SDS-PAGE analysis of the recombinantly expressed PvEH1 and
PvEH1L105I/V106I/L196D. Lane M, protein marker; lanes 1 and 2, the supernatants
of E. coli/pveh1 and /pveh1L105I/V106I/L196D cell lysates, respectively; lanes 3 and
4, the purified PvEH1 and PvEH1L105I/V106I/L196D, respectively.
Compared with the Km value (3.45 mM) of a wild-type PvEH1, that
(2.58 mM) of a triple-site mutant PvEH1L105I/V106I/L196D decreased by
25.2%, indicating that the substrate affinity of PvEH1 with rac-1 (or (S)-
1 considering the enhanced E value) was improved via its triple site-
directed mutagenesis. The purified PvEH1 and PvEH1L105I/V106I/L196D
displayed the kcat values of 9.38 and 17.15 s−1, respectively. In con-
sequence, the catalytic efficiency (6.65 mM−1 s−1) of the latter was 2.4
times higher than that (2.72 mM−1 s−1) of the former. These changes in
kinetic parameters implied that the combinatorial substitution of L105I,
V106I, and L196D in PvEH1 had a remarkably positive effect on its
substrate affinity and catalytic efficiency for rac-1 or (S)-1.
pveh1L105I/V106I/L196D, was amplified by PCR as designed theoretically,
and used to construct the corresponding E. coli transformant, that is, E.
coli/pveh1L105I/V106I/L196D. The purified PvEH1 and PvEH1L105I/V106I/
L196D
exhibited single-target protein bands with the same apparent
molecular weight of approximately 38.0 kDa (Fig. 3, lanes 3 and 4), in
accordance with their respective theoretical values (37,877 and
Fig. 4. Analysis of the source of PvEH1L105I/V106I/L196D with enhanced E value for rac-1 by MD simulation. The locally magnified 3-D conformation of a docked
complex PvEH1L105I/V106I/L196D-(S)-1 (A) or -(R)-1 (B) was compared with that of PvEH1-(S)-1 (C) or -(R)-1 (D).
5

C. Li, et al.
MolecularCatalysis476(2019)110517
3.6. Source of PvEH1L105I/V106I/L196D with enhanced E value
The catalytic mechanism of EHs, which are members of an α/β-
hydrolase fold superfamily, was that the ring-opening of an epoxide
molecule begins with the substrate activation via hydrogen bonds from
two specific Tyr residues served as proton donors. Then, a nucleophilic
Asp residue in the catalytic triad of EHs attacks the Cβ (a less hindered
C-atom in an oxirane ring) and Cα (a more hindered C-atom) with
different regioselectivities [8]. In the present work, Asp101 in PvEH1 or
PvEH1L105I/V106I/L196D was identified as a catalytic nucleophile. The Cβ
of (S)- or (R)-1 was mainly attacked by Asp101, affording the corre-
sponding (S)- or (R)-diols via the retention of configuration (Fig. S1).
Therefore, the through-space distance (dβ) between the nucleophilic O-
atom of Asp101 side chain and Cβ was considered as a crucial parameter
[25]. To gain insight into the source of PvEH1L105I/V106I/L196D with an
enhanced E value, the 3-D conformations of docked PvEH1L105I/V106I/
L196D-(S)-1 and -(R)-1 were compared with those of PvEH1-(S)-1 and
-(R)-1, respectively (Fig. 4). The lengths (l1 and l2) from the hydroxyl
groups of Tyr150 and Tyr234 (the proton donors in PvEH1 and
PvEH1L105I/V106I/L196D) to O-atom in an oxirane ring of (S)- or (R)-1
were not more than 3.5 Å, which was the prerequisite for ring-opening
Fig. 5. Process curves of various parameters in the scale-up kinetic resolution of
100 mM rac-1 at 25 °C until 3.5 h using 5 mg wet cells/mL E. coli/pveh1L105I/
V106I/L196D
.
pveh1L105I/V106I/L196D was conducted in a 100 mL phosphate buffer
system, and monitored by chiral HPLC at the given time intervals
(Fig. 5). After incubation at 25 °C for 1.5 h, (S)-1 was almost completely
hydrolyzed at 67.0% c of rac-1, retaining (R)-1 with 95.0% ees and
32.1% yield. The space-time yield (STY), which is defined as the
amount of obtained target product in the unit volume and time, of (R)-1
reached 3.52 g/L/h, higher than those by AnEH- and TlEH-expressing A.
niger and Trichosporon loubierii resting cells [30,31], but lower than that
by E. coli resting cells expressing VrEH3 [15] (Table S2). After the hy-
drolytic reaction of rac-1 proceeded for 3.5 h, the ees of (R)-1 had in-
significant increase, while its yield and STY obviously dropped to
23.3% and 1.02 g/L/h, respectively.
