P-aminosalicylic acid production by enzymatic kolbeschmitt reaction using salicylic acid decarboxylases improved through site-directed mutagenesis

Article abstract of DOI:10.1246/bcsj.20130006

A reversible salicylic acid decarboxylase (Sdc) catalyzes the carboxylation of m-aminophenol (m-AP) to paminosalicylic acid (PAS) as an antituberculous agent, through an enzymatic KolbeSchmitt reaction. To develop a highyield PAS production system through such an enzymatic reaction, we generated Sdc mutants by site-directed mutagenesis and succeeded in generating several mutants showing increased carboxylation specific activities. Among them, a Y64T-F195Y-Sdc mutant showed a 12-fold higher carboxylation specific activity toward m-AP than wild-type Sdc. By the whole-cell reaction of recombinant Escherichia coli BL21(DE3) expressing the gene encoding Y64T-F195Y-Sdc, 70mM PAS was produced from 100 mM m-AP within 2 h. This reaction time was shortened to one-twelfth that of the PAS production using E. coli BL21(DE3) expressing the gene encoding wild-type Sdc (24 h). Moreover, 140 mM PAS was produced from 200 mM m-AP within 9 h by the whole-cell reaction of recombinant E. coli BL21(DE3) expressing the gene encoding Y64T-F195Y-Sdc.

Full text of DOI:10.1246/bcsj.20130006

6
28
Bull. Chem. Soc. Jpn. Vol. 86, No. 5, 628­634 (2013)
© 2013 The Chemical Society of Japan
p-Aminosalicylic Acid Production by Enzymatic Kolbe­Schmitt
Reaction Using Salicylic Acid Decarboxylases Improved
through Site-Directed Mutagenesis
1
1
1
2
1
Saori Ienaga, Sachiyo Kosaka, Yuki Honda, Yoshitaka Ishii, and Kohtaro Kirimura*
1Department of Applied Chemistry, Faculty of Science and Engineering, Waseda University,
3
-4-1 Ohkubo, Shinjuku-ku, Tokyo 169-8555
2Environmental Information and Science Course, School of Social Information Studies, Otsuma Women’s University,
-7-1 Karakida, Tama, Tokyo 206-8540
2
Received January 8, 2013; E-mail: kkohtaro@waseda.jp
A reversible salicylic acid decarboxylase (Sdc) catalyzes the carboxylation of m-aminophenol (m-AP) to p-
aminosalicylic acid (PAS) as an antituberculous agent, through an enzymatic Kolbe­Schmitt reaction. To develop a high-
yield PAS production system through such an enzymatic reaction, we generated Sdc mutants by site-directed mutagenesis
and succeeded in generating several mutants showing increased carboxylation specific activities. Among them, a Y64T-
F195Y-Sdc mutant showed a 12-fold higher carboxylation specific activity toward m-AP than wild-type Sdc. By the
whole-cell reaction of recombinant Escherichia coli BL21(DE3) expressing the gene encoding Y64T-F195Y-Sdc, 70 mM
PAS was produced from 100 mM m-AP within 2 h. This reaction time was shortened to one-twelfth that of the PAS
production using E. coli BL21(DE3) expressing the gene encoding wild-type Sdc (24 h). Moreover, 140 mM PAS was
produced from 200 mM m-AP within 9 h by the whole-cell reaction of recombinant E. coli BL21(DE3) expressing the
gene encoding Y64T-F195Y-Sdc.
The Kolbe­Schmitt reaction1,2 is a well-known method of
synthesizing aromatic hydroxycarboxylic acid via a carbox-
ylation reaction of a phenol salt of an alkali metal under severe
reaction conditions at 125 °C and 4­7 atm carbon dioxide pres-
3
sure. Various salicylic acid derivatives are produced through
this reaction in industrial processes, which can be used as
medicines, herbicides, and industrial products. However, the
conventional Kolbe­Schmitt reaction generates large amounts
of by-products, which have to be separated,1 and its use of
high reaction temperature and pressure has a negative impact
on the environment. Recently, the enzymatic carboxylation
of phenolic compounds has been expected to be a novel eco-
logically benign method compared with the conventional
Kolbe­Schmitt reaction because it catalyzes regioselective
carboxylation under benign reaction conditions. To date,
several reversible decarboxylases, such as 4-hydroxybenzoate
m-AP
PAS
,2
Figure 1. Enzymatic reversible conversion of m-amino-
phenol (m-AP) to p-aminosalicylic acid (PAS).
