Structural basis for a highly (S)-enantioselective reductase towards aliphatic ketones with only one carbon difference between side chain

Article abstract of DOI:10.1007/s00253-019-10093-w

Aliphatic ketones, such as 2-butanone and 3-hexanone, with only one carbon difference among side chains adjacent to the carbonyl carbon are difficult to be reduced enantioselectively. In this study, we utilized an acetophenone reductase from Geotrichum candidum NBRC 4597 (GcAPRD) to reduce challenging aliphatic ketones such as 2-butanone (methyl ethyl ketone) and 3-hexanone (ethyl propyl ketone) to their corresponding (S)-alcohols with 94% ee and > 99% ee, respectively. Through crystallographic structure determination, it was suggested that residue Trp288 limit the size of the small binding pocket. Docking simulations imply that Trp288 plays an important role to form a C-H?π interaction for proper orientation of ketones in the pro-S binding pose in order to produce (S)-alcohols. The excellent (S)-enantioselectivity is due to a non-productive pro-R binding pose, consistent with the observation that the (R)-alcohol acts as an inhibitor of (S)-alcohol oxidation.

Full text of DOI:10.1007/s00253-019-10093-w

Applied Microbiology and Biotechnology
https://doi.org/10.1007/s00253-019-10093-w
BIOTECHNOLOGICALLY RELEVANT ENZYMES AND PROTEINS
Structural basis for a highly (S)-enantioselective reductase towards
aliphatic ketones with only one carbon difference between side chain
Afifa Ayu Koesoema1 & Yosuke Sugiyama1 & Zichang Xu2 & Daron M. Standley2 & Miki Senda3
Toshiya Senda3,4 & Tomoko Matsuda1
&
Received: 29 March 2019 /Revised: 5 August 2019 /Accepted: 5 August 2019
#
Springer-Verlag GmbH Germany, part of Springer Nature 2019
Abstract
Aliphatic ketones, such as 2-butanone and 3-hexanone, with only one carbon difference among side chains adjacent to the
carbonyl carbon are difficult to be reduced enantioselectively. In this study, we utilized an acetophenone reductase from
Geotrichum candidum NBRC 4597 (GcAPRD) to reduce challenging aliphatic ketones such as 2-butanone (methyl ethyl ketone)
and 3-hexanone (ethyl propyl ketone) to their corresponding (S)-alcohols with 94% ee and > 99% ee, respectively. Through
crystallographic structure determination, it was suggested that residue Trp288 limit the size of the small binding pocket. Docking
simulations imply that Trp288 plays an important role to form a C-H⋯π interaction for proper orientation of ketones in the pro-S
binding pose in order to produce (S)-alcohols. The excellent (S)-enantioselectivity is due to a non-productive pro-R binding pose,
consistent with the observation that the (R)-alcohol acts as an inhibitor of (S)-alcohol oxidation.
.
.
.
.
Keywords Alcohol dehydrogenase Enantioselectivity Asymmetric reduction Crystal structure Docking simulation
Introduction
enantioselectively if the difference in the size of the side chain
adjacent to the carbonyl carbon is large (Fig. 1a). However, if
the ketone side chains are similar in steric or electronic prop-
erties, it is harder for catalysts to distinguish between the pro-S
and pro-R poses. Only enzymes have been proven to
enantioselectively reduce ketones such as para-substituted
diaryl ketones (Truppo et al. 2007), 4-alkylidene cyclohexane
(Agudo et al. 2013), and 3-oxa- or 3-thiacyclopentanone
(Noey et al. 2015). The excellent enantioselectivity of en-
zymes can be explained by the architecture of the binding
pocket (Hamnevik et al. 2014). Besides the ketones mentioned
Asymmetric reduction of ketones by non-biological
(Yoshimura et al. 2014; Shende et al. 2018) and biological
catalysts (Matsuda et al. 2009; Ni and Xu 2012; Gröger
et al. 2012; Rosenthal and Lütz 2018) is one of the most
straightforward ways to access enantiomerically pure alcohols
beneficial for pharmaceutical and agrochemical intermediates.
In particular, biological catalysts or enzymes are considered a
sustainable catalyst for pharmaceutical intermediates. Both
classes of catalysts have been proven to reduce ketones
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s00253-019-10093-w) contains supplementary
material, which is available to authorized users.
2
* Daron M. Standley
Department of Genome Informatics, Genome Information Research
standley@biken.osaka-u.ac.jp
Center, Research Institute of Microbial Diseases, Osaka University,
3-1 Yamadaoka, Suita, Osaka 565-0871, Japan
* Toshiya Senda
toshiya.senda@kek.jp
3
Structural Biology Research Center, Institute of Materials Structure
* Tomoko Matsuda
Science, High Energy Accelerator Research Organization (KEK),
1-1 Oho Tsukuba, Ibaraki 305-0801, Japan
tmatsuda@bio.titech.ac.jp
1
4
Department of Life Science and Technology, School of Life Science
Department of Materials Structure Science, School of High Energy
and Technology, Tokyo Institute of Technology, 4259 Nagatsuta-cho
Midori-ku, Yokohama 226-8501, Japan
Accelerator Science, SOKENDAI (The Graduate University for
Advanced Studies), 1-1 Oho, Tsukuba, Ibaraki 305-0801, Japan

Appl Microbiol Biotechnol
Fig. 1 Enantioselective reduction
of an easy substrate with large
differences in the side chain size
(a), a challenging substrate with
small differences in side-chain
size (b) with its mechanism
leading to the observed
a
b
Methyl vs. phenyl
Large differences
Ethyl vs. propyl
One carbon difference among side chains (small differences)
This study
enantioselectivity
Non-ideal
hydride
NADH
NADH
Zn
Zn
Too far
Small
Binding
pocket
W288
transfer angle
from Zn
S47
θ
θ
S47
Large binding pocket
Productive pro-S binding pose
Non-productive pro-R binding pose
above, it is also hard to reduce aliphatic ketones with only one
carbon difference between side chains (here denoted “chal-
lenging aliphatic ketones”) enantioselectively (Fig. 1b) by
using non-biological catalysts (Harada and Osawa 1995;
Boukachabia et al. 2011; Zhang et al. 2015a). To the best of
our knowledge, only a few biological catalysts can reduce
challenging aliphatic ketones with excellent
enantioselectivity, and none of the published studies provided
a mechanistic explanation for how length differences of just
one carbon can be distinguished (Keinan et al. 1986; Velonia
et al. 1999; Matsuda et al. 2002; Shimoda et al. 2004). In order
to investigate the mechanism of such well-defined
enantioselectivity, we have carried out crystallographic and
docking studies of ketones with similar side-chain sizes and
properties of a highly enantioselective reductase.
observed enantioselectivity, we solved the crystal structure of
GcAPRD and conducted docking simulation. Crystal structure
of GcAPRD suggested that residue Trp288 limit the size of the
small binding pocket. Moreover, the C-H⋯π interaction
formed by Trp288 plays an important role in proper orientation
of ketones in the pro-S binding pose to produce (S)-alcohols.
The pro-R binding pose was found to be non-productive thus
resulting in excellent (S)-enantioselectivity. This explanation is
supported by the fact that the (R)-alcohol acts as an inhibitor of
(S)-alcohol oxidation.
