Engineering a Carbonyl Reductase as a Potential Tool for the Synthesis of Chiral α-Tetralinols

Article abstract of DOI:10.1002/cctc.202100890

Tailoring of enzyme toward α-tetralones, a class of bulky-bulky ketones, is still a challenge. In this work, the mutants of carbonyl reductase BaSDR1 with improved catalytic performance toward α-tetralone 1 a were obtained by adjusting the steric hindrance and hydrophobicity of the residues that affect the approach of α-tetralone with the catalytic residues. The designed mutants also showed enhanced catalytic performance toward halogenated α-tetralones 2 a–6 a. Remarkably, the activity of the mutant Q237V/I291F toward 7-fluoro-α-tetralone 5 a was 16.3-fold higher than the wildtype enzyme with improved stereoselectivity (98.8 % ee). More notably, the mutants Q139S and Q139S/V187S exhibited decreased or reversed stereoselectivity toward α-tetralone 1 a, 5-bromo-α-tetralone 2 a, 7-fluoro-α-tetralone 5 a and 7-chloro-α-tetralone 6 a, while the relatively high ee values were obtained in the presence of 6-chloro-α-tetralone 3 a and 6-bromo-α-tetralone 4 a as substrates. Further analysis showed the larger size of the substrates was beneficial for the substrates binding to the active cavity with a more specific binding mode, which endows the reaction with higher stereoselectivity. Moreover, the recombinant E. coli expressing the variant Q237V/I291F successfully catalyzed the reduction of a high concentration 7-fluoro-α-tetralone 5 a. These results not only offered a potential tool for chiral α-tetralols, but also provided guiding information for the enzyme engineering toward bulky-bulky ketones.

Full text of DOI:10.1002/cctc.202100890

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Engineering a carbonyl reductase as a potential tool for the
synthesis of chiral α-tetralinols
Authors: Aipeng Li, Ting Wang, Ke Yang, Xuanshuo Zhang, Dongming
Yin, Yong Qin, and Lianbing Zhang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202100890
Link to VoR: https://doi.org/10.1002/cctc.202100890
0
1/2020

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
Engineering a carbonyl reductase as a potential tool for the
synthesis of chiral α-tetralinols
[
a]
[a]
[a]
[a]
[a]
[a, b]
Aipeng Li, Wang Ting, Ke Yang, Xuanshuo Zhang, Dongming Yin, Yong Qin,*
Zhang*[a]
and Lianbing
[
a]
Dr. Aipeng Li, Wang Ting, Ke Yang, Xuanshuo Zhang, Dongming Yin, Prof. Yong Qin, Prof. Lianbing Zhang
School of Life Sciences
Northwestern Polytechnical University
710072 Xi'an, China
E-mail: lbzhang@nwpu.edu.cn
[
b]
Prof. Yong Qin
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry
Chinese Academy of Sciences
030001 Taiyuan, China
E-mail: qinyong@sxicc.ac.cn
Supporting information for this article is given via a link at the end of the document.
Abstract: Tailoring of enzyme toward α-tetralones, a class of bulky-
bulky ketones, is still a challenge. In this work, the mutants of carbonyl
reductase BaSDR1 with improved catalytic performance toward α-
tetralone 1a were obtained by adjusting the steric hindrance and
hydrophobicity of the residues that affect the approach of α-tetralone
with the catalytic residues. The designed mutants also showed
enhanced catalytic performance toward halogenated α-tetralones 2a-
Enzyme-mediated asymmetric reduction is a preferable
technology due to its mild reaction conditions and environmental
friendliness.[6] Several biocatalytic approaches have been
reported for the preparation of chiral α-tetralol.[7] However, the
biocatalytic asymmetric reduction is still far from feasible for the
preparation of chiral α-tetralol due to their insufficient catalytic
activity, unsatisfactory stereoselectivity or low substrate loading.
Moreover, previous studies mainly focused on the asymmetric
reduction of α-tetralone, specific biocatalysts for the reduction of
halogenated α-tetralones were seldom reported. The lack of
specific biocatalysts cannot satisfy the demand for the
halogenated chiral α-tetralols. Given the importance of chiral α-
tetralol and its halogenated derivatives in the synthesis of
bioactive molecules, it is necessary to develop enzymes with
robust capacity for the production of chiral α-tetralols.
6a. Remarkably, the activity of the mutant Q237V/I291F toward 7-
fluoro-α-tetralone 5a was 16.3-fold higher than the wildtype enzyme
with improved stereoselectivity (98.8% ee). More notably, the mutants
Q139S and Q139S/V187S exhibited decreased or reversed
stereoselectivity toward α-tetralone 1a, 5-bromo-α-tetralone 2a, 7-
fluoro-α-tetralone 5a and 7-chloro-α-tetralone 6a, while the relatively
high ee values were obtained in the presence of 6-chloro-α-tetralone
3a and 6-bromo-α-tetralone 4a as substrates. Further analysis
showed the larger size of the substrates was beneficial for the
substrates binding to the active cavity with a more specific binding
mode, which endows the reaction with higher stereoselectivity.
Moreover, the recombinant E. coli expressing the variant
Engineering of existing protein provides a convenient way to
create desired enzymes toward given substrates.[8] In particular,
rational design has been demonstrated as an effective strategy to
adjust enzyme catalytic performance, including activity
Q237V/I291F successfully catalyzed the reduction of
a
high
improvement,[9]
stereoselectivity
modification,[10]
stability
concentration 7-fluoro-α-tetralone 5a. These results not only offered a
potential tool for chiral α-tetralols, but also provided guiding
information for the enzyme engineering toward bulky-bulky ketones.
