One-Pot Preparation of d-Amino Acids Through Biocatalytic Deracemization Using Alanine Dehydrogenase and Ω-Transaminase

Article abstract of DOI:10.1007/s10562-018-2565-3

d-Amino acids are pharmaceutically important building blocks, leading to a great deal of research efforts to develop cost-effective synthetic methods. Preparation of d-amino acids by deracemization has been conceptually attractive owing to facile synthesis of racemic amino acids by Strecker synthesis. Here, we demonstrated biocatalytic deracemization of aliphatic amino acids into d-enantiomers by running cascade reactions; (1) stereoinversion of l-amino acid to a d-form by amino acid dehydrogenase and ω-transaminase and (2) regeneration of NAD⁺ by NADH oxidase. Under the cascade reaction conditions containing 100?mM isopropylamine and 1?mM NAD⁺, complete deracemization of 100?mM dl-alanine was achieved after 24?h with 95% reaction yield of d-alanine (> 99% ee^D, 52% isolation yield). Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-018-2565-3

Catalysis Letters
https://doi.org/10.1007/s10562-018-2565-3
One-Pot Preparation of d-Amino Acids Through Biocatalytic
Deracemization Using Alanine Dehydrogenase and ω-Transaminase
Sang‑Woo Han1 · Jong‑Shik Shin1
Received: 9 June 2018 / Accepted: 21 September 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
d-Amino acids are pharmaceutically important building blocks, leading to a great deal of research efforts to develop cost-
effective synthetic methods. Preparation of d-amino acids by deracemization has been conceptually attractive owing to facile
synthesis of racemic amino acids by Strecker synthesis. Here, we demonstrated biocatalytic deracemization of aliphatic
amino acids into d-enantiomers by running cascade reactions; (1) stereoinversion of l-amino acid to a d-form by amino acid
dehydrogenase and ω-transaminase and (2) regeneration of NAD+ by NADH oxidase. Under the cascade reaction conditions
containing 100 mM isopropylamine and 1 mM NAD+, complete deracemization of 100 mM dl-alanine was achieved after
24 h with 95% reaction yield of d-alanine (>99% eeD, 52% isolation yield).
Graphical Abstract
Keywords Cascade reaction · Deracemization · d-Amino acids · Alanine dehydrogenase · ω-Transaminase
Electronic supplementary material The online version of this
1 Introduction
article (https://doi.org/10.1007/s10562-018-2565-3) contains
supplementary material, which is available to authorized users.
Twenty canonical amino acids possess chirality on their
* Jong-Shik Shin
α-carbon, except glycine, and consequently exist in either
enzymo@yonsei.ac.kr
of the two stereoisomers differing in optical rotation, i.e.
1
levorotatory (l) and dextrorotatory (d). Although l-amino
Department of Biotechnology, Yonsei University,
Seoul 03722, South Korea
acids constitute building blocks of proteins and are major
Vol.:(0123456789)
1 3

S.-W. Han, J.-S. Shin
players in nature, d-amino acids also participate in a num-
ber of biological functions and therefore have been consid-
ered as one of the important target compounds in chemical
industry [1–3]. Indeed, d-amino acids have gained growing
attention in food, cosmetics and pharmaceutical industries
[4, 5]. For example, d-alanine is an important building
block of pharmaceutical drugs [6, 7] and an artificial
sweetener [8].
2.2 Molecular Cloning
Genes encoding AlaDH and NOX were cloned into a
pET28a(+) expression vector (Novagen) using an In-
Fusion® HD Cloning Kit (Clontech) according to an
instruction manual. The AlaDH gene was amplified by
colony PCR of Bacillus subtilis str168 (KCTC2217,
Korean Collection for Type Cultures, South Korea) using
forward and reverse primers, 5′-AGGAGATATACCATG
ATCATAGGGGTTCCTAAAG-3′ and 5′-GGTGGTGGT
GCTCGAGAGCACCCGCCACAGATGA-3′, respectively.
