Enantioselective synthesis of (S)-phenylephrine by recombinant Escherichia coli cells expressing the short-chain dehydrogenase/reductase gene from Serratia quinivorans BCRC 14811

Article abstract of DOI:10.1016/j.procbio.2013.07.006

Background An amino alcohol dehydrogenase gene (RE-AADH) from Rhodococcus erythropolis BCRC 10909 has been used for the conversion of 1-(3-hydroxyphenyl)- 2-(methylamino) ethanone (HPMAE) to (S)-phenylephrine [(S)-PE]. However RE-AADH uses NADPH as cofactor, and only limited production of (S)-PE from HPMAE is achieved. Methods A short-chain dehydrogenase/reductase gene (SQ-SDR) from Serratia quinivorans BCRC 14811 was expressed in Escherichia coli BL21 (DE3) for the conversion of HPMAE to (S)-PE. Results The SQ-SDR enzyme was capable of converting HPMAE to (S)-PE in the presence of NADH and NADPH, with specific activities of 26.5 ± 2.3 U/mg protein and 0.24 ± 0.01 U/mg protein, respectively, at 30 C and at a pH of 7.0. The E. coli BL21 (DE3), expressing NADH-preferring SQ-SDR, converted HPMAE to (S)-PE with more than 99% enantiomeric excess, a conversion yield of 86.6% and a productivity of 20.2 mmol/l h, which was much higher than our previous report using E. coli NovaBlue expressing NADPH-dependent RE-AADH as the biocatalyst. Conclusion The SQ-SDR enzyme with its high catalytic activity and strong preference for NADH as a cofactor provided a significant advantage in bioreduction.

Full text of DOI:10.1016/j.procbio.2013.07.006

Process Biochemistry 48 (2013) 1509–1515
Contents lists available at ScienceDirect
Process Biochemistry
journal homepage: www.elsevier.com/locate/procbio
Enantioselective synthesis of (S)-phenylephrine by recombinant
Escherichia coli cells expressing the short-chain
dehydrogenase/reductase gene from Serratia quinivorans BCRC 14811
Guan-Jhih Peng, Yen-Ching Cho, Tze-Kai Fu, Ming-Te Yang∗∗, Wen-Hwei Hsu∗
Institute of Molecular Biology, National Chung Hsing University, Taichung 402, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 May 2013
Received in revised form 8 July 2013
Accepted 12 July 2013
Available online 20 July 2013
Background: An amino alcohol dehydrogenase gene (RE AADH) from Rhodococcus erythropolis BCRC 10909
has been used for the conversion of 1-(3-hydroxyphenyl)-2-(methylamino) ethanone (HPMAE) to (S)-
phenylephrine [(S)-PE]. However RE AADH uses NADPH as cofactor, and only limited production of (S)-PE
from HPMAE is achieved.
Methods: A short-chain dehydrogenase/reductase gene (SQ SDR) from Serratia quinivorans BCRC 14811
was expressed in Escherichia coli BL21 (DE3) for the conversion of HPMAE to (S)-PE.
Results: The SQ SDR enzyme was capable of converting HPMAE to (S)-PE in the presence of NADH and
NADPH, with specific activities of 26.5 2.3 U/mg protein and 0.24 0.01 U/mg protein, respectively, at
30 ◦C and at a pH of 7.0. The E. coli BL21 (DE3), expressing NADH-preferring SQ SDR, converted HPMAE
to (S)-PE with more than 99% enantiomeric excess, a conversion yield of 86.6% and a productivity of
20.2 mmol/l h, which was much higher than our previous report using E. coli NovaBlue expressing NADPH-
dependent RE AADH as the biocatalyst.
Keywords:
Serratia quinivorans
Short-chain dehydrogenase/reductase
Enantioselective synthesis
Phenylephrine
Biocatalyst
Conclusion: The SQ SDR enzyme with its high catalytic activity and strong preference for NADH as a
cofactor provided a significant advantage in bioreduction.
© 2013 Elsevier Ltd. All rights reserved.
1. Introduction
PE contains one chiral carbon atom in the C␣ of the side chain,
and the R enantiomer of PE exhibits more potency than the S enan-
Phenylephrine (PE) is a sympathomimetic, which is widely used
as a nasal decongestant in common cold and flu medicines [1].
Other sympathomimetics, including phenylpropanolamine, pseu-
doephedrine (PDE) and ephedrine, are also used in cough and
cold medicines [2]. However, the over-the-counter (OTC) and pre-
scription drug products containing phenylpropanolamine were no
longer recommended for use by the Food and Drug Administra-
tion in the year 2000 because these drugs were associated with an
increased risk of hemorrhagic stroke [3]. PDE and PE are now com-
monly used worldwide as systemic nasal decongestants. Because
PDE is easily abused for the production of methamphetamine [4,5],
the Combat Methamphetamine Epidemic Act in the USA banned
OTC medicines containing PDE in 2005 to avoid the illicit conver-
sion of PDE into methamphetamine [6]. This restriction imposed
on PDE has led to a need for PE in common cold medicines.
tiomer for the activation of ␣1-adrenergic receptors [7]. PE can be
produced by various methods involving ring opening of 3-benzyl-
oxystyrene epoxide [8], Curtius rearrangement of a ␤-hydroxy acid
azide [9] and reduction of a mandelamide [10]. PE produced from
these methods is a racemic mixture; therefore, asymmetric hydro-
genation methods have been developed for the production of (R)-PE
[11–14]. However, these methods require high pressure, high tem-
perature and several environmentally unfriendly organic solvents.
