Rationalizing the Unprecedented Stereochemistry of an Enzymatic Nitrile Synthesis through a Combined Computational and Experimental Approach

Article abstract of DOI:10.1002/anie.202017234

In this contribution, the unique and unprecedented stereochemical phenomenon of an aldoxime dehydratase-catalyzed enantioselective dehydration of racemic E- and Z-aldoximes with selective formation of both enantiomeric forms of a chiral nitrile is rationalized by means of molecular modelling, comprising in silico mutations and docking studies. This theoretical investigation gave detailed insight into why with the same enzyme the use of racemic E- and Z-aldoximes leads to opposite forms of the chiral nitrile. The calculated mutants with a larger or smaller cavity in the active site were then prepared and used in biotransformations, showing the theoretically predicted decrease and increase of the enantioselectivities in these nitrile syntheses. This validated model also enabled the rational design of mutants with a smaller cavity, which gave superior enantioselectivities compared to the known wild-type enzyme, with excellent E-values of up to E>200 when the mutant OxdRE-Leu145Phe was utilized.

Full text of DOI:10.1002/anie.202017234

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Rationalizing the unprecedented stereochemistry of an enzymatic
nitrile synthesis through a combined computational and
experimental approach
Authors: Harald Gröger, Hilmi Yavuzer, and Yasuhisa Asano
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202017234
Link to VoR: https://doi.org/10.1002/anie.202017234

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
Rationalizing the unprecedented stereochemistry of an enzymatic
nitrile synthesis through a combined computational and
experimental approach
Hilmi Yavuzer,[a] Yasuhisa Asano[b] and Harald Gröger*,[a]
[a]
H. Yavuzer, Prof. H. Gröger
Faculty of Chemistry
Bielefeld University
Universitätsstraße 25, 33615 Bielefeld, Germany
E-mail: harald.groeger@uni-bielefeld.de
Prof. Y. Asano
[b]
Biotechnology Research Center
Toyama Prefectural University
5180 Kurokawa, Imizu, Toyama 939-0398, Japan
Supporting information for this article is given via a link at the end of the document.
Abstract: In this contribution, the unique and unprecedented
stereochemical phenomenon of an aldoxime dehydratase-catalyzed
enantioselective dehydration of racemic E- and Z-aldoximes with
formation of both opposite enantiomeric forms of a chiral nitrile has
been rationalized by means of molecular modelling, comprising in
silico mutations and docking studies. This theoretical investigation
gave a detailed insight why with the same enzyme the use of racemic
E- and Z-aldoximes leads to opposite forms of the chiral nitrile. The
calculated mutants with an increased and decreased cavity in the
active site have then been prepared and used in biotransformations,
showed the theoretically predicted decrease and increase of the
enantioselectivities in these nitrile syntheses. This validated model
also enabled the rational design of mutants with a decreased size of
the cavity, which gave superior enantioselectivities compared to the
known wild-type enzyme with excellent E-values of up to E>200 when
utilizing the mutant OxdRE-Leu145Phe.
it should be added that in practice such a mirror image protein
might not necessarily be required as often other proteins leading
to the opposite enantiomer products exist).3
However, very recently we identified a unique enzymatic
transformation which gives access to both enantiomeric forms of
a chiral nitrile although starting from the same racemic aldehyde
and utilizing the same enzyme.4 This unprecedented
stereochemical phenomenon turned out to have its origin in the
formation of racemic E- and Z-aldoximes as intermediates by
simple condensation of the aldehyde as starting material and
hydroxylamine. These racemic E- and Z-aldoximes serve as the
“real” substrate for the enzyme and are enantioselectively
dehydrated to the chiral nitrile. It is noteworthy that the
enantiopreference of this enzymatic resolution then surprisingly
does not depend on the absolute configuration of the stereogenic
center at the aldoxime, but on the E- or Z-conformation of the
aldoxime. In detail, in the presence of the same enzyme, using
the E-racemate as substrate then furnished the (S)-nitrile,
whereas the (R)-nitrile was formed when starting from the Z-
racemate (Scheme 1).4 Thus, when separating these E- and Z-
aldoximes, both racemates undergo dehydration with formation of
the opposite enantiomers of the nitrile products (whereas using
the non-separated racemic E/Z-mixtures would lead to more or
less racemic nitrile products).
