A Cyanide-free Biocatalytic Process for Synthesis of Complementary Enantiomers of 4-Chloro-3-hydroxybutanenitrile From Allyl Chloride

Article abstract of DOI:10.1002/cctc.202100835

A biocatalyst used for selective ring scission of (±)-5-(chloromethyl)-4, 5-dihydroisoxazole to synthesize chiral (R)-4-chloro-3-hydroxybutanenitrile (90 % ee, 39 % isolated yield) and (S)-5-(chloromethyl)-4, 5-dihydroisoxazole (99 % ee, 39 % isolated yield) was developed by site-saturated mutagenesis on aldoxime dehydratase derived from Pseudomonas chlororaphis B23 (OxdA). The positive mutant (OxdA-L318I, E=68) improved the enantiomeric ratio E by 6-fold as compared to the wild type enzyme (OxdA-wild, E=11). The racemic precursor of (±)-5-(chloromethyl)-4, 5-dihydroisoxazole, used in the reaction, can be synthesized from readily available allyl chloride without utilizing highly toxic cyanide. The enantiopure (S)-5-(chloromethyl)-4, 5-dihydroisoxazole remaining in the kinetic resolution can be transformed into corresponding chiral (S)-4-chloro-3-hydroxybutanenitrile without loss of enantiomeric excess by treating it with triethylamine in acetonitrile (99 % ee, 72 % isolated yield) or catalysis of OxdA-wild enzyme (99 % ee, 88 % isolated yield).

Full text of DOI:10.1002/cctc.202100835

Full Papers
doi.org/10.1002/cctc.202100835
ChemCatChem
A Cyanide-free Biocatalytic Process for Synthesis of
Complementary Enantiomers of 4-Chloro-3-
hydroxybutanenitrile From Allyl Chloride
Daijun Zheng[a] and Yasuhisa Asano*[a]
A
biocatalyst used for selective ring scission of (�)-5-
(chloromethyl)-4, 5-dihydroisoxazole, used in the reaction, can
be synthesized from readily available allyl chloride without
utilizing highly toxic cyanide. The enantiopure (S)-5-
(chloromethyl)-4, 5-dihydroisoxazole remaining in the kinetic
resolution can be transformed into corresponding chiral (S)-4-
chloro-3-hydroxybutanenitrile without loss of enantiomeric
excess by treating it with triethylamine in acetonitrile (99% ee,
72% isolated yield) or catalysis of OxdA-wild enzyme (99% ee,
88% isolated yield).
(chloromethyl)-4, 5-dihydroisoxazole to synthesize chiral (R)-4-
chloro-3-hydroxybutanenitrile (90% ee, 39% isolated yield) and
(S)-5-(chloromethyl)-4, 5-dihydroisoxazole (99% ee, 39% iso-
lated yield) was developed by site-saturated mutagenesis on
aldoxime dehydratase derived from Pseudomonas chlororaphis
B23 (OxdA). The positive mutant (OxdA-L318I, E=68) improved
the enantiomeric ratio E by 6-fold as compared to the wild type
enzyme (OxdA-wild, E=11). The racemic precursor of (�)-5-
Introduction
Chiral 4-chloro-3-hydroxybutanenitrile is a versatile building
block for the synthesis of L-carnitine[1] and Atorvastatin.[2]
Significant efforts have been made to synthesize this com-
pound mainly focused on the regioselective addition of cyanide
to epichlorohydrin.[3] The enzymatic methods for preparation of
chiral epichlorohydrin[4] include the enantioselective oxidation
of allyl chloride catalysed by chloroperoxidase,[5] enantioselec-
tive ring-closure of dichloropropanol catalysed by halohydrin
dehalogenase,[2a,6] kinetic resolution catalysed by epoxide
hydrolase[7] (Scheme 1A). In addition, a microbial resolution
method for producing chiral 4-chloro-3-hydroxybutanenitrile
has also been reported via enantioselective dehalogenation on
the racemic substrate.[8] Most of these enzymatic methods
inevitably use highly toxic cyanide as nucleophilic reagent to
construct the nitrile group. However, in 1973, a cyanide-free
strategy for accessing β-hydroxy nitriles, by treating the methyl
4, 5-dihydroisoxazole-5-carboxylate with trimethylamine, was
reported.[9] Subsequently, isomerization of 4, 5-dihydroisoxa-
zoles to the corresponding β-hydroxy nitriles was reported
using a variety of alkali conditions,[10] such as alkanethiolate,[10a]
sodium methoxide,[10b] DBU,[10c] alumina.[10d]
[a] D. Zheng, Prof. Y. Asano
Biotechnology Research Center and Department of Biotechnology
Toyama Prefectural University
5180 Kurokawa
Imizu, Toyama 939-0398 (Japan).