Materials and methods
Materials
Our group has previously utilized acetone powder of di-
morphic fungus Geotrichum candidum NBRC 4597 (APG4)
in order to reduce acetophenone and easy substrates
(Nakamura and Matsuda 1998) as well as challenging aliphat-
ic ketones with excellent enantioselectivity (Matsuda et al.
2002). However, these studies were still preliminary given
that APG4 consists of unpurified (S)- and (R)-alcohol dehy-
drogenase (ADH) mixture as well as other protein and cell
components. Following this study, one of the enzymes in
APG4, Geotrichum candidum acetophenone reductase
(GcAPRD), was purified (Nakata et al. 2010), overexpressed
in Escherichia coli, characterized (Yamamoto et al. 2013), and
crystallized (Sugiyama et al. 2015).
Our previous stud y showed that GcAPRD
enantioselectively reduced ketones with large differences be-
tween side chains such as acetophenone, tert-butyl acetoacetate,
4-phenyl-2-butanone, and 2,2,2-trifluoroacetophenone
enantioselectively (Yamamoto et al. 2013). Motivated by this
result, in this research, for the first time we utilized GcAPRD to
reduce challenging aliphatic ketones with side chains being
similar in sizes and properties (Fig. 2) resulting in 94–> 99%
ee (S). In order to understand the mechanism leading to the
Commercial grade solvents, ketones, and alcohol standards
(1-3b, 5-8b, (R)-1b, and (R)-2b) were purchased from
Nacalai Tesque (Kyoto, Japan) and Tokyo Chemical
Industry (Tokyo, Japan). Other alcohol standards were pre-
pared using sodium borohydride reduction of the correspond-
ing ketones (Johnson and Rickborn 1970). Chiral alcohol and
ester standards were prepared using Novoyzm 435®–cata-
lyzed kinetic resolution with vinyl propionate and/or vinyl
butyrylate as the acyl donor (Bandeira et al. 2016). Spectral
data of prepared standards are listed in Online Resource 1.
Chemicals for cultivation and enzyme purification were pur-
chased from Nacalai Tesque except for 4-(2-hydroxyethyl)-1-
piperazineethanesulfonic acid (HEPES), which was pur-
chased from Sigma-Aldrich (St. Louis, USA), and dithiothre-
itol (DTT) which was purchased from Wako (Tokyo, Japan).
Chemicals for protein quantification were purchased from
Nacalai Tesque, except for 29:1 acrylamide:bisacrylamide,
which was purchased from Bio-Rad (CA, USA), ammonium
persulfate, purchased from Tokyo Chemical Industry,
tetramethylethylenediamine purchased from Wako, and pro-
tein concentration measurement reagent from Bio-Rad. The

Appl Microbiol Biotechnol
Fig. 2 Asymmetric reduction of
challenging aliphatic ketones by
GcAPRD
GcAPRD
ee 94 ~ >99% (S)
1a-15a
1b-15b
12a, b R1 = n-Pr R2 = n-Pr
13a, b R1 = n-Pr R2 = n-Bu
14a, b R1 = n-Pr R2 = n-Pen
15a, b R1 = n-Bu R2 = n-Bu
1a, b R1 = Me R2 = Et
2a, b R1 = Me R2 = Pr
3a, b R1 = Me R2 = n-Bu
4a, b R1 = Me R2 = n-Pen
5a, b R1 = Me R2 = n-Hex
6a, b R1 = Me R2 = n-Hep
7a, b R1 = Et R2 = Et
8a, b R1 = Et R2 = Pr
9a, b R1 = Et R2 = n-Bu
10a, b R1 = Et R2 = n-Pen
11a, b R1 = Et R2 = n-Hex
accession numbers of the GcAPRD amino acid sequence in
GenBank and UNIPROT are BAN05122.1 and M5A8V4,
respectively. The protein structure of GcAPRD was deposited
to the Protein Data Bank (PDB) with an accession number of
6ISV.
9.5 μg for 12b, 190 μg for 13b-15b; for enantiomerically pure
oxidation: 0.16 μg for (S)-3b, (S)-4b, and (S)-5b; 0.32 μg for
(S)-8b and (R)-3b; 0.64 μg for (S)-9b; 2.6 μg for (S)-10b; and
12.8 μg for (R)-4b, (R)-5b, (R)-8b, (R)-9b, and (R)-10b). All
assays were performed using UV-1600-UV-Visible spectro-
photometer (Shimadzu). Initial velocities were recorded by
following NADH consumption (ketone reduction) or NADH
generation (alcohol oxidation) at 340 nm for 132 s. For the
Michaelis Menten constant [Km (mM)] and kcat (mM-1 s-1)
determination, activity towards substrate 1-15a, (S)-8b, and
(R)-8b were measured through variation of substrate concen-
tration from 0.075 to 15 mM. Inhibition constant [Ki (mM)]
was also determined by the reduction of 0.25 to 15 mM of 7a
in the presence of 0 to 1.0 mM of 12a or 0.25 to 15 mM of (S)-
8b and (S)-1-phenylethanol ((S)-16b) in the presence of 0 to
5.0 mM of (R)-8b and (R)-16b, respectively. All kinetic pa-
rameters were determined by using a non-linear regression
function in GraphPad Prism 7.0. Protein concentration was
determined by the Bradford method using bovine serum albu-
min as a standard.
Cloning, expression, and purification
GcAPRD was prepared according to a previous procedure
(Yamamoto et al. 2013). E. coli RosettaTM(DE3)plysS with
pET-21b(+)-APRD were grown at 37 °C in LB broth contain-
ing 125 μg/mL carbenicillin and 20 μg/mL chloramphenicol.
When the culture optical density (OD600) reached 0.6, protein
expression was induced by addition of isopropyl β-D-1-
thiogalactopyranosid (IPTG) with the final concentration of
1 mM. After incubation at 20 °C for 20 h, the cells were
harvested; suspended in a sonication buffer of 100 mM
HEPES-NaOH (pH 7.2), 2 mM imidazole, 1 mM
phenylmethylsulfonyl fluoride (PMSF), and 1 mM dithiothre-
itol (DTT); and sonicated on ice (20 min, 100 W) using
Insonator 201M (Kubota, Japan). The suspensions were cen-
trifuged and the supernatant was purified on HisTrapTM FF
crude (GE Healthcare) equipped on HPLC LC-20AD
(Shimadzu, Japan) using 20 mM HEPES-NaOH buffer (pH
7.4) containing 0.3 mM DTT with a gradient from 20 to
500 mM imidazole. The protein was concentrated by ultrafil-
tration using Amicon Ultra-15 10-K MWCO (Merck), and the
imidazole concentration was brought to 20 mM. The protein
was purified to near homogeneity, as shown by a single band
in SDS-PAGE (Fig. S1).
Analytical scale reduction
Analytical scale reductions for substrate 1-15a were per-
formed in a 3.0 mL of 100 mM HEPES-NaOH buffer (pH
7.2) consisting of 0.20 mM NAD+, 15% v/v 2-propanol, and
GcAPRD with the amount needed to achieve 1.0 μmol/min/
mL reduction of each substrate to its corresponding alcohol.