The NOX gene was amplified by colony PCR of Lacto-
bacillus brevis (KCTC3498) using forward and reverse
primers, 5′-TAAGAAGGAGATATACCATGAAAGTCA
CAGTTGTTGGT-3′ and 5′-GTGGTGGTGGTGGTGCTC
GAGAGCGTTAACTGATTGTTC-3′, respectively. The
PCR products were ligated with a linearized pET28a(+)
vector digested by NcoI and XhoI. Cloning was confirmed
by gel electrophoresis and DNA sequencing.
In addition to chemocatalytic methods [9], several bio-
catalytic approaches to prepare d-amino acids have been
developed [4, 10, 11]. d-Amino acids can be enzymatically
synthesized from keto acids by reductive amination using
d-amino acid dehydrogenase [12], d-amino acid transami-
nase [10, 13] and d-selective (or R-selective) ω-transaminase
(ω-TA) [14–16]. However, high prices of the keto acid pre-
cursors have been a major drawback of the asymmetric syn-
thesis approach. In contrast, kinetic resolution employing
racemic amino acid substrates is often economically more
feasible because the racemic amino acids can be readily
prepared by Strecker synthesis [17]. Therefore, a number
of kinetic resolution processes have been developed using
acylase [18, 19], amidase [20], oxidase [21] and hydantoin-
ase/carbamoylase [22]. A possible drawback of the kinetic
resolution processes is that a theoretical maximum yield is
only 50% unless an unwanted amino acid enantiomer is recy-
cled by racemization. In this regard, deracemization is an
attractive alternative since the maximum yield could reach
100% as well described in a review paper [23].
2.3 Expression and Purification of Enzymes
For expression of His-tagged enzymes, Escherichia coli
BL21(DE3) cells transformed with the expression vector
harboring AlaDH, ω-TA or NOX were cultivated in a LB
medium (typically 1 L) containing 50 µg mL−1 kanamycin
at 37 °C. In the case of ω-TA, a previously constructed
pET28a vector carrying a codon-optimized gene of an
engineered d-selective ω-TA from Arthrobacter sp. (ARTA
mut) was used [15]. When OD600 reached 0.4, IPTG was
added to induce protein expression and then cells were
allowed to grow for 15 h. The culture broth was centri-
fuged for 20 min at 10,000×g and 4 °C. The resulting cell
pellet was disrupted by a sonicator after being resuspended
in a 15 mL suspension buffer [50 mM sodium chloride,
1 mM EDTA, 1 mM β-mercaptoethanol, 0.1 mM PMSF in
50 mM Tris duffer (pH 8.0)], followed by centrifugation
for 1 h at 10,000×g and 4 °C to remove cell debris.
Recently, we reported deracemization of amino acids into
either l or d-form using two enantiocomplementary transam-
inases [24]. In this method, mandatory use of an expensive
keto acid cosubstrate (i.e., α-ketoglutarate) remained a prob-
lem. Here, we developed a new deracemization method for
production of d-amino acids using l-alanine dehydrogenase
(AlaDH), d-selective ω-TA and NADH oxidase (NOX).
Unwanted l-amino acid in the racemic mixture is converted
to a d-form after two consecutive reactions catalyzed by
AlaDH and ω-TA, leading to complete deracemization or
enantioenrichment depending on the substrate specificity of
the two enzymes.
Purification of the His-tagged enzymes was carried
out on an FPLC system (ÄKTAprime plus, GE Health-
care). In brief, the cell-free extract was loaded to a His-
Trap HP column (GE Healthcare) pre-equilibrated with a
buffer (20 mM imidazole, 500 mM sodium chloride and
20 mM sodium phosphate, pH 7.4). Elution of the bound
protein was done by increasing imidazole concentration
with a linear gradient from 20 to 500 mM. Removal of
imidazole from the protein solution was done on a HiTrap
desalting column (GE Healthcare) using a desalting buffer
(150 mM sodium chloride and 50 mM sodium phosphate,
pH 7). Protein concentrations of the purified enzyme solu-
tions were determined by measuring UV absorbance at
2 Materials and Methods
2.1 Chemicals
l-Alanine was obtained from Acros Organics (Geel, Bel-
gium) and dl-alanine was purchased from Hayashi Pure
Chemical Ind. (Osaka, Japan). Isopropylamine was pur-
chased from Junsei Chemical Co. (Tokyo, Japan). All other
chemicals were purchased from Sigma-Aldrich Co. (St.