T. moniliiforme WU-0401 has a novel metabolic pathway for
2
0
the enzymatic conversion of salicylic acid to phenol. In
T. moniliiforme WU-0401, we discovered a novel enzyme,
salicylic acid decarboxylase (Sdc), that reversibly catalyzes
the regioselective carboxylation of phenol to salicylic acid,
namely, the enzymatic Kolbe­Schmitt reaction. The purifi-
cation, characterization, gene cloning, and gene expression of
4
­7
8,9
decarboxylase,
resorcylic acid decarboxylase,
zoate decarboxylase,16 pyrrole-2-carboxylate decarboxyl-
3,4-dihydroxybenzoate decarboxylase, £-
10­15
21
vanillate/4-hydroxyben-
1
7,18
19
21
ase,
and indole-3-carboxylate decarboxylase, have been
Sdc have already been reported. Note that, p-aminosalicylic
reported. Since several decarboxylases reversibly catalyze the
carboxylation of aromatic hydroxylated compounds, in some
cases, valuable aromatic hydroxycarboxylic acids are pro-
acid (PAS) is an important derivatives of salicylic acid and
widely used as an antituberculous agent and that Sdc cata-
lyzes the carboxylation of m-aminophenol (m-AP) to PAS
4
­15,17,18
21
duced.
Among these decarboxylases, the X-ray crystal
(Figure 1). We have attempted to generate a mutant enzyme
structure was solved only for £-resorcylic acid decarboxylase
from Rhizobium sp. MTP-10005, and the amino acid residues
in the active site were determined.13
by site-directed mutagenesis toward Sdc and succeeded in
generating an F195Y-Sdc mutant showing a higher carboxyla-
tion specific activity toward m-AP than wild-type Sdc.
22
We previously isolated Trichosporon moniliiforme WU-
401 as a salicylic acid-degrading yeast and found that
We also succeeded in producing 70 mM PAS within 15 h by
the whole-cell reaction of recombinant E. coli BL21(DE3)
0
Published on the web March 23, 2013; doi:10.1246/bcsj.20130006

S. Ienaga et al.
Bull. Chem. Soc. Jpn. Vol. 86, No. 5 (2013)
629
expressing the gene encoding F195Y-Sdc.22 This is the first
report of a reversible decarboxylase being successfully im-
proved by site-directed mutagenesis.
(A)
In this study, several amino acid residues presumed to be
around the active site of Sdc were replaced with other amino
acid residues by site-directed mutagenesis. Through the experi-
ments, several genetically improved Sdc mutants with desired
properties were generated, and their carboxylation activities
toward m-AP were evaluated. Moreover, PAS production was
performed by the whole-cell reaction of recombinant E. coli
BL21(DE3) expressing the gene encoding one Sdc mutant, and
its reaction time was found to be shortened to one-twelfth that
of the PAS production by the whole-cell reaction of recombi-
nant E. coli BL21(DE3) expressing the gene encoding wild-
type Sdc. Moreover, 140 mM PAS was produced from 200 mM
m-AP within 9 h. Thus, using Sdc mutant improved through
site-directed mutagenesis, we succeeded in developing a high-
yield PAS production system through the enzymatic Kolbe­
Schmitt reaction.
(B)
His164
His169
His169
Asp287
Asp298
Asp298
His218
His224
His224
Results
Asn60
Site-Directed Mutagenesis of Sdc. As for the selection
of amino acid residues for site-directed mutagenesis, a 3D
structure model of Sdc was predicted. The amino acid sequence
of Sdc shows 50%, 39%, 40%, and 29% identities to the 2,3-
dihydroxybenzoic acid decarboxylase of Aspergillus oryzae
Tyr64
Thr64
γ-RA
Phe189
Phe195
Tyr195
(
DHBD, AP007151),23 the £-resorcylic acid decarboxylases of
10
R. radiobacter WU-0108 (Rdc, AB185333) and Rhizobium
sp. MTP-10005 (GRDC, AB170010),12 and the 5-carboxy-
vanillic acid decarboxylase of Sphingomonas paucimobilis
SYK-6 (LigW2, AB089690),24 respectively. Since the GRDC
structure was solved in both substrate-free and substrate-
complexed conformations among these decarboxylases, GRDC
was selected as the template for the structure modeling of Sdc.
The comparative model of Sdc was predicted by the SWISS
Figure 2. Comparison of structures (A) and active cites (B)
of GRDC (yellow), predicted Sdc (Blue), and predicted
Y64T-F195Y-Sdc (Green). The predicted structures of
wild-type Sdc and Y64T-F195Y-Sdc are merged into the
X-ray crystal structure of GRDC (substrate-complexed
conformation: 2DVU). The panel (B) shows the expanded
rectangular area in the panel (A) and indicates the pre-
sumed active site, in relation to the replaced amino acid
residues of Y64T-F195Y-Sdc and £-resorcylic acid (£-RA)
as the substrate of GRDC.
