Revealing Additional Stereocomplementary Pairs of Old Yellow Enzymes by Rational Transfer of Engineered Residues

Article abstract of DOI:10.1002/cbic.201600688

Every year numerous protein engineering and directed evolution studies are published, increasing the knowledge that could be used by protein engineers. Here we test a protein engineering strategy that allows quick access to improved biocatalysts with very little screening effort. Conceptually it is assumed that engineered residues previously identified by rational and random methods induce similar improvements when transferred to family members. In an application to ene-reductases from the Old Yellow Enzyme (OYE) family, the newly created variants were tested with three compounds, revealing more stereocomplementary OYE pairs with potent turnover frequencies (up to 660 h^?1) and excellent stereoselectivities (up to >99 %). Although systematic prediction of absolute enantioselectivity of OYE variants remains a challenge, “scaffold sampling” was confirmed as a promising addition to protein engineers' collection of strategies.

Full text of DOI:10.1002/cbic.201600688

Accepted Article
Title: Revealing Additional Stereocomplementary Pairs of Old Yellow
Enzymes by Rational Transfer of Engineered Residues
Authors: Nathalie Nett, Sabine Duewel, Alexandra Annelis Richter, and
Sabrina Hoebenreich
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: ChemBioChem 10.1002/cbic.201600688
Link to VoR: http://dx.doi.org/10.1002/cbic.201600688
A Journal of
www.chembiochem.org

10.1002/cbic.201600688
ChemBioChem
FULL PAPER
Revealing Additional Stereocomplementary Pairs of Old Yellow
Enzymes by Rational Transfer of Engineered Residues
Nathalie Nett,[a] Sabine Duewel,[a] Alexandra Annelis Richter and Sabrina Hoebenreich*
Abstract: Every year numerous protein engineering and directed
evolution studies are published, increasing the knowledge which
could be used by protein engineers. Here we test a protein
engineering strategy that allows quick access to improved
biocatalysts with very little screening effort. Conceptually it is
assumed that engineered residues, previously identified by rational
and random methods, induce similar improvements when transferred
to family members. Applied to ene-reductases from the Old Yellow
Enzyme family (OYE), the newly created variants were tested with
three compounds revealing more stereocomplementary OYE pairs
with potent turnover frequencies (up to 660 h-1) and excellent
stereoselectivities (up to >99%). Although systematic prediction of
absolute enantioselectivity still remains for OYE variants, ‘scaffold
sampling’ was confirmed as a promising addition to the protein
engineers’ collection of strategies.
bonds of α,β-unsaturated carbonyl, nitro and cyano compounds.
Hence, the family has attracted quite some attention as tool in
organic synthesis over the last years.[1,6–8] Because only medium
and low throughput screens based on gas or high-pressure
liquid chromatography are suitable for most OYE
transformations, effective strategies for engineering of family
members by avoiding screening-intensive protein engineering or
directed evolution are needed. The big challenge is the access
of both enantiomers, since stereocomplementary pairs of OYE
wild-types (wt) for control of facial selectivity are extremely rare
[9,10,11]
and substrate specific among the OYE family.
Mining the
existing OYE protein engineering literature reveals that OYEs
have been evolved towards three common compounds (Scheme
1).
Introduction
Implementing enzyme catalysis as a tool into our chemical
synthesis toolbox is an important contribution to achieve
sustainable synthesis routes for the 21th century.[1,2] Thus, the
need to adapt natures given catalyst is high and as a result,
protein engineering based on rational design as well as directed
evolution is often inevitable.[3]
Finding the best residues at hotspot positions with almost no
screening effort, be it in silico or experimentally, is currently the
big challenge in protein engineering. Recently two groups
simultaneously applied a fast protein engineering strategy
accessing significantly improved biocatalysts with virtually no
screening and creation of low numbers of variants. The strategy
assumes that residues obtained earlier by protein engineering,
either rational or random, will induce the same improvements of
traits when transferred to their hotspot positions in other family
members. Dunn et al.[4] transferred XNA polymerase activity to
three polymerase scaffolds, overcoming the limitation of the
originally evolved one. They named the strategy ‘scaffold
sampling’. Gober et al.[5] quickly identified new olefin
cyclopropanation biocatalysts that allow access to all four
stereoisomers by transferring engineered residues into twelve
P450 scaffolds. The overall success rate to obtain an improved
biocatalyst by this protein engineering strategy appears high, as
demonstrated by the results of both studies.
Scheme 1. Most common transformations in directed evolution and protein
engineering studies of OYE for control of facial selectivity of hydride attack.
The goal is to find complementary catalyst pairs allowing access to both
enantiomers with excellent stereoselectivities.
The first discovered OYE variant able to switch facial selectivity
was W116I in OYE1 from Saccharomyces pastorianus for the
reduction of
(S)-carvone
((S)-2-methyl-5-(prop-1-en-2-yl)
cyclohex-2-en-1-one, 3) to (2S,5S)-4.[12] Wt and W116F
enzymes produce (2R,5S)-4, as well as all other reported OYE
wild-types. Interestingly, W116X variants were identified by an
activity based screening for the reduction of 3-methylcyclohex-2-
en-1-one (1), but no flip in stereoselectivity was observed for
1.[12] The second reported stereocomplementary pair was C26G
and C26D/I69T variants of YqjM from Bacillus subtilis, allowing
access to both enantiomers for the reduction of 1.[13] They also
found that variants with mutations in A104, homolog to W116 in
hotspot position III (Figure 1), contribute to higher (R)-selectivity.
