Poly(ethylene glycol)-lipase complexes catalytically active in fluorous solvents

Article abstract of DOI:10.1039/b312212c

Lipase-catalyzed alcoholysis between vinyl cinnamate and benzyl alcohol in fluorous solvents was investigated. This is the first report of a lipase-catalyzed reaction in a fluorous solvent. Forming the poly(ethylene glycol) (PEG)-lipase PL complex enhanced lipase activity over 16-fold over that of native lipase powder. The PEG-lipase PL complex exhibited markedly higher alcoholysis activities in fluorous solvents than in conventional organic solvents such as isooctane and n-hexane. The optimum reaction temperature for FC-77 (perfluorooctane) was 55°C and the optimum pH for the preparation of the PEG-lipase complex was 9.0; similar to the conditions for lipase PL-catalyzed reaction in aqueous solution. The alcoholysis reaction in fluorous solvent requires the addition of a FC77-miscible organic solvent (isooctane) in order to dissolve non-fluorinated substrates. Lipase activity in the fluorous solvent was significantly influenced by the volume ratio of isooctane in the reaction medium. Vinyl cinnamate inhibition of the lipase-catalyzed reaction occurred at a much lower concentration in the fluorous solvent than in isooctane. These results can be explained by the localization of substrates around lipase molecules, induced by adsorption of the substrates to the PEG layer of the PEG-lipase complex.

Full text of DOI:10.1039/b312212c

View Article Online / Journal Homepage / Table of Contents for this issue
Poly(ethylene glycol)-lipase complexes catalytically active in
fluorous solvents
Tatsuo Maruyama, Takahiro Kotani, Hiroshi Yamamura, Noriho Kamiya and Masahiro Goto*
Department of Applied Chemistry, Graduate School of Engineering, Kyushu University,
6-10-1 Hakozaki, Fukuoka 812-8581, Japan
Received 2nd October 2003, Accepted 7th December 2003
First published as an Advance Article on the web 20th January 2004
Lipase-catalyzed alcoholysis between vinyl cinnamate and benzyl alcohol in fluorous solvents was investigated. This
is the first report of a lipase-catalyzed reaction in a fluorous solvent. Forming the poly(ethylene glycol) (PEG)-lipase
PL complex enhanced lipase activity over 16-fold over that of native lipase powder. The PEG-lipase PL complex
exhibited markedly higher alcoholysis activities in fluorous solvents than in conventional organic solvents such as
isooctane and n-hexane. The optimum reaction temperature for FC-77 (perfluorooctane) was 55 ЊC and the optimum
pH for the preparation of the PEG-lipase complex was 9.0; similar to the conditions for lipase PL-catalyzed reaction
in aqueous solution. The alcoholysis reaction in fluorous solvent requires the addition of a FC77-miscible organic
solvent (isooctane) in order to dissolve non-fluorinated substrates. Lipase activity in the fluorous solvent was
significantly influenced by the volume ratio of isooctane in the reaction medium. Vinyl cinnamate inhibition of the
lipase-catalyzed reaction occurred at a much lower concentration in the fluorous solvent than in isooctane. These
results can be explained by the localization of substrates around lipase molecules, induced by adsorption of the
substrates to the PEG layer of the PEG-lipase complex.
