Biocatalysis with thermostable enzymes: Structure and properties of a thermophilic 'ene'-reductase related to old yellow enzyme

Article abstract of DOI:10.1002/cbic.200900570

We report the crystal structure of a thermophilic "ene" reductase (TOYE) isolated from Thermoanaerobacter pseudethanolicus E39. The crystal structure reveals a tetrameric enzyme and an active site that is relatively large compared to most other structurally determined and related Old Yellow Enzymes. The enzyme adopts higher order oligomeric states (octamers and dodecamers) in solution, as revealed by sedimentation velocity and multiangle laser light scattering. Bead modelling indicates that the solution structure is consistent with the basic tetrameric structure observed in crystallographic studies and electron microscopy. TOYE is stable at high temperatures (T _m > 70°C) and shows increased resistance to denaturation in water-miscible organic solvents compared to the mesophilic Old Yellow Enzyme family member, pentaerythritol tetranitrate reductase. TOYE has typical ene-reductase properties of the Old Yellow Enzyme family. There is currently major interest in using Old Yellow Enzyme family members in the preparative biocatalysis of a number of activated alkenes. The increased stability of TOYE in organic solvents is advantageous for biotransformations in which water-miscible organic solvents and biphasic reaction conditions are required to both deliver novel substrates and minimize product racemisation.

Full text of DOI:10.1002/cbic.200900570

DOI: 10.1002/cbic.200900570
Biocatalysis with Thermostable Enzymes: Structure and
Properties of a Thermophilic ‘ene’-Reductase related to
Old Yellow Enzyme
Bjçrn V. Adalbjçrnsson,[a] Helen S. Toogood,[a] Anna Fryszkowska,[b] Christopher R. Pudney,[a]
Thomas A. Jowitt,[c] David Leys,[a] and Nigel S. Scrutton*[a]
We report the crystal structure of a thermophilic “ene” reduc-
tase (TOYE) isolated from Thermoanaerobacter pseudethanolicus
E39. The crystal structure reveals a tetrameric enzyme and an
active site that is relatively large compared to most other
structurally determined and related Old Yellow Enzymes. The
enzyme adopts higher order oligomeric states (octamers and
dodecamers) in solution, as revealed by sedimentation velocity
and multiangle laser light scattering. Bead modelling indicates
that the solution structure is consistent with the basic tetra-
meric structure observed in crystallographic studies and elec-
tron microscopy. TOYE is stable at high temperatures (Tm >
708C) and shows increased resistance to denaturation in
water-miscible organic solvents compared to the mesophilic
Old Yellow Enzyme family member, pentaerythritol tetranitrate
reductase. TOYE has typical ene-reductase properties of the
Old Yellow Enzyme family. There is currently major interest in
using Old Yellow Enzyme family members in the preparative
biocatalysis of a number of activated alkenes. The increased
stability of TOYE in organic solvents is advantageous for bio-
transformations in which water-miscible organic solvents and
biphasic reaction conditions are required to both deliver novel
substrates and minimize product racemisation.
Introduction
The use of enzymes as biocatalysts in asymmetric synthesis
has become an established manufacturing route[1] due to their
often high efficiency, as well as the stereo- and enantioselectiv-
ity of their reactions.[2] While enzymes can be found to catalyse
reactions of industrial importance, they rarely have properties
required for large-scale chemical production such as high ac-
tivity and selectivity for non-natural substrates.[3] In addition,
chemical synthesis is often carried out in the presence of or-
ganic solvents and high substrate conditions, and these are
often inhibitory to enzymes.[2]
tions.[8–10,12–18] Members of this family include OYE1 (brewers
yeast),[19] pentaerythritol tetranitrate reductase (PETNR; Entero-
bacter cloacae PB2),[20] morphinone reductase (MR; Pseudomo-
nas putida M10),[21] 12-oxophytodienoate reductase 1 (OPR1;
tomato)[22] and YqjM (Bacillus subtillis).[23] These enzymes have
structures based on an eight-stranded a/b barrel fold (TIM
barrel),[24] and can exist as a functional monomer (for example,
PETNR),[20] dimer (for example, MR)[25] or tetramer (YqjM).[26]
A search of the available sequence databases for potential
OYE family members from thermophilic microorganisms led us
to TOYE, a hitherto uncharacterised NAD(P)H-dependent oxi-
doreductase from Thermoanaerobacter pseudethanolicus E39.[27]
Here we describe the characterisation and biocatalytic poten-
tial of its ”ene-reductase“ activity, as well as the crystal struc-
tures of the apo and nicotinamide cofactor analogue-bound
TOYE. We provide evidence that TOYE is the first known OYE
Reactions carried out at high temperatures can be advanta-
geous in cases in which the substrate or product has poor sol-
ubility at low temperatures and/or the thermodynamics of the
reaction causes higher yield at higher temperatures. In addi-
tion, biotransformations in the presence of organic solvents
might lead to a change in enantioselectivity of the reaction, re-
versal of the thermodynamic equilibrium of hydrolysis reac-
tions and suppression of water-dependent side reactions.[4]
Thermo- and hyperthermophilic enzymes have great potential
for industrial use as their high resistance is not only to temper-
ature, but can also be towards chemicals, organic solvents and
extreme pH values at ambient temperatures.[5–7]
[a] B. V. Adalbjçrnsson, Dr. H. S. Toogood, Dr. C. R. Pudney, Dr. D. Leys,
Prof. N. S. Scrutton
Faculty of Life Sciences, University of Manchester
131 Princess Street, Manchester M1 7DN (UK)
Fax: (+44)161-306-8918
E-mail: nigel.scrutton@manchester.ac.uk
The biocatalytic potential of the Old Yellow Enzyme family
(OYE; EC 1.6.99.1) in the reduction of industrially useful activat-
ed a,b-unsaturated alkenes has been studied extensively.[8–11]
These NAD(P)H-dependent oxidoreductases are useful catalysts
in the reduction of a,b-unsaturated ketones, aldehydes, nitroal-
kenes, carboxylic acids and derivatives, yielding products with
a variety of biotechnological and pharmaceutical applica-
[b] Dr. A. Fryszkowska
School of Chemistry, University of Manchester
131 Princess Street, Manchester M1 7DN (UK)
[c] Dr. T. A. Jowitt
Faculty of Life Sciences, University of Manchester
Michael Smith Building, Oxford Road, Manchester M13 9PT (UK)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.200900570.
ChemBioChem 2010, 11, 197 – 207
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
197

N. S. Scrutton et al.
to exist in solution in large complexes of multiple oligomeric
states (for example, tetramers, octamers). The unique structural
features of TOYE are discussed with reference to both its ther-
mostability and biocatalytic potential.
Table 1. X-ray crystallographic data collection and refinement statistics.
[a]
[a]
Parameters
TOYE-His6
TOYE-NADH4
P21
Space group
Cell dimensions
a, b, c (ꢀ)
a, b, g (o)
P21
86.93, 97.38, 94.39 87.38, 98.56, 95.13
90.00, 92.34, 90.00 90.00, 92.40, 90.00
Results and Discussion
Resolution (ꢁ)
64.8–1.6
68.4–1.8
(1.64–1.60)
6.4 (35.1)
11.7 (3.0)
98.9 (95.0)
2.8 (3.0)
189805
15.2/17.9
(22.5/25.6)
(1.85–1.80)
10.9 (30.2)
11.0 (4.4)
98.9 (99.4)
3.3 (3.4)
141968
15.5/19.4
(18.5/21.8)
General
Rmerge (%)
I/sI
Completeness (%)
Redundancy
Unique reflections
Rwork/Rfree
We performed a database search for OYE enzymes from ther-
mophilic microorganisms in an effort to find a more stable
OYE for industrial biotransformations. TOYE (thermostable OYE)
from T. pseudethanolicus E39[27] was chosen, and recombinant
forms of the protein (native and His6-tagged) were produced
at high yield and purity (Figure S1 in the Supporting Informa-
tion).
