Exploring the synthetic applicability of a new carboxylic acid reductase from Segniliparus rotundus DSM 44985

Article abstract of DOI:10.1016/j.molcatb.2015.01.014

A new carboxylic acid reductase (CAR) gene from Segniliparus rotundus DSM 44985 was overexpressed in Escherichia coli. The recombinant enzyme exhibited high activity toward a variety of aromatic and aliphatic carboxylic acids. Especially, it effectively reduced 4-hydroxybenzoic acid (8a) and 4-nitrobenzoic acid (19a), toward which the known Nocardia CAR exhibited no or little activity. The recombinant E. coli cells co-expressing the Segniliparus CAR and Nocardia PPTase genes catalyzed the reductions of vanillic acid (20a) and 3,4-dihydroxyphenylacetic acid (25a) to give vanillyl alcohol (20c) and 3-hydroxytyrosol (25c) with high yield, respectively. The endogenous aldehyde reductases of E. coli should be responsible for the further reduction of the produced aldehydes. These results demonstrated that Segniliparus CAR was a useful addition to the biocatalyst tool-box for the reduction of carboxylic acids and might find applications in the synthesis of valuable bio-based chemicals from renewable resources.

Full text of DOI:10.1016/j.molcatb.2015.01.014

Journal of Molecular Catalysis B: Enzymatic 115 (2015) 1–7
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Exploring the synthetic applicability of a new carboxylic acid
reductase from Segniliparus rotundus DSM 44985
Yitao Duan, Peiyuan Yao, Xi Chen, Xiangtao Liu, Rui Zhang, Jinhui Feng, Qiaqing Wu∗,
Dunming Zhu∗
National Engineering Laboratory for Industrial Enzymes and Tianjin Engineering Center for Biocatalytic Technology, Tianjin Institute of Industrial
Biotechnology, Chinese Academy of Sciences, 32 Xi Qi Dao, Tianjin Airport Economic Area, Tianjin 300308, China
a r t i c l e i n f o
a b s t r a c t
Article history:
A new carboxylic acid reductase (CAR) gene from Segniliparus rotundus DSM 44985 was overexpressed
in Escherichia coli. The recombinant enzyme exhibited high activity toward a variety of aromatic
and aliphatic carboxylic acids. Especially, it effectively reduced 4-hydroxybenzoic acid (8a) and 4-
nitrobenzoic acid (19a), toward which the known Nocardia CAR exhibited no or little activity. The
recombinant E. coli cells co-expressing the Segniliparus CAR and Nocardia PPTase genes catalyzed the
reductions of vanillic acid (20a) and 3,4-dihydroxyphenylacetic acid (25a) to give vanillyl alcohol (20c)
and 3-hydroxytyrosol (25c) with high yield, respectively. The endogenous aldehyde reductases of E. coli
should be responsible for the further reduction of the produced aldehydes. These results demonstrated
that Segniliparus CAR was a useful addition to the biocatalyst tool-box for the reduction of carboxylic acids
and might find applications in the synthesis of valuable bio-based chemicals from renewable resources.
© 2015 Elsevier B.V. All rights reserved.
Received 11 November 2014
Received in revised form 26 January 2015
Accepted 26 January 2015
Available online 7 February 2015
Keywords:
Carboxylic acids
Biocatalysis
Reduction
Chemoselectivity
1. Introduction
whole-cell bioreductions of carboxylic acids by a number of
microorganisms have been reported, including Corynespora mel-
Carboxylic acids such as fatty acids, ferulic acid, and many
others are available in abundance from renewable resources. As
economic precursors, they can be used to prepare the bio-based
products [1]. For example, fatty acids are reduced to fatty alcohols,
which can be used as biofuels and intermediates in synthesizing
detergents, surfactants, and polymers [2]. The reduction of car-
boxylic acids is a useful and challenging task. Chemical methods
frequently depend on activated carboxylates by LiAlH4, NaBH4,
and their derivatives as reducing agents [3], which are unfavor-
able for large scale applications from the environmental and safety
points of view. More importantly, it is difficult to achieve the
chemoselective reduction of carboxylic acids by chemical meth-
ods, because other reducible moieties such as carbonyl group
and C C bond can be easily affected by such strong reducing
agents [4]. On the contrary, biocatalytic reductions of carboxylic
groups are often highly selective reactions under mild conditions,
thus overcoming the difficulties mentioned above. Furthermore,
enzymatic reactions are generally appreciated by producers of
pharmaceuticals and fine chemicals, and also biocatalytic reduction
processes are gaining increased attention [5]. In the past 20 years,
oni [6], Coriolus [6], Nocardia sp. [7], Pyrococcus furiosus [8,9],
and some plant cells (Actinidia chinensis, Convolvulus sepium, Dau-
cus carota, and so on) [10]. Although many potential sources for
carboxylic acid reductases (CAR) have been reported [1], only a
few CARs have been biochemically characterized, including those
from Nocardia [7,11], Mycobacterium [12], Streptomyces griseus [12],
Mycobacterium marinum [13], and subtype II nonribosomal peptide
synthetase (NRPS)-like proteins such as lnaA [14] and ATEG 03630
[15]. In addition, two reversible aldehyde oxidoreductases, which
not only oxidize aldehydes but also reduce carboxylic acids, were
identified [16–19]. These CARs possess the following consensus
sequence characteristics: (i) their N-terminals and C-terminals
show significant sequence similarity to AMP-binding proteins
and NAD(P)-dependent aldehyde dehydrogenases, respectively, (ii)
they all have a phosphopantetheine attachments site. CAR requires
post-translational phosphopantetheinylation catalyzed by phos-
phopantetheine transferases (PPTase) [20]. ATP, Mg2+, and NADPH
as cofactors are necessary for the reduction [11,20], and ATP and
NADPH are consumed in stoichiometric amounts (Fig. 1) [21].