E-mail: asano@pu-toyama.ac.jp
Scheme 1. (A). The enzymatic routes to chiral 4-chloro-3-hydroxybutaneni-
trile; (B). The proposed cyanide-free strategy to complementary enantiomers
of 4-chloro-3-hydroxybutanenitrile in this work.
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/cctc.202100835
© 2021 The Authors. ChemCatChem published by Wiley-VCH GmbH. This is
an open access article under the terms of the Creative Commons Attribution
Non-Commercial NoDerivs License, which permits use and distribution in
any medium, provided the original work is properly cited, the use is non-
commercial and no modifications or adaptations are made.
Moreover, the synthesis of the 4, 5-dihydroisoxazoles mainly
involved two methods: one is [3+2] cycloaddition of nitrile
oxides with alkenes,[11] which is a popular method to construct
2-isoxazoline ring except for that it’s lack of chirality control
ChemCatChem 2021, 13, 1–7
1
© 2021 The Authors. ChemCatChem published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100835
ChemCatChem
strategies on chiral 2-isoxazoline compounds synthesis; another
one is conjugate addition of oximes with α, β-unsaturated
aldehydes,[10c,12] which is an ideal synthetic route to synthesize
chiral 2-isoxazoline compounds if the availability and cost of
starting materials are not taken into consideration. The first
method enables the efficient synthesis of racemic β-hydroxy
nitriles via a 2-isoxazoline motif from readily available alkenes in
a cyanide-free reaction; however, the lack of chirality controlling
means limits its application in asymmetric synthesis of chiral β-
hydroxy nitriles. To solve this problem, we recently reported an
aldoxime dehydratase catalyzed kinetic resolution method for
the synthesis of chiral β-hydroxy nitriles from racemic 2-
isoxazolines substrates.[13]
Aldoxime dehydratase (Oxd) is a kind of heme-containing
enzyme that catalyses the dehydration of aldoximes to nitriles[14]
and is widely distributed in microorganism, as reported by
Asano et al.[15] It has been used in the asymmetric synthesis of
chiral nitriles,[16] and Kemp elimination.[17] In a recent study,[13]
we successfully developed a new application of the enzyme
wherein synthesis of chiral β-hydroxy nitriles was achieved by
enantioselective NÀ O bond cleavage in 4, 5-dihydroisoxazoles,
and the remaining chiral substrates can also be transformed
into complementary chiral β-hydroxy nitriles in alkali condition.
In the substrate scope investigation, we found that all three
enzymes exhibited poor enantioselectivity on aliphatic sub-
strates, restricting its application on aliphatic β-hydroxy nitriles
synthesis. Considering the usefulness of chiral 4-chloro-3-
hydroxybutanenitrile as a versatile building block in pharma-
ceuticals synthesis, we developed an efficient biocatalyst for the
asymmetric synthesis of this compound by enzyme engineering
(Scheme 1B).
existing with nitrile degrading enzymes in bacteria.[14a–c] The last
one, OxdRE, is derived from Rhodococcus sp. N-771, which is the
first aldoxime dehydratase that revealed the reaction mecha-
nism by X-ray crystal structure determination.[14d,19] A number of
subsequent studies established that all these three enzymes
exhibited better enantioselectivity on aromatic substrates as
compared to aliphatic substrates in the asymmetric synthesis of
nitriles from achiral aldoxime precursors. However, for aliphatic
chiral nitriles synthesis, the OxdA showed better selectivity than
OxdB and OxdRE.[16b]
The tested results on this new reaction showed that all
three enzymes are active on the selective NÀ O bond cleavage
of the precursor (�)-1 and prefer R configuration rather than S
configuration. This means (R)-2 and (S)-1 are produced in the
kinetic resolution of racemic substrate (�)-1.
Among them, the OxdA can catalyse the reaction with the
highest velocity (kcatR =173.2 sÀ 1, kcatS =12.7 sÀ 1) as compared to
OxdB (kcatR =26.8 sÀ 1, kcatS =5.4 sÀ 1) and OxdRE (kcatR =4.2 sÀ 1,
kcatS =2.9 sÀ 1). However, the enantiomeric ratio of all three
enzymes are not very high, the best enantiomeric ratio was
obtained from the OxdA and was only E=11(Table 1). In kinetic
resolution, the low enantiomeric ratio of the enzyme restricts its
application on the chiral product synthesis, but the enzyme can
be exploited to produce the substrate in high enantiopurity at
the expense of yield.[20] To improve the practicality of this
method in asymmetric organic synthesis, site-saturated muta-
genesis of OxdA on the residues around the substrate was
carried out to screen positive mutants possessing better
enantiomeric ratio (E) on this kinetic resolion.
