R-mandelic acid production with immobilized recombinant Escherichia coli cells in a recirculating packed bed reactor

Article abstract of DOI:10.1080/10242422.2016.1247822

A recirculating packed bed reactor (RPBR) was used for efficient production of R-mandelic acid (R-MA) by kinetic resolution of racemic R,S-mandelonitrile (R,S-MN) using the recombinant E. coli cells crosslinked with diatomite (DA)/glutaraldehyde (GA)/polyethyleneimine (PEI). The performance and productivity of RPBR were evaluated by several parameters, including cell load, substrate feeding rate, height diameter (H/D) ratio, reactor structures, and operation stability. The kinetic resolution process showed higher initial reaction rate (1.52 mM/min) and yield (100%) by recycling 100 mL of substrate solution (70 mM) through RPBR packed with 6.0 g immobilized cells at a substrate-feeding rate of 19 mL/min while the H/D ratio was 2.8. The immobilized cells were successfully applied into kinetic resolution of R,S-MN in the RPBR for 50 batches with an average productivity of 4.12 g/L/h for R-MA with >99% of enantiomeric excess.

Full text of DOI:10.1080/10242422.2016.1247822

Biocatalysis and Biotransformation
ISSN: 1024-2422 (Print) 1029-2446 (Online) Journal homepage: http://www.tandfonline.com/loi/ibab20
R-mandelic acid production with immobilized
recombinant Escherichia coli cells in a recirculating
packed bed reactor
Xin-Hong Zhang, Zhi-Qiang Liu, Ya-Ping Xue, Bo Yang, Ming Xu & Yu-Guo
Zheng
To cite this article: Xin-Hong Zhang, Zhi-Qiang Liu, Ya-Ping Xue, Bo Yang, Ming Xu & Yu-Guo
Zheng (2016) R-mandelic acid production with immobilized recombinant Escherichia coli cells
in a recirculating packed bed reactor, Biocatalysis and Biotransformation, 34:5, 205-211, DOI:
10.1080/10242422.2016.1247822
To link to this article: http://dx.doi.org/10.1080/10242422.2016.1247822
Published online: 18 Nov 2016.
Submit your article to this journal
Article views: 3
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ibab20
Download by: [FU Berlin]
Date: 10 December 2016, At: 04:29

Biocatalysis and Biotransformation, 2016; 34(5):205–211
RESEARCH ARTICLE
R-mandelic acid production with immobilized recombinant
Escherichia coli cells in a recirculating packed bed reactor
1
XIN-HONG ZHANG , ZHI-QIANG LIU , YA-PING XUE , BO YANG , MING XU
1
1
1
2
1
&
YU-GUO ZHENG
1
Key Laboratory of Bioorganic Synthesis of Zhejiang Province, College of Biotechnology and Bioengineering,
2
Zhejiang University of Technology, Hangzhou, PR China and Zhejiang Laiyi Biotechnology Co., Ltd.,
Shengzhou, Zhejiang, PR China
Abstract
A recirculating packed bed reactor (RPBR) was used for efficient production of R-mandelic acid (R-MA) by kinetic
resolution of racemic R,S-mandelonitrile (R,S-MN) using the recombinant E. coli cells crosslinked with diatomite (DA)/
glutaraldehyde (GA)/polyethyleneimine (PEI). The performance and productivity of RPBR were evaluated by several
parameters, including cell load, substrate feeding rate, height diameter (H/D) ratio, reactor structures, and operation
stability. The kinetic resolution process showed higher initial reaction rate (1.52 mM/min) and yield (100%) by recycling
1
1
00 mL of substrate solution (70 mM) through RPBR packed with 6.0 g immobilized cells at a substrate-feeding rate of
9 mL/min while the H/D ratio was 2.8. The immobilized cells were successfully applied into kinetic resolution of R,S-MN
in the RPBR for 50 batches with an average productivity of 4.12 g/L/h for R-MA with 499% of enantiomeric excess.
Keywords: R-mandelic acid, immobilized cells, nitrilase, packed bed reactor, kinetic, optimization
Introduction
volumetric productivity, and low output of waste.