enhancement [11] and substrate scope expansion.[12] Nevertheless,
rational engineering of enzymes for the asymmetric reduction of
bulky-bulky ketones, which bear two bulky substituents, is still a
major challenge.[13] As precursors for the asymmetric synthesis of
chiral α-tetralols, α-tetralones belong to the class of bulky-bulky
ketones. The elusive effect of the structure of the bulky-bulky
ketones, such as the substrate size and the presence of
substituent group, on the enzyme catalytic performance was the
main reason that hampers the obtaining of ideal enzymes by
rational design. Besides, the carbonyl group of α-tetralones is
located closer to the aromatic ring compare with their counterparts
such as β-tetralones. The large steric hindrance and unique
properties of aromatic ring might make the contact of the catalytic
residues of the enzyme with the carbonyl group of the substrates
more difficult, which impedes enzymatic performance. Moreover,
it was reported that the minor differences in the substrate
structure might exert great influence on the binding of the ligand
Introduction
Chiral α-tetralol and its halogenated derivatives are important
chiral scaffolds for the synthesis of pharmaceuticals. For instance,
enantiomerically pure α-tetralol is a versatile intermediate to
access sulfonamide compounds having TRPM8 antagonistic
activity,[1] inhibitors for treating central nervous system
disorders,[2] ORL-1 receptor agonist [3] and antidepressant;[4] (S)-
7-Fluoro-α-tetralol is used to synthesize purinone derivatives,
which are useful for indications such as leukemias and
lymphomas, organ and bone marrow transplant rejection, mast
cell-mediated allergic reactions and inflammatory diseases.[5]
Thus, considerable efforts have been devoted to the synthesis of
optically active α-tetralols.
and the protein receptor, leading to the obviously different enzyme
activity and stereoselectivity.[14] Therefore, it is more challenging
to tailor specific enzyme toward the asymmetric reduction of α-
1
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
tetralone and its halogenated derivatives. An in-depth
understanding of the effect of the substrate structure on the
enzyme catalytic performance is of great fundamental importance
in the rational design of the specific enzyme.
In our previous study, a carbonyl reductase (BaSDR1) from
Bacillus aryabhattai was discovered to be an excellent biocatalyst
for the reduction of aromatic ketones, especially ortho-
haloacetophenones.[15] However, the wildtype BaSDR1 showed
poor activity toward α-tetralone 1a (Scheme 1 and Table 1).
Herein, to obtain robust enzyme for the synthesis of chiral α-
tetralols, the reconstruction of the catalytic pocket of carbonyl
reductase BaSDR1 was performed by rational adjusting the steric
hindrance and hydrophobicity of the residues that affect the
approach of α-tetralone 1a with catalytic active site. The mutants
with significantly enhanced catalytic activity and high
stereoselectivity toward α-tetralone 1a were obtained. Then, the
stereoselectivity and activity of the designed mutants toward
halogenated α-tetralones (2a-6a), which bear various size
halogen atom at different position of the aromatic ring, was
evaluated to further clarify the effect of the structure of the bulky-
bulky ketones on enzyme catalytic performance. Furthermore, the
mechanism behind the changed catalytic performance was
analysed. And the effect of the structure of the bulky-bulky
ketones on enzyme catalytic performance, especially
stereoselectivity, was emphatically discussed. Finally, the whole
cell catalyzed synthesis of optical pure (S)-7-fluoro-α-tetralol,
which is an important building block for the synthesis of chiral
pharmaceuticals,[5] was presented as an example to demonstrate
the application potential of the designed mutants in the production
of chiral α-tetralols.
Figure 1. The complex model of wildtype BaSDR1 and substrate 1a. (a) and (b)
The typical binding modes of substrate 1a in the wildtype BaSDR1; (c) Substrate
1a with vertical binding pose in the active site; (d) Substrate 1a with parallel
binding pose in the active site.
engineering, the substrate 1a was docked into the catalytic pocket
of BaSDR1 (Figure 1). Substrate 1a showed two typical binding
modes in the catalytic cavity. In one binding mode, the aromatic
ring and the cyclohexane ring of the substrate 1a were vertical to
the bottom of the substrate-binding pocket (Figure 1c), while
these groups were almost parallel to the bottom of the substrate-
binding pocket in the other binding mode (Figure 1d). These
substrate binding modes shared some similarities. The
cyclohexane ring of substrate 1a directed to residues Q139, Q237
and F250, while its aromatic ring was oriented toward residues
V187 and I291. Such orientation ensures that NADH delivers its
hydride to Re-face of the substrate 1a, leading to the formation of
the corresponding (S)-alcohols. The docking results are in good
agreement with the experimental observations (Table 1 and Table
2).
It is noteworthy that the catalytic pocket space seems too
small to give enough steric flexibility for the bulky-bulky ketone 1a.
There are only narrow spaces between substrate 1a and the
residues Q139, V187, Q237, F250 and I291 (Figure 1c and 1d). It
might hamper the effective contact of the catalytic residues of the
enzyme with the carbonyl group of the substrates and further
caused the low catalytic activity of BaSDR1 toward α-tetralones.
Moreover, the limited structural flexibility and the relatively large
size of α-tetralone makes it difficult for the substrates to adjust
their structure spontaneously to gain a more productive binding
pose. To provide a suitable catalytic pocket for α-tetralone binding,
the residues Q139, V187, Q237, F250 and I291, which closely
adjacent to the substrate, were selected as the engineering
targets and re-designed.
Scheme 1. Asymmetric reduction of α-tetralones to the corresponding chiral
alcohols by BaSDR1.
As shown in Figure 1, the residues Q139, Q237 and F250 are
located on the side of the substrate cyclohexane ring and carbonyl
group. Obviously, the bulky side chain of these three residues
blocked the approach of the catalytic residue Y199 with the
carbonyl group of the substrate. Thus, Q139, Q237 and F250
were replaced by residues harboring a relatively small side chain
such as alanine, glycine, valine, leucine, serine, aspartate and
Results and Discussion
Rational design of BaSDR1 for asymmetric reduction of α-
tetralone
To clarify the molecular basis behind the low catalytic activity of
BaSDR1 toward α-tetralone 1a and identify the targets for protein
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
Table 1. Asymmetric reduction of α-tetralone 1a by the wildtype BaSDR1 and
its variants
asparagine. Moreover, previous study indicated that the proximity
of substrate to the catalytic residues could be driven by the
change of the side chain property or the increased steric
hindrance of residue in a special position.[15-16] The residues V187
and I291 positioned on the other side of the catalytic residue Y199.
