Synthesis of 3-cyclohexylpropyl caffeate from 5-caffeoylquinic acid with consecutive enzymatic conversions in ionic liquid

Article abstract of DOI:10.1016/j.molcatb.2011.01.012

We developed a convenient one-pot procedure for conversion of 5-caffeoylquinic acid to 3-cyclohexylpropyl caffeate, which exhibits an antiproliferative effect toward various human tumor cells. The procedure was comprised of two consecutive reactions by chlorogenate hydrolase (EC 3.1.1.42) from Aspergillus japonicus and Candida antarctica lipase B, and was performed using an ionic liquid, 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl) imide, as the reaction solvent. When various caffeoylquinic acids from coffee beans, namely, 3-caffeoylquinic acid, 4-caffeoylquinic acid, 5-caffeoylquinic acid, 3,5-dicaffeoylquinic acid, and 4,5-dicaffeoylquinic acid were used, the first alcoholysis reaction with methanol using chlorogenate hydrolase produced methyl caffeate with conversion yields of 60.0%, 61.3%, 86.0%, 92.7%, and 114.0%, respectively, to each individual substrate. Two caffeoyl groups of dicaffeoylquinic acids would be used for the synthesis of methyl caffeate. In the subsequent transesterification reaction by C. antarctica lipase B with 3-cyclohexyl-1-propanol, the methyl caffeate produced was converted to 3-cyclohexylpropyl caffeate under reduced pressure to remove the by-product methanol. In the one-pot synthesis, the methyl caffeate was transesterified efficiently to 3-cyclohexylpropyl caffeate by C. antarctica lipase B with deactivation of chlorogenate hydrolase by taking advantage of the difference between the optimum temperatures for the two enzymes. This system provided 12.8 mM 3-cyclohexylpropyl caffeate from 15 mM 5-caffeoylquinic acid with conversion yield of 85.3%.

Full text of DOI:10.1016/j.molcatb.2011.01.012

Journal of Molecular Catalysis B: Enzymatic 69 (2011) 161–167
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Synthesis of 3-cyclohexylpropyl caffeate from 5-caffeoylquinic acid with
consecutive enzymatic conversions in ionic liquid
Atsushi Kurataa,∗, Shintaro Takemotoa, Tokio Fujitaa, Kazuya Iwaib,
Mina Furusawab, Noriaki Kishimotoa,∗∗
a Department of Applied Biological Chemistry, Kinki University, 3327-204 Nakamachi, Nara City, Nara 631-8505, Japan
b R&D Center, UCC Ueshima Coffee Co. Ltd., 3-1-4 Zushi, Takatsuki-shi, Osaka 569-0036, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We developed
a convenient one-pot procedure for conversion of 5-caffeoylquinic acid to 3-
Received 26 October 2010
Received in revised form 25 January 2011
Accepted 25 January 2011
Available online 1 February 2011
cyclohexylpropyl caffeate, which exhibits an antiproliferative effect toward various human tumor cells.
The procedure was comprised of two consecutive reactions by chlorogenate hydrolase (EC 3.1.1.42)
from Aspergillus japonicus and Candida antarctica lipase B, and was performed using an ionic liq-
uid, 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide, as the reaction solvent. When
various caffeoylquinic acids from coffee beans, namely, 3-caffeoylquinic acid, 4-caffeoylquinic acid, 5-
caffeoylquinic acid, 3,5-dicaffeoylquinic acid, and 4,5-dicaffeoylquinic acid were used, the first alcoholysis
reaction with methanol using chlorogenate hydrolase produced methyl caffeate with conversion yields of
60.0%, 61.3%, 86.0%, 92.7%, and 114.0%, respectively, to each individual substrate. Two caffeoyl groups of
dicaffeoylquinic acids would be used for the synthesis of methyl caffeate. In the subsequent transesterifi-
cation reaction by C. antarctica lipase B with 3-cyclohexyl-1-propanol, the methyl caffeate produced was
converted to 3-cyclohexylpropyl caffeate under reduced pressure to remove the by-product methanol.
In the one-pot synthesis, the methyl caffeate was transesterified efficiently to 3-cyclohexylpropyl caf-
feate by C. antarctica lipase B with deactivation of chlorogenate hydrolase by taking advantage of the
difference between the optimum temperatures for the two enzymes. This system provided 12.8 mM
3-cyclohexylpropyl caffeate from 15 mM 5-caffeoylquinic acid with conversion yield of 85.3%.
© 2011 Elsevier B.V. All rights reserved.
Keywords:
Caffeic acid phenethyl ester analogue
Caffeoylquinic acid
Ionic liquid
Consecutive enzymatic conversions
1. Introduction
Immature green coffee beans are not marketed as coffee because
contamination of these beans negatively affects the flavor. How-
Caffeic acid esters are widely distributed in plants and propo-
lis [1,2]. Caffeic acid phenethyl ester (CAPE) especially has been
found in propolis and has a broad spectrum of biological activi-
ties, including antimicrobial, anti-inflammatory, antioxidant, and
antitumor activities [3]; it also has an inhibitory effect on HIV-1
integrase, cyclooxygenase, and lipooxygenase [4–7]. It has been
reported that the ester part of CAPE is important for the antipro-
liferative effect on various human tumor cells [8]. Additionally, it
was suggested that conversion of the phenyl group to a cyclohexyl
group in a CAPE analogue enhanced the antiproliferative effect [9].