between vinyl cinnamate and benzyl alcohol in a fluorous
Introduction
solvent (FC-77) at 40 ЊC (initial reaction rate was 0.156 mmol
In the last decade, fluorous solvents, especially perfluoro-
(hؒg-enzyme)Ϫ1). We then examined non-aqueous enzyme acti-
carbons and perfluorinated ethers, have attracted interest as
novel reaction or extraction solvents, due to their unique prop-
erties. They are nonpolar, inert and immiscible with water. The
miscibility of perfluorocarbons with conventional organic
solvents such as hexane and isooctane depends strongly on the
conditions (temperature etc.). Zhu summarized various organic
vation methods (enzymology in organic media). One involved
using a surfactant (dioleyl N--glucono--glutamate)-coated
lipase6 and another involved using a PEG-lipase complex.5 The
surfactant-coated lipase exhibited 5-fold higher alcoholysis
activity, while the PEG-lipase complex showed 16-fold higher
alcoholysis activity than the native lipase. These results mean
that the lipase preparations, which are catalytically active
reactions carried out in perfluorocarbons and described the
advantages of using perfluorocarbons as a reaction medium.1
in conventional organic solvents, are also very effective for
Horvath and Rabai developed a fluorous biphasic system (FBS)
enhancing the enzymatic activity in a fluorous solvent. Several
as an efficient reaction system with facile separation.2 They
demonstrated that the FBS enabled easy extractive recovery or
isolation of fluorous-tagged catalysts and reagents.
Fluorous-tagged techniques have also achieved excellent
enzyme-assisted optical resolution performance. Hungerhoff
et al. reported that combining the lipase-catalyzed enantio-
lipases in PEG-complex form were tested for alcoholysis in the
fluorous solvent (FC-77). Of the five lipases tested in this
study, lipase PL exhibited the highest activity (5.0 mmol (hؒg-
enzyme)Ϫ1) in FC-77 (Fig. 1). Further experiments were then
carried out employing the PEG-lipase PL complex as the
biocatalyst.
Fig. 2 depicts the time courses of alcoholysis catalyzed by the
PEG-lipase PL complex in FC-77 and in isooctane. The enzym-
atic reaction in FC-77 proceeded 10-fold faster than that in
isooctane. In isooctane, the reaction rate appears to gradually
reduce over time. This implies that the PEG-lipase complex was
selective reaction with extraction of the product by a fluorous
phase allowed for the efficient resolution of racemic alcohols.3
More recently, Beier and O’Hagan studied lipase-catalyzed
transesterification (alcoholysis) in homogeneous perfluoro-
carbon-hydrocarbon solvents followed by the enantiometric
partitioning of the products by phase separation.4 These studies
clearly demonstrated the usefulness of coupling fluorous
solvents with enzymatic reactions in order to achieve optical
resolution. However, no report has discussed lipase activity
in perfluorocarbons or in homogeneous perfluorocarbon-
hydrocarbon solvents.
In the present study we investigate the lipase-catalyzed
alcoholysis in perfluorocarbons and discuss the lipase activity
in perfluorocarbons in comparison with that in conventional
organic solvents. Since native lipase powder does not exhibit
significant activity in fluorous solvents, we used a poly(ethylene
glycol)-lipase complex,5 which exhibits high synthetic activity in
fluorous and organic solvents, to examine alcoholysis activity in
fluorous solvents.
Fig.
1
Alcoholysis activities of various lipases complexed with
Results and discussion
In a preliminary investigation, native lipase powder (lipase
PS) exhibited extremely low alcoholysis activity for the reaction
poly(ethylene glycol) in FC-77. Alcoholysis reactions were carried out
between vinyl cinnamate (1 mM) and benzyl alcohol (1 mM) at 40 ЊC
using lipase (0.1 mg mlϪ1).
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 2 4 – 5 2 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
524

View Article Online
Fig. 4 Effect of reaction temperature on the activity of PEG-lipase PL
in FC-77. Reaction conditions were the same as those given in Fig. 1.
Fig. 2 Time courses of benzyl cinnamate production by lipase-
catalyzed alcoholysis in FC-77 and in isooctane. Reaction conditions
were the same as those given in Fig. 1.
media. In particular, enzymes are known to remember the
buffer pH used during preparation, and an enzyme prepared at
an optimum pH exhibits maximal activity in a non-aqueous
solvent. This is called “pH memory”.10 In this work we investi-
gated the effect of the buffer pH used during preparation on
subsequent lipase activity in a fluorous solvent (Fig. 5). While
native lipase PL showed maximal activity around pH 8.5 in the
aqueous solution (data provided by enzyme manufacturer),
the buffer with pH 9 provided the highest activity for the
PEG-lipase PL complex in FC-77. That is, we found that “pH
memory” is observed in the lipase-catalyzed reaction in the
fluorous solvent as well as in other organic solvents.
relatively stable in the fluorous solvent, while some denatur-
ation of the lipase occurred in isooctane at 40 ЊC over the
course of a few hours.