The biocatalysts being capable of this unique transformation
are aldoxime dehydratases.5,6 These types of enzymes were
found three decades ago, but still their number is limited and not
much is known about them yet.7,8 A notable feature is their active
site which combines the structural unit of a catalytic triad
consisting of an arginine, histidine and serine with a heme moiety.
From a practical point of view, it is noteworthy that the conversion
of aldoximes to the corresponding nitriles by release of water
proceeds without the need of any additional cofactor. In addition,
the substrate scope proved to be very broad,4–13 and among the
resulting aliphatic and chiral nitriles are various important
products for the chemical industry.
Introduction
Designing stereochemical processes is of utmost importance
for access to chiral building blocks needed for the production of
chiral drugs.1–3 While over the last decades numerous impressive
examples with chiral chemocatalysts as well as enzymes, both
representing fascinating chiral catalysts,1–3 exist, at the same time
“by definition” there is one general limitation for any type of
stereochemical
process.
When
starting
from
one
(enantio)selective catalyst and a specific substrate (which can be
either prochiral or racemic), only one enantiomeric form of the
product will be obtained. This is true for any type of asymmetric
catalysis starting from prochiral substrates as well as for
resolutions starting from racemic substrates.1–3 In other words,
when having only one enantiomerically pure form of a chiral
catalyst in hand, only one enantiomeric form of the product will be
accessible in the corresponding transformation. This, however,
can turn out as a limitation in particular for the field of
biocatalysis,3 since only one enantiomeric form of the proteins
(based on the L-amino acids) is available whereas the mirror
image based on D-amino acids is not formed in nature (although
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
chiral) substrates. Furthermore, crystal structures of two different
conformations of the OxdRE are known, namely an “open” and “a
closed” conformation. Since the available co-crystals were only
found in “closed” conformation, the docking was performed using
the closed conformation of OxdRE as this conformation appeared
to be the one being relevant for the catalytic cycle.
Table 1. Biotransformations of various fluoro-substituted phenyl-2-propanal
oximes as E- or Z-isomers with OxdRE-WT
Scheme
1. Enantioselective dehydration of aldoximes towards the
corresponding nitrile.
To gain insight into this exciting and very unusual
stereochemical behavior of this class of aldoxime dehydratases
(which are at the same time attractive catalysts for chiral nitrile
synthesis without the need for highly toxic cyanides), molecular
modelling studies were performed, including in silico mutations
and docking studies. In order to prove such data on rationalizing
the selectivity of an aldoxime dehydratase, we combined these
computational studies with laboratory experiments by preparing
and characterizing the theoretically calculated enzyme mutants.
Based on a general postulated mechanism for an aldoxime
dehydratase by Nomura et al.14 we focused on rationalizing this
unusual switch in enzyme selectivity by means of docking
#
Name
OxdRE-WT
Con. /%
FPPOX
ee /%
E-
∆∆G /
kcal·mol-1
Value
1
45
50
96 (S)
112
2.6
experiments. As
a
software MOE (Molecular Operating
Environment)15 was used as a powerful tool to find suitable ligand-
protein conformations. The software has already proven highly
successful for a range of applications.16–21 With the combination
of wet experiments including a variety of mutants and substrates
we were able to understand this unique and unprecedented
behavior regarding the enantioselectivity of these enzymes.
Furthermore, based on this rational insight into the
stereochemical properties of such enzymes we were capable to
calculate mutants with superior and dramatically increased
enantioselectivities.
2
65 (R)
9
1.2
3
4
48
52
96 (S)
64 (R)
146
9
2.7
1.2
5
6
34
32
93 (S)
54 (R)
44
4
2.1
0.7
Results and Discussion
In order to characterize the stereoselectivity of the aldoxime
dehydratase from Rhodococcus sp. N-771 (OxdRE-WT)23 for a
model substrate as a basis for the molecular modelling studies, at
first ortho-, meta- and para-fluoro-substituted phenylpropanal
oximes (rac-2FPPOX, rac-3FPPOX, rac-4FPPOX) were
synthesized and purified as racemic E- and Z-isomers according
to previously developed protocols.4,22 The isolated racemic E- and
Z-isomers were subsequently used in biotransformations with E.
coli whole cells containing the OxdRE in recombinant form. The
resulting conversions and enantioselectivities of the formed chiral
nitriles are shown in Table 1. The reactions with this wild-type
enzyme proceed as expected, thus furnishing the (S)-nitrile when
starting from the E- isomer whereas utilizing the Z-isomer led to
the (R)-nitrile. The determined E-values and the corresponding
calculated energy values of these biotransformation experiments
are given in Table 1.
