Dynamic Kinetic Resolution for Asymmetric Synthesis of L-Noncanonical Amino Acids from D-Ser Using Tryptophan Synthase and Alanine Racemase

Article abstract of DOI:10.1002/ejoc.201901132

L-Ser is often used to synthesize some significant l-noncanonical α-amino acids(l-ncAAs), which are the prevalent intermediates and precursors for functional synthetic compounds. In this study, threonine aldolase from Escherichia coli k-12 MG1655 has been used to synthesize l-Ser. In contrast to the maximum catalytic capacity (20 g/L) for l-threonine aldolase(LTA), d-Ser was synthesized with high yield (240 g/L) from cheap Gly and paraformaldehyde using d-threonine aldolase (DTA) from Arthrobacter sp ATCC. In order to fully utilize d-Ser and expand the resource of l-Ser, a dynamic kinetic resolution system was constructed to convert d/dl-Ser to l-Ser through combining alanine racemase (Alr) from Bacillus subtilis with l-tryptophan synthase (TrpS) from Escherichia coli k-12 MG1655, and l-ncAAs including l-Trp and l-Cys derivatives were synthesized with excellent enantioselectivity and in high yields. The results indicated l-ncAAs could be efficiently synthesized from d-Ser using this original and green dynamic kinetic resolution system, and the reliable l-Ser resource has been established from simple and achiral substrates.

Full text of DOI:10.1002/ejoc.201901132

DOI: 10.1002/ejoc.201901132
Full Paper
Enzyme Catalysis
Dynamic Kinetic Resolution for Asymmetric Synthesis of L-Non-
canonical Amino Acids from D-Ser Using Tryptophan Synthase
and Alanine Racemase
Jinhai Yu,[a] Jing Li,[a] Xia Gao,[a] Shuiyun Zeng,[a] Hongjuan Zhang,*[b] Junzhong Liu,*[a] and
Qingcai Jiao*[a]
Abstract: L-Ser is often used to synthesize some significant system was constructed to convert
-noncanonical α-amino acids( -ncAAs), which are the prevalent combining alanine racemase (Alr) from Bacillus subtilis with
intermediates and precursors for functional synthetic com- -tryptophan synthase (TrpS) from Escherichia coli k-12 MG1655,
pounds. In this study, threonine aldolase from Escherichia coli and -ncAAs including -Trp and -Cys derivatives were synthe-
k-12 MG1655 has been used to synthesize -Ser. In contrast to sized with excellent enantioselectivity and in high yields. The
the maximum catalytic capacity (20 g/L) for -threonine aldol- results indicated -ncAAs could be efficiently synthesized from
ase(LTA), -Ser was synthesized with high yield (240 g/L) from -Ser using this original and green dynamic kinetic resolution
cheap Gly and paraformaldehyde using -threonine aldolase system, and the reliable -Ser resource has been established
(DTA) from Arthrobacter sp ATCC. In order to fully utilize -Ser from simple and achiral substrates.
-Ser, a dynamic kinetic resolution
D/DL-Ser to L-Ser through
L
L
L
L
L
L
L
L
L
D
D
D
L
D
and expand the resource of
L
Introduction
a building block chemical. The approaches to produce L-Ser in-
volve the extraction from keratin protein hydrolysates (e.g. hair
and feather) or poultry waste hydrolysates,[5] the fermentation
with sugar substrates (e.g. glucose and sucrose), the chemical
synthesis from α-haloacrylic acid derivatives,[6] and the enzy-
matic synthesis from Gly and formaldehyde using serine
hydroxymethyltransferase.[7] However, some disadvantages
make them less attractive such as the complicated process of
separation and purification, low productivity and stereoselecti-
vity, and the expensive cofactor (e.g. tetrahydrofolate) for serine
hydroxymethyltransferase catalysis.[7] Among these approaches,
L-noncanonical α-amino acids (L-ncAAs) are often used as inter-
mediates in biosynthesis and are prevalent precursors for func-
tional synthetic compounds, including over 12 % of the 200
top-selling pharmaceuticals, since they possess, at minimum,
two reactive functional groups (the amine and carboxylic acid)
and typically have at least one stereocenter. Therefore, increas-
ing attention has been devoted to the efficient synthesis of
L
-ncAAs in recent years.[1]
Enzymes are widely applied to synthesize
L
-ncAAs since they
not only can catalyze the reactions in mild conditions but also
could maintain highly stereoselective and obviate the need for
protecting groups, thereby decreasing the number of synthetic
steps.[2] Among these enzymes, tryptophan synthase (TrpS,
E.C. 4.2.1.20) is a pyridoxal 5-phosphate (PLP)-dependent en-
an alternative enzyme has been sought to synthesize L-Ser due
to the advantages of enzymatic synthesis. Threonine aldolase
has attracted considerable attention recently since it shows
good activity for preparation of ꢀ-hydroxy-α-amino compounds
from Gly and the corresponding aldehydes.[8] Since Ser pos-
sesses the characteristic of ꢀ-hydroxy-α-amino, we designed to
synthesize L-Ser using L-threonine aldolases (LTA) from E. coli
k-12 MG1655, and yet it showed weak activity only with the
maximum catalytic capacity 20 g/L. In contrast to LTA,
zyme that catalyzes the synthesis of
-Ser.[3] Frances H. Arnold et al. has reported that TrpS could
react with myriad indole analogs to provide a direct biocatalytic
route to -Trp derivatives, a kind of
-ncAAs.[4]
L-Trp from indole and
L
L
L
L-Ser as an essential substrate for TrpS is widely employed in
cosmetics, pharmaceuticals, food industries and also utilized as
D
-threonine aldolase (DTA) could synthesize
yield (240 g/L).[9] Therefore, how to efficiently convert
to -Ser and make full use of DL-Ser should be considered for
the better economic value of -Ser.