Lipase activities in various fluorous and organic solvents are
summarized in Fig. 3. Lipase activities in fluorous solvents were
remarkably high compared with those in organic solvents. In
general, the selection of solvents available for an enzymatic
reaction in organic media is severely constrained; because some
organic solvents have inherent ability to deactivate enzymes.
Laane et al. reported that enzymatic activity in organic solvents
is related to the hydrophobicity (such as log P) of the organic
solvents.7 They concluded that solvents with a log P greater
than four (i.e. hydrophobic solvents) tend to maintain high
enzymatic activity. Since the solubility of water in the fluorous
solvents tested here is less than 13 ppm, the fluorous solvents
are very hydrophobic. This hydrophobicity is one of the reasons
why the lipase exhibited markedly high activity in the fluorous
solvents. It should be noted that we did not find any significant
differences between the lipase activities in the fluorous solvents,
even though the properties of the fluorous solvents tested
varied.8 For example, the viscosity of FC-77 at 25 ЊC is 1.3 cp,
which is 2-fold higher than that of FC-72. The boiling points
of FC-72, FC-77 and FC-3255 are 56 ЊC, 97 ЊC and 101 ЊC,
respectively.
Fig. 5 Effect of pH during PEG-lipase complex preparation on
alcoholysis activity in FC-77. The buffers employed are KH2PO4–
NaOH buffer (pH 6–8, ᭹), borate buffer (pH 8-10, ᭺) and carbonate
buffer (pH 10-11, ᭝). The temperature of the alcoholysis reaction was
set at 55 ЊC and other reaction conditions were the same as those given
in Fig. 1.
The present reaction medium consisted of 99 vol% fluorous
solvent and 1 vol% isooctane. The substrates (vinyl cinnamate
and benzyl alcohol) were not fluorinated and were not soluble
in pure fluorous solvents. A small volume of isooctane was,
therefore, added to dissolve the non-fluorinated substrates. The
effect of the volume ratio of isooctane on the lipase activity in
FC-77 was examined (Fig. 6). The isooctane ratio was varied
from 1% to 100%, and the lipase activity in FC-77 increased
with decreasing isooctane ratio. This result implicitly suggests
Fig. 3 Alcoholysis activities of the PEG-lipase PL complex in fluorous
solvents and in ordinary organic solvents. Reaction conditions were the
same as those given in Fig. 1.
The effect of the reaction temperature on lipase activity in
FC-77 was studied (Fig. 4). The optimum temperature in FC-77
was 55 ЊC, while that for the hydrolysis reaction in an aqueous
buffer is 50 ЊC (data provided by enzyme manufacturer). Many
researchers have reported that the optimum reaction tem-
perature in non-aqueous enzymology is often higher than that
in aqueous buffer.9 However, in the present work, not much
difference was observed between the optimum temperatures in
fluorous solvent and in aqueous buffer, even though the lipase
molecule was coated with PEG-complex in the fluorous solvent.
In non-aqueous enzymology, enzyme preparation is of great
significance in obtaining an active biocatalyst in non-aqueous
Fig. 6 Effect of the volume ratio of isooctane on the activity of PEG-
lipase complex in FC-77. Reaction conditions were the same as those
given in Fig. 1.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 2 4 – 5 2 7
525

View Article Online
either the concentration effect of substrates that is induced by
localization of the substrates in a fluorous solvent or inhibition
by isooctane. Many papers have, however, described lipase-
catalyzed reactions using isooctane that showed no inhibition
of the enzymatic reactions by the isooctane co-solvent.11
containing 1 mM of pyrene exhibited emission at 475 nm (data
not shown). This indicated that an isooctane-in-FC-77 micro-
emulsion did not form in the reaction medium.