Thus, the change of the side chain property or increased steric
hindrance of these two residues might be a feasible strategy for
pushing the substrates close to the catalytic residues and further
enhancing the catalytic activity. Therefore, similar size residues
with different property (charge property and hydrophobicity of the
amino acid side chain) or residues with a bulky side chain were
introduced at those sites. Based on the above analysis and design,
a series of single-mutation variants were generated by site-
directed mutagenesis.
Enzyme
Specific activity (mU/mg)
ee (%)
WT
Q139S
21.25±3.20
170.37±10.57
125.20±5.48
113.47±6.65
35.68±3.87
23.63±2.49
25.25±1.68
74.60±5.87
50.21±2.55
63.25±3.86
41.23±5.33
18.25±2.22
14.36±4.02
25.66±1.87
12.31±2.65
15.22±3.68
20.05±4.13
18.25±2.20
15.65±3.69
41.84±4.07
32.67±1.89
26.78±3.88
36.23±4.04
22.29±4.21
38.40±3.29
15.22±1.87
16.98±3.47
19.29±1.08
28.21±4.33
156.09±11.18
94.26±9.56
42.31±1.33
40.02±5.12
32.94±2.25
26.64±3.21
98.7 (S)
29.2 (R)
16.5 (R)
10.3 (S)
90.5 (S)
98.5 (S)
97.6 (S)
98.5 (S)
98.2 (S)
97.9 (S)
98.3 (S)
98.5 (S)
98.7 (S)
98.0 (S)
98.2 (S)
98.6 (S)
98.3 (S)
97.9 (S)
98.3 (S)
96.8 (S)
97.5 (S)
97.9 (S)
98.0 (S)
97.8 (S)
92.7 (S)
93.2 (S)
92.8 (S)
94.2 (S)
92.9 (S)
14.4 (S)
97.8 (S)
18.0 (S)
14.2 (S)
98.3 (S)
97.7 (S)
Q139G
Q139A
Q139L
Q139D
Q139N
Q237V
Q237G
Q237A
Q237L
Q237D
Asymmetric reduction of α-tetralones
Q237N
The designed mutants were used to catalyze the asymmetric
reduction of α-tetralone 1a for evaluation of their catalytic
performance. As shown in Table 1 and Table S1, three variants
based on residue Q139 (Q139S, Q139G and Q139A) exhibited
significantly enhanced specific activities. Particularly, the variant
Q139S displayed the highest improvement among those of all the
variants tested, the specific activity enhanced from 21.25 to
Q237S
F250A
F250I
F250S
F250D
F250N
170.37 mU/mg toward substrate 1a, an 8.02-fold enhancement.
V187S
However, the trade-off between activity and stereoselectivity was
observed when Q139 was substituted. The mutants Q139S and
Q139G exhibited reversed enantiomeric preference towards
substrate 1a, while the stereoselectivity of the mutants Q139A,
Q139L, Q139I, Q139V, Q139T and Q139C was obviously
decreased as compared with the wildtype enzyme (Table 1 and
Table S1). Moreover, the substitutions at residues V187, Q237
and I291 also exhibited enhanced specific activity. In particular,
the mutants V187S, Q237V and I291F possessed the most
significant increase in the catalytic activity compared with other
mutants at corresponding sites, respectively. More importantly,
none of the variants based on residue V187 and Q237 caused an
adverse effect on stereoselectivity, though the mutation at residue
I291 slightly decreased the enzyme stereoselectivity (Table 1).
Furthermore, the combination of these single-mutations was
performed to investigate whether there would be a synergic effect
or not. As shown in Table 1 and Table S1, compared with the
corresponding single-point mutants, further increased activity was
observed when the mutant Q237V/I291F was utilized as catalyst,
reaching the specific activity values up to 94.26 mU/mg. Though
the variants Q139S/V187S and Q139G/V187S exhibited a lower
activity compared with their single mutant counterparts Q139S
and Q139G, respectively, both of them had significant
improvement compared with the wildtype BaSDR1 and the
mutant V187S. The specific activity of the other double-point
mutants was lower than that of the corresponding single point
mutants. Based on above results, though the mutants Q139S and
Q139S/V187S exhibited undesired stereoselectivity, their
catalytic activity was attractive. Therefore, the mutants Q139S,
V187S, Q139S/V187S, Q237V, I291F and Q237V/I291F were
selected to further evaluate their catalytic performance toward
halogenated α-tetralones (2a-6a) and investigate the effect of the
substrate structure on the enzyme catalytic performance.
V187A
V187D
V187T
V187N
I291F
I291N
I291D
I291S
I291L
Q139S/V187S
Q237V/I291F
Q139S/Q237V
Q139S/I291F
V187S/Q237V
V187S/I291F
I291F also improved the catalytic activity toward the majority of
halogenated α-tetralones. Exceptionally, the specific activity of
the mutant V187S toward substrate 6a decreased, and the
substitution of Q237 with valine lead to the loss of enzymatic
activity toward substrate 3a. Moreover, a synergic effect in the
enhancement of the catalytic activity was observed when the
variant V187S/I291F was utilized as the catalyst. The specific
activity of the variant V187S/I291F was further enhanced toward
the majority of the tested substrates compared with the single-
point mutants V187S and I291F (Table 2). And the variant
V187S/I291F exhibited excellent stereoselectivity, giving
corresponding alcohols with high ee values (97.8-99.7%). In
addition, the decreased catalytic activity was found in the
reduction of substrates 5a and 6a by the variant Q139S/V187S,
which was similar to its catalytic performance toward substrate 1a.
