Bioreductions catalyzed by an alcohol dehydrogenase in non-aqueous media

Article abstract of DOI:10.1002/cctc.201300841

Highly productive biocatalytic reductions were established using an isolated alcohol dehydrogenase (ADH) under water-deficient conditions. First, a solvent-free system was evaluated for the reduction of 2-butanone catalyzed by ADH evo-1.1.200 promoted by the "smart cosubstrate" 1,4-butanediol. ADH evo-1.1.200 excelled by its activity and stability under high reagent concentrations and hence was the enzyme of choice. However, conversion of 2-butanone was limited to <1 % in 10 days under the solvent-free conditions. Therefore, water-immiscible organic solvents were evaluated whereby the highest conversions were achieved in MTBE and toluene. MTBE was chosen as its different boiling point compared to other reaction components (e.g., 2-butanone, 2-butanol, diol cosubstrate, and lactone coproduct) would simplify the downstream processing. Further on, by tuning substrate loading, the productivity of the ADH evo-1.1.200 was successfully increased to a turnover number (TON) of 64 000. Practical water-deficient enzymology for bioreductions: The use of alcohol dehydrogenases (ADHs) in neat substrates and in water-immiscible organic solvents is explored. The ADH evo-1.1.200 excelled by its high stability, as it showed significant catalytic activity over days. Reductions are coupled with the "smart cosubstrate" 1,4-butanediol; hence, excess amounts of reductants are avoided.

Full text of DOI:10.1002/cctc.201300841

CHEMCATCHEM
COMMUNICATIONS
DOI: 10.1002/cctc.201300841
Bioreductions Catalyzed by an Alcohol Dehydrogenase in
Non-aqueous Media
[
a, b]
[c]
[c]
[c]
Selin Kara,*
Dominik Spickermann, Andrea Weckbecker, Christian Leggewie,
[b]
[b]
Isabel W. C. E. Arends, and Frank Hollmann*
Highly productive biocatalytic reductions were established
using an isolated alcohol dehydrogenase (ADH) under water-
deficient conditions. First, a solvent-free system was evaluated
for the reduction of 2-butanone catalyzed by ADH evo-1.1.200
promoted by the “smart cosubstrate” 1,4-butanediol. ADH evo-
1.1.200 excelled by its activity and stability under high reagent
concentrations and hence was the enzyme of choice. However,
conversion of 2-butanone was limited to <1% in 10 days
under the solvent-free conditions. Therefore, water-immiscible
organic solvents were evaluated whereby the highest conver-
sions were achieved in MTBE and toluene. MTBE was chosen
as its different boiling point compared to other reaction com-
ponents (e.g., 2-butanone, 2-butanol, diol cosubstrate, and lac-
tone coproduct) would simplify the downstream processing.
Further on, by tuning substrate loading, the productivity of the
ADH evo-1.1.200 was successfully increased to a turnover
number (TON) of 64000.
Scheme 1. ADH-catalyzed reductions running in non-aqueous media
through coupling of 1,4-butanediol as cosubstrate.
In the 1980s and 1990s, extensive studies on the use of
ADHs in organic solvents were done by Klibanov and co-work-
[2]
ers. However, since then very few examples of bioreductions
[3]
in predominantly organic media by using either whole cells
[4]
or isolated enzymes have been reported. As a consequence,
it would be highly desirable to investigate ADH-catalyzed re-
ductions in organic media under water-deficient conditions,
and hence, we drew our attention to neat substrates and to
organic solvents (Scheme 1).
Biotransformations in organic/non-aqueous media are enjoying
great interest, as most of the synthetically interesting sub-
strates and products are not soluble in water and product re-
covery from the aqueous medium is often tedious. Indeed, ex-
amples of lipase-catalyzed reactions in organic solvents can be
often found, whereas the use of alcohol dehydrogenases
Recently, we reported a “smart cosubstrate” approach to
overcome the thermodynamic challenge in ADH-catalyzed re-
[5]
ductions. Therein, by using 1,4-butanediol (1,4-BD) as a sacrifi-
cial electron donor the thermodynamic equilibrium could be
shifted to the side of the product. This approach offers a signifi-
cant environmental advantage, as the waste generated for
conversions of >95% is reduced 40-fold. Overall, herein we
report ADH-catalyzed reduction reactions under water-defi-
cient conditions (in neat substrates or in a water-immiscible
solvent) promoted by the “smart cosubstrate” 1,4-BD.