RMS deviations
Bond angles (o)
Bond lengths (ꢁ)
Ramachandran plot
Allowed region (%)
1.295
0.010
1.289
0.012
An alignment of eight OYE sequences (Figure S2) belonging
to the two family subclasses[26] shows each contain unique se-
quence and structural features. TOYE belongs to the YqjM sub-
class, and sequence comparisons suggest two other enzymes
from thermophilic microorganisms belong to this subclass,
namely the putative NADH-dependent oxidoreductase en-
zymes from Geobacillus kaustophilus HTA426,[28] and Coprother-
mobacter proteolyticus.[29] Given that YqjM was originally isolat-
ed from the mesophilic B. subtilis,[23] we describe this subclass
as “thermostable-like” OYEs in comparison to “classical” OYEs
such as PETNR and OYE1.
96.2
96.3
3.7
Additionally allowed region (%) 3.7
[a] Highest resolution shell is shown in parentheses. Rmerge =Shkl Si jIi(hkl)-
[I(hkl)]j/Shkl Si Ii(hkl), where Ii(hkl) is the intensity of the ith observation of
unique reflection hkl. Redundancy=total number of reflections/total
unique reflections. Rwork =SjjFobs j-jFcalc jj/SjFobs j, where Fobs and Fcalc are
observed and model structure factors, respectively. Rfree was calculated by
using a randomly selected set (5%) of reflections.
Y331 and A334 (backbone O), which are highly conserved
within the thermostable-like subclass, but not classical OYEs. In
addition, T45 residues of both subunits (S46 in YqjM) interact
through their OG1 atoms as well as side chain interactions be-
tween S28/H42/R46 and Q330/Y331/Y315. The total interaction
surface is 1054 ꢁ2 with a complexation significance score
(CSS)[32] of 1.0; this is consistent with a functional interaction.
While the nature of these interactions is similar in YqjM, the C-
terminal portion of TOYE encompassing helices 8-F is signifi-
cantly different in backbone orientation (Figure 1A). This
region forms an “arm” that interacts with the active site of its
neighbouring monomer to form the functional dimer. In YqjM,
the loop is extended by three amino acids, and is oriented to-
wards the turn T287M291 of the adjacent subunit. Residues
Q323 and L324 (absent in TOYE) interact with residues G289
and R3, respectively, of the adjacent subunit. In TOYE, the
equivalent loop is reoriented towards the central cavity so
these interactions are not possible and the C-terminus is disor-
dered within the central cavity. In spite of the loop differences,
TOYE still contains the key “arginine finger“[26] R333 (R336 in
YqjM), involved in substrate binding which extends into the
active site of the adjacent monomer close to the flavin. In the
holo-enzyme, the density of this residue suggests there are at
least three orientations of the side chain (results not shown).
The density of at least one position of R333 clashes significant-
ly with the FMN; this likely reflects the observation that the re-
combinant enzyme is not fully flavinated (FMN occupancy ap-
proximately 69%). In NADH4–TOYE the density for most of the
side chain in all but subunit B is disordered. This inherent
Overall structure of TOYE
We determined the X-ray crystal structures of both holo-TOYE
(His6-tagged) and OYE inhibitor 1,4,5,6-tetrahydroNAD
(NADH4)-bound enzyme to high resolution (1.6 and 1.8 ꢁ, re-
spectively). The data collection and refinement statistics can be
found in Table 1. Both structures show a homotetrameric ar-
rangement composed of a dimer of dimers (Figure 1A; Figure
S3A–B). Each monomer is comprised of a compact domain of
an (a/b)8-barrel fold (TIM barrel), typical for OYE en-
zymes.[24–26,30] Both structures are virtually identical, and are
very similar to the other known tetrameric OYE YqjM (50% se-
quence identity),[26] with RMSD values of 0.259 and 0.924, re-
spectively. However, the angle of interaction between mono-
mers within the functional dimers of TOYE and YqjM differ
with a typical separation distance of about 2 ꢁ (Figure 1A).
Other significant backbone changes between TOYE and Yqjm
are highlighted in Figure 1A.
Subunit interactions
Like YqjM, the tetramer contains a water-filled central hole or
cavity, with the active sites opening into this cavity. The (a/b)8-
domains (YqjM numbering of helices and sheets throughout)[26]
of each monomer within the functional dimer interact with
each other with same sides facing. As a consequence, a net-
work of hydrogen bonds that involve the same residues from
both subunits, holds the monomers together. Residues in-
volved in this interaction include the C-terminal residues Q330,
198
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 197 – 207

Characterisation of a Thermophilic Old Yellow Enzyme
as extensive hydrogen bonds
between the same residues on
opposite monomers (Y261,
G263, K267, T287, T288, E290
and N298). In addition, salt
bridges are present and involve
residues K267, E270 and R300.
The electron density for both
E270 and R300 shows two dis-
tinct conformations of each,
with R300 of each subunit inter-
acting with one conformation
of the other; this results in a
“diamond“ pattern of the two
interactions (results not shown).
The total interaction surface is
only 776 ꢁ2 and has a CSS value
of only 0.325.[32]
Active site
Both TOYE and YqjM contain
large, easily accessible hydro-
phobic active sites, with sub-
strate entry through the central
cavity.[26] This differs from the
classical OYEs that have signifi-
cantly smaller active sites due
to the shortening and reorienta-
tion of b-strand 6 to the turn
before b-strand 7, as well as a
reorientation of the slightly
larger turn between b-strand 5
and a-helix 5 (Figure 1B). The
active site contains a noncova-
lently bound FMN molecule
(Figures 1B, S3C), in a position
similar to other OYEs, with its si-
face directed towards the sol-
vent. The isoalloxazine ring ex-
Figure 1. X-ray crystal structures of TOYE. A) Superimposition of functional dimers of TOYE-His6 and YqjM (PDB
code 1Z41). TOYE is shown as a rainbow-coloured cartoon from N- (blue) to C-terminus (red) while YqjM is shown
as a pale pink ribbon. Helix and residue numbering refers to the YqjM and TOYE structures, respectively. B) Posi-
tion of the bound NADH4 cofactor analogue in the active site of TOYE superimposed with the PETNR structure
(PDB code 1H50). The omit jFo j–jFc j map of NADH4 is contoured at three s (green mesh). TOYE residues and
NADH4/FMN are shown as coloured sticks with green and magenta/yellow carbons, respectively, while water is
shown as red spheres. PETNR residues are shown as coloured lines with blue carbons. Residues of PETNR that
differ in amino acid identity are labelled. PETNR residues forming a loop overlapping the adenosyl-portion of the
NADH4 binding site are shown as a blue ribbon. C) Superimposition of TOYE and MR at the NADH4 binding re-
gions between two monomers. The two NADH4 molecules are shown as coloured sticks with magenta and grey
carbons for TOYE and MR, respectively. TOYE and MR are shown as blue and pale pink cartoons, respectively. The
two bright pink loops refer to MR residues absent in TOYE that are involved in creating a surface groove where
NADH4 is positioned. All figures were generated in Pymol.[31]
hibits
a nonlinear “butterfly-
bend” shape as seen in YqjM.[26]
The interactions between the
FMN and the surrounding pro-
tein/solvent
are
extensive
(Table S1), and are similar to
those in YqjM. Differences in-
clude backbone interactions be-
tween P23 and C25 with FMN
N1 and N5 atoms, respectively,
flexibility might have a functional consequence to enable it to
interact with a wide range of substrates.
while in YqjM N1 and N3 interact with the highly conserved
substrate-binding residues H167 and H164 (H166 and H163 in
TOYE), respectively. The electron density for the FMN showed
it was partially occupied (80%) with minor additional density
for an alternative conformation of the ribityl chain, except for
the terminal phosphate atoms that are anchored in place by
multiple interactions (Table S2).
The interactions between nonfunctional dimers involve
helix 7 (residues 287–300) of both subunits as well as helix 6
residues (Y261K274) of one subunit and Y261V266 of the adja-
cent subunit. These interactions encompass both hydrophobic
interactions (for example, Y261, P262, Y264 and L291) as well
ChemBioChem 2010, 11, 197 – 207
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
199

N. S. Scrutton et al.
The omit map jFo j–jFc j electron density suggests a small
molecule such as a formate or acetate ion might be bound on
the si-face of the isoalloxazine ring of FMN compared to the
presence of a sulphate ion in YqjM.[26] Therefore an acetate
molecule has been modelled into each subunit of holo-TOYE
(Figure S3C), in a position similar to holo-PETNR (PDB code:
1H50),[33] although the density suggests some mobility of the
molecule. In this position, an acetate oxygen atom interacts
with both H163 NE2 and H166 ND1 and the other oxygen with
water molecule/s (Figure S3C/Table S1).
teract with both phosphate groups through O1P/O1N atoms
of NADH4 in chains B and D (side chain is more disordered in
chains A and C). However, the quality of the density of the
later ribose suggests a higher mobility than the adjacent phos-
phate atoms; this is likely due to its location on the surface of
the central cavity.