Substrate concentration was varied as follows: 10 mM for 1-
5a, 7a-9a, and 12a; 6.3 mM for 10a and 13a; and 1.6 mM for
6a, 11a, and 14-15a. Reductions were done for 3 h at 30 °C
with a rotational speed of 250 rpm. Part of reaction mixture
was extracted with diethyl ether for reaction yield determina-
tion using chiral gas chromatography (GC-14 B (Shimadzu)
equipped with a CP-Chirasil-Dex-CB column (Varian
0.32 mm × 0.25 μm × 50 m)). The remaining reaction mixture
was extracted with dichloromethane; the organic layer was
dried over MgSO4 and used for propionylation by propionyl
chloride or butyrylation by butyryl chloride reaction. Absolute
configurations were determined by comparing GC retention
Activity assay
All activity assays were done at 37 °C in 100 mM HEPES-
NaOH buffer (pH 7.2). Reduction or oxidation activity was
measured in a 1.0-mL scale containing 2.5 mM of 1-15a or 1-
15b (rac or enantiomerically pure), 0.28 mM NADH or
NAD+, and purified enzyme (for reduction: 0.63 μg for 1a-
8a, 1.3 μg for 9a-13a, 2.6 μg for 14a-15a; for oxidation:
0.48 μg for 1b-8b, 0.95 μg for 9b, 1.9 μg for 10b-11b,

Appl Microbiol Biotechnol
times of 2b, propionates of 1b, 3b-6b, 8b-9b, and 11b, or
butyrylate of 10b with the corresponding authentic samples.
The chiral GC retention time of ketone, alcohol, and ester
standards is shown in Table S1.
crystal structure as the receptor. The sdf files for ligands were
obtained from PubChem and converted to pdb format.
Receptor and ligands were prepared by using
AutoDockTools 1.5.7rc1 and the docking was done by using
AutoDock Vina 1.1.12 (Trott and Olson 2010). The size of a
cubic gridbox centered on the receptor chain A catalytic zinc
was set to 1000 Å3. Productive binding poses were defined in
terms of distance and angle restraints as shown in Fig. S2.
Restraints 1, 2, and 3 were adapted from the literature
(Dhoke et al. 2015) and restraint 4 was adapted from several
reported ADH crystal structures in complex with substrate or
substrate analogs (PDB ID: 2VWH, 5ENV, 4DWV, and
1HLD) as follows:
Preparative scale reduction
Preparative scale reduction was performed in 100 mL
100 mM HEPES-NaOH buffer (pH 7.2) consisting of
1.4 mM NAD+, 3% v/v 2-propanol, and 0.85–1.1 mmol of
substrate. The amount of enzyme used was varied as follows:
90, 45, 18, 11, and 7 μmol/min for substrates 4a, 5a, 8a, 9a,
and 10a, respectively. The reaction was done for 12–40 h at 30
°C with rotational speed of 130 rpm. The reaction mixture was
extracted with diethyl ether 2 times and dried over MgSO4,
and the organic solvent was evaporated under reduced pres-
sure. The residue was purified by silica gel column chroma-
tography (hexane:ethyl acetate, 4:1) to give the corresponding
alcohols. The product was characterized by 1H-NMR analysis,
enantioselectivity was determined by GC analysis, and abso-
lute configurations were assigned by measuring the optical
rotation value using P-2200 Polarimeter (JASCO, Japan).
Spectral data are listed in Online Resource 1.
1. D1 (distance of ligand carbonyl or hydroxy oxygen to Zn)
is less than 2.5 Å
2. D2 (distance of ligand carbonyl or hydroxy oxygen to
hydroxy oxygen of Ser47) is less than 3.0 Å
3. D3 (distance of ligand carbonyl or alpha carbon to C4 of
NAD(H)) is less than 4.0 Å
4. θ (angle of C4 of NAD(H)—carbonyl or alpha carbon of
ligand—carbonyl or hydroxy oxygen of ligand) is within
70–90° (Ramaswamy et al. 1994; Baker et al. 2009; Plapp
and Ramaswamy 2012; Plapp et al. 2016)
Crystallization
Among the poses fulfilling restraints, we selected those
consistent with a given chirality and, from these, chose the
pose with the lowest binding energy. To ensure reproducibil-
ity, the docking simulation was repeated 10 times for each
ligand.
Crystallization of the GcAPRD-NAD+ complex was per-
formed by the hanging-drop vapor diffusion method at
293 K following previous procedures (Sugiyama et al.
2015). A 0.2 mM purified GcAPRD was mixed with 3 mM
NAD+ in a crystallization solution containing 20% PEG 3350
and 200 mM trisodium citrate dihydrate. GcAPRD-NAD+
complex crystal was cryoprotected using the cryo solution
containing 27% glycerol, 18% PEG 3350, 10 mM NAD+,
and 180 mM trisodium citrate dihydrate for 1 min.
Following cryoprotection, the crystals were picked using ny-
lon loops and flash cooled using liquid nitrogen.
Results
Asymmetric reduction of challenging aliphatic
ketones
Diffraction data of GcAPRD-NAD+ complex crystal was
collected at 95 K using Eiger X4M detector on beamline BL-
1A Photon Factory (PF, Tsukuba, Japan). The data set was
processed and scaled using XDS (Kabsch 2010) and
Aimless. Processed data was phased using zinc anomalous
scattering by the SAD method using Crank2. Model building
and electron density map inspection were done using the Coot
program (Emsley and Cowtan 2004). Following model build-
ing, refinement was carried out in Phenix refine (Adams et al.
2010). Validation of the structure and quality check was per-
formed using the PDB validation server.
We investigated GcAPRD activity of aliphatic ketone reduc-
tion (Fig. 3a). GcAPRD showed high activity in the reduction
of methyl ketone (1a-6a) followed by ethyl ketone (7a-11a),
and it had poor activity in the reduction of propyl and butyl
ketone (12a-15a). The kinetic parameters also support this
result (Table 1). Methyl ketone (1a-6a) reduction generally
had lower Km and high kcat compared with ethyl ketone (7a-
11a). The oxidation of corresponding racemic alcohols by
GcAPRD displayed a similar trend (Fig. 3b): it preferred the
oxidation of methyl (1b-6b) followed by ethyl alcohol (7b-
11b) with poor activity for propyl (12-14b) and butyl alcohol
(15b).
Docking simulations
Following analysis of the activities, we performed asym-
metric reduction of aliphatic ketone on analytical and prepar-
ative scales (Table 2). GcAPRD showed moderate to high
yield in reducing methyl (1a-6a) and ethyl (7a-11a) ketones.