When the precursor methyl-2-(hydroxymethyl)-acrylate (5) of the
industrial important ‘Roche ester’ 6 was introduced as an OYE
substrate, all tested wild-types produced (R)-6.[14] Access to (S)-
6 became possible upon discovery of OYE2.6 from Pichia
stipitis.[15] The same study also showed that single site
saturation mutagenesis of OYE1 at hotspot position W116
Here we examined the strategy with industrial interesting ene-
reductases from the Old Yellow Enzyme family (OYE,
EC 1.6.99.1). OYEs use flavin and a nicotinamide based hydride
source to catalyse the trans-specific hydrogenation of C=C
Dr. S. Hoebenreich, N. Nett, S. Duewel, A. Richter
Department of Chemistry
Philipps-Universität Marburg
Hans-Meerwein-Straße 4, 35032 Marburg
E-mail: sabrina.hoebenreich@chemie.uni-marburg.de
[a]
authors contributed equally
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201600688
ChemBioChem
FULL PAPER
likewise produced stereocomplementary variant pairs for 6. In a
follow up directed evolution study, additional complementary
pairs with improved conversion levels were discovered for the
OYE2.6 scaffold.[16] Lately, a rational design study identified
variants C26N/I69A and I69A/H167A of YqjM to induce the
same stereochemical flip.[17]
focused on accessing new stereocomplementary variants with
preparative useful turnover numbers and if the new OYE
biocatalysts are superior relative to their starting scaffolds and to
the original engineered wild-types.
Results and Discussion
Mutability and Recombinant Expression
Mutagenesis of selected wild-type genes turned out to be
challenging. High GC content and stable secondary DNA
structures were found in the sequence areas of interest (65-70%,
Table S3). We started with mutagenesis to introduce the
aspartate/threonine combination into the seven scaffold genes.
Surprisingly, nemA and xenA mutagenesis yielded either the wt
gene or the aspartate mutation alone, even after several rounds
of PCR optimization (see supporting information). Therefore we
did not further pursue the engineering of NemA and XenA.
After introducing aspartate/threonine at hotspot position I+II into
TsER, NCR, DrER, RmER and XenB, a first heterologous
expression test with wt and mutants were performed. To avoid a
later falsification of biotransformation data due to contaminations
with E. colis’ own OYE, we used a knockout strain for enzyme
production.[35] The OD600 normalized SDS-PAGE analysis
(Figure S2) reveals that tested variants were produced. Variant
XenB T25D/Y66T did not yield any protein after affinity
purification, confirming recent reports[30], XenB studies were not
further pursued. Mutagenesis and expression problems can
sometimes be solved by changing the plasmid backbone to
smaller or newest generation plasmids. This might be especially
promising for XenA, XenB (pGaston) and NCR (pET28).
Figure 1. Sequence alignment showing position and residues of hotspots in
primary and tertiary structure of wild-type OYEs, which have a low sequence
homology across group members (23-34%) and consist of two subgroups,
possessing distinct conserved residues at targeted hotspot positions.[6,8] An
important aspect is the >98% sequence identity at these empirically
determined hotspot positions, excluding the use of consensus based
engineering strategies for residue identification.[18] Alignment was created with
Clustal Omega[19] and sequences are sorted by pairwise identity. Structures of
TsER (3hgj, green)[20], PETNR (3p81, yellow)[21] and NCR (4a3u, cyan)[22] have
been aligned using PyMol. Residues at hotspot positions are shown as sticks,
FMN as lines. The Cα atoms of residues are shown as spheres. The side
chains of hotspot position II fill the same space, but residues have different
positions in primary structure.
Overall, site directed mutagenesis yielded 14 heterologous
expressed variants from group 1 and 2 members, which were
further tested for turnover and stereoselectivity.
Turnover number and frequency
All five purified wild-types and 14 variants were obtained as
active enzymes, being able to convert 10 mM of 3 containing an
electronic activated C=C double bond with turnover numbers
(TON) between 60 and 1000. With our catalyst loading of 10 µM
(0.1 mol%), the chosen reaction condition maximally lead to a
TON of 1000. Reduction of 3 achieved turnover frequencies
(TOF) between 40 and 667 molecules per hour, showing that
most of our new variants are faster than literature reported
variants (TOF, 18 - 300 h-1).[12,23,36,37] With our standardized,
non-optimized conditions, our TOF numbers are already close to
the best reported TOF for (S)-carvone (~1000 h-1) by LacER.[31]
As expected, the wild-types of group 2, TsER, RmER, DrER and
XenA, had a rather low activity (≤ 1%) for 1.[13,32,37,38,39] NCR wt,
TsER C25D/I67T, C25G/I67T and DrER C40D/I81T reach the
best TOF of 82-190 h-1 for reduction of 1. The electronic
deactivation of the beta position reduced TOF almost ten-fold. A
more significant effect on TOF was observed, when the
hydrogen bond acceptor potential of the anchoring group is
altered by switching from ketones to esters, as in the case of 5,
where only six (Table 1) of the 19 tested enzymes showed
detectable formation of 6. A reasonable active enzyme catalyst
In addition to aforementioned studies, a few single-residue
saturation mutagenesis,[21,23–26] rational design[27] or directed
evolution studies[28] exist, which allow extraction of engineered
residues. Taken all together, several hotspot positions have
been identified and especially positions I, II and III (Figure 1)
were found to control facial selectivity, activity and substrate
scope in OYE1, OYE2.6, PETNR, KYE and YqjM.[12,13,16,23,26]
Testing all combinatorial possibilities arising from empirical
identified hotspot positions, their amino acid combinations, the
screening substrates and available OYE scaffolds would be too
resource intensive and we therefore restricted ourselves to
seven scaffolds (Figure 2, green), three groups of variants
(Figure 2, red) and the three substrates (1-3). Engineered
residues were selected to represent both OYE groups.[6] Wild-
type scaffolds were partially selected based on proven potential
as biocatalyst in cascade reactions[29,30]
,
large scale
synthesis[9,31,32], use of NAD(P)H mimics[33] or stability.[20,34] With
this setup, we were interested in testing OYE family specific
response when applying the scaffold sampling strategy. We
This article is protected by copyright. All rights reserved.