A combination of these interactions cause the inhibitor to
fold back upon itself at the solvent-exposed central cavity in-
terface; this would position both the adenosine base and its
ribose directly beneath the d-ribose ring, with no density seen
for the adjoining atoms C5’ and O5’. We modelled one of the
possible positions of the O5’ and C5’ in subunit B and found
some constraint in the C4’ꢀC5’ꢀO5’ bond angle, although it
was within the allowable range (results not shown). These con-
straints are likely a consequence of the multiple interactions
holding both ends of the inhibitor in position. The poorer den-
sity for the adenosyl ribose in the remaining subunits and the
absence of density for C5’ in all subunits reflects the higher
mobility of the adenosyl portion of the inhibitor in this sol-
vent-exposed region.
A superimposition of members of the two subclasses of OYE
(TOYE and PETNR) show some important differences in the res-
idue composition lining the active site (Figure 1B) that are
often highly conserved within each subclass. The key substrate
binding residues H163/H166 were conserved within the TOYE
subclass, but were sometimes a histidine/asparagine pair
(H186/N189; MR numbering) in classical OYEs. The putative
proton donor is highly conserved (Y168; TOYE numbering)
while the conserved Y351 involved in substrate/inhibitor bind-
ing of classical OYEs is absent in the smaller molecular mass
TOYE subclass. Interestingly, residue Y27 is located near Y351
(PETNR), and is only conserved in thermostable-like OYEs. The
highly conserved T26 (PETNR numbering), known to interact
with the FMN N3 atom, has been replaced by a cysteine in the
TOYE subclass. Two other highly conserved hydrophobic resi-
dues W102 and Y68 (PETNR numbering), have been replaced
by smaller amino acids alanine/glycine and threonine/serine,
respectively, with lower sequence conservation. The differences
and conservation in the nature of the residues in the active
site between the two subclasses likely reflects the substrate
binding and/or shape and size of the active site. In particular,
the substitution of W102 in PETNR for alanine in TOYE might
be important in substrate binding as seen in recent work in
which the mutagenesis of this highly conserved residue in
OYE1 yielded enzymes with altered stereochemical out-
comes.[15]
These interactions are in stark contrast to the orientation of
binding of NADH4 to the classical OYE MR,[34] in which the
phosphate atoms have reoriented; this causes both the ribose
and its adenine base to lie above the nicotinamide portion
within a solvent-exposed groove on the surface of the protein
(Figure 1C). While the interactions with NADH4 in this region
are purely through the solvent, this groove is surrounded by
two “loops,” the size and orientation of which differ significant-
ly in the thermostable OYE subclass. Importantly, the small b-
strand C between b-strand 3 and a-helix 3 is considerably elon-
gated in classical OYEs, and this allows the “loop” to form. Sec-
ondly, the differences in the highly conserved C-terminal loop
Q330–A334, which is not conserved in classical OYEs, causes a
change in the orientation of this loop compared to MR. How-
ever, the position of the nicotinamide moiety in MR is similar
to TOYE, with interactions between the equivalent H186/N189
residues (MR) with the carbonyl-oxygen, and a third interaction
between the substrate interacting residue Y356 (absent in
TOYE) and the pyrophosphate moiety. Interestingly, the inhibi-
tor is unlikely to bind to classical OYEs, such as PETNR, in a
TOYE-like manner due to the altered position and elongation
of b-strand 6 that would overlap with the adenosine portion of
NADH4 in TOYE (blue ribbon in Figure 1B).
Inhibitor binding
The omit map jFo j–jFc j electron density of the NADH4-bound
TOYE crystal structure showed clear density for the inhibitor,
except for the phosphate atom O5’ and the adjacent carbon
atom C5’ in subunit B, which is disordered in this structure
(Figure 1B). In the remaining subunits there is poor density for
the adenosyl ribose so it was only modelled into subunit B.
The inhibitor makes multiple interactions between both the
protein and the solvent (Table S2). The highly conserved H163
and H166 residues interact with the carbonyl oxygen of the
nicotinamide moiety, while residues Y264, Q265 and S250 in-
teract with the N6A and N7A atoms of the adenine base. In
this position, the nicotinamide atom C4N (equivalent to the
hydride transferring carbon in NAD(P)H) is within 3.7 ꢁ of the
FMN N5 atom. Additional interactions include a hydrogen
bond between the O2’ atom of the adenosyl ribose and the
backbone O of V256. A significant difference between the
holo- and NADH4-bound structures is the position of the “argi-
nine finger” R333 the side chain atoms NH1/NH2 of which in-
Oligomeric states of TOYE
Both SDS PAGE and native PAGE analysis of TOYE showed the
presence of multiple bands, which were positively identified to
be TOYE. This suggests TOYE exists in vitro in multiple oligo-
meric states. To investigate this phenomenon we performed
multi-angle laser light scattering (MALLS) and sedimentation
velocity analysis on TOYE (Figure 2A; Table S3). These results
clearly showed the minimal quaternary structure of TOYE to be
a tetramer, but also higher species such as octamers and do-
decamers were present in lower proportions according to
MALLS and native PAGE, respectively (Figure 2A–B). Both
methods also detected a very minor species of a hexamer,
200
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 197 – 207

Characterisation of a Thermophilic Old Yellow Enzyme
clear axis of pseudo mirror symmetry. There was no evidence
for elongation of the basic tetramer into fibres or extended
chains. These data are consistent with a basic tetrameric parti-
cle that forms a 2-layered stack with a probable d-symmetrical
arrangement of rotational and flipped axes, consistent with
the bead modelling.
Solvent stability of TOYE
We investigated the thermostability of TOYE by monitoring the
secondary structure unfolding/aggregation with circular di-
chroism (CD), differential scanning calorimetry (DSC) and fluo-
rescence changes of FMN/Trp residues in TOYE (Figures S6–S7
and see discussion of “Thermostability determinants of TOYE”
in the Supporting Information). Observations with other
enzyme systems suggest that thermo- and hyperthermophilic
enzymes tend to have higher stability towards organic solvents
at ambient temperatures.[36–39] We therefore compared the sta-
bility of both the mesophilic PETNR and thermophilic TOYE in
the presence of water-miscible organic solvent:aqueous buffer
systems by using steady-state kinetics methods (Figure S8).
Steady-state reactions in the presence of ethanol showed a
steady decline in reaction rate with increasing solvent concen-
tration for both TOYE and PETNR (Figure S8), with 50% loss at
approximately 45% and 20% ethanol, respectively. PETNR pre-
cipitated significantly above 40% ethanol, while remarkably
TOYE still maintained approximately 20% of its activity in 80%
ethanol. In addition, the complete recovery of activity of TOYE
after a preincubation in up to 70% ethanol shows a dramatic
increase in solvent stability of TOYE compared to PETNR. This
can potentially be exploited in industrial biotransformations in
which organic solvents are often used to increase substrate
solubility and decrease unfavourable water-dependent reac-
tions.
Figure 2. A) MALLS analysis of TOYE-His6. Inset: Bead models of the quater-
nary structures for TOYE were generated by using the PDB file and the
SOMO solution bead modelling program.[35] The crystal structure oligomers
are shown as cartoons and were arranged in several orientations by using
Pymol.[31] B) Native PAGE analysis of TOYE-His6. Lane 1=native PAGE molec-
ular mass markers (Invitrogen) with the masses of the proteins indicated to
the left of the lane. Lane 2=TOYE. C) Two selected reference free projection
averages of tetrameric/octameric TOYE calculated from a complete dataset
(N=1450) with likely top-view (above) and side-view (below) projections.
Scale: the width of each box is 25 nm.
Substrate specificity
Members of the OYE family typically catalyse C=C bond reduc-
tion of a,b-unsaturated ketones, aldehydes, nitroalkenes, car-
boxylic acids and derivatives[8–10,12–18] through a net trans-addi-
tion of 2H[40] (Scheme 1). To benchmark the biocatalytic
which is presumed to be due to partially unfolded protein
and/or incorrectly assembled multimers. This is the first exam-
ple of an OYE demonstrated to exist in vitro in a higher order
than a tetramer.