Recently, we showed that 3-cyclohexylpropyl caffeate, one of the
CAPE analogues, exhibits a strong antiproliferative activity toward
various tumor cells that is comparable to that of 5-fluorouracil [10].
ever, they contain appreciable amounts of various caffeoylquinic
acids, for example, 4.8–5.8 g of 5-caffeoylquinic acid/100 g of
immature green coffee beans [11]. These immature beans are
notable among unused agricultural resources, and we are there-
fore currently investigating the enzymatic conversion of their
caffeoylquinic acids to valuable products. In recent work, we syn-
thesized CAPE using 5-caffeoylquinic acid and 2-phenylethanol as
substrates with chlorogenate hydrolase from Aspergillus japonicus
by a transesterification reaction in a biphasic aqueous-alcohol state
and elucidated the antibacterial, antimutagenic, and anti-influenza
virus activities of CAPE [12]. The procedure using chlorogenate
hydrolase provided various CAPE analogues, but the maximum con-
version yield of CAPE was 50%. The insufficient yield was probably
due to the hydrolysis of 5-caffeoylquinic acid by the enzyme to caf-
feic acid in the aqueous phase. Therefore, a new procedure for the
synthesis of CAPE analogues superior to that method in terms of
the conversion yield remained to be developed.
∗
Corresponding author. Tel.: +81 742 43 7028; fax: +81 742 43 1445.
Corresponding author. Tel.: +81 742 43 7338; fax: +81 742 43 1445.
E-mail addresses: kurata090401@nara.kindai.ac.jp (A. Kurata),
Ionic liquids (ILs), which are composed of a bulky asymmet-
ric cation and a small anion, are easily modified with respect to
the combination of cation and anion, and therefore, numerous IL
∗∗
kisimoto@nara.kindai.ac.jp (N. Kishimoto).
1381-1177/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2011.01.012

162
A. Kurata et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 161–167
1H NMR analysis according to reference [16]. 3-Cyclohexyl-
1-propanol (Tokyo Kasei, Tokyo, Japan) was purchased.
1-Butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide
([BMIM][NTf2]), 1-butyl-3-methylimidazolium tetrafluoroborate
([BMIM][BF4]), 1-butyl-3-methylimidazolium trifluoromethane-
sulfonate ([BMIM][CF3SO3]), N-methyl-N-propylpiperidinium
compositions are possible [13]. Unlike conventional organic sol-
vents used for biocatalytic reactions, ILs are able to dissolve many
compounds, have a wide temperature range for the liquid phase,
and possess no vapor pressures. Thus, ILs have good properties
for use as reaction solvents, and extensive studies of enzymatic
synthesis using ILs have been carried out [14,15]. We previously
developed an efficient procedure for conversion of methyl caf-
feate to produce various CAPE analogues with Candida antarctica
bis(trifluoromethylsulfonyl)imide
([MPPip][NTf2]),
and
N-
methyl-N-propylpyrrolidinium bis(trifluoromethylsulfonyl)imide
([MPPro][NTf2]) were purchased from Kanto Kagaku (Tokyo,
Japan). All other reagents were purchased from Merck (Darmstadt,
Germany), Sigma–Aldrich Japan (Tokyo, Japan), and Nacalai Tesque
(Kyoto, Japan).
lipase
B using an ionic liquid, 1-butyl-3-methylimidazolium
bis(trifluoromethylsulfonyl)imide ([BMIM][NTf2]), as a solvent
[10].
In this study, we found that chlorogenate hydrolase from
A. japonicus efficiently catalyzed the alcoholysis reaction of
caffeoylquinic acids purified from coffee beans, namely, 3-
caffeoylquinic acid, 4-caffeoylquinic acid, 5-caffeoylquinic acid,
3,5-dicaffeoylquinic acid, 3,4-dicaffeoylquinic acid, and 4,5-
dicaffeoylquinic acid, with methanol to produce methyl caffeate
in an IL, [BMIM][NTf2], as the solvent. By using two consecutive
reactions by chlorogenate hydrolase and C. antarctica lipase B in
[BMIM][NTf2] as the solvent, we developed a convenient one-pot
procedure for an enzymatic synthesis of 3-cyclohexylpropyl caf-
feate from 5-caffeoylquinic acid (Fig. 1). Firstly, methyl caffeate
(compound 2) was prepared from 5-caffeoylquinic acid (compound
1) and methanol using chlorogenate hydrolase with the IL. Then,
the unreacted methanol was removed in vacuo (14 hPa) at 80 ◦C for
1 h, and 3-cyclohexylpropyl caffeate (compound 4) was obtained
using C. antarctica lipase B with methyl caffeate and 3-cyclohexyl-
1-propanol (compound 3) as the substrates. In order to accelerate
the reaction equilibrium to give the desired product, we performed
the C. antarctica lipase B-catalyzed reaction under reduced pressure
(845 hPa) to remove the by-product methanol from the reaction
mixture. Additionally, to take advantage of the different optimum
temperatures for the two enzymes in the one-pot reaction, the
conversion of methyl caffeate to 3-cyclohexylpropyl caffeate was
performed by C. antarctica lipase B with deactivation of chloro-
genate hydrolase. The one-pot two-step method in the IL is a
convenient economical preparation of 3-cyclohexylpropyl caffeate
with good yield.