Several enzymatic approaches have been developed
for the production of R-MA, using lipases (Yadav and
Sivakumar 2004), esterases (Zhang et al. 2011),
glyoxalases (Patterson et al. 1981), mandelate
dehydrogenases (Vasic-Racki et al. 1989), and nitri-
lases (Xue et al. 2013a) as biocatalysts. Nitrilase-
mediated reaction offers significant advantages over
other routes owing to its excellent enantioselectivity,
high productivity and low cost. This method can
provide theoretically a 100% yield of the product
since the unreacted substrate can be spontaneously
into R,S-MN (Liu et al. 2014).
Recently, to improve productivity of R-MA in the
bioprocess, an immobilized cells hydrolysis process
in a recirculating packed-bed reactor (RPBR) has
been reported (Xue et al. 2013a). The RPBR using
immobilized cells reaction is superior to a stirred
tank reactor (STR), because enzyme activity loss and
liquid shearing can be reduced by the absence of
R-mandelic acid (R-MA) is a key precursor for the
synthesis of semisynthetic penicillin, cephalosporin,
anti-obesity drugs, and antitumor agents (Ju et al.
2010). Due to the important applications of enantio-
merically pure R-MA, most attention has focused on
establishing enantioselective routes for its production
including chemical and biocatalytic methods (Xue
et al. 2013a). Chemical synthesis of R-MA including
dichlorocarbene, benzaldehyde, acetophenone, and
glyoxylate routes typically requires high temperature,
strong acidic or basic conditions, and usually pro-
duces unwanted byproducts (Chen et al. 2002).
Though the chemical resolution methods could
efficiently obtain R-MA and avoid byproducts, it
still needs expensive resolving agents and results in
low yield, which is unfavorable for commercial
production. Enzymes are highly valuable catalysts
and allow the manufacture of chiral chemicals on an
industrial scale with high enantioselectivity, yield,
Correspondence: Yu-Guo Zheng. Key Laboratory of Bioorganic Synthesis of Zhejiang Province, College of Biotechnology and Bioengineering, Zhejiang
University of Technology, Hangzhou, PR China. Tel: +86 57188320630. Fax: +86 57188320630. E-mail: zhengyg@zjut.edu.cn
(Received 3 November 2015; revised 10 July 2016; accepted 4 August 2016)
ISSN 1024-2422 print/ISSN 1029-2446 online ß 2016 Informa UK Limited, trading as Taylor & Francis Group
DOI: 10.1080/10242422.2016.1247822

206 X. H. Zhang et al.
collisions between immobilized cells and impellers
Zhang et al. 2010; Zhao et al. 2012; Xue et al.
013a). However, the productivity reported in the
Effect of substrate concentration on the initial
reaction rate
(
2
The effects of different R,S-MN concentrations
RPBR reaction using Ca-alginate/PEI/GA immobi-
lized cells still cannot meet the requirement of
industrial production of R-MA (Xue et al. 2013a).
In this work, a RPBR was selected for the
production of R-MA using the recombinant E. coli
cells cross-linked with DA/GA/PEI, which had been
reported in our lab (Zhang et al. 2016) to improve
the reaction batch and the successive production.
The performance and productivity of the immobi-
lized cells in RPBR were evaluated. The operational
performances between RPBR and STR were further
compared, and the parallel and series operation of
RPBRs were also investigated in this study.
(5–300 mM) on the initial reaction rate and enan-
tioselectivity were investigated. The experiments
ꢀ
were carried out at 35 C in the water adjusted to
pH 8.0 by 6 M NaOH with 0.2 g of the immobilized
cells in 10 mL reaction system. The initial reaction
rate and ee were evaluated during first 10 min,
respectively.
Semi-continuous production of R-MA in RPBR
Immobilized cells were packed into a glass column
with a water jacket for temperature control. The
ꢀ
column temperature was maintained at 35 C using
circulator, and the reservoir was at the same tem-
perature using a water bath (Figure 1A). Substrate
solution was recirculated through packed bed reactor
Materials and methods
(
PBR) from the bottom by operating a peristaltic
Materials
pump to contact these immobilized cells. The
amount of immobilized cells (2–12 g), enzyme col-
umns with different ratios of H/D ratios (0.7–9),
and feeding rate (4.0–22.5 mL/min) were further
optimized.
