Enantiodifferentiation of ketoprofen by Japanese firefly luciferase from Luciola lateralis

Article abstract of DOI:10.1016/j.molcatb.2011.01.008

Recently, we found that firefly luciferase exhibited (R)-enantioselective thioesterification activity toward 2-arylpropanoic acids. In the case of Japanese firefly luciferase from Luciola lateralis (LUC-H), the E-value for ketoprofen was approximately 20. In this study, we used a spectrophotometric method to measure the catalytic activity of LUC-H. Using this method allowed us to judge the reaction efficiency easily. Our results confirmed that LUC-H exhibits enantioselective thioesterification activity toward a series of 2-arylpropanoic acids. The highest activity was observed with ketoprofen. We also observed high enzymatic activity of LUC-H toward long-chain fatty acids. These results were reasonable because LUC-H is homologous with long-chain acyl-CoA synthetase. To obtain further information about the enantiodifferentiation mechanism of the LUC-H catalyzed thioesterification of ketoprofen, we determined the kinetic parameters of the reaction relative to each of its three substrates: ketoprofen, ATP, and coenzyme A (CoASH). We found that whereas the affinities of each compound are not affected by the chirality of ketoprofen, enantiodifferentiation is achieved by a chirality-dependent difference in the k_cat parameter.

Full text of DOI:10.1016/j.molcatb.2011.01.008

Journal of Molecular Catalysis B: Enzymatic 69 (2011) 140–146
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Enantiodifferentiation of ketoprofen by Japanese firefly luciferase
from Luciola lateralis
Dai-ichiro Kato∗, Tomohiro Tatsumi, Asami Bansho, Keisuke Teruya,
Hiromitsu Yoshida, Masahiro Takeo, Seiji Negoro
Department of Materials Science and Chemistry, Graduate School of Engineering, University of Hyogo, 2167 Shosha, Himeji, Hyogo 671-2280, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Recently, we found that firefly luciferase exhibited (R)-enantioselective thioesterification activity toward
2-arylpropanoic acids. In the case of Japanese firefly luciferase from Luciola lateralis (LUC-H), the E-value
for ketoprofen was approximately 20. In this study, we used a spectrophotometric method to measure
the catalytic activity of LUC-H. Using this method allowed us to judge the reaction efficiency easily. Our
results confirmed that LUC-H exhibits enantioselective thioesterification activity toward a series of 2-
arylpropanoic acids. The highest activity was observed with ketoprofen. We also observed high enzymatic
activity of LUC-H toward long-chain fatty acids. These results were reasonable because LUC-H is homolo-
gous with long-chain acyl-CoA synthetase. To obtain further information about the enantiodifferentiation
mechanism of the LUC-H catalyzed thioesterification of ketoprofen, we determined the kinetic param-
eters of the reaction relative to each of its three substrates: ketoprofen, ATP, and coenzyme A (CoASH).
We found that whereas the affinities of each compound are not affected by the chirality of ketoprofen,
enantiodifferentiation is achieved by a chirality-dependent difference in the kcat parameter.
© 2011 Elsevier B.V. All rights reserved.
Received 15 November 2010
Received in revised form 18 January 2011
Accepted 19 January 2011
Available online 2 February 2011
Keywords:
Firefly luciferase
Enantioselective thioesterification
2-Arylpropanoic acid
Kinetic resolution
Luciola lateralis
1. Introduction
tioselective thioesterification of 2-arylpropanoic acid [8,9]. Based
on the sequence homology and the similarity of the reaction mech-
Firefly luciferase catalyzes the oxidation of d-luciferin with
molecular oxygen in the presence of ATP and Mg2+, producing
light [1–3]. This bioluminescence reaction occurs in two steps:
adenylation of d-luciferin followed by oxidation. It has been shown
that firefly luciferase can also catalyze thioesterification. In the
presence of ATP, Mg2+, and coenzyme A (CoASH), this enzyme
converts the non-luminescent substrates l-luciferin and dehy-
droluciferin into the corresponding thioesters [4–6]. This catalytic
activity allows firefly luciferase to recover from binding a non-
luminescent substrate, resulting in an overall enhancement of the
rate of the bioluminescence reaction. In addition, Oba et al. reported
that firefly luciferases from Luciola cruciata and Photinus pyralis
exhibit acyl-CoA synthetase-like activity toward long-chain fatty
acids such as arachidonic acid [7]. These reports inspired us to
investigate the application of firefly luciferase to synthetic sub-
strates. We were intrigued by the resemblance between firefly
luciferases and long-chain acyl-CoA synthetase (LACS). LACS is
involved in a deracemization reaction and can catalyze the enan-
anisms between firefly luciferase and LACS [10], we predicted that
the thioesterification activity of firefly luciferase would display
enantioselectivity. We have demonstrated that firefly luciferases
from Luciola lateralis, Luciola cruciata, and Photinus pyralis cat-
alyze the thioesterification of a series of 2-arylpropanoic acids
in an enantioselective manner, and the (R)-form is transformed
more efficiently than the (S)-form (Fig. 1) [11]. Utilization of firefly
luciferase would, therefore, be a viable new option for the prepara-
tion of optically active 2-arylpropanoic acids. In the case of Japanese
firefly luciferase from L. lateralis (LUC-H), the E-value for ketopro-
fen was approximately 20. The thioesterification reaction occurs
in two steps via the formation of an acyl-AMP intermediate, and
LUC-H can distinguish the absolute configuration of ketoprofen in
either step. The detailed enantiodifferentiation mechanism, how-
ever, remained unknown.