2
5
MODEL based on the structure obtained by X-ray analysis
on GRDC (substrate-complexed conformation: Protein Data
Bank accession No. 2DVU) as the template. On the basis of
the X-ray crystal structure of GRDC, it has been reported
2+
that His164 and Asp287 of GRDC interact with the Zn ion
located at the active site, and that His218 and Asp287 interact
with two water molecules to which 2-hydroxy group of the
substrate is hydrogen-bonded.13
GRDC correspond to His10, Tyr27, Tyr64, Ser65, Pro191, and
Phe195 of Sdc, respectively. Thus, we selected these six amino
acid residues as candidates for site-directed mutagenesis.
Carboxylation Activities toward m-AP to Produce PAS
by Sdc Mutants. For confirmation of the confidence of the
presumed active site of Sdc, His169, His224, and Asp298 of
Sdc, corresponding to His164, His218, and Asp287 of GRDC,
respectively, were replaced with Ala. The resulting Sdc mutants
H169A-Sdc, H224A-Sdc, and D298A-Sdc were detected in
the soluble fraction, but with no debris, and showed neither
decarboxylation nor carboxylation activities, indicating that
His169, His224, and Asp298 of Sdc are essential for the cata-
lytic activities of decarboxylase and carboxylase.
From the information described above, with a view to
obtaining Sdc mutants with higher carboxylation activities
toward m-AP, we generated 43 His-tagged Sdc mutants by site-
directed mutagenesis as follows: Ala10 was replaced with His;
Tyr27 with Ala, Asn, Gln, Phe, Ser, and Thr; Tyr64 with Ala,
Asn, Asp, Cys, Gln, Gly, Phe, Ser, Thr, and Val; Ser65 with
Ala, Asn, Asp, Cys, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Thr,
On the other hand, we have generated a mutant enzyme by
site-directed mutagenesis toward Rdc, which shows a 98%
amino acid sequence identity to GRDC, and confirmed that
H164Q-Rdc and H218Q-Rdc generated by site-directed muta-
genesis show no decarboxylation or carboxylation activity.
Therefore, His169, His224, and Asp298 of Sdc, corresponding
to His164, His218, and Asp287 of GRDC, respectively, are
considered essential amino acid residues for the expression of
reactivity (Figure 2). On the basis of the predicted structure of
Sdc and the alignment of Sdc, GRDC, DHBD, and LigW2, we
focused on amino acid residues in the region around the active
site of Sdc. Many amino acid residues in the region around the
GRDC active site are conserved among these four enzymes,
but several amino acid residues, such as His10, Phe23, Asn60,
Ala61, Pro185, and Phe189 of GRDC, are not (Figure S1). As
for Sdc, Ala10, Phe23, Asn60, Ala61, Pro185, and Phe189 of

6
30
Bull. Chem. Soc. Jpn. Vol. 86, No. 5 (2013)
Production by Enzymatic Kolbe­Schmitt Reaction
(A)
Sdc mutant
(B)
Sdc mutant
Figure 3. Carboxylation specific activities toward m-AP (A) and PAS production from m-AP for 24 h at 30 °C (B) by wild-type Sdc
and Sdc mutants. One unit (U) of enzyme activity is defined as the amount of enzyme necessary to produce 1 nmol of PAS per min.
and Tyr; Pro191 with Ala and Tyr; and Phe195 with Ala, Asp,
His, Ile, Ser, Thr, Trp, Tyr, and Val. We confirmed that all
the Sdc mutants were detected in the soluble fraction by SDS-
PAGE (data not shown). These His-tagged Sdc mutants were
Table 1. Kinetic Parameters for Carboxylation toward
m-AP
a)
kcat/Km)/mM s¹1
¹
1
(
¹1
Enzyme
m
K /mM
cat
k /s
(Relative)
2+
purified using a Ni -affinity column, and their carboxylation
activities toward m-AP were evaluated. The carboxylation
specific activities toward m-AP for 30 min at 30 °C and PAS
production from m-AP for 24 h at 30 °C by the Sdc mutants
were measured (Figure 3). Since the Sdc mutants S65G-Sdc,
S65H-Sdc, S65Y-Sdc, P191A-Sdc, P191Y-Sdc, F195A-Sdc,
F195S-Sdc, and F195T-Sdc showed no carboxylation activities,
they are not shown in Figure 3. Among the Sdc mutants, Y64T-
Sdc and F195Y-Sdc showed specific activities 5-fold and 3.5-
fold higher than that of wild-type Sdc, respectively. In addition,
Y64T-Sdc and F195Y-Sdc produced PAS in 24 h at amounts 4-
fold and 5-fold higher than that of wild-type Sdc, respectively.