This aldoxime dehydratase from Rhodococcus sp. N-771
(OxdRE-WT)23 has been chosen for our studies since crystal
structures are available for this enzyme (PDB 3a15, 3a16,
3a17)24 ,which served as an ideal starting point for the molecular
modelling of the transition states for the conversion of the various
E- and Z-isomers. In detail, three different crystal structures are
available, including two structures being co-crystallized with (non-
In order to find out which step in the reaction determines the
selectivity, normally all intermediate states must be calculated. As
in our case the enantioselectivity of the enzyme depends on the
E- or Z-isomer of the substrate and due to the fact that this E/Z
information only can play a role in the initial step of the catalytic
mechanisms during the binding phase of the ligand, we calculated
the binding energies of the ligands to rationalize the
stereoselectivity of this enzyme class. The MM force field
calculation was performed using data generated from preliminary
experiments8 and from this work. To find the correct protein-
ligand-structures (pose) we used the structure of the co-crystal24
with n-propanal oxime as ligand as a basis for our study. Based
on the distances and angles of the co-crystal structure and the
mechanism for the dehydration,14 we determined cut off values for
a correct pose (Figure 1a). With the redocking of the n-propanal
oxime we could evaluate and prove our docking model and
calculation.
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
The ligand shows a hydrogen donation to the Ser219 and the
His320 donates a hydrogen towards the oxygen of the ligand,
while the nitrogen is coordinated by the FeII-metal center of the
heme. The phenyl group of the ligand interacts with the porphyrin
ring via π-π-stacking and the aldoxime function for each isomer is
always in the same position. It is noteworthy that by means of this
modelling tool we were then not only able to predict the enantio-
preference, but also to determine quantitatively the enantio-
selectivity of any aldoxime substrate of this study (Table 2).
a
Table 2. Comparison of docking experiments with OxdRE (3a17) and 2’-, 3’-,
and 4’-fluorophenyl-2-propanal oximes FPPOX and the experimental
biotransformation of OxdRE-WT with the various aldoximes.
#
Aldo-
xime
∆G /
kcal·mol-1
∆∆G /
kcal·mol-1
Pred.
(R)
Exp.
(R)
1
2F-ZS
2F-ZR
2F-ES
2F-ER
3F-ZS
3F-ZR
3F-ES
3F-ER
4F-ZS
4F-ZR
4F-ES
4F-ER
-7.30
-9.17
-6.24
-4.60
-7.68
-9.66
-6.28
-4.77
-7.07
-9.72
-5.88
-4.96
1.87
1.63
1.98
1.51
2.65
0.93
2
b
3
(S)
(S)
4
5
(R)
(S)
(R)
(S)
(R)
(S)
(R)
(S)
Figure 1. (a) Evaluating the docking results with redocking of n-propanal oxime;
(b) pharmacophore generated with n-propanal oxime.
6
7
To further refine, simplify and automate our docking we used
pharmacophores to obtain the correct pose of our ligands (Figure
1b). The pharmacophore can be used to define the binding motif
such as the position of the functional group and its orientation. For
example, a good pose is not just be described by the presence of
the aldoxime ligand inside the pocket but by the orientation of this
ligand, in which the aldoxime function shows a specific position
with specified angles (His-O-Ser ~115°), dihedrals (C-N-O-O,
~85°) and distances (Fe-N, ~2.5 Å; O-Ser, ~2.8 Å; O-His, 2.7 Å)
towards the heme group and the catalytic triad. Only when these
prerequisites are fulfilled, the subsequent dehydration step can
8
9
10
11
12
ΔG: binding energy from docked ligand, ΔΔG difference of binding
energies between two enantiomers of one isomer. Lower binding energy
of one enantiomer corresponds to the formed enantiomer.