Louis, Missouri, USA) and of the highest purity available.
1 3

One-Pot Preparation of d-Amino Acids Through Biocatalytic Deracemization
280 nm using a molar extinction coefficient (ε) obtained
by a protein extinction coefficient calculator (http://www.
biomol.net/en/tools/proteinextinction.htm). The ε val-
ues of AlaDH (homohexamer), ARTAmut (homodimer)
and NOX (homotetramer) were 131,718, 102,135 and
146,572 M−1 cm−1, respectively.
NOX. Reaction was carried under magnetic stirring at 37 °C.
Aliquots (typically 10 µL) of the reaction mixture were taken
at predetermined reaction time and mixed with 40 µL of 5 N
HCl after 20-fold dilution with water. Quantitative chiral
analysis of dl-1a was performed by HPLC for determina-
tion of eeD. When eeD of d-1a exceeded 99%, reaction was
stopped and then product isolation was performed.
2.4 Gel Electrophoresis
The pH of the reaction mixture was adjusted to 1.0 by
adding 5 N HCl and then filtered using a glass-fritted filter
funnel to remove protein precipitates. The filtrate was loaded
on a glass column packed with Dowex 50WX8 cation-
exchange resin (40 g), followed by washing with 0.1 N HCl
(100 mL) and then water (50 mL). Elution was performed
by 10% (v/v) aqueous ammonia solution (150 mL) and then
the eluate was evaporated at 70 °C and 0.25 bar. The result-
ing white solid was washed with ethanol (20 mL) and dried
overnight at 37 °C.
SDS-PAGE was carried out using standard protocols. Pro-
teins were separated on 12% polyacrylamide gel in the pres-
ence of 0.1% SDS. Proteins were visualized by Coomassie
blue staining.
2.5 Enzyme Assay
Unless otherwise specified, all enzyme assays were per-
formed at 37 °C and pH 7 (50 mM Tris buffer). One unit
of ARTAmut activity was defined as the enzyme amount
required for formation of 1 µmol acetophenone in 1 min at
10 mM (R)-α-methylbenzylamine [(R)-α-MBA] and 10 mM
sodium pyruvate. Acetophenone produced was monitored
by a UV-spectrophotometer (UV-1650PC, Shimadzu Co.)
at 285 nm as described elsewhere [25]. One unit of AlaDH
was defined as the enzyme amount that produces 1 µmol
pyruvate in 1 min at 10 mM l-alanine and 2 mM NAD+.
One unit of NOX was defined as the amount of the enzyme
that oxidized 1 µmol NADH in 1 min at 0.2 mM NADH.
Reaction progresses of the AlaDH and NOX assays were
monitored by the UV-spectrophotometer at 340 nm using
εNADH = 6220 M−1 cm−1. In all the enzyme assays, initial
rate measurements were carried out by linear curve fitting of
the reaction progress data showing a time-dependent linear
change in the UV absorbance.
2.8 HPLC Analysis
All the HPLC analyses were carried out using a Waters
HPLC system or an Agilent 1260 Infinity HPLC. Analysis
of pyruvate was performed with an Aminex HPX-87H col-
umn (Bio-Rad) under an isocratic elution of 5 mM H2SO4
solution at 0.4 mL min−1. Column oven temperature was set
to 40 °C and UV detection was done at 210 nm. Quantita-
tive chiral analysis of amino acids were carried out using a
Symmetry C18 column (Waters) after derivatization with
Marfey’s reagent [26]. Details of the HPLC analyses are
described in the Supplementary Materials.