From the docking simulation of OxdA with (R)-1 and (S)-1
using MOE platform (Figure 1), we found that they were bound
in the catalytic pocket in different orientations, which resulted
in the different distance of oxygen (O1) in the substrate to
His320-Nδ. However, in aldoxime dehydration reaction mecha-
nism, the first step is protonation of hydroxyl group in aldoxime
by proton migration from Nδ of His320.[14g] The distance of Nδ
to the O1 in (R)-1 (3.2 Å) is much closer than that in (S)-1 (4.2 Å)
(Figure 1), which suggests that (R)-1 is the preferred substrate
of OxdA rather than (S)-1. If the substrate can twist more in
clockwise direction, the distance of O1 in (S)-1 to His320-Nδ will
further increase and may result in a decrease in activity.
Results and Discussion
First, the precursor of (�)-1 was synthesized using [3+2]
cycloaddition of silyl nitronate with allyl chloride by minor
modification of the published method (SI, section 5.1).[11b] And
the (�)-2 was synthesized in 88% yield from (�)-1 by treating it
with triethylamine in acetonitrile (SI, section 5.2). However, the
desired product was not obtained when NaOMe or K2CO3 were
used as alkali reagents in the ring-opening reaction for (�)-2
synthesis. Under these conditions, a by-product with higher
polarity than the target product was mainly formed, which may
be caused by the instability of 4-chloro-3-hydroxybutanenitrile
in strong alkali conditions.[3a]
Based on this tentative idea, the residues of Met29, Leu145,
Ala147 and Leu318, which are located around the 5-
chloromethyl group of the substrate, were selected as the
targets for saturated mutagenesis. Compared to the wild type
Next, three aldoxime dehydratases were selected for testing
the enantioselective ring scission of 5-(chloromethyl)-4, 5-
dihydroisoxazole. One is OxdA that derived from Pseudomonas
chlororaphis B23. This strain was first isolated as an excellent
whole cell catalyst to produce acrylamide in industry by Asano
as early as 1982[18] and subsequently identified to be one of the
strains possessing aldoxime dehydration activity in 2000.[15]
Then further studies to clone and characterize the aldoxime
dehydratase were carried out, in which the enzyme was named
as OxdA.[14f,g] Another enzyme is OxdB that derived from Bacillus
sp.OxB-1, which is the first report of aldoxime dehydratase co-
Table 1. The kinetic parameters for Oxds on the kinetic resolution of (�)-1.
Enzymes[a]
kR [b] [SÀ 1
]
KRM [mM]
kS [c] [SÀ 1
]
KSM [mM]
E[d]
cat
cat
OxdA
OxdA-L318I
OxdB
OxdRE
173.2
193.0
26.8
4.2
9.6
12.7
1.9
5.4
7.6
8.8
9.5
5.1
11
68
3
13.1
14.1
3.1
2.9
2
[a] The racemic (�)-1 was used for determination by monitoring the
consumption of R/S substrate and the details were available in
experimental section. [b] R means for R configuration of substrate. [c] S
means for S configuration of substrate. [d] E=(kRcat/KRM)/(kScat/KSM).[20]
ChemCatChem 2021, 13, 1–7
www.chemcatchem.org
2
© 2021 The Authors. ChemCatChem published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100835
ChemCatChem
the addition of co-solvent DMSO, the substrate tends to
aggregate together to form oil droplets in KPB solution, and the
static reaction mode further deteriorates the mass transfer
process. Moreover, the reductive additive is essential to recover
enzyme activity from inactive state.[14e,h] In this work, the
reductant Na2S2O4, which has been identified as the best
reductive additive for OxdA,[13] was used in the experiments. As
shown in Table 2, no any inhibitory effect on the enzyme
activity was observed up to 50 mM Na2S2O4, and slight activity
enhancement was detected with the increase of Na2S2O4
concentration. In the absence of Na2S2O4, almost no activity can
be detected, which implies that the reaction occurred in a
ferrous-Heme catalytic pocket, as previously reported.[13,14h,17]
Furthermore, no substrate inhibition effect was observed up to
250 mM substrate concentration (SI, section 4), which suggests
the feasibility of this method in asymmetric organic synthesis.
Though the enantiomeric ratio of mutant OxdA-L318I (E=
68) improved 6-fold as compared to the wild type, it was still
not as effective as infinite E. To quench the reaction at a proper
time is an effective mean to synthesize chiral substrate or chiral
product with acceptable yield when the enantiomeric excess
has been improved to the desired value. Therefore, the reaction
was monitored for 3 h and the results showed that 98% ee of
(R)-2 could be obtained, but in low yield of 16%. With an
acceptable 47% conversion, (R)-2 and (S)-1 were formed with
90% ee and 80% ee in 10 min, respectively. To produce (S)-1
with 99% ee, the reaction can be stopped at 30 min with 57%
conversion and 74% ee of (R)-2 remaining (Figure 2B). This
result was a significant improvement over the use of OxdA-wild
type as catalyst (Figure 2A), in which the best result for (R)-2
synthesis is 90% ee in 9% yield and for (S)-1 is 99% ee in 34%
yield, respectively.