Meanwhile, it has been found that two genes in the grixazone
biosynthesis gene cluster constituted a CAR [22]. The substrate
profile of an enzyme provides guidance for its synthetic applica-
tion, but this important information of the known CARs is limited
[12,13,23,24].
∗
Corresponding authors. Tel.: +86 22 84861963; fax: +86 22 84861996.
E-mail addresses: wu qq@tib.cas.cn (Q. Wu), zhu dm@tib.cas.cn (D. Zhu).
http://dx.doi.org/10.1016/j.molcatb.2015.01.014
1381-1177/© 2015 Elsevier B.V. All rights reserved.

2
Y. Duan et al. / Journal of Molecular Catalysis B: Enzymatic 115 (2015) 1–7
Fig. 1. Catalytic cycle of carboxylic acid reduction by CAR [20]. (A) Transfer of a phosphopantetheine moiety from the donor (CoA) to an acceptor apo-CAR conserved serine
residue catalyzed by PPTase. (B) Adenylation of acid at the N terminus by ATP yielding acyl-AMP intermediate. (C) Nucleophilic attack of the phosphopantetheine thiol at the
carbonyl carbon of acyl-AMP intermediate, releasing AMP and forming covalently bound acyl thioester. (D) The benzoyl thioester-phosphopantetheinyl “arm” swings from
the adenylating domain to the reduction domain, leading to the reduction of thioester by NADPH, releasing aldehyde, NADP+, and free holo-CAR ready for the next reduction
cycle.
Therefore, in an effort to establish a tool-box of carboxylic acid
reductases with instructive information, a new carboxylic acid
reductase gene from Segniliparus rotundus DSM 44985 was selected
by NCBI BLAST search (http://www.ncbi.nlm.nih.gov/BLAST/) using
the Nocardia CAR (accession number AAR91681.1) as template.
The Segniliparus CAR effectively catalyzed the reduction of a
series of aromatic and aliphatic carboxylic acids including 4-
hydroxybenzoic acid (8a) and 4-nitrobenzoic acid (19a), which
were not reduced by the known Nocardia CAR. The Segnili-
parus CAR gene was thus co-expressed with a Nocardia PPTase
gene in Escherichia coli. The resulting whole-cell biocatalyst was
evaluated for the reduction of carboxylic acids without adding
the expensive cofactors, in an effort to exploring its synthetic
applicability.
NMR spectra were recorded on a Bruker AVANCE-III 500 or 600 MHz
NMR spectrometer (Bruker BioSpin, Rheinstetten, Germany).
2.2. Expression and purification of CAR and PPTase
The optimized N-terminal His-tagged Segniliparus CAR and
Nocardia PPTase genes were synthesized and cloned into the
NdeI/XhoI sites of pET32a(+), respectively, by Shanghai Xuguan
Biotechnological Development Co., Ltd. (Shanghai, China). The
resulting plasmids [pET32a(+)-CAR and pET32a(+)-PPTase] were
transformed into E. coli BL21 (DE3), respectively. A culture of
E. coli BL21 (DE3) cells harboring pET32a(+)-CAR or pET32a(+)-
PPTase was grown overnight in LB-ampicillin (100 ␮g/mL) medium
(5 mL) at 37 ◦C, and then inoculated into 1 L of LB-ampicillin
(100 ␮g/mL) medium. The resulting culture was incubated con-
tinually at 200 rpm in a rotary shaker at 37 ◦C until cells reach
mid-log growth (OD600 of 0.5–1.0), which was followed by the addi-
tion of 0.5 mM IPTG and further incubation for 12 h at 25 ◦C. Cells
were harvested by centrifugation at 12,000 × g for 10 min at 4 ◦C,
and disrupted by high pressure homogenizer after re-suspension
in binding buffer (20 mM sodium phosphate buffer, 0.5 M NaCl,
20 mM imidazole, pH 7.4). His-CAR or His-PPTase fusion protein
in the supernatant fraction was collected from the crude cell lysate
by centrifugation at 12,000 × g for 20 min. Protein purification was
performed on a HisTrapTM FF crude column (GE Healthcare, Piscat-
away, USA), and the protein was desorbed with an elution buffer
(20 mM sodium phosphate, 0.5 M NaCl, 0.5 M imidazole, pH 7.4).
The purified proteins His-CAR or His-PPTase were dialyzed in a
sodium phosphate buffer (50 mM, pH 7.5) and then stored at −20 ◦C
for further use.