In detail, we calculated the difference (∆∆G) of the aldoxime
ligand binding energies (∆G) in order to predict the enantio-
preference and enantioselectivity. For example, the binding
energy of (S,E)-3F of -6.28 kcal·mol-1 is lower than the binding
energy of the opposite enantiomer (R,E)-3F with -4.77 kcal·mol-1
(Table 2, entries 7,8). Notably, the resulting calculated energy
difference (ΔΔG) of 1.51 kcal·mol-1 corresponds very well with the
experimental value of the enantioselectivity and the preferred
formation of the S-enantiomer. In general, the ∆∆G-values were
used to determine the enantioselectivity and even though our
method is based on force field it is noteworthy that with this type
of molecular modelling it was not only possible to rationalize the
enantiopreference but also to predict the degree of the
enantioselectivity for any investigated substrate in an accurate
fashion. The high accuracy is underlined when comparing the
∆∆G-values from the docking experiments (Table 2) with those
obtained from the biotransformation experiments (Table 1)
proceed. With these defined parameters in hand,
a
pharmacophore query was created which then allowed us to find
ligand poses with this specified binding motif (see Supporting
Information). Even though metalloenzymes are not well
parameterized25,26 a combination of MMF94 force field and the
scoring functions London dG and Affinity dG proved to be suited
for this purpose. The protein model was prepared using the
preparation kit given by MOE. As a representative example, the
2D-figure in Figure 2 shows the interaction pattern of the enantio-
and diastereomerically pure ligand (Z,R)-3FPPOX with the active
site of the enzyme OxdRE-WT.
demonstrating
a
perfect agreement of the predicted
enantiopreferences and a good agreement of the quantitative
∆∆G-values.
By looking deeper into the final poses of our protein-ligand-
structures and docked aldoxime ligands we could see similarities
between each substrate and its conformers (Figure 3a). The
methyl group is exposed to Met29, Leu145 and the His320. The
general pose of the docked ligand does not vary in dependency
of the position or the type of the halogen substituent at the phenyl
ring. Therefore, the halogen substituent has no direct effect upon
the selectivity. However, it has a certain influence on the
acceptance of a ligand.
Figure 2: Ligand interaction pattern with (Z,R)-3FPPOX as an example bound
in the active site of OxdRE-WT
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
a
b
a
c
d
Figure 4. Protein-ligand-structure (pose) of 3-FPPOX conformers in the active
center of OxdRE-WT. (a) 3-(E,R); (b) 3-(E,S); (c) 3-(Z,S); (d) 3-(Z,R).
c
b
Figure 3. 3-(Z,R)-FPPOX bound in the catalytic center of OxdRE-WT (a),
OxdRE-Leu145Phe (b) and OxdRE-3M (c). The methyl group is in the cavity,
which is formed by the amino acids Met29, Leu145, Ala147 and His320. The
figure shows the docked position of the ligand 3-(Z,R)-FPPOX in three different
OxdRE variants. The Leu145Phe has a smaller cavity, whereas the triple point
mutant 3M (Met29Ala, Leu145Ala, Ala147Gly) has an increased cavity size.
In order to prove our hypothesis that the methyl group is the
main cause for the selectivity and in order to validate our model-
ling study further, we decided to calculate and experimentally
evaluate various mutants with (i) a more enlarged or alternatively
(ii) a tighter cavity in the active site. Accordingly, these mutants
then should lead to a decreased (in case of (i)) or increased (in
case of (ii)) differentiation of the substrate enantiomers and, thus,
to a decrease or increase of the enantioselectivity.
To start with the design of mutants with an enlarged active site
(case (i)), our in silico studies showed that mutants with such an
increased size of the cavity lead to a significant decrease of the
∆∆G-value (see Supporting Information). Consequently, such
mutants then should lead to a lower selectivity. As described
above, the cavity is formed by four amino acids. Leu145 appeared
for us to be the most promising position for mutations due to the
highest steric contribution to the cavity compared to the amino
acids Ala147 and Met29. Thus, replacement of Leu145 with
alanine and glycine should lead to a dramatically enhanced cavity.
Mutating the proximal His320 to glycine or alanine also would
increase the size of the cavity tremendously, but as a part of the
catalytic triad it is not possible to mutate His320 without loss of
the catalytic activity.27
However, the single point mutations of each position did not
lead to a significant decrease of the ∆∆G-value except for the
His320 mutation (in silico). Therefore, we designed double and
triple point mutations, and the triple point mutation showed a more
significant decrease in the ∆∆G-value. Based on the in silico study,
the triple mutant Met29Gly/Leu145Gly/Ala147Gly (OxdRE-3G)
would have represented the most promising mutant. However, in
wet experiments this mutant was found to be unstable and the
expressed protein was only observed in the insoluble fraction (see
Supporting Information). As an alternative we calculated the
mutant Met29Ala/Leu145Ala/Ala147Gly (OxdRE-3M; Table 3 and
Figure 3c). This mutant was then also studied in wet experiments,
and the results turned out to be in excellent agreement with our
theoretical predictions. These data are shown in Table 3 and
Figure 5 and will be discussed in detail below.