3 Results and Discussion
3.1 Reaction Design of Deracemization
2.6 Cascade Enzyme Reactions
We envisioned that deracemization of racemic amino acid
into d-enantiomer could be implemented by stereoinver-
sion of l-amino acid while keeping d-amino acid intact
(Scheme 1). The stereoinversion of unwanted l-enantiomer
was designed to proceed through two-step enzyme reac-
tions; (1) oxidative deamination of l-amino acid (l-1a–e) by
l-alanine dehydrogenase (AlaDH) at the expense of NAD+
and (2) reductive amination of the resulting keto acid (2a–e)
back to the amino acid in an enantiomerically opposite form
by d-selective ω-transaminase (ω-TA) using isopropylamine
(3) as an amino donor cosubstrate. Note that 3 is one of the
preferred amino donors for ω-TA reactions owing to easy
removal of 4 [27]. In our previous study for deracemization
using enantiocomplementary transaminases, the oxidative
deamination step was carried out by α-transaminase such
as branched-chain transaminase and d-amino acid transami-
nase [24]. It is notable that substitution of α-transaminase
Enzyme reactions were carried out at 37 °C in water bath
without stirring. Reaction volume was 0.2 mL and the reac-
tions were carried out in 50 mM Tris buffer (pH 7.0). 10-µL
aliquots of the reaction mixture were taken at predetermined
reaction times and diluted fivefold with water, followed by
addition of 10 µL of 5 N HCl to stop the reaction. The result-
ing mixture was centrifuged at 13,000 rpm for 15 min and
then the supernatant was used for quantitative chiral HPLC
analysis of amino acid.
2.7 Preparative‑Scale Deracemization
Deracemization of dl-1a in a preparative-scale was carried
out in a 25 mL reaction mixture (50 mM Tris buffer, pH 7.0)
charged with 0.22 g dl-1a (2.5 mmol), 0.15 g 3 (2.5 mmol),
17 mg NAD+, 1750 U AlaDH, 40 U ARTAmut and 250 U
1 3

S.-W. Han, J.-S. Shin
Fig. 1 SDS-PAGE analysis of AlaDH and NOX. Lanes: 1 marker, 2
cell-free extract control prepared with E. coli BL21(DE3) cells not
carrying an expression vector, 3 cell-free extract prepared with cells
expressing AlaDH, 4 purified AlaDH, 5 cell-free extract prepared
with cells expressing NOX, 6 purified NOX
Scheme 1 Deracemization cascade to produce d-amino acids
with AlaDH eliminates the need for an expensive keto acid
cosubstrate. The oxidative deamination of amino acid can be
also carried out using amino acid oxidase as demonstrated
elsewhere [28]. However, generation of chemically reactive
H2O2 as a coproduct necessitated an auxiliary enzyme, such
as catalase, to neutralize the toxic coproduct.
Protein mobility of AlaDH in SDS–PAGE analysis was
in agreement with the molecular weight deduced from the
amino acid sequence (i.e., 39.7 kDa) while that of NOX
was a little lower than the deduced molecular weight (i.e.,
50 kDa).
The stereoinversion reaction requires two substrates (i.e.,
l-1a–e and 3) and a NAD+ cofactor. Because the AlaDH
reaction consumes NAD+, produced NADH should be
converted back to NAD+. We used NADH oxidase (NOX)
from L. brevis for the NAD+ recycling because NOX
does not require an externally added cosubstrate for oxi-
dation of NADH. Note that the NOX from L. brevis does
not form H2O2 but produce H2O [29]. Exploitation of the
H2O-forming NOX eliminates the need for catalase. As a
result, the overall deracemization consists of the stereoinver-
sion reaction performed by AlaDH and d-selective ω-TA and
the cofactor regeneration reaction by NOX.
To examine functional expression of AlaDH and NOX,
the purified enzymes were subjected to activity assays by
real-time monitoring of changes in NADH concentration at
340 nm (Fig. S1). Initial rates were determined by linear
curve fitting of Abs340 data showing a linear increase with
respect to reaction time, i.e. up to 2 min for both enzymes.
Based on the initial rate measurements, specific activities
of AlaDH and NOX were 4 and 20.5 U mg−1, respectively.
3.3 Substrate Range for Deracemization
Feasibility of the stereoinversion reaction is dependent on
enzyme activities of AlaDH and ω-TA for l-amino acids and
its keto acids, respectively. Substrate specificity of ARTAmut
for various keto acids were examined in the previous study
[15]. Owing to a steric constraint in a small substrate binding
pocket which is a general property for all the ω-TAs known
to date, activities of ARTAmut showed an inverse correlation
with the size of side chains of keto acid substrates (Table 1).
The steric constraint for substrate was found to be more
striking for AlaDH, which led to less than 1% enzyme activ-
ity, relative to l-1a, for l-amino acids carrying a side chain
larger than a methyl substituent (i.e., l-1b–e). Eventually,
activity for l-1e carrying a n-butyl substituent was below a
detection limit. These results indicate that stereoinversion
3.2 Cloning and Functional Expression of AlaDH
and NOX
To carry out the deracemization reaction, three enzymes
(i.e., AlaDH, ω-TA and NOX) were required to be puri-
fied. In the case of d-selective ω-TA, we used an engineered
d-selective (or R-selective) ω-TA from Arthrobacter sp.
(ARTAmut) which was cloned and expressed in a previ-
ous study [15]. For preparation of the other two enzymes,
we cloned AlaDH from B. subtilis (ald, Gene ID: 936557)
and NOX from L. brevis (nox, GenBank accession number:
AF536177) in a pET28a(+) expression vector. His-tagged
AlaDH and NOX were purified to homogeneity (Fig. 1).
1 3

One-Pot Preparation of d-Amino Acids Through Biocatalytic Deracemization
Table 1 Substrate specificity of AlaDH for amino acids and ARTAmut
for their cognate keto acids
Substrates (AlaDH/ARTAmut
)
Relative activity (%)a
AlaDHb
c
ARTAmut
1a/2a
1b/2b
1c/2c
1d/2d
1e/2e
100
0.9
0.7
0.05
n.d.
100
33
6
2
2
n.d. not detectable
aRelative activities of AlaDH and ARTAmut represent initial rates nor-
malized by those for 1a and 2a, respectively. Specific initial rate of
AlaDH for 1a was 0.51 0.02 Mm−1 min−1 µM−1-enzyme
Fig. 2 Proof-of-concept of the deracemization strategy. Reaction con-
ditions: 10 mM dl-1a, 10 mM 3, 1 mM NAD+, 0.14 U mL−1 AlaDH,
1 U mL−1 ARTAmut and 5 U mL−1 NOX in 50 mM Tris buffer (pH
7.0) at 37 °C
bReaction conditions: 10 mM dl-1a–e and 2 mM NAD+ in 50 mM
Tris buffer (pH 7.0) at 37 °C
cRelative activities were taken from a previous report where initial
6 h with >99% eeD and 89% reaction yield of d-1a. 2a was
kept no higher than 0.7 mM throughout the reaction, indicat-
ing that reductive amination of 2a by ARTAmut occurred as
efficiently as the oxidative deamination of l-1a by AlaDH.
To expand the applicability of the proposed method, we
tried deracemization of 50 mM dl-1b–d under the same
reaction conditions as those for Fig. 3 (Table 2; entries 1,
3 and 5). No activity of AlaDH for l-1e led us to exclude
dl-1e from the substrate list. Due to the low reactivities of
l-1b–d for AlaDH as well as 2b–d for ARTAmut, complete
deracemization was not achieved for any of the three amino
acids. After 28-h reaction, eeD of the resulting d-1b and d-1c
reached 82 and 80%, respectively, and that of d-1d was only
19%. To achieve complete deracemization of these amino
acids, increase in the enzyme levels or use of other enzymes
showing matching substrate specificity would be required.
Therefore, we tried the cascade reactions under modified
reaction conditions where concentration of AlaDH was ten-
fold increased because of the too low activities for l-1b–d
rates were measured at 20 mM 2a–e, 20 mM (R)-α-MBA [15]
employing AlaDH and ARTAmut should be optimal for l-1a
and much slower reaction progresses would be observed for
l-1b–e.
3.4 In Situ NAD+ Regeneration
To examine whether the proposed deracemization worked
as we intended, we first carried out a cascade reaction with-
out NOX. Starting with 10 mM dl-1a, 10 mM 3 and 1 mM
NAD+, supplementation of the reaction mixture with AlaDH
and ARTAmut enabled time-dependent decreases in the con-
centration of l-1a with increases in the concentration of d-1a
by stoichiometric amounts of the l-1a depletion (Fig. S2).
However, the absence of NAD+ regeneration resulted in only
10% conversion of l-1a to d-1a after 12-h reaction. Enantio-
meric excess was only 9.7% eeD at 12 h, corresponding to
partial enantioenrichment rather than decent deracemization.
Therefore, the reaction mixture was additionally charged
with NOX, so NAD+ could be regenerated in situ and con-
sequently conversion exceeded beyond the stoichiometric
amount of the NAD+ dose (Fig. 2). Indeed, we achieved
complete deracemization at 12 h with >99% eeD and 96%
reaction yield of d-1a. Taken together, these results clearly
indicate that stereoinversion from l to d-amino acid using
AlaDH and ARTAmut can lead to complete deracemization
contingent upon NAD+ recycling.
3.5 One‑Pot Deracemization
The complete deracemization of 10 mM dl-1a using the
cascade reactions led us to move on to use of a higher
substrate concentration with the activity levels of enzyme
components elevated (Fig. 3). Under the cascade reaction
conditions, 50 mM dl-1a was completely deracemized after
Fig. 3 One-pot production of d-1a by biocatalytic deracemization of
dl-1a. Reaction conditions: 50 mM dl-1a, 50 mM 3, 1 mM NAD+,
14 U mL−1 AlaDH, 16 U mL−1 ARTAmut and 5 U mL−1 NOX in Tris
buffer (50 mM, pH 7.0)
1 3

S.-W. Han, J.-S. Shin
Table 2 One-pot
Entry
Substrate
Reaction
Reaction
time (h)
Conversionb (%)
Reaction
eeD (%)
deracemization of dl-1b-d
conditionsa
yieldc (%)
1
2
3
4
5
6
dl-1b
dl-1b
dl-1c
dl-1c
dl-1d
dl-1d
A
B
A
B
A
B
28
9
28
20
28
48
82
>99
81
>99
10
88
90
93
85
95
66
90
82
>99
80
>99
19
88
aA: 50 mM dl-1b–d, 50 mM 3, 1 mM NAD+, 14 U mL−1 AlaDH, 16 U mL−1 ARTAmut and 5 U mL−1
NOX. B: 10 mM dl-1b-d, 50 mM 3, 1 mM NAD+, 140 U mL−1 AlaDH, 8 U mL−1 ARTAmut and
5 U mL−1 NOX
bConversion represents a ratio of consumption of l-amino acid to the initial amount of l-amino acid, i.e.
([l-amino acid]t=0 −[l-amino acid]t)/[l-amino acid]t=0
cReaction yield=[d-amino acid]t/[dl-amino acid]t=0
Acknowledgements This work was funded by the National Research
(i.e. less than 1% activity relative to that of l-1a). Under
the reaction conditions, 10 mM dl-1b and dl-1c were com-
pletely deracemized and led to 93 and 95% reaction yields
of the desired d-amino acid products with >99% eeD after 9
and 20 h, respectively (Table 2; entries 2 and 4). However,
l-1d was not completely converted to d-1d even after 48-h
reaction and the resulting eeD was 88% (entry 6).
Foundation of Korea under the Basic Science Research Program
(2016R1A2B400840). We thank Mrs Sae-Rom Park for technical
assistance in the cloning of NOX.
Compliance with Ethical Standards
Conflict of interest All authors declare that they have no conflict of
interest to disclose.
For demonstration of practical applicability, we per-
formed preparative-scale deracemization of dl-1a in a
25 mL reaction mixture under the reaction conditions of
100 mM dl-1a, 100 mM 3, 1 mM NAD+, 70 U mL−1 AlaDH
(73 µM), 1.6 U mL−1 ARTAmut (200 µM) and 10 U mL−1
NOX (2.4 µM). Complete deracemization was achieved after
24 h, leading to 95% reaction yield and >99% eeD of d-1a.
Product isolation using cation-exchange chromatography led
to recovery of optically pure d-1a (0.11 g, 52% isolation
yield). The purified d-1a was structurally characterized by
1H NMR, 13C NMR and MS (see Supplementary Materials).