2. Materials and methods
2.1. Materials
PrimeSTAR HS DNA polymerase and the DNA ligation kit were
from TaKaRa (Dalian, China). FastDigest Restriction Enzymes were
purchased from Fermentas (Shenzhen, China). Vectors [pET32a(+)
and pETDuet-1] were purchased from Novagen (Schwalbach,
Germany). The plasmid extraction kit was from CWBIO (Beijing,
China). The gel extraction kit was from Tiangen Biotech (Beijing,
China). The lyophilized powder of glucose dehydrogenase (GDH)
from Bacillus subtilis, with a specific activity of 1.1 U/mg, was pre-
pared in our laboratory. All carboxylic acids were purchased from
Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China), Alfa Aesar
(Shanghai, China) or Sigma–Aldrich (St. Louis, USA). 1H and 13C

Y. Duan et al. / Journal of Molecular Catalysis B: Enzymatic 115 (2015) 1–7
3
2.3. Molecular weight of Segniliparus CAR
from 1 M stock solution in DMSO), NADP+ (0.9 mM), GDH (1 U),
glucose (60 mM), ATP (15 mM) and MgCl2 (0 mM or 10 mM) (Sup-
plementary Table T1).
The molecular mass of Segniliparus CAR was determined by gel
filtration chromatography using a Superdex 200 10/300 GL column
and the high-molecular-weight gel filtration calibration kit (GE
Healthcare, Piscataway, USA) was used for calibration. The sodium
phosphate buffer (50 mM, pH 7) containing 150 mM NaCl was used
as the eluent. The flow rate was 0.4 mL/min and the absorbance at
280 nm was monitored.
2.7. Substrate specificity
The reduction of a series of carboxylic acids was carried out
by following the standard reduction procedure. The yields were
determined by GC analysis and the products were characterized by
GC–MS.
2.4. Standard reduction procedure
2.8. Analytical procedure
The His-CAR (1.5 mg) was incubated with His-PPTase (295 ␮g)
in the presence of CoA (1 mM) as a cofactor for 1 h at 28 ◦C in a final
volume of 600 ␮L of sodium phosphate buffer (100 mM, pH 7.5)
containing 10 mM of MgCl2. The resulting enzyme mixture (holo-
CAR, 50 ␮g) was mixed with NADP+ (0.9 mM), GDH (1 U, one unit
corresponds to the amount of enzyme which could reduce 1 ␮mol
NADP+ to NADPH per minute), glucose (60 mM), MgCl2 (10 mM),
substrate (5 or 10 mM, from 1 M stock solution in DMSO), and ATP
(15 mM) in Tris–HCl buffer (100 mM, pH 9) with a final volume of
1 mL. The reaction mixture was incubated at 100 rpm in a rotary
shaker at 35 ◦C for 16 h, and extracted with 1 mL of ethyl acetate
after the pH was adjusted to 2–3 with 1 M HCl solution. The organic
extracts were dried over anhydrous sodium sulfate and analyzed
by gas chromatography (GC) and gas chromatography–mass spec-
trometry (GC–MS) to determine the amount of substrate (a) and
products (aldehyde b, alcohol c) in the mixture. All experiments
were conducted in triplicate.
GC was performed using
a
CP-Chirasil-DEXCB column
(30 m × 0.25 mm × 0.25 ␮m) with flame ionization detector.
The carrier gas was helium at a flow rate of 2 mL/min. The column
temperature was controlled as follows: 120 ◦C – 10 ◦C/min – 160 ◦C
(8 min) – 40 ◦C/min – 180 ◦C (5 min). The identity of products
was confirmed with authentic standards. The retention times
for phenylacetaldehyde (23b), phenethyl alcohol (23c) and 23a
were 3.2, 4.7 and 9.0 min, respectively. Yield was expressed as a
molar ratio of product to product plus substrate determined by
GC. GC–MS was performed using an Agilent 7890 GC, equipped
with a HP-5 capillary column (30 m × 0.25 mm × 0.25 ␮m) and a
mass spectrometer 5975C as detector. The carrier gas was helium
at a flow rate of 1 mL/min. Column temperature was controlled
as follows: 60 ◦C (4 min) – 20 ◦C/min – 200 ◦C (6 min) – 20 ◦C/min
– 250 ◦C (4 min). For MS detection, the electron ionization was
performed with an ionization energy of 70 eV. The products were
identified by comparing the fragmentation patterns with the NIST
mass spectral library.
2.5. Determination of kinetic parameters
The enzyme mixtures were prepared by following the standard
assay procedure. The reaction was performed in Tris–HCl buffer
(100 mM, pH 9) containing enzyme mixture (holo-CAR, 20 ␮L,
5.87 ␮g), MgCl2 (10 mM), varying concentrations of benzoic acid
(5a, 0.05–15 mM), ATP (1.2 mM), and NADPH (0.4 mM) in a final
volume of 200 ␮L. For ATP, 5a (4 mM), NADPH (0.4 mM), enzyme
mixture (holo-CAR, 20 ␮L, 5.87 ␮g) and ATP in the range of
0.01–0.5 mM were used for the activity assay. For NADPH, 5a
(4 mM), ATP (1.2 mM), enzyme mixture (holo-CAR, 20 ␮L, 5.87 ␮g)
and NADPH in the range of 0.01–0.5 mM were used for the
activity assay. The activity of CAR was determined by spec-
trophotometrically measuring the oxidation of NADPH at 340 nm
(ε = 6.22 mM−1 cm−1). The specific activity was defined as the num-
ber of ␮mol of NADPH converted in 1 min by 1 mg of enzyme
(␮mol min−1 mg−1). All experiments were conducted in triplicate.