In addition, it was noticed that the ee of product (R)-2
decreased approximately 16% when the ee of substrate (S)-1
increased from 80% to 99%, which is greatly different from the
initial reaction stage in which the product ee decreased by only
about 10% when the substrate ee increased from 0 to 80%.
According to the time course of the reaction, it can be
calculated that approximately 10 mM (R)-1 and 119 mM (S)-1
remained when the substrate ee increased to 80%. However,
the Km value of OxdA-L318I on R and S configurations of the
substrate (table 1) showed that the reaction velocity of (R)-1
decreased to half of the maximum reaction velocity when (R)-1
concentration decreased to 13.1 mM. This means when the
substate ee reaches about 80%, the difference in reaction
velocity for (R)-1 and (S)-1 will decrease, resulting in a
significant decline on the ee value of (R)-2 in the final stage of
conversion to synthesize enantiopure (S)-1 (ee>99%). Further-
more, it was also observed that the reaction velocity on (S)-1
became faster when the competition from (R)-1 became
weaker, which further deteriorates the enantioselective con-
version of the substrate in the final stage of reaction.
Figure 1. (A). Docking simulation of (R)-1 with OxdA; (B). Docking simulation
of (S)-1 with OxdA.
enzyme, several positive mutants of M29N, L318I and L318T
were obtained from the initial screening by GC analysis. Then
the best mutant of L318I was purified and its kinetic parameter,
enantiomeric ratio were measured (Table 1). As shown in
Table 1, the significant difference between the OxdA-wild
(kcatS =12.7 sÀ 1) and OxdA-L318I (kcatS =1.9 sÀ 1) is the kcat value
for the S configuration of substrate. An approximately 6-fold
decrease in activity was determined, which resulted in about 6-
fold improvement of enantiomeric ratio E (E=68) on the kinetic
resolution of (�)-1. At the same 50% conversion rate, the
substrate ee using OxdA-L318I as catalyst was significantly
improved from 72% to 88% as compared to the use of OxdA-
wild type as catalyst.
The optimization of the reaction conditions showed that the
OxdA-L318I can tolerate a maximum of 30% DMSO in the
reaction mixture with approximately 77% residual activity
compared with the non-DMSO addition (100% activity). How-
ever, the addition of DMSO greatly improved the ee of substrate
and product as shown in Table 2, which may be due to the
better dispersion of substrate in the presence of DMSO. Without
Table 2. DMSO and Na2S2O4 effects on enzyme activity.
Entry[a] DMSO[b] content Na2S2O4 conc
ee of substrate
Conv
[%][d]
[%]
[mM]
[%][c]
1
2
3
4
5
6
7
8
0
5
5
5
5
5
5
0
10
20
30
40
50
46
97
96
86
1
0
1
99
99
98
99
99
70
67
62
54
0
0
0
76
76
77
79
80
10
20
30
40
50
10
10
10
10
10
10
9
10
11
12
To verify the practicability of this method in asymmetric
organic synthesis, gram-scale reactions were performed with
40 mL reaction mixture at different scheduled times. With a
reaction time of 10 min, (R)-2 was isolated in 39% yield with
90% ee and (S)-1 was isolated in 50% yield with 83% ee by
[a] The reactions were performed in 400 μL mixture containing 100 mM of
°
(�)-1, 10 U of OxdA-L318I, 100 mM KPB (pH 7.0), and incubated at 30 C
for 30 min. [b] The content of DMSO is volume ratio (V/V). [c] The ee of the
substrate was measured by GC. [d] The conversion of the substrate was
measured by GC.
ChemCatChem 2021, 13, 1–7
www.chemcatchem.org
3
© 2021 The Authors. ChemCatChem published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100835
ChemCatChem
Table 3. Gram-scale reactions using enzyme or whole cell as catalyst in
different scheduled times.
Entry[a]
Catalyst[b]
Reaction time [min]
(S)-1[c]
Yield
[%]
(R)-2[d]
Yield
[%]
ee
[%]
ee
[%]
1
2
3
4
5
6
enzyme
enzyme
cell
cell
cell
10
30
40
60
120
180
50
39
46
37
37
30
83
99
74
80
96
99
39
47
43
44
50
53
90
80
87
86
80
69
cell
[a] The details are available in biotransformation part of experimental
section. [b] The catalysts are purified enzyme of OxdA-L318I or lyophilized
whole cell of OxdA-L318I. [c] Isolated yield and the ee was detected using
chiral GC. [d] Isolated yield and the ee was detected by chiral GC after
derivatization using Et3N and CH3COCl in dichloromethane (details are
available in supporting information, section 6).