In this study, we used a spectrophotometric method to ana-
lyze the substrate specificity of recombinant LUC-H; this technique
allowed us to judge the reaction efficiency easily. We also deter-
mined the kinetic parameters of the thioesterification reaction
relative to each of its three substrates: ketoprofen, ATP, and CoASH.
We compared the kinetic parameters obtained using each enan-
tiomer of ketoprofen separately. We also conducted inhibition
experiments using an acyl-AMP intermediate analogue, 5ꢀ-O-(N-
ketoprofenylsulfamoyl)adenosine. Based on these results, we have
Abbreviations: CoASH, coenzyme A; LACS, long-chain acyl-CoA synthetase; LUC-
H, Japanese firefly luciferase from Luciola lateralis; PPB, potassium phosphate buffer.
∗
Corresponding author. Tel.: +81 79 267 4969; fax: +81 79 267 4969.
E-mail address: kato@eng.u-hyogo.ac.jp (D.-i. Kato).
1381-1177/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2011.01.008

D.-i. Kato et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 140–146
141
O
O
CH
O
CH
3
3
Fast
CO H
COSCoA
2
(R)-acid
(S)-acid
(R)-thioester
LUC-H
ATP, Mg2+, CoASH
CH
O
CH
3
3
Slow
CO H
COSCoA
2
(S)-thioester
Fig. 1. LUC-H catalyzed enantioselective thioesterification of ketoprofen.
1H, adenine H-2); 13C NMR (DMSO-d6) ı: 25.58 (CH3), 27.37 (CH3),
68.57 (ribose C-5ꢀ), 81.64 (ribose C-3ꢀ), 83.81 (ribose C-2ꢀ), 84.10
(ribose C-4ꢀ), 89.59 (ribose C-1ꢀ), 114.07 (C), 119.56 (adenine C-5),
140.32 (adenine C-8), 149.20 (adenine C-4), 153.34 (adenine C-2),
156.62 (adenine C-6).
proposed a rationale for enantiodifferentiation in LUC-H-catalyzed
thioester formation.
2. Materials and methods
Ketoprofen N-hydroxysuccinimide ester (3): 1-Ethyl-3-(3-
dimethylaminopropyl)-carbodiimide, hydrochloride (0.929 g.
4.85 mmol) and N-hydroxysuccinimide (0.570 g, 4.96 mmol) was
dissolved in 1,4-dioxane (5 ml) under a nitrogen atmosphere. After
stirring for 10 min at room temperature, a solution of ketoprofen
(0.827 g, 3.25 mmol) in 1,4-dioxane (15 ml) was added dropwise
to the solution at room temperature. The mixture was stirred at
room temperature for 24 h. The reaction mixture was poured into
CHCl3 and washed successively with 1 N HCl, sat. NaHCO3 and sat.
NaCl, and dried over Na2SO4. The organic layer was evaporated
to dryness, and the residue was purified by silica gel column
chromatography (hexane/EtOAc/acetic acid, 4:6:1.2%) to give 3 as
an colorless oil (0.961 g, 84%): IR (KBr) 1811, 1782, 1736, 1647,
2.1. General materials
The pHLfLK plasmid encoding the LUC-H gene was a kind
gift from Kikkoman Corporation Ltd. [12]. (R)-Ketoprofen was
prepared by enantioselective decarboxylation of 2-methyl-2-(3-
benzoylphenyl)malonic acid using arylmalonate decarboxylases
from Mesorhizobium sp. BNC1 [13]. Other chemicals were
purchased from standard vendors and used without further
purification unless otherwise noted. Protein concentrations were
determined by a Bradford assay with bovine serum albumin as a
standard [14].
2.2. Synthesis of the ketoprofenyl-AMP intermediate analogue,
5ꢀ-O-(N-ketoprofenylsulfamoyl)adenosine
1597, 1283, 1202, 1119, 1063, 986, 964, 856, 714 and 702 cm−1
;
1H NMR (CDCl3) ı: 1.66 (d, 3H, J = 7.1, CH3), 2.81 (s, 4H, CH2CH2),
4.12 (q, 1H, J = 7.1, CH), 7.47–7.83 (m, 9H, 3-benzoylphenyl); 13C
NMR (CDCl3) ı: 18.91 (CH3), 25.62 (CH2CH2), 42.83 (C), 128.44,
129.00, 129.30, 129.69, 130.23, 131.61, 132.67, 137.28, 138.20,
138.53 (C6H5COC6H4), 169.53 (CO), 196.36 (CO).