We performed bead modelling of the tetramer and octamer
to ascertain if the hydrodynamics data were consistent with
the crystal structure (Figure 2A inset). The modelled tetramer
had an area of 108 nm2, consistent with the crystal structure.
Modelled radii (RH) and sedimentation coefficients (So) of both
the tetramer and octamer are consistent with experimental
data (Table S3). Some models of the octamer were rejected
based on poor fitting to the hydrodynamic data (Figure S4).
The modelled octamer is assembled as a dimer of tetramers
with the central hole remaining intact. We also analysed TOYE
by transmission electron microscopy (TEM), which showed a
homogeneous population of clear squared single particles with
a diameter of 10–11 nm (Figures 2C, S5). Image analysis of the
data identified a particle with clear C4 rotational symmetry
with a corresponding projection with a similar diameter, but a
Scheme 1. General mechanism and stereochemistry of a,b-unsaturated sub-
strate reduction by OYEs. a) Substrates 4a–9a; b) Substrates 1a–3a; panel
c) Substrates 10a–11 a; d) Substrates 12a–15a; e) Substrate 18a.
ChemBioChem 2010, 11, 197 – 207
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
201

N. S. Scrutton et al.
potential of TOYE against other OYEs, we investigated the
anaerobic reduction of a variety of a,b-unsaturated activated
olefins. Initially, we determined the kinetic constants of the re-
ductive half-reactions (klim and ks) with NAD(P)H (Table S4 and
Figure S5) and steady-state parameters (apparent kcat/KM) with
oxidising substrates to determine the substrate specificity of
TOYE (Table 2).
Table 2. Steady-state kinetics of TOYE with a,b-unsaturated alkene sub-
strates.
[b]
Substrate[a]
Rate [sꢀ1
]
kcat/KM [sꢀ1 mmꢀ1
]
1a
2a
0.74ꢁ0.07
0.17ꢁ0.01
0.6
0.1
ND
In stopped-flow studies, NADPH preferentially reduces TOYE
over NADH, with klim values of 31.7ꢁ2.1 and 2.8ꢁ0.1 sꢀ1, re-
spectively, at 258C. These results compare to YqjM, which has
a comparable klim value,[23] although a tenfold higher Ks value
than TOYE (Table S4 and Figure S9). In addition, the klim and Ks
values of TOYE-His6 with NADH at 508C were 18.4ꢁ0.2 sꢀ1 and
186.1ꢁ8.1 mm, respectively; these values show a greater than
sixfold rate acceleration and threefold increased Ks value com-
pared to reactions at 258C. As the NADH4-TOYE crystal struc-
ture shows a relatively high mobility for the adenosyl-ribose
moiety of the NADH analogue, the mechanism of discrimina-
tion between NADH and NADPH is not clear.
3a
4a
<0.01
1.18ꢁ0.09
(2.29ꢁ0.35)
0.5
(1.9)
5a
6a
ND
ND
ND
ND
(5R)-7a
99% ee
2.23ꢁ0.10
0.83ꢁ0.07
1.5
0.6
When His6-tagged at the C-terminus, TOYE maintains both
its tetrameric/octameric structural integrity and its catalytic
ability as seen by the near identical kinetic parameters of the
reductive half-reaction with NADPH compared to enzyme lack-
ing the His6 tag (Table S4). This is in contrast to YqjM-His6,
which loses both structural integrity and catalytic efficiency.[41]
These differences might be due to the significant reorientation
of the C-terminal region of TOYE (helices 8-F; Figure 1A),
which in YqjM are involved in subunit–subunit interactions.
The location of the last visible amino acid K336 at the surface
of the central cavity in the crystal structure of TOYE suggests
the His6 tag is disordered within this cavity and therefore does
not disrupt the quaternary structure.
(5S)-8a
99% ee
9a
0.02ꢁ0.01
ND
10a
26.85ꢁ3.02
194.4
The steady-state parameters for TOYE with oxidising sub-
strates showed on average lower rates and specificity (kcat/KM)
than PETNR[42] or YqjM[26] at 258C as expected for thermophilic
enzyme catalysis at mesophilic temperatures (Table 2). For ex-
ample, the rate with 2-cyclohexenone 4a with TOYE at 258C
was 1.18ꢁ0.09 sꢀ1 compared to 6.1 and 4.4 sꢀ1 for PETNR and
YqjM, respectively.[26,42] However, at 508C this rate had in-
creased to 2.29ꢁ0.35 sꢀ1. TOYE showed a marked preference
for 2-methyl cyclopentenone 2a over the 3-methyl substituted
substrates 3a, 5a, and 6a, with almost no detectable activity
against the latter three substrates (Table 2). This is similar to
other OYEs such as OPR1, OPR3 from Lycopersicon esculen-
tum,[8] and YqjM, which did not significantly reduce substrates
3a and 5a.[8] In contrast, classical OYEs such as OYE1-3[14] and
PETNR[42] can reduce substrates 3a and 5a, although to a
lower yield than the equivalent 2-methyl substituted com-
pounds. A superimposition of the structures of TOYE and
PETNR suggests that this lack of activity against 3-methyl sub-
stituted substrates by TOYE might be due to a possible clash
between the side chain of I67 and the substituent of the sub-
strate, an interaction not possible in PETNR due to a shift in
the position of the backbone of the protein chain and the sub-
stitution of this residue for glycine.
11 a
16.90ꢁ0.61
213.6
12a
13a
14a
0.33ꢁ0.02
0.39ꢁ0.01
0.22ꢁ0.01
0.3
ND
0.14
15a
0.04ꢁ0.01
0.05
16a
17a
ND
ND
ND
<0.01
[a] Reactions (0.3 mL) were performed in 50 mm KH2PO4/K2HPO4 pH 7.0
containing 100 mm NADPH, 1 mm oxidising substrate and 0.1–2 mm TOYE
for 1 min at 258C. [b] Specificity constants (apparent kcat/KM) were deter-
mined from the slope of the rate versus substrate concentration plot at
substrate concentrations well below the estimated KM (5–20 mm). ND=No
significant activity detected under steady-state conditions.
ity for the phenyl substituted substrate 11 a. These rates were
surprisingly higher than PETNR, in spite of the reactions being
conducted 448C below the optimum growth temperature for
the source organism (698C). Other differences include the
Like other OYEs,[8,11,14] TOYE is highly active towards the mal-
eimides 10a–11 a, with a slightly lower rate but higher specific-
202
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 197 – 207

Characterisation of a Thermophilic Old Yellow Enzyme
relatively poor rate of TOYE towards ketoisophorone 9a under
steady-state conditions; this contrasts with other OYEs, such as
PETNR, that have high activity towards this substrate.[42] TOYE
also has a relatively high rate of catalysis with both (5R)- and
(5S)-carvone compared to PETNR. The origin of these differen-
ces is likely multifactorial such as the change in the shape and
composition of residues lining the active site combined with
the differences in the relative flexibility of the thermophilic
versus mesophilic proteins.
rable enantiopurities (45–64% ee).[8] The reduced enantiopurity
of the products is not surprising given the observation that 2-
methyl-substituted products are subject to nonenzymatic race-
misation.[42] However, in the case of YqjM, the use of the cofac-
tor regeneration system NADP+/glucose-6-phosphate dehydro-
genase (G6PDH) dramatically improved the enantiopurity of
the product from 55% (NADH) to 94%;[8] this highlights the
importance of optimising the reaction conditions, including
the source of reducing equivalents.
TOYE did not significantly reduce substrates 3a and 5a, sim-
ilar to the enzymes YqjM, OPR1 and OPR3, while other classical
OYEs produced low–medium yields of the respective (S)-enan-
tiomeric products with>99% ee.[8,14] The reduction of carvones
7a to 8a proceeded with high conversion rates and high
enantiopurity (85 to 95% de for 8b and 7b, respectively), al-
though the product yields were lower than expected. These
two diastereoisomeric products had the same configuration on
carbon C2 (2R; new chiral carbon generated); this suggests
that the presence of bulky substituents in this position does
not affect the binding mode of the substrate in the active site
of TOYE.