reaction [15]. The dβ value of PvEH1-(S)-1 (or PvEH1L105I/V106I/L196D
-
-
(S)-1) was shorter than that of PvEH1-(R)-1 (or PvEH1L105I/V106I/L196D
(R)-1), suggesting that PvEH1 or its best mutant preferentially hydro-
lyzes (S)-1. Although four docked complexes had approximate binding
energies ranging from –18.95 to –17.47 kJ/mol, the differences in dβ
values were still obvious. Through the triple site-directed mutagenesis
of PvEH1, the dβ to (S)-1 was shortened to 3.1 from 3.4 Å, while the dβ
to (R)-1 increased from 3.7 to 4.0 Å. The analytical results above were
in accordance with those of the experimental measurements, and si-
milar to the conclusions drawn by other research groups [10,26,27].
3.7. Preparation of (R)-1 from rac-1 by E. coli/pveh1L105I/V106I/L196D
4. Conclusions
The rac-1 concentration (10 or 20 mM) that was used for the assay
of EH activity or E value, was too low to realize the scale-up preparation
of (R)-1. Therefore, the kinetic resolutions, in a 1 mL 50 mM phosphate
buffer (pH 7.0) system containing 5 mg wet cells/mL of E. coli/
pveh1L105I/V106I/L196D and rac-1 at 50, 80, 100, and 120 mM were car-
ried out, respectively, at 25 °C for 4 h. As shown in Table 3, rac-1 was
enantioselectively hydrolyzed up to 100 mM, retaining (R)-1 with over
95.0% ees with only 12.0%–21.4% yield, suggesting that the excessive
hydrolysis of rac-1 may occur. Whereas, in the case of 120 mM rac-1,
the ees of (R)-1 was only 74.6%, and still lower than 85% even though
the reaction time was prolonged to 8 h. The result indicated that sub-
strate (rac-1) or its corresponding product (vicinal diol) at elevated
concentration severely inhibited the EH activity of E. coli/pveh1L105I/
V106I/L196D. This phenomenon was also observed in other stereoselective
hydrolytic reactions of rac-epoxides in an aqueous phase system by
various EHs, such as TpEH1 and AnEH [17,28]. Consequently, the
maximum allowable concentration of rac-1 was confirmed to be
100 mM. Using whole resting cells instead of purified EH as the bio-
catalyst, to our knowledge, was because that the former was easily
accessible, and displayed higher stability, especially at elevated sub-
strate concentration, than the latter during the hydrolytic reaction [29].
The scale-up kinetic resolution of 100 mM rac-1 catalyzed by E. coli/
PvEH1 showed the highest specific activity and E value for rac-1
among five tested rac-aryl glycidyl ethers (1–5). To improve the cata-
lytic properties, especially the E value, of PvEH1 for rac-1 efficiently, its
single or triple site-directed mutagenesis was carried out on the basis of
rational design. The residue substitutions of L105I and V106I in PvEH1
positively affected its specific activity, while L196D enhanced its E
value. Among all the purified PvEH1 and its mutants tested, PvEH1L105I/
V106I/L196D
possessed the highest specific activity and E value, sug-
gesting that the combinatorial substitution of L105I, V106I, and L196D
in PvEH1 had a positive synergistic impact on its catalytic properties.
The analytical result of PvEH1L105I/V106I/L196D with enhanced E value by
MD simulation was consistent with that of the experimental determi-
nation. Using whole resting cells of E. coli/pveh1L105I/V106I/L196D as the
biocatalyst, the kinetic resolution of rac-1 at elevated concentration was
carried out, retaining (R)-1 with high ees, yield, and STY. This work
developed the application of PvEH1 in the kinetic resolution of rac-aryl
glycidyl ethers, especially the rac-1, and also provided an efficient
technical strategy to customize the desired EHs for preparing target
chiral epoxides.
Declaration of Competing Interest
All of the authors declare that they have no conflict of interest.
Table 3
Kinetic resolutions of rac-1 at elevated concentrations by E. coli/pveh1L105I/
V106I/L196D
.
Acknowledgments
Rac-1 (mM)
Time (h)
c (%)
ees (%)
Yield (%)
This work was financially supported by the National Natural Science
Foundation of China (No. 21676117), the National First-Class
Discipline Program of Food Science and Technology of China
(JUFSTR20180101), and the Natural Science Foundation of Jiangsu
Province for Youth of China (No. BK20180622).