Scheme 2. The conversion of chiral (S)-1 to (S)-2. For enzymatic trans-
formation, the reaction was performed in 4 mL mixture containing 1 mmol
of (S)-1, 250 U of OxdA-wild, 100 mM KPB (pH 7.0), 25% DMSO and stirred at
400 rpm. For chemical transformation, the reaction was performed in 1 mL
acetonitrile containing 1 mmol of (S)-1, 0.2 mmol triethyl amine.
Alternatively, it can be also achieved using OxdA-wild as
catalyst with 100% conversion and without formation of any
by-products at room temperature (Scheme 2, details in SI,
section 7.2), in which the product can be easily isolated by a
simple work-up process.
Figure 2. (A). The time course of the kinetic resolution of (�)-1 catalysed by
OxdA-wild. (B). The time course of the kinetic resolution of (�)-1 catalysed
by OxdA-L318I. The reactions were performed in 4 mL reaction mixture
containing 250 mM substrate (�)-1, 25% DMSO, 25 U of enzymes, 100 mM
KPB (pH 7.0), 20 mM Na2S2O4, room temperature, stirred at 400 rpm. The
substrate ee and product ee were assayed by GC and the conversion was
evaluated by Conv=ees/(ees +eep). [20]
Conclusion
In summary, we have developed a biocatalyst OxdA-L318I (E=
68) by enzyme engineering for the kinetic resolution of racemic
5-(chloromethyl)-4, 5-dihydroisoxazole ((�)-1). The OxdA-L318I
catalysed method furnished (R)-4-chloro-3-hydroxybutanenitrile
((R)-2) in 39% isolated yield with 90% ee and (S)-5-(chlorometh-
yl)-4, 5-dihydroisoxazole ((S)-1) in 39% isolated yield with 99%
ee, respectively. Furthermore, two strategies to transform the
(S)-5-(chloromethyl)-4, 5-dihydroisoxazole ((S)-1) into (S)-4-
chloro-3-hydroxybutanenitrile ((S)-2) without loss of enantio-
meric excess, have aslo been described. The above disclosed
synthetic route uses readily available allyl chloride as starting
material and avoids highly toxic cyanide for nitrile group
construction. It thus provides an environmentally friendly and
green synthetic strategy for accessing complementary enan-
tiomers of 4-chloro-3-hydroxybutanenitrile from alkene.
silica gel column chromatography (Table 3, entry 1). Alterna-
tively, with a reaction time of 30 min, (S)-1 was isolated in 39%
yield with 99% ee and (R)-2 was isolated in 47% yield with 80%
ee by silica gel column chromatography (Table 3, entry 2).
Furthermore, we attempted to carry out the reaction using
lyophilized whole cell as catalyst in gram-scale reactions, which
is a common reaction mode in industry. As summarized in
table 3, (S)-1 can be isolated with more than 96% ee when the
reaction time was prolonged more than 2 h. In less than 1 h of
reaction time, (R)-2 can be obtained with more than 86% ee in
the current reaction conditions.
In addition, (S)-2 with 99% ee was successfully synthesized
°
in 72% yield from enantiopure (S)-1 by treating it at 60 C with
triethylamine in acetonitrile (Scheme 2, details in SI, section 7.1).
ChemCatChem 2021, 13, 1–7
www.chemcatchem.org
4
© 2021 The Authors. ChemCatChem published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100835
ChemCatChem
°
°
°
°
ature program: 100 C-2 min, 10 C/min to 160 C, 20 C/min to
Experimental Section
°
°
200 C, 200 C-10 min. Retention time for (R)-1 is 10.393 min and
10.676 min for (S)-1. 1 U enzyme was defined as the amount of
enzyme to consume 1 μmol substrate in 1 min. The kinetic
parameters for (R) and (S) configuration of the substrate were
measured in the same way using racemic substrate with substrate
concentration varying from 0.5 mM to 50 mM. The initial reaction
velocity at each concentration was measured by monitoring the
reaction at different time.
Site-saturated mutagenesis and screening: The site-saturated
mutagenesis of OxdA was performed using the PrimeSTAR® Muta-
genesis Basal kit (Takara) with forward and reverse primers of 27-
mer oligonucleotide designed as the kit manual indicated (SI,
section 2). The PCR reaction was performed for 30 cycles: (denatura-
°
°
°
tion 98 C/10 s, annealing 55 C/15 s, elongation 72 C/40 s). The
amplified PCR product was purified by Wizard® SV gel and PCR
clean-up system (Promega). The resulting PCR product was trans-
formed into JM109 E.coli competent cell. Then the plasmids were
extracted from the JM109 E.coli and transformed into BL21DE3 E.coli
for expression. The colonies picked up from plate were inoculated
For whole cell activity assay, the reaction was performed in 2 mL
containing 100 mM (�)-1, 25% DMSO, 20 mM Na2S2O4, 100 mM
KPB (pH 7.0) and 10 mg lyophilized whole cell. The reaction mixture
was stirred in 400 rpm at room temperature and monitored by GC
analysis within 6 min.