The kinetic parameters were obtained by measuring the initial
velocities of the enzymatic reaction and curve-fitting according to
the Michaelis–Menten equation using GraphPad Prism 5 software
(GraphPad Software Inc., San Diego, USA).
2.9. Preparation of E. coli cells co-expressing CAR and PPTase
A CAR DNA fragment was acquired from the vector pET32a(+)-
CAR by digesting at the restriction sites NdeI and XhoI, and then
ligated by T4 DNA ligase into pETDuet-1 at the same restriction sites
to generate the expression vector pETDuet-1-CAR. The PPTase gene
was amplified by PCR using plasmid pET32a(+)-PPTase as template
and primers containing the restriction sites NcoI and SalI, respec-
tively, PPTase-F 5ꢀ-CATGCCATGGCAATGATCGAAACCATCCTGCCG-
3ꢀ and PPTase-R 5ꢀ-ACGCGTCGACTTATTAAGCGTAAGCGATCGCGG-
3ꢀ. The PCR fragment was digested with NcoI and SalI, and then
ligated with pETDuet-1-CAR at the same restriction sites to afford
the expression vector pETDuet-1-PPTase-CAR, which was con-
firmed by DNA sequencing. The confirmed recombinant vector
was transformed into E. coli BL21(DE3). A 100 mL of LB-ampicillin
(100 ␮g/mL) medium harboring E. coli BL21(DE3)/pETDuet-1-
PPTase-CAR was grown overnight and then inoculated into 9 L of
LB-ampicillin (100 ␮g/mL) medium in a fermenter. When the cells
reached the mid-log growth period (OD600 of 0.5–1.0), the induc-
tion was initiated by the addition of 0.5 mM IPTG and the culture
was incubated for another 12 h at 25 ◦C. Cells were harvested by
centrifugation at 12,000 × g for 10 min at 4 ◦C, and washed once
with phosphate buffer (100 mM, pH 8).
2.6. Mechanism of Segniliparus CAR
To determine whether the mechanism is consistent with that
of the Nocardia CAR (Fig. 1), the following reaction mixtures of
sodium phosphate buffer (1 mL, 100 mM, pH 7.5) were incubated
at 30 ◦C or 35 ◦C for 12 h (Supplementary Fig. S2 and Table T1).
For the trapping of possible intermediate phenylacetyl thioester in
Fig. 1, the reaction mixture contained enzyme mixture (holo-CAR,
100 ␮g), phenylacetic acid (23a, 10 mM) from 1 M stock solution in
methanol (Supplementary Fig. S2a) or DMSO (Supplementary Fig.
S2b), NADP+ (0.9 mM), GDH (1 U), glucose (60 mM), ATP (15 mM)
and MgCl2 (10 mM). For the effect of Mg2+ and post-translational
phosphopantetheinylation, the reaction mixture contained His-
CAR (apo-CAR) or enzyme mixture (holo-CAR) (30 ␮g), 23a (10 mM,
2.10. Reduction with whole cell biocatalyst
Glucose (22.2 mM), E. coli BL21(DE3)/pETDuet-1-PPTase-CAR
(wet cells, 10 g), and the substrate (20a, 5.0 mM) were mixed in
the sodium phosphate buffer (100 mL, 100 mM, pH 8). The result-
ing mixture was incubated at 200 rpm in a rotary shaker at 30 ◦C,
and the reaction was monitored by GC. After 29 h, the pH of the
reaction mixture was adjusted to 2–3 with 2 M HCl and the mix-
ture was filtered through a Celite pad to remove the biomass. The
resulting aqueous solution was extracted with ethyl acetate. The

4
Y. Duan et al. / Journal of Molecular Catalysis B: Enzymatic 115 (2015) 1–7
organic phase was dried over anhydrous sodium sulfate, filtered,
and concentrated under reduced pressure. The residue was purified
by a silica gel column to give the product.
compared to the rate of acid reduction D (Fig. 1), which may be the
rate-limiting step in the whole reaction. The N-terminal domain
of CAR is responsible for adenylation and thioesterification of car-
boxylic acid, and it may play a role in determining the substrate
specificity, which is similar to the adenylation domain of NRPS [28].
No reaction was observed when apo-CAR was used in the reaction
system (23a as substrate) or the reaction system did not contain
Mg2+ (Supplementary Table T1), indicating that post-translational
phosphopantetheinylation and Mg2+, ATP, and NADPH as cofactors
were necessary for this enzymatic reduction.
The optimum pH of the enzymatic reduction of 23a was between
pH 9 and 9.5, but yield of 23a did not significantly decrease at pH
7–8 and kept at more than 80% (Supplementary Fig. S3). This is
probably because phosphate had better buffering effect for the pH
change caused by the phosphoric acid formed from the reaction
of water and PPi during reaction process (Fig. 1). The yield of the
enzymatic reduction increased slightly from 25 to 35 ◦C.