Stereochemistry of C=C reduction
We carried out biotransformations of TOYE at 308C against a
variety of activated a,b-unsaturated alkenes, by using NADH as
the source of reducing equivalents, to determine the efficiency
and enantioselectivity of C=C reduction (Table 3). Like YqjM,
the (S)-enantiomer of 2b was produced with moderately high
enantioselectivity (70%).
This is the opposite enantiomeric product as expected from
Scheme 1, but was also seen with classical OYEs such as PETNR
and likely arises from an alternative or “flipped” binding mode
of 2a in the enzyme active site.[43] Classical OYEs such as OYE1-
The reduction of ketoisophorone 9a by TOYE yielded the ex-
pected (R)-9b product with a low enantiopurity (26%) as seen
with other OYEs.[8,14] The poor ee likely arises from significant
solvent-catalysed racemisation
3
produce racemic mixtures of product 2b by using
NAD(P)H,[14] while OPR1 and OPR3 produce (S)-2b with compa-
of the product, although recent
reaction optimisation studies
with PETNR have shown that a
reduction of the reaction time
(thereby reducing substrate:wa-
ter exposure time) can dramati-
cally improve the enantiopurity
to>93%.[42] In the case of YqjM,
substitution of NADPH for the
Table 3. Product determination of the reduction of a,b-unsaturated alkene substrates by TOYE.
Substrate
Product[a]
Conv.[b,c]
[%]
Yield[c]
[%]
ee[d]
[%]
2a
(S)-2b
77
94
73
61
70
(5R)-7a
99% ee
(2R,5R)-7b
99% ee
95 de
85 de
26
cofactor
recycling
system
NADP+/G6PDH was sufficient to
improve the enantiopurity of
(R)-9b from 37% to 97%.[8] In
comparison, the reduction of
the N-phenyl substituted malei-
mide 10a by TOYE yielded the
(R)-enantiomer with characteris-
tically high yields and enantio-
purity. This is due to both the
lack of significant product race-
misation and the observation
that at least two likely modes of
binding of this substrate would
yield the same enantiomeric
product.[42]
(5S)-8a
99% ee
(2R,5S)-8b
99% ee
89
77
9a
(R)-9b
99
74
90
10a
(R)-10b
>99
>99
14a
15a
(S)-14b
(S)-15b
(S)-18b
97
74
82
82
23
12
55
91
61
The reduction of 2-methyl
pentenal 14a proceeded with
high efficiency and yield, al-
though with only moderate
enantiopurity of (S)-14b (55%)
as seen with PETNR.[42] Such
(E)-18a
[a] Standard reactions (1.0 mL) were performed in buffer (50 mm KH2PO4/K2HPO4 pH 7.0) containing alkene
(5 mm; added as a 2% DMF solution), NADH (6 mm) and TOYE (10 mm) at 308C for 24 h at 130 rpm. [b] Conv.=
% conversion. [c] by GC using DB-Wax column. [d] by GC or HPLC.
poor
stereoselectivity
is
ChemBioChem 2010, 11, 197 – 207
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
203

N. S. Scrutton et al.
common with this substrate as seen with OPR3 and YqjM,
which produced the (S)- and (R)-14b enantiomeric products
with only 19 and 10% ee, respectively, due to product racemi-
sation.[8] Biotransformations of the citral 15a generated citro-
nellal (S)-15b in high optical purity (91% ee), but in low yield
(23%). GC-MS analysis did not reveal any volatile byproducts;
this suggests the low yield might result from possible side re-
actions that give rise to condensation products for example,
nonenzymatic substrate/product decomposition in aqueous
media.
Table 4. Product determination of the reduction of ketoisophorone 9a
to (R)-9b[a] by TOYE in the presence of water-miscible solvent systems.
Solvent [%][b]
Sub. [mm] Time [h] Conv. [%][c] Yield [%][c] ee[d]
[%]
Time course
DMF 2
DMF 2
DMF 2
DMF 2
5
5
5
5
5
5
0.5
1.0
2
3
6
13
37
71
89
96
97
12
34
68
83
88
85
>99
95
92
88
76
DMF 2
DMF 2
9
64
The reduction of the nitroalkene (E)-18a by TOYE proceeded
with a good conversion (82%) and moderate enantiopurity
(61%) of the product (S)-18b, although with poor yield. By-
products detected in the assays with TOYE were identified
mostly as 2-phenylpropioaldehyde (that is, a Nef reaction prod-
uct) with its equivalent oxime, plus some other as yet unidenti-
fied products. In addition, acetophenone was detected in both
enzyme assays and the respective enzyme-free controls. Its
presence suggests that substrate 18a is relatively unstable
under aqueous reaction conditions and possibly undergoes hy-
dration followed by a retro-Henry reaction. The predominant
formation of the (S)-enantiomer is consistent with the reactions
with PETNR, NCR, OPR3 and YqjM, but opposite to those of
OYE1-3 and OPR1, which produced the (R)-enantiomer with
higher ees.[8,11,14]
Substrate
concentration
DMF 4
10
2
54
43
93
Solvent concentration
DMF 5
DMF 10
DMF 15
DMF 20
5
5
5
5
5
2
2
2
2
2
66
48
40
38
32
64
42
32
23
12
92
92
92
93
92
DMF 30
Solvent comparison
DMF 10
THF 10
MeOH 10
EtOH 10
PrOH 10
5
5
5
5
5
5
5
2
2
2
2
2
2
2
44
76
74
70
63
41
81
37
69
66
65
54
36
66
93
93
92
95
90
93
93
Acet. 10
ACN 10
[a] Standard reactions (1.0 mL) were performed in buffer (50 mm KH2PO4/
K2HPO4 pH 7.0) containing water-miscible organic solvents (2–50%), ke-
toisophorone 9a (5-10 mm), NADH (6 mm) and TOYE (10 mm) at 308C for
0.5–9 h at 130 rpm. Substrate was added to all reactions as a 2% final
DMF solution. Sub.=substrate concentration; Conv.= % conversion.
[b] DMF=dimethyl formamide; THF=tetrahydrofuran; MeOH=metha-
nol; EtOH=ethanol; PrOH=propanol; Acet.=acetone; ACN=acetoni-
trile. [c] by GC using a DB-Wax column. [d] by GC using a Chirasil-DEX CB
column.
Biotransformations in water-miscible organic solvents
Comparative biotransformations of TOYE at 30 and 508C
showed a reduction in both product yield and enantiopurity at
high temperatures (Table S5 and see discussion “High tempera-
ture biotransformations” in the Supporting Information). There-
fore we decided to exploit the improved solvent stability of
thermostable TOYE. Previous studies with PETNR have shown
the enantiopurity of the product (R)-9b can be dramatically
improved by reaction optimisation such as a decrease in reac-
tion time[42] or by using biphasic solvent systems.[42] Given the
markedly increased solvent stability of TOYE over PETNR, we
investigated biotransformations of TOYE with substrate 9a at
308C varying the time, substrate concentration, solvent con-
centration and the nature of the solvent (Table 4). We used
water-miscible organic solvents rather than a biphasic system
to avoid the decrease in reaction rate attributed to poor mass
transfer.
ried out further comparative work in the presence of organic
solvents by using a 2 h incubation time.
A comparison of the reduction of 9a by TOYE in the pres-
ence of 10% water-miscible organic solvents shows that the
yield of products was significantly affected by dimethylforma-
mide (DMF), acetone and to a lesser extent by propanol (36–
54% yield) compared to a 68% yield under standard reaction
conditions of 2% DMF. In contrast, the presence of 10% tetra-
hydrofuran, methanol, ethanol and acetonitrile gave compara-
tive conversions, yields and enantiopurities as standard reac-
tions; this suggests the choice of cosolvent is important not
only for substrate solubility but also for enzyme activity. Reac-
tions with higher substrate and solvent concentration showed
lower percent conversion for the same enzyme concentration,
but increased overall yield of product; this suggests enzyme
concentration optimisation is required.