2.2. Immobilization of chlorogenate hydrolase
Quaternary ammonium sepabeads (SEPABEADS EC-QA) was
donated by Mitsubishi Chemical Co. (Tokyo, Japan). Chlorogenate
hydrolase (0.1 g) and quaternary ammonium sepabeads (1.0 g)
were dissolved in 4 ml of 20 mM sodium phosphate buffer (pH 6.5),
and the mixture were incubated with stirring at 300 rpm using
an Invitro shaker Mix-VR (TAITEC, Tokyo, Japan) at 24 ◦C for 20 h.
Then, the supernatant was removed by decanting, and the immobi-
lized enzyme fraction was washed three times with 10 ml of 20 mM
sodium phosphate buffer (pH 6.5). The immobilized chlorogenate
hydrolase was assayed for protein concentration and enzyme activ-
ity. The protein concentration was determined using a Bradford
assay kit (Nacalai Tesque) with bovine serum albumin as the stan-
dard and by measuring absorbance at 595 nm. The procedure for
determining chlorogenate hydrolase activity is described below.
The immobilized enzyme (0.02 U mg−1) was prepared with the
immobilization yield of 48.8%.
2.3. Enzyme activity
The standard reaction for chlorogenate hydrolase was per-
formed at 40 ◦C for 4 h with shaking at 200 rpm (Magnetic stirrer
SW-RS777D, Nissin, Tokyo, Japan) under ambient pressure in a
1-ml reaction mixture consisting of 15 mM 5-caffeoylquinic acid,
2200 mM methanol, 3.6 U of immobilized chlorogenate hydro-
lase, 10 ␮l of 50 mM sodium phosphate buffer (pH 6.5), and
[BMIM][NTF2] as the reaction solvent. Fifty microliters of the reac-
tion mixture was removed, and the reaction with chlorogenate
hydrolase was terminated by adding 950 ␮l of methanol. For
quantitative analyses of the substrate and product formed, high-
performance liquid chromatography (HPLC) was performed using
a Cosmosil 5C18-ARII column (4.6 mm × 250 mm, Nacalai Tesque)
on a Hitachi D-2000 Elite HPLC system (Hitachi High-Technologies
Corp., Tokyo, Japan) equipped with a UV detector. The substrate and
product formed were detected at 330 nm. The column was equili-
brated with 0.2% acetate containing 30% methanol at a flow rate
of 1.0 ml/min at 40 ◦C. Elution was performed in a linear gradient
of 30–60% methanol for 10 min, followed by an isocratic elution
with 100% methanol for 15 min. 5-Caffeoylquinic acid and methyl
caffeate were eluted at retention times of 5.5 min and 13.1 min,
respectively. The synthesis of 3-cyclohexylpropyl caffeate with
methyl caffeate and 3-cyclohexy-1-propanol by Novozyme435 was
described previously [10]. The method of quantitative analysis of
the 3-cyclohexylpropyl caffeate formed was same as that of methyl
caffeate described above. The product 3-cyclohexylpropyl caffeate
was eluted at the retention time of 18.5 min.
2. Experimental
2.1. Enzymes and materials
Chlorogenate hydrolase (without glucose as
a stabilizing
agent, 0.36 U mg−1) from A. japonicus and C. antarctica lipase B
(Novozyme435, 3530 U mg−1) were kindly donated by Kikko-
man (Chiba, Japan) and Novozymes (Bagsvaerd, Denmark),
respectively. One unit of chlorogenate hydrolase activity was
defined as the amount of enzyme that catalyzed the formation
of 1 ␮mol of caffeic acid from 5-caffeoylquinic acid per min
at 40 ◦C [11]. One unit of Novozyme435 activity was defined
as the amount of enzyme that catalyzes the production of
1 ␮mol of 2-phenylethyl acetate from vinyl acetate and 2-
phenylethanol per min at 25 ◦C [10]. Caffeoylquinic acids, namely,
3-caffeoylquinic acid, 4-caffeoylquinic acid, 5-caffeoylquinic acid,
3,5-dicaffeoylquinic acid, 4,5-dicaffeoylquinic acid, and a mix-
ture of 3,4-dicaffeoylquinic acid and 4,5-dicaffeoylquinic acid,
were purified from green coffee beans [16]. The purities of the
compounds were confirmed with HPLC analysis according to
reference [16]. The HPLC analysis of these compounds showed
that single peak was occurred at 1.45 min (3-caffeoylquinic acid),
17.7 min (4-caffeoylquinic acid), 14.8 min (5-caffeoylquinic
acid), 38.6 min (3,5-dicaffeoylquinic acid), 44.0 min (4,5-
dicaffeoylquinic acid), and 35.8 min (3,4-dicaffeoylquinic acid).