References
1. Fuchs SA, Berger R, Klomp LW, de Koning TJ (2005) Mol Genet
Metab 85:168–180
2. Kiriyama Y, Nochi H (2016) Scientifica (Cairo) 2016:6494621
3. Genchi G (2017) Amino Acids 49:1521–1533
4. Gao X, Ma Q, Zhu H (2015) Appl Microbiol Biotechnol
99:3341–3349
5. Martinez-Rodriguez S, Martinez-Gomez AI, Rodriguez-Vico F,
Clemente-Jimenez JM, Heras-Vazquez LFJ (2010) Chem Biodiv
7:1531–1548
6. Tsai G, Coyle J (2001) US Patent 6,228,875 B1
7. Huirne JA, Lambalk CB (2001) Lancet 358:1793–1803
8. Yuasa Y, Nagakura A, Tsuruta H (2001) J Agric Food Chem
49:5013–5018
4 Conclusions
9. Ma JA (2003) Angew Chem Int Ed 42:4290–4299
10. Breuer M, Ditrich K, Habicher T, Hauer B, Kesseler M, Sturmer
R, Zelinski T (2004) Angew Chem Int Ed 43:788–824
11. Xue Y-P, Cao C-H, Zheng Y-G (2018) Chem Soc Rev
47:1516–1561
In this study, we developed a deracemization cascade for
preparation of d-amino acids using l-selective amino acid
dehydrogenase, d-selective ω-TA and NOX. Compared to the
previous deracemization processes using amino acid oxidase
or α-transaminase for oxidative deamination of an unwanted
amino acid enantiomer [24, 28], the proposed method ben-
efits from no generation of hydrogen peroxide and no need
for an expensive keto acid cosubstrate, respectively. In
addition, the deracemization method might be modified for
preparation of l-amino acids by switching enantioselectivity
of amino acid dehydrogenase and ω-TA. However, future
direction of the proposed strategy should be expansion of a
product pipeline by exploiting diverse amino acid dehydro-
genases and ω-TAs with a broad substrate range.
12. Vedha-Peters K, Gunawardana M, Rozzell JD, Novick SJ (2006)
J Am Chem Soc 128:10923–10929
13. Ager DJ, Fotheringham IG (2001) Curr Opin Drug Discov Dev
4:800–807
14. Park ES, Dong JY, Shin JS (2013) Org Biomol Chem
11:6929–6933
15. Park ES, Dong JY, Shin JS (2013) ChemCatChem 5:3538–3542
16. Park ES, Dong JY, Shin JS (2014) Appl Microbiol Biotechnol
98:651–660
17. Merino P, Marqués-López E, Tejero T, Herrera RP (2009) Tetra-
hedron 65:1219–1234
18. Yano S, Haruta H, Ikeda T, Kikuchi T, Murakami M, Moriguchi
M, Wakayama M (2011) J Chromatogr B 879:3247–3252
1 3

One-Pot Preparation of d-Amino Acids Through Biocatalytic Deracemization
19. Carboni C, Gardossi L, Tamiola K, Janssen DB, Quaedflieg PJLM
(2006) Tetrahedron Asymmetry 17:245–251
20. Krieg L, Ansorge-Schumacher MB, Kula MR (2002) Adv Synth
Catal 344:965–973
24. Park ES, Shin JS (2014) Adv Synth Catal 356:3505–3509
25. Schatzle S, Höhne M, Redestad E, Robins K, Bornscheuer UT
(2009) Anal Chem 81:8244–8248
26. Bhushan R, Bruckner H (2004) Amino Acids 27:231–247
27. Malik MS, Park ES, Shin JS (2012) Appl Microbiol Biotechnol
94:1163–1171
21. Isobe K, Tamauchi H, Fuhshuku K, Nagasawa S, Asano Y (2010)
Enzyme Res 2010:567210
22. Wegman MA, Janssen MHA, van Rantwijk F, Sheldon RA (2001)
Adv Synth Catal 343:559–576
28. Seo YM, Mathew S, Bea HS, Khang YH, Lee SH, Kim BG, Yun
H (2012) Org Biomol Chem 10:2482–2485
23. Servi S, Tessaro D, Pedrocchi-Fantoni G (2008) Coord Chem Rev
252:715–726
29. Geueke B, Riebel B, Hummel W (2003) Enzyme Microb Technol
32:205–211
1 3