2.10.1. 4-Hydroxy-3-methoxybenzyl alcohol (20c)
20c (53.8 mg, 0.35 mmol, 69%) was obtained from vanillic acid
(20a). 1H NMR (600 MHz, DMSO-d6): ı = 8.75 (s, 1H, 4-OH), 6.87
(s, 1H, 2-H), 6.65–6.73 (m, 2H, 5-H, 6-H), 4.97 (t, J = 5.7 Hz, 1H,
CH2OH), 4.37 (d, J = 5.5 Hz, 2H, CH2 ), 3.75 (s, 3H, OCH3). 13C
NMR (150 MHz, DMSO-d6): ı = 147.34 (C-3), 145.27 (C-4), 133.47
(C-1), 119.07 (C-6), 115.01 (C-5), 111.03 (C-2), 62.97 ( CH2OH),
55.49 (OCH3).
2.10.2. 2-(3,4-Dihydroxyphenyl)ethanol (25c)
25c (32.9 mg, 0.21 mmol, 42%) was obtained from 3,4-
dihydroxyphenylacetic acid (25a). 1H NMR (500 MHz, DMSO-d6):
d = 8.70 (br. s., 1H, 3ꢀ-OH), 8.60 (br. s., 1H, 4ꢀ-OH), 6.59 (d, J = 8.3 Hz,
1H, 5ꢀ-H), 6.56 (d, J = 1.0 Hz, 1H, 2ꢀ-H), 6.41 (d, J = 7.8 Hz, 1H, 6ꢀ-H),
4.55 (br. s., 1H, 1-OH), 3.48 (br. s., 2H, 1-CH2), 2.51 (t, J = 7.3 Hz, 2H,
2-CH2). 13C NMR (125 MHz, DMSO-d6): ı = 145.54 (C-3ꢀ), 143.97 (C-
4ꢀ), 130.81 (C-1ꢀ), 120.11 (C-6ꢀ), 116.93 (C-2ꢀ), 116.02 (C-5ꢀ), 63.32
(C-1), 39.21 (C-2).
The Km and catalytic efficiencies (kcat/Km) of Segniliparus CAR
toward 5a were 404 37 ␮M and 0.32 ␮M−1 min−1, respectively.
They fell in the range of those for Nocardia CAR [23], and M. marinum
CAR (Supplementary Table T2) [13].
2.10.3. 3-(3-Methoxy-4-hydroxyphenyl)-1-propanol (29d)
29d (20.0 mg, 0.11 mmol, 22%) was obtained from trans-ferulic
acid (29a). 1H NMR (500 MHz, DMSO-d6): ı = 8.63 (s, 1H, 4ꢀ-OH),
6.71 (s, 1H, 2ꢀ-H), 6.63 (d, J = 7.8 Hz, 1H, 5ꢀ-H), 6.54 (d, J = 7.8 Hz,
1H, 6ꢀ-H), 4.43 (t, J = 5.1 Hz, 1H, 1-OH), 3.72 (s, 3H, OCH3), 3.39
(m, 2H, 1-CH2), 2.43–2.50 (m, 2H, 3-CH2), 1.65 (quin, J = 7.0 Hz, 2H,
2-CH2). 13C NMR (125 MHz, DMSO-d6): ı = 148.00 (C-3ꢀ), 144.98 (C-
4ꢀ), 133.62 (C-1ꢀ), 120.95 (C-6ꢀ), 115.90 (C-5ꢀ), 113.11 (C-2ꢀ), 60.84
(C-1), 56.16 (OCH3), 35.30 (C-2), 31.91 (C-3).
3.2. Substrate specificity
The substrate specificity of Segniliparus CAR was examined
with the purified enzyme. The results in Table 1 showed that a
series of aromatic and aliphatic carboxylic acids were reduced to
their corresponding aldehydes. For the aliphatic acids, the reduc-
tion of nonanoic acid (2a) resulted in higher yields than those of
hexanoic acid (1a) and myristic acid (3a). Segniliparus CAR also
showed high activity toward aromatic carboxylic acids. For exam-
ples, 5a, 23a, 3-phenylpropanoic acid (26a), and 4-phenylbutyric
acid (30a) were reduced completely. Ortho-substituted benzoic
acids [hydroxyl (6a), methoxyl (9a), fluoro (12a), chloro (14a),
and bromo (16a)] were the poorer substrates than their corre-
sponding meta-substituted ones (7a, 10a, 13a, 15a, 17a). This
enzyme was less active toward 2-phenylpropionic acid [24a, (S)-
24a] and 3-phenylbutyric acid (27a) than 23a and 26a, respectively.
These results suggested that the activity of Segniliparus CAR might
be strongly influenced by the steric demand surrounding car-
boxylic groups, which was consistent with the results for P.
furiosus (whole-cell) [8] and Nocardia CAR [23]. Segniliparus CAR
showed good chemoselectivity for the reduction of some car-
boxylic acids containing C C or C O double bonds, such as linoleic
acid (4a), trans-cinnamic acid (28a), 29a, and 3-benzoylpropionic
acid (31a). These acids were reduced, with C C or C O double
bonds remaining unaffected. However, this enzyme was less active
toward 28a and 3-benzoylacrylic acid (32a) than their saturated
counterparts (26a and 31a). Interestingly, Segniliparus CAR could
effectively catalyze the reduction of 8a and 19a, toward which
Nocardia CAR exhibited no or little activity [23]. The complete
conversion of both rac-24a and (S)-24a (5 mM) under identical con-
ditions suggested that Segniliparus CAR had low enantioselectivity
toward racemic 24a. For ibuprofen [rac-33a and (S)-33a], the con-
version of rac-33a (14% yield of rac-33b) was higher than (S)-33a
[3% yield of (S)-33b], suggesting that Segniliparus CAR might be
more active toward the (R)-enantiomer than (S)-enantiomer, but
the enantioselectivity was also low. Similarly, rac-24a, rac-27a and
rac-33a were also almost completely reduced by M. marinum CAR
[13] in our experiments (data not shown). However, by compar-
ing Vmax/Km ratios of Nocardia CAR toward (R)-33a and (S)-33a, the
enantiomeric ratio was calculated to be 40.4 for the (R)-isomer over
the (S)-isomer [23]. The conversion of rac-33a catalyzed by Nocar-
dia sp. indicated that Nocardia CAR was R enantioselective, giving
the chiral alcohol product with an enantiomeric excess of 61.2% [7].