We have cloned an amino alcohol dehydrogenase gene
(RE AADH) from Rhodococcus erythropolis BCRC 10909 into
Escherichia coli, and the cells expressing RE AADH were used for the
conversion of 1-(3-hydroxyphenyl)-2-(methylamino) ethanone
(HPMAE) to (S)-PE [15], which can be further converted to (R)-PE
by the Walden inversion reaction. However, RE AADH used NADPH
as cofactor, and only limited production of PE from HPMAE was
achieved. In this study, we have described the cloning of a short-
chain dehydrogenase/reductase gene from Serratia quinivorans
BCRC 14811 and expression in E. coli BL21 (DE3). The biochemical
properties of recombinant SQ SDR were characterized. We found
that the SQ SDR enzyme could utilize both NADH and NADPH as
a cofactor for its catalytic activity. The yield and productivity of
the conversion of HPMAE to PE by recombinant E. coli expressing
SQ SDR were also evaluated.
∗
Corresponding author. Tel.: +886 4 2285 1885; fax: +886 4 2287 4879.
Corresponding author. Tel.: +886 4 2284 0485x227; fax: +886 4 2287 4879.
E-mail addresses: mtyang@dragon.nchu.edu.tw (M.-T. Yang),
∗∗
whhsu@dragon.nchu.edu.tw (W.-H. Hsu).
1359-5113/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.procbio.2013.07.006

1510
G.-J. Peng et al. / Process Biochemistry 48 (2013) 1509–1515
Table 1
2. Materials and methods
Conversion of HPMAE to PE by microorganismsa.
2.1. Materials
Microorganism
Concentration (mM)
All solvents were LC grade and purchased from Merck (Darmstadt, Germany) and
Sigma–Aldrich (St. Louis, MO, USA). Restriction enzymes were obtained from New
England Biolabs (Ipswich, MA, USA). DNA polymerase, ExTaq, and T4 DNA ligase
were purchased from TaKaRa (Tokyo, Japan). Culture media were obtained from
Becton, Dickinson and Company (Sparks, MD, USA).
Residual HPMAE PE production
R. erythropolis BCRC 10909
R. erythropolis BCRC 13743
3.38
3.65
3.75
4.73
3.63
3.53
3.08
0.15
0
0
0
0
Streptomyces clavuligerus BCRC 11518
Streptomyces griseslus BCRC 13677
Streptomyces lividans M2
S. lividans TK24 BCRC 51705
Streptomyces spp. NCHU-1151
2.2. Bacterial strains, plasmid and culture conditions
0
0
E. coli BL21 (DE3) was cultivated in Luria-Bertani (LB) medium at 37 ◦C. All bac-
terial strains were obtained from the Bioresource Collection and Research Center
(Hsinchu, Taiwan) and cultivated in LB medium at either 25 ◦C or 30 ◦C. The plasmids
pET30a (Novagen, Inc., Madison, WI, USA) and pQE-30 (QIAGEN, Hilden, Germany)
were used for the expression of the gene in E. coli.
Streptomyces thermovvulgaris BCRC 12488 2.63
0.05
0
1.25
1.84
0
Agrobacterium tumefaciens LB4404
Serratia marcescens BCRC 10948
S. quinivorans BCRC 14811
Pseudomonas putida GB-1
Pseudomonas fluorescens BCRC 10304
P. fluorescens BCRC 11028
4.72
2.27
2.12
4.95
4.80
4.79
0
0
2.3. Analysis of HPMAE and PE
a
Reaction was performed in 100 mM sodium phosphate buffer (pH 7.0) contain-
ing 10% (wet weight) cell, 2% glucose and 5 mM HPMAE at 30 ◦C for 24 h.
HPMAE and PE were analyzed by high performance liquid chromatography
(HPLC) with a reverse-phase INERTSIL 10 ODS column (3.2 mm × 250 mm) (VER-
COPAK, Taipei, Taiwan). The mobile phase consisted of methanol and 0.5% sodium
acetate (pH 5.5) at a ratio of 2:98, with a flow rate of 0.8 ml/min. The absorbance of
the products was detected at a wavelength of 215 nm.
The chirality of PE was analyzed by HPLC with the chiral column CYCLOBOND
I 2000 AC (4.6 mm × 250 mm) (Aztec, NJ, USA). The mobile phase consisted of
methanol and 0.5% sodium acetate (pH 5.5) at a ratio of 5:95, with a flow rate of
0.4 ml/min. The absorbance of the products was detected at a wavelength of 215 nm.
The retention times were 13.2 min for (R)-PE and 14.9 min for (S)-PE.
100 mM NaCl and 10 mM imidazole, pH 8.0), the adherent His-tagged SQ SDR pro-
tein was eluted with a buffer containing 20 mM Tris-HCl, 100 mM NaCl and 100 mM
imidazole (pH 8.0).
2.8. Enzyme activity assay of purified SQ SDR
2.4. Screening of bacterial strains for the conversion of HPMAE to PE
We found that HPMAE, PE and NAD(P)H have various absorbances at 340 nm the
wavelength typically used for monitoring NAD(P)H concentration [16]. Therefore,
in this study, the production of PE from HPMAE was determined by measuring the
decrease in absorbance at 370 nm caused by oxidation of NAD(P)H at 30 ◦C.
A steady-state kinetics study of the conversion of HPMAE to PE by purified
SQ SDR enzyme was performed. The reaction mixture (1 ml) containing the appro-
priately diluted enzyme in 100 mM sodium phosphate buffer (pH 7.0), 0.1–5 mM
NAD(P)H and 0.1–10 mM HPMAE was incubated at 30 ◦C. The Km and kcat values for
HPMAE and Km value for NAD(P)H were calculated by fitting the rates, as a function
of substrate concentration, to the Michaelis–Menten equation.
Bacterial colonies selected from agar plates were inoculated into 5 ml of cul-
ture medium and incubated at the desired temperature with shaking at 150 rpm.
One milliliter of overnight culture was inoculated into 100 ml of the same medium
and incubated until the OD600 reached 2. Cells were collected by centrifugation at
9000 × g for 10 min and then washed with 10% glycerol. A cell pellet of 0.1 g (wet
weight) was resuspended in 1 ml of 100 mM sodium phosphate buffer (pH 7.0) con-
taining 5 mM HPMAE and 2% glucose. After incubation at 30 ◦C for 24 h, the cells
were removed by centrifugation. The supernatant was filtered through a 0.45-␮m
membrane and analyzed by HPLC.