These compounds were identified with FAB-MS analysis and
2.4. Consecutive conversions
For synthesis of 3-cyclohexylpropyl caffeate from 5-
caffeoylquinic acid by two consecutive reactions, methyl caffeate
was produced from 5-caffeoylquinic acid with shaking at 200 rpm

A. Kurata et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 161–167
163
Fig. 1. Consecutive enzymatic reactions for synthesis of 3-cyclohexylpropyl caffeate from 5-caffeoylquinic acid, and chemical structures of caffeoylquinic acids from cof-
fee beans. (A) In all reaction steps, [BMIM][NTf2] was used as the solvent. Compound 1: 5-caffeoylquinic acid, 2: methyl caffeate, 3: 3-cyclohexyl-1-propanol, and 4:
3-cyclohexylpropyl caffeate. (B) Structures of substrates for synthesis of methyl caffeate with chlorogenate hydrolase.
(Magnetic stirrer SW-RS777D, Nissin) under ambient pressure
in a 1-ml reaction mixture consisting of 15 mM 5-caffeoylquinic
acid, 2200 mM methanol, 3.6 U of immobilized chlorogenate
hydrolase, 10 ␮l of 50 mM sodium phosphate buffer (pH 6.5),
and [BMIM][NTF2] as the solvent. After the reaction at 40 ◦C for
4 h, the reaction mixture was gently stirred by a rotary evap-
orator (RE600, Yamato Science Co., Ltd., Tokyo, Japan) under
reduced pressure (14 hPa) at 80 ◦C for 1 h to remove unre-
acted methanol. Next, 400 mM 3-cyclohexane-1-propanol and
120,000 U of Novozyme435 were added to the reaction mixture,
and 3-cyclohhexylpropyl caffeate was synthesized at 80 ◦C under
reduced pressure (845 hPa) to remove methanol, which was
produced as a byproduct. Chlorogenate hydrolase is deactivated at
80 ◦C. Then, fifty microliters of the reaction mixture was removed,
and the reaction was terminated by adding 950 ␮l of methanol.
The conditions for quantitative analysis of 5-caffeoylquinic acid,
methyl caffeate, and 3-cyclohexylpropyl caffeate were described
above.
2.6. Structure determination
The structure of 3-cyclohexylpropyl caffeate was previously
determined by 1H and 13C NMR analysis [10]. For determination of
the structure of the methyl caffeate produced, the 1H and 13C NMR
spectra were recorded on a JEOL FT-NMR JNM-270EX (270 MHz)
spectrometer with tetramethylsilane as the internal standard.
2.6.1. Methyl caffeate (compound 2)
1H NMR (CD3OD): ı 3.72 (d, 3H, J = 1.00 Hz), 6.25 (d, 1H,
J = 1.31 Hz), 6.90 (d, 1H, J = 1.86 Hz), 6.92 (d, 1H, J = 1.86 Hz), 6.93 (d,
1H, J = 1.86 Hz), 7.54 (d, 1H, J = 2.03 Hz); 13C NMR (CD3OD): ı 51.99
(C1), 114.83 (C3), 115.10 (C7), 116.48 (C6, C10), 127.68 (C5), 146.82
(C4), 146.95 (C9), 149.59 (C8).
3. Results and discussion
3.1. Effect of aqueous solution on chlorogenate hydrolase in IL
We found that immobilized chlorogenate hydrolase catalyzes
the conversion of 5-caffeoylquinic acid to methyl caffeate with
methanol using [BMIM][NTf2] as a solvent under a non-aqueous
condition. Accordingly, we investigated an enzymatic synthesis of
3-cyclohexylpropyl caffeate from 5-caffeoylquinic acid via methyl
caffeate. In various enzymatic syntheses, water can shift the equi-
librium towards the direction of hydrolysis; therefore, methods
with lipase for removal of the water produced as a by-product
were applied [17,18]. However, the catalytic activity in a non-
conventional solvent depends on the amount of water associated
with the enzyme forming its native conformation [19]. First, the
effect of the addition of an aqueous solution on the production of
methyl caffeate was examined in [BMIM][NTf2] by varying the addi-
tional volume of 50 mM sodium phosphate buffer (pH 6.5) from 0
to 5% (v/v) (Fig. 2). When the concentration of the aqueous solution
was 1% or lower, the production of methyl caffeate increased. How-
2.5. Purification of reaction products
The purification procedures for methyl caffeate and 3-
cyclohexylpropyl caffeate were almost identical. As a represen-
tative example, the purification procedure for methyl caffeate is
described. The reaction mixture was extracted with diethyl ether,
and the extract containing the substrates and methyl caffeate was
concentrated in vacuo. The concentrate was applied to a silica gel
column (26 mm × 400 mm, Wako gel C-200, Wako Pure Chemical)
equilibrated with chloroform. Then, the product was eluted with
chloroform:methanol (9:1). The product formed was detected by
HPLC as described above and thin-layer chromatography (TLC) with
a silica gel 60F254 plate (Merck no. 5715, Darmstadt, Germany)
using chloroform:methanol (9:1) and detected by UV-radiation at
254 nm.