Sulfamoyl chloride (1): Chlorosulfamoyl isocyanate (3.10 g,
21.9 mmol) was placed in a dry two-necked flask equipped with
a CaCl2 drying tube. Formic acid (1.00 g, 21.7 mmol) was added
dropwise via syringe to the flask with ice cooling. Vigorous reac-
tion occurred, and effervescence was observed. After the addition of
HCOOH, the mixture was stirred at room temperature until no more
gas was evolved. After stirring at room temperature for 1 h, color-
less crystals were accumulated in the flask. Benzene (10 ml) was
added to the reaction mixture, and insoluble material was removed
by filtration. The solution was evaporated to give 1 as colorless
crystals (1.86 g, 73%).
2ꢀ,3ꢀ-O-Isopropilidene-5ꢀ-O-(N-ketoprofenylsulfamoyl)adenosine
(4): A solution of 2 (0.692 g, 1.79 mmol) and 3 (0.943 g, 2.68 mmol)
in acetonitrile (50 ml) was cooled at 0 ◦C, and cesium carbonate
(0.705 g, 2.16 mmol) was added under a nitrogen atmosphere.
The mixture was stirred at 0 ◦C for 1 h, and the ice bath was
removed. The mixture was stirred at room temperature for 19 h.
The reaction mixture was evaporated to dryness, and the residue
was purified by open column chromatography on Chromatorex NH
to give 4 as a colorless crystals (0.781 g, 70%): Rf = 0.66 (Merk TLC,
EtOAc:2-butanone:EtOH:H2O = 5:3:1:1, anisaldehyde), Rf = 0.57
(NH-TLC, H2O:MeCN = 1:5): IR (KBr) 3341, 3202, 2936, 1649, 1576,
2ꢀ,3ꢀ-O-Isopropylidene-5ꢀ-O-sulfamoyladenosine (2): To a sus-
pension of NaH (0.174 g, 55% in paraffin, 3.94 mmol, washed
with n-hexane) in 1,4-dioxane (50 ml) was added 2ꢀ,3ꢀ-O-
isopropylideneadenosine (0.806 g, 2.62 mmol) at 0 ◦C under a
nitrogen atmosphere. After stirring for 30 min at 0 ◦C, a solution
of 1 (0.481 g, 4.16 mmol) in 1,4-dioxane (5 ml) was added drop-
wise to the suspension at 0 ◦C. The mixture was stirred at 0 ◦C
for 30 min, and ice bath was removed. The mixture was stirred
at room temperature for 24 h. Anhydrous methanol (20 ml) was
added to the reaction mixture, and the resulting mixture was fil-
tered through Celite pad. The filtrate was evaporated to dryness,
and the residue was purified by silica gel column chromatography
(CHCl3/MeOH, 9:1) to give 2 as a colorless crystals (0.87 g, 86%): IR
(KBr) 3354, 3188, 2943, 1651, 1601, 1576, 1373, 1182, 1074, 858,
1475, 1375, 1288, 1148, 1076, 1001, 845, 777, 721 and 644 cm−1
;
1H NMR (DMSO-d6) ı: 1.24 (s, 1.5H, CH3, diastereomer), 1.25 (s,
1.5H, CH3, diastereomer), 1.27 (d, 1.5H, J = 7.0, CH3, diastereomer),
1.28 (d, 1.5H, J = 7.0, CH3, diastereomer), 1.49 (s, 3H, CH3), 3.49
(q, 1H, J = 7.0, CH), 3.89 (m, 2H, ribose H-5ꢀ), 4.32 (m, 1H, ribose
H-4ꢀ), 4.90 (m, 1H, ribose H-3ꢀ), 5.25 (dd, 0.5H, J = 6.0 and 3.0,
ribose H-2ꢀ, diastereomer), 5.28 (dd, 0.5H, J = 6.0 and 3.0, ribose
H-2ꢀ, diastereomer), 6.10 (d, 1H, J = 3.0, ribose H-1ꢀ), 7.33 (br s,
2H, adenine NH2), 7.38–7.72 (m, 9H, 3-benzoylphenyl), 8.12(s,
0.5H, adenine H-8, diastereomer), 8.13 (s, 0.5H, adenine H-8,
diastereomer), 8.35(s, 0.5H, adenine H-2, diastereomer), 8.36 (s,
0.5H, adenine H-2, diastereomer); 13C NMR (DMSO-d6) ı: 19.67
(CH3), 25.55 (CH3), 27.45 (CH3), 49.22 (C), 67.60 (ribose C-5ꢀ),
82.03 (ribose C-3ꢀ), 83.96 (ribose C-2ꢀ), 84.24 (ribose C-4ꢀ), 89.71
(ribose C-1ꢀ), 113.62 (C), 119.30 (adenine C-5), 128.01, 128.61,