The effect of pH on the specific activity of SQ SDR was determined by measuring
enzyme activity over a pH range from 4 to 9 using different buffers containing 1 mM
HPMAE and 0.4 mM NADH at 30 ◦C. To determine the effect of temperature on the
purified recombinant SQ SDR activity, various assays were performed in sodium
phosphate buffer (pH 7.0) at 25–60 ◦C. One unit (U) of enzyme activity was defined
as the amount of enzyme required for production of 1 ␮mol PE from HPMAE per
minute under assay conditions.
2.5. Cloning of genes from S. quinivorans BCRC 14811
The aminoketone asymmetric reductase (akr), short-chain dehydroge-
nase/reductase (sdr) and alcohol dehydrogenase (adh) genes, which encode gene
products with considerable identity to R. erythropolis BCRC 10909 RE AADH pro-
tein, were amplified from the genomic DNA of S. quinivorans BCRC 14811 with PCR
using gene-specific primers. The PCR products were digested with BamHI and HindIII
and cloned into pQE-30 or digested with NdeI and XhoI and cloned to pET30a. The
recombinant plasmids were introduced into E. coli NovaBlue or E. coli BL21 (DE3).
2.9. Toxicity of HPMAE and PE to E. coli BL21 (DE3)
E. coli BL21 (DE3) (pET30a) was cultivated in LB at 37 ◦C to an OD600 of 0.8. Cells
were collected by centrifugation at 9000 × g for 10 min. To maintain the osmotic
cell equilibrium [17,18] and stabilize the cell proteins [19], cell pellets were washed
with 10% glycerol and then resuspended in reaction mixture. The reaction mixture
(15 ml), containing 1% (wet weight) cells, 100 mM sodium phosphate buffer (pH 7.0),
2% glucose and 70 mM HPMAE or 70 mM PE, was incubated at 30 ◦C with shaking at
150 rpm. The cells were collected after incubating for 3, 6, 9, 12 and 15 h and plated
on LB agar at an appropriate dilution. The agar plate was incubated at 37 ◦C for 24 h,
and individual colonies were counted as viable cells.
2.6. Cell enzyme activities of the recombinant E. coli BL21 (DE3) harboring the
akr, sdr or adh genes
E. coli BL21 (DE3) and E. coli NovaBlue harboring recombinant vectors contain-
ing either the akr, sdr or adh genes were cultivated in LB at 37 ◦C to an OD600 of 0.8.
Isopropyl-␤-d-thiogalactopyranoside (IPTG) was added to the culture broth to a final
concentration of 1 mM, and cultivation was continued at 28 ◦C for an additional 6 h
for gene expression. Cells were collected by centrifugation at 9000 × g for 10 min.
Cell pellets were washed with 10% glycerol and then resuspended in reaction mix-
ture. The reaction mixture (15 ml), containing 1% (wet weight) cells, 100 mM sodium
phosphate buffer (pH 7), 10 mM HPMAE and 2% glucose, was incubated at 30 ◦C with
shaking. After the reaction, the cells were removed by centrifugation at 4 ◦C, and the
supernatant was subjected to HPLC analysis. One unit (U) of cell enzyme activity
was defined as the amount of recombinant cells required for production of 1 ␮mol
PE from HPMAE per minute under assay conditions.
3. Results and discussion
3.1. Screening of bacterial strains for the conversion of HPMAE to
PE
2.7. Expression and purification of recombinant SQ SDR proteins
In a previous report, we found that R. erythropolis BCRC 10909
converted just detectable levels of HPMAE to (S)-PE and that E. coli
NovaBlue expressing RE AADH from R. erythropolis BCRC 10909
converted HPMAE to PE with limited productivity [15]. Therefore,
13 strains of bacteria were screened for their ability to convert
HPMAE to PE. As shown in Table 1, S. quinivorans BCRC 14811 was
the most promising bacterium in this test, with a yield of 36.8%.
E. coli BL21 (DE3) harboring pET30a-sdr-39729 was cultivated in LB at 37 ◦
C
until an OD600 of 0.8 was reached. The expression of the sdr gene was induced by
the addition of IPTG to a final concentration of 1 mM. The culture was incubated at
28 ◦C with shaking for 6 h. Cells were collected by centrifugation at 9000 × g and 4 ◦C,
resuspended in sonication buffer (20 mM Tris–HCl and 100 mM NaCl, pH 8.0) and
disrupted by sonication. The cell extract was clarified by centrifugation at 12,000 × g
for 30 min at 4 ◦C, and the resulting supernatant was loaded onto a Ni-NTA col-
umn (QIAGEN, Hilden, Germany). After washing with wash buffer (20 mM Tris–HCl,

G.-J. Peng et al. / Process Biochemistry 48 (2013) 1509–1515
1511
Fig. 2. Effect of pH on SQ SDR activity. The reaction was performed by incubating the
enzyme with 1 mM HPMAE in 100 mM sodium citrate buffer (ꢀ, pH 4–6), sodium
phosphate buffer (ꢁ, pH 6–8) or Tris–HCl buffer (᭹, pH 7–9) containing 0.4 mM
NADH at 30 ◦C.
Fig. 1. SDS-PAGE analysis of SQ SDR purified from E. coli BL21 (DE3) (pET30a-sdr-
39729) by Ni-NTA column. Lane 1, protein size markers; lane 2, crude cell extract;
lane 3, supernatant of cell lysate; lane 4, pellet of cell lysate; lane 5, purified SQ SDR.