Docking simulations were done for 6 ligands 1a, (S)-1b, (R)-
1b, 8a, (S)-8b, and (R)-8b with the GcAPRD-NAD+ complex

Appl Microbiol Biotechnol
a
b
100
80
60
40
20
0
100
methyl alcohol
ethyl alcohol
propyl alcohol
butyl alcohol
methyl ketone
ethyl ketone
propyl ketone
80
60
40
20
0
butyl ketone
1b 2b 3b 4b 5b 6b 7b 8b 9b 10b 11b 12b 13b 14b 15b
1a 2a 3a 4a 5a 6a 7a 8a 9a 10a 11a 12a 13a 14a 15a
Substrate
Substrate
Fig. 3 Relative specific activity of aliphatic ketone reduction and racemic
alcohol oxidation. a Aliphatic ketone reduction, absolute specific activity
for substrate 1a was determined to be 34.7 μmol/min/mg enzyme. b
Aliphatic racemic alcohol oxidation, absolute specific activity for
substrate 3b was determined to be 19.9 μmol/min/mg enzyme. Assays
were done under standard assay conditions using 2.5 mM of aliphatic
ketones 1a-15a or rac-aliphatic alcohols 1b-15b. Substrate owning R1
of methyl, ethyl, propyl, and butyl are indicated with dark gray, gray,
white, and black bars, respectively
As expected, the reduction of propyl (12a-14a) and butyl ke-
tone (15a) had very low yields. In terms of enantioselectivity,
GcAPRD displayed excellent enantioselectivity in aliphatic
ketone reduction following Prelog’s rule to produce the corre-
sponding (S)-alcohols (Prelog 1964). The enantioselectivity
was > 99% (S) for reduction of most of methyl (2a-6a) and
ethyl (7a-10a) ketones except in the reduction of 1a and 11a,
which were 94% and 98%, respectively. This excellent S-se-
lectivity was supported by the preference of GcAPRD to ox-
idize (S)-alcohol by more than 60~160 fold higher activity
than the (R)-alcohol (Table S2). GcAPRD reduced challeng-
ing aliphatic ketones 2-butanone (1a) and 3-hexanone (8a) to
their corresponding (S)-alcohols with 94% ee and > 99% ee,
respectively. Previously, we reported the enantioselective re-
duction of ketones with large differences between side chains
such as acetophenone, tert-butyl acetoacetate, 4-phenyl-2-
butanone, and 2,2,2-trifluoroacetophenone by GcAPRD
(Yamamoto et al. 2013). This study shows the capability of
GcAPRD to enantioselectively reduce challenging ketones
with only one carbon difference in the length of the side chain
(1a, methyl vs. ethyl or 8a, ethyl vs. propyl) among side
chains adjacent to the carbonyl carbon. Comparison of (S)-
8b and (R)-8b kinetic parameters shows similar Km, but the
kcat of (R)-8b is approximately 10 times smaller than that of
(S)-8b. Thus, we investigated the mechanism leading to the
observed enantioselectivity of GcAPRD through its crystal
structure and docking simulation.
Table 1 Kinetic parameters of GcAPRD towards aliphatic ketones and
alcohols
Substrate
Km
(mM)
kcat
kcat/Km
(s-1)
(mM-1 s-1)
1a
0.54 0.03
0.50 0.05
0.32 0.03
0.25 0.03
0.41 0.05
0.53 0.04
3.20 0.42
2.42 0.16
1.66 0.20
1.32 0.09
0.76 0.08
N.d
49.36 1.31
47.88 2.78
39.38 1.98
38.2 1.76
37.38 1.92
44.57 1.68
28.15 1.73
27.73 0.82
42.02 2.01
37.59 0.93
14.51 0.97
N.d
91.41 3.92
95.76 7.76
123.10 9.25
152.80 11.47
90.98 7.67
84.10 4.99
8.79 0.85
11.46 0.42
26.26 2.27
28.48 1.32
19.09 1.64
N.d
2a
3a
Crystal structure of GcAPRD
4a
5a
Crystallization was done for the GcAPRD binary complex
with NAD+. The GcAPRD crystal was refined to 2.5-Å reso-
lution. Final refinement resulted in a Rcryst and Rfree of 0.219
and 0.271, respectively (Table S3). GcAPRD possesses the
same structural fold as many other medium-chain ADHs
(Vidal et al. 2018): its monomer consists of Rossman-like
NAD+ binding domain and catalytic domain (Fig. 4a). One
asymmetric unit of the crystal contains a GcAPRD dimer (Fig.
S3a). SEC-MALS analysis revealed the molecular mass in
solution to be ~ 154 kDa (Fig. S4), and SDS-PAGE analysis
revealed the molecular mass of the monomer to be 37 kDa
(Fig. S1). These results indicate that GcAPRD forms a tetra-
mer consisting of subunit A-B and subunit C-D (Fig. S3b), in
contrast to the classical dimeric ADH (Plapp 2011). Other
examples of tetrameric ADHs are yeast alcohol dehydroge-
nase (Raj et al. 2014), E. coli alcohol dehydrogenase
6a
7a
8a
9a
10a
11a
12a
13a
14a
15a
(S)-8b
(R)-8b
N.d
N.d
N.d
N.d
N.d
N.d
N.d
N.d
N.d
2.18 0.09
3.91 0.34
11.56 0.19
0.14 0.005
5.31 0.15
0.04 0.002
Kinetics assays were done under standard assays condition using 0.075–
10 mM of ketones or alcohols. The ee for all alcohols used were > 99%.
N.d, not determined due to the low activity

Appl Microbiol Biotechnol
Table 2 Reduction of aliphatic
ketones by GcAPRDa
Substrate
Yield (%)a, b
ee (%)a,b,c
1a
62
94 (S)
> 99 (S)
> 99 (S)
> 99 (S)
> 99 (S)
> 99 (S)
–
2a
> 99
99
3a
4ad
5ad
6a
99 (53)
> 99 (85)
> 99
98
7a
8ad
9ad
10ad
11a
12a
13a
14a
15a
85 (46)
91 (54)
98 (85)
> 99
2
> 99 (S)
> 99 (S)
> 99 (S)
98 (S)
–
3
N.d
1
N.d
2
–
a Reactions were performed in standard reaction conditions in a total volume of 3 mL for 3 h at 30 °C. b Yields and
ee values were determined using chiral GC. c The absolute configurations of the products were determined by
comparing GC retention times of 2b, propionates of 1b, 3b-6b, 8b-9b, and 11b or butyrylate of 10b with authentic
samples. d Preparative scale reductions were also done for substrate 4a, 5a, 8a, 9a, and 10a with isolated yield
shown in parentheses and following result. 121 mg (1.06 mmol) of 4a was converted to 65 mg of (S)-4b, 53%,
22
25
[α]D = +11.34 (c = 1.15, CHCl3) > 99% ee (S), ([α]D = +10.40 (c = 0.52, CHCl3) 99% ee (S) (Shimoda et al.
23
2004)). 104 mg of 5a (0.81 mmol) was converted to 89 mg of (S)-5b, 85%, [α]D = +9.63 (c 0.91, CHCl3) > 99%
ee (S), ([α]D = +7.90 (c 0.60, CHCl3) > 99% ee (S) (Ferreira et al. 2012)). 101 mg of 8a (1.01 mmol) was
28
21
20
converted to 48 mg of (S)-8b, 46%, [α]D = +12.04 (c 0.95, CHCl3) > 99% ee (S), ([α]D = +8.50 (c 1.00,
CHCl3) 97% ee (S) (Keinan et al. 1986)). 102 mg of 9a (0.89 mmol) was converted to 56 mg of (S)-9b, 54%,
21
20
[α]D = +10.05 (c 1.01, CHCl3) > 99% ee (S) ([α]D = +8.24 (c 0.20, CHCl3) > 99% ee (S) (Ferreira et al.
21
2012)). 103 mg (0.80 mmol) of 10a was converted to 89 mg of (S)-10b, 85%, [α]D = +10.46 (c 0.95, CHCl3)
98% ee (S) ([α]D = +10.0 (c 1.00, CHCl3) > 99% ee, (S) (Murai et al. 2006)). N.d, not determined due to the low
20
amount of alcohol produced
(Karlsson et al. 2003), and CPCR2 from Candida parapsilosis
(Man et al. 2014).