129.00, 129.27, 130.14, 132.50, 133.04, 137.03, 137.56, 145.02
799 and 556 cm−1 1H NMR (DMSO-d6) ı: 1.34 (s, 3H, CH3), 1.55 (s,
;
3H, CH3), 4.13 (dd, 1H, J = 10.5 and 6.3 Hz, ribose H-5ꢀa), 4.24 (dd, 1H,
J = 10.5 and 5.3 Hz, ribose H-5ꢀb), 4.40 (m, 1H, ribose H-4ꢀ), 5.08 (dd,
1H, J = 6.3 and 3.3 Hz, ribose H-3ꢀ), 5.43 (dd, 1H, J = 6.3 and 2.3 Hz,
ribose H-2ꢀ), 6.23 (d, 1H, J = 2.3, ribose H-1ꢀ), 7.38 (br s, 2H, adenine
NH2), 7.61 (br s, 2H, SO2NH2), 8.17 (s, 1H, adenine H-8), 8.30 (s,

142
D.-i. Kato et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 140–146
oxidation at 340 nm (6220 M−1 cm−1) with a spectrophotometer
(U-2800A, Hitachi) [15–17]. The standard reaction mixture for this
assay contained 100 mM PPB (pH 7.0), 10 mM ATP, 10 mM MgCl2,
0.25 mM (R)-ketoprofen, 2 mM CoASH, 1 mM phosphoenolpyruvic
acid, 0.2 mM NADH, 40 mg/ml adenylate kinase, 20 mg/ml pyru-
vate kinase, 20 mg/ml lactate dehydrogenase, and 100 mg of LUC-H.
The total volume was brought to 500 ␮l. The mixture containing
all components except LUC-H was incubated at room temperature
(27 2 ◦C) for 3 min. The reaction was then initiated by addition of
the enzyme.
(C6H5COC6H4), 140.00 (adenine C-8), 149.49 (adenine C-4), 153.28
(adenine C-2), 156.56 (adenine C-6), 179.05 (CO), 196.35 (CO).
5ꢀ-O-(N-Ketoprofenylsulfamoyl)adenosine (5): Compound
4
(0.108 g, 0.173 mmol) was dissolved in anhydrous methanol
(3.8 ml). Conc. HCl (0.2 ml) was added to the solution, and the
mixture was stirred for 1 h at 60 ◦C. The pH of the reaction mixture
was adjusted to 8.5 by adding 28% NH3 aq at 0 ◦C. The resulting
mixture was evaporated to dryness, and the residue was purified
by open column chromatography on Chromatorex NH to give 5 as a
white solid (0.101 g). By comparison with the 1H NMR integration
value of internal standard, 3-trimethylsilyl-1-propanesulfonic
acid, the purity was determined as 46%: Rf = 0.43 (Merk TLC,
EtOAc:2-butanone:EtOH:H2O = 5:3:1:1, anisaldehyde), Rf = 0.29
(NH-TLC, H2O:MeCN = 1:5): [M−H]−1 observed 581.1453, pre-
dicted 581.1455; IR (KBr) 3352, 3211, 2974, 1651, 1599, 1479,
2.5. Kinetic analysis and determination of Ki value
Thekinetic parameters were determinedbyspectrophotometry.
The kcat and Km values were evaluated by Michaelis–Menten anal-
ysis using GraphPad Prism, version 5.01 (GraphPad, USA). The kcat
values were expressed as the turnover numbers per subunit (Mr of
the subunit, 60,163). The kinetic studies were performed with vary-
ing concentrations of the test substrate while the concentrations of
all other components were fixed. The default concentrations of the
three cofactors, ketoprofen, ATP, and CoASH, were 0.25 mM, 10 mM,
and 2 mM respectively. The concentrations of substrate for mea-
suring each parameter were 0.05–1 mM ((R)- or (S)-ketoprofen),
0.1–2.5 mM (ATP), and 0.1–3 mM (CoASH) respectively. Each assay
was repeated five times.
Inhibition studies were performed with a ketoprofenyl-AMP
intermediate analogue. Concentrations of the analogue were 0.05
and 0.1 mM. Each set of assays was performed five times. The Ki
value was determined by fitting initial velocity data obtained by the
spectrophotometric method described above with the appropriate
rate equation in GraphPad Prism.