The enantiomeric excess can differ because the binding
energies vary with the type and position of the substituent at the
phenyl ring. The methyl group is the only group which differs
among the compared conformers in space and alignment. Based
on that information we propose that the arrangement of the methyl
group has a major impact on the selectivity for a specific aldoxime
isomer. The methyl group is always positioned in a cavity being
formed by the amino acids Met29, Leu145, Ala147 and His320,
but for each conformer in a specific arrangement (Figure 4).
In general, all PPOX-derivatives follow a certain alignment.
The methyl group of the PPOX-derivatives is always in the above-
mentioned cavity. However, while the methyl group of the ligands
with E,R- and Z,S-conformations (not preferred) are closer to the
His320 and Leu145, the methyl group of the ligand with E,S- and
Z,R-conformation (preferred) are more in the middle of the pocket
directing towards the cavity (Figure 4). The available data suggest
that the position and orientation of the methyl group in relation to
the cavity determines the selectivity (Figure 4a). The necessary
alignment is only possible for the E-isomer in S-configuration and
for the Z-isomer only in R-configuration. Due to the scoring
function, however, it is not possible to determine the single
contribution of each Van-der-Waals-(VdW)-interaction. Thus, a
direct determination of the energy contribution of the methyl group
was not possible.
Another contribution leading to the observed selectivity could
be related to the aldoxime function. Due to their small variation in
the alignment, the aldoxime moiety could cause the energy gap
(∆∆G) and consequently the degree of the selectivity. It has been
successfully demonstrated, however, that no correlation between
energy contribution, selectivity and conformation exists.
Therefore, it was concluded that the arrangement of the aldoxime
function is essential for the reaction, but insignificant for the
differentiation between the enantiomeric pairs of an isomer (see
Supporting Information).
As a next step, we focused on a complementary mutation
strategy, which now should create enzyme mutants with a
decreased size of the cavity (case (ii)), thus making the active site
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
more tight. Accordingly, for the resulting biotransformation then
an increased enantioselectivity can be expected. In order to
decrease the size of the cavity, the position Leu145 turned out as
the only option for a suitable mutation. In contrast, the amino acids
Met29 and Ala147 are not suitable for rationally decreasing the
size of the cavity. The Ala147 mutations were all in the wrong
orientation and not exposed to the active site. Met29 was already
one of the longest amino acids, which could reach the cavity. For
Leu145 the most promising in silico mutation was found to be
Leu145Phe (Figure 3b). Other amino acids were also tested in
silico, but they were either too small, too large or changed the
environment of the catalytic triad too much. Our in silico mutation
and the performed docking predicted that the Leu145Phe mutant
would be a promising candidate for a more enantioselective
conversion of PPOX derivatives (Table 3 and Figure 3b).
With these two optimized mutants in hand (addressing case
(i) and (ii)), a detailed theoretical and experimental study was
conducted (Table 3, Figure 5). The docking results with the two
identified most promising mutants with an increased and
decreased cavity, respectively, are shown in Table 3 and Figure
5, Part A). As expected, the ΔΔG-values are much higher for the
Leu145Phe mutant compared to the 3-M mutant, which is in good
agreement with our hypothesis. For example, for the two
enantiomers of the 2FE-PPOX substrate, a much higher ∆∆G-
value of 3.6 kcal·mol-1 was determined in the docking experiment
for the OxdRE-Leu145Phe mutant compared to the mutant
OxdRE-3M (1.7 kcal·mol-1) and also to the wild-type enzyme (2.6
kcal·mol-1) (Table 3, entries 1,2 and Table 2, entry 1). For a better
comparison of the theoretical data with the experimental ones
determined from the biotransformations, we calculated the E-
values from the calculated energy values (∆∆G). Accordingly, in
case of 2FE-PPOX these in silico experiments predict an
excellent E value for the mutant Leu145Phe with E>200, thus
significantly outperforming the wild type enzyme (E=112) as well
as the 3M-mutant (E=22).