R,S-MN, R-MA, and S-MA were purchased from
J&K Chemical Co., Ltd. (Shanghai, PR China). All
other reagents and chemicals were commercially
available and of analytic grade.
Activity assay of immobilized cells
Time course of R-MA production by kinetic resolution
of R,S-MN in RPBR and STR
The activities of immobilized cells were assayed in a
standard reaction mixture (10 mL) containing 0.2 g
immobilized cells suspended in water adjusted to pH
Immobilized cells (6.0 g) were packed into a glass
column under the optimal conditions. Reactions
ꢀ
8.0 by 6 M NaOH and 64 mM R,S-MN in 50 mL
were carried out at 35 C for 5 h by recycling 100 mL
ꢀ
flasks. The reactions were carried out at 35 C,
180 rpm for 10 min, and then were stopped by
of substrate solution (70 mM) through RPBR at a
feeding rate of 19 mL/min or performed in a STR at
180 rpm. The time course of R-MA production by
kinetic resolution of R,S-MN for one batch recycle
was investigated.
adding 200 mL 2 M HCl. The amounts of R-MA
were determined by HPLC as described below. One
unit of immobilized cells activity was defined as the
amount of immobilized cells that produce 1 mmol of
R-MA per minute under the conditions described
above.
Parallel and series operation of RPBRs
Immobilized cells (12 g) were packed into two glass
columns under the optimal conditions. Reactions
Preparation of immobilized cells
ꢀ
were performed at 35 C for 0.5–5 h by recycling
The immobilized cells DA/GA/PEI were prepared
according to the method as described previously
200 mL of substrate solution (70 mM) through
RPBR at a feeding rate of 19 mL/min. The time
courses of R-MA production by two types of reactor
structures (Figures 1B and 1C) were studied.
(Zhang et al. 2016). To perform the reaction in a
recirculating packed bed reactor (RPBR), the immo-
bilized cells were air-dried to a constant weight at
ꢀ
2
5 C, and the shapes including strip and granular,
Operation stability of RPBR
and sizes of the dry immobilized cells including 510
mesh, 10–18 mesh, and418 mesh were studied. The
experiments were carried out at 35 C in water (pH
Long-term stability of the immobilized cells was
evaluated for 125 h under the optimal conditions.
For each batch of reaction, 100 mL of 70 mM
R,S-MN in water (pH 8.0) was recycled through
RPBR with a substrate feeding rate of 19 mL/min
ꢀ
8.0) using 64 mM R,S-MN as the substrate with
0.2 g of the immobilized cells in 10 mL reaction
system, and the relative activity was evaluated.

R-mandelic acid production with immobilized recombinant Escherichia coli cells 207
Figure 1. Experiment set-up of RPBR used for semi-continuous production of R-MA by kinetic resolution of R,S-MN.
for 2.5 h. The substrate solution was subsequently
withdrawn by a peristaltic pump in opposite direc-
tion. The next reaction was started by supplying a
fresh substrate solution into the reservoir. The
average productivity of the RPBR for R-MA was
with 0.01 M NH H PO :CH OH ¼13:7 (v/v, 1 mL/
4
2
4
3
min) as mobile phase, and absorbance at 228 nm
was measured. The optical purity of R-MA was
determined by analysis of the enantiomers on a
CHIRALEL-OD-H column (250 ꢁ 4.6 mm) (Daicel
Chemical Industries, Shinzaike, Japan) at a flow rate
of 0.8 mL/min with a mobile phase containing
hexane, 2-propanol, and trifluoroacetic acid
(90:10:0.1, v/v) at wavelength of 228 nm. The
retention time for R-MA was 3.2 min. The ee of
R-MA was calculated by the equation: ee¼ [(RꢂS)/
calculated according to the formula (S ꢁ M ꢁ
w
P
Y%)/T, where
S was the concentration of
R,S-MN, M was the molecular weight of R-MA,
w
P
Y% was the total yield of 50 batches reaction, and
T was the total reaction time.