1288, 1148, 999, 849, 777, 721, 644 and cm−1 1H NMR (DMSO-d6)
;
ı: 1.27 (d, 1.5H, CH3, diastereomer), 1.28 (d, 1.5H, CH3, diastere-
omer), 3.49 (m, 1H, diastereomer), 3.92 (m, 1H, ribose H-5ꢀa),
3.98 (m, 1H, ribose H-5ꢀb), 4.03 (m, 1H, ribose H-4ꢀ), 4.10 (m, 1H,
ribose H-3ꢀ), 4.57 (m, 1H, ribose H-2ꢀ), 5.29 (d, 1H, OH), 5.47 (d,
1H, OH), 5.88 (d, 1H, ribose H-1ꢀ), 7.28 (br s, 2H, adenine NH2),
7.40–7.73 (m, 9H, 3-benzoylphenyl), 8.12 (s, 1H, adenine H-8),
8.38 (s, 1H, adenine H-2); 1H NMR (D2O) ı: 1.40 (d, 1H, CH3),
4.04–4.39 (m, 5H, ribose H-5ꢀ, H-4ꢀ, H-3ꢀ, H-2ꢀ), 5.81 (d, 0.5H, ribose
H-1ꢀ, diastereomer), 5.78 (d, 0.5H, ribose H-1ꢀ, diastereomer),
7.14–7.59 (m, 11H, 3-benzoylphenyl and adenine NH2), 7.96 (s,
0.5H, diastereomer), 7.97 (s, 0.5H, diastereomer), 8.02 (s, 0.5H,
diastereomer), 8.04 (s, 0.5H, diastereomer); 13C NMR (DMSO-d6)
ı: 19.86 (CH3), 49.17 (C), 67.74 (ribose C-5ꢀ), 71.28 (ribose C-3ꢀ),
74.10 (ribose C-2ꢀ), 83.02 (ribose C-4ꢀ), 87.52 (ribose C-1ꢀ), 119.38
(adenine C-5), 128.02, 128.67, 129.04, 129.33, 130.15, 132.56,
133.07, 137.04, 137.56, 145.08 (C6H5COC6H4), 139.93 (adenine
C-8), 150.05 (adenine C-4), 153.17 (adenine C-2), 156.47 (adenine
C-6), 179.00 (CO), 196.45 (CO).
3. Results and discussion
2.3. Plasmid construction and expression of LUC-H protein
3.1. Substrate specificity of recombinant LUC-H
The LUC-H gene was amplified from pHLfLK by PCR with for-
ward (TTTAACATATGGAAAACATGGAGAACG) and reverse (TGAT-
TAAACTCTAGATTGACATTTACATC) primers. These primers were
designed with Nde I and Xba I sites (italicized) at the 5ꢀ and 3ꢀ ends
of the LUC-H gene, respectively. The underlined ATG represents
the start codon. The amplified 1.7 kbp fragment was subcloned to
a TA vector, which was prepared from pXCmkn12 digested with
Xcm I (National Institute of Genetics, Japan). The subcloned vector
was first digested with Nde I and Xba I, then purified and inserted
into a pCold I expression vector (Takara, Japan). The resultant plas-
mid, pColdI-LUC-H, was used to express the recombinant LUC-H
protein. Escherichia coli BL21(DE3) cells were transformed with
pColdI-LUC-H, cultured, and induced to express LUC-H according to
manufacturer’s instructions. After production of the recombinant
enzyme, the cells were collected, disrupted by sonication (20 kHz,
30 s ×10 times) in 5 ml of 50 mM potassium phosphate buffer (PPB)
(pH 7.0) containing 300 mM NaCl, and centrifuged (14,500 × g for
10 min, 4 ◦C). The enzyme was purified from the supernatant using
TALON Metal Affinity Resin (2 ml) (Clontech) according to manu-
facturer’s instructions and identified by SDS–PAGE analysis. The
fractions containing LUC-H were combined and dialyzed overnight
against 100 mM Tris–HCl buffer (pH 7.5).
The specific activity of recombinant LUC-H was measured using
a spectrophotometric method (Table 1). LUC-H exhibited thioester-
ification activity toward a series of 2-arylpropanoic acids, including
flurbiprofen, ibuprofen, ketoprofen, and naproxen. Among these 2-
arylpropanoic acids, ketoprofen was the best substrate. When the
concentration of ( )-ketoprofen was 0.25 mM, the specific activ-
ity of LUC-H was 20.5 nmol/min/mg, which was comparable to the
previously reported value obtained with an HPLC assay [11]. The
same spectrophotometric method was used to analyze substrate
specificity. 2-Phenylbutanoic acid was accepted by LUC-H, although
the specific activity was very low. When 2-phenylpentanoic acid,
2-phenyl-3-methylbutanoic acid, 2-(4-chlorophenoxy)propanoic
acid, and 2-methyl-3-phenylpropanoic acid were used, the cat-
alytic activity of LUC-H was below the limit of detection of the
assay. Fatty acids with various chain lengths were also tested with
this assay. As shown in Table 1, LUC-H exhibited thioesterification
activity only toward fatty acids with 8–18 carbons and the high-
est activity was detected with dodecanoic acid (C12). These results
indicate that the substrate specificity of LUC-H toward fatty acids is
similar to that of the LACS family. 6-Mercaptohexanoic acid and 11-
mercaptoundecanoic acid, both of which have a thiol group at the
␻-position, were also converted to the corresponding coenzyme A
thioesters. 6-Aminohexanoic acid was not converted to a coenzyme
A thioester; however, the catalytic activity of LUC-H recovered
when the ␻-amino group was protected by a tert-butoxycarbonyl
(Boc) or benzyloxycarbonyl (Cbz) group. This result indicates that
the identity of substituents at the ␻-position of the fatty acid
affects the substrate accessibility of this enzyme. Benzoic acid, 2-
anthracenecarboxylic acid, anthraquinone-2-carboxylic acid, and
2.4. Spectrophotometric assay for LUC-H activity
The activity of LUC-H was determined by UV–vis spectropho-
tometry. In this assay, we measured the initial rate of AMP
formation by coupling the reaction of LUC-H with adenylate kinase,
pyruvate kinase, and lactate dehydrogenase and following NADH

D.-i. Kato et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 140–146
143
Table 1
Substrate specificity of the LUC-H catalyzed thioesterification reaction.