°
into 96-well plate with 1 mL auto-induction medium, at 37 C,
1000 rpm for 24 h. The cell were harvested by centrifugation (3,
°
000 rpm, 30 min, 4 C) and re-suspended into 200 μL reaction
The product 4-chloro-3-hydroxybutanenitrile (�)-2 can’t be sepa-
rated in the GC analysis condition described above, the retention
time for both (R)-2 and (S)-2 are 14.820 min. To measure the ee of
the product, it was needed to transform 4-chloro-3-hydroxybutane-
nitrile (�)-2 into corresponding 1-chloro-3-cyanopropan-2-yl
acetate (�)-3 by derivatization using triethylamine and acetyl
chloride as the reagents. The retention time for (R)-1-chloro-3-
cyanopropan-2-yl acetate ((R)-3) is 11.551 min and for (S)-1-chloro-
3-cyanopropan-2-yl acetate ((S)-3) is 11.355 min. The absolute
configuration of product was confirmed using commercial chiral 4-
chloro-3-hydroxybutanenitrile compounds (TCI).
mixture containing 100 mM KPB (pH=7.0), 50 mM (�)-1, 4 mM
Na2S2O4 and 10% DMSO. The resulting reaction mixture was
°
incubated at 30 C, 1000 rpm for 5 h, followed by extraction using
250 μL ethyl acetate. The contents of substrate and product in the
reaction mixture were evaluated by thin layer chromatography
(TLC). The candidates were analysed by gas chromatography after
derivatization using triethylamine and acetyl chloride. The positive
mutants were sequenced by Genetic Analyzer 3500 (ThermoFisher)
using T7 promoter primer (5’-TAATACGACTCACTATAGGG-3’) and T7
terminator primer (5’-ATGCTAGTTATTGCTCAGCGG-3’).
Enzyme purification: The recombinants of BL21DE3-pET-28a-N-His-
OxdA, BL21(DE3)-pET-22b-C-His-Oxd-B and BL21(DE3)-pET-28a-N-His-
Oxd RE were cultivated in the same way as previous work.[13] The
enzymes were purified with minor modification. The recombinant
Biotransformation and isolation: In the gram-scale reactions, the
reactions were carried out in 40 mL reaction mixture containing
1.19 g (�)-1, 25% DMSO, 100 mM KPB (pH 7.0), 20 mM Na2S2O4,
250 U enzyme or 65 mg lyophilized whole cell (5 U/mg). The
reactions were performed at room temperature with stirring at
400 rpm. In the scheduled times, the reactions were stopped by
addition of 5 mL 6 N HCl and extracted by 50 mL×3 ethyl acetate
or diethyl ether. The organic solvent layer was dried against
anhydrous MgSO4, followed by concentration in vacuum, the
residue was purified by silica column chromatography.
°
cells were harvested by centrifugation (6, 000×g, 15 min, 4 C) and
the pellet was re-suspended into lysis buffer (20 mM Tris-HCl,
300 mM NaCl, 20 mM imidazole, pH 8.0), which was applied to
sonication (ice-bath, 15 min) for cell disruption. The debris was
°
removed by centrifugation (20,000×g, 30 min, 4 C) and the super-
natant was loaded onto Ni SepharoseTM 6 Fast flow column (GE
Healthcare) equilibrated by lysis buffer (20 bed volumes). The
unbound protein was washed with lysis buffer (30 bed volumes)
and the bound protein was eluted by elution buffer (20 mM Tris-
HCl, 300 mM NaCl, 300 mM imidazole, pH 8.0), the active fractions
Further experimental details and NMR data, HRMS data, optical
rotation data, GC analysis chromatograms are available in the
supporting information.
°
were pooled and dialyzed against 20 mM KPB (pH 7.0, 5 L, 4 C).
Then the resulting enzyme solution was loaded onto the manual
Toyopearl DEAE-650 M column (TOSOH BIOSCIENCE), which was
equilibrated by equilibration buffer (20 mM KPB, pH 7.0, 10 bed
volumes). After loading the sample, the unbounded protein was
washed by equilibration buffer (5 bed volumes), following with
further washing by 6 bed volume of 10% elution buffer (20 mM
KPB, pH 7.0, 500 mM NaCl)/90% equilibration buffer. Then the
bound protein was eluted by 6 bed volume of 20% elution buffer/
80% equilibration buffer, 6 bed volume of 30% elution buffer/70%
equilibration buffer, 6 bed volume of 40% elution buffer/60%
equilibration buffer. The purity of the fractions was checked by
SDS-PAGE containing 12% acrylamide (SI, section 3), and pure
fractions were pooled and dialyzed against KPB (20 mM, pH 7.0,
5 L×2).
Acknowledgements
We are grateful for the support from Grant-in-Aid for Scientific
Research (S) from the Japan Society for Promotion of Sciences
(Grant No. 17H06169) awarded to Y. Asano.