(
R + S)] * 100%, where R and S represent the
concentration of R-MA and S-MA, respect-
ively (Liu et al. 2014; Zhang et al. 2016).
Analytical methods
The amount of R-MA was determined by a LC-
2
0AD prominence liquid chromatography instru-
ment (Shimadzu, Kyoto, Japan) using a C18 column
Elite Analytical
Results
(
5 mm ꢁ 250 nm
ꢁ4.6 nm,
Preparation of immobilized cells
Instruments Co., Ltd., Dalian, PR China) and
equipped with a SPD-2A prominence UV/Vis
To perform the reaction in the RPBR, the shapes and
sizes of the dry immobilized cells were studied.
ꢀ
detector. Each sample (20 mL) was eluted at 30 C

208 X. H. Zhang et al.
Table 1 and Figure 2 indicated that the granular
shape of dry immobilized cells with 10–18 mesh was
the best and the relative activity was 75%. However,
the dry immobilized cells with 418 mesh could not
be used to prevent clogging the reactor though the
relative activity was 100%.
Effect of immobilized cells amount on initial reaction
rate
The effects of immobilized cell amounts (2.0–12.0 g)
on the initial reaction rate and the yield of R-MA in
RPBR and STR were investigated in a 100 mL
reaction mixture containing 70 mM R,S-MN. Figure
4
showed that the initial reaction rate in STR was
obviously higher than that in RPBR in the range of
.0–12.0 g, and the biocatalyst in the STR was
The effect of substrate concentration on the initial
reaction rate
2
unstable due to the collisions between the immobi-
lized cells and the impellers (Zhang et al. 2010). As
the biocatalyst amount was increased from 2.0 to
Higher substrate concentration can accelerate the
reaction rate from the perspective of chemical
equilibrium, while it also may cause inhibition of
enzyme activity. The initial reaction rates at lower
substrate concentrations ranging from 5 to 50 mM
were found to be directly proportional to the
substrate concentration, while at higher substrate
concentrations in the range of 50–70 mM, the initial
reaction rate tended to a maximum value (Figure 3),
and the highest initial reaction rate of 97 mmol/min/g
was obtained at the substrate concentration of
4
.0 g, there was a linear increase of the initial rate.
Mass transfer limitations appeared at higher loading
beyond 4.0 g. In all experiments, the ee of the
product remained more than 99%. Considering the
1
00
1
00
80
8
6
0
0
7
0 mM. The reaction obeyed with the Michaelis–
Menten kinetics indicating that substrate inhibition
does not occur when the substrate concentration was
lower than 70 mM. The values of Km and Vmax were
calculated to be 22 mM and 118 mmol/min/g,
respectively. When the substrate concentration
increased beyond 140 mM, the initial reaction rate
decreased because of substrate inhibition. The sub-
strate concentration did not affect the enantioselec-
tivity of immobilized cells, and ee values of all the
products remained above 99%.
6
4
2
0
0
0
0
40
20
0
Initial rate V0
ee
0
50
100
150
200
250
300
Mandelonitrile concentration [S] (mM)
Figure 3. The effect of substrate concentration on the initial
ꢀ
Table 1. The relative activity comparison of immo-
bilized cells dried before and after.
reaction rate. Reactions were carried out at 35 C and 180 rpm for
1
0 min in 10 mL of water (pH 8.0) containing various concentra-
Immobilized cells
Relative activity
tion of R,S-MN with 0.2 g of the immobilized cells in 50 mL flasks.