164
A. Kurata et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 161–167
caffeate and 3.2 mM caffeic acid were produced with conversion
yields of 75.4% and 21.5%, respectively (Fig. 3A), whereas 13.1 mM
caffeic acid and 1.21 mM 3-cyclohexylpropyl caffeate were pro-
duced with conversion yields of 87.6% and 8.1%, respectively,
using 15 mM 5-caffeoylquinic acid and 2200 mM 3-cyclohexyl-1-
propanol (Fig. 3B). The result showed that methanol is a better
substrate for chlorogenate hydrolase than 3-cyclohexylpropyl caf-
feate to produce methyl caffeate with a high conversion yield.
Additionally, Novozyme435 did not catalyze the conversion of
5-caffeoylquinic acid to 3-cyclohexylpropyl caffeate or methyl
caffeate (data not shown), but efficiently catalyzed the conver-
sion of methyl caffeate to 3-cyclohexylpropyl caffeate with a
conversion yield of 93.8% [10]. Thus, consecutive conversions of
5-caffeoylquinic acid to 3-cyclohexylpropyl caffeate via methyl
caffeate using chlorogenate hydrolase and Novozyme435 would
be a more effective procedure than a single conversion of 5-
caffeoylquinic acid to 3-cyclohexylpropyl caffeate.
Fig. 2. Effects of water concentration on conversion of 5-caffeoylquinic acid to
methyl caffeate by chlorogenate hydrolase. The reaction was performed according
to the standard method described in Section 2, except that the reaction was carried
out with a 0–5% (v/v) aqueous solution of 50 mM sodium phosphate (pH 6.5). Each
symbol indicates methyl caffeate (closed circle) and caffeic acid (open circle).
3.3. Selection of IL for chlorogenate hydrolase
In order to investigate the activities of chlorogenate hydro-
lase in various ILs, the alcoholysis of 5-caffeoylquinic acid
with methanol was examined (Fig. 4). The reactions were
performed in five ILs, namely, [BMIM][NTf2], two [BMIM]
cation-containing ILs, 1-butyl-3-methylimidazolium tetraflu-
oroborate ([BMIM][BF4]) and 1-butyl-3-methylimidazolium
trifluoromethanesulfonate ([BMIM][CF3SO3]), and two [NTf2]
ever, the production of methyl caffeate was decreased in greater
than 2% aqueous solutions. The production of caffeic acid was
increased with addition of the buffer, indicating that the enzyme
probably catalyzed hydrolysis of 5-caffeoylquinic acid to produce
caffeic acid rather than the alcoholysis to produce methyl caffeate.
Thus, our results indicated that addition of 1% aqueous solution
was suitable for the production of methyl caffeate. Using lipase
from C. antarctica for synthesis of isoamylacetate in a biphasic
1-butyl-3-methylimidazolium hexafluorophosphate [BMIM][PF6]-
alcohol system produced a similar optimum curve, and thus the 1%
(w/w) initial water concentration is suitable for the reaction [20]. It
was suggested that ILs are able to maintain the active structures of
enzymes with the monomolecular layer of water [21]. Thus, chloro-
genate hydrolase maintains the active structure with the layer of
the buffer in [BMIM][NTf2].
anion-containing
bis(trifluoromethylsulfonyl)imide
ILs,
N-methyl-N-propylpyrrolidinium
([MPPro][NTf2]) and N-
methyl-N-propylpiperidinium bis(trifluoromethylsulfonyl)imide
([MPPip][NTf2]). Although chlorogenate hydrolase catalyzed
the alcoholysis reaction in [BMIM][NTf2] to produce 13.7 mM
methyl caffeate, the reaction scarcely proceeded in [BMIM][BF4]
and [BMIM][CF3SO3]. Although Novozyme435 catalyzed trans-
esterification of methyl caffeate to 3-cyclohexylpropyl caffeate
in [BMIM][NTf2], the transesterification scarcely proceeded in
[BMIM][BF4] and [BMIM][CF3SO3] [10]. Lipase from Candida rugosa
was found to be active in [BMIM][PF6] for the transesterification,
but inactive in ILs including [BMIM][acetate] and [BMIM][nitrate]
[22]. Lipase from Pseudomonas aeruginosa was more stable in
[BMIM][PF6] than in [BMIM][BF4] [23]. Thus, the nature of the
anion in IL plays a critical factor in determining the enzyme
activity and stability. It was reported that the hydrogen-bond
basicities of [BMIM][BF4] and [BMIM][CF3SO3] are larger than that
of [BMIM][NTf2] [22,24]. Additionally, it was suggested that the
3.2. Conversion of 5-caffeoylquinic acid with chlorogenate
hydrolase
We examined the production of methyl caffeate or 3-
cyclohexylpropyl caffeate using immobilized chlorogenate hydro-
lase in [BMIM][NTf2] (Fig. 3A, B). After a 10-h reaction using 15 mM
5-caffeoylquinic acid and 2200 mM methanol, 11.3 mM methyl
Fig. 3. Conversion of 5-caffeoylquinic acid to methyl caffeate and caffeic acid (A) and. 3-cyclohexylpropyl caffeate and caffeic acid (B) by chlorogenate hydrolase. The reaction
was performed according to the standard method described in Section 2. The reaction mixture consisted of methanol (A) or 3-cyclohexyl-1-propanol (B) as the substrate.
Symbols indicate 5-caffeoylquinic acid (open circle), caffeic acid (open triangle), methyl caffeate (open square), and 3-cyclohexylpropyl caffeate (closed circle).