SQ SDR was performed. In the absence of either cofactor NADH or
NADPH, no enzyme activity was observed for SQ SDR. When either
NADH or NADPH was present in the reaction mixture, PE could
be produced from HPMAE, indicating that SQ SDR required either
NADH or NADPH as a cofactor (data not shown). SQ SDR was most
active at pH 7.0 in 100 mM sodium phosphate buffer containing
0.4 mM NADH as cofactor. About 50% of the maximal activity was
retained at pH 6 and pH 9. The activity was dramatically reduced at
pHs below 6 (Fig. 2). The optimal temperature for SQ SDR was 45 ◦C,
and approximately 60% of maximal activity was observed at 30 ◦C
(Fig. 3). The specific activities of SQ SDR in the presence of NADH
and NADPH were 26.5 2.3 and 0.24 0.01 U/mg protein, respec-
tively, at 30 ◦C and pH 7.0. The specific activity (26.5 2.3 U/mg
protein) of SQ SDR was 1.4 × 108-fold higher than that of RE AADH
(0.19 ␮U/mg protein) [15]. The kinetic parameters for HPMAE in
the presence of these two cofactors were analyzed. The appar-
ent kcat and Km of SQ SDR for HPMAE were 74.7 3.3 min−1 and
1.76 0.08 mM, respectively, in the presence of 0.4 mM NADH,
whereas only 0.85 0.04 min−1 and 2.41 0.09 mM were obtained
in the presence of NADPH, respectively. The apparent Km value for
NAD(P)H was also determined. The apparent Km values for NADH
and NADPH were 0.32 0.07 mM and 2.28 0.03 mM, respectively.
These results revealed that NADH was superior to NADPH in the
conversion of HPMAE to PE.
3.2. Cloning of the gene responsible for HPMAE reduction from S.
quinivorans BCRC 14811
The genomic sequence of S. quinivorans (formerly Serratia
proteamaculans subsp. quinovora) has been reported (GenBank:
CP000826.1). To identify the enzyme responsible for the conver-
sion of HPMAE to PE, the genes encoding proteins with amino acid
sequences similar to R. erythropolis BCRC 10909 RE AADH were
searched using BLAST. Ten AKRs, 3 SDRs and 7 ADHs proteins
exhibiting 35–49% sequence identities with RE AADH were found.
The genes encoding these proteins were amplified with PCR, cloned
into pQE-30 or pET30a and expressed in the E. coli strains NovaBlue
or BL21 (DE3). The ability to convert HPMAE to PE by these trans-
formants was monitored in a reaction mixture (15 ml) containing
100 mM sodium phosphate buffer (pH 7.0), 10% cells (wet weight),
2% glucose and 10 mM HPMAE. After incubating at 30 ◦C for 24 h, the
supernatants obtained from the reaction mixtures were subjected
to HPLC analysis, and only E. coli BL21 (DE3) harboring pET30a-sdr-
39729 exhibited enzyme activity for the conversion of HPMAE to
PE (3.23 U/g wet cells). The chirality of the resulting PE was ana-
lyzed by HPLC using a CYCLOBOND I 2000 AC chiral column, and a
99% enantiomeric excess of (S)-PE was obtained in this conversion
(data not shown). The sdr gene (designated SQ SDR with GenBank
accession number ABV39729) responsible for this conversion con-
tained a 783-bp ORF encoding the SQ SDR protein with a calculated
molecular mass of 26.61 kDa. The amino acid sequence of SQ SDR
showed 37.36% identity to the amino acid sequence of RE AADH
[15] from R. erythropolis BCRC 10909 (data not shown).
The deduced amino acid sequence of SQ SDR was compared
with the sequence of NADPH-specific RE AADH (with 37.36%
3.3. Biochemical characterization of SQ SDR protein
The SQ SDR protein was purified from the crude cell extract of
E. coli BL21 (DE3) cells harboring the SQ SDR gene by Ni-NTA affin-
ity chromatography and then subjected to SDS-PAGE analysis. The
molecular mass of recombinant SQ SDR was approximately 26 kDa,
which compared well with the calculated mass of the affinity-
tag translated product from the corresponding gene. About 75% of
SQ SDR expressed in E. coli BL21 (DE3) was in a soluble form, which
was desirable for biocatalysis. As shown in the lane 5 of Fig. 1, the
purity of eluted SQ SDR was more than 99%.
Fig. 3. Effect of temperature on SQ SDR activity. The reaction was performed by
incubating the enzyme with 1 mM HPMAE in 100 mM sodium phosphate buffer (pH
7.0) containing 0.4 mM NADH at different temperatures.
To investigate the coenzyme dependence and specificity of
SQ SDR, the conversion of HPMAE to PE by purified recombinant

1512
G.-J. Peng et al. / Process Biochemistry 48 (2013) 1509–1515
Fig. 4. Alignment of the amino acid sequences of SQ SDR with RE AADH and 3D structure available SDRs from E. coli (EC HSDH), Streptomyces hydrogenans (SH HSDH),
Cochliobolus lunatus (CL HSDH) and E. coli (EC SDH). The conserved Gly-motif for cofactor binding is indicated with a box. The catalytic tetrad is indicated with an arrow. The
cofactor-determined residue is marked with a star.
identity to SQ SDR) and also with the sequences of SDRs for which
3D structures were available, such as NAD(H)-specific 7-alpha-
hydroxysteroid dehydrogenase from E. coli (EC HSDH, Protein Data
Bank: 1FMC, with 23.08% identity to SQ SDR) [20], NAD(H)-specific
3␣,20␤-hydroxysteroid dehydrogenase from Streptomyces hydro-
genans (SH HSDH, Protein Data Bank: 2HSD, with 21.51% identity to
SQ SDR) [21], NADP(H)-specific serine dehydrogenase from E. coli
(EC SDH, Protein Data Bank: 3ASV, with 13.31% identity to SQ SDR)
[22] and NADP(H)-specific 17␤-hydroxysteroid dehydrogenase
from Cochliobolus lunatus (CL HSDH, Protein Data Bank: 3QWF,
with 11.11% identity to SQ SDR) [23]. A conserved TGXXXGXG
motif specific for NAD(P)H binding [15,24] was found in the N-
terminal region of the protein (residues 18–25, all numbering refers
to SQ SDR). The catalytic tetrad Asn120-Ser148-Tyr162-Lys166,
was also found in the protein. This tetrad, of which Tyr-162 is the
most conserved residue within the SDR family, fits the conserved
active site motif for the catalytic reaction (Fig. 4) [15,24–27].