Wet immobilized cells
Dry immobilized cells (granular)
Dry immobilized cells (strip)
100 ꢃ 2.4
76.2 ꢃ 0.5
76.1 ꢃ 0.2
3
3
2
.5
.0
.5
1
1
8
6
4
20
00
0
1
00
Initial reaction rate in a RPBR
Initial reaction rate in a STR
Yield in a RPBR
1
8
6
4
2
0
00
0
80
ee
Relative activity
1
00
ee
2.0
1.5
1.0
6
4
2
0
0
0
0
0
80
60
40
20
0
0
0
0
0
.5
.0
20
0
0
1
2
3
4
5
6
7
8
9
10 11 12 13
Immobilized cells amount (g)
0
Figure 4. Effect of immobilized cells amount loaded on the initial
ꢀ
reaction rate and yield. Reactions were carried out at 35 C for
<
10
10-18
>18
10 min – 2 h by recycling 100 mL of substrate solution (70 mM)
through RPBR packed with 2.0–12.0 g immobilized cells with a
feeding rate of 4 mL/min in a 300 ꢁ 20 mm reactor and in STR
with 180 rpm.
Mess size (mesh)
Figure 2. The selection of the sizes of the dry immobilized cells.

R-mandelic acid production with immobilized recombinant Escherichia coli cells 209
cost, 6.0 g of immobilized cells were used for 100 mL
of reaction solution for the subsequent study on the
factors affecting the bioreactor.
Time course of R-MA production by kinetic resolution
of R,S-MN in RPBR and STR
Immobilized cells (6.0 g) were added into a RPBR
and STR under the optimal conditions, respectively.
The time course of biotransformation showed that
after 0.5 h conversion, the yield (50%) in the RPBR
was obviously lower than that (69%) in the STR
Effect of different H/D ratio of reactor on initial
reaction rate
The effects of different H/D ratios on the initial
reaction rate and yield of R-MA in a RPBR were
investigated in 100 mL reaction mixture containing
(Figure 7). The max amount of R-MA with ee499%
reached 70 mM in 2.5 h and 2 h in a RPBR and STR,
respectively. The results indicated that there was no
significant difference in the final yield in RPBR
and STR.
70 mM R,S-MN. The results in Figure 5 showed
that the initial reaction rate increased as the H/D
ratio varied from 0.7 to 9.0, and the ee of the product
remained above 99% in all the tests. The substrate
diffusion was considered as a key factor in the RPBR
and the H/D ratio greatly influenced the substrate
diffusion (Zhang et al. 2010; Xue et al. 2013b). In
order to reduce the energy of the peristaltic pump
and to avoid the possibility of clogging the enzyme
column, the H/D ratio of 2.8 was chosen for further
studies.
Effect of parallel and series operation of RPBRs on
yield
Immobilized cells (total 12 g) were packed into two
glass columns under the optimal conditions.
100
1.6
1
1
.4
.2
8
6
4
2
0
0
0
0
0
Effect of feeding rate on initial reaction rate
The effects of feeding rates on the initial reaction rate
in a RPBR were investigated in 100 mL reaction
mixture containing 70 mM R,S-MN. The results
showed that the initial rate raised with a linear
trend when the flow rate was increased from 4 to
1.0
0
0
.8
.6
0.4
19 mL/min (Figure 6) because of a possible external
0
0
.2
.0
Initial reaction rate
diffusion resistance (Zhang et al. 2010). The external
diffusion resistance dispersed when the flow rate was
ee
5
10
15
20
25
419 mL/min. In all experiments with different flow
Feeding rate (mL/min)
rates, ee of the product remained 499%.
Figure 6. Effect of flow rate on the initial rate and ee of R,S-MN
hydrolysis in the RPBR. Reactions were carried out at 35 C for
ꢀ
1
0 min – 2 h by recycling 100 mL of substrate solution (70 mM)
through RPBR with a flow rate of 4.0–22.5 mL/min.
1
1
1
0
0
0
0
.50
.25
.00
.75
.50
.25
.00
120
100
80
60
40
20
0
Initial reaction rate
Yield ee
8
7
0
0
100
80
1
8
6
4
2
0
00
0
100
60
50
8
6
4
2
0
0
0
0
0
0
6
0
4
3
0
0
0
40
R-MA in a RPBR
R-MA in a STR
Yield in a RPBR
Yield in a STR
ee in a RPBR
0
20
2
0
0
1
0
0
ee in a STR
0
2
4
6
8
10
H/D
0
1
2
3
4
5
Time (h)
Figure 5. Effect of H/D in the RPBR on the initial reaction rate
ꢀ
and yield. Reactions were carried out at 35 C for 10 min – 2 h by
Figure 7. Time course of R,S-MN hydrolysis in the reactors.