Carboxylic acida Relative activityb (%)
(
(
(
(
(
(
(
(
(
)-Flurbiprofen
)-Ibuprofen
)-Ketoprofen
)-Naproxen
)-Phenylbutanoic acid
)-2-Phenylpentanoic acid
)-2-Phenyl-3-methylbutanoic acid
)-2-(4-Chlorophenoxy)propanoic acid
)-2-Methyl-3-phenylpropanoic acid
10
8
46
15
4
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
7
93
100
64
3
3
Acetic acid (C2)
Propanoic acid (C3)
Butanoic acid (C4)
Hexanoic acid (C6)
Octanoic acid (C8)
Decanoic acid (C10)
Dodecanoic acid (C12)
Butyldecanoic acid (C14)c
Hexadecanoic acid (C16)c
Octadecanoic acid (C18)c
6-Mercaptohexanoic acid
11-Mercaptoundecanoic acid
6-Aminohexanoic acid (Ahx)
Boc-Ahx
16
99
N.D.
58
Cbz-Ahx
23
Benzoic acid
32
2-Anthracenecarboxylic acid
Anthroquinone-2-carboxylic acid
1-Pyrenecarboxylic acid
Ferrocenecarboxylic acid
Phenylglycine
12
49
13
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
Mandelic acid
Dihydrocinnamic acid
trans-Cinnamic acid
Malonic acid
Retionic acid
␣-Lipoic acid
N.D. = not detected; the relative activity could not be calculated because the specific
activity was below the detection limit of the assay.
a
Final concentration of carboxylic acid was 0.25 mM.
The relative activity was calculated by comparison with the specific activity,
b
using the value for dodecanoic acid (C12) (44.6 nmol/min/mg) as a standard. The spe-
cific activity of thioester formation was determined using the spectrophotometric
assay.
c
0.2% Triton X-100 was added to the reaction mixture for this assay to ensure
complete dissolution of the substrate.
1-pyrenecarboxylic acid were also converted to the corresponding
thioesters in moderate efficiency. These results imply that LUC-H
can accept bulky substitutions at the ␣-position of the carboxyl
group. Ferrocenecarboxylic acid, phenylglycine, mandelic acid,
dihydrocinnamic acid, trans-cinnamic acid, malonic acid, retionic
acid, and ␣-lipoic acid were not accepted by this enzyme. It is
interesting that ␣-lipoic acid was not accepted as a thioesterifica-
tion substrate because Niwa et al. concluded that ␣-lipoic acid acts
as a competitive inhibitor of the bioluminescence reaction rather
than being converted into a thioester [18]. Our result supports their
conclusion and adds to knowledge of the inhibitory pattern of this
compound.
3.2. Determination of kinetic parameters of LUC-H catalyzed
enantioselective thioesterification of ketoprofen
To obtain information about the enantiodifferentiation mech-
anism of the LUC-H catalyzed thioesterification reaction, we
determined its kinetic parameters using ketoprofen as the car-
boxylic acid substrate. It has been proposed that the firefly
luciferase catalyzed thioesterification reaction occurs in two steps
(Fig. 2) [7,11]. First, the carboxylic acid substrate couples with
ATP to generate an acyl-AMP intermediate. Then CoASH attacks

144
D.-i. Kato et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 140–146
NH
2
N
N
N
ATP, Mg2+
O
N
O
CH
3
O
CH
3
P O
OH
O
O
CO H
2
O
OH OH
PPi
ketoprofen
ketoprofenyl-AMP
CoASH
O
CH
3
COSCoA
AMP
(R)-thioester
Fig. 2. Chemical conversion process of the firefly luciferase catalyzed thioesterification reaction.