Therefore, we generated a Y64T-F195Y-Sdc double mutant.
As shown in Figure 3, the specific activity of Y64T-F195Y-
Sdc was 12-fold higher than that of wild-type Sdc, and is the
highest one among those of the Sdc mutants generated in this
¹
¹3
Sdc
Y64T
F195Y
6.00 © 10 2.00 © 10 1 3.30 © 10 (1.00)
¹2
2.50 © 10
2.76 © 10 4.80 © 10
4.56
1.70
6.81 © 10 (20.0)
¹1
¹2
1.78 © 10 (5.39)
¹2
Y64T-F195Y 8.55 © 10
5.33 © 10 (16.2)
a) Reaction conditions: A reaction mixture containing the
purified enzyme (0.1 mg mL ), 5­50 mM m-AP, and 2 M
KHCO3 was incubated at 30 °C.
¹1
study. Moreover, Y64T-F195Y-Sdc also produced PAS in 24 h
at amounts of 2.8-fold than that of wild-type Sdc.
Characterization of Y64T-F195Y-Sdc. We determined
the kinetic parameters of wild-type Sdc, Y64T-Sdc, F195Y-
Sdc, and Y64T-F195Y-Sdc for carboxylation toward m-AP
(Table 1). We previously determined the kinetic parameters of
2
2
wild-type Sdc and F195Y-Sdc. The Michaelis constants (Km)

S. Ienaga et al.
Bull. Chem. Soc. Jpn. Vol. 86, No. 5 (2013)
631
BL21(DE3) expressing the gene encoding Y64T-F195Y-Sdc.
This reaction time (2 h) was shortened to one-twelfth that
of PAS production using recombinant E. coli BL21(DE3)
expressing the gene encoding wild-type Sdc (24 h).
PAS Production by the Whole-Cell Reaction of Recombi-
nant E. coli BL21(DE3) Expressing the Gene Encoding
Y64T-F195Y-Sdc. We succeeded in generating Y64T-F195Y-
Sdc with a 12-fold higher carboxylation specific activity toward
m-AP than wild-type Sdc. Therefore, we carried out PAS
production by the whole-cell reaction of recombinant E. coli
BL21(DE3) expressing the gene encoding Y64T-F195Y-Sdc.
The optimal conditions for this whole-cell reaction were
examined. The optimal pH cannot be determined since excess
KHCO concentration is needed for the carboxylation reaction
3
2
1
by Sdc as a reversible decarboxylase. The optimal temper-
ature is the same as that for PAS production by wild-type
Sdc (data not shown). However, the optimal substrate (m-AP)
concentrations for PAS production are different between
wild-type Sdc and Y64T-F195Y-Sdc, as shown in Figure 5.
Above 100 mM, the molar conversion yield of PAS production
from m-AP using recombinant E. coli BL21(DE3) expressing
the gene encoding wild-type Sdc gradually decreased with
increasing m-AP concentration. On the other hand, the molar
conversion yield of PAS production from m-AP using recom-
binant E. coli BL21(DE3) expressing the gene encoding Y64T-
F195Y-Sdc was maintained at 70% up to 200 mM m-AP, and
140 mM PAS was maximally produced from 200 mM m-AP.
We found that the optimal substrate concentration for PAS
production using recombinant E. coli BL21(DE3) expressing
the gene encoding Y64T-F195Y-Sdc is 200 mM, which is 2-
fold higher than that using recombinant E. coli BL21(DE3)
expressing the gene encoding wild-type Sdc (100 mM).
We investigated the time course of PAS production from
200 mM m-AP by the whole-cell reaction of recombinant
E. coli BL21(DE3) expressing the gene encoding Y64T-
F195Y-Sdc under optimal conditions. As shown in Figure 6,
140 mM PAS was produced from 200 mM m-AP within 9 h
using recombinant E. coli BL21(DE3) expressing the gene
encoding Y64T-F195Y-Sdc, although only 20 mM PAS was
produced from 200 mM m-AP using recombinant E. coli
BL21(DE3) expressing the gene encoding wild-type Sdc. Since
this reaction is reversible, the molar conversion yield of PAS
production using recombinant E. coli BL21(DE3) expressing
the gene encoding Y64T-F195Y-Sdc was the same (70%) as
that using recombinant E. coli BL21(DE3) expressing the gene
encoding wild-type Sdc.