Nevertheless, the specific activities of the variant Q139S/V187S
toward those substrates were still much higher than those of the
wildtype BaSDR1. More notably, the variant Q139S/V187S
showed excellent stereoselectivity toward substrates 3a and 4a,
As shown in Table 2, the single-mutation variants Q139S
exhibited obviously enhanced activity toward all halogenated α-
tetralones (2a-6a), the enhancement of the specific activity varied
from 1.9- to 14.6-fold. The designed mutants V187S, Q237V and
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
Table 2. Asymmetric reduction of prochiral α-tetralones by the wildtype BaSDR1 and its variants.
Enzyme
Substrate
Specific activity (mU/mg)
ee (%)
Substrate
Specific activity (mU/mg)
ee (%)
WT
21.25±3.20
170.37±10.57
41.84±4.07
156.09±11.18
74.60±5.87
38.40±3.29
94.26±9.56
20.31±4.46
39.02±3.31
57.04±2.80
99.71±6.28
98.7 (S)
29.2 (R)
96.8 (S)
14.4 (S)
98.5 (S)
92.7 (S)
97.8 (S)
97.4 (S)
85.6 (S)
99.4(S)
98.8 (S)
/
27.83±3.38
125.78±4.25
44.88±6.57
134.14±10.74
69.83±4.53
29.31±2.57
86.71±5.80
19.13±1.17
53.71±5.68
41.89±4.72
120.55±8.41
40.95±4.51
24.32±0.95
83.82±4.45
12.00±3.20
70.34±1.14
10.13±1.50
45.96±3.92
45.32±3.32
18.30±2.79
86.07±4.49
97.5 (S)
17.2 (R)
99.6 (S)
72.7 (S)
99.7 (S)
99.3 (S)
98.5 (S)
98.2 (S)
98.4 (S)
99.2 (S)
99.8 (S)
99.5 (S)
98.1 (S)
99.6 (S)
93.6 (S)
22.9 (S)
90.5 (S)
49.8 (R)
96.5 (S)
95.1 (S)
99.7 (S)
Q139S
V187S
Q139S/V187S
Q237V
1a
2a
I291F
Q237V/I291F
WT
Q139S
V187S
Q139S/V187S
Q237V
3a
5a
4a
6a
[
a]
Nr
I291F
20.05±1.77
Nr
98.5 (S)
/
Q237V/I291F
WT
12.90±3.74
187.72±7.90
57.10±3.05
171.63±9.54
54.75±4.48
33.52±3.54
210.59±7.35
70.0 (S)
22.6 (R)
55.8 (S)
56.9 (R)
95.7 (S)
92.5 (S)
98.8 (S)
Q139S
V187S
Q139S/V187S
Q237V
I291F
Q237V/I291F
[
a] No reaction was observed.
producing corresponding chiral alcohols with 98.8% and 99.8%
ee, respectively. Those results indicated that the presence of
halogen atoms and their position at the aromatic ring of substrates
had significant effects on the catalytic performance of the enzyme.
In order to elucidate the underlying mechanism contributing to
the enhanced catalytic activity toward α-tetralones, the kinetic
parameters of the double-point mutants and the wildtype enzyme
toward α-tetralone 1a and halogenated α-tetralones (2a, 3a and
simultaneous increase in the kcat value for substrates 1a when the
mutant Q139S/V187S was used as a catalyst. This result
revealed that the improved catalytic capability of Q139S/V187S
toward substrate 1a was mainly originated from the high degree
of productive binding but not from the enhanced enzyme-
substrate affinity.[13b,
17]
It is obviously different from the
determined kinetic parameters of the mutant Q139S/V187S
toward halogenated α-tetralones (2a, 4a and 5a), indicating the
halogen substitution at aromatic ring of the substrates result in
distinct effects on enzyme-substrate binding.
5
a) were determined. As shown in Table 3, the variants
Q139S/V187S and V187S/I291F exhibited enhanced catalytic
efficiency (kcat/K ) toward all tested substrates. In particular, the
values of these two mutants toward substrate 5a were
5.5- and 16.8-fold higher than the wildtype enzyme. Moreover,
these two variants exhibited lower K values toward the majority
m
kcat/K
m
Molecular basis of the desired activity and stereoselectivity
To explore the molecular basis of the enhanced catalytic activity
and the changed stereoselectivity from the perspective of the
structure-function relationship, docking of α-tetralones (1a, 2a, 4a
and 5a) into the structural model of the double-mutation variants
was performed (Figure 2 and Figure S2-S4). The substrate
binding modes in the mutants shared some similarities with those
observed in the wildtype enzyme. Similar with one of the typical
binding conformation in the wildtype enzyme, the substrates
mainly adopted a vertical conformation, namely the aromatic ring
and cyclohexane ring of the substrates were vertical to the bottom
of the substrate-binding pocket in the complex models of the
variant Q139S/V187S and α-tetralones (Figure 2e and Figure
S2b-S4b). The dominant substrate-binding modes in the variant
V187S/I291F was similar to another typical binding conformation
in the wildtype enzyme, in which the aromatic ring and
cyclohexane ring of the substrates were almost parallel to the
bottom of the substrate-binding pocket (Figure 2f and Figure S2c-
S4c). The difference is that the carbonyl moiety of the substrates
was positioned closer to the catalytic triad (Ser186-Tyr199-
Lys203) in both double-mutation variants than those in the
wildtype enzyme. As shown in Figure 3, the distances between
the Tyr199 hydroxyl and the oxygen atom of the substrate
carbonyl groups in the variants Q139S/V187S and V187S/I291F
were shorter than those in wildtype. Furthermore, compared with
those in wildtype, the shorter or almost equal distances between
1
m
of the tested substrates. The binding free energy between the
enzymes and α-tetralones are in good agreement with the
m
experimentally determined K . The binding free energies of the
mutant complex models were lower than those of wildtype
enzyme for the majority of the tested substrates (Figure S1),
indicating the higher affinity of the variants to the substrates than
the wildtype enzyme. These results demonstrated that the
improved enzyme-substrate affinity was a main driving force for
the elevated catalytic efficiency toward these substrates.