10.1002/cbic.201600688
ChemBioChem
FULL PAPER
Figure 2. Conceptual overview of protein engineering strategy applied in this study. Choosing engineered residues (red) identified in directed evolution and
protein engineering studies in hotspot positions (blue) and transferring them to OYE scaffolds (green).
should have at least a TON of 10-100 and a conversion of >90%
for industrial applications.[40] TsER wt and TsER C25D/I67T
achieved this with 94 and 98% conversion after 24h, albeit with
a moderate TOF of 39 and 41 h-1, respectively. Transfer of
engineered residues into group 1 OYE scaffold NCR did
increase conversion of 5, since NCR W100I improved from no
detectable to 46% production of 6. A similar trend was observed
for OYE1 W116X variants, but not for OYE2.6 variants, even
when mutating at the same three hotspot positions discussed
here.[16,41] In group 2, conversion of 5 was significantly reduced
for engineered variants of YqjM, where substrate engineering
circumvented the problem.[17] More trends are apparent; C26G
homologs showed either unaltered or decreased conversion,
irrespective of the substrate. Literature also reports the same
trend for YqjM and XenA, especially when substrates without Cα
activation are used,[13,32] and saturation mutagenesis studies of
OYE2.6[41] or PETNR[23] did not yield a homolog glycine variant
when selected by conversion. Since we noted during our data
mining that double/triple variants with mutations in hotspot
supports the importance of hotspot position III itself. Additionally,
data mining also reveals an importance for fine-tuning reactivity,
either alone[15,23,41,46] or in context with variations at other hotspot
positions.[13,16,36]
Results for reduction of 1 and 3 reflect that evolution of the
active site towards 1 was performed in a group 2 member.
Therefore, it was surprising that variant TsER C25D/I67T
converts the structurally different 5 equally well as the wt, and
the aspartate/threonine variant of DrER showed increased
conversions, but only marginally (from 2 to 6%).
Overall RmER wt and variants turned out to be the least active
OYE, possibly due to dimer-dimer instabilities.[39] Since, activity
improvements are reported to trade off against stability, melting
temperatures of RmER C25D/I66T, DrER C40D/I81T were
analysed and found to drop between 4 and 10°C compared to
the wild-types (see Figure S7). We could further confirm that
NCR wild-type and variants have a high biocatalytic potential
and shown that TsER variants are equally good as NCR variants,
when used for biotransformation. Additionally, all shown TsER
variants are still obtainable as active, when purified at 70°C for
90 min and ratios of conversion levels obtained with purified
(heat treated) or cell lysate (no heat treatment, see supporting
information) were maintained.
position II generally favour higher conversion levels,[13,16,42]
threonine at the respective position was introduced in all C26G
homologs and indeed conversion of and for
a
1
3
glycine/threonine variants increased compared to the single
C26G variants.
Control of Facial Selectivity
Interestingly the C26D/I69T pair did generally increase
conversion and TOF in group 2 members (Table 1). The
introduction into group 1 members NCR and XenB however
resulted in a loss of conversion from 47 to 3% or insolubility,
respectively. A similar phenomenon was observed when hotspot
position II in OYE2.6 (Y78) was exchanged to smaller residues
and abolished carvone conversion.[16] Vice versa, incorporation
of W116I homologs into group 2 member TsER was not able to
induce increased conversion levels. Mixed reports are apparent
in the literature demonstrating that the influence of hotspot III on
conversion might be weaker than other hotspots and highly
substrate dependent. No activity for 1 was observed with
PETNR W102I, but other variants at this hotspot position and
It appears that group 1 wild-types, with their bulky residues in
hotspot position II and III prefer a binding mode of 1 that leads to
(S)-2.[11,47] Reports of facial selectivity for group 2 members are
rare. XenA wt favours (S)-2 (>99%)[30] like all other OYE wt and
only YqjM wt produces (R)-2 with 76% ee.[13] We expected that
TsER wt and DrER wt will likewise produce the (R)-enantiomer
with their structurally almost identical active sites.[34] To access
accurate stereoselectivities for entries with such low conversions
(≤ 2%, Table 1), their reactions were performed in large scale
(10 mL). Surprisingly, group 2 members DrER wt and TsER wt
produced (S)-2 with 93% and 76% ee, respectively, and YqjM wt
reproducibly (R)-2 with 78% ee (Figure 3). It would be
interestingto understand what determines the re-face orientation
of 1 in YqjM wt compared to DrER and TsER. Turning to the fact
that all current known OYE wild-types reduced 3 with a facial
selectivity yielding exclusively (2R,5S)-4.[31,48] Our data confirms
this trend for TsER, DrER, RmER and XenB, and newly shows
other
substrates
showed
increased
conversion.[23,43]
Incorporation of W116I into circular permutated OYE1 variants
lowered conversion of 3,[25,44] a saturation mutagenesis library at
W116 showed no hits for 1[45] and biocatalytic characterisation of
all 20 variants showed that W116I is not the best engineered
residue for this position.[15,46] Nevertheless, our NCR data
This article is protected by copyright. All rights reserved.

10.1002/cbic.201600688
ChemBioChem
FULL PAPER
Table 1. OYE wt and variant activities under defined conditions given as turnover numbers (TON) and turnover frequencies (TOF) as product formation rate,
using 10 µM OYE and 10 mM substrate concentration.
conv. (a)
(%GC)
TOF(b)
conv. (a)
(%GC)
TOF(c)
conv. (a)
(%GC)
TOF(d)
TON
TON
TON
(h-1)
(h-1)
(h-1)
wild-types
NCR
XenB
RmER
DrER
47
10
n.c.
