Stereospecific alkyne reduction: Novel activity of old yellow enzymes

Article abstract of DOI:10.1002/anie.200605179

Old enzyme, new trick: An environmentally sustainable alternative for the regio-and stereoselective reduction of the alkyne bond is the biotransformation by "Old Yellow Enzyme" (OYE). This "old" enzyme, first isolated in 1932, catalyzes the trans-selective hydrogenation of the carbon-carbon triple bond under mild conditions. (Chemical Equation Presented).

Full text of DOI:10.1002/anie.200605179

Communications
DOI: 10.1002/anie.200605179
Enzyme Catalysis
Stereospecific Alkyne Reduction: Novel Activity of Old Yellow
Enzymes**
AndrØ Müller, Rainer Stürmer, Bernhard Hauer, and Bettina Rosche*
The use of oxidoreductase enzymes extends the repertoire of
methods for regio- and stereospecific synthesis remarkably.[1]
In particular, the flavin-containing “Old Yellow Enzyme”
(OYE) and a growing number of discovered homologues are
valued as a toolbox for asymmetric reduction of the carbon–
carbon double bond.[2] While this is one of the best studied
enzyme families with ongoing investigations of substrate
range, biotechnological application, active-site catalytic
mechanism, and natural function, their action on a triple
bond is reported here for the first time. As an example, we
studied the enzymatic reduction of the phenyl-substituted
ynone 1. To our knowledge, enzymatic reductions of ynones
to enones have not been reported to date.
Figure 1. Enzymatic reduction of ynones and enones. 20 mm ynone 1
(left)or enone 2 (right)was converted by 10 gL À1 E. coli cells
(containing the enzymes OYE 1–3 or NCR)at pH 6.8 and 30 8C over
1 h (shaded: product 2; unshaded: product 3). The average values of
three biotransformations are given; error bars indicate the lowest and
highest results.
The trans-selective reduction of ynones to enones can be
achieved by treatment with diisobutylaluminum hydride in
the presence of hexamethylphosphoric triamide.[3] However,
the yield and selectivity are only satisfactory. Alternatively,
the trans enone can be produced by treatment of the ynone
with a water-soluble triaryl phosphine, but this phosphine has
to be applied in excess amount as it is oxidized to the
corresponding oxide.[4] For the cis-selective reduction of
dependent 2-cyclohexen-1-one reductase (NCR) from the
bacterium Zymomonas mobilis, were able to catalyze the
reactions shown in Scheme 1. The enzyme glucose dehydro-
ynones,
a solid-supported palladium catalyst has been
used.[5] The enzymatic reduction outlined herein may be an
environmentally friendly alternative for trans-selective reduc-
tions of the alkyne bond: The nontoxic catalyst and co-
substrate glucose are cheaply produced from renewable
resources and are biodegradable. Furthermore, the reaction
has low energy requirements as it operates at room temper-
ature and atmospheric pressure.
Recombinant E. coli strains that each produce one of four
different enzymes of the OYE family were investigated as
whole-cell biocatalysts for the reduction of the ynone 1. The
product concentrations shown in Figure 1 illustrate that all
tested enzymes, OYE 1, 2, and 3 from yeasts and NAD(P)H-
Scheme 1. Reductions catalyzed by Old Yellow Enzymes (OYEs). 1: 4-
phenyl-3-butyne-2-one; 2: (E)-4-phenyl-3-butene-2-one; 3: 4-phenyl-2-
butanone; NAD(P)+: nicotinamide adenine (phosphate)dinucleotide.
genase concurrently oxidized glucose and thus recycled
NADP+ to NADPH. Additional experiments showed that
NADH can also serve as the reducing agent. It is apparent
from Figure 1 that OYE 1, 2, and 3 reduced 1 to 2 at different
rates (left), and 1H NMR spectroscopy revealed that only the
trans isomer of 2 was produced. However, when 2 was the
substrate (right) all three enzymes reduced 2 to 3 at similar
rates. As an exception, NCR can reduce 2 to 3 faster than the
other enzymes, but 1 was reduced to 3 very slowly and without
detectable intermediate. Knowledge of these catalytic differ-
ences may be useful for synthetic applications as well as for
further studies of the structure–function relationship in this
popular enzyme family. The observed progressive increase in
ynone reduction from OYE 1 to OYE 3, for example,
correlates with an amino acid sequence identity of 91%
between OYE 1 and 2 and 80% between OYE 1 and 3.
The reverse reactions, the oxidation of 3 to 2 or 2 to 1,
were not observed in control experiments at pH 6.8 with
NAD+ added as an electron acceptor in the absence of
[*] Dipl.-Biol. A. Müller, Dr. B. Rosche
School of Biotechnology and Biomolecular Sciences
University of New South Wales
Sydney NSW 2031 (Australia)
Fax: (+61)2-9313-6710
E-mail: b.rosche@unsw.edu.au
Homepage: http://www.babs.unsw.edu.au
Dr. R. Stürmer, Prof. Dr. B. Hauer
Fine Chemicals and Biocatalysis Research
BASF-AG
67056 Ludwigshafen (Germany)
[**] This study was financially supported by BASF-AG. The authors are
very grateful to Dr. Thomas Friedrich (BASF)for the purification of
OYE3, to Dr. James Hook (UNSW)for his expertise on NMR, and to
Martin Bucknall (BMSF, UNSW)for assistance with GC/MS
analysis.
3316
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 3316 –3318

Angewandte
Chemie
glucose dehydrogenase and glucose. Under the experimental
conditions as described in Figure 1, the E. coli host strain was
not able to catalyze the reduction of the carbon triple or
double bond, but the host had a very low background activity
for the reduction of the keto group, with less than 2.4% of the
substrate being converted into the corresponding alcohol.
As OYE 3 showed a much higher rate for the reduction of
1 (Figure 1, left) than for the reduction of 2 (Figure 1, right),
this enzyme was selected for a biotransformation. The
enzymes OYE 3 and glucose dehydrogenase were both
produced by recombinant E. coli strains, which were added
as whole cells because enzyme purification would otherwise
add to production costs. The biotransformation profile in
Figure 2 illustrates that 1 was rapidly reduced to 2, which
2. For an optimized enzyme, an increase of the kcat value for
the ynone could be the target of improvement, for example,
through evolutionary approaches. Alternatively, the desired
product may be protected from further conversion through in
situ removal as the established “ping-pong” mechanism with a
shared site for substrate and NADPH binding[7] would require
2 to leave the enzyme before further reduction of 2 to 3 can
occur. In situ product removal has been successfully estab-
lished for other biotransformations.[8] A search for other
enzymes that reduce a carbon triple bond only revealed
nitrogenase, which reduces acetylene to ethylene in an ATP-
dependent reaction.[9] However, the reduction of ynones to
enones has not been reported.
In conclusion, OYE and homologous enzymes display a
potential for environmentally sustainable regio- and stereo-
selective hydrogenation of the carbon–carbon triple bond
conjugated with a ketone or possibly with other polar groups
such as aldehydes, esters, or nitro groups. Furthermore, the
discovery of this novel activity of an “old enzyme” is also
relevant for ongoing investigations of the enzyme structure–
function relationship.
Experimental Section
Unless otherwise specified, chemicals and enzymes were purchased
from Sigma-Aldrich (Castle Hill, Australia). Recombinant E. coli
strains that express the genes for the enzymes OYE 1 (from
Saccharomyces pasteurianus formerly named S. carlsbergensis,
sequence accession number Q02899, E. coli Lu13667), OYE 2 (from
S. cerevisiae, Q03558, E. coli Lu13668), OYE 3 (from S. cerevisiae,
P41816, E. coli Lu13669), NCR (from Zymomonas mobilis, Q5NLA1,
E. coli Lu14079), and glucose dehydrogenase (Bacillus subtilis,
M12276, E. coli Lu11598) were obtained from BASF (Ludwigshafen,
Germany). The strains were grown in Luria Broth with the respective
antibiotics and inducers for enzyme synthesis at 378C for 22 h. Cells
were washed, and the wet biomass was stored at À208C until further
usage. Biomass concentrations are given based on dry mass, which
was determined after drying an aliquot of cells at 1008C. (MES: b-
morpholinoethanesulfonic acid; EDTA: ethylenediaminetetraacetic
acid.)
Initial productivities (Figure 1) were determined in 4-mL glass
vials containing a magnetic stirrer bar and 1 mL of the reaction
mixture at 308C with a composition of 20 mm 1 or 2, 10% propan-2-ol
(v/v), 5 mm EDTA, 2 mm NADP+, 50 mm d-glucose, 1 UmLÀ1 cell-
free glucose dehydrogenase from Thermoplasma acidophilum
(Sigma-Aldrich, Castle Hill, Australia), E. coli equivalent to
10 gLÀ1 dry biomass, and 50 mm MES/KOH pH 6.8. The batch
biotransformation (Figure 2) was performed in a 50-mL glass reactor
fitted with stirrer, pH controller, and water jacket for temperature
control (308C). The biotransformation mixture contained 42 mm
(6.1 gLÀ1) 1, 10% propan-2-ol (v/v), 300 mm d-glucose, 10 mm
EDTA, 2 mm NADP+, biomass equivalent to the dry mass of
5 gLÀ1 E. coli (OYE 3) and 5 gLÀ1 E. coli (glucose dehydrogenase),
300 mm MES/KOH at pH 6.8, controlled by the addition of 2.5m
KOH. For the determination of the activity of OYE 3, samples were
treated with 5% (v/v) toluene for 15 min at room temperature to
further permeabilize the cells. After phase separation by centrifuga-
tion at 3000 ” g, Micro Bio-Spin 6 chromatography columns from Bio-
Rad (Hercules, CA, USA) were used to remove reactants and any
compounds smaller than 6 kDa from the supernatant. The activity of
OYE 3 in this eluate was determined for the oxidation of NADPH in
the presence of 1 using a photometer at a wavelength of 365 nm.
Other samples were analyzed by GC after extraction with
chloroform using n-decane as internal standard. Concentrations
Figure 2. Biotransformation of 1 by recombinant E. coli cells. E. coli
(OYE 3)and E. coli (GDH)were used at 5 gL À1 each (pH 6.8, 308C).
^
The average values of two analyses from one reactor are given (1:
;
~
^
2: ; 3: ). Values were not corrected for dilution by addition of KOH.
Considering the dilution, the mass balance at the end of the reaction
was 5 mm lower than expected.
accumulated over the first 5 h at an overall rate of
0.76 gLÀ1 hÀ1 and with a theoretical molar yield of 65%.
The velocity of the further reduction of 2 to 3 increased with
decreasing concentrations of 1 and the conversion was
complete in 11.5 h. Only 2.4% of 1 was reduced to the
corresponding alcohol. The enzyme was very robust, as
activity assays revealed that OYE 3 did not lose any activity
over the 11.5 h.
The apparent kcat values in Table 1 illustrate that one
active site of purified OYE 3 (one subunit of the dimer) was
able to turn over 1.63 molecules of 1 per second, but the
enzyme was about four times slower with 2. The apparent
half-maximal velocities were reached at low substrate con-
centrations (Michaelis–Menten constant, Km), which indi-
cates a high affinity of the enzyme to the substrates. The
catalytic efficiency (kcat/Km) was 6.7-fold higher for 1 than for
Table 1: Kinetic properties of purified OYE 3 from S. cerevisiae.[a]
Substrate
kcat [sÀ1
]
Km [mm]
kcat/Km [sÀ1 mmÀ1
]
1
2
1.63Æ0.13
0.40Æ0.01
0.184Æ0.019
0.304Æ0.014
8.8
1.3
[a] Apparent values with 0.3 mm NADPH at 308C, pH 6.8. Niino et al.
estimated a Km value of 0.0058 mm for NADPH when measured with
cyclohexenone as oxidant at 258C, pH 7.[6]
Angew. Chem. Int. Ed. 2007, 46, 3316 –3318ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3317

Communications
were determined by using external standards that were extracted by
Keywords: alkynes · biotransformation · enzyme catalysis ·
green chemistry · reduction
.
the same method. An RTX5-MS column (30 m ” 0.25 mm, 0.25-mm
phase thickness) from RESTEK (Bellefonte, USA) was used with
nitrogen as carrier gas (0.6 mLminÀ1) and split injection (1:60). The
temperature program was 808C (1 min), then 168CminÀ1 to 2808C
(1 min). The injector was set at 2508C and the flame ionization
detector was set at 3108C. The identity of compounds was confirmed
by co-elution with standards and by GC/MS (Agilent Technologies
5971GC/MSD, Santa Clara, CA, USA). The retention times were as
follows: n-decane 5.51 min, 3 8.30 min, 1 8.82 min, and 2 9.48 min.
GC/MS was performed as described for GC with a flow rate of
[1] A. Schmid, J. S. Dordick, B. Hauer, A. Kiener, M. Wubbolts, B.
Witholt, Nature 2001, 409, 258.
[2] A. S. Bommarius, K. M. Polizzi, Chem. Eng. Sci. 2006, 61, 1004;
A. Mꢀller, B. Hauer, B. Rosche, J. Mol. Catal. B 2006, 38, 126; M.
Wada, A. Yoshizumi, Y. Noda, M. Kataoka, S. Shimizu, H.
Takagi, S. Nakamori, Appl. Environ. Microbiol. 2003, 69, 933;
R. E. Williams, N. C. Bruce, Microbiology 2002, 148, 1607.
[3] T. Tsuda, T. Yoshida, T. Kawamoto, T. Saegusa, J. Org. Chem.
1987, 52, 1624.
[4] C. Larpent, G. Meignan, Tetrahedron Lett. 1993, 34, 4331.
[5] M. Gruttadauria, L. F. Liotta, R. Noto, G. Deganello, Tetrahe-
dron Lett. 2001, 42, 2015.
1
0.7 mLminÀ1 and splitless injection. H NMR spectra were recorded
in CDCl3 on a 300 MHz (Bruker DPX 300) instrument. The trans
configuration of 2 was concluded from the coupling constant of the
protons of the double bond, which was identical (3H, J = 16.2 Hz) for
sample and standard.
To determine the kinetic properties of purified OYE 3 (kindly
provided by BASF), the oxidation of NADPH was monitored at
365 nm and 308C (e365nm = 3.515 ” 103 mÀ1 cmÀ1) by using a temper-
ature-controlled CaryBio100 spectrophotometer (Varian Inc., Palo
Alto, CA, USA). Control experiments showed that activities were not
limited by the NADPH concentration. The assay contained 2.5–
13 mgmLÀ1 OYE 3, 2% propan-2-ol (v/v), 5 mm EDTA, 0.3 mm
NADPH, 50 mm MES/KOH pH 6.8, and 26.5 mm–1.95 mm of 1 or
202 mm–2.99 mm of 2. To minimize oxidation of reduced OYE 3 by O2,
the buffer was degassed and an O2-consuming system was included
(10 UmLÀ1 glucose oxidase from Aspergillus niger and 10 mm d-
glucose).[10] Kinetic parameters were calculated from double-recip-
rocal plots using CaryWinUV software 02.00 (Varian Inc.). For the
calculation of kcat the protein concentrations of OYE 3 were
determined by the method of Bradford. [11]
[6] Y. S. Niino, S. Chakraborty, B. J. Brown, V. Massey, J. Biol.
Chem. 1995, 270, 1983.
[7] K. M. Fox, P. A. Karplus, Structure 1994, 2, 1089.
[8] B. Buhler, A. Schmid, J. Biotechnol. 2004, 113, 183; D. Meyer, B.
Buehler, A. Schmid, Adv. Appl. Microbiol. 2006, 59, 3.
[9] B. K. Burgess, D. J. Lowe, Chem. Rev. 1996, 96, 2983.
[10] T. B. Fitzpatrick, S. Auweter, K. Kitzing, T. Clausen, N. Amrhein,
P. Macheroux, Protein Expression Purif. 2004, 36, 280.
[11] M. M. Bradford, Anal. Biochem. 1976, 72, 248.
Received: December 22, 2006
Published online: March 23, 2007
3318
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 3316 –3318