Figure 4. PAS production from 100 mM m-AP by whole-
cell reaction of recombinant E. coli BL21(DE3) expressing
gene encoding wild-type Sdc or Y64T-F195Y-Sdc. Sym-
bols: open circle, wild-type Sdc; closed circle, Y64T-
F195Y-Sdc.
showed that Y64T-Sdc and F195Y-Sdc have higher affinities
toward m-AP than wild-type Sdc. Although Y64T-F195Y-
Sdc has a lower affinity toward m-AP (K = 8.55 © 10 mM)
m
than wild-type Sdc (K = 6.00 © 10 mM), Y64T-F195Y-Sdc
m
showed the highest carboxylation specific activity among
the Sdc mutants, and the molecular catalytic activity (k ) of
cat
¹
1
Y64T-F195Y-Sdc (4.56 s ) is the highest among those of the
Sdc mutants and 23-fold greater than that of wild-type Sdc
2.00 © 10 s ). The catalytic constants (k /K ) of Y64T-
¹1 ¹1
(
cat m
Sdc, F195Y-Sdc, and Y64T-F195Y-Sdc are also 20-fold, 5-fold,
and 16-fold greater than that of wild-type Sdc, as shown in
Table 1.
We also evaluated the carboxylation specific activity toward
phenol and salicylic acid production by Y64T-F195Y-Sdc
(
details not shown). The carboxylation specific activities
toward phenol of Y64T-F195Y-Sdc and wild-type Sdc were
¹
1
5
.82 and 4.48 © 10 U mg , respectively, under the conditions
where one unit of enzyme activity is defined as the amount of
enzyme necessary to produce 1 nmol of salicylic acid per min.
The carboxylation specific activity toward phenol of Y64T-
F195Y-Sdc decreased and was 0.14 times as much as that
of wild-type Sdc. Moreover, Y64T-F195Y-Sdc and wild-type
¹
2
¹1
Sdc produced 5.46 © 10 and 9.86 © 10 mM salicylic acid
within 24 h, respectively. The salicylic acid production of
Moreover, PAS in the reaction mixture was easily recovered
as follows: The reaction mixture was filtered through a 0.20-
¯m PTFE membrane filter (ADVANTEC Toyo, Tokyo, Japan),
and the filtrate was acidified with 6 M HCl to pH 2.0 with
stirring. Through this simple procedure, a white crystal was
recovered (Figure S2). We confirmed that the white crystal is
PAS of 100% purity, containing no m-AP, using an HPLC
LC-20 system (Shimadzu, Kyoto, Japan).
¹2
Y64T-F195Y-Sdc also decreased and was 5.56 © 10 times
as much as that of wild-type Sdc. These results indicate that
Y64T-F195Y-Sdc is suitable enzyme for PAS production, but
not for salicylic acid production.
We investigated the time course of PAS production from 100
mM m-AP by the whole-cell reaction of recombinant E. coli
BL21(DE3) expressing the gene encoding Y64T-F195Y-Sdc
under optimal conditions for wild-type Sdc (Figure 4). Using
recombinant E. coli BL21(DE3) expressing the gene encoding
wild-type Sdc, 70 mM PAS was produced within 24 h from
Discussion
The conventional Kolbe­Schmitt reaction is not regioselec-
tive, and its use of a high reaction temperature and pressure
has a negative impact on the environment. Therefore, the
1
00 mM m-AP. On the other hand, 70 mM PAS was produced
within 2 h from 100 mM m-AP using recombinant E. coli

6
32
Bull. Chem. Soc. Jpn. Vol. 86, No. 5 (2013)
Production by Enzymatic Kolbe­Schmitt Reaction
(A)
Figure 6. PAS production from 200 mM m-AP by whole-
cell reaction of recombinant E. coli BL21(DE3) expressing
gene encoding wild-type Sdc or Y64T-F195Y-Sdc. Sym-
bols: open circle, wild-type Sdc; closed circle, Y64T-
F195Y-Sdc.
(B)
aromatic hydroxycarboxylic acid production by a genetically
improved reversible decarboxylase has yet been reported so far,
2
2
except for our report an F195Y-Sdc. Moreover, in this study,
we succeeded in producing 140 mM PAS within 9 h at a molar
conversion yield of 70% in aqueous solution by the whole-
cell reaction of recombinant E. coli BL21(DE3) expressing
the gene encoding Y64T-F195Y-Sdc, which is a genetically
improved Sdc mutant based on the comparative model of Sdc.