1
465
100
n.a.
10
93
20
n.a.
2
84
68
93
99
100
840
680
930
990
1000
560
453
620
660
667
n.c.
n.c.
n.c.
2
n.a.
n.a.
n.a.
20
n.a
n.a
n.a
1
TsER
1
10
2
94
940
39
C26G homologs
NCR T25G
RmER C25G
DrER C40G
9
93
n.a.
n.a.
70
19
n.a.
n.a.
14
43
16
28
97
430
160
280
970
287
107
187
647
n.c.
n.c.
n.c.
n.c.
n.a.
n.a.
n.a.
n.a.
n.a
n.a
n.a
n.a
n.c.
n.c.
7
TsER C25G
C26G/I69T homologs
NCR T25G/W66T
RmER C25G/I66T
DrER C40G/I81T
TsER C25G/I67T
C26D/I69T homologs
NCR T25D/W66T
RmER C25D/I66T
DrER C40D/I81T
TsER C25D/I67T
W116I homologs
NCR W100I
1
2
5
14
20
50
3
4
10
82
6
60
40
n.c.
n.c.
n.c.
11
n.a.
n.a.
n.a.
110
n.a
n.a
n.a
5
19
85
99
190
850
990
127
567
660
41
410
3
7
73
95
34
70
730
950
7
14
146
190
69
96
99
690
960
990
460
640
660
667
n.c.
n.c.
6
n.a.
n.a.
60
n.a
n.a
3
100
1000
98
980
41
1
<1
6
4
1
1
31
81
310
810
207
540
46
n.c.
460
n.a.
19
n.a
TsER A102I
(a) mean from triplicates, standard deviation ±5%; (b) 5h; (c) 1.5h; (d) 24 h; n.c. no conversion under assay conditions, n.a. not applicable
Figure 3. Overview of stereoselectivity of purified OYE variants producing 2, 4 and 6. The selectivity on the left is the prevailing selectivity accessible with wt
OYEs. A selectivity switch for all three substrates is observed with the created variants in hotspot positions I+II. Assays were performed aerobically, in 200 µl
containing 10 µM OYE, 10 mM substrate, a NADP+, glucose, GDH based cofactor recycling system for 5 h (1), 1.5 h (3) or 24 h (5) in 100 mM KPi, pH = 7.4.
Enantiomeric excess was determined by GC. Error bars omitted for clarity, error from n=3 given in supporting information. n.c. no conversion, n.d. not determined.
that NCR wt is no exception. Additionally, our data newly
showed 76% or 97% ee for (R)-6 for TsER and DrER wt,
respectively, and confirms the trend that all current known OYE
wt scaffolds exclusively yield the (R)-enantiomer of 6,[14] where
the only known exception is OYE2.6 wt.[15]
We found that control of facial selectivity previously observed for
the YqjM variant pair C26G and C26D/I69T is transferable to
NCR, RmER, DrER and TsER for the reduction of 1. Glycine
variants in hotspot position I give access to (S)-2 and the
aspartate/threonine combination in hotspot position I+II allows
This article is protected by copyright. All rights reserved.

10.1002/cbic.201600688
ChemBioChem
FULL PAPER
access to (R)-2, irrespective of the scaffold or the selectivity of
the wild-type. It appears that for the substrate/residue pair 3-
methylcyclohex-2-en-1-one (1) and aspartate/threonine in
hotspot position I+II the flipped orientation is predominant and
the glycine based variants showed the same facial selectivity as
the majority of wt. Notably, TsER C25G, DrER C40D/I81T and
TsER C25D/I67T showed increased selectivity compared to
YqjM variants. Achieved selectivities are excellent, up to 98%
(R)-2, which is one of the best results obtained so far with OYEs.
Only engineering of group 2 achieved good TON, TOF and
enantioselectivity values for (R)-2 production. Since literature
residues and hotspot positions for future OYE protein
engineering.
Here we discovered that the Asp/Thr and Gly/(Thr or wt) set in
hotspot position I+II represent a reliable stereocomplementary
pair of OYE variants, best suitable for group 2 members. Further,
we showed that Ile incorporation into hotspot position III has
more pronounced effect in group 1 member NCR than with
group 2. Our results indicate that we need two sets of
engineered residues for the two OYE subfamilies. We already
have data for group 2 combination of hotspots I+II+III, but a
dataset for hotspot position I in context with the other two
positions for group 1 members is currently missing. But even
with the bigger systematisation of the OYE engineering
knowledge that our study provides, models based on empiric
data for systematic prediction of absolute facial selectivity seem
out of reach for OYEs.
reported engineered group 1 members all favour (S)-2.[12,23,36]
,
our NCR T25D/W66T variant is the first (R)-2 selective group 1
example, but with rather low activity. An exception might be
XenA C25G, where a report by Yanto et al.[32] indicates >99%
(R)-2 with a TOF of 53 h-1, but stands in sharp contrast to the
systematic appearance of (S)-2 preference of hotspot I glycine
variants reported here.
Though, it remains to be seen if other recently identified
stereocontrolling
hotspot
positions[45]
and
engineered
Further, the same pattern of stereocomplementarity for the
glycine/threonine and aspartate/threonine variants was observed
for the reduction of 3, allowing access to (2S,5S)-4 or (2R,5S)-4
with excellent diastereoselectivity of up to 92 or 96% de,
respectively. This is the first report of group 2 members
producing (2S,5S)-4 and besides OYE1 W116X variants from
group 1,[24] the only other set of variants so far for this
reaction.[48] All our other new variants led to (2R,5S)-4, even the
hotspot position III homologs NCR W100I and TsER A102I.
residues[16,36] will be equally transferable among OYE scaffolds.