SDR-family enzymes catalyze NAD(P)(H)-dependent oxida-
tion/reduction reactions [25,28–30]. SQ SDR belonged to the
classical SDR family, which possesses a conserved Gly-motif,
TGXXXGXG [15,24], for NAD(H) and NADP(H) binding in the
coenzyme-binding fold (Rossmann fold). The results of our study
demonstrated that SQ SDR had dual coenzyme specificity, with a
very strong preference for NADH over NADPH as a cofactor. NADH
differs structurally from NADPH in the absence of a phosphate
group esterified to the 2ꢀ-hydroxyl group of its AMP moiety, which
suggests that residues other than TGXXXGXG are responsible for
determining coenzyme specificity for SDRs. It has been reported
that NAD(H)-preferring enzymes typically have a conserved aspar-
tate residue present at the end of the second ␤-strand of the ␤␣␤
motif located at the beginning of the Rossmann fold [31–36]. Amino
acid alignment of SQ SDR with other NADH-preferring SDRs clearly
showed that the corresponding position in SQ SDR is occupied by
Asp-43, indicating that SQ SDR is a NAD(H)-preferring enzyme, in
agreement with our experimental data.
3.4. Conversion of HPMAE to PE by E. coli BL21 (DE3)
(pET30a-sdr-39729) expressing SQ SDR
In a cell-free bioconversion system, NAD(P)H is a relatively
unstable molecule and prohibitively expensive. Therefore, E. coli
BL21 (DE3) (pET30a-sdr-39729) expressing SQ SDR was used as a
biocatalyst to evaluate the conditions for the conversion of HPMAE
to PE. The effect of pH was investigated. The SQ SDR activity of
recombinant E. coli cells increased with pH up to a maximum of pH
7.0 and declined thereafter (data not shown). The effect of the car-
bon source on conversion was also determined. In the absence of a
carbon source, a yield of only 1.2% was obtained in a 1-h reaction.
Sucrose and xylose were ineffective carbon sources for bioconver-
sion (Table 2). The addition of either glucose, fructose, maltose or
lactose to the reaction mixture resulted in a great increase in the
conversion yield of HPMAE. Of these, glucose was the most effective
carbon source for the conversion (90.7% yield). The conversion yield
of HPMAE decreased as the glucose concentration increased from
2% to 10%, and 2% glucose was determined to be the optimal concen-
tration for conversion (Table 3). It has been reported that 2% glucose
is better than higher glucose concentrations for E. coli growth and
enzyme production [37]. In most cells the concentration of NAD+
is significantly higher than NADH [38–41], indicating that efficient
recycling of NADH is required to maintain the efficient reductive
activity of NADH-preferring enzymes such as SQ SDR during any
bioconversion process. Our results indicated that NADH could be
supplied effectively by the catabolic pathways of the E. coli cells
using glucose, fructose, maltose and lactose as carbon sources.

G.-J. Peng et al. / Process Biochemistry 48 (2013) 1509–1515
1513
Table 2
Table 4
Effect of carbon source on the conversion of HPMAE to PE by E. coli BL21 (DE3)
Effects of HPMAE concentration on the PE production.a
(pET30a-sdr-39729) expressing SQ SDR.a
HPMAE (mM)
PE
Yield (%)
Productivity
Carbon source (2%)
Concentration (mM)
Residual HPMAE
concentration
(mM)
(mmol/l, h)
PE production
10
20
40
60
70
9.67 0.78
16.26 1.07
32.52 0.05
52.69 0.12
60.60 0.88
96.7
81.3
81.3
87.8
86.6
9.67
13.01
16.26
17.56
20.20
Glucose
Sucrose
Fructose
Xylose
Maltose
Lactose
None
0.74 0.06
7.83 0.04
1.61 0.17
7.59 0.03
0.89 0.16
0.16 0.05
6.69 0.23
9.07 0.82
0.23 0.03
4.56 0.33
0.26 0.02
7.31 0.33
7.17 0.60
0.12 0.01
a
Reaction was performed in the reaction mixture (15 ml) containing 100 mM
sodium phosphate buffer (pH 7.0), 1% (wet weight) cell mass, 2% glucose and differ-
ent concentrations of HPMAE at 30 ◦C.
a
Reaction was performed in the reaction mixture (15 ml) containing 100 mM
sodium phosphate buffer (pH 7.0), 1% (wet weight) cell mass, 2% carbon source and
10 mM HPMAE at 30 ◦C for 1 h.
Because conversion yield [42–44] and conversion rate [45–47]
are usually affected by substrate concentration, the effects of
HPMAE concentration on the PE production were examined (Fig. 5).
The results showed that the productivity of PE was increased as
substrate concentration increased, and the highest concentration
of PE (60.6 mM) and level of productivity (20.2 mmol PE/l h) could
be obtained from a reaction mixture containing 70 mM HPMAE
(Table 4). HPMAE concentrations higher than 70 mM were dif-
ficult to achieve due to the limited solubility of HPMAE (18 g
dissolves in 1 L of 100 mM sodium phosphate buffer, pH 7.0, at
30 ◦C). In the presence of 70 mM, a HPMAE conversion yield of
86.6% could be obtained in a 3-h reaction (Table 4). In the previ-
ous report, we used E. coli NovaBlue expressing NADPH-dependent
RE AADH from R. erythropolis BCRC 10909 as catalyst to pro-
duce PE from HPMAE [15]. The cell enzyme activity of E. coli
NovaBlue expressing RE AADH was only 0.42 U/g wet cells. In addi-
tion, a conversion yield of 78% and a productivity of 3.9 mmol
PE/l h were obtained from a reaction mixture containing 60 mM
HPMAE in a 12-h reaction [15]. In this study, E. coli BL21 (DE3)
Table 3
Effect of glucose concentration on the conversion of HPMAE to PE by E. coli BL21
(DE3) (pET30a-sdr-39729) expressing SQ SDR.a
Glucose concentration (%)
Concentration (mM)
Residual HPMAE
PE production
0
2
4
6
8
6.84 0.05
0.74 0.06
1.29 0.18
2.18 0.02
3.33 0.09
4.37 0.00
0.02 0.02
9.07 0.82
7.79 1.02
7.87 0.22
6.94 0.13
5.64 0.01
10
a
Reaction was performed in the reaction mixture (15 ml) containing 100 mM
sodium phosphate buffer (pH 7.0), 1% (wet weight) cell mass, different concentra-
tions of glucose and 10 mM HPMAE at 30 ◦C for 1 h.