A. Kurata et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 161–167
165
of the product initially increased. However, at higher than 40 ◦C,
the amounts of the product decreased, indicating that 40 ◦C is the
optimum temperature for chlorogenate hydrolase. Additionally,
the enzyme is deactivated at temperatures higher than 80 ◦C. In
contrast, we previously showed that the optimum temperature of
Novozyme435 is 80 ◦C [10].
3.5. Effect of substrate concentration
For determination of the optimum concentrations of 5-
caffeoylquinic acid and methanol, the conversion yields of methyl
caffeate toward 5-caffeoylquinic acid were measured (Fig. 6).
Fig. 6A shows the conversion yields of 5–30 mM 5-caffeoylquinic
acid and 2200 mM methanol in [BMIM][NTf2]. When the 5-
caffeoylquinic acid concentration was 15 mM or lower, about 85%
conversion yield was observed. However, the conversion yield
decreased at concentrations of 5-caffeoylquinic acid greater than
15 mM. When the methanol concentration was 2200 mM or less,
the conversion yield was markedly increased. However, the con-
version yield was decreased at concentrations more than 2200 mM.
High concentrations of 5-caffeoylquinic acid and methanol possi-
bly have an inhibitory effect on chlorogenate hydrolase. Our results
indicated that 15 mM 5-caffeoylquinic acid and 2200 mM methanol
were suitable for the enzyme reaction.
Fig. 4. Selection of IL for chlorogenate hydrolase. The reaction was performed
according for the standard method described in Section 2.
[BF4] and [CF3SO3] anions are more nucleophilic than the [NTf2]
anion, and that the [BF4] and [CF3SO3] anions coordinate more
strongly to positively charged sites in the structure of an enzyme.
In consequence, the enzyme is deactivated by a conformation
change in the enzyme structure due to these anions. As shown
Fig. 4, chlorogenate hydrolase was deactivated in [BMIM][BF4] and
[BMIM][CF3SO3]. Additionally, the ability to decrease the enzyme
activity was in the order of cations: [MPPip]+ > [MPPro]+ > [BMIM]+.
In the case of chlorogenate hydrolase, the enzyme activity was
affected by the anions and the cations.
3.6. Substrate specificity
We examined the production of methyl caffeate with alco-
holysis reaction by chlorogenate hydrolase in [BMIM][NTf2].
Methyl caffeate were produced using various 15 mM caffeoylquinic
acids (15 ␮mol) and 2200 mM methanol (2200 ␮mol). Using 3-
caffeoylquinic acid, 4-caffeoylquinic acid, 5-caffeoylquinic acid,
3,5-dicaffeoylquinic acid, and 4,5-dicaffeoylquinic acid, methyl caf-
feate was synthesized at concentrations of 9.0 mM (9.0 ␮mol),
9.2 mM (9.2 ␮mol), 12.9 mM (12.9 ␮mol), 13.9 mM (13.9 ␮mol),
and 17.1 mM (17.1 ␮mol), respectively. Because dicaffeoylquinic
acid and caffeoylquinic acid have two and one caffeoyl groups,
respectively, the volume of methyl caffeate prepared from 3,5-
dicaffeoylquinic acid and 4,5-dicaffeoylquinic acid were greater
than that of methyl caffeate prepared from 3-caffeoylquinic acid,
4-caffeoylquinic acid, and 5-caffeoylquinic acid. In the cases of
3,5-dicaffeoylquinic acid and 4,5-dicaffeoylquinic acid, both caf-
feoyl groups would be used for synthesis of methyl caffeate. When
3-caffeoylquinic acid, 4-caffeoylquinic acid, 5-caffeoylquinic acid,
3,5-dicaffeoylquinic acid, and 4,5-dicaffeoylquinic acid were used
as the substrate, the conversion yields of methyl caffeate to each
substrate were 60%, 61.3%, 86.0%, 92.7%, and 114%, respectively.
Additionally, using a mixture of 3,4-dicaffeoylquinic acid and 4,5-
dicaffeoylquinic acid, which is a crude fraction prepared from coffee
beans, HPLC analysis showed that all peaks of caffeoylquinic acids
disappeared and that the peak of methyl caffeate occurred after a
4-h reaction with chlorogenate hydrolase in [BMIM][NTf2]. Chloro-
genate hydrolase acted on caffeoylquinic acids and dicaffeoylquinic
acids. Thus, using this procedure, methyl caffeate was produced
from various caffeoylquinic acids prepared from coffee beans.
Because both chlorogenate hydrolase and Novozyme435 were
active in [BMIM][NTf2], it is suitable for the consecutive conversion
of 5-caffeoylquinic acid to 3-cyclohexylpropyl caffeate via methyl
caffeate.