this intermediate, releasing the thioester product. LUC-H can dis-
tinguish the absolute configuration of ketoprofen in either of
these steps, and the (R)-acid is preferentially converted to the
thioester about 20 times faster than the (S)-antipode [11]. Three
substrates, ketoprofen, ATP, and CoASH, take part in this trans-
formation process. Thus the affinity and/or catalytic rate of these
compounds must be different when ketoprofen enantiomer is sub-
jected separately to the reaction. We therefore determined the
kinetic parameters of each of these three compounds (Table 2). The
Km values of the ketoprofen enantiomers show that the affinity
of (S)-ketoprofen, the less reactive substrate, is two fold less than
that of (R)-ketoprofen, the more reactive substrate. The affinities
NH
2
N
N
N
2',3'-O-isopropyli-
O
H2N S
deneadenosine
3
O
N
Cl
H N S
O
2
O
Cs2CO3 / MeCN
NaH / 1,4-dioxane
O
O
O
O
1
86%
70%
2
NH
2
N
N
N
O
S O
O
N
O
CH
3
H
N
conc. HCl / MeOH
46%
O
O
O
O
4
NH
2
N
N
N
O
S O
O
N
O
CH
3
H
N
O
O
OH OH
5'-O-(N-ketoprofenylsulfamoyl)adenosine (5)
Fig. 3. Synthesis of the ketoprofenyl-AMP intermediate analogue, 5ꢀ-O-(N-ketoprofenylsulfamoyl)adenosine (5).

D.-i. Kato et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 140–146
145
isopropilidene-5ꢀ-O-(N-ketoprofenylsulfamoyl)adenosine (4) was
obtained in good yield and the isopropylidene moiety was easily
deprotected under acidic condition to obtain target analogue (5).
In the presence of this analogue, we measured the initial rate of the
thioesterification reaction of (R)- or (S)-ketoprofen and compared
the patterns of inhibition. We found that the analogue compound
behaved as a competitive inhibitor with Ki values of 0.059 and
0.064 mM for (R)- and (S)-ketoprofen, respectively (Fig. 4). These
results indicate that both enantiomers of ketoprofen could bind
to the active site of LUC-H and be converted to the acyl-AMP
intermediate with similar reaction efficiency. The results of the
kinetic parameter determination demonstrated that the affinity of
all components of the reaction is independent of the chirality of
ketoprofen. We therefore concluded that the first adenylation step
is not enantioselective, and the enantiodifferentiation is achieved
in the second thioester-forming step.
Our results are consistent with the action of firefly luciferase on
the luciferin substrate, which behaves differently after the forma-
tion of the luciferyl-AMP intermediate, depending on the chirality
of luciferin [4–6]. When d-luciferin is the substrate, the oxidation
reaction occurs and produces bioluminescence, whereas when l-
luciferin is the substrate, the thioesterification reaction proceeds
in the presence of CoASH. A domain movement mechanism for
acyladenylate-forming enzymes has been proposed, based on avail-
able structural and functional information [20–23]. In the case of
the firefly luciferase-catalyzed bioluminescent reaction, there is
some evidence that movement of the C-terminal domain is nec-
essary for light production [24,25]. Therefore, in the case of the
thioesterification reaction of ketoprofen, it is not inconsistent that a
similar movement would occur after the formation of ketoprofenyl-
AMP intermediate. Different movement of the C-terminal domain
depending on substrate chirality may cause the difference in the
kcat value that results in enantiodifferentiation by this enzyme.
4. Conclusions
Fig. 4. Inhibition of LUC-H by a ketoprofenyl-AMP intermediate analogue with
respect to (R)-ketoprofen (A) and (S)-ketoprofen (B). The concentrations of the
analogue compound used were 0 (circle), 0.05 mM (triangle), 0.1 mM (square),
respectively.
We have analyzed the substrate specificity of recombinant LUC-
H using a spectrophotometric assay system to easily judge the
reaction efficiency. We also determined the kinetic parameters
of the enantioselective thioesterification reaction of ketoprofen.
Based on the kinetic parameters, we concluded that a difference
in kcat between (R)-ketoprofen and (S)-ketoprofen contributes sig-
nificantly to the enantiodifferentiation event. The results of our
inhibition experiment suggest that the enantiodifferentiation event
occurs after the formation of the acyl-AMP intermediate.
of the ATP and CoASH cofactors for LUC-H are independent of the
absolute configuration of the ketoprofen substrate. These results
suggest that all components of the reaction have an equal chance
to bind to the active site of enzyme with respect to the enantiomers
of the ketoprofen substrate before the enantiodifferentiation pro-
cess occurs. In contrast, the kcat values show an obvious difference
between the optical isomers of ketoprofen. The value of kcat for (R)-
ketoprofen is approximately 100 times larger than that observed
for (S)-ketoprofen. Because of this difference, the product from the
(R)-enantiomer dominates the reaction outcome.
Acknowledgments
We are deeply grateful to Dr. Hiromichi Ohta (University of
Nagasaki) for comments and for discussing this manuscript. We
extend special thanks to professor Jun Hiratake (Institute for Chem-
ical Research, Kyoto University) for technical assistance in the
synthesis of the ketoprofenyl-AMP intermediate analogue. This
work was partially supported by a Grant-in-Aid for Young Scientists
(B) (19750144) from the Japan Society for the Promotion of Science
(JSPS) and by the Hyogo Science and Technology Association.