Fig. 5. Effect of substrate concentration on the conversion of HPMAE to PE by E. coli BL21 (DE3) (pET30a-sdr-39729) expressing SQ SDR. The reaction was performed in
100 mM sodium phosphate buffer (pH 7.0) containing 1% cell mass, 2% glucose and various concentrations of HPMAE at 30 ◦C for 4 h. (A) 20 mM, (B) 40 mM, (C) 60 mM and
(D) 70 mM. Symbols: ꢁ, HPMAE consumption; ꢂ, (S)-PE production.

1514
G.-J. Peng et al. / Process Biochemistry 48 (2013) 1509–1515
Fig. 6. Repeated use of E. coli cells expressing SQ SDR for PE production. The reac-
tion mixture (15 ml), containing 100 mM sodium phosphate buffer (pH 7.0), 1%
(wet weight) E. coli BL21 (DE3) (pET30a-sdr-39729) expressing SQ SDR cells, 70 mM
HPMAE and 2% glucose, was incubated at 30 ◦C for 3 h in each cycle. Symbols: ꢁ, con-
version yield of HPMAE for PE; ꢂ, Log10 CFU/ml. Data are presented as mean SD
(n = 3).
Fig. 7. The toxicity of HPMAE and PE to E. coli BL21 (DE3) (pET30a) cells. The reac-
tion mixture (15 ml) containing 100 mM sodium phosphate buffer (pH 7.0), 1% (wet
weight) cells, 70 mM HPMAE or PE and 2% glucose was incubated at 30 ◦C. Symbols:
ꢂ, PE. ꢀ, HPMAE.
that PE caused only a slight reduction in the number of viable E. coli
cells, whereas high concentrations of HPMAE were toxic to E. coli.
Overall, these data indicated that HPMAE toxicity to E. coli cells was
an important limiting factor for the cell-recycling process.
expressing SQ SDR with a high HPMAE reduction activity (3.23 U/g
wet cells) was used as biocatalyst. This strain exhibited a cell
enzyme activity approximately 7.7-fold higher than that of E. coli
cells expressing RE AADH [15]. For the reaction mixture contain-
ing the highest substrate concentration (70 mM HPMAE), E. coli
BL21 (DE3) expressing NADH-preferring SQ SDR exhibited approx-
imately 370%-higher PE formation (60.6 mM), a 63.2%-higher yield
of HPMAE and a 367%-higher productivity than E. coli expressing
NADPH-dependent RE AADH [15] in a 3-h reaction.
Whole cell conversion of HPMAE to PE required a stoichiomet-
ric amount of the cofactor NAD(P)H, and thus practical applications
would need efficient recycling of NAD(P)H for the reduction activity
of SQ SDR. The concentration of NADH in cells is usually higher than
that of NADPH. Reiss et al. reported that the total amount of NAD+
and NADH is approximately 10-fold higher than the concentration
of NADP+ and NADPH in rat liver [48]. In exponentially growing
E. coli cell, the concentration of NAD(H) is roughly 3-fold higher
than that of NADP(H) [49]. Apparently, NADH is superior to NADPH
for its efficient recycling. Therefore, the specificity of the cofactor-
binding pocket of NADPH-dependent 2,5-diketo-d-gluconic acid
reductase has been modified to improve its ability to use NADH
as a cofactor for vitamin C production [50]. Obviously, recombi-
nant E. coli expressing SQ SDR with a 80-fold preference for NADH
over NADPH as a cofactor would be a valuable catalyst in industrial
production of PE from HPMAE.
4. Conclusion
The SQ SDR gene was cloned from S. quinivorans BCRC 14811 and
functionally expressed in E. coli BL21 (DE3). The cell enzyme activity
for the recombinant strain was 3.23 U/g wet cells. The recombi-
nant E. coli strain expressing NADH-preferring SQ SDR exhibited a
much higher conversion yield and productivity for the conversion
of HPMAE to (S)-PE than did our previous strain, E. coli NovaBlue
expressing NADPH-dependent RE AADH. We demonstrated that
the SDR enzyme with its high catalytic activity and strong pref-
erence for NADH as a cofactor provided a significant advantage in
bioreduction. Because (S)-PE must be further converted to (R)-PE
by a Walden inversion reaction for medical use, a novel biocatalyst
that is able to directly convert HPMAE to (R)-PE would be desir-
able for industrial production. An enzyme engineering approach to
develop such a biocatalyst is currently in progress.
Acknowledgements
This work was supported by the research grants NSC95-2313-
B-005-005 and NSC96-2628-B-005-002-MY3 from the National
Science Council, Taiwan.
3.5. Recycling of E. coli BL21 (DE3) (pET30a-sdr-39729) cells in
the conversion process
References
[1] Eccles R. Substitution of phenylephrine for pseudoephedrine as a nasal decon-
geststant. An illogical way to control methamphetamine abuse. British Journal
of Clinical Pharmacology 2007;63:10–4.
[2] PAGB. UK OTC Directory, Treatments for Common Ailments. London: Commu-
nications International Group; 2005/2006.