3. Results and discussion
3.1. Purification and biochemical characterization of Segniliparus
CAR
A putative carboxylic acid reductase from S. rotundus DSM
44985 (accession number YP 003658971.1), which possessed the
CAR consensus sequence characteristics, was chosen. The gene
was synthesized and expressed in E. coli [11]. Since PPTase
from S. rotundus has not been determined, a known Nocardia
PPTase (accession number ABI83656.1) was selected for the post-
translational phosphopantetheinylation of Segniliparus CAR [20].
Recombinant His-CAR and His-PPTase were produced as soluble
proteins under the same conditions (OD600 – 0.6, 0.5 mM IPTG,
25 ◦C, 12 h) and purified in one chromatographic step using a
HisTrapTM FF crude column (Supplementary Fig. S1). The molecular
mass of His-CAR was estimated to be about 158 kDa by gel filtration
chromatography. Since its theoretical value is 129 kDa, this enzyme
is a monomeric protein, which is consistent with Nocardia CAR [23].
The reaction mechanism of Segniliparus CAR was similar to that
of Nocardia CAR [20,25]. When methanol was added into the reac-
tion system (23a as substrate), methyl phenylacetate was observed
by GC (Supplementary Fig. S2). The methyl ester was probably pro-
duced from the reaction of methanol with phenylacetyl thioester
(Fig. 1). Similar aryl-aldehyde NADP+ oxidoreductase from Neu-
rospora crassa was demonstrated to catalyze a rapid exchange of
pyrophosphoric acid (PPi) with ATP during the reduction of 5a
[26]. The adenylation of some carboxylic acids involving ATP to
be converted to AMP and PPi was almost in equilibrium [26,27].
For Nocardia CAR, benzoyl-AMP had a significantly lower (67-
fold) apparent Km (9.66 0.71 ␮M) and a higher (1.3-fold) Vmax
(7.50 0.18 ␮mol min−1 mg−1) than those for 5a [23,25]. Therefore,
the reactions B and C at the N terminus of CAR seem to be rapid

Y. Duan et al. / Journal of Molecular Catalysis B: Enzymatic 115 (2015) 1–7
5
Table 1
Table 1 (Continued)
Substrate specificity of Segniliparus CAR.
Substrate
Yield (%) of aldehyde
37
Substrate
Yield (%) of aldehyde
1a (n = 3)
2a (n = 6)
3a (n = 11)
79
100
65
25aab
26a
4aa
6
100
94
5a
100
17
27a
28a
29ab
6a (ortho)
7a (meta)
8a (para)
100
100
51
9a (ortho)
2
52
10a (meta)
11a (para)
93
100
12a (ortho)
82
30a
31a
100
95
13a (meta)
14a (ortho)
86
18
15a (meta)
16a (ortho)
41
2
32a
0
17a (meta)
70
18
33a (R/S)
33a (S)
14
3
18ab
a
b
c
The reaction was performed in sodium phosphate buffer (100 mM, pH 7.5).
Silylation was performed before the GC analysis.
The substrate was taken from 500 mM stock solution in DMSO.
Substrate concentration (5 mM).
d
19ac
20a
100
93
3.3. Whole cell catalysis
In order to explore the application potential of Segniliparus CAR
as a biocatalyst for the reduction of carboxylic acids, the Segnili-
parus CAR and Nocardia PPTase genes were co-expressed in E.
coli BL21(DE3). Without externally adding the cofactors (ATP and
NADPH), the recombinant whole cells catalyzed the reduction of
23a with 48% yield, which was much higher than that using the
E. coli cells expressing only the Segniliparus CAR gene (Table 2)
[24,29]. The optimum pH for the whole-cell biotransformation was
21ab
62
Table 2
22ab
23a
52
Reduction of 23a catalyzed by E. coli BL21(DE3)/pETDuet-1-PPTase-CAR and E. coli
BL21(DE3)/pET32a(+)-CAR.a
Conditions
Yield (%) of 23c
100
E. coli BL21(DE3)/pET32a(+)-CAR + glucose
E. coli BL21(DE3)/pETDuet-1-PPTase-CAR
E. coli BL21(DE3)/pETDuet-1-PPTase-CAR + glucose
14
14
48
24a (R/S)
24a (S)
32; 100d
55; 100d
a
Reaction conditions: sodium phosphate buffer (1 mL, 100 mM, pH 7.5) contained
glucose (22.2 mM), E. coli BL21(DE3)/pET32a(+)-CAR or E. coli BL21(DE3)/pETDuet-
1-PPTase-CAR (wet cells, 100 mg), 23a (10 mM); 16 h, 30 ◦C, 200 rpm.