3.3. Inhibitory effect of an acyl-AMP intermediate analogue
The kinetic parameters indicated that there are no differences
in the affinities for all substrates regardless of the chirality of
ketoprofen. Next, we focused our attention on the acyl-AMP inter-
mediate and investigated the inhibitory effect of an analogue
on the enzyme. We used the N-acyl sulfamate compound 5ꢀ-O-
(N-ketoprofenylsulfamoyl)adenosine (5), a stable analogue of the
intermediate ketoprofenyl-AMP (Fig. 3). We expected this com-
pound to function as an inhibitor of thioester formation because a
similar compound was used for enzyme-analogue structural analy-
sis of L. cruciata firefly luciferase [19]. The synthesis was undertaken
with commercially available 2ꢀ,3ꢀ-O-isopropylideneadenosine and
this compound was converted into 2ꢀ,3ꢀ-O-isopropylidene-5ꢀ-O-
sulfamoyladenosine (2) by reaction with sulfamoyl chloride (1).
By cesium carbonate mediated condensation of 2 and 3, 2ꢀ,3ꢀ-O-
References
[1] W.D. McElroy, A. Green, Arch. Biochem. Biophys. 64 (1956) 257–271.
[2] W.C. Rhodes, W.D. McElroy, Science 128 (1958) 253–254.
[3] W.C. Rhodes, W.D. McElroy, J. Biol. Chem. 233 (1958) 1528–1537.
[4] R.L. Airth, W.C. Rhodes, W.D. McElroy, Biochim. Biophys. Acta 27 (1958)
519–532.
[5] H. Fraga, D. Fernandes, R. Fontes, J.C.G. Esteves de Silva, FEBS J. 272 (2005)
5206–5216.
[6] M. Nakamura, S. Maki, Y. Amano, Y. Ohkita, K. Niwa, T. Hirano, Y. Ohmiya, H.
Niwa, Biochem. Biophys. Res. Commun. 331 (2005) 471–475.

146
D.-i. Kato et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 140–146
[7] Y. Oba, M. Ojika, S. Inouye, FEBS Lett. 540 (2003) 251–254.
[8] R.D. Knihinicki, K.M. Williams, R.O. Day, Biochem. Pharmacol. 38 (1989)
4389–4395.
[17] Y. Hisanaga, H. Ago, N. Nakagawa, K. Hamada, K. Ida, M. Yamamoto, T. Hori,
Y. Arii, M. Sugahara, S. Kuramitsu, S. Yokoyama, M. Miyano, J. Biol. Chem. 279
(2004) 31717–31726.
[9] S.D. Hall, Q. Xiaotao, Chem. Biol. Interact. 90 (1994) 235–251.
[10] H. Suzuki, Y. Kawarabayashi, J. Kondo, T. Abe, K. Nishikawa, S. Kimura, T.
Hashimoto, T. Yamamoto, J. Biol. Chem. 256 (1990) 8681–8685.
[11] D. Kato, K. Teruya, H. Yoshida, M. Takeo, S. Negoro, H. Ohta, FEBS J. 274 (2007)
3877–3885.
[12] N. Hattori, N. Kajiyama, M. Maeda, S. Murakami, Biosci. Biotech. Biochem. 66
(2002) 2587–2593.
[13] K. Okrasa, C. Levy, B. Hauer, N. Baudendistel, D. Leys, J. Micklefield, Chem. Eur.
J. 14 (2008) 6609–6613.
[18] K. Niwa, Y. Ohmiya, Biochem. Biophys. Res. Commun. 323 (2004) 625–
629.
[19] T. Nakatsu, S. Ichiyama, J. Hiratake, A. Saldanha, N. Kobashi, K. Sakata, H. Kato,
Nature 440 (2005) 372–376.
[20] A.M. Gulick, V.J. Starai, A.R. Horswill, K.M. Homick, J.C. Escalante-Semerena,
Biochemistry 42 (2003) 2866–2873.
[21] G. Jogl, L. Tong, Biochemistry 43 (2004) 1425–1431.
[22] A.S. Reger, J.M. Carney, A.M. Gulick, Biochemistry 46 (2007) 6536–6546.
[23] A.M. Gulick, ACS Chem. Biol. 4 (2009) 811–827.
[14] M.M. Bradford, Anal. Biochem. 78 (1976) 248–254.
[15] B.L. Horecker, A. Kornberg, J. Biol. Chem. 175 (1948) 385–390.
[16] M.-X. Wu, K.A.W. Hill, Anal. Biochem. 211 (1993) 320–323.
[24] B.R. Branchini, T.L. Southworth, M.H. Murtiashaw, S.R. Wilkinson, N.F. Khattak,
J.C. Rosenberg, M. Zimmer, Biochemistry 44 (2005) 1385–1393.
[25] A. Szarecka, Y. Xu, P. Tang, Biophys. J. 93 (2007) 1895–1905.