m
Specifically, increases in the K value was observed with a
Table 3. Apparent kinetic parameters of the wildtype BaSDR1 and the variants
toward α-tetralones
k
cat
K
m
k
cat/K
m
[
a]
Enzyme
Sub
−
1
−1
−1
(
min
)
(mM)
(min mM )
WT
5.02±0.96
50.02±2.44
18.03±1.68
4.56±1.22
16.10±1.53
8.64±1.08
28.98±1.20
46.51±1.26
10.62±0.87
9.39±1.33
48.45±1.47
55.33±1.68
22.11±1.02
48.29±2.24
21.96±0.83
79.62±2.07
30.09±2.31
18.45±1.41
387.15±5.22
217.26±4.33
56.32±2.11
195.64±1.83
77.73±0.98
82.61±1.39
0.23±0.03
1.04±0.05
0.82±0.06
0.08±0.02
0.54±0.03
0.47±0.02
0.07±0.01
0.21±0.07
0.19±0.04
0.04±0.02
0.62±0.06
0.67±0.05
Q139S/V187S
Q237V/I291F
WT
1a
Q139S/V187S
Q237V/I291F
WT
2a
4a
5a
Q139S/V187S
Q237V/I291F
WT
Q139S/V187S
Q237V/I291F
the C
4
atom of the nicotinamide ring of NADH and the carbon
atom of the substrate carbonyl groups were observed in both
[
a] Abbreviation for substrate.
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
Figure 2. Structure comparison and enzyme-substrate complex models. (a) The active site of the wildtype BaSDR1; (b) and (e) Superimposition of the wildtype
BaSDR1 and the variant Q139S/V253S complex models; (c) and (f) Superimposition of the wildtype BaSDR1 and the variant Q237V/I291F complex models; (d)
The complex models of the wildtype BaSDR1 and substrate 1a; (g) and (h) The interactions between the wildtype BaSDR1 and substrate 1a; (i) The interactions
between the variant Q139S/V253S and substrate 1a; (j) The interactions between the variant Q237V/I291F and substrate 1a. The Ser186-Tyr199-Lys203 catalytic
triad is highlighted in light purple. The specific residues are indicated by labels. Green dash lines, hydrogen bonds; Black dash line, hydrophobic interactions.
double-mutation variants. The proposed catalytic mechanism of
the carbonyl reductase suggests that the hydrogen atoms of the
residue at this site could push the substrates close to the catalytic
residues, and thus, the corresponding mutants displayed
improved activity. Interestingly, the enzyme activity of the mutants
also increased when V187 was substituted with a serine, though
this residue also located on the opposite side of Y199 and the
mutation of this site did not have significant steric hindrance
change as other sites (Figure 2). In this case, the proximity of
substrates to the catalytic residues was probably driven by the
change of the residue hydrophobicity at this site. The aromatic
ring of the substrates oriented toward residues V187. The
hydrophobic aromatic ring is preferably close to hydrophobic
region of the substrate-binding pocket. However, the side chain of
serine is hydrophilic. The tendency of the hydrophobic benzene
ring away from the hydrophilic region become a pushing force for
shortening the distance between the key functional groups of the
substrates and enzymes. These results indicated that it was
effective to improve the enzyme catalytic performance by
adjusting the steric hindrance and hydrophobicity of the residues
which affect the approach of the substrate with the catalytic active
site.
4
C atom of NADH nicotinamide ring and the tyrosine hydroxyl
were respectively transferred to the carbon atom and oxygen
atom of the substrate carbonyl group during the reaction.[18]
Shortening of those distances in the mutants facilitated the
hydride transfer from Tyr199 and NADH to the substrates, and
thus, the engineered enzyme-mediated asymmetric reduction
proceeded more rapidly.[19]
The primary causes for the change of the substrate position
could be ascribed to the changes of the steric hindrance and
hydrophobicity of the corresponding residues in the mutants. As
shown in Figure 2, the residues Q139 and Q237 were located on
the Y199 side of the substrate-binding pocket. The introduction of
residues harboring a smaller side chain at those sites carved out
a bigger space for the proximity of the substrates to the catalytic
residue Y199 and consequently contributing to the enhanced
catalytic activity. However, the increased catalytic activity also
was observed while I291 was substituted with a relatively bulky
phenylalanine. Similar results were obtained in other enzyme
engineering studies.[16, 20] This is probably due to the special
location of the residue I291, which located on the opposite side of
the catalytic residue Y199 (Figure 2). The increased size of the
Furthermore,
the
protein-ligand
interaction
profile
demonstrated that the dominant binding conformations of the
substrates were mainly facilitated by the hydrophobic interactions
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
stereoselectivity. As shown in Scheme 1, the special position of
the halogen atom on the aromatic ring of the substrates 3a and
4
a makes their size relatively larger than those of substrates 1a,
a, 5a and 6a in horizontal and/or vertical direction. The relative
2
larger size restricted the binding of the substrates 3a and 4a to
more defined conformations in the substrate-binding pocket of the
mutants. It makes the mutants Q139S and Q139S/V187S
retained high stereoselectivity toward the substrates 3a and 4a.
The similar effect of substrate size on the enzyme
stereoselectivity was also confirmed by the results of substrate
reduction catalyzed by the wildtype enzyme and other mutants.
For example, with the increase of the halogen atom size of the
substrates 3a and 4a, an obvious increase of the stereoselectivity
was observed when the mutant Q139S was utilized as catalyst.
Similar results were obtained in the reduction of substrates 5a and
6a by the wildtype enzyme, Q139S, V187S and I291F (Table 2).
It is generally accepted that the lower steric hindrance is beneficial
for substrate binding and product release, resulting in a higher
catalytic activity.[9, 12b, 22] However, above results indicated that the
relative larger size of the substrate was beneficial for the
substrates binding to the enzyme active cavity with a more
specific binding mode, and further endows the corresponding
asymmetric reduction with higher stereoselectivity. Excessive
expansion of the substrate-binding pocket might lead to the
undesired trade-off between enzyme activity and stereoselectivity.
The relative size of the substrate and the substrate-binding pocket
should be balanced to obtain ideal enzymes. Previous study also
found that the structural flexibility of the substrates was an
important factor determining the enzyme stereoselectivity.[23]
Therefore, the reconstruction of the substrate-binding pocket
should be performed based on a comprehensive analysis of the
structural properties of the substrate to obtain enzyme with
desired activity and stereoselectivity.
Figure 3. The distances between the key atoms in the complex models of
substrates and enzymes. (a) Distance Osub-OY199, the distance between
the tyrosine hydroxyl and the oxygen atom of the substrate carbonyl. (b)
Distance Csub-C4NADH, the distance between the C atom of the
4
nicotinamide ring of NADH and the carbon atom of the substrate carbonyl.
and the hydrogen bonds. The hydrophobic interactions were
mainly formed between the cyclohexane ring of the substrates
and the pocket-forming amino acids such as Q139, Q188, L196,
Y199, P231, Q/V237, L236 and F250. The hydrogen bonds
occurred more frequently between the oxygen atoms of the
substrate carbonyl and the residues such as Q139 and Q239
besides the catalytic residues Y199 and S186 (Figure 2 and
Figure S2-S4). It is worth noting that the introduced mutations
obviously changed the enzyme-substrate interaction network in
the enzyme-substrate complex models. Some of the original
enzyme-substrate interactions were disrupted, while novel
interactions were constructed for stabilization of the productive
conformation, though the main types of enzyme-substrate
interactions did not change significantly. The enzyme-substrate
interactions were important driving forces for shaping substrate
binding conformation and played critical roles in determining the
enzyme catalytic performance.[21] Further research on the precise
introduction/elimination of enzyme-substrate interactions and the
evaluation of their contributions should be conducted.
Potential application of the designed mutant under high
substrate loading
(S)-7-Fluoro-α-tetralol is applicable to the synthesis of purinone
derivatives, a class of chemical compounds with the potential for
medicinal use.[5] Therefore, the mutant V187S/I291F and the
corresponding fluorinated α-tetralone (5a) were selected as an
example to test the potential of the BaSDR1 variants. Previous
study found that the whole cells expressing BaSDR1 could
effectively utilize glucose to recycle the cofactor and accelerate
the reaction without exogenous NADH addition.[15] Thus, the
recombinant E. coli whole cells expressing the wildtype BaSDR1
and the mutant V187S/I291F were used to catalyse the
asymmetric reduction of substrate 5a at different concentrations
for evaluation of their catalytic performances.
As shown in Figure 4, in the reduction of 10 mM substrate 5a,
the mutant V187S/I291F achieved >99% conversion within 60 min.
In contrast, a very low conversion (17.5%) conversion was
obtained after 60 min upon using the whole cells of the wildtype
enzyme, and the conversion reached only 41.5% after 300 min.
The effectiveness of the mutant V187S/I291F was further
evaluated with increased substrate loading. The whole cells of the
mutant V187S/I291F also could efficiently catalyzed the reduction
of substrate 5a at substrate concentrations of 30 mM and 50 mM,
the conversion reached >99% after 90 min and 300 min,
respectively. More importantly, chiral (S)-7-Fluoro-α-tetralol was
obtained with high ee value (>98%) at different substrate
concentrations. These results suggested that the designed
Another point of concern is that the well-known trade-off
between enzyme activity and stereoselectivity was observed. The
mutants Q139S and Q139S/V187S exhibited significantly
decreased or even reversed stereoselectivity toward substrates
1
a, 2a, 5a and 6a. However, the relatively high ee values were
obtained in the presence of α-tetralones 3a and 4a as substrates
Table 2). Further analysis showed that the size of the substrate
might play an important role in determining the enzyme
(
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
variants could be potential tools for the synthesis of chiral α-
tetralols.
sequenced to exclude the undesired mutations. The expression and
purification of the wildtype BaSDR1 and its mutants were conducted as
previously described.[16] The protein concentrations were determined
according to Bradford method employing bovine serum albumin as the
standard protein.
50
45
40
35
30
25
20
15
10
5
0
Enzyme assay and kinetic analysis
Activity of wildtype BaSDR1 and its mutants was assayed with α-tetralones
as the substrate. One unit of activity was defined as the amount of enzyme
catalysing the reduction of 1.0 μmol of substrate per minute under
measurement conditions. All activity assays were performed in 100 mM
sodium phosphate buffer (pH 7.5) at 35 °C and 200 rpm. The standard
assay mixture (1.0 mL) contained 10 mM α-tetralones, 5% (w/w) glucose,
2.0 mM NADH, 1.0 mM Co2+ and an appropriate amount of the purified
enzyme. The reaction solutions were extracted with the same volume of
ethyl acetate. The conversion and stereoselectivity were measured by a
GC-9790II gas chromatography system (Wenling, China), which equipped
with an FID detector and chiral column Hydrodex β-TBDAc chiral column
10 mM WT
10 mM Q237V/I291F
30 mM Q237V/I291F
50 mM Q237V/I291F
(Macherey-Nagel, Germany, 25 m × 0.25 mm, 0.25 μm film thickness).
30
60 90 120 150 180 210 240 270 300
Time (min)
The injector and detector temperatures were set at 250 °C. The apparent
kinetic parameters of wildtype BaSDR1 and the selected mutants were
determined by measuring the enzyme activity at various substrate
Figure 4. Asymmetric reduction of fluorinated α-tetralone 5a by recombinant E.
coli whole cells expressing the wildtype BaSDR1 and the variant Q237V/I291F.
m
concentrations. The Michaelis-Menten constants (K ) were determined
from plots of the initial velocity versus the substrate concentration using
non-linear regression. The kcat values were obtained by dividing the
maximum velocity values with the enzyme concentration. All
measurements were conducted in triplicate if not specified.
Conclusion
In conclusion, the catalytic activity of the carbonyl reductase
BaSDR1 toward α-tetralones was successfully improved by
rational reconstruction of the substrate-binding pocket. The
designed mutants exhibited potential as efficient tools for the
synthesis of chiral α-tetralols. The changes of the steric hindrance
and hydrophobicity of the substrate-binding pocket-forming
residues turned out to be crucial factors connected to the
improved activity. More notably, it was found that the substrate
size played an important role in determining the stereoselectivity
of enzyme-catalyzed asymmetric reduction. This was a reminder
that the structural properties of the substrate should be
considered as important factor in the rational design of enzymes.
Precise and efficient engineering of enzyme needs to fully match
the substrate and the substrate-binding pocket. This study not
only provided a successful case of rational enzyme design for the
asymmetric reduction of bulky-bulky ketones but also offered
valuable clues for understanding the basis of the enzyme activity
and stereoselectivity control.
Molecular docking and dynamics simulation
The three-dimensional homology model of wildtype BaSDR1 was
generated using the oxidoreductase YghA from Salmonella enterica (67%
identity, PDB: 3R3S) as the template in our previous study.[15] The
structure models of mutants were constructed using Swiss-PDB Viewer
and refined by energy minimization. Docking studies were performed using
AutoDock Vina 1.1.2 program [24] and the enzyme-substrate complexes
were visualized by Chimera.[25] The binding conformation that the
substrate carbonyl oxygen forms hydrogen bonds with Y199 hydroxyl and
S186 hydroxyl at the same time was regarded as an effective conformation
and used in the calculation of the average distances. Molecular dynamics
simulation was performed as previously described.[26] The conformation of
each simulation and the binding free energy of each complex were
analysed by the ptraj program and mmpbsa program in AMBER
package,[27] respectively.
Asymmetric reduction of 7-fluoro-α-tetralone by recombinant whole-
cell
The asymmetric reduction of 7-fluoro-α-tetralone by recombinant whole-
cell was studied with substrate concentrations ranging from 10 mM to 50
mM. The reaction mixture contained 0.4 g wet resting cells, 5% (w/w)
glucose, 1.0 mM Co2+, Na
HPO -NaH PO buffer (100 mM, pH 7.5) in a
2 4 2 4
Experimental Section
total volume of 10 ml. The reactions were carried out at 35 °C with shaking
at 220 rpm. The reaction medium was sampled regularly and extracted
Materials
with ethyl acetate. The organic layer was dried over anhydrous Na
and analysed by chiral gas chromatography.
2 4
SO
Commercial grade solvents, α-tetralone, halogenated α-tetralones and
chiral alcohols used in this study were obtained from J&K Scientific Ltd.
(Beijing, China) or Sigma-Aldrich (Shanghai, China). Nicotinamide
adenine dinucleotide (NADH) disodium salt and isopropyl β-D-1-
thiogalactopyranoside (IPTG) were obtained from Sangon, China.
PrimerSTAR Max DNA Polymerase and restriction enzyme DpnI were
procured from TaKaRa Biotechnology, China. Plasmid miniprep kit and
polymerase chain reaction (PCR) purification kit were purchased from
Transgen Biotech (Beijing, China). Primers synthesis and DNA
sequencing were performed by Sangon (Shanghai, China).
Acknowledgements
The authors greatly acknowledge the financial support from the
Natural Science Basic Research Plan in Shaanxi Province of
China (2020JQ-145), the Postdoctoral Research Project of
Shaanxi Province (2018BSHQYXMZZ27), the National Natural
Science Foundation of China (31700700 and 31971315), the
Science, Technology and Innovation Commission of Shenzhen
Municipality (JCYJ20170815154418207).
Mutant construction, expression and purification
The recombinant plasmid pET30a-BaSDR1 constructed in our previous
work was used as template.[15] Variants were generated by PCR-based
site-directed mutagenesis and E. coli BL21(DE3) was used as the host cell
for the overexpression of the mutants. The entire mutant gene was
7
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
[
[
21] a) Y. Wang, B. Zhang, J. Zhu, C. Yang, Y. Guo, C. Liu, F. Liu, H. Huang,
Keywords: Protein engineering • Chirality • Carbonyl reductase
S. Zhao, Y. Liang, R. Jiao, R. Tan, H. Ge, J. Am. Chem. Soc. 2018, 140,
•
α-Tetralols • Asymmetric reduction
10909-10914; b) A. Su, T. Tyrikos-Ergas, A. N. Shirke, Y. Zou, A. L.
Dooley, I. V. Pavlidis, R. A. Gross, ACS Catal. 2018, 8, 7944-7951.
22] F. Chen, G. Zheng, L. Liu, H. Li, Q. Chen, F. Li, C. Li, J. Xu, ACS Catal.
2
018, 8, 2622-2628.
[23] A. Li, T. Wang, Q. Tian, X. Yang, D. Yin, Y. Qin, L. Zhang, Chem-Eur. J.
021, 27, 6283-6294.
[
[
[
[
[
[
1]
2]
T. Yasuyuki, S. Daisuke, S. Toshiaki, K. Taku, N. Yasuki, A. Nobumasa,
WO2012124825A1, 2012.
2
[
[
24] O. Trott, A. J. Olson, J. Comput. Chem. 2010, 31, 455-461.
25] E. F. Pettersen, T. D. Goddard, C. C. Huang, G. S. Couch, D. M.
Greenblatt, E. C. Meng, T. E. Ferrin, J. Comput. Chem. 2004, 25, 1605-
M. L. R. Heffernan, L. W. Hardy, S. P. Brown, L. W. Herman,
US20190194163A1, 2019.
3]
4]
5]
6]
N. Masakazu, M. Masanori, H. Tokushi, O. Shin-Ichiro, H. Hideki, T. Koji,
EP2128154A1, 2009.
1612.
G. Wang, C. Zheng, G. Zhao, Tetrahedron: Asymmetry 2006, 17, 2074-
[26] J. Gu, H. Yu, J. Biomol. Struct. Dyn. 2012, 30, 585-593.
[
27] D. A. Case, T. E. Cheatham, 3rd, T. Darden, H. Gohlke, R. Luo, K. M.