D
-Ser in excellent
D
/DL-Ser
L
D/
L
[a] Nanjing University,
State Key Laboratory of Pharmaceutical Biotechnology,
School of Life Science, Nanjing 210093, P. R. China
E-mail: junzhongliu2414@163.com
In the past decades, kinetic resolution (KR) has been widely
studied through using stereoselective lipase, acylase, amidase,
and carbamoylase to hydrolyze the racemates to obtain opti-
cally pure compounds. However, the theoretical yield of prod-
uct (≤50 %) for KR limits its further application in industry, and
consequently dynamic kinetic resolution (DKR) has been devel-
oped.[10] Compared with the traditional KR, DKR can provide
jiaoqc@163.com
[b] Nanjing Medical University,
School of Pharmacy, Nanjing 211166, China
E-mail: zhanghj@njmu.edu.cn
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under https://doi.org/10.1002/ejoc.201901132.
Eur. J. Org. Chem. 2019, 6618–6625
6618
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 1. Enzymatic synthesis of L-ncAAs from Gly and (CH2O)n using DTA, Alr and TrpS.
complete conversion to enantiomerically pure product by com-
bining the classical KR with continuous racemization of the un- acids including
reacted enantiomer, and now DKR is one of the most powerful (Scheme 1). The aim of this study was to develop a green, eco-
D
-Ser was racemized to
D, L-Ser to synthesize eight L-α-amino
L
-Trp, -Cys and their corresponding derivatives
L
and elegant methods for efficient synthesis of one enantiomer nomic enzymatic synthesis route of L-ncAAs from simple and
from racemic starting materials.[11]
achiral substrates.
Racemases have attracted considerable attention recently
because of their remarkable potency of racemization in DKR. As
one of the members of racemase family, alanine racemase (Alr,
EC 5.1.1.1) possesses the similar characteristics of racemase,
which is PLP-dependent enzyme that catalyze the reversible
Results and Discussion
Enzymatic Synthesis of Serine Using Threonine Aldolase
interconversion of
the structures of Ser and alanine are similar, we attempted to factors, such as pH, substrate concentration, and temperature.
perform the racemization of -Ser to -Ser using Alr, simulta- Herein the enzymatic activities were tested by varying one of
neously with the conversion of -Ser to -Trp and -Cys deriva- the bioconversion conditions at a time while keeping others
tives using TrpS. We expected that the source of
-Ser will be constant to investigate the influential factors systematically.[13]
expanded by DKR using the combination of two enzymes.
All the enzymatic reactions were optimized and the data
In this report, we focused on -Ser as starting compounds
-ncAAs synthesis through DKR because -Ser can be easily
obtained from the cheap Gly and paraformaldehyde ((CH2O)n)
using DTA. By combining Alr with TrpS in a one-pot reaction, perature (20–50 °C) and substrate concentration (0.5–4
D- and L
-forms of alanine.[12] Considering that Enzymatic activities can be affected by many environmental
D
D,L
L
L
L
L
weren't shown except DTA since the synthesis of
was rarely reported.
D-Ser by DTA
D
for
L
D
For DTA, the reaction conditions including pH (5–10), tem-
) were
M
Figure 1. The effect of pH and temperature on DTA activity. The reaction (volume: 10 mL) was carried out at 200 rpm, in which DTA dry whole cell (5 mg),
Gly (2.5 ), (CH2O)n (2.5 ) and PLP (0.4 m ) were contained. The mixture was shaken for 2 h at 30 °C for determination of the optimal pH and was shaken
at pH 7 for determination of the optimal temperature. 100 % relative activity denoted that 16 % conversion of Gly or (CH2O)n.
M
M
M
Eur. J. Org. Chem. 2019, 6618–6625
www.eurjoc.org
6619
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. a) The effect of substrate concentration on threonine aldolase activity. The mixture (volume: 10 mL) including threonine aldolase dry whole cell
(5 mg), Gly, (CH2O)n, PLP (0.4 m
1:1, the substrate concentrations varied as shown. b) Time course for the production of
pH 7.0) was carried out at 30 °C, 200 rpm, in which DTA dry whole cell (50 mg), Gly (2.5
M
) was reacted for 20 h at 40 °C, pH 8.0 for LTA and at 30 °C, pH 7.0 for DTA. The mol ratio of Gly and (CH2O)n was fixed at
-Ser from Gly and (CH2O)n using DTA. The reaction (volume: 100 mL,
), (CH2O)n (2.5 ) and PLP (0.4 m ) were contained.
D
M
M
M
optimized. The pH was fixed at the designated value by con- mulated in 14 h and increased very little after the subsequent
stantly adding sodium hydroxide solution (30 %, m/v) since pH 14 h, which indicated -Ser could be efficiently synthesized by
would gradually decrease during the reaction. Figure 1 and Fig- DTA in high yield after appropriate reaction time.
D
ure 2a show that the optimal pH, temperature and substrate
concentration for the reaction catalyzed by DTA were 7.0, 30 °C
and 2.5 M, respectively.
In terms of the reaction catalyzed by threonine aldolase,
formaldehyde as the substrate will result in the inhibition of
DTA to some extent, and the inhibition has been overcome
As described in Figure 2a, all conversions reached above through adding formaldehyde by fed batch. Compared with
91 % between 0.5 and 2.5 , and then the conversions de- formaldehyde, (CH2O)n could slowly release formaldehyde in
creased while the substrate concentration increased from 3.0
solution, which was equal to the fed batch. The scale-up reac-
to 4.0 for DTA catalysis. In the case of LTA, the highest conver- tion described that (CH2O)n was a good substrate for DTA since
sion was only 19 % (that was 20 g/L for 1.0 ), and the conver- the conversion of (CH2O)n (91.7 %, that was 240 g/L) was satis-
M
M
M
M
M
sion decreased above 1 M substrate concentration. The enzyme factory after 28 h (Figure 2b).
activities of DTA and LTA were 6.7 U/mg and 0.43 U/mg respec-
tively (Table 1), which indicated that LTA exhibited a weak activ-
ity towards Gly and (CH2O)n compared with DTA. Therefore, DTA
D- Ser Racemization with Alr
was used to synthesize
to 100 mL to obtain
D-Ser, and the reaction was scaled up
-Ser as the starting material for the follow- Though
D
D
-Ser could be synthesized in a good yield using DTA,
ing DKR process. Figure 2b shows that
D
-Ser was mainly accu- it is still not meaningful for the enzymatic reactions with -Ser
L
Table 1. Enzymes activities.