Conflict of Interest
The authors declare no conflict of interest.
Activity assay and reaction analysis by GC: The purified OxdA and
its mutant were measured with appropriate amount of enzyme
containing 10 mM (�)-1, 10% DMSO, 5 mM Na2S2O4, 100 mM KPB
Keywords: aldoxime dehydratase · asymmetric synthesis ·
biocatalysis · chiral resolution · halohydrin
°
(pH 7.0) in 500 μL, which was incubated at 30 C for 5 min and
followed by addition of 100 μL 6 N HCl to quench the reaction. The
resulting solution was extracted using 900 μL ethyl acetate
containing 20 mM octanenitrile as internal standard substance. The
mixture was centrifuged at 12,000×g for 1 min, 500 μL supernatant
was used for gas chromatograph analysis (Shimadazu gas chroma-
tograph equipped with BGB-174 column (BGB Analytic), carrier gas:
[1] Y. Yang, W. Wang, A. Wumaier, R. Sheng, X. Zhang, T. Zhang, J. Chem.
Res. 2011, 35, 371–372.
[2] a) N.-W. Wan, Z.-Q. Liu, F. Xue, Z.-Y. Shen, Y.-G. Zheng, ChemCatChem
2015, 7, 2446–2450; b) S. K. Ma, J. Gruber, C. Davis, L. Newman, D. Gray,
A. Wang, J. Grate, G. W. Huisman, R. A. Sheldon, Green Chem. 2010, 12,
°
He, flow rate: 50 mL/min, detector temperature: 230 C, temper-
ChemCatChem 2021, 13, 1–7
www.chemcatchem.org
5
© 2021 The Authors. ChemCatChem published by Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100835
ChemCatChem
81–86; c) M. Muller, Angew. Chem. Int. Ed. 2005, 44, 362–365; Angew.
Chem. 2005, 117, 366–369.
158, 185–190; d) Y. Kato, S. Yoshida, S. X. Xie, Y. Asano, J. Biosci. Bioeng.
2004, 97, 250–259; e) E. Pinakoulaki, C. Koutsoupakis, H. Sawai, A.
Pavlou, Y. Kato, Y. Asano, S. Aono, J. Phys. Chem. B 2011, 115, 13012–
13018; f) K. Oinuma, Y. Hashimoto, K. Konishi, M. Goda, T. Noguchi, H.
Higashibata, M. Kobayashi, J. Biol. Chem. 2003, 278, 29600–29608; g) J.
Nomura, H. Hashimoto, T. Ohta, Y. Hashimoto, K. Wada, Y. Naruta, K.
Oinuma, M. Kobayashi, Proc. Natl. Acad. Sci. USA 2013, 110, 2810–2815;
h) K. Kobayashi, S. Yoshioka, Y. Kato, Y. Asano, S. Aono, J. Biol. Chem.
2005, 280, 5486–5490.
[3] a) C. C. J. Culvenor, W. Davies, F. G. Haley, J. Chem. Soc. 1950, 3123–
3125; b) D. Mitchell, T. M. Koenig, Tetrahedron Lett. 1992, 33, 3281–3284;
c) N. Iranpoor, M. Shekarriz, Synth. Commun. 1999, 29, 2249–2254.
[4] H.-X. Jin, X.-K. OuYang, RSC Adv. 2015, 5, 92988–92994.
[5] J. Wu, C. Liu, Y. Jiang, M. Hu, S. Li, Q. Zhai, Catal. Commun. 2010, 11,
727–731.
[6] a) T. Nakamura, T. Nagasawa, F. Yu, I. Watanabe, H. Yamada, Tetrahedron
1994, 50, 11821–11826; b) T. Nakamura, T. Nagasawa, Y. Fujio, I.
Watanabe, H. Yamada, Biochem. Biophys. Res. Commun. 1991, 180, 124–
130.
[7] a) E. Y. Lee, J. Ind. Eng. Chem. 2007, 13, 159–162; b) H.-X. Jin, Z.-Q. Liu,
Z.-C. Hu, Y.-G. Zheng, Eng. Life Sci. 2013, 13, 385–392; c) J.-H. Woo, Y.-O.
Hwang, J.-H. Kang, H. S. Lee, S.-J. Kim, S. G. Kang, J. Biosci. Bioeng. 2010,
110, 295–297.
[8] N. Kasai, T. Suzuki, Y. Furukawa, Chirality 1998, 10, 682–692.
[9] R. Huisgen, M. Christl, Chem. Ber. 1973, 106, 3291–3311.