We then examined the adsorption of substrates and product
to PEG (PEG control). Since the PEG-lipase complex could
hydrolyse the substrate and the product, the PEG control,
which was prepared without lipase by the same procedure as the
PEG-lipase complex, was used for the adsorption tests. Fig. 8
shows that the substrates and the product employed here were
all adsorbed to the PEG control, both in FC-77 and in isooc-
tane. In particular, the adsorption of benzyl alcohol and benzyl
cinnamate was much greater in FC-77 than in isooctane, which
would be due to the differing solubilities of these components
in the different solvents. The adsorption experiments revealed
that the substrates in FC-77 were localized at the PEG layer
of the PEG-lipase complex, and suggest that the high activity
of the PEG-lipase complex and the results in Fig. 7 can be
explained by this localization.
Considering these facts, we investigated the effects of
substrate concentration on lipase activity in FC-77 and in
isooctane (Fig. 7). Lipase activity in both solvents was
strongly influenced by the substrate concentrations. In FC-77,
lipase activity sharply increased with increasing substrate
concentrations. However, when the substrate concentrations
exceeded a certain value (10 mM for vinyl cinnamate and
1 mM for benzyl alcohol), the activity started to decrease.
Chulalaksananukul et al. demonstrated that a lipase-catalyzed
esterification in an organic solvent proceeds by a Ping-Pong
Bi–Bi mechanism with inhibition by excess alcohol.12 In the
present study, we confirmed inhibition of lipase activity by both
benzyl alcohol and vinyl cinnamate in the fluorous solvent.
Fig. 7b depicts the effects of substrate concentrations on lipase
activity in isooctane. As in the fluorous solvent, lipase activity
in isooctane depends on substrate concentrations, but the max-
imum activity was obtained at a concentration of 200 mM vinyl
cinnamate, which is much higher than that in the fluorous sol-
vent. These results indicate that the substrates were localized
around lipase molecules in the fluorous solvent, which resulted
in a much higher substrate concentration in the environment of
the lipase molecules than the apparent one.
Fig. 8 Adsorption of vinyl cinnamate, benzyl alcohol and benzyl
cinnamate on PEG control (see Experimental section for details)
suspended in FC-77 (filled bars) and in isooctane (open bars). Substrate
and product concentrations were 1 mM, and the concentration of the
PEG control was 9 mg mlϪ1
.
Conclusions
Lipase-catalyzed alcoholysis between vinyl cinnamate and
benzyl alcohol was investigated in fluorous solvents. Lipase PL
in a complex form with PEG 20000 was highly active in fluor-
ous solvents. The optimum conditions for the lipase-catalyzed
reaction in the fluorous solvents were similar to those in
aqueous solution. pH memory was also observed in lipase in
fluorous solvents. Fluorous solvents are hydrophobic, but are
unlikely to dissolve non-fluorinated substrates, which induced
adsorption of the substrates to the PEG layer, and could
explain the higher activity achieved for lipases in fluorous
solvents. The present study suggests that methodologies
for enzyme reactions in ordinary organic solvents are also
applicable to those in fluorous solvents.
Fig. 7 Effect of substrate concentration on activity of PEG-lipase
complex in FC-77 (a) and in isooctane (b). Reaction conditions were
the same as those given in Fig. 1, except for the isooctane ratio (which
was 10 vol%).
Experimental
Lipase PL (from Alcaligence sp.) and lipase QLM (from
Alcaligence sp.) were kindly supplied by Meito Sangyo Co.
Ltd. (Nagoya, Japan). Lipase AY (from Candida rugosa), lipase
M (from Mucor javanicus) and lipase PS (from Pseudomonas
cepacia) were kindly provided by Amano Enzyme Inc (Nagoya,
Japan). FC-72 (perfluorohexane), FC-77 (perfluorooctane) and
FC-3255 (perfluorooctane) were generously donated by Sumi-
tomo 3M Co. Ltd. Benzyl cinnamate, vinyl cinnamate and
benzyl alcohol were purchased from Tokyo Chemical Industry,
Japan. All other chemicals were purchased from Wako Pure
Chemicals Ltd., Japan.