We could confirm that scaffold sampling by transfer of
engineered residues into wild-types is an extremely valuable
protein engineering strategy with high success rates, allowing
quick access to improved biocatalysts. It allows us protein
engineers to start using the knowledge obtained earlier by
protein engineering and evolution systematically, thereby
avoiding a waste of resources when creating new and improved
biocatalysts.
Lastly, reduction of
5
confirms the generality of the
stereocomplementary pair, since TsER C25G/I67T and
C25D/I67T gave 79% (R)-6 or 97% (S)-6 ee, respectively. The
combination of 5 and glycine based variants showed the same
facial selectivity as the wt and the aspartate/threonine
combination flipped it. This represents the second variant of
group 2 that is able to access (S)-6. Rüthlein et al. discovered
YqjM variants C26N/I69A and H167A/I69A for production of (S)-
6,[17] albeit with TOFs of 10.3 and ~0.3 h-1, respectively,
compared to 41 h-1 (TsER C26D/I67T) and 19 h-1 (NCR W100I).
A combinatorial library of TsER (unpublished data) contained
some of the in silico designed variants of Rüthlein et al., but they
showed no conversion of 5 with our reaction conditions (Table
S6).
Experimental Section
Details
about
mutagenesis
PCR,
protein
expression
and
biotransformation reactions can be found in the supporting information.
Acknowledgements
The work was supported by the LOEWE Research Cluster
SynChemBio of the Federal State of Hesse (Germany) and the
Philipps-Universität Marburg. We thank M.T Reetz for access to
instruments and reference chemicals, D.J. Opperman for RmER
and DrER, B. Hauer for NCR, U. Bornscheuer for XenA, XenB
and NemA plasmids, M. Mihovilovic for E. coli BL21 (DE3)
ΔnemA strain and Evocatal for the GDH-60 gift. Our thanks goes
to Saskia Döhring and the mass spectrometry department for
excellent technical assistance.
Turning to isoleucine based variants at hotspot position III, we
found no flip in facial selectivity for the reduction of 3, which is
contrary to the original report,[12] but consistent with cpOYE1 and
OYE2.6 engineering results.[36,44] Nevertheless, NCR W100I
showed >99% (S)-6 selectivity, presumably representing a
flipped facial selectivity compared to the wt.[14] This supports
follow up studies by Stewart and coworkers, where it was shown
that hotspot III in group 1 indeed controls stereoselectivity, but
highly depending on amino acid substitution in this position and
Keywords: ene reductases • enzyme based stereocontrol •
protein engineering • residue transfer • scaffold sampling
tested substrate.[15,16,46]
A
combinatorial library of NCR
(unpublished data) yielded W100V, a homolog of one of the
most promising OYE W116X variants.[48] Interestingly, this NCR
W100V variant was not able to switch facial selectivity for
reduction of 1 or 3 (Table S5).
[1]
[2]
K. Faber, W.-D. Fessner, N. Turner, Science of synthesis. Biocatalysis
in organic synthesis 1-3, Georg Thieme, Stuttgart, 2015.
a) R. A. Sheldon, Green Chem 2017, 19, 18–43; b) H. Gröger, W.
Hummel, Curr. Opin. Chem. Biol. 2014, 19, 171–179; c) G. Torrelo, U.
Hanefeld, F. Hollmann, Catal. Lett. 2015, 145, 309–345; d) J.-M. Choi,
S.-S. Han, H.-S. Kim, Biotechnol. Adv. 2015; e) A. Wells, H.-P. Meyer,
ChemCatChem 2014, 6, 918–920;
Conclusion
[3]
a) U. T. Bornscheuer, G. W. Huisman, R. J. Kazlauskas, S. Lutz, J. C.
Moore, K. Robins, Nature 2012, 485, 185–194; b) M. T. Reetz, Directed
evolution of selective enzymes. Catalysts for organic chemistry and
biotechnology, Wiley-VCH Verlag GmbH et Co KGaA, Weinheim, 2017;
Until we are able to understand on molecular level why
engineered variants of OYE are better biocatalysts and thereby
will be able to rationally design them, it should be possible to
simply learn how to use the current knowledge of engineered
This article is protected by copyright. All rights reserved.

10.1002/cbic.201600688
ChemBioChem
FULL PAPER
c) H. Renata, Z. J. Wang, F. H. Arnold, Angew. Chem. Int. Ed. 2015, 54,
3351–3367;
[29] a) S. Reich, B. M. Nestl, B. Hauer, ChemBioChem 2016, 17, 561–565;
b) E. Brenna, F. G. Gatti, D. Monti, F. Parmeggiani, A. Sacchetti, J.
Valoti, J. Mol. Catal. B: Enzym. 2015, 114, 77–85; c) T. Knaus, F. G.
Mutti, L. D. Humphreys, N. J. Turner, N. S. Scrutton, Org. Biomol.
Chem. 2014, 13, 223–233; d) R. Agudo, M. T. Reetz, Chem. Commun.
2013, 49, 10914–10916; e) C. Zhang, D. K. Schneiderman, T. Cai, Y.-S.
Tai, K. Fox, K. Zhang, ACS Sustainable Chem. Eng. 2016, 4, 4396–
4402;
[4]
[5]
[6]
[7]
[8]
M. R. Dunn, C. Otto, K. E. Fenton, J. C. Chaput, ACS Chem. Biol. 2016,
11, 1210–1219.
J. G. Gober, A. E. Rydeen, E. J. Gibson-O’Grady, J. B. Leuthaeuser, J.
S. Fetrow, E. M. Brustad, ChemBioChem 2016, 17, 394–397.
H. S. Toogood, J. M. Gardiner, N. S. Scrutton, ChemCatChem 2010, 2,
892–914.
H. S. Toogood, N. S. Scrutton, Curr. Opin. Chem. Biol. 2014, 19, 107–
115.
[30] L. Skalden, C. Peters, L. Ratz, U. T. Bornscheuer, Tetrahedron 2015.
[31] X. Chen, X. Gao, Q. Wu, D. Zhu, Tetrahedron: Asymmetry 2012, 23,
734–738.
E. Brenna (Ed.) Synthetic Methods for Biologically Active Molecules.
Exploring the Potential of Bioreductions, Wiley-VCH Verlag GmbH & Co.
KGaA, Weinheim, Germany, 2013.
[32] Y. Yanto, H.-H. Yu, M. Hall, A. S. Bommarius, Chem. Commun. 2010,
46, 8809–8811.
[9]
T. Reß, W. Hummel, S. P. Hanlon, H. Iding, H. Gröger, ChemCatChem
2015, 7, 1302–1311.
[33] A. Geddes, C. E. Paul, S. Hay, F. Hollmann, N. S. Scrutton, J. Am.
Chem. Soc. 2016, 138, 11089–11092.
[10] a) D. J. Bougioukou, A. Z. Walton, J. D. Stewart, Chem. Commun. 2010,
46, 8558–8560; b) C. Stueckler, M. Hall, H. Ehammer, E. Pointner, W.
Kroutil, P. Macheroux, K. Faber, Org. Lett. 2007, 9, 5409–5411;
[11] M. Hall, C. Stueckler, H. Ehammer, E. Pointner, G. Oberdorfer, K.
Gruber, B. Hauer, R. Stuermer, W. Kroutil, P. Macheroux, K. Faber,
Adv. Synth. Catal. 2008, 350, 411–418.
[34] D. J. Opperman, L. A. Piater, E. van Heerden, J. Bacteriol. 2008, 190,
3076–3082.
[35] N. Oberleitner, C. Peters, J. Muschiol, M. Kadow, S. Saß, T. Bayer, P.
Schaaf, N. Iqbal, F. Rudroff, M. D. Mihovilovic, U. T. Bornscheuer,
ChemCatChem 2013, 5, 3524–3528.
[36] A. Patterson-Orazem, B. Sullivan, J. D. Stewart, Bioorg. Med. Chem.
2014, 22, 5628–5632.
[12] S. K. Padhi, D. J. Bougioukou, J. D. Stewart, J. Am. Chem. Soc. 2009,
131, 3271–3280.
[37] B. V. Adalbjörnsson, H. S. Toogood, A. Fryszkowska, C. R. Pudney, T.
A. Jowitt, D. Leys, N. S. Scrutton, ChemBioChem 2010, 11, 197–207.
[38] a) X.-Q. Pei, M.-Y. Xu, Z.-L. Wu, J. Mol. Catal. B: Enzym. 2015, 123,
91–99; b) C. Peters, R. Kölzsch, M. Kadow, L. Skalden, F. Rudroff, M.
D. Mihovilovic, U. T. Bornscheuer, ChemCatChem 2014, 6, 1021–1027;
c) X. Gao, J. Ren, Q. Wu, D. Zhu, Enzyme Microb. Technol. 2012, 51,
26–34; d) A. Riedel, M. Mehnert, C. E. Paul, A. H. Westphal, W. J. H.
van Berkel, D. Tischler, Front. Microbiol. 2015, 6, 1–14;
[13] D. J. Bougioukou, S. Kille, A. Taglieber, M. T. Reetz, Adv. Synth. Catal.
2009, 351, 3287–3305.
[14] C. Stueckler, C. K. Winkler, M. Bonnekessel, K. Faber, Adv. Synth.
Catal. 2010, 352, 2663–2666.
[15] A. Z. Walton, W. C. Conerly, Y. Pompeu, B. Sullivan, J. D. Stewart,
ACS Catal. 2011, 1, 989–993.
[16] A. Z. Walton, B. Sullivan, A. C. Patterson-Orazem, J. D. Stewart, ACS
Catal. 2014, 4, 2307–2318.
[39] S. Litthauer, S. Gargiulo, E. van Heerden, F. Hollmann, D. J. Opperman,
J. Mol. Catal. B: Enzym. 2014, 99, 89–95.
[17] E. Rüthlein, T. Classen, L. Dobnikar, M. Schölzel, J. Pietruszka, Adv.
Synth. Catal. 2015, 357, 1775–1786.
[40] P. Tufvesson, J. Lima-Ramos, N. A. Haque, K. V. Gernaey, J. M.
Woodley, Org. Process Res. Dev. 2013, 17, 1233–1238.
[41] Y. A. Pompeu, B. Sullivan, A. Z. Walton, J. D. Stewart, Adv. Synth.
Catal. 2012, 354, 1949–1960.
[18] a) A. S. Bommarius, Nat. Chem. Biol. 2010, 6, 793–794; b) B. T.
Porebski, A. M. Buckle, Protein Eng. Des. Sel. 2016, 29, 245–251; c) J.
Bendl, J. Stourac, E. Sebestova, O. Vavra, M. Musil, J. Brezovsky, J.
Damborsky, Nucleic Acids Res. 2016; d) R. K. Kuipers, H.-J. Joosten,
W. J. H. van Berkel, N. G. H. Leferink, E. Rooijen, E. Ittmann, F. van
Zimmeren, H. Jochens, U. T. Bornscheuer, G. Vriend, V. A. P. M. dos
Santos, P. J. Schaap, Proteins: Struct., Funct., Bioinf. 2010, 78, 2101–
2113; e) H. Jochens, U. T. Bornscheuer, ChemBioChem 2010, 11,
1861–1866;
[42] S. Kille, Flavoproteins in Directed Evolution. Iterative CASTing to
evolve YqjM and P450BM3, PhD Thesis, Ruhr Universität Bochum,
2010.
[43] H. Khan, T. Barna, R. J. Harris, N. C. Bruce, I. Barsukov, A. W. Munro,
P. C. E. Moody, N. S. Scrutton, J. Biol. Chem. 2004, 279, 30563–30572.
[44] A. B. Daugherty, J. R. Horton, X. Cheng, S. Lutz, ACS Catal. 2015, 5,
892–899.
[19] F. Sievers, A. Wilm, D. Dineen, T. J. Gibson, K. Karplus, W. Li, R.
Lopez, H. McWilliam, M. Remmert, J. Soding, J. D. Thompson, D. G.
Higgins, Mol. Syst. Biol. 2011, 7, 539.
[45] L. T. Quertinmont, S. Lutz, Tetrahedron 2016, 72, 7282–7287.
[46] E. Brenna, M. Crotti, F. G. Gatti, D. Monti, F. Parmeggiani, R. W.
Powell, S. Santangelo, J. D. Stewart, Adv. Synth. Catal. 2015, 357,
1849–1860.
[20] D. J. Opperman, B. T. Sewell, D. Litthauer, M. N. Isupov, J. A.
Littlechild, E. van Heerden, Biochem. Biophys. Res. Commun. 2010,
393, 426–431.
[47] a) A. Müller, B. Hauer, B. Rosche, Biotechnol. Bioeng. 2007, 98, 22–29;
b) M. A. Swiderska, J. D. Stewart, Org. Lett. 2006, 8, 6131–6133; c) N.
J. Mueller, C. Stueckler, B. Hauer, N. Baudendistel, H. Housden, N. C.
Bruce, K. Faber, Adv. Synth. Catal. 2010, 352, 387–394; d) D. Monti, M.
C. Forchin, M. Crotti, F. Parmeggiani, F. G. Gatti, E. Brenna, S. Riva,
ChemCatChem 2015, 7, 3106–3109;
[21] H. S. Toogood, A. Fryszkowska, M. Hulley, M. Sakuma, D. Mansell, G.
M. Stephens, J. M. Gardiner, N. S. Scrutton, ChemBioChem 2011, 12,
738–749.
[22] S. Reich, H. W. Hoeffken, B. Rosche, B. M. Nestl, B. Hauer,
ChemBioChem 2012, 13, 2400–2407.
[23] M. E. Hulley, H. S. Toogood, A. Fryszkowska, D. Mansell, G. M.
Stephens, J. M. Gardiner, N. S. Scrutton, ChemBioChem 2010, 11,
2433–2447.
[48] E. D. Amato, J. D. Stewart, Biotechnol. Adv. 2015, 33, 624–631.
[24] Y. A. Pompeu, B. Sullivan, J. D. Stewart, ACS Catal. 2013, 3, 2376–
2390.
[25] L. T. Quertinmont, R. Orru, S. Lutz, Chem. Commun. 2015, 51, 122–
124.
[26] J. T. Park, L. M. Gómez Ramos, A. S. Bommarius, ChemBioChem
2015, 16, 811–818.
[27] a) C. K. Winkler, D. Clay, S. Davies, P. O’Neill, P. McDaid, S. Debarge,
J. Steflik, M. Karmilowicz, J. W. Wong, K. Faber, J. Org. Chem. 2013,
78, 1525–1533; b) S. Horita, M. Kataoka, N. Kitamura, T. Nakagawa, T.
Miyakawa, J. Ohtsuka, K. Nagata, S. Shimizu, M. Tanokura,
ChemBioChem 2015, 16, 440–445;
[28] a) A. Fryszkowska, H. Toogood, M. Sakuma, G. M. Stephens, J. M.
Gardiner, N. S. Scrutton, Catal. Sci. Technol. 2011, 1, 948; b) A. B.
Daugherty, S. Govindarajan, S. Lutz, J. Am. Chem. Soc. 2013, 135,
14425–14432;
This article is protected by copyright. All rights reserved.

10.1002/cbic.201600688
ChemBioChem
FULL PAPER
FULL PAPER
Nathalie Nett, Sabine Duewel,
Alexandra Annelis Richter and Sabrina
Hoebenreich*
Page No. – Page No.
Revealing Additional
Stereocomplementary Pairs of Old
Yellow Enzymes by Rational Transfer
of Engineered Residues
Transfer of engineered residues in ene-reductase wild-type scaffolds revealed
additional stereocomplementary pairs and provide a fast engineering method for
Old Yellow Enzymes.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201600688
ChemBioChem
targeted transformations in OYE engineering
O
O
OYE variant
NAD(P)+
GDH/glucose
1
2
O
O
OYE variant
NAD(P)+
GDH/glucose
(S)-carvone
4
3
O
O
OYE variant
HO
OMe
HO
OMe
NAD(P)+
GDH/glucose
'Roche' ester
6
5
This article is protected by copyright. All rights reserved.