[10] a) M. H. Seo, Y. Y. Lee, Y. M. Goo, Synth. Commun. 1994, 24, 1433–1439;
b) A. Yashiro, Y. Nishida, K. Kobayashi, M. Ohno, Synlett 2000, 2000, 361–
362; c) A. Pohjakallio, P. M. Pihko, J. Liu, J. Org. Chem. 2010, 75, 6712–
6715; d) S. S. Mochalov, A. N. Fedotov, E. V. Trofimova, R. A. Gazzaeva,
N. S. Zefirov, Russ. J. Org. Chem. 2016, 52, 397–403.
[15] Y. Kato, R. Ooi, Y. Asano, Appl. Environ. Microbiol. 2000, 66, 2290–2296.
[16] a) R. Metzner, S. Okazaki, Y. Asano, H. Gröger, ChemCatChem 2014, 6,
3105–3109; b) T. Betke, J. Higuchi, P. Rommelmann, K. Oike, T. Nomura,
Y. Kato, Y. Asano, H. Gröger, ChemBioChem 2018, 19, 768–779; c) T.
Betke, P. Rommelmann, K. Oike, Y. Asano, H. Gröger, Angew. Chem. Int.
Ed. 2017, 56, 12361–12366; Angew. Chem. 2017, 129, 12533–12538; d) C.
Plass, A. Hinzmann, M. Terhorst, W. Brauer, K. Oike, H. Yavuzer, Y. Asano,
A. J. Vorholt, T. Betke, H. Gröger, ACS Catal. 2019, 9, 5198–5203; e) A.
Hinzmann, S. Glinski, M. Worm, H. Gröger, J. Org. Chem. 2019, 84, 4867–
4872; f) K. Chen, Z. Wang, K. Ding, Y. Chen, Y. Asano, Green Synth. Catal
2021, 2, 179–186.
[17] Y. Miao, R. Metzner, Y. Asano, ChemBioChem 2017, 18, 451–454.
[18] Y. Asano, T. Yasuda, Y. Tani, H. Yamada, Agric. Biol. Chem. 1982, 46,
1183–1189.
[11] a) A. N. Boa, D. A. Dawkins, A. R. Hergueta, P. R. Jenkins, J. Chem. Soc.
[Perkin 1]. 1994, 953–960; b) F. J. F. Castro, M. M. Vila, P. R. Jenkins, M. L.
Sharma, G. Tustin, J. Fawcett, D. R. Russell, Synlett 1999, 6, 798–800.
[12] a) A. Pohjakallio, P. M. Pihko, Synlett 2008, 2008, 827–830; b) A.
Pohjakallio, P. M. Pihko, Chem. Eur. J. 2009, 15, 3960–3964; c) A.
Pohjakallio, P. M. Pihko, U. M. Laitinen, Chem. Eur. J. 2010, 16, 11325–
11339.
[19] H. Sawai, H. Sugimoto, Y. Kato, Y. Asano, Y. Shiro, S. Aono, J. Biol. Chem.
2009, 284, 32089–32096.
[20] A. J. J. Straathof, J. A. Jongejan, Enzyme Microb. Technol. 1997, 21, 559–
571.
[13] D. Zheng, Y. Asano, Green Chem. 2020, 22, 4930–4936.
[14] a) Y. Kato, K. Nakamura, H. Sakiyama, S. G. Mayhew, Y. Asano,
Biochemistry 2000, 39, 800–809; b) Y. Kato, Y. Asano, Protein Expression
Purif. 2003, 28, 131–139; c) Y. Asano, Y. Kato, FEMS Microbiol. Lett. 1998,
Manuscript received: June 9, 2021
Revised manuscript received: July 23, 2021
Version of record online: ■■■, ■■■■
ChemCatChem 2021, 13, 1–7
www.chemcatchem.org
6
© 2021 The Authors. ChemCatChem published by Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPERS
D. Zheng, Prof. Y. Asano*
1 – 7
A Cyanide-free Biocatalytic
Process for Synthesis of Comple-
mentary Enantiomers of 4-Chloro-
3-hydroxybutanenitrile From Allyl
Chloride
An efficient biocatalyst of OxdA-L318I
has been developed for enantioselec-
tive NÀ O bond cleavage of (�)-5-
(chloromethyl)-4, 5-dihydroisoxazole
to synthesize (R)-4-chloro-3-hydroxy-
butanenitrile by enzyme engineering.
The remaining chiral (S)-5-
ing (S)-4-chloro-3-hydroxybutaneni-
trile in alkali condition or under the
catalysis of OxdA-wild type enzyme.
This enzymatic method established a
cyanide-free synthetic route to
produce complementary enantiomers
of 4-chloro-3-hydroxybutanenitrile
from readily available allyl chloride.
(chloromethyl)-4, 5-dihydroisoxazole
can be transformed into correspond-