In this study, the comparative model of Sdc was predicted
2
5
by the SWISS MODEL based on the structure obtained by X-
1
3
ray analysis on GRDC (Figure 2). The predicted structure of
Sdc shows a TIM-barrel structure and is similar to the X-ray
structure of GRDC. On the basis of the multiple alignment of
Sdc, GRDC, DHBD, and LigW2, the six amino acid residues
Ala10, Tyr27, Tyr64, Ser65, Pro191, and Phe195 of Sdc are not
conserved among these decarboxylases, and thus were selected
as target points for site-directed mutagenesis in this study.
Among them, Y64T-Sdc and F195Y-Sdc produced PAS in
Figure 5. PAS production (A) and molar conversion rate
of PAS production (B) at various substrate concentrations
by whole-cell reaction of recombinant E. coli BL21(DE3)
expressing gene encoding wild-type Sdc or Y64T-F195Y-
Sdc under optimal conditions. Symbols: open circle, wild-
type Sdc; closed circle, Y64T-F195Y-Sdc.
2
4 h at amounts 4-fold and 5-fold amounts higher that of wild-
production of aromatic hydroxycarboxylic acids by a reversi-
ble decarboxylase has recently attracted interest, and several
trials for aromatic hydroxycarboxylic acid production from
phenolic compounds through the enzymatic Kolbe­Schmitt
reaction using reversible decarboxylases have been carried
type Sdc, respectively. Among the Tyr64 mutants, as shown in
Figure 3, not only Y64T-Sdc but also Y64A-Sdc, Y64G-Sdc,
and Y64S-Sdc show higher carboxylation specific activities
than wild-type Sdc. The molecular weights of the amino acid
residues at position 64 of these Sdc mutants are lower than
that of wild-type Sdc. These results suggest that the molec-
ular weight of the amino acid residue at position 64 of Sdc is
important for enhancement of the carboxylation activity toward
m-AP. Tyr64 of Sdc corresponds to Asn60 of GRDC, which is
located near the entrance of the substrate in GRDC (Figure 2),
suggesting that the access of m-AP to the active site is
improved by Tyr64 mutants of Sdc. Phe189 of GRDC, corre-
sponding to Phe195 of Sdc, seems likely to limit the substrate
11,15,18,21,22
out.
We previously carried out £-resorcylic acid
11
production by Rdc from R. radiobacter WU-0108. By the
whole-cell reaction of recombinant E. coli BL21(DE3) ex-
pressing the gene encoding Rdc, 20 mM resorcinol was con-
verted to £-resorcylic acid within 7 h at a molar conversion
yield of 44%, and the recombinant E. coli cells were recycla-
ble for use as biocatalysts at least five times.11 Matsui et al.
reported £-resorcylic acid production by a reversible decar-
boxylase of Pandoraea sp. 12B-2 and succeeded in producing
1
3
to having a planar shape (Figure 2). If Phe195 also plays
the same role in Sdc, the amino acid residue at position 195 of
Sdc must be one of the keys to enhancing the affinity toward
m-AP. From predicted structure of Sdc, the para-position of
£
-resorcylic acid from 3 M resorcinol at a molar conver-
sion rate of 48% within 120 h by the whole-cell reaction of
Pandoraea sp. 12B-2.15 On the other hand, no enhancement of

S. Ienaga et al.
Bull. Chem. Soc. Jpn. Vol. 86, No. 5 (2013)
633
¹
1
the phenyl ring of Phe195 is directed toward the amino group
of m-AP. The hydroxy group of Tyr195 in F195Y-Sdc might
interact with the amino group of m-AP, suggesting that this
interaction is important for carboxylation toward m-AP. In
Phe195 mutants, only F195Y-Sdc was markedly enhanced its
carboxylation activity toward m-AP. This suggests that the
critical amino acid residue at position 195 of Sdc is Tyr for
carboxylation toward m-AP.
In this study, we succeeded in generating a Y64T-F195Y-Sdc
mutant with a 12-fold higher carboxylation specific activity
toward m-AP than wild-type Sdc. Moreover, 140 mM PAS
was produced within 9 h at a molar conversion yield of 70%
in aqueous solution by the whole-cell reaction of recombi-
nant E. coli BL21(DE3) expressing the gene encoding Y64T-
F195Y-Sdc. As described above, by site-directed mutagenesis
through genetic engineering, reaction time and PAS production
were improved.
at 120 strokes min . After cultivation for 18 h, the recombi-
nant E. coli BL21(DE3) cells were harvested by centrifugation
at 5000 © g for 10 min at 4 °C and washed twice with 50 mM
K HPO ­KH PO buffer (pH 6.0).