50
80
100
120
4
4
4
4
88.2
81.1
78.0
57.2
> 95.0
> 95.0
> 95.0
74.6
12.0
18.7
21.4
37.4
6

C. Li, et al.
MolecularCatalysis476(2019)110517
Appendix A. Supplementary data
[15] D. Hu, C.D. Tang, C. Li, T.T. Kan, X.L. Shi, L. Feng, M.C. Wu, Stereoselective hy-
drolysis of epoxides by reVrEH3, a novel Vigna radiata epoxide hydrolase with high
enantioselectivity or high and complementary regioselectivity, J. Agric. Food Chem.
65 (2017) 9861–9870.
[16] N. Bala, S.S. Chimni, H.S. Saini, B.S. Chadha, Bacillus alcalophilus MTCC10234
catalyzed enantioselective kinetic resolution of aryl glycidyl ethers, J. Mol. Catal. B
Enzym. 63 (2010) 128–134.
[17] K. Wu, H.L. Wang, H.H. Sun, D.Z. Wei, Efficient kinetic resolution of phenyl glycidyl
ether by a novel epoxide hydrolase from Tsukamurella paurometabola, Appl.
Microbiol. Biotechnol. 99 (2015) 9511–9521.
[18] C.S. Chen, Y. Fujimoto, G. Girdaukas, C.J. Sih, Quantitative analyses of biochemical
kinetic resolutions of enantiomers, J. Am. Chem. Soc. 104 (1982) 7294–7299.
[19] J. Sanchis, L. Fernández, J.D. Carballeira, J. Drone, Y. Gumulya, H. Höbenreich,
D. Kahakeaw, S. Kille, R. Lohmer, J.J.-P. Peyralans, J. Podtetenieff, S. Prasad,
P. Soni, A. Taglieber, S. Wu, F.E. Zilly, M.T. Reetz, Improved PCR method for the
creation of saturation mutagenesis libraries in directed evolution: application to
difficult-to-amplify templates, Appl. Microbiol. Biotechnol. 81 (2008) 387–397.
[20] M. Kotik, A. Archelas, V. Faměrová, P. Oubrechtová, V. Křen, Laboratory evolution
of an epoxide hydrolase ― towards an enantioconvergent biocatalyst, J. Biotechnol.
156 (2011) 1–10.
[21] L.F. Zhang, J.M. Wu, H. Feng, Homology modelling and site-directed mutagenesis
studies of the epoxide hydrolase from Phanerochaete chrysosporium, J. Biochem.
149 (2011) 673–684.
[22] Z.T. Sun, R. Lonsdale, L. Wu, G.Y. Li, A.T. Li, J.B. Wang, J.H. Zhou, M.T. Reetz,
Structure-guided triple-code saturation mutagenesis: efficient tuning of the stereo-
selectivity of an epoxide hydrolase, ACS Catal. 6 (2016) 1590–1597.
[23] M.T. Reetz, Biocatalysis in organic chemistry and biotechnology: past, present, and
future, J. Am. Chem. Soc. 135 (2013) 12480–12496.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.mcat.2019.110517.
References
[1] P. Saini, D. Sareen, An overview on the enhancement of enantioselectivity and
stability of microbial epoxide hydrolases, Mol. Biotechnol. 59 (2017) 98–116.
[2] J. Zaugg, Y. Gumulya, M. Bodén, A.E. Mark, A.K. Malde, Effect of binding on en-
antioselectivity of epoxide hydrolase, J. Chem. Inf. Model. 58 (2018) 630–640.
[3] G.S. de Oliveira, P.P. Adriani, F.G. Borges, A.R. Lopes, P.T. Campana,
F.S. Chambergo, Epoxide hydrolase of Trichoderma reesei: biochemical properties
and conformational characterization, Int. J. Biol. Macromol. 89 (2016) 569–574.
[4] A. Bisi, A. Rampa, R. Budriesi, S. Gobbi, F. Belluti, P. Ioan, E. Valoti, A. Chiarini,
P. Valenti, Cardiovascular hybrid drugs: new benzazepinone derivatives as brady-
cardic agents endowed with selective β1-non-competitive antagonism, Bioorg. Med.
Chem. 11 (2003) 1353–1361.
[5] M. Kotik, A. Archelas, R. Wohlgemuth, Epoxide hydrolases and their application in
organic synthesis, Curr. Org. Chem. 16 (2012) 451–482.
[6] M.P. Kamble, G.D. Yadav, Kinetic resolution of (R,S) phenyl glycidyl ether by red
mung beans (Vigna angularis) epoxide hydrolases, Biocatal. Agric. Biotechnol. 12
(2017) 260–265.
[7] D. Hu, R. Wang, X.L. Shi, H.H. Ye, Q. Wu, M.C. Wu, J.J. Chu, Kinetic resolution of
racemic styrene oxide at a high concentration by recombinant Aspergillus usamii
epoxide hydrolase in an n-hexanol/buffer biphasic system, J. Biotechnol. 236
(2016) 152–158.