3.4. Effect of temperature
The effect of temperature on the activity of chlorogenate
hydrolase for alcoholysis reaction with 5-caffeoylquinic acid and
methanol was examined in [BMIM][NTf2] at temperatures from
20 to 100 ◦C (Fig. 5). With increased temperature, the amounts
3.7. Effect of methanol on transesterification by Novozyme435
During conversion of 5-caffeoylquinic acid with methanol, the
reaction mixture probably contained unreacted methanol. Addi-
tionally, when methyl caffeate prepared from caffeoylquinic acids
was used as the acyl donor to produce 3-cyclohexylpropyl caffeate,
methanol was produced as a byproduct. In the conversion of methyl
caffeate to 3-cyclohexylpropyl caffeate in the consecutive synthe-
sis, therefore, the reverse reaction with methanol would occur
easily to decrease the amount of the desired product. We tested the
Fig. 5. Effect of temperature on conversion of 5-caffeoylquinic acid to methyl caf-
feate by chlorogenate hydrolase. The reaction was performed according to the
standard method described in Section 2.

166
A. Kurata et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 161–167
Fig. 6. Effects of 5-caffeoylquinic acid (A) and methanol (B) concentrations on conversion yields of methyl caffeate. The reaction with chlorogenate hydrolase was performed
according to the standard method described in Section 2, except that the reaction was carried out using 7.5–30 mM 5-caffeoylquinic acid (A) and 120–4400 mM methanol
(B).
effect of methanol on Novozyme435, and observed that methanol
had an inhibitory effect on the production of 3-cyclohexylpropyl
caffeate (Fig. 7). For development of the two-step procedure, we
attempted to avoid this difficulty by removing the unreacted
methanol in vacuo (14 hPa) and the transesterification of methyl
caffeate by Novozyme435 under reduced pressure (845 hPa), so
that methanol could be removed immediately from the reaction
mixture [25].
3.8. Consecutive conversion of 5-caffeoylquinic acid with
chlorogenate hydrolase and Novozyme435
For the development of a convenient procedure for synthe-
sis of 3-cyclohexylpropyl caffeate from caffeoylquinic acids via
methyl caffeate, we tested the consecutive enzymatic reactions
by chlorogenate hydrolase and Novozyme435 using [BMIM][NTf2]
as the solvent (Fig. 8). In the one-pot synthesis, methyl caffeate
Fig. 8. Consecutive reactions for conversion of 5-caffeoylquinic acid to 3-
was first prepared from 5-caffeoylquinic acid by a 4-h reac-
tion at 40 ◦C. Next, the unreacted methanol was removed under
reduced pressure (14 hPa) for 1 h at 80 ◦C, and chlorogenate hydro-
lase was deactivated at 80 ◦C. Then, 3-cyclohexyl-1-propanol and
Novozyme435 were added to the reaction mixture. The trans-
cyclohexylpropyl caffeate via methyl caffeate. The reaction conditions are described
in Section 2. The symbols indicate 5-caffeoylquinic acid (open circle), methyl caffeate
(open square), and 3-cyclohexylproyl caffeate (closed circle).
esterification of methyl caffeate to 3-cyclohexylpropyl caffeate
was performed under reduced pressure (845 hPa) for removal
of methanol as the by-product. In the one-pot two-step syn-
thesis, 15 mM 5-caffeoylquinic acid (15 ␮mol) was converted,
via 13.1 mM methyl caffeate (13.1 ␮mol, 87.3%), to 12.8 mM 3-
cyclohexylpropyl caffeate (12.8 ␮mol, 97.7%). The conversion yield
of 3-cyclohexylpropyl caffeate toward 5-caffeoylquinic acid was
Fig. 7. Effect of methanol on conversion of methyl caffeate to 3-cyclohexylpropyl
caffeate by Novozyme435. Transesterification of methyl caffeate was performed
at 40 ◦C for 24 h with shaking in a 1-ml reaction mixture consisting of 50 mM
methyl caffeate, 400 mM 3-cyclohexyl-1-propanol, 120,000 U of Novozyme435,
0–2200 mM methanol, and [BMIM][NTf2] as the reaction solvent. After the reac-
tion, aliquots of the reaction mixture were removed, and the 3-cyclohexylpropyl
caffeate produced was quantified by HPLC analysis as described in Section 2.
Fig. 9. One-pot two-step synthesis of various CAPE analogues from various caf-
feoylquinic acids prepared from coffee beans via methyl caffeate. Using chlorogenate
hydrolase, Novozyme435, and [BMIM][NTf2] as the reaction solvent, caffeoylquinic
acids from coffee beans were converted to various CAPE analogues, which have
antimicrobial, anti-inflammatory, antioxidant, and antitumor activities.

A. Kurata et al. / Journal of Molecular Catalysis B: Enzymatic 69 (2011) 161–167
167
85.3%. In the synthesis of CAPE analogues, the conversion yield of
[3] V. Bankova, eCAM 2 (2005) 29–32.
[4] M.R. Fesen, K.W. Kohn, F. Leteurtre, Y. Pommier, Proc. Natl. Acad. Sci. U.S.A. 90
(1993) 2399–2403.
[5] P. Michaluart, J.L. Masferrer, A.M. Carothers, K. Subbaramaiah, B.S. Zweifel,
C. Koboldt, J.R. Mestre, D. Grunberger, P.G. Sacks, T. Tanabe, A.J. Dannenberg,
Cancer Res. 59 (1999) 2347–2352.