2
4
2
4
2+
Purification of Enzyme Using Ni -Affinity Column.
The E. coli cell extract was obtained by sonicating (20 kHz,
0 °C, 2 min © 5 cycle) the cells after appropriate cultiva-
tion time and removing the cell debris by centrifugation at
15000 © g for 30 min at 4 °C. The resulting cell-free extracts
of the His-tagged wild-type Sdc or Sdc mutants were purified
using a HisTrap· HP column (GE Healthcare Japan, Tokyo,
Japan) equilibrated with 50 mM Tris-HCl buffer (pH 8.0),
and His-tagged Sdc was eluted with 50 mM Tris-HCl buffer
(pH 8.0) containing 500 mM imidazole and 500 mM NaCl. The
His-tagged wild-type Sdc or Sdc mutants were desalted with a
PD-10 column (GE Healthcare Japan, Tokyo, Japan) and eluted
with 3.5 mL of 50 mM K HPO ­KH PO buffer (pH 6.0).
2
4
2
4
Enzyme Assay of Carboxylation. A reaction mixture (500
L) containing the purified enzyme (0.1 mg mL ), 40 mM
Conclusion
¹1
¯
In summary, we designed and generated many Sdc mutants
by site-directed mutagenesis on the basis of predicted structure
of Sdc and the alignment of Sdc with GRDC, DHBD, and
m-AP, and 2 M KHCO in 50 mM K HPO ­KH PO buffer
3 2 4 2 4
(pH 6.0) in a 1.5 mL microcentrifuge tube was incubated at
30 °C. The reaction was stopped by heating for 10 min at 90 °C.
The mixture was then centrifuged at 6000 © g for 10 min at
4 °C, and the supernatant was filtered using a 0.2-¯m PTFE
membrane. The amounts of PAS and m-AP in the filtrate were
¹1
LigW2. We succeeded in producing 70 mM (10.7 g L ) PAS
within 2 h by the whole-cell reaction of recombinant E. coli
BL21(DE3) expressing the gene encoding Y64T-F195Y-Sdc.
¹1
22
Moreover, we succeeded in producing 140 mM (21.4 g L )
PAS within 9 h from 200 mM m-AP. By the whole-cell reac-
tion of recombinant E. coli BL21(DE3) expressing the gene
encoding the Sdc mutant Y64T-F195Y-Sdc, the reaction time of
PAS production was shortened to one-twelfth (24 to 2 h) and
PAS production increased 2-fold (70 to 140 mM). To the best
of our knowledge, this is the first report on the enhancement of
aromatic hydroxycarboxylic acid production using a geneti-
cally improved reversible decarboxylase by a combination of
comparative modeling and site-directed mutagenesis. We have
succeeded in developing an effective bioprocess for PAS
production through the enzymatic Kolbe­Schmitt reaction in
aqueous solution.
determined by HPLC as described previously.
Whole-Cell Reaction of Recombinant E. coli BL21(DE3).
The whole-cells reaction of recombinant E. coli BL21(DE3)
was carried out basically as described previously.1
1,21,22
A
reaction mixture (500 ¯L) containing recombinant E. coli
BL21(DE3) cells (O.D.660 = 30), 20­300 mM m-AP, and 2 M
KHCO3 in 50 mM K HPO ­KH PO buffer (pH 6.0) in a
2
4
2
4
1.5 mL microcentrifuge tube was incubated at 30 °C with recip-
¹1
rocal shaking at 280 strokes min . The reaction was stopped
by heating for 10 min at 90 °C. The resulting mixture was then
centrifuged at 15000 © g for 10 min at 4 °C, and the amounts
of PAS and m-AP in the supernatant were determined by HPLC
2
2
as described previously.
Experimental
Supporting Information
Construction and Cultivation of Recombinant E. coli
Amino acid sequence alignment of Sdc, GRDC, DHBD, and
LigW2, and precipitation of PAS by acidification. These
materials are associated with this article can be found, in the
online version, at http://www.csj.jp/journals/bcsj/.
BL21(DE3).
The gene encoding Sdc was amplified as
described previously.21 The gene encoding Sdc was subcloned
into pET-21a(+) (Novagen, WI, USA) for the generation of
the plasmid to express in E. coli BL21(DE3) cells as host.