Substrate
Concentration
[m
Reaction
Enzyme
Enzyme[a]
M
]
time [min]
activity(U/mg)
C8H7N(1a) + L-Ser
2-CH3-C8H6N(1b) + L-Ser
5-CH3-C8H6N(1c) + L-Ser
7-Cl-C8H6N(1d) + L-Ser
NaSH(1e) + L-Ser
n-C3H7SH(1f) + L-Ser
C6H5SNa(1g) + L-Ser
C6H5CH2SH(1h) + L-Ser
D-Ser
20
20
20
20
20
20
20
20
20
20
20
15
15
15
15
15
15
15
15
15
15
15
TrpS
TrpS
TrpS
TrpS
TrpS
TrpS
TrpS
TrpS
Alr
1.2
1.5
1.4
0.46
5.4
5.3
3.7
4.5
9.4
6.7
0.43
+ Gly + (CH2O)n
Gly+(CH2O)n
DTA
LTA
[a] One unit (U) of enzyme activity was defined as the amount of dry whole cells catalyzing the conversion of substrate at a rate of 1 μmol/min.
Eur. J. Org. Chem. 2019, 6618–6625 www.eurjoc.org © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6620

Full Paper
as substrate if
D
-Ser couldn't be effectively converted to
L-Ser.
Table 2. Enzymatic synthesis of L-cystine, L
-Trp and their derivatives.[c]
Therefore, an enzyme was needed to racemize
D
-Ser to D,L-Ser
and simultaneously avoid racemizing the amino acid products.
Serine racemase was firstly considered, however, according to
the reported protocol,[14] serine racemase from Dictyostelium
discoideum could catalyze not only serine racemization but also
serine dehydration (α,ꢀ-elimination) to yield pyruvate and am-
Substrate (1a–1h)
Concentration
(M)
Reaction
Conversion[a] ee[b]
monia, which brings the consumption of
the side-product, consequently, decreases the conversion of
-Ser to the target -ncAAs.
Another amino acid racemase (AAR) from Pseudomonas
putida KT2440 (ATCC 47054) was shown to convert -Ser to
-Ser,[15] nevertheless, it also could racemize the other amino
D-Ser and produces
time(h)
[%]
(% ee)
C8H7N (1a)
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
14
6
6
24
24
4
>77
>85
>84
>34
>20
>95
>93
>94
>99
ND[d]
>99
>99
>99
>99
>99
>99
D
L
2-CH3-C8H6N (1b)
5-CH3-C8H6N(1c)
7-Cl-C8H6N(1d)
NaSH(1e)
n-C3H7SH (1f)
C6H5SNa(1g)
C6H5CH2SH(1h)
D
L
acids including the products, which would influence not only
the purity of products but also the dynamic kinetic resolution
process. Alr was selected because its substrate alanine is similar
to Ser and it does not catalyze racemization of the products. It
8
8
[a] The conversions were determined by HPLC and calculated according to
the substrates (1a–1h) concentration. [b] The enantiomeric excesses were
determined by the derivation of amino acid products with the Marfey's rea-
gent. [c] The reactions (volume: 10 mL, pH 8.5) were carried out at 40 °C and
was first used to racemize
their corresponding derivatives through DKR.
Figure 3 shows that the racemization had been almost com-
pleted within 1 h by Alr for low concentration -Ser (e.g. 0.2
and 0.4 ), and there was no inhibition over time even for high
concentration -Ser (e.g. 1.5 ) since its optical rotation steadily
D-Ser and synthesize L-Trp, L-Cys and
catalyzed by Alr (5 mg) and TrpS (5 mg) dry whole cells, in which 0.2 M D
Ser, PLP (0.4 m ) and 0.2 of each substrate was added. [d] Not detected.
-
M
M
D
M
D
M
Enzymatic Synthesis of L-2a, L-2b, L-2c, L-2d
As shown in Figure 4, the concentration of -Ser was the same
reached to zero after 10 h. Alr (9.4 U/mg) could catalyze the
racemization quickly for low concentration D-Ser and has the
D
as -Ser (0.084 ) in the mixture after 1 h, and the racemization
was well performed in accordance with converting -Ser only
by Alr (Figure 3). On the other hand, -Trp concentration in-
creased slowly which was still approximately 0.06 after 6 h
and subsequently almost kept unchanged. The low concentra-
tion of -Trp in solution was attributed to two factors, one was
the poor solubility of indole which made the efficiency of TrpS
decrease, the other was the low solubility of -Trp at pH 8.5
L
M
potential to racemize high concentration
hand, Alr is inactive with the target. Therefore, Alr was chosen
as the optimal racemase to catalyze the racemization of
in DKR system.
D-Ser. On the other
D
L
D-Ser
M
L
L
which quickly precipitated to decrease its concentration in solu-
tion and shift the reaction equilibrium toward the product for-
mation. Finally, the total yield of
ration and purification.
L-Trp was 65 % after the sepa-
Figure 3. The effect of
(100 mL, pH 8.5) containing
reacted at 40 °C, 200 rpm.
D
-Ser concentration on Alr activity. The mixture
-Ser, Alr dry whole cell (50 mg) and PLP (0.4 m
D
M)
Considering insolubility of the substrates (1a–1d), low con-
centration substrate (0.2 ) was used to synthesize -ncAAs
while investigating DKR system. Thereby -Ser (0.2 ) was used
M
L
D
M
according to the optimal reaction conditions and long reaction
time was avoided since Alr could almost completely catalyze
Figure 4. Time course for the enzymatic synthesis of
L-Trp using Alr and TrpS.
the racemization of low concentration
D
-Ser within 1 h (Table 2).