Merz, Jr., A. Onufriev, C. Simmerling, B. Wang, R. J. Woods, J. Comput.
Chem. 2005, 26, 1668-1688.
2081.
C. Kingsbury, M. Ohlmeyer, V. M. Paradkar, H. Park, J. Quintero, Y.
Shao, WO2009048474A1, 2009.
a) J. C. Mayr, J. H. Grosch, L. Hartmann, L. F. M. Rosa, A. C. Spiess, F.
Harnisch, ChemSusChem 2019, 12, 1631-1634; b) C. Aguirre-Pranzoni,
R. D. Tosso, F. R. Bisogno, M. Kurina-Sanz, A. A. Orden, Process
Biochem. 2019, 79, 114-117.
[
7]
a) S. M. Husain, A. Präg, A. Linnenbrink, A. Bechthold, M. Meller,
ChemBioChem 2020, 21, 780-784; b) S. A. Serapian, M. W. van der
Kamp, ACS Catal. 2019, 9, 2381-2394; c) R. Patil, L. Banoth, A. Singh,
Y. Chisti, U. C. Banerjee, Biocatal. Biotransfor. 2013, 31, 123-131; d) T.
Janeczko, A. Panek, A. Świzdor, J. Dmochowska-Gładysz, E.
Kostrzewa-Susłow, Curr. Microbiol. 2012, 65, 189-194; e) V. Pace, Á. C.
Cabrera, V. Ferrario, J. V. Sinisterra, C. Ebert, L. Gardossi, P. Braiuca,
A. R. Alcántara, J. Mol. Catal. B: Enzym. 2011, 70, 23-31; f) A. Świzdor,
T. Janeczko, J. Dmochowska-Gładysz, J. Ind. Microbiol. Biot. 2010, 37,
1121-1130; g) A. Pennacchio, B. Pucci, F. Secundo, F. La Cara, M. Rossi,
C. A. Raia, Appl. Environ. Microb. 2008, 74, 3949-3958.
[
8]
9]
a) K. Chen, F. H. Arnold, Nature Catal. 2020, 3, 203-213; b) A. Z. Khan,
M. Bilal, T. Rasheed, H. M. N. Iqbal, Chinese J. CATAL. 2018, 39, 1861-
1
868.
D. Wang, H. Li, S. Xia, Y. Xue, Y. Zheng, Catal. Sci. Technol. 2019, 9,
961-1969.
10] a) F. Xue, L. Zhang, Q. Xu, Appl. Microbiol. Biotechnol. 2020, 104, 2067-
077; b) K. Wu, Z. Yang, X. Meng, R. Chen, J. Huang, L. Shao, Catal.
[
[
1
2
Sci. Technol. 2020, 10, 1650-1660.
[
[
11] X. Gong, Z. Qin, F. Li, B. Zeng, G. Zheng, J. Xu, ACS Catal. 2019, 9,
147-153.
12] a) Y. Nie, S. Wang, Y. Xu, S. Luo, Y. L. Zhao, R. Xiao, G. T. Montelione,
J. F. Hunt, T. Szyperski, ACS Catal. 2018, 8, 5145-5152; b) M. M. Musa,
O. Bsharat, I. Karume, C. Vieille, M. Takahashi, S. M. Hamdan, Eur. J.
Org. Chem. 2018, 2018, 798-805.
[
13] a) J. Zhou, G. Xu, Y. Ni, ACS Catalysis 2020, 10, 10954-10966; b) B. Su,
Z. Shao, A. Li, M. Naeem, J. Lin, L. Ye, H. Yu, ACS Catal. 2020, 10, 864-
876; c) G. Xu, Y. Wang, M. Tang, J. Zhou, J. Zhao, R. Han, Y. Ni, ACS
Catal. 2018, 8, 8336-8345; d) O. Bsharat, M. M. Musa, C. Vieille, S. A.
Oladepo, M. Takahashi, S. M. Hamdan, ChemCatChem 2017, 9, 1487-
1493.
[
14] E. L. Noey, N. Tibrewal, G. Jiménez-Osés, S. Osuna, J. Park, C. M. Bond,
D. Cascio, J. Liang, X. Zhang, G. W. Huisman, Y. Tang, K. N. Houk, P.
Natl. Acad. Sci. USA 2015, 112, E7065-E7072.
[
[
[
[
15] A. Li, Q. Yuchi, X. Li, W. Pang, B. Li, F. Xue, L. Zhang, Int. J. Biol.
Macromol. 2019, 138, 781-790.
16] A. Li, X. Li, W. Pang, Q. Tian, T. Wang, L. Zhang, Catal. Sci. Technol.
2020, 10, 2462-2472.
17] E. Hamnevik, D. Maurer, T. R. Enugala, T. Chu, R. Lofgren, D.
Dobritzsch, M. Widersten, Biochemistry 2018, 57, 1059-1062.
18] C. Filling, K. D. Berndt, J. Benach, S. Knapp, T. Prozorovski, E. Nordling,
R. Ladenstein, H. Jornvall, U. Oppermann, J. Biol. Chem. 2002, 277,
25677-25684.
[
[
19] X. Luo, Y. Wang, W. Shen, Y. Zheng, J. Biotechnol. 2016, 224, 20-26.
20] a) X. L. Tang, J. Q. Jin, Z. M. Wu, L. Q. Jin, R. C. Zheng, Y. G. Zheng,
Appl. Environ. Microbiol. 2019, 85, e02471-02418; b) N. Chen, Y. Chen,
Y. Tang, Q. Zhao, C. Liu, W. Niu, P. Huang, F. Yu, Z. Yang, G. Ding,
Process Biochem. 2019, 85, 78-83.
8
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202100890
FULL PAPER
Entry for the Table of Contents
The activity of carbonyl reductase BaSDR1 toward α-tetralones, a class of bulky-bulky ketones, was successfully improved by rational
design. The relative larger size of the substrates turned out to be crucial factor connected to the higher enzyme stereoselectivity.
9
This article is protected by copyright. All rights reserved.