[3] Meadows M. FDA issues public health advisory on phenylpropanolamine in
drug products. FDA Consumer 2001;35:9.
[4] Tanne JH. Methamphetamine epidemic hits middle America. British Medical
Journal 2006;332:382.
E. coli BL21 (DE3) (pET30a-sdr-39729) expressing SQ SDR
showed high activity in the conversion of HPMAE to PE. However,
from an economical standpoint, the reusability of the recombinant
cells in the conversion process should be determined. The reac-
tions were performed with 70 mM HPMAE at 30 ◦C for 3 h for each
cycle. After conversion, the cells were collected, washed and then
used immediately in the next cycle. As shown in Fig. 6, the conver-
sion yield remained at 77% in the fourth run. After 13 cycles, the
conversion yield declined to 3%. The reduction of the conversion
yield occurred simultaneously with the loss of viable cells in the
reaction mixture (Fig. 6). To study the toxicity of HPMAE and PE to
E. coli BL21 (DE3), cells were incubated with 70 mM of HPMAE or PE
in 100 mM sodium phosphate buffer (pH 7.0) at 30 ◦C. Fig. 7 shows
[5] Roehr B. Half a million Americans use methamphetamine every week. British
Medical Journal 2005;331:476.
[6] Registration requirements for importers and manufacturers of prescrip-
tion drug products containing ephedrine, pseudoephedrine, or phenyl-
propanolamine. Final rule. Federal Register 2010;75:4973–82.
[7] Westfall TC, Westfall DP. Adrenergic agonists and antagonists. In: Brunton LL,
Lazo JS, Parker KL, editors. Goodman and Gilman’s the pharmacological basis
of therapeutics. 11th ed. New York: McGraw-Hill; 2006. p. 237–95.

G.-J. Peng et al. / Process Biochemistry 48 (2013) 1509–1515
1515
[8] Britten AZ.
A
new route to d,1-phenylephrine. Chemistry
&
Industry
[30] Moon HJ, Tiwari MK, Singh R, Kang YC, Lee JK. Molecular determinants
of the cofactor specificity of ribitol dehydrogenase, short-chain dehy-
1968;24:771–2.
a
[9] Bergmann ED, Sulzbacher M. A new synthesis of 1-(m- and p-hydroxyphenyl)-
2-methylaminoethanol (m- and p-sympathol). The Journal of Organic
Chemistry 1951;16:84–9.
[10] Russell PB, Childress SJ. New route to phenylephrine. Journal of Pharmaceutical
Sciences 1961;50:713.
drogenase/reductase. Applied and Environmental Microbiology 2012;78:
3079–86.
[31] Scrutton NS, Berry A, Perham RN. Redesign of the coenzyme specificity of a
dehydrogenase by protein engineering. Nature 1990;343:38–43.
[32] Bocanegra JA, Scrutton NS, Perham RN. Creation of an NADP-dependent
pyruvate dehydrogenase multienzyme complex by protein engineering. Bio-
chemistry 1993;32:2737–40.
[11] Klinger FD, Wolter L, Dietrich W. Method for preparing of l-phenylephrine
hydrochloride. U.S. Patent. US 6,187,956 B1; 01.02.13.
[12] Kumar P, Naidu V, Gupta P. Application of hydrolytic kinetic resolution (HKR)
in the synthesis of bioactive compounds. Tetrahedron 2007;63:2745–85.
[13] Pandey RK, Upadhyay PK, Kumar P. Enantioselective synthesis of (R)-
phenylephrine hydrochloride. Tetrahedron Letters 2003;44:6245–6.
[14] Takeda H, Tachinami T, Aburatani M, Takahashi H, Morimoto M, Achiwa K. Prac-
tical asymmetric synthesis of (R)-(−)-phenylephrine hydrochloride catalyzed
by (2R,4R)-MCCPM-rhodium complex. Tetrahedron Letters 1989;30:367–70.
[15] Lin WD, Hsu WH, Chen CY, Chen HC. Enantioselective synthesis of (S)-
phenylephrine by whole cells of recombinant Escherichia coli expressing the
amino alcohol dehydrogenase gene from Rhodococcus erythropolis BCRC 10909.
Process Biochemistry 2010;45:1529–36.
[16] Steen IH, Lien T, Madsen MS, Birkeland NK. Identification of cofactor discrimina-
tion sites in NAD-isocitrate dehydrogenase from Pyrococcus furiosus. Archives
of Microbiology 2002;178:297–300.
[17] Henin J, Tajkhorshid E, Schulten K, Chipot C. Diffusion of glycerol through
Escherichia coli aquaglyceroporin GlpF. Biophysical Journal 2008;94:832–9.
[18] Froger A, Rolland JP, Bron P, Lagree V, Le Caherec F, Deschamps S, et al.
Functional characterization of a microbial aquaglyceroporin. Microbiology
2001;147:1129–35.
[19] Na GC. Interaction of calf skin collagen with glycerol: linked function analysis.
Biochemistry 1986;25:967–73.
[20] Tanaka N, Nonaka T, Tanabe T, Yoshimoto T, Tsuru D, Mitsui Y. Crystal structures
of the binary and ternary complexes of 7 alpha-hydroxysteroid dehydrogenase
from Escherichia coli. Biochemistry 1996;35:7715–30.
[21] Ghosh D, Wawrzak Z, Weeks CM, Duax WL, Erman M. The refined three-
dimensional structure of 3 alpha, 20 beta-hydroxysteroid dehydrogenase and
possible roles of the residues conserved in short-chain dehydrogenases. Struc-
ture 1994;2:629–40.
[22] Yamazawa R, Nakajima Y, Mushiake K, Yoshimoto T, Ito K. Crystal struc-
ture of serine dehydrogenase from Escherichia coli: important role of the
C-terminal region for closed-complex formation. Journal of Biochemistry
2011;149:701–12.
[33] Chen Z, Lee WR, Chang SH. Role of aspartic acid 38 in the cofactor speci-
ficity of Drosophila alcohol dehydrogenase. European Journal of Biochemistry
1991;202:263–7.