The reaction was performed as described in Table 2.
Eur. J. Org. Chem. 2019, 6618–6625
www.eurjoc.org
6621
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 5. Time course for the enzymatic synthesis of L-2e (a), L-2f (b). The reactions were performed as described in Table 2.
As observed for L-Trp, the other L-ncAAs (L-2b, L-2c L-2d) all selectivities by quantifying the enantiomeric purity of the prod-
quickly precipitated during the reactions and were synthesized ucts. In this study, analysis of the amino acids' enantiomeric
from 1b– d in a similar procedure, so their time courses weren't purity was divided into two parts including natural and unnatu-
shown again. The indole derivatives, except L-1d (>34 %), were ral amino acids. The enantiomers of natural amino acids could
all well converted with the excellent enantiomeric purities be obtained commercially, so we determined their enantio-
(Table 2).
meric purities by HPLC through modifying the conversion solu-
tion with Marfey's reagent and then comparing with the stan-
dards.[16]
Enzymatic Synthesis of L-2e, L-2f, L-2g, L-2h
For the
because they cannot be commercially bought as standards es-
pecially -isomers. Fortunately, we found amino acid racemase
Pseudomonas putida KT2440 (ATCC 47054) could racemize the
L-ncAAs, it was difficult to assay their optical purities
In this study, L-Cys derivatives could be synthesized through
DKR system with high yields, which showed that TrpS could act
on the wide range of mercaptans as substrates (Table 2). The
mercaptan derivatives are the important chiral intermediates
to prepare a variety of pharmaceuticals including HIV protease
inhibitors, anti-inflammatory analgesics, and oxytocin, there-
fore, it is important to efficiently synthesize these compounds.
In contrast to L-2f, L-2g, and L-2h, L-2e was produced in
low yield by DKR with sodium hydrosulfide as sulfur donating
reagent. As indicated in Figure 5a, at the beginning of the reac-
D
15]
L-ncAAs products,
which made it unsuitable to catalyze the
racemization of
D
-Ser in DKR. Therefore, the products were race-
mized by amino acid racemase and subsequently modified as
diastereomers with Marfey's reagent. The - and -amino acid
derivatives were separated by HPLC and the -amino acid deriv-
-iso-
D
L
L
ative was usually eluted before the corresponding
D
mer,[16,17] and thereby the absolute configuration was deter-
mined. The related results were shown in supporting informa-
tion,
D
-Ser decreased from 0.2
M to 0.17 M with L-Cys increasing
at half an hour, but after then the conversion in-
to 0.029
M
tion, which indicated all the analyzed products were L-configu-
creased slowly, even if the reaction time extended to 8 h, the
conversion only increased from 15 % to 22.5 %. Moreover, after
ration with >99 % ee (Figure S8-S22). The absolute configura-
tion of the product 2b was not detected since it couldn't be
racemized by amino acid racemase (Table 2).
2 h, Alr showed the difficulty in racemizing
it at 0.154 from 2 h to 8 h, and thereby the conversion of
-Cys wasn't very good in this DKR system. Though TrpS activity
(5.4 U/mg) for sodium hydrosulfide (1e) was the highest among
the substrates (1a–1h) and Alr (9.4 U/mg) could racemize -Ser
well (Table 1), the preparation of -Cys by combining TrpS with
Alr form -Ser couldn't be smoothly performed.
D-Ser with keeping
M
L
Conclusions
D
L
L-Ser is not only an important amino acid but also a key build-
ing block, so it is significant to obtain it economically and envi-
ronmentally friendly. In this study, we have synthesized D-Ser
with high yield (240 g/L) from Gly and (CH2O)n using DTA.
Herein a DKR system was reported through using the combina-
D
As described in Figure 5b, 1f was quickly converted into L-
2f with high conversion (>95 %) and ee value (>99 %) after 4 h.
Since L-2g and L-2h also continuously precipitated during the
reactions, the time courses for L-2g and L-2h were similar to
that of L-2a and not described again.
tion of Alr and TrpS to convert
ously convert -Ser to other -ncAAs, which could make full use
of -Ser and provide a desirable source of -Ser. Furthermore,
some important -Trp and -Cys derivatives could be efficiently
D/DL-Ser to L-Ser and simultane-
L
L
D
L
Analysis of the Amino Acids' Enantiomeric Purity
L
L
New chiral stereocenters are generally formed under the cataly- synthesized with excellent enantiomeric purities and in good
sis of the enzymes, so it is necessary to evaluate the enzyme yields by the DKR system.
Eur. J. Org. Chem. 2019, 6618–6625
www.eurjoc.org
6622
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
whole cells catalyzing the conversion of substrate at a rate of
1 μmol/min.
This work highlights not only an improved and complemen-
tary route to the conversion of -Ser to -Ser, but also the effec-
tive enzymatic synthesis of -Trp, -Cys and their derivatives by
D
L
L
L
Definition of TrpS Activity: TrpS activity was determined by de-
DKR. In addition, the DKR strategy reported in this study would tecting the concentrations of
derivatives. The mixture (final volume, 10 mL) containing 20 m
-Ser, 20 m the other corresponding substrates, PLP (0.4 m ) and
L-Trp, L-Cys and their corresponding
M
encourage us to uncover more DKR systems in which
a key substrate is used to efficiently synthesize optical purity
canonical or -ncAAs.
L-Ser as
L
M
M
5 mg TrpS whole cell was performed at 40 °C, pH 8.5 for 15 min.
One unit (U) of enzyme activity was defined as the amount of dry
whole cells catalyzing the conversion of substrate at a rate of
1 μmol/min.