Reactions were carried out at 35 C for 0.5–5.0 h by recycling
ꢀ
recycling 100 mL of substrate solution (70 mM) through RPBR
with a feeding rate of 4 mL/min, while the H/D ratio was varied
from 0.7 to 9.0.
100 mL of substrate solution (70 mM) through RPBR with a flow
rate of 19 mL/min (solid) or performed in a STR (hollow).

210 X. H. Zhang et al.
80
70
60
50
40
30
20
10
0
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
1
8
6
4
00
0
100
80
60
40
20
0
1
8
6
00
0
1
00
8
0
0
0
0
0
6
4
40
0
R-MA in a series RPBR
R-MA in a parallel RPBR
Yield in a series RPBR
Yield in a parallel RPBR
ee in a series RPBR
2
0
0
20
0
20
0
R-MA
ee
Relative activity
ee in a parallel RPBR
0
1
2
3
4
5
0
10
20
30
40
50
Time (h)
Batches
Figure 8. Effect of reactor structures on yield. Reactions were
ꢀ
Figure 9. Semi-continuous production of R-MA in a packed bed
reactor. For each batch of the reaction in RPBR, 100 mL of
substrate emulsion containing 70 mM R,S-MN was recycled with
performed at 35 C for 0.5–5 h by recycling 200 mL of substrate
solution (70 mM) through two kinds of RPBRs at a feeding rate of
1
9 mL/min, series RPBR (solid), and parallel RPBR (hollow).
a flow rate of 19 mL/min through the column reactor packed with
ꢀ
6
.0 g immobilized cells at 35 C for 2.5 h.
ꢀ
Reactions were performed at 35 C for 0.5–5 h by
recycling 200 mL of substrate solution (70 mM)
through RPBR at a feeding rate of 19 mL/min. The
trends of the R-MA production in the parallel and
series RPBRs were the same, although their rates
differed considerably. After 0.5 h conversion, the
yield (54%) in the parallel RPBR was obviously
higher than that (41%) in the series RPBR, and the
max amount of R-MA reached 70 mM and the yield
was 100% in 2.5 h and 3 h in the series RPBR and
the parallel RPBR, respectively, and ee of R-MA was
catalysts (Kirk et al. 2002; Straathof et al. 2002;
O’Reilly and Turner 2003; Liu et al. 2014; Zhang
et al. 2015). Recently, immobilized nitrilase has
attracted much attention because enzyme immobil-
ization could make recovery and reuse of the enzyme
easy, facilitate the separation of production, and
reduce the operation cost (Zhang et al. 2014; Zhang
et al. 2015; Zhang et al. 2016).
Though many immobilization methods had been
reported (Vejvoda et al. 2006; Xue et al. 2013a;
Zhang et al. 2014; Zhang et al. 2015; Zhang et al.
2016), the stability of the immobilized cells in STR is
still defective because of the collisions by the forces
of impellers resulting in the low productivity (Zhang
et al. 2010). The RPBR is preferred owing to
high-conversion efficiency and long reaction time,
especially, facilitating the industrial solid-liquid sep-
aration (Manolov et al. 2005; Zhang et al. 2010;
Zhao et al. 2012; Xue et al. 2013b), So far, a few
studies have been reported about the production of
R-MA by kinetic resolution of R,S-MN in RPBR
(Xue et al. 2013a). In this study, a RPBR packed
with the direct cross-linking immobilized cells using
DA/GA/PEI was used for efficient production of
R-MA by kinetic resolution of racemic R,S-MN.
To obtain high-reaction rate and reduce the
production cost, the performance and productivity
of RPBR were determined by several critical param-
eters (Zhang et al. 2010; Xue et al. 2013b). The
kinetic resolution process showed higher initial
reaction rate (1.52 mM/min) and yield (100%) by
recycling 100 mL of substrate solution (70 mM)
through RPBR packed with a substrate feeding rate
of 19 mL/min while the H/D ratio was 2.8 (Zhang
et al. 2010; Xue et al. 2013b).