Site-directed mutagenesis toward Sdc was performed using
KOD plus DNA polymerase (Toyobo, Osaka, Japan) by PCR
based on Quick Change Site-Directed Mutagenesis (Stratagene,
CA, USA). Plasmids containing the gene encoding wild-type
Sdc or Sdc mutants with a His-tag were constructed for easy
purification and characterization. Plasmids containing the gene
encoding wild-type Sdc or Sdc mutants without a His-tag were
constructed for the use of whole-cell reaction. Each plasmid
was introduced into E. coli BL21(DE3). Recombinant E. coli
BL21(DE3) was precultivated at 37 °C in 5 mL of LB medium
References
1
2
A. S. Lindsey, H. Jeskey, Chem. Rev. 1957, 57, 583.
R. Schmitt, J. Prakt. Chem. 1885, 31, 397.
3
4
I. Omae, Coord. Chem. Rev. 2012, 256, 1384.
X. Zhang, J. Wiegel, Appl. Environ. Microbiol. 1994, 60,
4
182.
5
Z. He, J. Wiegel, Eur. J. Biochem. 1995, 229, 77.
6
7
J. Huang, H. He, J. Wiegel, J. Bacteriol. 1999, 181, 5119.
T. Matsui, T. Yoshida, T. Hayashi, T. Nagasawa, Arch.
Microbiol. 2006, 186, 21.
¹
1
supplemented with 10 ¯g mL ampicillin and cultivated at
5 °C in 500 mL Erlenmeyer flasks with 200 mL of LB medium
8
9
Z. He, J. Wiegel, J. Bacteriol. 1996, 178, 3539.
T. Yoshida, Y. Inami, T. Matsui, T. Nagasawa, Biotechnol.
2
¹
1
supplemented with 10 ¯g mL ampicillin and 0.1 mM isopro-
pyl-¢-D-thiogalactopyranoside (IPTG) with reciprocal shaking
Lett. 2010, 32, 701.
10 Y. Ishii, Y. Narimatsu, Y. Iwasaki, N. Arai, K. Kino, K.

6
34
Bull. Chem. Soc. Jpn. Vol. 86, No. 5 (2013)
Production by Enzymatic Kolbe­Schmitt Reaction
Kirimura, Biochem. Biophys. Res. Commun. 2004, 324, 611.
Y. Iwasaki, K. Kino, H. Nishide, K. Kirimura, Biotechnol.
Lett. 2007, 29, 819.
M. Yoshida, N. Fukuhara, T. Oikawa, J. Bacteriol. 2004,
86, 6855.
M. Goto, H. Hayashi, I. Miyahara, K. Hirotsu, M. Yoshida,
T. Oikawa, J. Biol. Chem. 2006, 281, 34365.
T. Yoshida, Y. Hayakawa, T. Matsui, T. Nagasawa, Arch.
Microbiol. 2004, 181, 391.
T. Matsui, T. Yoshida, T. Yoshimura, T. Nagasawa, Appl.
Microbiol. Biotechnol. 2006, 73, 95.
B. Lupa, D. Lyon, L. N. Shaw, M. Sieprawska-Lupa, J.
Wiegel, Can. J. Microbiol. 2008, 54, 75.
H. Omura, M. Wieser, T. Nagasawa, Eur. J. Biochem. 1998,
53, 480.
M. Wieser, N. Fujii, T. Yoshida, T. Nagasawa, Eur. J.
Biochem. 1998, 257, 495.
19 T. Yoshida, K. Fujita, T. Nagasawa, Biosci., Biotechnol.,
Biochem. 2002, 66, 2388.
20 Y. Iwasaki, H. Gunji, K. Kino, T. Hattori, Y. Ishii, K.
Kirimura, Biodegradation 2010, 21, 557.
21 K. Kirimura, H. Gunji, R. Wakayama, T. Hattori, Y. Ishii,
Biochem. Biophys. Res. Commun. 2010, 394, 279.
22 K. Kirimura, S. Yanaso, S. Kosaka, K. Koyama, T. Hattori,
Y. Ishii, Chem. Lett. 2011, 40, 206.
23 R. Santha, N. A. Rao, C. S. Vaidyanathan, Biochim.
Biophys. Acta, Protein Struct. Mol. Enzymol. 1996, 1293, 191.
24 X. Peng, E. Masai, D. Kasai, K. Miyauchi, Y. Katayama,
M. Fukuda, Appl. Environ. Microbiol. 2005, 71, 5014.
25 L. Bordoli, F. Kiefer, K. Arnold, P. Benkert, J. Battey, T.
Schwede, Nat. Protoc. 2008, 4, 1.
1
1
1
2
1
1
3
1
4
1
5
1
6
1
7
2
1
8