The fluorous reaction medium containing the substrates
appeared homogeneous, except for the PEG-lipase complex. If
the substrates are localized in the fluorous reaction medium,
there are two possibilities for this localization. One is isooctane-
in-FC-77 microemulsions where isooctane droplets dissolve the
substrates at high concentrations, and the other is adsorption of
the substrate to the PEG-lipase complex. To ascertain whether
or not a microemulsion formed, fluorescent measurement of
pyrene was carried out. Pyrene above 1 mM has intermolecular
excimer emission at a wavelength of 475 nm (excitation 345 nm)
and is dissolved in isooctane much more readily than in FC-77.
One milliliter of isooctane containing 1 mM of pyrene was
added to 99 ml of FC-77, followed by fluorescent measurements.
We did not observe excimer emission of pyrene, while isooctane
Preparation of poly(ethylene glycol) (PEG)-lipase complexes
The preparation of the PEG-lipase complex has been described
previously.5 Each lipase (10 mg) and poly(ethylene glycol)
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 2 4 – 5 2 7
526

View Article Online
20000 (PEG) were dissolved in phosphate buffer (5 ml, pH 7.0)
with a typical molar ratio (PEG : lipase) of 10. In order to
effectively coat the surfaces of lipase molecules with PEG13 and
provide an oil–water interface for lipase,14 toluene (13 ml) was
added to the lipase solution, which was then homogenized at
20000 rpm for 3 min using an Ultra-Turrax T25 in an ice bath,
to prepare water-in-oil (w/o) emulsions. The w/o emulsions
were immediately frozen in liquid nitrogen, followed by lyophil-
ization for 24 h using a freeze-dryer. The PEG-lipase complexes
were obtained as white powders.
complex without lipase (abbreviated as PEG control) was pre-
pared by the same preparation procedure as the PEG-lipase
complex, except that lipase was not added. PEG control (9 mg,
a 10-fold amount of the PEG-lipase complex for the reaction)
was added to 3 different 1 ml portions of FC-77 containing
10 µl of isooctane and 1 mM of each substrate or product. The
mixture was incubated for 30 min at 55 ЊC and then filtered
through a 0.5 µm filter. The free vinyl cinnamate and benzyl
cinnamate in FC-77 were extracted into 1 ml of acetonitrile.
The acetonitrile phase was then analyzed by HPLC to deter-
mine the concentration of each compound. The free benzyl
alcohol in FC-77 was extracted into 1 ml of n-hexane, followed
by GC analysis (Hewlett Packard HP 6890 equipped with a
capillary HP-FFAP column (Agilent Technologies, Palo Alto,
CA) and a flame-ionization detector).
Enzymatic reaction in perfluorocarbons and conventional organic
solvents
Alcoholysis between vinyl cinnamate (1 mM) and benzyl
alcohol (1 mM) to irreversibly produce benzyl cinnamate and
acetaldehyde was typically conducted using PEG-lipase com-
plexes (0.1 mg of original enzyme per 1 ml of a reaction
medium) at 40 ЊC in the fluorous and organic solvents. The
PEG-lipase complex (containing 0.1 mg of original enzyme)
was added to 0.99 ml of a fluorous solvent. The reaction was
started by the addition of an isooctane solution (10 µl) contain-
ing vinyl cinnamate (100 mM) and benzyl alcohol (100 mM) to
the fluorous solvent. Aliquots were periodically withdrawn
from the reaction medium and filtered through a 0.5 µm filter
(Millex, Millipore Co.). 1 ml of acetonitrile was added to the
filtrate to extract substrates and products into the acetonitrile
phase. The reaction product (benzyl cinnamate) was monitored
at 254 nm using an HPLC system (JASCO 2000 series)
equipped with a 4.6 × 250 mm column packed with 5 µm ODS.