499% in these two kinds of reactors. The results also
indicated that structures of reactors had no signifi-
cant effect on the yield (Figure 8).
Operation stability of RPBR
The repeated use and operation stability of the
immobilized biocatalyst are the key factors for
industrial application. The semi-continuously oper-
ation of immobilized cells in RPBR was evaluated for
125 h under the optimal conditions. According to the
results in Figure 7, the first run of the reaction
afforded R-MA in 100% yield and 499% ee at the
operation time of 2.5 h and the productivity was
28 mM/h, which was 3.04-fold higher than previous
report (9.2 mM/h) (Xue et al. 2013a). Figure 9
showed that there was a gradual drop in the activity
of immobilized cells after 125 h conversion, but the
relative activity still could remain above 78% after 50
batches in RPBR and the average productivity of the
RPBR for R-MA was 4.12 g/L/h.
Discussion
Nitrilase, as one of the important industrial enzymes
for the conversion of R,S-MN to R-MA, has more
significant improvements than traditional chemical
The catalytic efficiencies of immobilized cells in
RPBR and STR are different. The yield in the STR

R-mandelic acid production with immobilized recombinant Escherichia coli cells 211
was obviously higher than that in the RPBR after
.5 h conversion. However, the immobilized cells in
References
0
Chen HB, Lin YB, Liu ZP. 2002. Study of synthesis of mandelic
acid. Chin J Synth Chem 10:185.
a STR was unstable, which was accordance with
previous report about immobilized lactonase in STR
and fibrous bed reactor (FBR) showing that the half-
time of immobilized F. proliferatum lactonases
Ju X, Yu HL, Pan J, Wei DZ, Xu JH. 2010. Bioproduction of
chiral mandelate by enantioselective deacylation of alpha-
acetoxyphenylacetic acid using whole cells of newly
isolated Pseudomonas sp ECU1011. Appl Microbiol
Biotechnol 86:83–91.
Kirk O, Borchert TV, Fuglsang CC. 2002. Industrial enzyme
applications. Curr Opin Biotechnol 13:345–351.
Liu ZQ, Zhang XH, Xue YP, Xu M, Zheng YG. 2014.
Improvement of Alcaligenes faecalis nitrilase by gene site
saturation mutagenesis and its application in stereospecific
biosynthesis of (R)-(-)-mandelic acid. J Agric Food Chem
(
IFPL-A) in FBR was at least 6-fold higher than
that in STR (Zhang et al. 2010), and the productivity
in RPBR was 3.04-fold higher than the previous
report (9.2 mM/h) (Xue et al. 2013a). The further
scale-up study of R-MA production in series and
parallel RPBRs showed that two kinds of reactor
structures had no significant effect on the yield
and ee.
Reusability of immobilized cells was one of the
important characteristics to reduce fermentation
workload and cost (Zhang et al. 2014; Zhang et al.
6
2:4685–4694.
Manolov T, Kristina H, Benoit G. 2005. Continuous acetonitrile
degradation in packed-bed bioreactor. Appl Microbiol
a
Biotechnol 66:567–574.
OReilly C, Turner PD. 2003. The nitrilase family of CN
hydrolysing enzymes - a comparative study. J Appl Microbiol
9
5:1161–1174.
Patterson MAK, Szajewski RP, Whitesides GM. 1981.
Enzyme conversion of a-keto aldehydes to optically active
a-hydroxy acids using glyoxalase I and II. J Org Chem 46:
4682–4685.
Straathof AJJ, Panke S, Schmid A. 2002. The production of fine
chemicals by biotransformations. Curr Opin Biotechnol
2015; Zhang et al. 2016). The relative activity still
could remain above 78% after 50 batches with high-
optical purity (ee499%) in RPBP, which was higher
than 61.8% after 30 batches for the immobilized cells
and 8.79% after 10 reuse cycles for the free cells in a
shake flask (Zhang et al. 2016). Moreover, the
successive production of R-MA could reach 3.39
M, which was higher than any previous reports (Xue
et al. 2013a; Zhang et al. 2014; Zhang et al. 2015;
Zhang et al. 2016).