(R)-specific carbonyl reductase from Candida parapsilosis
(RCR) with 47% amino acid sequence identity and a root-
mean square deviation of 1.6 Å over 332 residues (Wang
et al. 2014, PDB ID:3WLE). Differences in the two structures
were observed not only in the surface loops of the protein but
also near the active site. In particular, the shapes and entrances
of the binding pockets were different in terms of both se-
quence and structure. Several key differences in the amino
acids composing the large binding pocket of these two en-
zymes were Ile56 (Val50 in RCR), Phe56 (Leu55 in RCR),
and Asn119 (Trp116 in RCR). Structurally, Leu122 (Leu119
in RCR) had a significantly different orientation. In the small
binding pocket, the side chain of Trp288 (Trp286 in RCR)
was considerably different. Moreover, the active site entrance
of GcAPRD had bulkier amino acid residues compared with
RCR.
One GcAPRD subunit consists of a catalytic and a non-
catalytic zinc ion. The catalytic zinc which resided in the ac-
tive site is coordinated to Cys45, His66, and Asp156 (Fig. 4b).
Meanwhile, the non-catalytic zinc is coordinated to Cys98,
Cys101, Cys104, and Cys112. The nicotinamide ring of
NAD+ is positioned with its C4 towards the active site.
NADH in this orientation would point its pro-R hydride to-
wards the active site. Ser47, which is believed to have an
important role in proton transfer (Man et al. 2014), also orients
its hydroxy group toward the active site. Neighboring the
active site are the large and small binding pockets of
GcAPRD (Fig. 4c). The large binding pocket is composed
of Ser47, His50, Ile51, Phe56, His66, Asn119, Leu122, and
Leu264. The small binding pocket is composed of Cys45,
Asp156, Thr160, Phe287, and Trp288. Especially, Trp288
was located on the edge of the small binding pocket; thus, it
played an important role to limit its size.
Docking simulations
We compared GcAPRD with its closely related homologs
by using the DALI server calculation (Holm and Laakso
2016). The closest homolog of GcAPRD was chain C of
To understand the enantioselectivity of GcAPRD, we con-
ducted docking simulations of 1a, (S)-1b, (R)-1b, 8a, (S)-8b,
and (R)-8b. We chose the best productive binding poses that

Appl Microbiol Biotechnol
Fig. 4 GcAPRD crystal structure
monomer. a Overall fold.
a
b
NAD+
Catalytic domain and NAD+
(Rossmann-like) binding domain
are shown in cyan and blue,
respectively. b Active site of
GcAPRD crystal structure
Catalytic zinc
(derived from subunit A). Amino
acids coordinated to catalytic zinc
(C45, H66, and D156) and S47
(responsible for the proton
S47
S47
N-term
Zn
transfer) are shown in stick. c
Binding pocket illustration of
GcAPRD crystal structure. Pink
and yellow sphere represent the
small and large binding pocket.
Catalytic zinc and NAD+ are
shown in grey sphere and purple
stick, respectively. The C4 of
NAD+ that will accept a hydride
from the alcohol substrate is
highlighted with white circle.
D156
C45
C-term
Non-catalytic zinc
H66
c
fulfilled the distance and angle restraint parameters described
in the materials and methods sections. The restraints corre-
spond to ideal binding distances of substrate with catalytic
Zn, ideal hydrogen transfer distances to Ser47, and favored
distances and angles for hydride transfer from C4 of NADH to
the carbonyl carbon of substrate (Fig. S2).
The binding energies, distance restraints, and angle re-
straints are described in Table S4. The pro-S binding pose of
8a (Fig. 5a) showed that the carbonyl oxygen of 8a ligated to
the catalytic zinc with a distance of 2.5 Å and was located
within proton transfer distance from hydroxy oxygen of
Ser47 (2.7 Å). The observed distance between C4 of NADH
pro-S
pro-R
a
b
S47
S47
F56
S47
8a
8a
W288
W288
F56
L122
L122
Fig. 5 Comparison of pro-S and pro-R binding poses of challenging
aliphatic ketone 8a in docking model of GcAPRD active site. Catalytic
zinc and NADH are shown in grey sphere and purple line, respectively.
S47 is shown as blue line. The C4 of NADH with hydride to be
transferred to the ketone substrate is highlighted with white circle. a
pro-S binding pose of 8a shown in yellow stick (productive, 2.5 Å
distance of 8a carbonyl oxygen to catalytic zinc (black dashed-line)).
The important C-H⋯π interaction of that stabilizes ethyl side chain to
Trp288 is shown in red dashed-line. b pro-R binding pose of 8a shown in
orange stick (non-productive, 4.2 Å distance of 8a carbonyl oxygen to
catalytic zinc (black dashed-line))

Appl Microbiol Biotechnol
and carbonyl carbon was 3.6 Å, and the angle between C4 of
NADH carbonyl carbon and carbonyl oxygen of 8a was also
ideal for hydride nucleophilic attack (Ramaswamy et al. 1994;
Baker et al. 2009; Plapp and Ramaswamy 2012; Plapp et al.
2016). It could be seen that the (smaller) ethyl side chain was
located inside the small binding pocket in the vicinity of res-
idue Trp288. The C1 atom of 8a was located 3.3 Å from the
center of the phenyl ring of Trp288 so that a C-H⋯π interac-
tion may occur. The (larger) propyl side chain was located
inside the large binding pocket where it was stabilized through
hydrophobic interactions with Phe56 and Leu122. We could
also obtain a pro-R binding pose of 8a with reasonable energy
(Fig. 5b). In this binding pose, 8a docked with an inverted
binding orientation where the small binding pocket was occu-
pied by propyl instead of ethyl. However, the carbonyl oxygen
of substrate was located approximately 4.2 Å from the cata-
lytic zinc (Fig. 5b), rendering it non-productive.
The binding pose of (S)-8b (Fig. S5a) showed that the
hydroxy oxygen of 8b was ligated to the catalytic zinc with
a distance of 2.4 Å and it was located within proton transfer
distance from Ser47 (3.0 Å from the hydroxy group). The
observed distance between C4 of NAD+ and the alpha carbon
was 3.7 Å, and the angle between C4 of NAD+, alpha carbon,
and hydroxy oxygen of 8b was also ideal for hydride transfer
to NAD+. In contrast, for (R)-8b, the distance of substrate
hydroxy oxygen to the catalytic zinc (3.9 Å) was large, ren-
dering it non-productive (Fig. S5b). Similar docking results
were obtained for pro-S and pro-R 1a (Fig. S6), (S)-1b, and
(R)-1b (Fig. S7). Because the docking simulations suggested
that both (S)-8b and (R)-8b could bind to GcAPRD, we con-
ducted inhibition studies and found out that (R)-8b acts as a
competitive inhibitor for (S)-8b oxidation with an inhibition
constant of 1.04 mM (Table S5). Additionally, we examined if
the bulky-bulky ketone 12a can be an inhibitor for the reduc-
tion of 7a and found out that 12a did not act as a competitive
inhibitor (Table S5).
bulky ketone was not found to be the inhibitor of ethyl ketone
reduction. Thus, the inactiveness of bulky-bulky ketones was
likely caused by the steric hindrance of these substrates
(Lavandera et al. 2008), hindering their binding to the active
site. The enantioselectivity was > 99% (S) ee for most of the
substrates tested. Remarkably, GcAPRD reduced 1a and 8a to
their corresponding (S)-alcohols with 94% ee and > 99% ee,
respectively. The enantioselectivity obtained in this study was
higher than any reported in the literature for these two sub-
strates (Table S6). Although some of the earlier reports sug-
gested the high enantioselectivity could be due to the ability of
the enzymes to distinguish one carbon difference, none pro-
vided a mechanism supported by experimental structure de-
termination and computational docking.