6
Y. Duan et al. / Journal of Molecular Catalysis B: Enzymatic 115 (2015) 1–7
Fig. 3. Time courses for the reductions of 20a (square) to 20c (circle) and 25a (uptri-
angle) to 25c (downtriangle) by the whole cells.
Table 3
Reduction of 29a by various amount of whole cells.a
Conditions
Yield (%) of 29b and 29d
Ratio of 29d/29b
100 mg cells + 5 mM 29a
100 mg cells + 2.5 mM 29a
200 mg cells + 2.5 mM 29a
>90
>90
>90
12/88
46/54
76/24
a
Reaction conditions: sodium phosphate buffer (1 mL, 100 mM, pH 8) contained
glucose (22.2 mM), E. coli BL21(DE3)/pETDuet-1-PPTase-CAR (wet cells), and 29a;
16 h, 30 ◦C, 200 rpm.
cells co-expressing Nocardia CAR and a PPTase from E. coli, only 19%
yield and 2 mg L−1 h−1 productivity were achieved after 144 h. The
endogenous aldehyde reductases were proposed to be responsible
for the further reduction of the aldehyde product [24,29].
Fig. 2. Effects of pH (a) and temperature (b) on the reduction of 23a to 23c by
the whole cells. (a) Reaction conditions: different buffers [1 mL, square 100 mM
sodium phosphate buffer (pH 6.0–8.0), round 100 mM Tris–HCl buffer (pH 7.1–9.0)]
contained glucose (22.2 mM), E. coli BL21(DE3)/pETDuet-1-PPTase-CAR (wet cells,
50 mg), 23a (5 mM); 16 h, 30 ◦C, 200 rpm. (b) The reaction at the optimum pH was
carried out at different temperatures (25 ◦C, 30 ◦C, 37 ◦C and 42 ◦C).
In contrast to 20a and 25a, coniferyl aldehydes (29b) and 29d
were produced when 29a was reduced by the whole cells co-
expressing the Segniliparus CAR and Nocardia PPTase genes. This
is different from the reduction catalyzed by the recombinant E. coli
cells harboring the Nocardia CAR and PPTase, in which 29b and
coniferyl alcohol were obtained [24]. The purified Segniliparus CAR
did not catalyze the reduction of C C bond of 29a, so we specu-
lated that 29a was first reduced to 29b by the recombinant CAR and
29b was further reduced to 29d by endogenous ene reductase and
aldehyde reductase (Scheme 1). This is consistent with the obser-
vation that the ratio of 29d/29b increased when more whole cell
biocatalyst was used (Table 3).
at pH 8 (Fig. 2), while the reaction temperature did not exert signif-
icant effect on the reaction rate in the range of 25–37 ◦C, although
the yield decreased at 42 ◦C. The reductions of 20a and 25a were
carried out at pH 8 and 30 ◦C. As shown in Fig. 3, the conver-
sions of 20a and 25a gradually increased over time, reaching up
to 87% and 69% at 29 h. 20c (18.6 mg L−1 h−1 productivity) and 25c
(11.3 mg L−1 h−1 productivity) were isolated in 69% and 42% yields,
respectively. The lower isolated yield of 25c was due to its higher
solubility in aqueous solution and the instability of 25a and proto-
catechualdehyde (25b) under alkaline conditions (pH 8). The E. coli
cells co-expressing Nocardia CAR and PPTase catalyzed the reduc-
tion of 20a to give vanillin (20b) as the major product together with
20c [24]. Recently, Napora-Wijata et al. [29] reported the synthesis
of 25c through the reduction of 25a using whole E. coli BL21 (DE3)
Carboxylic acid reductases require Mg2+, ATP, and NADPH as
cofactors, and the use of whole cells could avoid the introduction
of external cofactor regeneration system into the reaction. How-
ever, the diffusion of the substrate into the cell and the insufficient
intracellular cofactors may decrease the efficiency of the reaction.
Therefore, high cell loading or low conversion until now has been
observed for the reduction of carboxylic acids with whole cell bio-
catalysts [24,29]. Although the studies on ATP/NADPH regeneration
O
O
O
E.coli ene
E.coli aldehyde
O
Segniliparus CAR
NADPH/ATP/Mg2+
O
O
reductase
NAD(P)H
O
reductase
OH
OH
H
H
NAD(P)H
HO
HO
HO
HO
29a
29b
29c
29d
E.coli BL21(DE3)/pETDuet-1-PPTase-CAR
Scheme 1. Proposed process for the reduction of 29a by the whole cells.

Y. Duan et al. / Journal of Molecular Catalysis B: Enzymatic 115 (2015) 1–7
7
have been carried out [30,31], and the cofactor NADPH regenera-
References
tion by GDH has been brought into the reduction of carboxylic acid
[24], the challenge in the recycling of these cofactors has not been
effectively addressed. Our attempt to bring ATP regeneration into
the reaction with crude enzymes was also unsuccessful.
[1] K. Napora-Wijata, G.A. Strohmeier, M. Winkler, Biotechnol. J. 16 (2014)
822–843.
[2] K. Noweck, W. Grafahrend, Ullmann’s Encyclopedia of Industrial Chemistry,
Wiley-VCH, Weinheim, 2006.