In order to evaluate if these predictions and hypothesis are
correct, as next we performed in vitro these mutations of OxdRE
and used the mutants in biotransformations (Figure 5, Part B). We
were pleased to find that the results from this experimental study
proved to be in good agreement with the theoretical data and
revealed interesting results in terms of strongly mutants with
improved enantioselectivities (Figure 5, Part B)). It is noteworthy
that in all biotransformations the theoretically predicted tendency
of the change of the enantioselectivity (E value) when replacing
the wild-type enzyme by the prioritized mutant Leu145Phe as a
biocatalyst was experimentally observed (Figure 5, Parts A and
B), thus also confirming the validity of the theoretical study.
Furthermore, the improved, high enantioselectivities calculated
for the prioritized Leu145Phe-mutant have been found in the
experimental study, leading to excellent E values of >200 in the
biotransformations when starting from the aldoxime substrates
2FE-PPOX and 3FE-PPOX (Figure 5, Parts A and B). From a
practical point it should be added that the mutant OxdRE-
Leu145Phe, which is still able to convert both E- and Z-isomers of
the aldoxime substrate, shows an increased enantioselectivity in
particular for the E-isomer.
Table 3. Comparison of docking result for the FPPOX derivatives with
OxdRE-Leu145Phe and OxdRE-3M. Only isomer pairs with ∆∆G values
are shown.
Name
Leu145Phe
∆∆G /
3M
∆∆G /
#
1
2
3
4
5
6
PPOX
2FE
2FZ
3FE
3FZ
4FE
4FZ
kcal·mol-1
E
>200
65
kcal·mol-1
E
22
1
Pred
(S)
(R)
(S)
(R)
(S)
(R)
3.64
1.73
2.32
0.19
2.90
184
1
1.87
29
6
0.70
1.00
3.85
>200
11
1.54
16
11
1.35
1.34
Figure 5. Part A): Calculated E-values from the ∆∆G values of the docking data
for the FPPOX aldoximes with each enzyme. Part B): Experimental E-values
determined from the biotransformation of the FPPOX aldoximes to their corres-
ponding nitriles with the use of OxdRE-WT, OxdRE-Leu145Phe and OxdRE-
3M. The graphs are normalized to a maximum E-value of 200 (thus, an E value
of 200 is graphically shown even when the determined E-value was >200).
In Figure 5, Part A), a comparison of the docking data for all
FPPOX derivatives with the three different is given. As can be
seen in Figure 5, Part A), the Leu145Phe mutant has in general
higher E-values than the wild type, while the 3M-mutant has lower
E-values. The docking data further predict that in general the
enantioselectivity for the conversion of the Z-isomer is lower than
the conversion of the E-isomer.
Thus, these data also underline the plausibility of our
hypothesis of the impact of active site modification on the
enantioselectivity. Furthermore, we could demonstrate that based
on this rational modelling in silico approach with the MOE
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
[2]
[3]
[4]
[5]
U. T. Bornscheuer, G. W. Huisman, R. J. Kazlauskas, S. Lutz, J. C.
Moore, K. Robins, Nature 2012, 485, 185–194.
software more enantioselective aldoxime dehydratases can be
generated.
K. Drauz, H. Gröger, O. May, (eds.), Enzyme Catalysis in Organic
Synthesis, vol. 1-3, 3. ed., Wiley-VCH, Weinheim, 2012.
T. Betke, P. Rommelmann, K. Oike, Y. Asano, H. Gröger, Angew. Chem.
2017, 129, 12533-12538; Angew. Chem. Int. Ed. 2017, 56, 12361–12366.
T. Betke, J. Higuchi, P. Rommelmann, K. Oike, T. Nomura, Y. Kato, Y.
Asano, H. Gröger, ChemBioChem 2018, 19, 768–779.
H. Gröger, Y. Asano, J. Org. Chem. 2020, 85, 6243–6251.
Y. Kato, K. Nakamura, H. Sakiyama, S. G. Mayhew, Y. Asano,
Biochemistry 2000, 39, 800–809.
Conclusion
In conclusion, the stereochemically unique and to the best of
our knowledge unprecedented stereochemical phenomena of
aldoxime dehydratases, which are able to enantioselectively
dehydrate E- and Z-aldoximes to the opposite enantiomeric forms
of a chiral nitrile, has been rationalized by means of a molecular
modelling study utilizing an MOE software. This modelling gave a
detailed insight why with the same enzyme the use of racemic E-
and Z-aldoximes led to the opposite forms of the chiral nitrile.
Furthermore, mutants with an increased and decreased cavity
have been calculated, which subsequently showed the expected
and theoretically predicted decrease and increase of the
enantioselectivities also in the experimental biotransformations.