In this study, we explained the mechanism leading to the
observed enantioselectivity of 1a and 8a reduction by first
solving the crystal structure of GcAPRD. Comparison of
GcAPRD with its closest homolog (RCR) showed that al-
though both enzymes have similar folds, the amino acids near
the active site that form its binding pockets are unique. These
differences resulted in different binding pocket shapes and
may result in different substrate binding properties. One of
the most apparent differences was the rotamer angle of
Trp288 in the active site of GcAPRD limiting the size of the
small pocket (Trp286 in the active site of RCR), which was
approximately 180°. Trp286 of RCR was determined to be an
important residue for RCR enantioselectivity (Wang et al.
2014). GcAPRD has better enantioselectivity in reducing ali-
phatic ketones 2a-5a (ee > 99% (S)) compared with RCR (ee
78–95% (S)) and this difference might be attributed to the
Trp288 rotamer angle, binding pocket shape, or both (Nie
et al. 2018). These structural and catalytic property differences
suggested that the GcAPRD structure is novel among other
medium-chain ADHs. Active site Trp has also been investi-
gated in Thermoanaerobium brockii ADH (TbADH) (Maria-
Solano et al. 2017) and Sporobolomyces salmonicolor reduc-
tase (SSCR) (Zhang et al. 2015b) as a crucial residue for
substrate recognition size-determinant of the binding pocket,
affecting the enzyme’s substrate specificity and
enantioselectivity.
Discussion
Ketones possessing two small aliphatic side chains adjacent to
the carbonyl group have been challenging substrates for asym-
metric reduction by biological or non-biological catalyst.
Substrates such as 2-butanone (1a, methyl ethyl ketone) and
3-hexanone (8a, ethyl propyl ketone) can be considered to be
very challenging because they only have one carbon differ-
ence in the side chain lengths. In an effort to reduce challeng-
ing aliphatic ketones, we utilized GcAPRD which is known to
reduce ketone substrates possessing large side-chain size dif-
ferences with high enantioselectivity (Yamamoto et al. 2013).
In this study, it was found that GcAPRD was able to reduce
methyl (1a-6a) and ethyl (7a-11a) ketones with high yields,
but not propyl (12a-14a) or butyl (15a) ketones. The bulky-
Following crystal structure determination, we docked 1a
and 8a to the active site of GcAPRD. The hydride in NADH
is transferred to the re-face of to produce (S)-alcohol. In ac-
cordance with the literature (Keinan et al. 1986; Loderer et al.
2015), to form (S)-alcohol (pro-S binding pose), the smaller
and larger side chain of ketone have to be inside the small and
large pocket of GcAPRD, respectively. It is suggested that
although 1a and 8a can bind in both pro-S and pro-R pose,
only pro-S binding of 8a is productive. One of the important
interactions in the pro-S pose is the C-H⋯π interaction be-
tween C1 of 8a and Trp288. The distance calculated for the C-
H⋯π interaction (3.3 Å) is less than the average distance for
the C-H⋯π interaction (3.7 Å) calculated in over 400 other

Appl Microbiol Biotechnol
crystal structures with a resolution of 2.0–2.5 Å (Brandl et al.
2001). It can be concluded that the C-H⋯π interaction of
Trp288 with the smaller substrate side chain plays an impor-
tant role in the stabilization of the side chain located in the
small binding pocket. This ensures proper orientation of ke-
tones to their re-face such that the (S)-alcohol will be pro-
duced. Similar result of substrate stabilization by the C-
H⋯π interaction contributed by active site Trp has been re-
ported in the reduction of substituted 4-alkylidene cyclohexa-
none by Thermoanaerobium brockii ADH active site Trp
(Maria-Solano et al. 2017).
In contrast, in pro-R binding poses, 1a and 8a bind in a
reversed orientation such that the small binding pocket is oc-
cupied with the larger side chain instead of the smaller side
chain. The flip of the side-chain position has been recently
described as a mechanism to produce enantioselectivity inver-
sion in ADHs (Nie et al. 2018). However, in the pro-R binding
poses of 1a and 8a, the carbonyl oxygen is more than 4.0 Å
away from catalytic zinc, making it impossible to coordinate
with the catalytic zinc. Coordination of catalytic zinc with
carbonyl oxygen is crucial in the catalysis mechanism of
ADHs since the zinc ion will stabilize the formation of an
alkoxide ion during catalysis (Baker et al. 2009). Moreover,
the observed angles between C4 of NADH, carbonyl carbon,
and carbonyl oxygen of 1a and 8a were not ideal for hydride
transfer (less than 50°). A combination of these non-ideal
restraints makes the pro-R binding pose non-productive.
We also docked the corresponding alcohols (S)-1b, (R)-1b,
(S)-8b, and (R)-8b. All of the (S)-alcohol binding poses with
reasonable binding energies were productive, but not the (R)
binding poses, due to their large distance between the sub-
strate hydroxy oxygen and catalytic zinc. The similar Km value
for (S)-8b and (R)-8b (Table 1) indicates that both substrates
can bind to GcAPRD; however, the difference in their Kcat
shows that the (R)-8b reaction is not efficient due to non-
productive binding. This result is supported by the fact that
(R)-8b acts as a competitive inhibitor for (S)-8b oxidation
(Table S5).
The structure of GcAPRD is globally similar but locally dis-
tinct from other members of the medium-chain ADH family.
This is the first study to describe the mechanism of substrate
recognition that can distinguish one carbon difference be-
tween two side chains adjacent to the carbonyl carbon in chal-
lenging aliphatic ketones. It is suggested that although chal-
lenging aliphatic ketones can bind with pro-S and pro-R pose,
the pro-R pose is not able to yield (R)-alcohols, thus resulting
in excellent (S)-enantioselectivity. Residue Trp288 was con-
sidered to play important role for the excellent
enantioselectivity of GcAPRD by limiting the size of the small
binding pocket as well as forming a C-H⋯π interaction sta-
bilizing the pro-S binding pose. We clarified that a common
mechanistic explanation is applicable to ketones with both
small and large side-chain size differences. Knowing the
mechanism leading to the observed enantioselectivity of
GcAPRD will be very beneficial for the future implementation
of GcAPRD as a sustainable catalyst to produce pharmaceu-
tical intermediates.
Funding information This study was funded by the Japan Society for the
Promotion of Science (grant number JP16K05864)
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
Ethical approval This article does not contain any studies with human
participants or animals performed by any of the authors.
References
Adams PD, Afonine PV, Bunkóczi G, Chen VB, Davis IW, Echols N,
Headd JJ, Hung LW, Kapral GJ, Grosse-Kunstleve RW, McCoy AJ,
Moriarty NW, Oeffner R, Read RJ, Richardson DC, Richardson JS,
Terwilliger TC, Zwart PH (2010) PHENIX: a comprehensive
Python-based system for macromolecular structure solution. Acta
Crystallogr Sect D Biol Crystallogr 66:213–221. https://doi.org/10.