[8] E. Serrano-Hervás, M. Garcia-Borràs, S. Osuna, Exploring the origins of selectivity in
soluble epoxide hydrolase from Bacillus megaterium, Org. Biomol. Chem. 15 (2017)
8827–8835.
[9] C. Li, D. Hu, X.C. Zong, C. Deng, L. Feng, M.C. Wu, Asymmetric hydrolysis of
styrene oxide by PvEH2, a novel Phaseolus vulgaris epoxide hydrolase with ex-
tremely high enantioselectivity and regioselectivity, Catal. Commun. 102 (2017)
57–61.
[10] S.P. Zou, Y.G. Zheng, Q. Wu, Z.C. Wang, Y.P. Xue, Z.Q. Liu, Enhanced catalytic
efficiency and enantioselectivity of epoxide hydrolase from Agrobacterium radio-
bacter AD1 by iterative saturation mutagenesis for (R)-epichlorohydrin synthesis,
Appl. Microbiol. Biotechnol. 102 (2018) 733–742.
[11] C. Li, J. Zhao, D. Hu, B.C. Hu, R. Wang, J. Zang, M.C. Wu, Multiple site-directed
mutagenesis of a Phaseolus vulgaris epoxide hydrolase to improve its catalytic per-
formance towards p-chlorostyrene oxide based on the computer-aided re-design,
Int. J. Biol. Macromol. 121 (2019) 326–332.
[12] M.T. Reetz, C. Torre, A. Eipper, R. Lohmer, M. Hermes, B. Brunner, A. Maichele,
M. Bocola, M. Arand, A. Cronin, Y. Genzel, A. Archelas, R. Furstoss, Enhancing the
enantioselectivity of an epoxide hydrolase by directed evolution, Org. Lett. 6 (2004)
177–180.
[13] M. Widersten, A. Gurell, D. Lindberg, Structure-function relationships of epoxide
hydrolases and their potential use in biocatalysis, BBA-Gen. Subjects 1800 (2010)
316–326.
[14] H.H. Ye, D. Hu, X.L. Shi, M.C. Wu, C. Deng, J.F. Li, Directed modification of a novel
epoxide hydrolase from Phaseolus vulgaris to improve its enantioconvergence to-
wards styrene epoxides, Catal. Commun. 87 (2016) 32–35.
[24] Q. Luo, Y. Chen, J. Xia, K.Q. Wang, Y.J. Cai, X.R. Liao, Z.B. Guan, Functional ex-
pression enhancement of Bacillus pumilus CotA-laccase mutant WLF through site-
directed mutagenesis, Enzyme Microb. Tech. 109 (2018) 11–19.
[25] M.T. Reetz, M. Bocola, L.W. Wang, J. Sanchis, A. Cronin, M. Arand, J.Y. Zou,
A. Archelas, A.L. Bottalla, A. Naworyta, S.L. Mowbray, Directed evolution of an
enantioselective epoxide hydrolase: uncovering the source of enantioselectivity at
each evolutionary stage, J. Am. Chem. Soc. 131 (2009) 7334–7343.
[26] F. Xue, Z.Q. Liu, N.W. Wan, H.Q. Zhu, Y.G. Zheng, Engineering the epoxide hy-
drolase from Agromyces mediolanus for enhanced enantioselectivity and activity in
the kinetic resolution of racemic epichlorohydrin, RSC Adv. 5 (2015) 31525–31532.
[27] Z.Q. Liu, L.P. Zhang, F. Cheng, L.T. Ruan, Z.C. Hu, Y.G. Zheng, Y.C. Shen,
Characterization of a newly synthesized epoxide hydrolase and its application in
racemic resolution of (R,S)-epichlorohydrin, Catal. Commun. 16 (2011) 133–139.
[28] H.X. Jin, Z.C. Hu, Y.G. Zheng, Enantioselective hydrolysis of epichlorohydrin using
whole Aspergillus niger ZJB-09173 cells in organic solvents, J. Biosci. 37 (2012)
695–702.
[29] E.Y. Lee, M.L. Shuler, Molecular engineering of epoxide hydrolase and its appli-
cation to asymmetric and enantioconvergent hydrolysis, Biotechnol. Bioeng. 98
(2007) 318–327.
[30] W.J. Choi, E.C. Huh, H.J. Park, E.Y. Lee, C.Y. Choi, Kinetic resolution for optically
active epoxides by microbial enantioselective hydrolysis, Biotechnol. Tech. 12
(1998) 225–228.
[31] Y. Xu, J.H. Xu, J. Pan, Y.F. Tang, Biocatalytic resolution of glycidyl aryl ethers by
Trichosporon loubierii: cell/substrate ratio influences the optical purity of (R)-ep-
oxides, Biotechnol. Lett. 26 (2004) 1217–1221.
7