In addition, based on this validated model it was possible to
rationally design the mutant OxdRE-Leu145Phe with a decreased
size of the cavity, which then gave superior enantioselectivities
compared to the known wild-type enzyme with excellent E-values
of up to E>200. Thus, this theoretical study will also serve as a
basis for predicting aldoxime dehydratase mutants with improved
stereochemical properties for novel substrates in the future.
[6]
[7]
[8]
[9]
S. X. Xie, Y. Kato, Y. Asano, Biosci. Biotechnol. Biochem. 2001, 65,
2666–2672.
Y. Kato, R. Ooi, Y. Asano, J. Mol. Catal. B: Enzym. 1999, 6, 249–256.
[10] R. Metzner, S. Okazaki, Y. Asano, H. Gröger, ChemCatChem 2014, 6,
3105–3109.
[11] J. E. Choi, S. Shinoda, Y. Asano, H. Gröger, Catalysts 2020, 10, 362.
[12] A. Hinzmann, S. Glinski, M. Worm, H. Gröger, J. Org. Chem. 2019, 84,
4867–4872.
[13] C. Plass, A. Hinzmann, M. Terhorst, W. Brauer, K. Oike, H. Yavuzer, Y.
Asano, A. J. Vorholt, T. Betke, H. Gröger, ACS Catal. 2019, 9, 5198–
5203.
[14] J. Nomura, H. Hashimoto, T. Ohta, Y. Hashimoto, K. Wada, Y. Naruta,
K.-I. Oinuma, M. Kobayashi, Proc. Nat. Acad. Scienc. USA 2013, 110,
2810–2815.
[15] Molecular Operating Environment (MOE) 2019.01 (Chemical Computing
Group ULC, Montreal, QC, Canada, 2019).
[16] J. Davila-Calderon, N. N. Patwardhan, L.-Y. Chiu, A. Sugarman, Z. Cai,
S. R. Penutmutchu, M.-L. Li, G. Brewer, A. E. Hargrove, B. S. Tolbert,
Nat. Commun. 2020, 11, 4775.
[17] A. B. Bueno, B. Sun, F. S. Willard, D. Feng, J. D. Ho, D. B. Wainscott, A.
D. Showalter, M. Vieth, Q. Chen, C. Stutsman, B. Chau, J. Ficorilli, F. J.
Agejas, G. R. Cumming, A. Jiménez, I. Rojo, T. S. Kobilka, B. K. Kobilka,
K. W. Sloop, Nat. Chem. Biol. 2020, 16, 1105–1110.
Acknowledgements
We gratefully acknowledge generous support from the
Fachagentur Nachwachsende Rohstoffe (FNR) and the German
Federal Ministry of Food and Agriculture (BMEL), respectively,
within the funding program on the utilization of biorenewables
(Grant No. 22001716). Thanks are also due to ERATO Asano
Active Enzyme Molecule Project of Japan Science and
Technology Agency (Grant No. JPMJER1102). We also
gratefully acknowledge generous support by a grant-in-aid for
Scientific Research (S) from The Japan Society for Promotion of
Sciences to Y.A. (Grant No. 17H06169). Furthermore, generous
support from the German Academic Exchange Service (DAAD)
and Japan Society for the Promotion of Science (JSPS) within
the joint bilateral DAAD-JSPS funding program “PPP Japan
2017” (DAAD grant no. 57345566) is gratefully acknowledged,
and in addition H.Y. thanks the German Academic Exchange
Service (DAAD) for a travel grant. We thank Dr. Fumihiro
Motojima for technical assistance and help in introducing the
MOE software to H.Y..
[18] F.-A. Khan, N. Nasim, Y. Wang, A. Alhazmi, M. Sanam, Z. Ul-Haq, D.
Yalamati, M. Ulanova, Z.-H. Jiang, Eur. J. Med. Chem. 2020, 209,
112863.
[19] W.-L. Ye, C. Shen, G.-L. Xiong, J.-J. Ding, A.-P. Lu, T.-J. Hou, D.-S. Cao,
J. Chem. Inform. Mod. 2020, 60, 4216–4230.
[20] J. K. Leman, B. D. Weitzner, S. M. Lewis, J. Adolf-Bryfogle, N. Alam, R.