1107/S0907444909052925
The mechanism leading to the observed enantioselectivity
for ketones or alcohols with large side chains size differences,
such as the oxidation 1-phenylethanol (16b), has been de-
scribed previously (Hamnevik et al. 2014). The oxidation of
(S)-16b is inhibited by the presence of (R)-16b (Höffken et al.
2006; Dudzik et al. 2014) and this phenomenon was also
observed in oxidation of (S)-16b by GcAPRD. The preference
of oxidizing (S)-alcohol is caused by productive S-alcohol
binding pose and non-productive R-alcohol binding pose
(Hamnevik et al. 2014). We clarified that the same mechanis-
tic explanation is applicable for alcohol with similar side
chains where (R)-alcohol binding is non-productive and in-
hibits (S)-alcohol oxidation.
Agudo R, Roiban GD, Reetz MT (2013) Induced axial chirality in bio-
catalytic asymmetric ketone reduction. J Am Chem Soc 135:1665–
1668. https://doi.org/10.1021/ja3092517
Baker PJ, Britton KL, Fisher M, Esclapez J, Pire C, Bonete MJ, Ferrer J,
Rice DW (2009) Active site dynamics in the zinc-dependent medi-
um chain alcohol dehydrogenase superfamily. Proc Natl Acad Sci
106:779–784. https://doi.org/10.1073/pnas.0807529106
Bandeira PT, Alnoch RC, De Oliveira ARM, De Souza EM, De OPF,
Krieger N, Piovan L (2016) Enzymatic kinetic resolution of aliphatic
sec-alcohols by LipG9, a metagenomic lipase. J Mol Catal B Enzym
125:58–63. https://doi.org/10.1016/j.molcatb.2015.12.010
Boukachabia M, Zeror S, Collin J, Fiaud JC, Zouioueche LA (2011)
Screening method for the evaluation of asymmetric catalysts for
the reduction of aliphatic ketones. Tetrahedron Lett 52:1485–1489.
https://doi.org/10.1016/j.tetlet.2011.01.112
Brandl M, Weiss MS, Jabs A, Su È, Jena D (2001) C-H…π interactions in
proteins. J Mol Biol 307:357–377. https://doi.org/10.1006/jmbi.
2001.4473
In conclusion, we found that GcAPRD has excellent
enantioselectivity to reduce challenging aliphatic ketones.

Appl Microbiol Biotechnol
Dhoke GV, Loderer C, Davari MD, Ansorge-Schumacher M,
Schwaneberg U, Bocola M (2015) Activity prediction of substrates
in NADH-dependent carbonyl reductase by docking requires cata-
lytic constraints and charge parameterization of catalytic zinc envi-
ronment. J Comput Aided Mol Des 29:1057–1069. https://doi.org/
10.1007/s10822-015-9878-8
dehydrogenase. Org Biomol Chem:11–14. https://doi.org/10.1039/
C7OB00482F
Matsuda T, Nakajima Y, Nakamura K (2002) Asymmetric reduction of
simple aliphatic ketones with dried cells of Geotrichum candidum.
Tetrahedron Asymmetry 13:971–974. https://doi.org/10.1016/
S0957-4166(02)00226-4
Dudzik A, Snoch W, Borowiecki P, Opalinska-Piskorz J, Witko M,
Heider J, Szaleniec M (2014) Asymmetric reduction of ketones
and β-keto esters by (S)-1-phenylethanol dehydrogenase from
denitrifying bacterium Aromatoleum aromaticum. Appl Microbiol
Biotechnol 99:5055–5069. https://doi.org/10.1007/s00253-014-
6309-z
Emsley P, Cowtan K (2004) Coot: model-building tools for molecular
graphics. Acta Crystallogr Sect D Biol Crystallogr 60:2126–2132.
https://doi.org/10.1107/S0907444904019158
Ferreira HV, Rocha LC, Severino RP, Porto ALM (2012) Syntheses of
enantiopure aliphatic Secondary alcohols and acetates by
bioresolution with lipase b from Candida antarctica. Molecules
17:8955–8967. https://doi.org/10.3390/molecules17088955
Gröger H, Hummel W, Borchert S, Kraußer M (2012) Reduction of
ketones and aldehydes to alcohols. In: Enzyme catalysis in organic
synthesis, 3rd edn. Wiley-VCH, Weinheim, pp 1037–1110
Hamnevik E, Blikstad C, Norrehed S, Widersten M (2014) Kinetic char-
acterization of Rhodococcus ruber DSM 44541 alcohol dehydroge-
nase A. J Mol Catal B Enzym 99:68–78. https://doi.org/10.1016/j.
molcatb.2013.10.023
Harada T, Osawa T (1995) Enantio-differentiating hydrogenation of 2-
butanone : distinction between CH3 and C2H5 with a modified nick-
el catalyst. In: Jannes G, Dubois V (eds) Chiral reactions in hetero-
geneous catalysis. Springer US, Boston, pp 83–88
Höffken HW, Duong M, Friedrich T, Breuer M, Hauer B, Reinhardt R,
Rabus R, Heider J (2006) Crystal structure and enzyme kinetics of
the (S)-specific 1-phenylethanol dehydrogenase of the denitrifying
bacterium strain EbN1. Biochemistry 45:82–93. https://doi.org/10.
1021/bi051596b
Holm L, Laakso LM (2016) Dali server update. Nucleic Acids Res 44:
W351–W355. https://doi.org/10.1093/nar/gkw357
Johnson MR, Rickborn B (1970) Sodium borohydride reduction of con-
jugated aldehydes and ketones. J Organomet Chem 35:1041–1045.
https://doi.org/10.1021/jo00829a039
Matsuda T, Yamanaka R, Nakamura K (2009) Recent progress in bioca-
talysis for asymmetric oxidation and reduction. Tetrahedron
Asymmetry 20:513–557. https://doi.org/10.1016/j.tetasy.2008.12.
035
Murai T, Matsuoka D, Morishita K (2006) 1,1’-Binaphthyl-2,2’-diyl
phosphoroselenoyl chloride as a chiral molecular tool for the prep-
aration of enantiomerically pure alcohols and amines. J Am Chem
Soc 128:4584–4585. https://doi.org/10.1021/ja060308b
Nakamura K, Matsuda T (1998) Asymmetric reduction of ketones by the
acetone powder of Geotrichum candidum. J Organomet Chem 63:
8957–8964. https://doi.org/10.1021/jo9812779
Nakata Y, Fukae T, Kanamori R, Kanamaru S, Matsuda T (2010)
Purification and characterization of acetophenone reductase with
excellent enantioselectivity from Geotrichum candidum NBRC
4597. Appl Microbiol Biotechnol 86:625–631. https://doi.org/10.
1007/s00253-009-2329-5
Ni Y, Xu JH (2012) Biocatalytic ketone reduction: a green and efficient
access to enantiopure alcohols. Biotechnol Adv 30:1279–1288
Nie Y, Wang S, Xu Y, Luo S, Zhao YL, Xiao R, Montelione GT, Hunt JF,
Szyperski T (2018) Enzyme engineering based on x-ray structures
and kinetic profiling of substrate libraries: alcohol dehydrogenases
for stereospecific synthesis of a broad range of chiral alcohols. ACS
Catal 8:5145–5152. https://doi.org/10.1021/acscatal.8b00364
Noey EL, Tibrewal N, Jiménez-Osés G, Osuna S, Park J, Bond CM,
Cascio D, Liang J, Zhang X, Huisman GW, Tang Y, Houk KN
(2015) Origins of stereoselectivity in evolved ketoreductases. Proc
Natl Acad Sci 112(51):E7065–E7072. https://doi.org/10.1073/pnas.