[34] Kallberg Y, Oppermann U, Jornvall H, Persson B. Short-chain dehydroge-
nases/reductases (SDR): the 2002 update. Chemico-Biological Interactions
2002;269:4409–17.
[35] Rossmann MG, Liljas A, Branden CI, Banaszak LJ. Evolutionary and structural
relationships among dehydrogenases. In: Boyer PD, editor. The enzymes, vol.
11. New York: Academic Press; 1975. p. 61–102.
[36] Belyaeva OV, Johnson MP, Kedishvili NY. Kinetic analysis of human enzyme
RDH10 defines the characteristics of a physiologically relevant retinol dehy-
drogenase. Journal of Biological Chemistry 2008;283:20299–308.
[37] Kazan D, C¸ amurdan A, Hortac¸ su A. The effect of glucose concentration on the
growth rate and some intracellular components of a recombinant E. coli culture.
Process Biochemistry 1995;30:269–73.
[38] Matschinsky FM. Quantitative histochemistry of nicotinamide adenine
nucleotides in retina of monkey and rabbit. Journal of Neurochemistry
1968;15:643–57.
[39] Giblin FJ, Reddy VN. Pyridine nucleotides in ocular tissues as determined by
the cycling assay. Experimental Eye Research 1980;31:601–9.
[40] Bublitz C, Lawler CA. The levels of nicotinamide nucleotides in liver micro-
somes and their possible significance to the function of hexose phosphate
dehydrogenase. Biochemical Journal 1987;245:263–7.
[41] Adams Jr JD, Klaidman LK, Chang ML, Yang J. Brain oxidative stress—analytical
chemistry and thermodynamics of glutathione and NADPH. Current Topics in
Medicinal Chemistry 2001;1:473–82.
[42] Chartrain M, White R, Goegelman R, Gbewonyo K, Greasham R. Bioconver-
sion of avermectin into 27-OH avermectin. Journal of Industrial Microbiology
1990;6:279–84.
[43] Rottava I, Cortina PF, Martello E, Cansian RL, Toniazzo G, Antunes OA, et al.
Optimization of alpha-Terpineol production by the biotransformation of R-
(+)-limonene and (−)-beta-pinene. Applied Biochemistry and Biotechnology
2011;164:514–23.
[44] Zhu Y, Li J, Liu L, Du G, Chen J. Production of alpha-ketoisocaproate via free-
whole-cell biotransformation by Rhodococcus opacus DSM 43250 with l-leucine
as the substrate. Enzyme and Microbial Technology 2011;49:321–5.
[45] Doong RA, Wu SC. Effect of substrate concentration on the biotransformation
of carbon tetrachloride and 1,1,1-trichloroethane under anaerobic condition.
Water Research 1996;30:577–86.
[23] Cassetta A, Krastanova I, Kristan K, Brunskole Svegelj M, Lamba D, Lanisnik
Rizner T, et al. Insights into subtle conformational differences in the substrate-
binding loop of fungal 17beta-hydroxysteroid dehydrogenase: a combined
structural and kinetic approach. Biochemical Journal 2012;441:151–60.
[24] Oppermann U, Filling C, Hult M, Shafqat N, Wu X, Lindh M, et al. Short-chain
dehydrogenases/reductases (SDR): the 2002 update. Chemico-Biological Inter-
actions 2003;143-144:247–53.
[25] Filling C, Berndt KD, Benach J, Knapp S, Prozorovski T, Nordling E, et al. Critical
residues for structure and catalysis in short-chain dehydrogenases/reductases.
Journal of Biological Chemistry 2002;277:25677–84.
[46] van der Werf MJ, Keijzer PM, van der Schaft PH. Xanthobacter sp. C20 con-
tains a novel bioconversion pathway for limonene. Journal of Biotechnology
2001;84:133–43.
[26] Schlieben NH, Niefind K, Muller J, Riebel B, Hummel W, Schomburg D. Atomic
resolution structures of R-specific alcohol dehydrogenase from Lactobacillus
brevis provide the structural bases of its substrate and cosubstrate specificity.
Journal of Molecular Biology 2005;349:801–13.
[47] Shen YB, Wang M, Li HN, Wang YB, Luo JM. Influence of hydroxypropyl-
beta-cyclodextrin on phytosterol biotransformation by different strains of
Mycobacterium neoaurum. Journal of Industrial Microbiology and Biotechnol-
ogy 2012;39:1253–9.
[27] Jornvall H, Persson B, Krook M, Atrian S, Gonzalez-Duarte R, Jeffery J, et al. Short-
chain dehydrogenases/reductases (SDR). Biochemistry 1995;34:6003–13.
[28] Lee SA, Belyaeva OV, Kedishvili NY. Biochemical characterization of
human epidermal retinol dehydrogenase 2. Chemico-Biological Interactions
2009;178:182–7.
[29] Sola-Carvajal A, Garcia-Garcia MI, Sanchez-Carron G, Garcia-Carmona F,
Sanchez-Ferrer A. Functional assignment of gene AAC16202.1 from Rhodobacter
capsulatus SB1003: new insights into the bacterial SDR sorbitol dehydrogenases
family. Biochimie 2012;94:2407–15.
[48] Reiss PD, Zuurendonk PF, Veech RL. Measurement of tissue purine, pyrimi-
dine, and other nucleotides by radial compression high-performance liquid
chromatography. Analytical Biochemistry 1984;140:162–71.
[49] Lundquist R, Olivera BM. Pyridine nucleotide metabolism in Escherichia coli. I.
Exponential growth. Journal of Biological Chemistry 1971;246:1107–16.
[50] Banta S, Swanson BA, Wu S, Jarnagin A, Anderson S. Alteration of the specificity
of the cofactor-binding pocket of Corynebacterium 2,5-diketo-d-gluconic acid
reductase A. Protein Engineering 2002;15:131–40.