L
Experimental Section
Enzymatic Synthesis of D- Ser: The reaction conditions were opti-
mized as shown in Figure 1 and Figure 2a. The scaled-up synthesis
Chemicals: Lactose and PLP were purchased from Sigma (St. Louis,
Mo. USA). Silica gel GF254 plate (5.0 × 10.0 cm) was purchased from
Haiyang Chemical Co. Ltd. (Qingdao, China). All other chemicals and
reagents used in this work were of analytical grade and purchased
from Sinopharm Chemical Reagent Co. Ltd (Shanghai, China). Re-
striction enzymes and kits for genetic manipulation were purchased
from Takara Bio (Dalian, China).
of
mixed with a solution (100 mL) containing Gly (18.8 g), (CH2O)n
(7.5 g) and PLP (0.4 m ) under pH 7.0, 30 °C at 200 rpm for 16 h.
D-Ser was operated as follows: DTA dry whole cells (50 mg) were
M
The reaction mixture was centrifuged to remove bacteria cells. And
then, the supernatant was concentrated under vacuum and the resi-
due was recrystallized from water/ethanol (1:2 v/v).
D-Ser (22 g) was
Microorganisms and Shake Flask Fermentation: LTA (Uniprot ID:
P75823) was amplified from genomic DNA of E. coli str. K-12 substr.
MG1655, two primers were designed as follows: forward primer is:
5′-TGGGGATCCATGCAATTCTCTCGC-3′ (BamHI site underlined); re-
verse primer is: 5′-GGGGAATTCTTAGTCAACCAGGATTTCGG-3′ (EcoRI
site underlined). The PCR fragment was inserted into pETDuet plas-
mid to construct the recombinant plasmid pETDuet-LTA.
obtained as white crystals in >99 % ee and 84 % yield.
Enzymatic Synthesis of L-Ser: -Ser was prepared from Gly and
L
(CH2O)n using LTA, the mixture (volume: 10 mL) containing LTA dry
whole cell (5 mg), Gly (1.88 g), (CH2O)n (0.75 g) and PLP (0.4 m
was reacted at 40 °C, pH 8.0 and 200 rpm.
M)
General Scaled-up Procedure for DKR Employing Alr and TrpS
Whole Cells: The mixture (100 mL, pH 8.5) containing -Ser (0.2 ),
the other corresponding substrates (0.2 ), PLP (0.4 m ), Alr
D
M
DTA from Arthrobacter sp ATCC 21022 (Uniprot ID: O82872) and
amino acid racemase from Pseudomonas putida KT2440 (ATCC
47054) (Uniprot ID: B8K1S4) in pETDuet plasmid, and Alr from Bacil-
lus subtilis (Uniprot ID: P10725) in pGEX-KG plasmid were all total
synthesized by Genscript Biotech Co. (Nanjing, China) with codons
optimized for expression in E. coli. The recombinant plasmids pET-
Duet-DTA, pETDuet-AAR, pGEX-KG-Alr, and pETDuet-LTA were all
transformed into E. coli BL21 (DE3) for the expression of DTA, AAR,
Alr and LTA respectively. TrpS from E. coli k-12 MG1655 (Uniprot ID:
P0A879) in pET28a plasmid was obtained according to our previous
studies.[3a]
M
M
(50 mg), and TrpS dry whole cells (50 mg) was shaken at 200 rpm
and 40 °C. Finally, the reaction was quenched by adding ammonia
water. 1H-NMR and 13C-NMR of the products were all provided in
supporting information (Figure S23-S38).
Enzymatic Synthesis of L-Tryptophan and its Derivatives: The
reaction solution was quenched by adjusting pH to 11 using ammo-
nia water (25 %–28 %) and then centrifuged at 10000 rpm for
10 min. The supernatant was decolorized by active charcoal (1 g)
at 50 °C for 20 min and subsequently filtered. The filtrate was con-
centrated under vacuum, and the crude product was recrystallized
All the media used for inoculums preparation were adjusted to pH
7.0 before autoclaving. For DTA, AAR, Alr, and LTA, a loopful of strain
was inoculated in a 100 mL Erlenmeyer flask containing 30 mL of
LB medium in which ampicillin (50 mg/L) was added in advance.
For TrpS, kanamycin (50mg/L) was added into LB medium. These
flasks incubated on a rotary shaker at 200 rpm and 30 °C for 12 h.
with water to obtain
L-Trp as white crystals in >99 % ee and 65 %
yield. The spectral characterization of
L-Trp was shown as follows:
1H-NMR (D2O) (Bruker DRX600, Germany) δ: 7.62 (d, 1H, J =8.0 Hz),
7.39 (d, 1H, J =8.2 Hz), 7.11–7.14 (m, 2H), 7.05 (dd, 1H, J =7.3 Hz,
7.7 Hz), 3.45 (dd, 1H, J =5.4 Hz, 7.1 Hz), 3.06 (dd, 1H, J =5.3 Hz,
14.5 Hz), 2.91 (dd, 1H, J =7.2 Hz, 14.5 Hz). 13C-NMR (D2O, Bruker
DRX600, Germany) δ: 182.9, 136.1, 127.2, 124.2, 121.7, 119.0, 118.8,
111.7, 110.5, 56.5, 30.3.
The obtained bacterial strains in LB medium were inoculated and
cultivated according to our previous protocols.[18] And then fer-
mented broth was collected at the end of the fermentation process
and centrifuged at 10000 rpm and 4 °C for 10 min to obtain wet
cells, which were lyophilized and stored at –20 °C for further
study.[18]
The other
L-Trp derivatives were obtained in the similar operation.
2-CH3-
L
-tryptophan: white powder, yield, 74 %. 1H-NMR (D2O)
(Bruker DRX600, Germany) δ: 7.50 (d, 1H, J =7.7 Hz), 7.27 (d, 1H, J =
8.0 Hz), 6.97–7.04 (m, 2H), 3.39 (dd, 1H, J =5.9 Hz, 7.3 Hz), 2.98 (dd,
1H, J =5.7 Hz, 14.5 Hz), 2.81 (dd, 1H, J =7.5 Hz, 14.4 Hz), 2.24 (s, 3H).
