Efficient whole-cell biotransformation in a biphasic ionic liquid/water system

Article abstract of DOI:10.1002/anie.200460241

Biocompatible ionic liquids act as a substrate reservoir and an in situ extracting agent in biotransformations with whole cells (see scheme). In the reduction of 4-chloroacetophenone to (R)-1-(4-chlorophenyl)ethanol, high product concentration (82 g L

Full text of DOI:10.1002/anie.200460241

Angewandte
Chemie
Bioengineering
results were compared to those of a spectrum of organic
solvents, among which tert-butyl methyl ether (MTBE) and
diisopropyl ether are known to help stabilize alcohol dehy-
drogenase from Lactobacillus brevis.[7] For organic solvents,
membrane integrity clearly depends on their logP values
(Figure 1), as found earlier.[8,9] For n-decane (logP = 5.0)
Efficient Whole-Cell Biotransformation in a
Biphasic Ionic Liquid/Water System**
Holger Pfruender,* Maya Amidjojo, Udo Kragl, and
Dirk Weuster-Botz*
Whole-cell biocatalysis represents an elegant way of produc-
ing fine chemicals.[1] Industrial processes benefit from the high
product selectivity, simple catalyst preparation, and, in the
case of redox reactions, the recycling system inherent to the
cells. Problems arise from substrates and products that are
poorly soluble in water and/or display inhibitory or toxic
effects on the biocatalyst. Multiphase processes have been
established to address these limitations,[2] but the conven-
tional organic solvents used are often, themselves, toxic for
the cells, possibly explosive, and environmentally harmful.
Green solvents such as supercritical fluids and ionic liquids
(ILs) are promising alternatives.[1i,3]
Here we demonstrate the application of ILs as a substrate
reservoir and in situ extracting agent in an efficient multi-
phase process for the whole-cell-catalyzed synthesis of fine
chemicals exemplified by the asymmetric reduction of 4-
chloroacetophenone to (R)-1-(4-chlorophenyl)ethanol with
Lactobacillus kefir.[4] We have shown for the first time that a
cellular cofactor regeneration system is active in the presence
of ILs, and consequently high product concentrations can be
achieved without cofactor supplements.
Economically interesting whole-cell-catalyzed redox reac-
tions require a solvent that is nontoxic for the biocatalyst, a
prerequisite for efficient cofactor regeneration,[1e] and that
dissolves poorly water-soluble substrates. We focused on
water-immiscible ILs as the second liquid phase because with
a virtually nonexistent vapor pressure they are safe and easy
to handle, and first reports indicated whole-cell biocatalytic
activity in their presence.[3a,b,5] The ILs used in this study were
Figure 1. Membrane integrity (MI) of Lactobacillus kefir cells after 5 h
exposure to biphasic systems consisting of: a) buffer and organic sol-
vents, b) buffer and ionic liquids compared to pure buffer system
(aqueous), c) buffer containing glucose and ionic liquids containing 4-
chloroacetophenone compared to pure buffer systems containing glu-
cose with (“aqueous”) and without (“aqueous -S”) dispersed 4-chloro-
acetophenone.
1-n-butyl-3-methylimidazolium
hexafluorophosphate
BMIM-bis(trifluoromethanesulfonyl)imide
and methyltrioctylammonium-[Tf2N]
(BMIM[PF6]),
(BMIM[Tf2N]),
(OMA[Tf2N]).
In the absence of substrate, the effect of different ILs
present at 20% by volume on the L. kefir cell membrane, the
primary target of solvent toxicity, was determined.[6] The
membrane integrity decreased only to 52.7%. With MTBE
(logP = 0.9), diisopropylether (1.5), n-octanol (3.0), and n-
decanol (4.6) membrane integrity was 10% or less.
[*] H. Pfruender, M. Amidjojo, Prof. Dr.-Ing. D. Weuster-Botz
Institute of Biochemical Engineering
Munich University of Technology
ILs alone do not damage the cell membrane of L. kefir
(Figure 1b). Membrane integrity rather increases, an effect
that can be explained by a change in morphology similar to
that observed in growing cells (data not shown). We also
investigated the ability of ILs to circumvent substrate and
product toxicity. Without the addition of a solvent phase the
presence of 150 mm substrate and product led to nearly
complete membrane destruction (Figure 1c), whereas
whereas with BMIM[Tf2N] the membrane remained almost
intact (89.8%) (Figure 1c). BMIM[PF6] and OMA[Tf2N]
reduced toxicity to only around 30% although distribution
coefficients for the substrate and the product are nearly the
Boltzmannstrasse 15, 85748 Garching (Germany)
Fax: (+49)89-289-15714
E-mail: h.pfruender@lrz.tum.de
d.weuster-botz@lrz.tum.de
Prof. Dr. U. Kragl
Department of Chemistry
University of Rostock
Albert-Einstein-Strasse 3a, 18059 Rostock (Germany)
[**] We are grateful to Solvent Innovation for providing us with the ionic
liquids. This work was supported by the German Federal Ministry of
Education and Research (grant no. 0312737E).
Angew. Chem. Int. Ed. 2004, 43, 4529 –4531
DOI: 10.1002/anie.200460241
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4529

Communications
Table 1: Distribution coefficients logD of 4-chloroacetophenone and 1-(4-chlorophenyl)ethanol between the aqueous phase and different solvents,
chemical yields Y of biotransformation reactions in such biphasic systems with different cell densities and product purity (as enantiomeric excess).
Aqueous
BMIM[PF6]
BMIM[Tf2N]
OMA[Tf2N]
MTBE
n-Decane[a]
logD (4-Cl-AP)
solubilityꢀ7 mm
solubilityꢀ25 mm
42.1(Æ6.5)
2.65(Æ0.17)
1.93(Æ0.08)
82.7(Æ2.1)
88.2(Æ4.7)
99.8(Æ0.0)
2.71(Æ0.10)
2.03 (Æ0.09)
89.0(Æ6.2)
92.8(Æ3.4)
99.7(Æ0.1)
2.77(Æ0,.7)
1.98(Æ0.12)
80.1(Æ17.1)
88.4(Æ6.0)
99.4(Æn.d.)
2.73(Æ0.38)
2.54(Æ0.37)
2.4(Æ0.0)
2.38(Æ0.16)
1.54(Æ0.10)
n.d.
n.d.
n.d.
logD (1-4-ClÀ-P1E)
Y [%] (25 gL cells)
Y [%] (50 gLÀ1 cells)
ee [%] (50 gLÀ1 cells)
46.2(Æ3.7)
4.2(Æ0.1)
98.1(Æ0.2)
96.3(Æ0.2)
[a] n.d.=not determined.
same (Table 1). Hence, suitability of ILs for whole-cell
biocatalytic processes cannot be deduced by their distribution
coefficients and biocompatibility in the absence of toxic
reactants alone. Tests under the actual process conditions are
necessary to choose the right IL for a given system of
biocatalyst and reactants.
Equilibrium distribution coefficients of substrate and
product are nevertheless interesting with regard to process
efficiency. Inter alia, they determine the potential process
yield. Due to its low logD value for 1-4-Cl-PE (1.54) decane,
for example, would reduce the process yield in a 20% solvent
setup by about 11.5%, which is the fraction remaining in the
aqueous phase (Table 1). With BMIM[Tf2N] this loss only
accounts for 3.7%.
!
Figure 2. Concentrations of 4-chloroacetophenone ( ) and (R)-1-(4-
~
chlorophenyl)ethanol ( ) in BMIM[Tf2N] during biotransformation with
Lactobacillus kefir in a biphasic system.
Biotransformation reactions with different solvents were
conducted on the 2.8-mL scale to determine the chemical
yield and product purity. With 25 gLÀ1 L. kefir (dry weight) in
the aqueous phase, the reaction proceeded for 6 h and yielded
a little less product than reactions with 50 gLÀ1 that were
finished after 3 h (Table 1). Employing a pH control did not
significantly increase the yield (data not shown). The
advantage of ILs used as solvent phase was clearly demon-
strated. With BMIM[Tf2N] the chemical yield of 1-4-Cl-PE
was doubled from 46.2% to 92.8% relative to the process
with dispersed substrate and product, and the product purity
was pushed to an excellent 99.7% ee.
pervaporation, nanofiltration, or extraction with supercritical
fluids.[11] Practically no interfering by-products accumulated
in the IL (data not shown). The IL may consequently be
recycled. Membrane integrity of the cells at the end of the
reaction was 101.7% (Æ 5.8), which indicates that the
biocatalyst could potentially be reused. The cofactor turnover
number of this process is infinite, as the reaction did not
require additional NADPH or other redox equivalents. Thus,
overall costs of such a process are estimated to be very
competitive.
Despite the affected cell membranes, the results with
BMIM[PF6] and OMA[Tf2N] were also very good (Table 1).
With MTBE the yield is very low. Cells obviously disintegrate
even faster in the presence of MTBE than in the absence of a
solvent so that the cofactor regeneration breaks down before
significant amounts of product can be formed.
In summary, it could be shown that the ionic liquid
BMIM[Tf2N] exhibits very good solvent properties for 4-
chloroacetophenone and its benzyl alcohol reduction product
without destructive effects on the cell membranes of L. kefir.
Therefore, BMIM[Tf2N] could be used successfully as a
substrate reservoir and in situ extracting agent for an efficient
whole-cell biocatalytic process with inherent cofactor regen-
eration. We are currently testing this process design with
other reactions in order to evaluate its application range. If
positive, it could possibly boost the industrial use of whole-
cell catalysts for the production of fine chemicals.
We conducted the reaction as a 200-mL batch process with
BMIM[Tf2N] in a stirred-tank reactor (600 rpm, 2.3 WLÀ1).[10]
The reaction kinetics and process parameters were monitored
(Figure 2). The chemical yield (93.8% Æ 1.0) and product
purity (99.6% ee Æ 0.1) are identical to those obtained on the
2.8-mL scale. The process yield (88.3% Æ 1.4), volumetric
productivity (20.4 gLÀ1 hÀ1 Æ 0.4), and final product concen-
tration (81.6 gLÀ1 Æ 1.6) are much higher than reported so far
for whole-cell-catalyzed reductions of ketones in IL.[5d]
Furthermore, these key figures are also very good compared
to industrial biocatalytic processes.[1b,g] No emulsification of
the two-phase medium was observed. Therefore, the phases
could be separated easily by sedimentation or centrifugation.
For GC analysis, the product was extracted from the ILs with
hexane. Quantitative and environmentally benign recovery of
product from the IL should be possible by distillation,
Experimental Section
For biotransformations the IL or organic solvent (20% by volume)
containing 600 mm 4-chloroacetophenone, the buffer (0.2m KPi,
pH 6.5, 0.2m glucose, 5 mm MgCl2) and Lactobacillus kefir
DSM20587 cells were combined and stirred in vials (2.8 mL) or in a
200-mL bioreactor, resulting in IL droplets with diameters < 1 mm
dispersed in buffer.[10] Reference batches without a solvent phase
contained the same amount of substrate dispersed in buffer. Samples
were quenched with HCl if necessary, extracted with ethyl acetate or
4530
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 4529 –4531

Angewandte
Chemie
hexane, and analyzed by GC on a CP-3800 GC system (Varian) fitted
with a chiral BGB-174 column (BGB Analytik AG).
Science 2003, 302, 792 – 793; d) A. J. Mesiano, E. J. Beckman,
A. J. Russell, Chem. Rev. 1999, 99, 623 – 633; e) T. Matsuda, T.
Harada, K. Nakamura, Chem. Commun. 2000, 15, 1367 – 1368;
f) S. V. Dzyuba, R. A. Bartsch, Angew. Chem. 2003, 115, 158 –
160; Angew. Chem. Int. Ed. 2003, 42, 148 – 150.
Received: April 6, 2004 [Z460241]
[4] W. Hummel, Appl. Microbiol. Biotechnol. 1990, 34, 15 – 19.
[5] a) M. Freemantle, Chem. Eng. News 2003, 81, 9; b) B. Jastorff, R.
Stormann, J. Ranke, K. Molter, F. Stock, B. Oberheitmann, W.
Hoffmann, J. Hoffmann, M. Nuchter, B. Ondruschka, J. Filser,
Green Chem. 2003, 5, 136 – 142; c) S. G. Cull, J. D. Holbrey, V.
Vargas-Mora, K. R. Seddon, G. J. Lye, Biotechnol. Bioeng. 2000,
69, 227 – 233; d) J. Howarth, P. James, J. Dai, Tetrahedron Lett.
2001, 42, 7517 – 7519.
Keywords: asymmetric catalysis · biotransformations ·
cofactors · ionic liquids · whole-cell methods
.
[1] a) K. Faber, Biotransformations in Organic Chemistry: A Text-
book, Springer, Berlin, 1997, pp. 8 – 10; b) A. Liese, K. Seelbach,
C. Wandrey, Industrial Biotransformations, Wiley-VCH, Wein-
heim 2000, pp. 25 – 27; c) A. Liese, M. Villela Filho, Curr. Opin.
Biotechnol. 1999, 10, 595 – 603; d) B. Schulze, M. G. Wubbolts,
Curr. Opin. Biotechnol. 1999, 10, 609 – 615; e) A. Schmid, J. S.
Dordick, B. Hauer, A. Kiener, M. Wubbolts, B. Witholt, Nature
2001, 409, 258 – 268; f) A. Zaks, Curr. Opin. Chem. Biol. 2001, 5,
130 – 136; g) A. J. J. Straathof, S. Panke, A. Schmid, Curr. Opin.
Biotechnol. 2002, 13, 548 – 556; h) R. N. Patel, Enzyme Microb.
Technol. 2002, 31, 804 – 826; i) M. Bertau, Curr. Org. Chem. 2002,
6, 987 – 1014.
[6] Membrane integrity determination was based on the viability
test kit LIFE/DEADꢀ Baclight (Molecular Probes).
[7] M. Villela Filho, T. Stillger, M. Muller, A. Liese, C. Wandrey,
Angew. Chem. 2003, 115, 3101 – 3104; Angew. Chem. Int. Ed.
2003, 42, 2993 – 2996.
[8] Values for logP (n-octanol/water) were obtained from http://
esc-plaza.syrres.com/interkow/kowdemo.htm
data).
(experimental
[9] A. Inoue, K. Horikoshi, J. Ferment. Bioeng. 1991, 3, 194 – 196.
[10] D. Weuster-Botz, S. Stevens, A. Hawrylenko, Biochem. Eng. J.
2002, 11, 69 – 72.
[11] a) L. A. Blanchard, D. Hancu, E. J. Beckman, J. F. Brennecke,
Nature, 1999, 399, 28 – 29; b) T. Schafer, C. M. Rodrigues,
C. A. M. Afonso, J. G. Crespo, Chem. Commun. 2001, 17,
1622 – 1623; c) S. H. Schofer, N. Kaftzik, P. Wasserscheid, U.
Kragl, Chem. Commun. 2001, 5, 425 – 426; d) J. Krockel, U.
Kragl, Chem. Eng. Technol. 2003, 26, 1166 – 1168.
[2] a) C. Laane, S. Boeren, K. Vos, C. Veeger, Biotechnol. Bioeng.
1987, 30, 81 – 87; b) G. J. Salter, D. B. Kell, Crit. Rev. Biotechnol.
1995, 15, 139 – 177; c) R. Leon, P. Fernandes, H. M. Pinheiro,
J. M. S. Cabral, Enzyme Microb. Technol. 1998, 23, 483 – 500;
d) A. J. J. Straathof, A. J. J. Biotechnol. Prog. 2003, 19, 755 – 762.
[3] a) K. R. Seddon, J. Chem. Technol. Biotechnol 1997, 68, 351 –
356; b) U. Kragl, M. Eckstein, N. Kaftzik in Ionic Liquids in
Synthesis (Eds.: P. Wasserscheid, T. Welton), Wiley-VCH,
Weinheim, 2003, pp. 336 – 346; c) R. D. Rogers, K. R. Seddon,
Angew. Chem. Int. Ed. 2004, 43, 4529 –4531
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4531