[
1]
(
ADHs) in non-aqueous media has been limited. Hence, ADH-
catalyzed biotransformations often run at relatively low sub-
strate loadings and can still be considered economically unat-
tractive. In addition, from an environmental point of view, bio-
transformations with relatively low substrate loadings suffer
from the enormous amount of wastewater generated.
In a first set of experiments, we analyzed the reduction of 2-
butanone in neat substrates coupled with 1,4-BD as a cosub-
strate. We chose 2-butanone as the substrate because in aque-
ous media the product 2-butanol would form an azeotrope
and, thus, downstream processing would be tedious. The
enzyme of choice was the commercial ADH evo-1.1.200, as it
initially showed no loss in its activity at elevated diol concen-
trations up to approximately 4m (Figure S1, Supporting Infor-
[
a] Dr. S. Kara
Institute of Microbiology
Chair of Molecular Biotechnology
Technische Universitꢀt Dresden, 01062 Dresden (Germany)
Fax: (+49)0351-463-39520
E-mail: selin.kara@tu-dresden.de
[
b] Dr. S. Kara, Prof. Dr. I. W. C. E. Arends, Dr. F. Hollmann
Department of Biotechnology, Biocatalysis Group
Delft University of Technology
m
a
t
i
o
n
)
;
t
h
e
v
a
l
u
e
s
o
f
V
(
m
a
x
i
m
u
m
r
e
a
c
t
i
o
n
r
a
t
e
)
a
n
d
K
m
a
x
M
À1
Julianalaan 136, 2628 BL Delft (The Netherlands)
Fax: (+31)015-278-1415
E-mail: f.hollmann@tudelft.nl
(Michaelis–Menten constant) were found to be 2400 Ug and
84 mm for 1,4-BD, respectively.
Reactions running in neat substrates (2-butanone and 1,4-
[
c] D. Spickermann, Dr. A. Weckbecker, Dr. C. Leggewie
evocatal GmbH
Alfred-Nobel-Str. 10, 40789 Monheim am Rhein (Germany)
BD) revealed a linear increase in the concentration of 2-butanol
over 10 days, however, with a very limited specific activity
À1
(
15 Ug , Figure S2) and <1% conversion in 10 days (Figure 1).
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300841.
We attributed this significantly diminished enzyme activity in
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 973 – 976 973

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
Figure 1. Conversion of 2-butanone by using 1,4-BD as a cosubstrate run-
ning in neat substrates catalyzed by evo-1.1.200. Reaction conditions: c(2-
Figure 3. Conversion of 2-butanone catalyzed by evo-1.1.200 at different
substrate loadings in MTBE (5 mm dodecane) under water-deficient condi-
tions. Reaction conditions: c(2-butanone)=0.5–2m, c(1,4-BD)=0.25–1m,
+
butanone)=7.3m, c(1,4-BD)=3.7m, c(NAD )=0.5 mm, 2.5% (v/v) external
À1
water (50 mm Tris-HCl, pH 7.0), c(evo-1.1.200)=0.3 gL . Reaction mixtures
+
c(NAD )=0.5 mm, 71–91% (v/v) MTBE (5 mm dodecane), 2.5% (v/v) external
(1.5 mL) were kept at 308C and 1000 rpm.
À1
water (50 mm Tris-HCl, pH 7.0), c(evo-1.1.200)=0.3 gL , 308C, 1000 rpm. 2-
Butanone and 1,4-BD: 0.5 and 0.25 (
), 0.75 and 0.375 (*), 1 and 0.5 (~), 2
&
and 1m (^).