[6] M.C. Nicklaus, N. Neamati, H. Hong, A. Mazumder, S. Sunder, J. Chen, G.W. Milne,
Y. Pommier, J. Med. Chem. 40 (1997) 920–929.
[7] G.F. Sud’ina, O.K. Mirzoeva, M.A. Pushkareva, G.A. Korshunova, N.V. Sumbatyan,
S.D. Varfolomeev, FEBS Lett. 329 (1993) 21–24.
[8] T. Nagaoka, A.H. Banskota, Y. Tezuka, I. Saiki, S. Kadota, Bioorg. Med. Chem. 10
(2002) 3351–3359.
[9] S. Kadota, Y. Tezuka, Honeybee Sci. 25 (2004) 107–112.
[10] A. Kurata, Y. Kitamura, S. Irie, S. Takemoto, Y. Akai, Y. Hirota, T. Fujita, K. Iwai,
M. Furusawa, N. Kishimoto, J. Biotechnol. 148 (2010) 133–138.
[11] N. Kishimoto, Y. Kakino, K. Iwai, T. Fujita, Appl. Microbiol. Biotechnol. 68 (2005)
198–202.
3-cyclohexylpropyl caffeate toward 5-caffeoylquinic acid (85.3%)
was superior to that of CAPE toward 5-caffeoylquinic acid (50%)
[11]. CAPE analogues were obtained by single transesterification
systems catalyzed by Novozyme435 using [BMIM][NTf2] or isooc-
tane as the solvent with conversion yields of 93.8% and 91.65%,
respectively [10,26]. The conversion yield of 3-cyclohexylpropyl
caffeate produced by the consecutive conversion system was com-
parable to those of CAPE produced by single transesterification
systems.
4. Conclusion
[12] N. Kishimoto, Y. Kakino, K. Iwai, K. Mochida, T. Fujita, Biocontrol Sci. 10 (2005)
155–161.
We developed a convenient procedure for conversion of 5-
caffeoylquinic acid to 3-cyclohexylpropyl caffeate, which exhibits
an antiproliferative activity toward various human tumor cells
[10]. The procedure uses [BMIM][NTf2] as the reaction solvent, and
is composed of consecutive reactions catalyzed by chlorogenate
hydrolase and Novozyme435 (Fig. 1). In this study, we showed that
chlorogenate hydrolase acts on the caffeoylquinic acids in coffee
beans, namely, 5-caffeoylquinic acid, 4-caffeoylquinic acid, 3-
caffeoylquinic acid, 3,5-dicaffeoylquinic acid, 3,4-dicaffeoylquinic
acid, and 4,5-dicaffeoylquinic acid, to produce methyl caffeate
in [BMIM][NTf2]. Additionally, Novozyme435 catalyzed the con-
version of methyl caffeate to various CAPE analogues [10]. The
procedure developed in this study may be useful for the exploita-
tion of immature coffee beans that contain various caffeoylquinic
acids (Fig. 9).
[13] M. Sureshkumar, C.K. Lee, J. Mol. Catal. B: Enzyme 60 (2009) 1–12.
[14] F. van Rantwijk, R.A. Sheldon, Chem. Rev. 107 (2007) 2757–2785.
[15] M. Moniruzzaman, N. Kamiya, M. Goto, Org. Biomol. Chem. 8 (2010) 2887–2899.
[16] K. Iwai, N. Kishimoto, Y. Kakino, K. Mochida, T. Fujita, J. Agric. Food. Chem. 52
(2004) 4893–4898.
[17] K. Bélafi-Bakó, N. Dörmõ, O. Ulbert, L. Gubicza, Desalination 149 (2002)
267–268.
[18] L. Gubicza, A. Kabiri-Badr, E. Keoves, K. Bélafi-Bakó, J. Biotechnol. 84 (2000)
193–196.
[19] D. Barahona, P.H. Pfromm, M.E. Rezac, Biotechnol. Bioeng. 93 (2006) 318–324.
[20] E. Fehér, V. Illeová, I. Kelemen-Horváth, K. Bélafi-Bakó, M. Polakovicˇ, L. Gubicza,
J. Mol. Catal. B: Enzyme 50 (2008) 28–32.
[21] E. Fehér, B. Major, K. Bélafi-Bakó, L. Gubicza, Biochem. Soc. Trans. 35 (2007)
1624–1627.
[22] J.L. Kaar, A.M. Jesionowski, J.A. Berberich, R. Moulton, A.J. Russell, J. Am. Chem.
Soc. 125 (2003) 4125–4131.
[23] M. Singh, R. Singh, U. Banerjee, J. Mol. Catal. B: Enzyme 56 (2009) 294–299.
[24] J.L. Anderson, J. Ding, T. Welton, D.W. Armstrong, J. Am. Chem. Soc. 124 (2002)
14247–14254.
[25] T. Itoh, in: T. Matsuda (Ed.), Future Directions in Biocatalysis, Elsevier Bio-
science, The Netherlands, 2007, pp. 3–20.
References
[26] H.C. Chen, H.Y. Ju, Y.K. Twu, J.H. Chen, C.M. Chang, Y.C. Liu, C. Chang, C.J. Shieh,
N. Biotechnol. 27 (2010) 89–93.
[1] M. Tagashira, Y. Ohtake, Planta Med. 64 (1998) 555–558.
[2] M. Ysrael, M. Nonato, Acta Manilana. Ser. A 47 (1999) 41–52.