F. Alford, M. Aprahamian, D. Baker, K. A. Barlow, P. Barth, B. Basanta,
B. J. Bender, K. Blacklock, J. Bonet, S. E. Boyken, P. Bradley, C. Bystroff,
P. Conway, S. Cooper, B. E. Correia, B. Coventry, R. Das, R. M. de Jong,
F. DiMaio, L. Dsilva, R. Dunbrack, A. S. Ford, B. Frenz, D. Y. Fu, C.
Geniesse, L. Goldschmidt, R. Gowthaman, J. J. Gray, D. Gront, S. Guffy,
S. Horowitz, P.-S. Huang, T. Huber, T. M. Jacobs, J. R. Jeliazkov, D. K.
Johnson, K. Kappel, J. Karanicolas, H. Khakzad, K. R. Khar, S. D. Khare,
F. Khatib, A. Khramushin, I. C. King, R. Kleffner, B. Koepnick, T.
Kortemme, G. Kuenze, B. Kuhlman, D. Kuroda, J. W. Labonte, J. K. Lai,
G. Lapidoth, A. Leaver-Fay, S. Lindert, T. Linsky, N. London, J. H. Lubin,
S. Lyskov, J. Maguire, L. Malmström, E. Marcos, O. Marcu, N. A. Marze,
J. Meiler, R. Moretti, V. K. Mulligan, S. Nerli, C. Norn, S. Ó'Conchúir, N.
Ollikainen, S. Ovchinnikov, M. S. Pacella, X. Pan, H. Park, R. E.
Pavlovicz, M. Pethe, B. G. Pierce, K. B. Pilla, B. Raveh, P. D. Renfrew,
S. S. R. Burman, A. Rubenstein, M. F. Sauer, A. Scheck, W. Schief, O.
Schueler-Furman, Y. Sedan, A. M. Sevy, N. G. Sgourakis, L. Shi, J. B.
Siegel, D.-A. Silva, S. Smith, Y. Song, A. Stein, M. Szegedy, F. D. Teets,
S. B. Thyme, R. Y.-R. Wang, A. Watkins, L. Zimmerman, R. Bonneau,
Nat. Methods 2020, 17, 665–680.
Conflicts of Interest
The authors declare no conflicts of interest.
[21] Clairfeuille, T.; Buchholz, K. R.; Li, Q.; Verschueren, E.; Liu, P.;
Sangaraju, D.; Park, S.; Noland, C. L.; Storek, K. M.; Nickerson, N. N.;
Martin, L.; Dela Vega, T.; Miu, A.; Reeder, J.; Ruiz-Gonzalez, M.; Swem,
D.; Han, G.; DePonte, D. P.; Hunter, M. S.; Gati, C.; Shahidi-Latham, S.;
Xu, M.; Skelton, N.; Sellers, B. D.; Skippington, E.; Sandoval, W.; Hanan,
E. J.; Payandeh, J.; Rutherford, S. T. Structure of the essential inner
membrane lipopolysaccharide-PbgA complex. Nature 2020, 584, 479–
483. DOI: 10.1038/s41586-020-2597-x.
Keywords: aldoxime dehydratase • chiral nitriles • docking •
enantioselectivity • rational protein design
References
[1]
A. Kleemann, J. Engel, B. Kutscher, D. Reichert, Pharmaceutical
substances: Syntheses, patents, applications; Thieme, Stuttgart, New
York, 2001.
[22] C. Czekelius, E. M. Carreira, Angew. Chem. 2005, 117, 618–621; Angew.
Chem. Int. Ed. 2005, 44, 612–615.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
[23] Y. Kato, S. Yoshida, S.-X. Xie, Y. Asano, J. Biosci. Bioeng. 2004, 97,
250–259.
[24] H. Sawai, H. Sugimoto, Y. Kato, Y. Asano, Y. Shiro, S. Aono, J. Biol.
Chem. 2009, 284, 32089–32096.
[25] G. Sciortino, J. Rodríguez-Guerra Pedregal, A. Lledós, E. Garribba, J.-
D. Maréchal, J. Comput. Chem. 2018, 39, 42–51.
[26] V. Muñoz Robles, E. Ortega-Carrasco, L. Alonso-Cotchico, J. Rodriguez-
Guerra, A. Lledós, J.-D. Maréchal, ACS Catal. 2015, 5, 2469–2480.
[27] Y. Miao, R. Metzner, Y. Asano, ChemBioChem 2017, 18, 451–454.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202017234
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
8
This article is protected by copyright. All rights reserved.