13C-NMR (D2O, Bruker DRX600, Germany) δ: 183.0, 135.1, 134.5,
128.4, 120.7, 118.9, 117.9, 110.7, 106.3, 56.8, 29.4, 10.7.
Definition of DTA and LTA Activity: Activities of DTA and LTA were
measured by detecting the formation of
tively. The standard assay solution (final volume, 10 mL) containing
20 m Gly, 20 m paraformaldehyde, PLP (0.4 m ) and dry whole
D-Ser and L-Ser respec-
M
M
M
5-CH3-L
-tryptophan: light gray solid, yield, 75 %. 1H-NMR (D2O)
cell (5 mg) was incubated for 15 min at 40 °C, pH 8.0 for LTA and
at 30 °C, pH 7.0 for DTA. One unit (U) of enzyme activity was defined
as the amount of dry whole cells catalyzing the conversion of sub-
strate to the product at a rate of 1 μmol/min.
(Bruker DRX600, Germany) δ: 7.42 (s, 1H), 7.28 (d, 1H, J =8.3 Hz),
7.07 (s, 1H), 6.98 (d, 1H, J =8.3 Hz), 3.44 (dd, 1H, J =5.3 Hz, 7.2 Hz),
3.03 (dd, 1 H, J = 5.2Hz, 14.4 Hz), 2.85 (dd, 1 H, J =7.4 Hz, 14.4 Hz),
2.32 (s, 3H). 13C-NMR (D2O, Bruker DRX600, Germany) δ: 182.9,
134.4, 128.7, 127.5, 124.4, 123.2, 118.1, 111.5, 110.1, 56.5, 30.3, 20.4.
Definition of Alr Activity: Activity of Alr was measured by detect-
ing the concentrations of
D-Ser and L-Ser. The standard assay solu-
tion (final volume, 10 mL) containing 20 mM D-Ser, PLP (0.4 m
M)
7-Cl-L
-tryptophan: white solid, yield, 20 %. 1H-NMR (D2O) (Bruker
and dry whole cell (5 mg) was incubated at 40 °C, pH 8.5 for 15 min.
One unit (U) of enzyme activity was defined as the amount of dry
DRX600, Germany) δ: 11.25 (s, 1H), 7.55 (d, 1H, J =7.9 Hz), 7.28 (d,
1H, J =2.0 Hz), 7.15 (d, 1H, J =7.5 Hz), 6.99 (t, 1H, J =7.7 Hz), 3.42
Eur. J. Org. Chem. 2019, 6618–6625
www.eurjoc.org
6623
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(dd, 1H, J =4.0 Hz, 8.5 Hz), 3.27 (dd, 1H, J =4.0 Hz, 15 Hz), 2.99 (dd,
1H, J =8.6 Hz, 15 Hz). 13C-NMR (D2O, Bruker DRX600, Germany) δ:
(FDAA).[16,17a] To a solution (1 mL, pH 8.5) of amino acid racemase
dry whole cells (1 mg) and PLP (0.4 mM), the unnatural amino acid
170.0, 133.5, 129.8, 125.9, 120.9, 119.8, 118.1, 116.2, 111.6, 55.1, 27.5. products (10 mg) were added respectively. After the reaction mix-
ture had been incubated whilst stirring at 40 °C for 6 h, the reac-
Enzymatic Synthesis of L-Cys: The reaction mixture was centri-
tions were centrifuged to remove bacteria cells, and then the super-
fuged and the supernatant was adjusted to pH 2.0 with HCl (6 M)
natants were derivatized with Marfey's reagent. The derivatives with
to convert the excessive sodium hydrosulfide to hydrogen sulfide,
FDAA were prepared according to the previously reported proto-
which was released from the solution and was trapped with NaOH
col.[16]
(6 M). Due to the instability of
L-Cys in the air, L-Cys was completely
oxidized into -cystine to calculate its yield. After hydrogen sulfide
was exhausted, the acidic solution was adjusted to pH 4.6 with
ammonia water, and hydrogen peroxide solution (2 mL, 6 %, m/v)
L
The diastereomers with FDAA were analyzed by the same Waters
e2695 instrument at 335 nm and 30 °C using 40 % v/v CH3CN in
H2O (0.1 % v/v trifluoroacetic acid) as the mobile phase with a flow
rate of 1 mL/min. The HPLC chromatograms were shown in Figure
S8-S22.
was subsequently added to oxidize
L-Cys to
L-cystine, which precipi-
tated and was collected by centrifugation.
L-cystine was purified as
follows: it was dissolved in HCl solution (pH 0.5) and decolorized
with active carbon at 70 °C for 30 min, followed by filtration. The
filtrate was neutralized to pH 4.6 with ammonia water to precipitate
The concentration and ee value of L-Cys and D/L-Ser were all de-
tected by HPLC through modifying with Marfey's reagent.[18] In this
study, all assays were performed in triplicates.
L
-cystine again.[18]
L-cystine was obtained as white crystals in
ee>99 % and 13 % yield. 1H-NMR (D2O) (Bruker DRX600, Germany)
δ: 3.47 (dd, 2H, J =4.7 Hz, 7.6 Hz), 3.01 (dd, 2H, J =4.7 Hz, 13.6 Hz),
2.80 (dd, 2H, J =7.6 Hz, 13.6 Hz) (m, 4H). 13C-NMR (D2O, Bruker
DRX600, Germany) δ: 180.8, 54.8, 43.5.
Acknowledgments
This work was financially supported by National Key R&D Pro-
gram of China (2017YFC0506005) and National Natural Science
Foundation of China (No. 21302100) and the Open Fund of
State Key Laboratory of Pharmaceutical Biotechnology of
Nanjing University, China.
Enzymatic Synthesis of S-Propyl-L-cysteine: The mixture was
centrifuged and the supernatant was concentrated under vacuum.
The residue was recrystallized with water-acetone (1:1, v/v).