Figure 2. Synthesis of 2-butanol under water-deficient conditions running in
organic solvents. Reaction conditions: c(2-butanone)=1m, c(1,4-BD)=0.5m,
+
Figure 4. Specific activity of evo-1.1.200 depending on the concentration of
2-butanone in MTBE. Reaction conditions: c(2-butanone)=0.5–2m, c(1,4-
c(NAD )=0.5 mm, 84% (v/v) organic solvent, 2.5% (v/v) external water
À1
(
(
(
50 mm Tris-HCl, pH 7.0), c(evo-1.1.200)=0.3 gL , 308C, 1000 rpm. logP-
MTBE)=1.0, logP(DIPE)=1.4, logP(toluene)=2.5, logP(DBE)=2.9, logP-
octane)=4.5, logP(decane)=5.6.
+
BD)=0.25–1m, c(NAD )=0.5 mm, 2.5% (v/v) external water (50 mm Tris-HCl,
À1
pH 7.0), 71–91% (v/v) MTBE, c(evo-1.1.200)=0.3 gL , 308C, 1000 rpm.
neat substrates to the relatively high hydrophilicity of 1,4-BD
logP=À0.81, logarithmic value of octanol-water partition co-
efficient), which may strip the essential “structural water”
the organic solvents screened, MTBE and toluene gave the
highest conversions (Figure 2), whereby logP(MTBE)=1.0 and
logP(toluene)=2.5 (Table S2). Thus, in the present context,
logP cannot be a direct criterion to choose the organic sol-
(
[
6]
[1a]
of the enzyme, and this would result in limited molecular flexi-
bility.
[7]
vent. Owing to its lower boiling point, which also differs sig-
nificantly from the boiling points of the other reaction compo-
nents (Table S3), we chose MTBE as the solvent for our further
investigations.
In our further studies, we focused our attention on organic
solvents, as we could avoid the aforementioned removal of
“
structural water” by means of a water-immiscible organic sol-
vent. We chose MTBE (methyl tert-butyl ether), DIPE (diisoprop-
yl ether), toluene, DBE (di-n-butyl ether), decane, and octane,
as these have been successfully applied in ADH-catalyzed reac-
Next, we evaluated the effect of substrate loading on con-
À1
version (Figure 3). The specific activities in MTBE (80–150 Ug )
were determined to be up to 10-fold higher than those in neat
[
7]
À1
tions. In addition, we chose these solvents on the basis of
substrates (15 Ug , Figure S3). Our results showed significant-
their significantly different polarities (logP=1.0–5.6). Among
ly decreased enzyme activity at a 2-butanone concentration of
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 973 – 976 974

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
fluoroacetoacetate (6a). Moder-
ate conversions were obtained
for acetophenone (1a), and no-
tably, the conversions decreased
with Cl (see 2a), and MeO (see
Table 1. Conversion and ee values of alcohol products 1b–11 b after 24 h and 14 days synthesized under
water-deficient conditions (2.5% v/v water) in MTBE (5 mm dodecane) catalyzed by evo-1.1.200.
[
a]
Substrate
Product
24 h
14 d
Conv. [%]
ee [%]
Conv. [%]
ee [%]
3
a) substituents in the para po-
sition. Aliphatic ketones (see 8a–
1 a), however, were accepted to
2
1
2.6
9.1
99.0
40.7
36.9
18.4
42.1
96.5
1
1
2
a
1b
a lesser extent, and conversions
decreased with increasing chain
length. The complete conversion
of 6a and 2-butanone (8a) was
99.7
>99.9
>99.9
98.4
97.8
a
2b
achieved in
a maximum of
[
[
b]
b]
[b]
9
.8
14 days (Figure S5). The enantio-
meric excess (ee) values of evo-
3
4
5
a
a
a
3b
4b
5b
1
.1.200 for the products 2-chlor-
ophenylethanol (4b) and ethyl-
,4,4-trifluoro-3-hydroxybuta-
[b]
1
2.4
>99.9
4
noate (6b) were always at a max-
imum (>99.9%ee). The ee
values of phenylethanol (1b)
and its para-substituted deriva-
tives 2b and 3b were ꢀ99% ee;
however, these values slightly
decreased over the course of the
reaction. The ee values of secon-
dary aliphatic alcohols increased
with the chain length. Most in-
terestingly, in all cases we ob-
served a significant decrease in
the ee values with increasing
conversions. In fact, the ee
values decreased steeply if they
were low already at the begin-
ning, as shown in the case of 2-
butanol (8b) and 2-pentanol
4
3
5.8
9.2
–
89.3
99.9
–
[
[
b]
b]
[b]
[b]
>99.9
>99.9
6
7
8
9
a
a
a
a
6b
7b
1
1
1
4.8
5.1
2.9
95.0
21.1
76.5
98.0
39.4
95.6
62.5
34.9
31.2
93.5
7.0
8b
[
[
[
b]
b]
b]
[b]
[b]
[b]
49.9
96.7
96.2
9b
11.2
1
0a
1 a
10b
11 b
9
.4
96.8
(
9b, Figure S6). We attribute this
1
decrease in the ee values of the
alcohol products to enzyme-cat-
alyzed racemization through re-
oxidation of the alcohol prod-
+
À1
[
3
a] Reaction conditions: c(substrate)=0.5m, c(1,4-BD)=0.25m, c(NAD )=0.5 mm, c(evo-1.1.200)=0.3 gL
,
08C, 1000 rpm. Conversions were determined by GC. Absolute configurations were confirmed by authentic
standards. [b] Absolute configurations were assumed because of similar chromatographic behavior; for 4b and
b, the configuration was reversed because of the switch in substituent priorities according to Cahn–Ingold–
6
[8]
ucts. In the oxidation of 1,4-bu-
Prelog rules.
tanediol, this might be due to
the “half-reversibility” of the re-
action; hence, the first oxidation
step of 1,4-butanediol to the
2
m (Figure 4); however, the initial rates were essentially the
corresponding hydroxy aldehyde might still be regarded as
“quasireversible”. In fact, this might be the case if the second
oxidation step is rate limiting. At the beginning of the reaction,
only 2-butanone and 1,4-butanediol, as substrate and cosub-
strate, respectively, are present in the reaction mixture, and
hence, the ee value of (R)-2-butanol is determined by the enan-
tioselectivity of evo-1.1.200. However, with time a mixture of
(R)- and (S)-2-butanol, the hydroxy aldehyde intermediate, and
a lactone coproduct is formed, and consequently, there is
a competition between the (R)- and (S)-alcohol for their reoxi-
dation, which determines the ee of the alcohol product. To
same up to 1m 2-butanone (Figure 4). The turnover numbers
TONs) for the enzyme and for the cofactor were calculated to
(
be 64000 and 960, respectively. However, the reactions were
stopped at a point at which the enzyme was still active;
hence, the total turnover numbers (TTNs) would be still higher.
Owing to the imperfect enantioselectivity of evo-1.1.200 to-
wards the reduction of 2-butanone (see Table 1, entry 8), we
focused our further studies on screening the substrates.
Among the substrates screened (Table 1), the highest conver-
sions were achieved for cyclohexanone (5a) and ethyl-4,4,4,-tri-
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 973 – 976 975

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
sum up, as long as tiny amounts of “unpreferred” enantiomer
are present, the racemization will take place, as it is a thermo-
Keywords: alcohol dehydrogenases · biotransformations ·
organic media · reduction · smart cosubstrates
[
8b]
dynamically downhill process. The essential information for
the course of the reactions can be found in the Supporting In-
formation (Figure S6).
In this study, we demonstrated that ADHs applied in sub-
strate-coupled reductions can go far beyond the traditional
substrate loadings. For the reduction of 2-butanone coupled
with 1,4-butanediol as a cosubstrate in neat substrates, conver-
sion of <1% was detected, still with a significant TON of ap-
proximately 7800. Productivity of the enzyme was further opti-
mized by using MTBE as the organic solvent, whereby a TON
of 64000 was achieved.
[
[
1] a) K. Faber, Biotransformations in Organic Chemistry, 6th ed., Springer,
Berlin, 2011; b) P. Adlercreutz, Organic Synthesis with Enzymes in Non-
Aqueous Media (Ed.: G. Carrea, S. Riva), Wiley-VCH, Weinheim, 2008,
pp. 3–34.
2] a) J. Grunwald, B. Wirz, M. P. Scollar, A. M. Klibanov, J. Am. Chem. Soc.
1
986, 108, 6732–6734; b) C. R. Wescott, A. M. Klibanov, Biochim. Biophys.
Acta Protein Struct. Mol. Enzymol. 1994, 1206, 1–9; c) L. Dai, A. M. Kliba-
nov, Proc. Natl. Acad. Sci. USA 1999, 96, 9475–9478; d) A. M. Klibanov,
Nature 2001, 409, 241–246; e) A. M. Klibanov, Curr. Opin. Biotechnol.
2
003, 14, 427–431.
Notably, a suspension of the enzymes in non-aqueous media
resembles heterogeneous catalysis (Figure S7); thus, the solid
enzyme aggregate can be recycled without immobilization.
Ongoing research in our laboratories focuses on recycling the
enzyme and application of this approach.
[
3] a) H. Grçger, F. Chamouleau, N. Orologas, C. Rollmann, K. Drauz, W.
Hummel, A. Weckbecker, O. May, Angew. Chem. 2006, 118, 5806–5809;
Angew. Chem. Int. Ed. 2006, 45, 5677–5681; b) G. de Gonzalo, I. Lavan-
dera, K. Faber, W. Kroutil, Org. Lett. 2007, 9, 2163–2166; c) A. Jakoblin-
nert, R. Mladenov, A. Paul, F. Sibilla, U. Schwaneberg, M. B. Ansorge-Schu-
macher, P. D. de Maria, Chem. Commun. 2011, 47, 12230–12232; d) A.
Hibino, H. Ohtake, Process Biochem. 2013, 48, 838–843.
[
4] a) M. M. Musa, K. I. Ziegelmann-Fjeld, C. Vieille, J. G. Zeikus, R. S. Phillips,
Angew. Chem. 2007, 119, 3151–3154; Angew. Chem. Int. Ed. 2007, 46,
Experimental Section
3
091–3094; b) M. M. Musa, K. I. Ziegelmann-Fjeld, C. Vieille, R. S. Phillips,
The chemicals used in this study were commercially obtained in
analytical-grade quality and were used as received. The ADH evo-
Org. Biomol. Chem. 2008, 6, 887–892.
[
5] S. Kara, D. Spickermann, J. H. Schrittwieser, C. Leggewie, W. J. H. van Ber-
kel, I. W. C. E. Arends, F. Hollmann, Green Chem. 2013, 15, 330–335.
6] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR. Fundamentals and Appli-
cations in Chemistry and Biology, Vol. 2, 1st ed., American Chemical Soci-
ety, Washington, DC, 1995.
1
.1.200, recombinantly expressed in Escherichia coli, was commer-
cially available from evocatal GmbH (Monheim am Rhein, Germa-
ny). A detailed description of the experimental procedures as well
as the analytical protocols is given in the Supporting Information.
[
[
7] M. Villela Filho, T. Stillger, M. Mꢁller, A. Liese, C. Wandrey, Angew. Chem.
2
003, 115, 3101–3104; Angew. Chem. Int. Ed. 2003, 42, 2993–2996.
Acknowledgements
[
8] a) H. Yang, A. Jçnsson, E. Wehtje, P. Adlercreutz, B. Mattiasson, Biochim.
Biophys. Acta Gen. Subj. 1997, 1336, 51–58; b) M. M. Musa, R. S. Phillips,
M. Laivenieks, C. Vieille, M. Takahashi, S. M. Hamdan, Org. Biomol. Chem.
The authors thank Deutsche Bundesstiftung Umwelt (DBU) for fi-
nancial support of the project (AZ 13261). Joerg H. Schrittwieser
is gratefully acknowledged for fruitful discussions and proofread-
ing of the manuscript. It is a pleasure to acknowledge stimulat-
ing discussions with Andre Jakoblinnert and Doerte Rother.
2
013, 11, 2911–2915.
Received: October 4, 2013
Revised: December 8, 2013
Published online on February 12, 2014
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 973 – 976 976