[3] J. Seyden-Penne, Reductions by the Alumino- and Borohydrides in Organic
Synthesis, 2nd ed., Wiley-VCH, Pittsburgh, 1997.
[4] P.A. Dub, T. Ikariya, ACS Catal. 2 (2012) 1718–1741.
[5] F. Hollmann, I.W.C.E. Arends, D. Holtmann, Green Chem. 13 (2011) 2285–2314.
[6] H.A. Arfmann, W.R. Abraham, Z. Naturforsch. C: Biochem. Biophys. Biol. Virol.
48 (1993) 52–57.
[7] Y. Chen, J.P. Rosazza, Appl. Environ. Microbiol. 60 (1994) 1292–1296.
[8] E.C.D. van den Ban, H.M. Willemen, H. Wassink, C. Laane, H. Haaker, Enzyme
Microb. Technol. 25 (1999) 251–257.
[9] F. Hollmann, Y. Ni, P.L. Hagedoorn, I.W.C.E. Arends, J.H. Xu, Chem. Commun. 48
(2012) 12056–12058.
[10] R. Villa, F. Molinari, J. Nat. Prod. 71 (2008) 693–696.
[11] A.M. He, T. Li, L. Daniels, I. Fotheringham, J.P.N. Rosazza, Appl. Environ. Micro-
biol. 70 (2004) 1874–1881.
[12] B. Behrouzian, R. Mcdaniel, X. Zhang, L. Clark, World Patent WO2010135624A2
(2010).
4. Conclusion
A gene encoding the thioester reductase domain-containing
protein from S. rotundus DSM 44985 was successfully overex-
pressed in E. coli. The purified recombinant His-CAR exhibited a
broad substrate spectrum and was active toward both aliphatic
and aromatic carboxylic acids. Other functional groups such as
keto groups and C C double bonds remained unaffected. The Seg-
niliparus CAR and Nocardia PPTase genes were co-expressed in E.
coli BL21(DE3), and the recombinant whole cells convert vanil-
lic acid (20a) and 3,4-dihydroxyphenylacetic acid (25a) to give
vanillyl alcohol (20c) and 3-hydroxytyrosol (25c), respectively,
with the concurrent function of the endogenous aldehyde reduc-
tases of E. coli. As such, Segniliparus CAR may serve as a useful
biocatalyst for the reduction of carboxylic acids and find appli-
cations in the synthesis of valuable compounds such as vanillyl
alcohol (20c) and 3-hydroxytyrosol (25c), and/or be integrated
into the metabolic pathways for producing carboxylic acid-derived
biochemicals.
[13] M.K. Akhtar, N.J. Turner, P.R. Jones, Proc. Natl. Acad. Sci. U. S. A. 110 (2013)
87–92.
[14] R.R. Forseth, S. Amaike, D. Schwenk, K.J. Affeldt, D. Hoffmeister, F.C. Schroeder,
N.P. Keller, Angew. Chem. Int. Ed. 52 (2013) 1590–1594.
[15] M. Wang, M. Beissner, H. Zhao, Chem. Biol. 21 (2014) 257–263.
[16] H. White, C. Huber, R. Feicht, H. Simon, Arch. Microbiol. 159 (1993) 244–249.
[17] C. Huber, J. Caldeira, J. Jongejan, H. Simon, Arch. Microbiol. 162 (1994) 303–309.
[18] G. Strobl, R. Feicht, H. White, F. Lottspeich, H. Simon, Biol. Chem. Hoppe-Seyler
373 (1992) 123–132.
[19] H. White, R. Feicht, C. Huber, F. Lottspeich, H. Simon, Biol. Chem. Hoppe-Seyler
372 (1991) 999–1005.
[20] P. Venkitasubramanian, L. Daniels, J.P.N. Rosazza, J. Biol. Chem. 282 (2007)
478–485.
Acknowledgements
[21] K. Kameda, W.D. Nunn, J. Biol. Chem. 256 (1981) 5702–5707.
[22] H. Suzuki, Y. Ohnishi, S. Horinouchi, J. Antibiot. 60 (2007) 380–387.
[23] T. Li, J.P. Rosazza, J. Bacteriol. 179 (1997) 3482–3487.
[24] P. Venkitasubramanian, L. Daniels, S. Das, A.S. Lamm, J.P. Rosazza, Enzyme
Microb. Technol. 42 (2008) 130–137.
[25] T. Li, J.P. Rosazza, J. Biol. Chem. 273 (1998) 34230–34233.
[26] G.G. Gross, Eur. J. Biochem. 31 (1972) 585–592.
[27] A. Rodriguez, E. Meighen, J. Biol. Chem. 260 (1985) 771–774.
[28] M. Wang, H. Zhao, ACS Catal. 4 (2014) 1219–1225.
[29] K. Napora-Wijata, K. Robins, A. Osorio-Lozada, M. Winkler, ChemCatChem 6
(2014) 1089–1095.
This work was financially supported by National Basic Research
Program of China (973 Program, No. 2011CB710800) and the Chi-
nese Academy of Sciences (KSZD-EW-Z-015).
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcatb.
2015.01.014.
[30] H. Zhao, W.A. van der Donk, Curr. Opin. Biotechnol. 14 (2003) 583–589.
[31] W.A. van der Donk, H. Zhao, Curr. Opin. Biotechnol. 14 (2003) 421–426.