The effect of reaction time on the enantiopurity of the prod-
uct (R)-9b shows a time-dependent loss of ee; this has been
previously shown to be due to nonenzymatic racemisation of
the product.[42] The reactions were near completion after about
6 h; this reflects the overall slower rate of reaction seen under
steady-state conditions (Table 2) compared to other OYEs[42] as
expected for a thermostable enzyme. Reaction optimisation
studies of PETNR showed that for products susceptible to race-
misation, a decrease in the reaction time from 24 h to 2 h was
sufficient to improve the ee from 29% to 98%.[42] Similar results
were seen for TOYE, which showed an improvement of ee
from 26% to 92% under these conditions. Therefore, we car-
We have shown that the reaction rate of TOYE decreases
with increasing solvent concentration under steady-state con-
ditions (2 min; FigureS8). To test the effect of solvent concen-
tration under biotransformation conditions we carried out
204
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 197 – 207

Characterisation of a Thermophilic Old Yellow Enzyme
GTGGTGCTCTCATTTTTTGAAGGCACGTTC-3’ (Eurofins MWG Operon,
Ebersberg, Germany). All constructs were transformed into the
E. coli strain Arctic Express (Stratagene) for soluble protein overex-
pression according to the manufacturers protocol.
reductions of 9a in the presence of 2–30% DMF for 2 h at
308C (Table 4). As expected the percent conversion and yields
decreased as the solvent concentration increased, with no
effect on the product enantiopurity. While DMF is not the sol-
vent of choice for TOYE, these results show that biotransforma-
tions can be carried out in the presence of significant quanti-
ties of water-miscible organic solvents to yield a highly enan-
tiopure product with potentially high yields after compensato-
ry enzyme concentration increase. Thus the higher solvent sta-
bility of TOYE compared to classical OYEs can be exploited in
industrial biotransformations by using solvent concentrations
normally incompatible with mesophilic enzyme structural in-
tegrity to facilitate higher substrate/product solubility.
Protein production and purification: Native and His6-tagged
TOYE cultures were grown in Luria-Bertani medium containing the
antibiotics carbenicillin (50 mgmLꢀ1) and gentamicin (20 mgmLꢀ1).
Cultures were incubated at 258C until OD600 reached 0.5, followed
by a 12 h induction with isopropyl b-d-1-thiogalactopyranoside
(IPTG; 0.4 mm) at 258C. Cells were harvested by centrifugation at
6000 g for 30 min at 58C (Beckman Coulter, Avanti J-26 XP). Cell
pellets were resuspended in lysis buffer (50 mm KH2PO4/K2HPO4
pH 8.0) containing the EDTA-free complete protease inhibitor cock-
tail (Roche), DNase I (0.1 mgmLꢀ1) and lysozyme (1 mgmLꢀ1) and
incubated for 30 min at 48C. Cells were disrupted by sonication
(Sonics Vibra Cell, Newtown, CT, USA) followed by extract clarifica-
tion by centrifugation for 90 min at 26600 g. To improve the
degree of flavination of recombinant TOYE, excess free FMN was
added to the cell extracts.
Conclusions
We have identified and solved the structure of a novel thermo-
stable old yellow enzyme (TOYE) and characterized its poten-
tial as an industrial biocatalyst for the asymmetric bioreduction
of a,b-unsaturated ketones, aldehydes and (E)-1-nitro-2-phenyl-
propene. The enzyme forms multiple oligomeric states (tetram-
ers to dodecamers), previously unseen among the OYE family.
The relatively large active site and increased general protein
and solvent stability make this enzyme a natural target for in-
dustrial-scale biotransformations. TOYE can reduce a variety of
a,b-unsaturated ketones, aldehydes and the nitroalkene 18a in
a manner similar to other OYEs, although with a slightly re-
duced substrate range. TOYE has increased stability in water-
miscible organic solvents that should be advantageous for bio-
transformations where water-miscible organic solvents and bi-
phasic reaction conditions are required to deliver novel sub-
strates to this ene-reductase catalyst.
Native TOYE was purified by binding to a Q-Sepharose column
(50 mL; GE Healthcare), pre-equilibrated in buffer A (50 mm Tris,
pH 8.0, containing 25 mm NaCl). Proteins were separated by elu-
tion in a gradient of NaCl (25–200 mm) in buffer A. TOYE-His6 was
purified by binding to a Ni-NTA column (25 mL; QIAGEN), pre-equi-
librated in buffer B (50 mm KH2PO4/K2HPO4, pH 8.0, containing
40 mm imidazole and 0.3m NaCl). TOYE was eluted in a step to
buffer C (50 mmK H2PO4/K2HPO4 pH 8.0 containing 250 mm imida-
zole and 0.3m NaCl). Purity was assessed by SDS-PAGE and native
PAGE and the ratio of total protein versus flavinated (active) pro-
tein was determined by the BioRad (Hertfordshire, U.K.) protein
assay kit according to the manufacturers protocol and using its
flavin extinction coefficient, respectively. PETNR was prepared as
described previously.[11] The methods of the positive identification
of all the major protein bands of purified TOYE on SDS-PAGE are
described in the Supporting Information Methods.
Crystallogenesis and data collection: Crystals of oxidised TOYE
were grown by using the sitting-drop method in magnesium for-
mate (50 mm) or NaOAc (0.1m pH 4.6) containing CaCl2 (0.2m), iso-
propanol (20%) and ethylene glycol (12%) for three days at 208C.
The crystals were soaked in mother liquor containing FMN and
PEG 200 (10%), the later as a cryoprotectant, +/ꢀ saturating levels
of the inhibitor NADH4 and flash frozen in liquid nitrogen. Full
TOYE-His6 (1.6 ꢁ) and inhibitor-bound TOYE (1.8 ꢁ) X-ray diffraction
data sets were collected from single crystals at the European Syn-
chrotron Radiation Facility (Grenoble, France) on Station ID 14.4
(wavelength 0.97 ꢁ; 100 K) by using an ADSC CCD detector.
Experimental Section
General: All reagents were of analytical grade. All the solutions for
enzyme kinetics were made anaerobic and the reactions were set-
up in an oxygen free environment (<5 ppm O2). TOYE was deoxy-
genated by passage through a BioRad 10DG column equilibrated
in anaerobic reaction buffer to avoid unproductive flavin reoxida-
tion during oxidizing substrate reduction. The substrate extinction
coefficients used were as described previously.[44] The flavin extinc-
tion coefficient (e456) used for TOYE was 11300mꢀ1 cmꢀ1. The inhibi-
tor NADH4 was synthesised as described previously.[34] Substrates
and products were either obtained from Aldrich or synthesised ac-
cording to the methods previously described.[42]
Structure determination and refinement: Data was processed
and scaled by using the programs MOSFLM[45] and Scala.[46] The
structures were solved through molecular replacement by using
the program MolRep[46] with the coordinates for a Swiss-MODEL-
generated pdb file of TOYE (http://swissmodel.expasy. org). This
model was generated based on the known 3-D structure of the
OYE Yqjm from B. subtilis.[26] Model rebuilding and water addition
was performed automatically by using REFMAC combined with
ARP/warp.[47] Positional and isotropic B-factor refinement was per-
formed by using REFMAC5[48] (hydrogens included in the refine-
ment for the holo-TOYE structure), with alternate rounds of
manual rebuilding of the model in COOT.[48] The final models were
refined to 1.60 ꢁ and 1.80 ꢁ resolution to give final Rfactor/Rfree of
15.2/17.9 and 15.5/19.4 for His6- and inhibitor-bound TOYE, respec-
tively. The interactions involved in the subunit–subunit interfaces
Gene synthesis and mutagenesis: The protein sequence for the
thermostable NADH:flavin oxidoreductase enzyme TOYE (accession
number: ZP 00777979) from T. pseudethanolicus (ATCC 33223) was
obtained from PubMed (http://www.ncbi.nlm.nih.gov). The gene
sequence was designed and synthesised by Entelechon GmbH (Re-
gensburg, Germany), which incorporated codon optimisation tech-
niques of rare codon removal for optimal expression in Escherichia
coli. The gene was cloned into pET21b (Novagen) through NdeI/
XhoI restriction sites without a stop codon to incorporate a C-ter-
minal His6 tag. To produce a non-His6-tag version of TOYE, a stop
codon was incorporated by using the QuikChange mutagenesis
strategy (Stratagene) with the primers 5’-GAACGTGCCTT-
CAAAAAATGAGAGCACCACCACCACCAC-3’ and 5’-GTGGTGGTG-
ChemBioChem 2010, 11, 197 – 207
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
205

N. S. Scrutton et al.
were analysed by the protein interfaces, surfaces and assemblies
service (PISA) at the European Bioinformatics Institute (http://
www.ebi.ac.uk/msd-srv/prot_int/pistart.html).[32] The atomic coordi-
nates and structure factors (PDB codes 3KRU and 3KRZ for His6-
and NADH4-PETNR structures, respectively) have been deposited in
the Protein Data Bank, Research Collaboratory for Structural Bioin-
formatics, Rutgers University, New Brunswick, NJ (http://
www.rcsb.org/).
rates were determined by using the rapid equilibrium formalism of
Strickland et al[52] (Equation 1):
k3½Sꢂ
ð1Þ
kox
¼
Kd þ ½Sꢂ
for the kinetic scheme (Equation 2).