S-propyl-L-cysteine was obtained as white crystals in ee>99 % and
84 % yield. 1H-NMR (D2O) (Bruker DRX600, Germany) δ: 3.31 (dd,
1H, J =5.1 Hz, 6.7 Hz), 2.76 (dd, 1H, J =5.0 Hz, 13.4 Hz), 2.67 (dd, 1H,
J =6.8 Hz, 13.4 Hz), 2.46 (t, 2H, J =7.3 Hz), 1.46–1.53 (m, 2H), 0.85 (t,
3H, J =7.4 Hz). 13C-NMR (D2O, Bruker DRX600, Germany) δ: 181.2,
55.1, 36.8, 33.7, 22.3, 12.6.
Keywords: Alanine Racemase · Dynamic Kinetic Resolution ·
D
-Threonine Aldolase ·
L-Noncanonical Amino Acids ·
Tryptophan Synthase
Enzymatic Synthesis of S-Phenyl- and S-Benzyl-L-cysteine: After
the reaction mixture had been incubated whilst stirring at 40 °C for
8 h, the pH of the reaction solution was adjusted to 11 to make
the product dissolve completely, and then the mixture was centri-
fuged to remove the whole cells and the other insoluble impurities.
The supernatant was concentrated under vacuum and the residue
[1] a) P. J. Almhjell, C. E. Boville, F. H. Arnold, Chem. Soc. Rev. 2018, 47, 8980–
8997; b) W. Leuchtenberger, K. Huthmacher, K. Drauz, Appl. Microbiol.
Biotechnol. 2005, 69, 1–8; c) E. S. Park, J. S. Shin, J. Mol. Catal. B: Enzym.
2015, 121, 9–14; d) M. A. T. Blaskovich, J. Med. Chem. 2016, 59, 10807–
10836; e) D. D. Young, P. G. Schultz, ACS Chem. Biol. 2018, 13, 854–870.
[2] a) R. J. M. Goss, P. L. A. Newill, Chem. Commun. 2006, 47, 4924–4925;
b) F. Zhang, Q. Z. Zheng, Q. C. Jiao, J. Z. Liu, G. H. Zhao, Amino Acids
2010, 39, 1177–1182; c) H. J. Zhang, J. Z. Liu, Y. P. Zhan, J. J. Chang, L. S.
Xu, Z. Y. Wang, Q. Liu, Q. C. Jiao, J. Mol. Catal. B: Enzym. 2013, 90, 128–131;
d) H. J. Zhang, Y. P. Zhan, J. J. Chang, J. Z. Liu, L. S. Xu, Z. Y. Wang, Q. Liu,
Q. C. Jiao, Biotechnol. Lett. 2012, 34, 1931–1935.
[3] a) G. H. Zhao, J. Z. Liu, K. Dong, F. Zhang, H. J. Zhang, Q. Liu, Q. C. Jiao,
Bioresour. Technol. 2011, 102, 3554–3557; b) D. R. M. Smith, T. Willemse,
D. S. Gkotsi, W. Schepens, B. U. W. Maes, S. Ballet, R. J. M. Goss, Org. Lett.
2014, 16, 2622–2625; c) L. S. Xu, Z. Y. Wang, J. Z. Liu, P. T. Mao, H. J.
Zhang, J. Gao, Q. Liu, Q. C. Jiao, Catal. Lett. 2012, 142, 282–286.
[4] a) C. E. Boville, R. A. Scheele, P. Koch, S. B. Chen, A. R. Buller, F. H. Arnold,
Angew. Chem. Int. Ed. 2018, 57, 14764–14768; b) M. Herger, P. V. Roye,
D. K. Romney, S. B. Chen, A. R. Buller, F. H. Arnold, J. Am. Chem. Soc.
2016, 138, 8388–8391; c) D. K. Romney, J. M. Calles, J. E. Wehrmuller, F. H.
Arnold, J. Am. Chem. Soc. 2017, 139, 10769–10776.
was recrystallized with water/methanol (1:1, v/v). S-phenyl-L-
cysteine was obtained as white crystals in ee>99 % and 76 % yield.
1H-NMR (D2O) (Bruker DRX600, Germany) δ: 7.35 (d, 2H, J =7.7 Hz),
7.26 (t, 2H, J =7.6 Hz), 7.18 (t, 1H, J =7.4 Hz), 3.29 (dd, 1H, J =4.9 Hz,
7.3 Hz), 3.21 (dd, 1H, J =4.8 Hz, 13.4 Hz), 2.98 (dd, 1H, J =7.4 Hz,
13.4 Hz). 13C-NMR (D2O, Bruker DRX600, Germany) δ: 180.6, 134.8,
129.6, 129.2, 126.8, 55.1, 39.0.[18]
Based on the similar operation, S-benzyl-L-cysteine was obtained as
white crystals in >99 % ee and 83 % yield, and its structure was
1
confirmed by H-NMR and 13C-NMR. 7.19–7.28 (m, 5H), 3.66 (s, 2H),
3.26 (dd, 1H, J =5.3 Hz, 6.7 Hz), 2.65 (dd, 1H, J =5.2 Hz, 13.5 Hz),
2.58 (dd, 1H, J =6.8 Hz, 13.5 Hz). 13C-NMR (D2O, Bruker DRX600,
Germany) δ: 181.0, 138.6, 129.0, 128.8, 127.3, 54.8, 36.3, 35.4.
Analytical Methods
[5] W. Jiang, B. Z. Xia, Z. D. Liu, Microbiol. Res. 2013, 168, 477–484.
[6] F. Effenberger, G. Zoller, Tetrahedron 1988, 44, 5573–5582.
Detection of the Product Concentrations by HPLC: Products ex-
cept cysteine could be directly analyzed by HPLC using Waters
e2695 separations module coupled with Waters 2998 photodiode
array detector, and XBridge C18 column (250 × 4.6 mm, 5 μm) at
210 nm and 30 °C. The mobile phase composed of 30 % v/v meth-
anol and 70 % v/v 0.05
at a flow rate of 0.8 mL/min. The HPLC chromatograms were shown
in Figure S1-S7.