1507910112
Plapp BV (2011) Conformational changes and catalysis by alcohol dehy-
drogenase. Arch Biochem Biophys 493:3–12. https://doi.org/10.
1016/j.abb.2009.07.001.Conformational
Plapp BV, Ramaswamy S (2012) Atomic-resolution structures of horse
liver alcohol dehydrogenase with NAD+ and fluoroalcohols define
strained michaelis complexes. Biochemistry 51:4035–4048. https://
doi.org/10.1021/bi300378n
Plapp BV, Charlier HA, Ramaswamy S (2016) Mechanistic implications
from structures of yeast alcohol dehydrogenase complexed with
coenzyme and an alcohol. Arch Biochem Biophys 591:35–42.
https://doi.org/10.1016/j.abb.2015.12.009
Kabsch W (2010) XDS. Acta Crystallogr Sect D Biol Crystallogr 66:125–
132. https://doi.org/10.1107/S0907444909047337
Karlsson A, El-Ahmad M, Johansson K, Shafqat J, Jörnvall H, Eklund H,
Ramaswamy S (2003) Tetrameric NAD-dependent alcohol dehy-
drogenase. Chem Biol Interact 143-144:239–245. https://doi.org/
10.1016/S0009-2797(02)00222-3
Prelog V (1964) Specification of the stereospecificity of some oxido-
reductases by diamond lattice sections. Pure Appl Chem 9:119–130
Raj SB, Ramaswamy S, Plapp BV (2014) Yeast alcohol dehydrogenase
structure and catalysis. Biochemistry 53:5791–5803. https://doi.org/
10.1021/bi5006442
Ramaswamy S, Eklund H, Plapp BV (1994) Structures of horse liver
alcohol dehydrogenase complexed with NAD+ and substituted ben-
zyl alcohols. Biochemistry 33:5230–5237. https://doi.org/10.1021/
bi00183a028
Rosenthal K, Lütz S (2018) Recent developments and challenges of bio-
catalytic processes in the pharmaceutical industry. Curr Opin Green
Sustain Chem 11:58–64. https://doi.org/10.1016/j.cogsc.2018.03.
015
Keinan E, Hafeli EK, Seth KK, Lamed R (1986) Thermostable enzymes
in organic synthesis. 2. Asymmetric reduction of ketones with alco-
hol dehydrogenase from Thermoanaerobium brockii. J Am Chem
Soc:162–169. https://doi.org/10.1021/ja00261a026
Lavandera I, Kern A, Ferreira-silva B, Glieder A, de Wildeman S, Kroutil
W (2008) Stereoselective bioreduction of bulky-bulky ketones by a
novel adh from Ralstonia sp. J Organomet Chem 73:6003–6005.
https://doi.org/10.1021/jo800849d
Loderer C, Dhoke GV, Davari MD, Kroutil W, Schwaneberg U, Bocola
M, Ansorge-Schumacher MB (2015) Investigation of structural de-
terminants for the substrate specificity in the zinc-dependent alcohol
dehydrogenase CPCR2 from Candida parapsilosis. ChemBioChem
16:1512–1519. https://doi.org/10.1002/cbic.201500100
Shende VS, Singh P, Bhanage BM (2018) Recent trends in
organocatalyzed asymmetric reduction of prochiral ketones. Catal
Sci Technol 8:955–969. https://doi.org/10.1039/c7cy02409f
Shimoda K, Kubota N, Hamada H, Yamane SY, Hirata T (2004)
Asymmetric transformation of enones with Synechococcus sp.
PCC 7942. Bull Chem Soc Jpn 77:2269–2272. https://doi.org/10.
1246/bcsj.77.2269
Man H, Loderer C, Ansorge-Schumacher MB, Grogan G (2014)
Structure of NADH-dependent carbonyl reductase (CPCR2) from
Candida parapsilosis provides insight into mutations that improve
catalytic properties. ChemCatChem 6:1103–1111. https://doi.org/
10.1002/cctc.201300788
Maria-Solano MA, Romero-Rivera A, Osuna S (2017) Exploring the
reversal of enantioselectivity on a zinc-dependent alcohol

Appl Microbiol Biotechnol
Sugiyama Y, Senda M, Senda T, Matsuda T (2015) Crystallization and
preliminary crystallographic analysis of acetophenone reductase
from Geotrichum candidum NBRC 4597. Acta Crystallogr Sect F
Struct Biol Commun 71:320–323. https://doi.org/10.1107/
S2053230X15002265
Trott O, Olson A (2010) Autodock vina: improving the speed and accu-
racy of docking. J Comput Chem 31:455–461. https://doi.org/10.
1002/jcc.21334.AutoDock
Truppo MD, Pollard D, Devine P (2007) Enzyme-catalyzed
enantioselective diaryl ketone reductions. Org Lett 9:335–338.
https://doi.org/10.1021/ol0627909
Velonia K, Tsigos I, Bouriotis V, Smonou I (1999) Stereospecificity of
hydrogen transfer by the NAD+-linked alcohol dehydrogenase from
the Antarctic psychrophile Moraxella sp. TAE123. Bioorg Med
Chem Lett 9:65–68. https://doi.org/10.1016/S0960-894X(98)
00678-7
Vidal LS, Kelly CL, Mordaka PM, Heap JT (2018) Review of NAD(P)H-
dependent oxidoreductases: properties, engineering and application.
Biochim Biophys Acta, Proteins Proteomics 1866:327–347. https://
doi.org/10.1016/j.bbapap.2017.11.005
Wang S, Nie Y, Xu Y, Zhang R, Ko T-P, Huang C-H, Chan H-C, Guo R-T,
Xiao R (2014) Unconserved substrate-binding sites direct the
stereoselectivity of medium-chain alcohol dehydrogenase. Chem
Commun (Camb) 50:7770–7772. https://doi.org/10.1039/
c4cc01752h
Yamamoto T, Nakata Y, Cao C, Sugiyama Y, Asanuma Y, Kanamaru S,
Matsuda T (2013) Acetophenone reductase with extreme stability
against a high concentration of organic compounds or an elevated
temperature. Appl Microbiol Biotechnol 97:10413–10421. https://
doi.org/10.1007/s00253-013-4801-5
Yoshimura M, Tanaka S, Kitamura M (2014) Recent topics in catalytic
asymmetric hydrogenation of ketones. Tetrahedron Lett 55:3635–
3640. https://doi.org/10.1016/j.tetlet.2014.04.129
Zhang AL, Da YZ, Yang LW, Yang NF (2015a) Synthesis of several
polyethers derived from BINOL and their application in the asymmet-
ric borane reduction of prochiral ketones. Tetrahedron Asymmetry 26:
173–179. https://doi.org/10.1016/j.tetasy.2014.12.012
Zhang D, Chen X, Chi J, Feng J, Wu Q, Zhu D (2015b) Semi-rational
engineering a carbonyl reductase for the enantioselective reduction
of β-amino ketones. ACS Catal 5:2452–2457. https://doi.org/10.
1021/acscatal.5b00226
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.