Multi-angle light scattering of TOYE-His6: TOYE was purified from
its aggregates on a Supredex-200 24/30 gel filtration column (GE
Healthcare) equilibrated in buffer (50 mm KH2PO4/K2HPO4, pH 7.0)
on a Dionex BioLC HPLC at 0.71 mLminꢀ1. TOYE was passed
through a Wyatt EOS 18-angle laser photometer with the 13th de-
tector replaced with a Wyatt QELS detector for the simultaneous
measurement of hydrodynamic radius. This was coupled to a
Wyatt Optilab rEX refractive index detector and the hydrodynamic
radius, molecular weight moments and concentration of the result-
ing peaks was analysed by using Astra 5.2.
Steady-state kinetics: Steady-state reactions with TOYE-His6 were
performed anaerobically by using either a BioTek Synergy HT mi-
crotiter plate reader with a 0.84 cm path length or a Cary 50 Bio
spectrophotometer (Varian) fitted with a temperature controlled
curvette holder. Standard reactions (0.3–1.0 mL) were performed in
buffer (50 mm KH2PO4/K2HPO4 pH 7.0) containing NADPH (100 mm),
oxidising substrate (0.01–1 mm; stock solutions in 100% ethanol)
and TOYE (0.1–2 mm). The reactions were followed continuously by
monitoring NADPH oxidation at 340 nm (365 nm for trans-cinna-
maldehyde) for 1 min at 258C. The specificity constants (apparent
kcat/KM) were determined from the slope of the initial rates at oxi-
dising substrate concentrations considerably below the estimated
KM (5–20 mm). Steady-state reactions to determine solvent stability
were carried out as above (1.0 mL) in buffer (50 mm KH2PO4/
K2HPO4 pH 7.0) containing NADPH (100 mm), 2-cyclohexenone
(1 mm) and TOYE/PETNR (200/40 nm). The enzyme was preincubat-
ed in ethanol solutions (0–60% ethanol in 50 mm KH2PO4/K2HPO4
pH 7.0) at 258C for 5 min prior to the assays. In all steady-state re-
actions, the final concentration of ethanol was kept constant (5%)
except in the case of activity in the presence of increasing concen-
trations of ethanol (5–80%).
Sedimentation velocity and bead modelling: All experiments
were performed in a Beckman XL-A ultracentrifuge with an An60Ti
four-hole rotor with the standard two-sector epon-filled centrepie-
ces with quartz glass windows. All samples were purified by using
gel filtration immediately prior to loading. Sedimentation of TOYE
was performed in buffer (50 mm KH2PO4/K2HPO4, pH 7.0), with a
rotor speed of 129024 g at 208C. The sedimenting boundary was
monitored every 90 s at 230 nm until full sedimentation was
reached. Data was interpreted with the model-based distribution
of Lamm equation solutions c(s) using the software Sedfit[49] and
the data corrected for standard conditions of water at 208C by
using a n˜ of 0.714 calculated from amino acid composition within
the program Sednterp.[50] Frictional ratios for the monomer and
dimer were calculated directly from the light scattering derived
mass and the sedimentation coefficient. Bead models of the qua-
ternary structures for TOYE were generated by using the PDB file
and the SOMO solution bead modelling program.[35] The motifs
were arranged in several orientations using Pymol.[31] Multiple
bead models of the TOYE domains were created in various confor-
mations. The hydrodynamic parameters generated for the models
were compared to the experimental results until a best-fit was ach-
ieved.
Biotransformations: Standard reactions (1.0 mL) were performed
in buffer (50 mm KH2PO4/K2HPO4 pH 7.0) containing alkene (5 mm;
added as a solution in DMF, 2% final concentration), NAD(P)H
(6 mm) and TOYE (10 mm) at 30 or 508C for 24 h. Additional reac-
tions were performed as above for 2 h at 308C with NADH with
the inclusion of the following organic solvents: dimethyl forma-
mide, tetrahydrofuran, ethanol, methanol, propanol, acetone and
acetonitrile (2–30% v/v). The products were extracted and ana-
lysed for percent yield, percent conversion and enantiomeric
excess according to methods previously described.[42]
Transmission electron microscopy: TOYE-His6 samples were ab-
sorbed onto glow discharged carbon-coated copper TEM grids and
negatively stained as previously described.[51] Data were recorded
digitally by using a Joel 1220 coupled to a Gatan Orius CCD and
subsequent reference free projection averaging of observed single
particles was performed using EMAN (http://blake.bcm.tmc.edu/
eman/) and strategies previously described.[51]
Circular dichroism: Circular dichroism (CD) measurements of
TOYE-His6 were performed by using a Chirascan CD spectrometer
(Applied Photophysics, UK) with a thermoelectrically controlled cell
holder and a 2 mm path length cuvette. Secondary structure un-
folding of TOYE (5–10 mm in 50 mm KH2PO4/K2HPO4 pH 7.0) was
monitored by absorbance change at 222 nm over time (1 h) be-
tween 70–808C and analysed by the Pro-Data Chirascan software.
Data is corrected for background solvent effects and expressed as
percent signal strength.
Fast reaction kinetics: Reductive half-reaction kinetic experiments
with both TOYE and TOYE-His6 were performed anaerobically by
using an Applied Photophysics SX.18MV-R stopped-flow spectro-
photometer. The transients were analysed by using nonlinear least
squares regression analysis on an Acorn Risc PC microcomputer
using Spectrakinetics software (Applied Photophysics). Reactions
(1.0 mL) were performed in buffer (50 mm KH2PO4/K2HPO4 pH 7.0)
containing NADH or NADPH (0.05–5 mm) and TOYE (20 mm). Ab-
sorption change of flavin reduction was monitored continuously at
464 nm at 258C. The concentrations of substrates were always at
least tenfold greater than the enzyme concentration; this ensured
pseudo-first-order conditions. Each data point is the average of at
least three transients. The transients were fitted to a double expo-
nential plot (triple exponential for NADPH) from which observed
Differential scanning calorimetry: TOYE (0.5 mgmlꢀ1) was dena-
tured under controlled conditions to monitor the melting tempera-
ture (Tm) of the protein and its domains and the enthalpy of the
unfolding event (DHV). This was performed using a VP-DSC instru-
ment (MicroCal) with a scan rate of 608Chꢀ1 from 208C to 1008C.
Protein samples were purified by using gel filtration chromatogra-
phy and referenced using the column buffer. Data was analyzed
using Microcal ORIGIN software (Microcal Inc., Amherst, MA).
206
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 197 – 207

Characterisation of a Thermophilic Old Yellow Enzyme
[17] M. A. Swiderska, J. D. Stewart, J. Mol. Catal. B 2006, 42, 52–54.
[18] M. A. Swiderska, J. D. Stewart, Org. Lett. 2006, 8, 6131–6133.
[19] O. Warburg, W. Christian, Naturwissenschaften 1932, 20, 688.
[20] P. R. Binks, C. E. French, S. Nicklin, N. C. Bruce, Appl. Environ. Microbiol.
1996, 62, 1214–1219.
[21] C. E. French, N. C. Bruce, Biochem. J. 1995, 312, 671–678.
[22] B. A. Vick, D. C. Zimmerman, Plant Physiol. 1986, 80, 202–205.
[23] T. B. Fitzpatrick, N. Amrhein, P. Macheroux, J. Biol. Chem. 2003, 278,
19891–19897.