hotspot position I
hotspot position II
10.1002/cbic.201600688
ChemBioChem
E A S A V N PQG R I T DQ - D L G I WS
E A T G V T PQG R I S E R - D L G I WS
E A T A V E S RG R I T D H - D L G I WN
E A T A V E P L G R I S P Y - D L G I WS
E A T A V S P EG R I T P E - D L G L WD
E A T A V A P EG R I T PG - C AG I WS
E A T A V S P EG R I T P T - D L G L Y N
E A T Q I S AQ A KG Y AG - A PG L H S
E A T S V S AMG VG Y P D - T PG I WN
E A T G I S R EG L GWP F - A PG I WS
E A T G I SQ EG L GWP Y - A PG I WS
E A T Q I S F Q A KG Y SG - S PG I H S
E A T V I S E T G I G Y K D - V PG I WT
E A T F V S PQ A SG Y EG A A PG I WT
EG T F P S PQ SGG Y D N - A PG I WS
EG T M I S P T S AG F P H - V PG I F T
EG A F I S PQ AGG Y D N - A PG VWS
EG A F P S AQ SGG Y D N - A PG VWS
E G T F I S PQ AGG Y D N - A PG I WS
YqjM
GkOYE
TOYE
TsER
DrER
XenA
RmER
PETNR
XenB
Gox
T L K N R I VM S PM CM Y S S H E K DG
T L K N R I VM S PM CM Y S C D T K DG
T I K N R I MM S PM CM Y S A S - T DG
R L K N R L AM S PM CQ Y S A T - L EG
E L P N R V V V S PM C T Y S A T - - DG
T L R N R I A I P PM CQ YM A E - - DG
E L A N R I A I A PM CQ Y S AQ - - EG
T A P N R V F M A P L T R L R S I E PG D
Q L P N R I I M A P L T R C R A D - EG R
H A K NR I VM S P L T RG R A D - K E A
T A K N R I WM A P L T RG R A T - R D H
T L P N R V F M A P L T R L R S L E PG D
E L C H R V V L A P L T RQ R S Y - - G Y
T L Q T K I V Y P P T T R F R A L E D H T
E L L H R A V I P P L T RM R AQ H PG N
N L S H R V V L A PM T R C R A L - - N N
E L L H R A V I P P L T RM R A L H PG N
E L K H R V VM P A L T RM R A L H PG N
Q L A H R A VM P P L T RM R A T H PG N
group 2
NCR
NemA
OPR1
OYE2.6
OYE2
OPR3
OYE1
KYE1
OYE3
group 1
hotspot position III
YqjM
K I G I Q L A H AG R K A E L E
A I G I Q L A H AG R K SQ V P
VMG I Q L A H AG R K C N I S
V PG I Q L A H AG R K AG T A
H I G VQ L A H AG R K A S T Y
V PG I Q I A H AG R K A S A N
A V T I Q L A H AG R K A S S E
R I A VQ L WH T G R I S H S S
R I F L Q L WH VG R I S H P S
K I V CQ L WHMG RM V H S -
L I F AQ L WHMG RM V P S -
hotspot
position II
GkOYE
TOYE
TsER
DrER
XenA
RmER
PETNR
XenB
Gox
Gly
Gly
PETN Y68
NCR W66
TsER I67
group 2
group 2
Tyr
Trp
Ile
Ser
group 1
NCR
hotspot
NemA
OPR1
OYE2.6
OYE2
OPR3
OYE1
KYE1
OYE3
H S A VQ VWH T G R V S H T S
I F F CQ I WH VG R V S N K D
F V S T Q L I F L G R V A D P V
F AWVQ L WV L GWA A F P D
V I F CQ L WH VG R A S H E V
F VWVQ L WV L GWA A F P D
F VWVQ L WV L G RQ A F A D
F AWVQ L WS L GWA S F P D
position III
group 1
PETN W102
NCR W100
TsER A102
hotspot
position I
PETN T26
NCR T25
TsER C25
This article is protected by copyright. All rights reserved.

empirically identified hotspot positions
OYE wild type scaffolds
Are traits from evolved OYE variants
within active site of OYE scaffolds
with biocatalytic potential
transferable to other OYE scaffolds?
10.1002/cbic.201600688
ChemBioChem
group 2
‘thermophilic-like‘
group 1
‘classical’
group 1
‘classical’
group 2
‘thermophilic-like‘
selected
engineered
residues
C26
T37
Y78
W116
H194
gene mutability ?
expression ?
PETNR
OYE2.6
NCR
NemA
OYE3
LacER
YqjM
OYE1
XenB
I69
TsER
DrER
for transfer
OYE1
YqjM
C26D/I69T
C26G
W116I
A104
XenA
LeOPR3
turnover ?
TOYE
Gox
A60
H167
KYE1
stereoselectivity ?
P295
F250
OYE active site
GkOEY
LeOPR1
selected scaffolds
RmER
selected hotspot positions
This article is protected by copyright. All rights reserved.

O
O
10.1002/cbic.201600688
ChemBioChem
O
O
O
O
HO
OMe
HO
OMe
(S)-2
(R)-2
(2R,5S)-4
(2S,5S)-4
(R)-6
(S)-6
n.c.
n.c.
n.c.
NCRwt
XenBwt
RmERwt
DrERwt
TsERwt
wt
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
NCRT25G
RmERC25G
DrERC40G
TsERC25G
n.c.
n.c.
NCRT25G/W66T
RmERC25G/I66T
DrERC40G/I81T
TsERC25G/I67T
n.c.
n.c.
n.c.
NCRT25D/W66T
RmERC25D/I66T
DrERC40D/I81T
TsERC25D/I67T
n.c.
n.c.
NCRW100I
TsERA102I
n.d.
n.d.
n.c.
100 80
40
(2R,5S)-4
0
40
80 100 100 80
40
(R)-6
0
40
80 100
100 80
40
0
40
80 100
(S)-2
(R)-2
(2S,5S)-4
(S)-6
ee(%)
de(%)
ee(%)
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201600688
O
O
O
O
O
O
HO
OMe
HO
OMe
(S)
(R)
(2R,5S)
(2S,5S)
(R)
(S)
NCR T25G
NCR W100I
expected but
not obtained
NCR T25G
no conversion
NCR W100I
Group 1
Group 2
?
NCR T25D/W66T
TsERC25D/I67T
DrERC40D/I81T
TsERC25G
TsERC25G
DrERC40G
TsERC25G/I67T
TsERC25D/I67T
TsERC25D/I67T TsERC25G/I67T
This article is protected by copyright. All rights reserved.