An acetonitrile : water : acetic acid (80 : 20 : 0.1) solvent system
was used as the mobile phase, at a flow rate of 1.0 ml minϪ1. The
initial reaction rate was taken to be the initial rate of benzyl
cinnamate production.
The effect of the volume ratio of isooctane in the reaction
medium on the enzymatic activity was investigated as follows:
Isooctane/FC-77 mixtures with various isooctane ratios were
prepared. Enzymatic reactions were started by the addition of
isooctane solution (10 µl) containing the dissolved substrate
(100 mM each) to 0.99 ml of isooctane/FC-77 mixtures
containing the PEG-lipase complex (with 0.1 mg of original
enzyme).
The effects of the substrate concentrations on the enzymatic
activity in fluorous solvent were investigated using 1 ml solu-
tions of FC-77 containing 0.1 ml of isooctane and various con-
centrations of the substrates. When the concentration of one
substrate was varied, that of the other substrate was fixed at
1 mM.
Acknowledgements
This work was partly supported by Kyushu University
Interdisciplinary Programs in Education and Projects
I
Research Development. We would like to thank Meito Sangyo
Co. Ltd., Amano Pharmaceutical Co. Ltd. and Sumitomo
3M Co. Ltd. for generously providing the lipases and
perfluorocarbons.
References
1 D. W. Zhu, Synthesis, 1993, 10, 953–954.
2 I. T. Horvath and J. Rabai, Science, 1994, 266, 72–75.
3 (a) B. Hungerhoff, H. Sonnenschein and F. Theil, J. Org. Chem.,
2002, 67, 1781–1785; (b) B. Hungerhoff, H. Sonnenschein and
F. Theil, Angew. Chem. Int. Ed., 2001, 40, 2492–2494.
4 P. Beier and D. O’Hagan, Chem. Commun., 2002, 1680–1681.
5 T. Maruyama, S. Nagasawa and M. Goto, Biotechnol. Lett., 2002,
24, 1341–1345.
6 N. Kamiya, M. Goto and F. Nakashio, Biotechnol. Prog., 1995, 11,
270–275.
7 C. Laane, S. Boeren, K. Vos and C. Veeger, Biotechnol. Bioeng.,
1987, 30, 81–87.
8 http://cms.3m.com/cms/US/en/2-178/iilzRFS/view.jhtml.
9 A. Zaks and A. M. Klibanov, Science, 1984, 224, 1249–1251.
10 A. M. Klibanov, Nature, 1995, 374, 596–596.
11 Enzymatic Reactions in Organic Media, ed. A. M. P. Koskinen
and A. M. Klibanov, Blackie Academic & Professional, Glasgow,
1996.
12 (a) W. Chulalaksananukul, J. S. Condor, P. Delorme and R. M.
Willmot, FEBS Lett., 1990, 276, 181–184; (b) A. E. M. Janssen,
B. J. Sjursnes, A. V. Vakurov and P. J. Halling, Enzyme Microb.
Technol., 1999, 24, 463–470.
13 S. Okazaki, N. Kamiya, K. Abe, M. Goto and F. Nakashio,
Biotechnol. Bioeng., 1997, 55, 455–460.
14 (a) I. Mingarro, C. Abad and L. Braco, Proc. Natl. Acad. Sci. USA,
1995, 92, 3308–3312; (b) T. Maruyama, M. Nakajima, S. Ichikawa,
H. Nabetani, S. Furusaki and M. Seki, J. Am. Oil. Chem. Soc., 2000,
77, 1121–1126; (c) T. Maruyama, M. Nakajima and M. Seki,
J. Biosci. Bioeng., 2001, 92, 242–247.
Adsorption of the substrates and the product to PEG-lipase
complex
To examine the adsorption of benzyl alcohol, vinyl cinnamate
and benzyl cinnamate to the PEG-lipase complex, PEG
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 2 4 – 5 2 7
527