Fluorescence spectroscopy: Temperature-induced unfolding of
the protein was monitored by using either tryptophan or flavin in-
trinsic fluorescence on a Jasco FP750 spectrofluorimeter fitted with
an external heating/cooling water bath. Tryptophan fluorescence
was studied with an excitation of 295 nm and scanning between
310 and 400 nm. Flavin fluorescence was studied by using an exci-
tation of 430 nm and scanning emissions between 460 and
600 nm. Temperature-induced unfolding was achieved by a 20 min
incubation of 5 mm TOYE within the instrument at 10 degree inter-
vals between 208C and 808C with each protein sample scanned
once only.
[24] P. A. Karplus, K. M. Fox, V. Massey, FASEB J. 1995, 9, 1518–1526.
[25] T. Barna, H. L. Messiha, C. Petosa, N. C. Bruce, N. S. Scrutton, P. C. E.
Moody, J. Biol. Chem. 2002, 277, 30976–30983.
[26] K. Kitzing, T. B. Fitzpatrick, C. Wilken, J. Sawa, G. P. Bourenkov, P. Macher-
oux, T. Clausen, J. Biol. Chem. 2005, 280, 27904–27913.
[27] R. U. Onyenwoke, V. V. Kevbrin, A. M. Lysenko, J. Wiegel, Int. J. Syst. Evol.
Microbiol. 2007, 57, 2191–2193.
Acknowledgements
[28] T. N. Nazina, T. P. Tourova, A. B. Poltaraus, E. V. Novikova, A. A. Grigoryan,
A. E. Ivanova, A. M. Lysenko, V. V. Petrunyaka, G. A. Osipov, S. S. Belyaev,
M. V. Ivanov, Int. J. Syst. Evol. Microbiol. 2001, 51, 433–446.
[29] F. A. Rainey, E. Stackebrandt, Int. J. Syst. Bacteriol. 1993, 43, 857–859.
[30] P. R. Binks, C. E. French, S. Nicklin, N. C. Bruce, Appl. Environ. Microbiol.
1996, 62, 1214–1219.
[31] W. L. DeLano, DeLano Scientific, Palo Alto, CA, USA, 2002.
[32] E. Krissinel, K. Henrick, J. Mol. Biol. 2007, 372, 774–797.
[33] T. M. Barna, H. Khan, N. C. Bruce, I. Barsukov, N. S. Scrutton, P. C. E.
Moody, J. Mol. Biol. 2001, 310, 433–447.
We thank Richard Collins for the electron microscopy data and
Emma Keevill for the mass spectrometry protein identification
analysis. This work was funded by UK Biotechnology and Biologi-
cal Sciences Research Council (BBSRC). N.S.S. is a BBSRC Professo-
rial Research Fellow.
Keywords: asymmetric
bioreduction
·
biocatalysis
·
thermostable enzymes
·
unsaturated alkenes
·
UV/Vis
[34] C. R. Pudney, S. Hay, J. Y. Pang, C. Costello, D. Leys, M. J. Sutcliffe, N. S.
Scrutton, J. Am. Chem. Soc. 2007, 129, 13949–13956.
[35] N. Rai, M. Nollmann, B. Spotorno, G. Tassara, O. Byron, M. Rocco, Struc-
ture 2005, 13, 723–734.
[36] R. M. Daniel in Protein Structure,Folding and Design. (Ed.: D. L. Oxender),
Alan R. Liss, New York, 1986, pp. 291–296.
[37] A. Guagliardi, G. Manco, M. Rossi, S. Bartolucci, Eur. J. Biochem. 1989,
183, 25–30.
[38] H. Huber, K. O. Stetter, J. Biotechnol. 1998, 64, 39–52.
[39] R. K. Owusu, D. A. Cowan, Enzym. Microb. Technol. 1989, 11, 568–574.
[40] R. Stuermer, B. Hauer, M. Hall, K. Faber, Curr. Opin. Chem. Biol. 2007, 11,
203–213.
[41] T. B. Fitzpatrick, S. Auweter, K. Kitzing, T. Clausen, N. Amrhein, P. Macher-
oux, Protein Expression Purif. 2004, 36, 280–291.
[42] A. Fryszkowska, H. S. Toogood, M. Sakuma, J. M. Gardiner, G. M. Ste-
phens, N. S. Scrutton, Adv. Synth. Catal. 2009; DOI: 10.1002/
adsc.200900574.
[43] R. C. Stewart, V. Massey, J. Biol. Chem. 1985, 260, 13639–13647.
[44] A. Fryszkowska, K. Fisher, J. M. Gardiner, G. M. Stephens, J. Org. Chem.
2008, 73, 4295–4298.
[45] A. G. W. Leslie in Joint CCP4 & ESRF-EACBM Newsletter on Protein Crystal-
lography, Serc Laboratory, Daresbury, Warrington, U.K. 1992.
[46] Collaborative Computational Project, Number 4, Acta Crystallogr. Sect.
D: Biol. Crystallogr. 1994, 50, 760–763.
[47] A. Perrakis, R. Morris, V. S. Lamzin, Nat. Struct. Biol. 1999, 6, 458–463.
[48] P. Emsley, K. Cowtan, Acta Crystallogr. Sect. D 2004, 60, 2126–2132.
[49] P. Schuck, Biophys. J. 2000, 78, 1606–1619.
[50] D. Hayes, J. J. Philo, Sednterp, University of New Hampshire, New Hamp-
shire, 1992.
[51] R. F. Collins, M. Saleem, J. P. Derrick, J. Bacteriol. 2007, 189, 6389–6396.
[52] S. Strickland, G. Palmer, V. Massey, J. Biol. Chem. 1975, 250, 4048–4052.
spectroscopy · X-ray crystallography
[1] S. Panke, M. Held, M. Wubbolts, Curr. Opin. Biotechnol. 2004, 15, 272–
279.
[2] J. A. Littlechild, J. Guy, S. Connelly, L. Mallett, S. Waddell, C. A. Rye, K.
Line, M. Isupov, Biochem. Soc. Trans. 2007, 35, 1558–1563.
[3] S. Luetz, L. Giver, J. Lalonde, Biotechnol. Bioeng. 2008, 101, 647–653.
[4] G. R. Castro, T. Knubovets, Crit. Rev. Biotechnol. 2003, 23, 195–231.
[5] C. Breithaupt, EMBO Rep. 2001, 2, 968–971.
[6] M. Champdorꢂ, M. Staiano, M. Rossi, S. D’Auria, J. R. Soc. Interf. 2007, 4,
183–191.
[7] L. D. Unsworth, J. van der Oost, S. Koutsopoulos, FEBS J. 2007, 274,
4044–4056.
[8] 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.
[9] M. Hall, C. Stueckler, B. Hauer, R. Stuermer, W. Kroutil, P. Macheroux, K.
Facer, in Biotrans-8th International Symposium on Biocatalysis and Bio-
transformations, Oviedo, Spain, 2007.
[10] M. Hall, C. Stueckler, W. Kroutil, P. Macheroux, K. Faber, Angew. Chem.
2007, 119, 4008–4011; Angew. Chem. Int. Ed. 2007, 46, 3934–3937.
[11] H. S. Toogood, A. Fryszkowska, V. Hare, K. Fisher, A. Roujeinikova, D.
Leys, J. M. Gardiner, G. M. Stephens, N. S. Scrutton, Adv. Synth. Catal.
2008, 350, 2789–2803.
[12] D. J. Bougioukou, J. D. Stewart, J. Am. Chem. Soc. 2008, 130, 7655–
7658.
[13] J. F. Chaparro-Riggers, T. A. Rogers, E. Vazquez-Figueroa, K. M. Polizzi,
A. S. Bommarius, Adv. Synth. Catal. 2007, 349, 1521–1531.
[14] M. Hall, C. Stueckler, B. Hauer, R. Stuermer, T. Friedrich, M. Breuer, W.
Kroutil, K. Faber, Eur. J. Org. Chem. 2008, 1511–1516.
[15] S. K. Padhi, D. J. Bougioukou, J. D. Stewart, J. Am. Chem. Soc. 2009, 131,
3271–3280.
Received: September 11, 2009
[16] C. Stueckler, M. Hall, H. Ehammer, E. Pointner, W. Kroutil, P. Macheroux,
K. Faber, Org. Lett. 2007, 9, 5409–5411.
Published online on November 26, 2009
ChemBioChem 2010, 11, 197 – 207
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
207