Development of a multi-enzymatic desymmetrization and its application for the biosynthesis of L-norvaline from DL-norvaline

Article abstract of DOI:10.1016/j.procbio.2017.01.022

Perindopril is an effective antihypertensive drug in strong demand used to treat hypertension. L-norvaline is a vital intermediate of Perindopril production mainly produced by chemical synthesis with low purity. We developed an environmentally friendly method to produce L-norvaline with high purity based on a desymmetrization process. D-Norvaline was oxidized to the corresponding keto acid by D-amino acid oxidase from the substrate DL-norvaline. Asymmetric hydrogenation of the keto acid to L-norvaline was carried out by leucine dehydrogenase with concomitant oxidation of NADH to NAD⁺. A NADH regeneration system was introduced by overexpressing a formate dehydrogenase. The unwanted H₂O₂by-product generated during D-norvaline oxidation was removed by adding catalase. A total of 54.09?g/L of L-norvaline was achieved, with an enantiomeric excess over 99% under optimal conditions, with a 96.7% conversion rate. Our desymmetrization method provides an environmental friendly strategy for the production of enantiomerically pure L-norvaline in the pharmaceutical industry.

Full text of DOI:10.1016/j.procbio.2017.01.022

Accepted Manuscript
Title: Development of a multi-enzymatic desymmetrization
and its application for the biosynthesis of L-norvaline from
DL-norvaline
Authors: Yunlong Qi, Taowei Yang, Junping Zhou, Junxian
Zheng, Meijuan Xu, Xian Zhang, Zhiming Rao, Shang-Tian
Yang
PII:
S1359-5113(16)30881-9
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.procbio.2017.01.022
PRBI 10922
To appear in:
Process Biochemistry
Received date:
Revised date:
Accepted date:
16-11-2016
15-12-2016
23-1-2017
Please cite this article as: Qi Yunlong, Yang Taowei, Zhou Junping, Zheng Junxian,
Xu Meijuan, Zhang Xian, Rao Zhiming, Yang Shang-Tian.Development of a multi-
enzymaticdesymmetrizationanditsapplicationforthebiosynthesisofL-norvalinefrom
DL-norvaline.Process Biochemistry http://dx.doi.org/10.1016/j.procbio.2017.01.022
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Development of a multi-enzymatic desymmetrization and its application for
the biosynthesis of L-norvaline from DL-norvaline
a
a,b*
a
a
a
Yunlong Qi , Taowei Yang ytw1228@163.com, Junping Zhou , Junxian Zheng , Meijuan Xu , Xian
Zhang , Zhiming Rao raozhm@jiangnan.edu.cn, Shang-Tian Yangc
a
a,b*
aThe Key Laboratory of Industrial Biotechnology, Ministry of Education; Laboratory of Applied
Microorganisms and Metabolic Engineering, School of Biotechnology, Jiangnan University, Wuxi, Jiangsu
Province 214122, P. R. China
bJiangnan University (Rugao) Food Biotechnology Research Institute, Rugao, Jiangsu Province 226500,
China
cDepartment of Chemical and Biomolecular Engineering, The Ohio State University, Columbus, Ohio
4
3210, United States
*Corresponding author at: School of Biotechnology, Jiangnan University, 1800 Lihu Avenue, Wuxi,
Jiangsu 214122, P. R. China. Tel: +86-510-85916881.
1

Graphical Abstract
2

Highlights:
·
·
·
Development of a multi-enzymatic desymmetrization of L-norvaline from DL-norvaline.
The catalytic system obtained various L-amino acids with varied optical purities.
The system produced L-norvaline with a high yield, enantiomeric purity and titer.
3

Abstract
Perindopril is an effective antihypertensive drug in strong demand used to treat hypertension. L-norvaline
is a vital intermediate of Perindopril production mainly produced by chemical synthesis with low purity. We
developed an environmentally friendly method to produce L-norvaline with high purity based on a
desymmetrization process. D-Norvaline was oxidized to the corresponding keto acid by D-amino acid oxidase
from the substrate DL-norvaline. Asymmetric hydrogenation of the keto acid to L-norvaline was carried out by
+
leucine dehydrogenase with concomitant oxidation of NADH to NAD . A NADH regeneration system was
introduced by overexpressing a formate dehydrogenase. The unwanted H
2
2
O by-product generated during
D-norvaline oxidation was removed by adding catalase. A total of 54.09 g/L of L-norvaline was achieved, with
an enantiomeric excess over 99% under optimal conditions, with a 96.7% conversion rate. Our
desymmetrization method provides an environmental friendly strategy for the production of enantiomerically
pure L-norvaline in the pharmaceutical industry.
Keywords: Desymmetrization; D-amino acid oxidase; Leucine dehydrogenase; NADH regeneration.
4

Introduction
Hypertension, or high blood pressure, is a long-term medical condition that is partially caused by a
high-salt, high-fat diet or a stressful lifestyle. Hypertension plagues more than 25% of adults globally and is
estimated to be a factor in over 7 million death each year [1]. Several antihypertensive drugs have been
developed to treat high blood pressure. Perindopril is an effective antihypertensive drug that is commonly used
and is in strong demand [2]. Perindopril treatment decreased the relative risks of strokes and intracerebral
hemorrhages in patients by 26% and 49%, respectively, during a 3.9-year period [3]. L-Norvaline is a vital
intermediate in the chemical synthesis of Perindopril. Several strategies have been developed for production of
L-norvaline. The conventional chemical synthesis of L-norvaline is expensive and has many disadvantages, such
as a complicated process, low enantiomeric excess (e.e.<75%) and negative environmental impacts [4] [5].
L-Norvaline can also be produced by bacterial fermentation with an appropriate substrate. Stoynova et al.
devised a method to produce L-norvaline using a bacterium from the Enterobacteriaceae family by inactivating
its acetohydroxy acid synthases [6]. Soriano et al. reported the production of L-norvaline from racemic
N-Formyl-DL-norvaline and N-Carbamoyl-DL-norvaline catalyzed by L-N-carbamoylase and
N-succinyl-amino acid racemase [7]. Han et al. employed ω-Transaminase to carry out the asymmetric synthesis
of L-norvaline from a keto acid using amino donors [8]. Chiriac et al. developed an enzymatic synthesis of
L-norvaline using Leucine dehydrogenase as a catalyst and glucose dehydrogenase plus galactomutarotase as a
NADH regeneration system [9].
Enzyme-mediated chiral resolution is a common method for the large-scale production of a single enantiomer
of a chiral compound because this method is high efficiency and has high stereo-, regio- and enantioselectivity
[
10]. Deracemization of chiral compounds via a enantioselective amine oxidases and unspecific reduction
process was first introduced by Turner [11], [12]. D-Amino acid oxidase (DAAO, EC 1.4.3.3, GenBank
5

accession number: AAR98816.1) is a well-known flavoenzyme that catalyzes the oxygen-dependent oxidative
deamination of D-amino acids to corresponding α-keto acids with absolute stereo-specificity, which also yields
hydrogen peroxide (H
2
O
2
) and NH as by-products [13]. DAAO has been attracting considerable attention
3
because of its prospects for the production of 7-aminocephalosporanic acid (7-ACA), the precursor of other
semi-synthetic cephalosporins [14]. DAAO is also widely used for chiral resolutions to produce single
enantiomers by catalyzing the oxidative deamination of D-amino acids to corresponding α-keto acids [15], [16]
and [17]. DAAO from porcine kidneys or Trigonopsis variabilis was employed to prepare a keto acid by
oxidation of racemic 6-hydroxy norleucine, and subsequent reductive amination of the keto acid catalyzed by
glutamate dehydrogenase transformed the mixture completely to give L-6-hydroxy norleucine exclusively [18],
[
19]. Leucine dehydrogenase (LeuDH, EC 1.4.1.9, GenBank accession number: KPU53554.1) is an
NAD(H)-dependent oxidoreductase that is widely used with NADH as a coenzyme for the enantioselective
production of L-leucine and other branched-chain L-amino acids [20]. LeuDH can catalyze the conversion of
aliphatic 2-oxo-acids to the corresponding L-amino acids via reductive amination [21]. NADH acts as an
important coenzyme during catalytic reactions of LeuDH. NADH is costly and its stoichiometric use is
economically unfeasible. Various in situ NADH regeneration strategies have been developed, such as
enzymatical, biological, electrochemical, chemical and photochemical methods [22]. Formate dehydrogenase
(FDH, EC 1.2.1.2, GenBank accession number: 013437) is one enzyme capable of regenerating of NADH,
+
which catalyzes the oxidation of formate to carbon dioxide followed by the reduction of NAD to NADH. FDH
is extensive used as an enzymatic NADH regeneration system in industrial production because formate is an
innocuous substrate and the CO produced in the regeneration off gasses, which drives the process towards
2
irreversible NADH regeneration [23]. Jiang et al. constructed a regulatory expression system of LeuDH and
FDH for the synthesis of L-Tle, which could be used to synthesize other L-amino acid and chiral drug
6

intermediates [24].
We proposed an alternative strategy for production of L-norvaline through enzyme-mediated kinetic
resolution combined with asymmetric hydrogenation of DL-norvaline based on the functions of these three
enzymes. In addition, catalase (EC 1.11.1.6, GenBank accession number: KT963080) was added to degrade the
unwanted H
2
2
O by-product generated during the oxidation of D-norvaline [16], [25] and [26]. This paper
provides an effective method for converting DL-norvaline to L-norvaline without any keto acid by-product.
Materials and methods
Chemicals
DL-Norvaline (content>98.5%, D-norvaline: L-norvaline ratio, 1:1) was purchased from Kebang Chemical
Co., Ltd (Henan, China). D-Norvaline (purity 99%) and L-norvaline (purity 99%) were purchased from
Sigma-Aldrich Co. LLC (Shanghai, China). All other chemicals were of reagent grade.
Bacterial strains and plasmids construction
The bacterial strains, plasmids and primers used in this study are listed in Table s1. The gene daao, encoding
D-amino acid oxidase (DAAO), was cloned from T. variabilis using primers P1/P2 or P3/P4 (primers P3/P4
were for pXMJ19-daao construction). The gene leudh, encoding LeuDH, was amplified from Bacillus cereus
using primers P5/P6. The gene fdh, encoding FDH, was cloned from Candida boidinii with primers P7/P8. The
plasmids pET-28a (+)-leudh and pET-28a (+)-fdh were separately transformed into Escherichia coli BL21 (DE3)
cells by the calcium chloride method. The plasmid pXMJ19-daao was transformed into Corynebacterium
glutamicum ATCC 13032 using the method described by Xu et al. [27].
Preparation of DAAO, LeuDH and FDH
The recombinant E. coli and C. glutamicum strains were cultivated in 100 mL of Terrific Broth (TB)
medium (24 g/L yeast extract, 12 g/L peptone, 4 g/L glycerol, 2.31 g/L KH
2
PO
4
, 12.54 g/L K
2
HPO ) at 37 °C
4
7

and 30 °C, respectively, until the OD600 reached 0.6–0.8. Isopropyl β-D-Thiogalactoside (IPTG, 0.8 mM) was
added and incubation continue for 12 h. Cells were harvested by centrifugation at 10000 rpm and 4 °C for 10
min and washed twice with 0.1 M sodium phosphate buffer (pH 7.5). The cells were resuspended with 10 mL of
0
.1 M sodium phosphate buffer (pH 7.5). The dry cell weights of ATCC13032/pXMJ19-daao, BL21
(DE3)/pET-28a (+)-leudh and BL21 (DE3)/pET-28a (+)-fdh were 6.32, 5.14 and 3.06 g/L, respectively.
Lysozyme (100 mg/mL) was used to aid C. glutamicum cell lysis before ultrasonic decomposition. Cells were
sonicated for 3 s 100 times with 7 s cooling intervals. Cell disruption by high-pressure homogenizer was
conducted at a pressure of 1000 bar for 20 min at 4 °C. The lysed cells were centrifuged at 12,000 rpm and 4 °C
for 25 min, and the supernatant was used for enzyme assays.
Assays of DAAO, LeuDH and FDH activities
DAAO activity was assessed spectrophotometrically by measuring the formation of 2-oxovaleric acid at
5
50 nm [28]. LeuDH activity was measured by spectrophotometrically monitoring the consumption of NADH in
−1
−1
the reductive amination at 340 nm (ɛ=6220 M cm ) [21]. FDH activity was assayed by the method described
by Berrios-Rivera et al. [29]. A Bradford Protein Assay Kit (Sangon Biotech Co, Ltd) was used to determine
protein concentration.
Production of L-norvaline by enzyme-mediated desymmetrization
DAAO, LeuDH and FDH were added according to the optimal dosage ratio of the enzymes into a 50-mL
enzymatic conversion mixture (0.1 M sodium phosphate buffer, pH 7.0) containing 1 M ammonium formate, 5%
glycerol, 5 g/L DL-norvaline and 30 U/mL catalase (prepared by our laboratory [30]). The enzymatic resolution
was performed on a magnetic stirrer at 25 °C and 160 rpm. Formate (20%) or ammonium hydroxide (50%) were
added to maintain the pH at 7.0 during the process.
Production of L-norvaline by fed-batch
8

The three crude enzymes, DAAO, LeuDH and FDH, were released using a high-pressure homogenizer.
The crude enzymes were added according to their optimal dosages to a 5-L reactor containing 2 L of 0.1 M
sodium phosphate buffer (pH 7.5). An initial DL-norvaline charge of 5.0 g/L was added to the reactor, with a
further 5.0 g/L added into the transformation system to continue the enzymatic reaction when almost all the
D-norvaline was converted to L-norvaline. The temperature, pH and rotor speed were kept at 30 °C, 7.5 and 160
rpm, respectively.
Analysis of products
The concentrations of D- and L-norvaline were assayed by HPLC (Agilent 1260, Agilent Technologies,
Palo Alto, CA, USA) with a UV detector at 338 nm as described by Denton et al. [31] using a Chiralcel OD-3R
column (150×4.6 mm, 3-Micron, Daicel). The eluent was composed of buffer A (0.1% H
3
PO in water) and
4
buffer B (100% MeCN), with gradient elution at 40 °C and a flow rate of 1 mL/min [31]. Conversion rate was
calculated by the following equation:
produced mL-norvaline
Conversion rate 
100%
mD-norvaline
where mL-norvaline is the titer of L-norvaline produced from D-norvaline and mD-norvaline is the titer of D-norvaline
contained in the substrate DL-norvaline.
Results and Discussion
Preparation of DAAO, LeuDH and FDH
LeuDH from Bacillus cereus and FDH from Candida boidinii were successfully overexpressed in E. coli
BL21 (DE3). The specific activity of crude LeuDH in the recombinant strain was 2.96 U/mg, which was
5
9.2-fold higher than that of the parent strain (Table s2). This activity was a little lower than that of LeuDH from
Bacillus sphaericus (GenBank accession number: BAD11020) [32] likely due to the differences of amino acid
sequence (sequence identity was 79.12%). The specific activity of crude FDH from the engineered E. coli was
9

0
.39 U/mg, and no FDH activity was detected in the wild type E. coli BL21 (DE3). DAAO expression failed in
E. coli BL21 (DE3). Expression of DAAO in E. coli BL21 (DE3) has previously been reported to resulted in a
marked decrease of cell viability owing to the toxicity of DAAO activity toward D-alanine, an essential amino
acid for cell wall construction [33], [34]. We screened other species for a candidate to host DAAO production
and the gram-positive bacteria C. glutamicum ATCC 13032 was chosen as a suitable host for DAAO expression.
The activity of crude DAAO in the engineered C. glutamicum was 0.3 U/mL, which was slightly lower than that
of other DAAOs likely due to the different substrates and amino acid sequence differences [35], [36]. No
DAAO activity was detected in the parent C. glutamicum ATCC 13032 strain.
Substrate spectrum of the multi-enzymatic desymmetrization system
Both DAAO and LeuDH have broad substrate spectrums, including various D-amino acids and ɑ-keto
acids [37], [38]. Desymmetrizations of a variety of DL-amino acids were attempted to produce enantiomeric
pure L-amino acids with our system. L-Methionine, L-valine, L-arginine, L-leucine, L-histidine L-norvaline and
L-phenylalanine were all produced form their respective racemates (Table 1). The system was highly effective
for the production of L-valine, obtaining an e.e of >99%. However, DAAO was more active towards D-valine
than D-histidine, and the conversion rate of D-histidine to L-histidine was much slower than that of D-valine to
L-valine [37].
Production of L-norvaline by the enzyme-mediated desymmetrization
The lysate supernatants of recombinant strains E. coli BL21 (DE3)/pET-28a (+)-fdh, E. coli BL21
(DE3)/pET-28a (+)-leudh and C. glutamicum ATCC 13032/pXMJ19-daao were used as catalysts. The dosages
of DAAO, LeuDH, and FDH in the conversion mixture were 0.6, 1 and 0.8 U/mL, respectively. Due to little
activity of LeuDH was detected in the parent strain E. coli BL21 (DE3), so a reaction blank without the
recombinant LeuDH extract was also investigated, D-norvaline was oxidased to corresponding keto acid
1
0

completely but only 0.04 g/L L-norvaline was produced from D-norvaline during 2 h of biocatalysis. When
recombinant LeuDH extract was employed as catalyst, 0.48 g/L D-norvaline was transformed to L-norvaline at
2
h. It is probably that there were some LeuDH homologs produced in E. coli BL21 resulting in LeuDH activity
detected in the parent strain, but these LeuDH homologs showed much lower catalytic capacity for the keto acid
oxidased from D-norvaline than that of the recombinant LeuDH. However, by-product H
2
2
O was generated
during the oxidation of D-norvaline by DAAO, which has deleterious effects on enzyme catalysts and
decarboxylates the intermediate keto acid [39], [40]. This degradation of the intermediate keto acid would have
decrease the product yield drastically. Catalase (30 U/mL) was added to degrade the unwanted by-product H
2
2
O ,
eliminating its toxic effects on the catalysts and intermediate. With this addition, 1.66 g/L D-norvaline was
converted into L-norvaline after 2 h of biocatalysis.
Optimization of enzyme-mediated L-norvaline biocatalyzed desymmetrization conditions
Each enzyme catalyst involved in the conversion has its own optimal pH and temperature conditions. The
optimal temperatures of DAAO and LeuDH are 25 °C and 55 °C, respectively, and the optimal pH of DAAO
and LeuDH are 8.0 and 8.5, respectively [41], [42], [43]. The optimal pH of FDH is around 7.5–8.5, with an
optimal temperature of 37 °C [44]. The pH and temperature of the biocatalysis system should be optimized to
achieve the highest product yield and productivity because the optimal pH and temperature conditions are not
the same for all the enzyme catalysts. The effects of pH on the enzymatic resolution were investigated between
5
7
.5–10.5. The conversion rate improved as pH increased until the maximum conversion rate was reached at pH
.5, then the conversion rate declined gradually with further increases in pH (Fig. 1a). Temperature can also
affect the efficiency of enzymatic resolutions processes. The effect of temperature on our enzymatic resolution
process was investigated by carrying out six enzymatic resolutions at temperatures varying from 20 to 45 °C at
pH 7.5. The conversion rate gradually increased with temperature up to 30 °C and then decreased at higher
11

temperatures, probably because of thermal inactivation of the enzymes at high temperature (Fig. 1b). The
optimal conversion conditions were at pH 7.5 and a temperature of 30 °C, which were employed in the
following experiments.
DAAO and LeuDH act as the main catalysts that participate in the kinetic resolution process directly. The
catalytic rates of DAAO and LeuDH were different, and the dosage ratio of DAAO and LeuDH in the
conversion mixture should be optimized in order to balance the individual catalytic steps. The activities of
DAAO, LeuDH and FDH were considered stable according to previous reports [41], [45], [46]. The initial rates
of DAAO and LeuDH are shown in Fig. s1a and b, respectively, and the slopes of these plots were used to
calculate the conversion rates and compare the catalytic rates of DAAO and LeuDH. The catalytic rates of
DAAO to convert D-norvaline to the keto acid and LeuDH to convert the keto acid to L-norvaline were 0.0015
g/(L·min·U) and 0.003 g/(L·min·U), respectively. Thus, the optimal ratio of DAAO to LeuDH activity was
theoretically 2:1 because the ratio of catalytic rates was 1:2, and the highest conversion rate was achieved at this
ratio (Fig. 2a). The rate dropped gradually when the ratio exceeded 2:1 because DAAO suffered from product
inhibition at a high concentration of keto acid (Fig. s2) [47], and a high concentration of the keto acid also
inhibited the LeuDH activity (Fig. s3) [48]. The optimal dosage ratio of DAAO to LeuDH activity of 2:1 was
employed in the following experiments.
The optimal dosage ratio of LeuDH to FDH activity was also investigated, with the ratio of DAAO to
LeuDH activity kept at 2:1. Only 0.7 g/L of L-norvaline was produced in 2 h when no FDH was added (Fig.
2
b-f). The amount of L-norvaline produced in 2 h was 2.55 g/L when the dosage ratio of LeuDH to FDH activity
was 1:0.5, which was 3.5-fold higher than that of no FDH added. This indicates that FDH activity is the
rate-limiting step. However, further increases in the proportion of FDH activity failed to improve the production
of L-norvaline. There is a sequentially ordered ternary/binary kinetic mechanism for the reductive amination
1
2

process catalyzed by LeuDH, in which NADH, keto acid and ammonia bind to LeuDH in that order [49].
NADH showed noncompetitive inhibition with the keto acid and a high concentration of NADH prevented the
keto acid binding to LeuDH, leading to a low production of L-norvaline. We used the optimal dosage ratio of
LeuDH to FDH activity of 1:0.5 in subsequent experiments.
Fed-batch conversion under optimal conditions for biocatalysis of L-norvaline
A fed-batch strategy could effectively improve the concentration of the desired product. High DL-norvaline
concentration inhibits the conversion, and high L-norvaline production could be attained using a fed-batch
strategy to mitigate the inhibition. We used an initial DL-norvaline concentration of 5.0 g/L and added a further
5
g/L charge into the conversion system when the D-norvaline was all almost converted to L-norvaline. As
shown in Fig. 3, the L-norvaline yield increased rapidly during the nine fed-batches and 23.74 g/L of
D-norvaline was converted into L-norvaline, with an average productivity of 1.187 g/(L·h). However, total
L-norvaline production increased slowly after the tenth fed-batch and the productivity gradually decreased
because of inactivation of the enzyme catalysts (Fig. s4) and limitation caused by high product concentration
(
Fig. s5). Almost no intermediate keto acid accumulated during the conversion process, which is favorable for
separation and purification of the product. A total of 26.49 g/L of D-norvaline was consumed in 25 h, and the
L-norvaline titer reached 54.09 g/L with a conversion rate of 96.3% and an e.e. over 99%.
Several chemical synthesis and biosynthesis methods for L-norvaline production are listed in Table 2.
ω-Transaminase was adopted for the asymmetric synthesis of L-norvaline from a keto acid using the amino
donor isopropylamine and gave almost 50 mM (5.8 g/L) of L-norvaline in 2 h with a conversion rate of 99.3%
[
8]; however, an unwanted byproduct isopropanol was generated during the deamination of amino donor
isopropylamine, which would increase the cost of product purification. Taking advantage of the catalytic
promiscuity of L-N-carbamoylase and N-succinyl-amino acid racemase, Soriano et al evaluated the production
1
3

of L-norvaline starting from racemic N-formyl-DL-norvaline and N-Carbamoyl-DL-norvaline [7], but only 7.5
mM (0.87 g/L) of substrate was converted in 3 h with a conversion rate of 95%. Furthermore, a stereospecific
enzymatic synthesis of L-norvaline from the corresponding keto acid catalyzed by Leucine dehydrogenase
combined with glucose dehydrogenase/galactose mutarotase has also been reported and 0.11M (12.3 g/L) keto
acid was transformed to L-norvaline with the yield of 92% in 2-3h [9]. Galkin et al described a method for the
conversion of keto acid to L-norvaline with recombinant E.coli cells which contain plasmids with leudh and fdh
genes necessary for biotransformation, 0.38 M (44.4 g/L) L-norvaline was synthesized in 12 h [50]. In this study,
we employed DL-norvaline as substrate for biosynthesis of L-norvaline, and the titer of L-norvaline reached to
5
4.09 g/L in 25 h with a productivity of 1.06 g/ (L·h) and a conversion rate of 96.7%. Furthermore, no
byproduct was generated. The L-norvaline titer, and yield obtained in this work could be comparable to the most
efficient strategy (Table 2).
In conclusion, an enzyme-mediated desymmetrization system for L-norvaline production was successfully
developed by applying DAAO, LeuDH, FDH and catalase. Optimization of conversion conditions and a
fed-batch strategy were investigated to improve the conversion yield. This study provides a promising
bioconversion method for enantiomerically pure L-norvaline production. However, the preparation of the
conversion system is time-consuming because bacteria do not naturally secrete these enzymes extracellularly
and cell wall disruption is necessary. Extracellular expression of the enzyme catalysts is more desirable and our
group is undertaking work to achieve this goal.
Acknowledgements
This work was supported by the National Natural Science Foundation of China (863 Program)
(2015AA021004), Jiangsu Province Science Fund for Distinguished Young Scholars (BK20150002), Natural
Science Foundation of Jiangsu Province (BK20161292), National Natural Science Foundation of China
1
4

(31570085), the 111 Project (111-2-06), and the Priority Academic Program Development of Jiangsu Higher
Education Institution.
1
5

References
[
1] R.H. Fagard, Resistant hypertension, Heart 98(3) (2012) 254-261.
[
2] J.R. Sowers, M. Epstein, E.D. Frohlich, Diabetes, hypertension, and cardiovascular disease an update,
Hypertension 37(4) (2001) 1053-1059.
[
3] N. Chapman, R. Huxley, C. Anderson, M. Bousser, J. Chalmers, S. Colman, S. Davis, G. Donnan, S.
MacMahon, B. Neal, Effects of a Perindopril-based blood pressure–lowering regimen on the risk of
recurrent stroke according to stroke subtype and medical history The progress trial, Stroke 35(1) (2004)
116-121.
[
4] C. Qian, L. Gong, X.Z. Chen, Preparation of optical active aliphatic α-amino acid from fatty acid:
synthesis of L-norvaline and D-norvaline, Res Chem Intermed 35(1) (2009) 117-121.
[
5] R. Xie, L.Y. Chu, J.G. Deng, Membranes and membrane processes for chiral resolution, Chem Soc
Rev 37(6) (2008) 1243-63.
[
6] N. Stoynova, E. Sycheva, E. Preobrazhenskaya, A. Novikova, N. Matrosov, Method for producing
abnormal amino acids using a bacterium of the Enterobacteriaceae family having all acetohydroxy acid
synthases inactivated, US. Patent, US 20060057685A1.
[
7] P. Soriano-Maldonado, F.J. Las Heras-Vazquez, J.M. Clemente-Jimenez, F. Rodriguez-Vico, S.
Martínez-Rodríguez, Enzymatic dynamic kinetic resolution of racemic N-formyl- and
N-carbamoyl-amino acids using immobilized l-N-carbamoylase and N-succinyl-amino acid racemase,
Appl Microbiol Biotechnol 99 (1) (2015) 283-291.
[
8] S.-W. Han, E.-S. Park, J.-Y. Dong, J.-S. Shin, Active-site engineering of ω-Transaminase for
production of unnatural amino acids carrying a side chain bulkier than an ethyl substituent, Appl Environ
Microbiol 81(20) (2015) 6994-7002.
[
9] M. Chiriac, I. Lupan, N. Bucurenci, O. Popescu, N. Palibroda, Stereoselective synthesis of L-[15N]
amino acids with glucose dehydrogenase and galactose mutarotase as NADH regenerating system, J
Labelled Comp Radiopharm 51(4) (2008) 171-174.
[
10] A. Kamal, M.A. Azhar, T. Krishnaji, M.S. Malik, S. Azeeza, Approaches based on enzyme mediated
kinetic to dynamic kinetic resolutions: a versatile route for chiral intermediates, Coord Chem Rev 252(5)
(
[
2008) 569-592.
11] N.J. Turner, Enzyme catalysed deracemisation and dynamic kinetic resolution reactions, Curr Opin
Chem Biol 8(2) (2004) 114-9.
[
12] M. Alexeeva, R. Carr, N.J. Turner, Directed evolution of enzymes: new biocatalysts for asymmetric
synthesis, Org Biomol Chem 1(23) (2003) 4133-4137.
1
6

[
13] H. Han, B. Zhu, X. Fu, S. You, B. Wang, Z. Li, W. Zhao, R. Peng, Q. Yao, Overexpression of
D-amino acid oxidase from Bradyrhizobium japonicum, enhances resistance to glyphosate in Arabidopsis
thaliana, Plant Cell Rep 34(12) (2015) 2043-51.
[
14] R. Li, J. Sun, Y. Fu, K. Du, M. Cai, P. Ji, W. Feng, Immobilization of genetically-modified D-amino
acid oxidase and catalase on carbon nanotubes to improve the catalytic efficiency, Catalysts 6(5) (2016)
6.
6
[
15] D. Koszelewski, K. Tauber, K. Faber, W. Kroutil, ω-Transaminases for the synthesis of non-racemic
α-chiral primary amines, Trends Biotechnol 28(6) (2010)
[
16] A. Caligiuri, P. D'Arrigo, E. Rosini, D. Tessaro, G. Molla, S. Servi, L. Pollegioni, Enzymatic
conversion of unnatural amino acids by Yeast D-amino acid oxidase, Adv Synth Catal 348(15) (2006)
183-2190.
2
[
17] P. Benz, R. Wohlgemuth, Amino acid oxidase-catalysed resolution and pictet-spengler reaction
towards chiral and rigid unnatural amino acids, J Chem Technol Biotechnol 82(12) (2007) 1082-1086.
[
18] R.L. Hanson, M.D. Schwinden, A. Banerjee, D.B. Brzozowski, B.-C. Chen, B.P. Patel, C.G.
McNamee, G.A. Kodersha, D.R. Kronenthal, R.N. Patel, L.J. Szarka, Enzymatic synthesis of
l-6-hydroxynorleucine, Bioorg Med Chem 7(10) (1999) 2247-2252.
[
19] R.N. Patel, Enzymatic synthesis of chiral intermediates for Omapatrilat, an antihypertensive drug,
Biomol Eng 17(6) (2001) 167-182.
[
20] W. Zhu, Y. Li, H. Jia, P. Wei, H. Zhou, M. Jiang, Expression, purification and characterization of a
thermostable leucine dehydrogenase from the halophilic thermophile Laceyella sacchari, Biotechnol Lett
8(5) (2016) 855-861.
3
[
21] M.B. Ansorge, M.R. Kula, Production of recombinant L-leucine dehydrogenase from Bacillus
cereus in pilot scale using the runaway replication system E. coli [piet98], Biotechnol Bioeng 68(5) (2000)
57-562.
5
[
22]R. Wichmann, D. Vasic-Racki, Cofactor regeneration at the lab scale, Technology transfer in
biotechnology, Springer 2005, pp. 225-260.
[
23] E.B. Ordu, E. Yelboğa, F. Secundo, R.B. Sessions, N.G. Karagüler, The effect of methionine to
cysteine substitution on the stability of formate dehydrogenase from Candida methylica, J Mol Catal B
Enzym 82 (2012) 109-114.
[
24] W. Jiang, B. Fang, Construction of a tunable multi-enzyme-coordinate expression system for
biosynthesis of chiral drug intermediates, Sci Rep 6 (2016) 30462.
[
25] R. Fernández-Lafuente, V. Rodrı́ guez, C. Mateo, G. Fernández-Lorente, P. Arminsen, P. Sabuquillo,
1
7

J.M. Guisán, Stabilization of enzymes (D-amino acid oxidase) against hydrogen peroxide via
immobilization and post-immobilization techniques, J Mol Catal B Enzym 7(1) (1999) 173-179.
[
26] M.E. Matlashov, V.V. Belousov, G. Enikolopov, How much H
2
O is produced by recombinant
2
D-amino acid oxidase in mammalian cells?, Antioxid Redox Signal 20(7) (2014) 1039-44.
[
27] M. Xu, Z. Rao, W. Dou, J. Yang, J. Jin, Z. Xu, Site-directed mutagenesis and feedback-resistant
N-acetyl-L-glutamate kinase (NAGK) increase Corynebacterium crenatum L-arginine production, Amino
Acids 43(1) (2012) 255-266.
[
28] G. Molla, L. Piubelli, F. Volontè, M.S. Pilone, Unnatural amino acids. In: P. Loredano(ed),
Enzymatic detection of D-amino acids, Humana Press, New York,2012, pp 273-289.
[
29] S. Berríos-Rivera, Metabolic Engineering of Escherichia coli: Increase of NADH availability by
+
overexpressing an NAD -dependent formate dehydrogenase, Metab Eng 4(3) (2002) 217-229.
[
30] T. Philibert, Z. Rao, T. Yang, J. Zhou, G. Huang, K. Irene, N. Samuel, Heterologous expression and
characterization of a new heme-catalase in Bacillus subtilis 168, J Ind Microbiol Biotechnol (2016) 1-12.
[
31] J.R. Denton, R.K. Dermenjian, B. Mao, Development of an enantiomeric separation of D & L valine
as their corresponding isoindole adducts by RP-HPLC for utilization of the L-valine toward
pharmaceutically relevant materials, Anal. Methods 6(16) (2014) 6565-6570.
[
32] H. Li, D. Zhu, B.A. Hyatt, F.M. Malik, E.R. Biehl, L. Hua, Cloning, protein sequence clarification,
and substrate specificity of a leucine dehydrogenase from Bacillus sphaericus ATCC4525, Appl Biochem
Biotechnol 158(2) (2009) 343-51.
[
33] L.J. Chien, J.M. Wu, I. Kuan, C.K. Lee, Coexpression of Vitreoscilla hemoglobin reduces the toxic
effect of expression of D-amino acid oxidase in E. coli, Biotechnol Prog 20(5) (2004) 1359-1365.
[
34] L.-J. Chien, C.-K. Lee, Synergistic effect of co-expressing D-amino acid oxidase with T7 lysozyme
on self-disruption of Escherichia coli cell, Biochem Eng J 28(1) (2006) 17-22.
[
35] V.I. Tishkov, S.V. Khoronenkova, D-amino acid oxidase: structure, catalytic mechanism, and
practical application, Biochemistry, Moscow, 70(1) (2005) pp 40-54.
[
36] L.-L. Lin, H.R. Chien, W.-C. Wang, T.-S. Hwang, H.-M. Fu, W.-H. Hsu, Expression of Trigonopsis
variabilis D-amino acid oxidase gene in Escherichia coli and characterization of its inactive mutants,
Enzyme Microb Technol 27(7) (2000) 482-491.
[
37] M. Gabler, M. Hensel, L. Fischer, Detection and substrate selectivity of new microbial D-amino acid
oxidases, Enzyme and Microbial Technology 27(8) (2000) 605-611.
[
38] K. Kataoka, K. Tanizawa, Alteration of substrate specificity of leucine dehydrogenase by
1
8

site-directed mutagenesis, J Mol Catal B Enzym 23(2-6) (2003) 299-309.
[
39] I. de la Mata, F. Ramón, V. Obregón, M.P. Castillón, C. Acebal, Effect of hydrogen peroxide on
D-amino acid oxidase from Rhodotorula gracilis, Enzyme Microb Technol 27(3) (2000) 234-239.
[
40] W. Märkle, S. Lütz, Electroenzymatic strategies for deracemization, stereoinversion and asymmetric
synthesis of amino acids, Electrochim. Acta 53(7) (2008) 3175-3180.
[
(
41] B.D. Sanwal, M.W. Zink, L-leucine dehydrogenase of Bacillus cereus, Arch Biochem Biophys 94(3)
1961) 430-435.
[
42] T. Schräder, J.R. Andreesen, Properties and chemical modification of D-amino acid oxidase from
Trigonopsis variabilis, Arch Microbiol 165(1) (1996) 41-47.
[
43] I. Dib, B. Nidetzky, The stabilizing effects of immobilization in D-amino acid oxidase from
Trigonopsis variabilis, BMC Biotechnol 8 (2008) 72-72.
[
44] H. Schütte, J. Flossdorf, H. Sahm, M.R. Kula, Purification and properties of formaldehyde
dehydrogenase and formate dehydrogenase from Candida boidinii, Eur J Biochem 62(1) (1976) 151-160.
[
45] S.-J. Wang, C.-Y. Yu, C.-K. Lee, M.-K. Chern, I.-C. Kuan, Subunit fusion of two yeast D-amino acid
oxidases enhances their thermostability and resistance to H , Biotechnol Lett 30(8) (2008) 1415-1422.
2
O
2
[
46] H. Slusarczyk, S. Felber, M.R. Kula, M. Pohl, Stabilization of NAD-dependent formate
dehydrogenase from Candida boidinii by site-directed mutagenesis of cysteine residues, Eur J Biochem
67(5) (2000) 1280-1289.
2
[
47] L. Pollegioni, A. Falbo, M.S. Pilone, Specificity and kinetics of Rhodotorula gracillis D-amino acid
oxidase, Biochim Biophys Acta 1120(1) (1992) 11-16.
[
48] A. Galkin, L. Kulakova, T. Yoshimura, K. Soda, N. Esaki, Synthesis of optically active amino acids
from alpha-keto acids with Escherichia coli cells expressing heterologous genes, Applied and
environmental microbiology 63(12) (1997) 4651-4656.
[
49] S. Nagata, S. Bakthavatsalam, A. Galkin, H. Asada, S. Sakai, N. Esaki, K. Soda, T. Ohshima, S.
Nagasaki, H. Misono, Gene cloning, purification, and characterization of thermostable and halophilic
leucine dehydrogenase from a halophilic thermophile, Bacillus licheniformis TSN9, Applied microbiology
and biotechnology 44(3-4) (1995) 432-438.
[
50] A. Galkin, L. Kulakova, T. Yoshimura, K. Soda, N. Esaki, Synthesis of optically active amino acids
from alpha-keto acids with Escherichia coli cells expressing heterologous genes, Appl Environ Microbiol
3(12) (1997) 4651-4656.
6
[
51] B.B. De, N.R. Thomas, Optimisation of the retroracemisation procedure for α-amino acids using
1
9

(
(
S)-2-[(N-alkylprolyl)amino]benzophenones, recyclable chiral auxiliaries, Tetrahedron Asymmetry 8(16)
1997) 2687-2691.
[
52] M.C. Daga, M. Taddei, G. Varchi, Rapid microwave-assisted deprotection of N-Cbz and N-Bn
derivatives, Tetrahedron Lett. 42(31) (2001) 5191-5194.
[
53] H.K. Chenault, J. Dahmer, G.M. Whitesides, Kinetic resolution of unnatural and rarely occurring
amino acids: Enantioselective hydrolysis of N-acyl amino acids catalyzed by acylase I, ChemInform
11(47) (1989) 6354-6364.
1
[
54] J.E. Baldwin, R.L. Dyer, S.C. Ng, A.J. Pratt, M.A. Russell, Application of E. Coli aspartate
transaminase to amino acid synthesis, Tetrahedron Lett. 28(32) (1987) 3745-3746.
2
0

Fig. 1 Optimization of enzymatic conversion conditions for enantiomerically pure L-norvaline production a
Effects of pH on the enzyme-mediated desymmetrization. The following pH buffers were employed for the
optimal pH experiments: 0.1 M acetate-sodium acetate (pH 4.5–5.5); sodium phosphate (Na
2
HPO
4
-NaH
2
PO )
4
(pH 5.5–8.5); Na
2
B
4
7
O -NaOH (pH 8.5–10.5). b Effects of temperature on the enzymatic desymmetrization. All
assays were performed in triplicate with three independent measurements. Results are presented as the mean
with standard deviation error bars for each condition.
2
1

Fig. 2 Optimization of dosage of DAAO, LeuDH and FDH a Effects of different DAAO to LeuDH ratios on the
−1
conversion rate of L-norvaline. The conversion was carried out with various dosages of DAAO (1.2 U·mL , 1.8
U·mL , 2.4 U·mL , 3 U·mL , 3.6 U·mL ), 1.2 U·mL LeuDH, 1.2 U·mL−1 FDH, 30 U·mL−1 catalase and 5
g/L DL-norvaline. b-f Effects of different ratios of LeuDH to FDH activity on L-norvaline production. The
conversion was carried out with 2.4 U·mL−1 DAAO, 1.2 U·mL−1 LeuDH, 30 U·mL−1 catalase and various
−1
−1
−1
−1
−1
−1
−1
−1
−1
−1
dosages of FDH (0 U·mL , 0.6 U·mL , 1.2 U·mL , 2.4 U·mL , 3.6 U·mL ). Ratios of DAAO, LeuDH and
FDH: b, 2:1:0; c, 2:1:0.5; d, 2:1:1; e, 2:1:2; f, 2:1:3.
2
2

Fig. 3 Time profiles of a fed-batch enzyme-mediated desymmetrization of DL-norvaline to L-norvaline
2
3

Table 1. Substrate spectrum of the multi-enzymatic desymmetrization system
Substrate
D-met
D-val
Product
L- met
L-val
Conversion rate (%)
56.8
99.7
92.8
30.4
16.4
32.0
77.4
D-arg
D-leu
L-arg
L-leu
D-his
L-his
D-phe
D-nor
L-phe
L-nor
Conversions were carried out with 25 mM of the different substrates in 0.1 M buffer (pH 7.5) at 30 °C for 2 h,
.4 U·mL−1 DAAO, 1.2 U·mL−1 LeuDH, 0.6 U·mL−1 FDH and 30 U·mL−1 catalase were contained in the
conversion mixture.
2
2
4

Table 2. Different strategies for production of L-norvaline
Methods
Substrate
Conditions
Conversion rate
or yield (%)
70
Titer
(g/L)
nm
Productivity
(g/L/h)
nm
e.e.
(%)
72
R
[
Chemical
synthesis
Racemic ɑ-valer-
(S)-2-[(Nalkylprolyl)amino]-
ic acid
benzophenones as chiral auxiliary
(1S,2S)-2-(1-phenylet S-ɑ-Aminophenethylamine
as
75
nm
nm
n.m.
[
hylamino)
butyronitrile
Valeric acid
chiral auxiliary and Cyanide,
Methanol were necessary
α-brominevaleryl chloride and Br
2
26.7
77
23.3
nm
nm
nm
2.9
<75
>90
99
[
[
were required.
Biosynthesis
N-acyl-α-amino-valer Aminoacylase from pig kidney
ic acid
was employed as catalyst
α-keto acid
Transaminases from different
sources were used as catalysts and
amino donor was necessary
99
5.8
[
[
D-Formyl-D, L
Immobilized L-N-carbamoylase
and N-succinyl -amino acid
>95
0.87
0.29
>99
[
-
norvaline and
N-Carbamoyl-D,
L-norvaline
racemase was acted as the catalyst
α-keto acid and
ammonium formate
LeuDH
and
FDH
were
95
92
44.4
12.3
0.9
3.70
4.10
100
n.m.
n.m.
[
[
[
simultaneous expressed in E. coli
and used as catalysts
α-keto acid
Couple the reactions catalyzed by
LeuDH and GluDH/galactose
mutarotase
Glucose
Fermentation of bacterium from
the Enterobacteriaces family
whose acetohydroxy acid
n.m.
0.012
synthases were inactivated
DAAO, LeuDH combined with
FDH were actd as catalysts,
catalase was used to remove
DL-norvaline
96.3
54.09
1.06
>99
T
s
unwanted H
2
O
2
n.m. not mentioned in the reference, e.e. enantiomeric excess, DAAO D-amino acid oxidase; LeuDH leucine
dehydrogenase, FDH formate dehydrogenase, GluDH glucose dehydrogenase
2
5