1
3:548–556.
Vasic-Racki D, Jonas M, Wandrey C, Hummei W, Kula M.
989. Continuous (R)-mandelic acid production in an
enzyme membrane reactor. Appl Microbiol Biotechnol
1:215–222.
1
3
Vejvoda V, Kaplan O, Kubac D, Kren V, Martinkova L. 2006.
Immobilization of fungal nitrilase and bacterial amidase -
two enzymes working in accord. Biocatal Biotransform
In summary, immobilized cells catalyzed enantio-
selective resolution of R,S-MN for the efficient
production of R-MA in RPBR was first successfully
performed in this study. The immobilized cells
exhibited a strict enantiospecificity for the R,S-MN
during the hydrolysis at different substrate concen-
tration. The RPBR could lead to a significant
increase in the immobilized cells stability with the
relative activity of above 78% after 50 batches
reaction, and it is easy to be applied in the large-
scale production with high efficiency and low cost.
The results obtained in this study could contribute to
the commercial production of R-MA with high
efficiency from R,S-MN using immobilized cells
in RPBR.
2
4:414–418.
Xue YP, Jiang T, Liu X, Zheng YG. 2013b. Efficient production
of S-(+)-2-chlorophenylglycine by immobilized penicillin G
acylase in a recirculating packed bed reactor. Biochem Eng J
7
Xue YP, Xu M, Chen HS, Liu ZQ, Wang YJ, Zheng YG. 2013a.
A novel integrated bioprocess for efficient production of (R)-
(-)-mandelic acid with immobilized Alcaligenes faecalis
ZJUTB10. Org Process Res Dev 17:213–220.
Yadav GD, Sivakumar P. 2004. Enzyme-catalysed optical reso-
lution of mandelic acid via RS(-/+)-methyl mandelate in non-
aqueous media. Biochem Eng J 19:101–107.
Zhang GY, Tao J, Zheng LY, Cao SG. 2011. Molecular basis for
stereospecific hydrolysis of ethyl mandelate by thermophilic
esterase. J Chem Eng Chin Uni 27:841–844.
Zhang X, Xu JH, Liu DH, Pan J, Chen B. 2010. Construction and
operation of a fibrous bed reactor with immobilized lactonase
for efficient production of (R)-alpha-hydroxy-gamma-butyro-
lactone. Biochem Eng J 50:47–53.
4:88–94.
Zhang XH, Liu ZQ, Xue YP, Xu M, Zheng YG. 2016. Nitrilase-
catalyzed conversion of (R,S)-mandelonitrile by immobilized
recombinant E. coli cells harboring nitrilase. Biotechnol Appl
Biochem 63:479–489.
Declaration of interest
Zhang ZJ, Pan J, Li CX, Yu HL, Zheng GW, Ju X, Xu JH. 2014.
Efficient production of (R)-(-)-mandelic acid using glutaralde-
hyde cross-linked Escherichia coli cells expressing Alcaligenes sp.
nitrilase. Bioprocess Biosyst Eng 37:1241–1248.
Zhang ZJ, Yu HL, Imanaka T, Xu JH. 2015. Efficient production
of (R)-(-)-mandelic acid by isopropanol-permeabilized recom-
binant E. coli cells expressing Alcaligenes sp nitrilase. Biochem
Eng J 95:71–77.
The authors report no conflicts of interest. The
authors alone are responsible for the content and
writing of this article. This work was financially
supported by National Natural Science Foundation
of China (No. 21476210), National High
Technology Research and Development Program
of China (863 Program) (No. 2011AA02A210), and
Natural Science Foundation of Zhejiang Province
Zhao T, Kim BH, Hong SI, Yoon SW, Kim CT, Kim Y, Kim IH.
2
012. Lipase-catalyzed production of pinolenic acid concen-
trate from pine nut oil using a recirculating packed bed reactor.
J Food Sci 77:C267–C271.
(
Nos. R3110155, Z4090612, and Y4110409).