[7] a) X. M. Zhang, L. P. Yao, G. Q. Xu, J. F. Zhu, X. J. Zhang, J. S. Shi, Z. H. Xu,
Biochem. Eng. J. 2017, 118, 113–122; b) Y. Li, G. K. Chen, X. W. Tong, H. T.
Zhang, X. G. Liu, Y. H. Liu, F. P. Lu, Biotechnol. Lett. 2012, 34, 1525–1530;
c) K. Hernandez, I. Zelen, G. Petrillo, I. Uson, C. M. Wandtke, J. Bujons, J.
Joglar, T. Parella, P. Clapes, Angew. Chem. Int. Ed. 2015, 54, 3013–3017; d)
Q. J. Zhu, X. M. Zhang, Y. C. Luo, W. Guo, G. Q. Xu, J. S. Shi, Z. H. Xu, Appl.
Microbiol. Biotechnol. 2015, 99, 1665–1673.
[8] a) J. Q. Liu, T. Dairi, N. Itoh, M. Kataoka, S. Shimizu, H. Yamada, J. Biol.
Chem. 1998, 273, 16678–16685; b) K. Fesko, Appl. Microbiol. Biotechnol.
2016, 100, 2579–2590; c) J. Steinreiber, K. Fesko, C. Reisinger, M. Schür-
mann, F. V. Assema, M. Wolberg, D. Mink, H. Griengl, Tetrahedron 2007,
63, 918–926.
M potassium dihydrogen phosphate solution
Analysis of the Enantiomeric Purities of the Products: The enan-
tiomeric purities of products were determined by derivation with
Marfey's reagent, 1-fluoro-2,4-dinitrophenyl-5-
L-alanine amide
Eur. J. Org. Chem. 2019, 6618–6625
www.eurjoc.org
6624
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[9]
[13] H. J. Zhang, Y. Wei, Y. Lu, S. P. Wu, Q. Liu, J. Z. Liu, Q. C. Jiao, Appl.
Microbiol. Biotechnol. 2016, 100, 1691–1700.
A. Tadashi, H. Tomonori, W. Seiichi, N. Keita, N. Kiyoteru, K. Mitsuo, (Mitsui
Chemicals Inc), EP Patent 180, 640, 1A1, 2005.
[14] a) T. Ito, M. Maekawa, S. Hayashi, M. Goto, H. Hemmi, T. Yoshimura, Amino
Acids 2013, 44, 1073–1084; b) T. Ito, H. Murase, M. Maekawa, M. Goto, S.
Hayashi, H. Saito, M. Maki, H. Hemmi, T. Yoshimura, Amino Acids 2012,
43, 1567–1576.
[15] a) A. D. Radkov, L. A. Moe, J. Bacteriol. 2013, 195, 5016–5024; b) M. Bech-
told, S. Makart, R. Reiss, P. Alder, S. Panke, Biotechnol. Bioeng. 2007, 98,
812–824.
[10]
a) A. Kirschner, U. T. Bornscheuer, Angew. Chem. Int. Ed. 2006, 45, 7004–
7006; b) W. Adam, Z. Lukacs, C. R. Saha-Möller, P. Schreier, J. Am. Chem.
Soc. 2000, 122, 4887–4892; c) O. Pàmies, J.-E. Bäckvall, J. Org. Chem.
2002, 67, 1261–1265; d) M. Pickl, A. Swoboda, E. Romero, C. K. Winkler,
C. Binda, A. Mattevi, K. Faber, M. W. Fraaije, Angew. Chem. Int. Ed. 2018,
57, 2864–2868; e) W. J. Choi, Appl. Microbiol. Biotechnol. 2009, 84, 239–
247.
[16] J. H. Yu, J. Li, S. Y. Cao, T. Wu, S. Y. Zeng, H. J. Zhang, J. Z. Liu, Q. C. Jiao,
Catal. Commun. 2019, 120, 28–32.
[17] a) R. Bhushan, H. Bruckner, Amino Acids 2004, 27, 231–247; b) D. R. Good-
lett, P. A. Abuaf, P. A. Savage, K. A. Kowalski, T. K. Mukherjee, J. W. Tolan,
J. W. Tolan, N. Corkum, G. Goldstein, J. B. Crowther, J. Chromatogr. A
1995, 707, 233–244.
[11] a) A. G. Applegatea, D. B. Berkowitza, Adv. Synth. Catal. 2015, 357, 1619–
1632; b) A. Cuetos, A. R. Martínez, F. R. Bisogno, B. Grischek, I. Lavandera,
G. D. Gonzalo, W. Kroutil, V. Gotor, Adv. Synth. Catal. 2012, 354, 1743–
1749; c) Y. Asano, S. Yamaguchi, J. Am. Chem. Soc. 2005, 127, 7696–7697;
d) K. Yasukawa, Y. Asano, Adv. Synth. Catal. 2012, 354, 3327–3332.
[12] a) T. Yoshimura. K. Soda, in: Molecular Aspects of Enzyme Catalysis (Eds.:
T. Fukui, K. Soda), Wiley-Blackwell, 1994, pp. 147–163; b) K. K. Nambu, Y.
Yasuda, K. Tochikubo, Amino Acids 2000, 18, 375–387; c) E. Ferrari, D. J.
Henner, M. Y. Yang, Nat. Biotechnol. 1985, 3, 1003–1007; d) K. J. Pierce,
S. P. Salifu, M. Tangney, FEMS Microbiol. Lett. 2008, 283, 69–74.
[18] J. Z. Liu, S. P. Wu, Y. Lu, Q. Liu, Q. C. Jiao, X. Z. Wang, H. J. Zhang, Chem.
Eng. J. 2016, 302, 146–154.
Received: August 1, 2019
Eur. J. Org. Chem. 2019, 6618–6625
www.eurjoc.org
6625
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim