Biocatalytic Properties and Structural Analysis of Eugenol Oxidase from Rhodococcus jostii RHA1: A Versatile Oxidative Biocatalyst

Article abstract of DOI:10.1002/cbic.201600148

Eugenol oxidase (EUGO) from Rhodococcus jostii RHA1 had previously been shown to convert only a limited set of phenolic compounds. In this study, we have explored the biocatalytic potential of this flavoprotein oxidase, resulting in a broadened substrate scope and a deeper insight into its structural properties. In addition to the oxidation of vanillyl alcohol and the hydroxylation of eugenol, EUGO can efficiently catalyze the dehydrogenation of various phenolic ketones and the selective oxidation of a racemic secondary alcohol—4-(1-hydroxyethyl)-2-methoxyphenol. EUGO was also found to perform the kinetic resolution of a racemic secondary alcohol. Crystal structures of the enzyme in complexes with isoeugenol, coniferyl alcohol, vanillin, and benzoate have been determined. The catalytic center is a remarkable solvent-inaccessible cavity on the si side of the flavin cofactor. Structural comparison with vanillyl alcohol oxidase from Penicillium simplicissimum highlights a few localized changes that correlate with the selectivity of EUGO for phenolic substrates bearing relatively small p-substituents while tolerating o-methoxy substituents.

Full text of DOI:10.1002/cbic.201600148

DOI: 10.1002/cbic.201600148
Full Papers
Biocatalytic Properties and Structural Analysis of Eugenol
Oxidase from Rhodococcus jostii RHA1: A Versatile
Oxidative Biocatalyst
Quoc-Thai Nguyen+,[a, c] Gonzalo de Gonzalo+,[b] Claudia Binda,[c] Ana Rioz-Martínez,[d]
Andrea Mattevi,*[c] and Marco W. Fraaije*[a]
Eugenol oxidase (EUGO) from Rhodococcus jostii RHA1 had pre-
viously been shown to convert only a limited set of phenolic
compounds. In this study, we have explored the biocatalytic
potential of this flavoprotein oxidase, resulting in a broadened
substrate scope and a deeper insight into its structural proper-
ties. In addition to the oxidation of vanillyl alcohol and the
hydroxylation of eugenol, EUGO can efficiently catalyze the de-
hydrogenation of various phenolic ketones and the selective
oxidation of a racemic secondary alcohol—4-(1-hydroxyethyl)-
2-methoxyphenol. EUGO was also found to perform the kinetic
resolution of a racemic secondary alcohol. Crystal structures of
the enzyme in complexes with isoeugenol, coniferyl alcohol,
vanillin, and benzoate have been determined. The catalytic
center is a remarkable solvent-inaccessible cavity on the si side
of the flavin cofactor. Structural comparison with vanillyl alco-
hol oxidase from Penicillium simplicissimum highlights a few
localized changes that correlate with the selectivity of EUGO
for phenolic substrates bearing relatively small p-substituents
while tolerating o-methoxy substituents.
Introduction
Biocatalysts have become a valuable tool for performing selec-
tive oxidation reactions that usually require harsh conditions
and reagents when performed by conventional chemistry ap-
proaches.[1,2] Oxidases make up an emerging biocatalyst class
for carrying out biotransformations that involve two- or four-
electron oxidations of substrates with use of molecular oxygen
as a mild oxidant.[3] Among them, the so-called flavoprotein
oxidases can be valuable for synthetic applications, because
they can participate in a broad variety of oxidative procedures
with high chemo-, regio-, and/or enantioselectivity.[4–6] Catalysis
for these enzymes occurs in two half-reactions: 1) flavin cofac-
tor reduction by oxidation of the substrate, and 2) reoxidation
of the flavin by molecular oxygen.[7] A wide set of flavoprotein
oxidases have been extensively employed in biotransforma-
tions, with glucose oxidase,[8] d-amino acid oxidase,[9] and cho-
lesterol oxidase[10] as well-known examples.
The vanillyl alcohol oxidase (VAO) family is a group of flavo-
protein oxidases that share a common FAD-binding domain
and present similar structural features.[11–13] In addition to alco-
hol oxidations, leading in most cases to aldehydes and ke-
tones, the enzymes of this family can also catalyze amine oxi-
dations, thiol oxidations, hydroxylations, and even CÀC bond-
formation reactions.[5] A representative family member is VAO
from Penicillium simplicissimum, an enzyme that acts on 4-hy-
droxybenzylic compounds.[14,15] Kinetic studies have shown
that the action of VAO involves an initial hydride transfer from
the substrate to the FAD; this is facilitated by residues of the
protein and generates a p-quinone methide intermediate. Af-
terwards, the reduced flavin is reoxidized by molecular oxygen,
yielding hydrogen peroxide, whereas the p-quinone methide
undergoes hydration in the VAO active site, yielding the corre-
sponding alcohol or aldehyde.[16,17] However, for some sub-
strates the hydration of the p-quinone methide is inefficient,
resulting in the formation of the corresponding p-alkenylphe-
nol: that is, formally, a dehydrogenation process.[18]
[a] Q.-T. Nguyen,+ Prof. M. W. Fraaije
Molecular Enzymology, Groningen Biomolecular Sciences and
Biotechnology Institute, University of Groningen
Nijenborgh 4, 9747 AG Groningen (NL)
E-mail: m.w.fraaije@rug.nl
[b] Dr. G. de Gonzalo+
Departmento de Química Orgµnica, Universidad de Sevilla
c/Profesor García Gonzµlez 1, 41012 Sevilla (Spain)
[c] Q.-T. Nguyen,+ Dr. C. Binda, Prof. A. Mattevi
Department of Biology and Biotechnology, University of Pavia
Via Ferrata 1, 27100 Pavia (Italy)
E-mail: andrea.mattevi@unipv.it
[d] Dr. A. Rioz-Martínez
Stratingh Institute for Chemistry, University of Groningen
Nijenborgh 4, 9747 AG Groningen (NL)
[+] These authors contributed equally to this work.
The genome of Rhodococcus jostii RHA1 contains a remarka-
bly large number of genes coding for oxidative enzymes.[19]
Among them, one gene codes for a eugenol oxidase (EUGO)
that displays significant sequence identity with VAO (ꢀ40%)
and a similar but not fully overlapping pattern of substrate
preferences.[20] Unlike VAO, which expresses poorly in Escheri-
Supporting information for this article can be found under http://
dx.doi.org/10.1002/cbic.201600148.
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA.
This is an open access article under the terms of the Creative Commons At-
tribution License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited.
ChemBioChem 2016, 17, 1359 – 1366
1359 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
chia coli, recombinant EUGO can be produced in large quanti-
ties (a rather impressive 160 mg of pure protein from 1 L of
E. coli culture), which makes it an attractive target for large-
scale oxidative transformations.
plete conversion after 4 h at room temperature. If the reaction
was performed in the presence of 10% DMSO (v/v) as cosol-
vent, an increase in the enzymatic activity was observed, with
only 2 h being required to reach complete conversion. Vanillyl
alcohol (3) was very effectively converted by EUGO. After 4 h,
vanillin (4) was recovered with complete conversion when the
oxidation was performed at pH 8.0 either in aqueous medium
or in buffer containing 10% (v/v) DMSO. To explore the poten-
tial of EUGO we also tested other substrates.
In this report, the EUGO-catalyzed oxidation of eugenol and
vanillyl alcohol, the dehydrogenation of various a,b-unsaturat-
ed ketones, and the kinetic resolution of a racemic secondary
alcohol were examined. Optimal conditions for these biocata-
lytic conversions were determined, revealing that the enzyme
is quite tolerant towards organic solvents. Valuable insights
into the substrate selectivity of EUGO were obtained by resolv-
ing enzyme crystal structures in complexes with isoeugenol,
coniferyl alcohol, vanillin, and benzoate at 1.7–2.6 Š resolution.
Inspection of the active-site cavity structure reveals a conserved
substrate-binding mode with respect to that of VAO but also
highlights specific features that fine-tune substrate specificity.
An interesting result was observed when zingerone (5), a
ginger-derived natural product,[21] was incubated with EUGO in
Tris·HCl buffer (pH 8.0). After 4 h, 58% conversion was ob-
served, with dehydrozingerone (6) being obtained as the
single product. When the reaction was carried out for 12 h,
complete conversion was achieved. Again, the presence of
10% (v/v) DMSO in the reaction medium increased the enzy-
matic activity, with the reaction being complete after 8 h. This
shows that EUGO is also able to catalyze dehydrogenation
processes. Interestingly, dehydrozingerone is also present in
ginger and exhibits interesting medicinal properties.[22]
Analogously, raspberry ketone (7) was incubated with EUGO
at room temperature. This led to the formation of 4-phenyl-
but-3-en-2-one (8) as the sole dehydrogenation product. The
enzyme presented significantly lower activity with this com-
pound than with zingerone, with only 48% conversion being
observed after 24 h at pH 8.0.
Results and Discussion
Substrate specificity of eugenol oxidase
Previous experiments had shown that EUGO is active on phe-
nolic substrates, with the identification of eugenol, vanillyl al-
cohol, vanillylamine, and indan-5-ol as good substrates (kcat
=
0.26–12 sÀ1), whereas 4-ethylguaiacol and 4-(methoxymethyl)-
phenol were poorly converted (kcat =0.004–0.026 sÀ1).[20] Al-
though it was shown that EUGO displays oxidase activity to-
wards these aromatic compounds, no product analysis had
been carried out for any of the above substrates. Therefore, as
a first step in our project, we performed conversions and sub-
sequent product analyses to establish which phenolic products
can be prepared with the aid of EUGO (Scheme 1). Incubation
of eugenol (1) with the enzyme in Tris·HCl buffer (pH 8.0) con-
taining 1% (v/v) DMSO afforded coniferyl alcohol (2) with com-
Previous experiments have shown that indan-5-ol (9) is a
EUGO substrate,[20] which is interesting in view of the cyclic
substituent of this compound. After 24 h incubation of this
substrate at room temperature and pH 8.0, 34% conversion
was achieved. The resulting mixture contained two different
products in a 7:3 proportion. Analysis by MS/GC and NMR
revealed that the major product was 1H-inden-6-ol (10), the
dehydrogenation product, whereas 5-hydroxyindan-1-one (11)
was obtained as the minor product. Again, when the oxidation
was performed in the presence of 10% (v/v) DMSO, the enzy-
matic activity was doubled, leading to 63% conversion after
24 h.
Other potential substrates, such as 4-(1,3-dithiolan-2-yl)phe-
nol, capsaicin, (4-hydroxyphenyl)acetonitrile, 3-(4-hydroxyphe-
nyl)propanoic acid, 2-methoxy-4-methylphenol, and 4-(hept-1-
enyl)-2-methoxyphenol, did not show any significant oxidase
activity nor yield any product after incubation with EUGO for
24 h.
These data were complemented by measurements of the
initial oxidase rates for some of the EUGO substrates (see
Table S1 in the Supporting Information). Next to vanillyl alco-
hol, eugenol (1) and its analogue 4-(1-hydroxyethyl)-2-meth-
oxyphenol (12) were found to be the best substrates for this
biocatalyst (2.1 and 4.3 sÀ1, respectively), whereas raspberry
ketone (7) was by far the most slowly reacting substrate
(0.02 sÀ1) of all tested. In essence, EUGO prefers phenolic sub-
strates with ortho-methoxy substituents and a relatively small
group at the para position.
Scheme 1. Reactions catalyzed by EUGO from R. jostii RHA1. a) EUGO,
Tris·HCl (50 mm, pH 7.5), DMSO, RT, 200 rpm.
ChemBioChem 2016, 17, 1359 – 1366
www.chembiochem.org
1360 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Enzymatic dehydrogenation of raspberry ketone (7): Effect
of organic cosolvent, temperature, and pH on EUGO activity
The data above show that raspberry ketone (7) was only
slowly converted by EUGO. For this reason, we set out to opti-
mize the reaction conditions using 7 as a substrate. Because
we had observed that EUGO is quite tolerant towards DMSO
[10% (v/v) of DMSO resulted in higher levels of conversion for
all tested substrates], several solvents were tested in the enzy-
matic dehydrogenation of this ketone. The reaction carried out
in the presence of 10% (v/v) DMSO afforded the final product
8 with a slightly higher level of conversion than achieved in
the aqueous medium (60% conversion after 24 h; Table 1,
entry 2). The use of other cosolvents, with different physico-
chemical properties, led to much lower levels of conversion
(entries 3–7).
*
Figure 1. Effect of DMSO concentration on the degree of conversion ( ) in
EUGO-catalyzed oxidation of raspberry ketone (7, 2.0 mm) after 24 h.
Dehydrogenation of raspberry ketone was also performed in
Tris·HCl buffer with 10% (v/v) DMSO at different pH values
(from pH 7.0 to 10.0; entries 13–16). The enzyme showed
higher activity at higher pH values (from 8.0 to 10.0), with
levels of conversion around 50–60% after 24 h. Lower pH
values led to a rapid decrease in conversion, with only 23%
conversion at pH 7.0.
Table 1. Effect of organic cosolvents, temperature, and pH on EUGO ac-
tivity in the catalysis of the dehydrogenation of raspberry ketone (7)
under different reaction conditions after 24 h.
Cosolvent (v/v)
pH
8.0
T [8C]
c [%][a]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
none
25
25
25
25
25
25
25
17
30
37
45
60
25
25
25
25
48
60
10
ꢁ3
17
ꢁ3
7
15
43
38
33
ꢁ3
23
38
53
57
Collectively, this set of experiments indicated that EUGO is
highly solvent-tolerant and maximally active at room tempera-
ture, whereas it is rather sensitive to variation of the pH of the
medium.
10% DMSO
10% 1,4-dioxane
10% CH3CN
10% EtOAc
10% CH2Cl2
10% tBuOMe
10% DMSO
10% DMSO
10% DMSO
10% DMSO
10% DMSO
10% DMSO
10% DMSO
10% DMSO
10% DMSO
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
7.0
7.5
9.0
10.0
Eugenol oxidase stability
EUGO is a robust enzyme, being active over a wide range of
pH values and tolerating the use of several cosolvents. We de-
termined its thermostability by determining the apparent melt-
ing temperatures (Tm values) under different conditions by the
ThermoFAD method.[23] At pH values from 5.0 to 8.0 the
enzyme is exceptionally thermostable, exhibiting Tm values
around 658C (Table S2). Moreover, the presence of 10% (v/v)
DMSO or ethyl acetate does not significantly affect the en-
zyme’s thermostability. Only extreme pH conditions or the ad-
dition of some organic cosolvents decrease the Tm below
608C. This high thermostability is in line with the observation
that EUGO can be employed in reaction mixtures including an
organic cosolvent. As would be expected, enzyme-stabilizing
agents such as glycerol and salts such as NaCl have a positive
effect on the thermostability of EUGO.
[a] Conversion determined by GC.
With DMSO selected as the best cosolvent for the enzyme,
the effect of its concentration on the enzymatic activity was
analyzed (Figure 1). The conversion of 7 catalyzed by EUGO is
optimal at 10% (v/v) cosolvent. Higher DMSO concentrations
led to progressive deactivation of the biocatalyst, although it is
noteworthy that a moderate level of conversion (30%) was still
obtained at 50% (v/v) DMSO after 24 h. This shows that EUGO
is fairly solvent-tolerant.
Kinetic resolution of 4-(1-hydroxyethyl)-2-methoxyphenol
[(Æ)-12] catalyzed by EUGO
The effect of temperature on the EUGO-catalyzed bioconver-
sion of 7 in Tris·HCl (pH 8.0)/10% (v/v) DMSO medium was also
analyzed (Table 1, entries 8–12). At 178C, only 15% conversion
was obtained after 24 h. Higher temperatures led to increased
levels of conversion, with 60% of 8 being achieved at 258C. A
slightly lower level of conversion was observed on working at
higher temperatures but EUGO remained active at 458C, with
33% conversion after 24 h. In contrast, no reaction was ob-
served at 608C.
Because vanillyl alcohol [3] is an excellent substrate for EUGO,
its racemic analogue 4-(1-hydroxyethyl)-2-methoxyphenol [(Æ)-
12] was studied in a secondary alcohol kinetic resolution pro-
cedure (Scheme 2). In this process, one enantiomer of the sub-
strate is oxidized to the ketone while the other enantiomer
hence becomes enantiomerically enriched. EUGO is a true ox-
idase, so molecular oxygen consumption leads to the forma-
tion of hydrogen peroxide. For this reason, the reactions were
ChemBioChem 2016, 17, 1359 – 1366
www.chembiochem.org
1361 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
The overall three-dimensional structure of EUGO
The structure of EUGO in complex with the inhibitor isoeuge-
nol was solved at 1.7 Š resolution by molecular replacement
with use of the VAO monomer (PDB ID: 2VAO)[11] as a search
model. The asymmetric unit contains two enzyme subunits
forming a compact dimer that is also observed in solution (Fig-
ure 2A and B). The dimer interface area is quite large, approxi-
mately 3500 Š2, accounting for 16.5% of the monomer’s sur-
face as calculated by PISA.[25] The overall structure of EUGO is
highly similar to that of VAO, as indicated by a 1.1 Š rmsd for
490 Ca atom pairs (45% sequence identity).[11] As predicted,
EUGO also shares with VAO the presence of a characteristic co-
valent bond between His390 (equivalent to His422 of VAO)
and the C8M atom of the flavin ring (Figure 3). Nevertheless,
the oligomeric organization in the two enzymes is partly differ-
ent, because, though both form essentially the same functional
dimer (1.3 Š rmsd for 999 Ca atom pairs), EUGO lacks the
dimer–dimer interacting loop that in VAO mediates the octa-
meric structure of the enzyme.
Scheme 2. EUGO-catalyzed kinetic resolution of racemic 4-(1-hydroxyethyl)-
2-methoxyphenol [(Æ)-12]. a) EUGO, Tris·HCl (50 mm, pH 7.5), cosolvent,
308C, 200 rpm.
performed in the presence of catalase, which catalyzes the
decomposition of hydrogen peroxide, thereby regenerating
molecular oxygen. It was observed that the presence of small
amounts of catalase indeed led to slightly higher levels of con-
version with no effect on the selectivity (data not shown). Rac-
emic (Æ)-12 was found to be selectively oxidized by EUGO into
ketone 13, as shown in Table 2. Initial experiments performed
with a Tris·HCl buffer (50 mm, pH 9.0) showed that the kinetic
resolution occurred rapidly and with very low selectivity
(Table 2, entry 1). Use of lower pH values (see pH 7.5, entry 2)
led to significantly higher enantioselectivity (E=35), resulting
in the enantioenrichment of (S)-12. Conversely, the presence of
10% (v/v) DMSO in the reaction medium decreased the level
of conversion and the selectivity of the reaction (entry 3). For
this reason, the amount of this organic cosolvent was reduced
to 5% (v/v), resulting in a higher level of conversion, as shown
in entry 4.
The effects of alternative organic cosolvents (5%) in the
EUGO-catalyzed kinetic resolution of (Æ)-12 were studied (en-
tries 5–11). The use of ethyl acetate (Table 2, entry 5) led to
similar results in terms of activity and selectivity, whereas the
best result was afforded with iPr2O, in a process with a promis-
ing selectivity (E=45) and a 39% level of conversion after 4 h
(entry 7). Kinetic resolutions in reaction media containing 5%
(v/v) tBuOMe, hexane, or octan-2-ol were much faster, with up
to 50% conversion after 1.5 h, but the selectivity was from
moderate to poor. Oxidations with cosolvents such as toluene
or CH2Cl2 occurred with very low conversion efficiencies (en-
tries 9 and 10).
Table 2. EUGO-catalyzed kinetic resolution of the racemic alcohol 4-(1-hy-
droxyethyl)-2-methoxyphenol [(Æ)-12].[a]
Cosolvent
t [h]
c [%][b]
ee [%][c]
E[d]
1[e]
2
3
4
5
6
7
8
none
none
4
4
4
2
4
1.5
4
4
4
49
33
29
30
31
40
39
48
ꢁ3
6
23
45
17
35
49
55
59
13
–
2.0
35
3.0
14
38
18
45
1.5
–
3.7
9.1
Figure 2. Crystal structure of eugenol oxidase from R. jostii RHA1 in complex
with the inhibitor isoeugenol. A) Ribbon diagram of the EUGO dimer with
the two monomers in ice blue (monomer A) and gold (monomer B), respec-
tively. B) Close-up sliced view of the EUGO surface in the vicinity of the cata-
lytic pocket. The inhibitor isoeugenol is buried in a pocket in close contact
with a deep invagination originating at the top of the dimer and being
formed by residues at the dimer interface, which might represent the path-
way for ligand access to the active site. The orientation and atom coloring is
the same as that of the structure shown in (A). The FAD carbon atoms are
shown in yellow, the ligand carbon atoms in lawn green, oxygen atoms in
red, nitrogen atoms in blue, and phosphorus atoms in magenta.
10% DMSO
5% DMSO
5% EtOAc
5% tBuOMe
5% iPr2O
5% hexane
5% CH2Cl2
5% toluene
5% octan-2-ol
9
10
11
4
1.5
12
65
50
[a] For reaction conditions, see the Experimental Section. [b] Determined
by GC. [c] Determined by HPLC. [d] Enantiomeric ratio, E=ln[(1Àc)/
(1Àees)]/ln[(1Àc)/(1+ees)].[24] [e] Reaction performed at pH 9.0.
ChemBioChem 2016, 17, 1359 – 1366
www.chembiochem.org
1362 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 3. The EUGO active site in complex with A) isoeugenol, B) coniferyl alcohol, C) vanillin, and D) benzoate. Unbiased 2FoÀFc omit electron density maps
calculated at 1.7–2.6 Š resolution and contoured at 1.0s are drawn in chicken-wire style. Hydrogen bond contacts are shown as dashed lines. Residues in
direct contact with the bound ligands are labeled. Protein, ligand, and FAD carbon atoms are colored in gray, lawn green, and yellow, respectively, oxygen
atoms are in red, nitrogen atoms are in blue, and phosphorus atoms are in magenta. The orientation is the same as that of monomer A in Figure 2.
cluster forms an anion-binding site is supported by the struc-
ture obtained from native crystals (i.e., not soaked in a ligand
solution), which exhibited a clear electron density peak close
to the flavin rings of both monomers present in the asymmet-
ric unit. This residual density was tentatively assigned to a ben-
zoate ion bound with its carboxylate group directly interacting
with the Tyr-Tyr-Arg residues (Figure 3D). Indeed, we found
benzoate to be a weak EUGO inhibitor, probably captured by
the enzyme in the cell and retained during purification. Collec-
tively, these features are fully consistent with the substrate
being bound in the phenolic form and oxidized through a re-
action mechanism involving a hydride transfer from the Ca to
the flavin N5, coupled to the stabilization of a p-quinone me-
thide intermediate as demonstrated for VAO.[16]
The active site of EUGO
The active site of EUGO consists of a round-shaped cavity lo-
cated in front of the flavin cofactor (Figure 2B). As in VAO, this
substrate-binding site is completely solvent-inaccessible. How-
ever, inspection of the dimeric structure of EUGO reveals a pu-
tative admission path for the ligands. The active-site cavity of
each monomer is in close contact with a funnel-shaped large
chamber that runs along the dimer interface to the top of the
structure and might represent a passageway for the diffusion
of the substrate into the binding site (in the orientation of Fig-
ure 2B). The structural bases of substrate and inhibitor binding
by the cavity were explored by solving the crystal structures of
EUGO in complexes with isoeugenol (inhibitor), coniferyl alco-
hol (2, reaction product and inhibitor), and vanillin (4, reaction
product); these were obtained by crystal soaking, resulting in
highly defined electron density maps (Figure 3A–C).
However, comparative analysis of the three-dimensional
structures also revealed striking differences between EUGO
and VAO. Above all, inhibitor/product binding differs in a 1808
flipped orientation. This can be visualized by superimposing
the isoeugenol complexes of the two enzymes. As shown in
Figure 4A, the ligand methoxy groups are in opposed orienta-
tions and occupy different niches in the active sites of the two
enzymes. In EUGO, the methoxy moiety is in close contact
with Gly392, which replaces the Phe424 residue of VAO. This
bulkier aromatic side chain of VAO likely creates steric hin-
drance, impeding the binding orientation found in EUGO. Con-
All these ligands bind with their aromatic moiety stacking
against the pyrimidine ring of the flavin and the Ca atom of
the aliphatic substituent lying right above the flavin N5 atom
at a distance of 3.2 Š. In this orientation, the ligand 4-hydroxy
group is H-bonded to Tyr91 and Tyr471 and is also in close
contact with Arg472 (3.5 Š). These three residues are involved
in the stabilization of the phenolate form of the substrate and
are strictly conserved in VAO. The fact that such a Tyr-Tyr-Arg
ChemBioChem 2016, 17, 1359 – 1366
www.chembiochem.org
1363 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
The other relevant element in EUGO that partially differs
from the situation in VAO involves the section of the active-
site cavity that is involved in the binding of the substrate alkyl
chain. Here, despite a set of conserved residues (Asp151,
Tyr168, Arg278, and Glu378, corresponding to Asp170, Tyr187,
Arg312, and Glu410 in VAO), there are a few major differing
amino acids between the two enzymes: Thr457, Trp413, and
Leu316 in VAO are replaced by Gln425, Leu381, and Met282,
respectively, in EUGO (Figure 4B). In particular, these replace-
ments collectively reshape this part of the substrate site, which
becomes narrower in EUGO than in VAO. As a result, it is more
difficult for EUGO to accept long aliphatic chains and hydro-
phobic substituents than it is for VAO (Figure 4B). Indeed,
whereas VAO is reactive towards alkylphenols and can accom-
modate aliphatic substituents containing up to seven carbon
atoms,[11,26] we found no activity of EUGO with such com-
pounds.
Conclusion
The synthetic potential of eugenol oxidase (EUGO) in oxidative
procedures has been demonstrated. The data show that EUGO
is a robust biocatalyst, tolerating a wide range of organic sol-
vents and high temperatures. Interestingly, the addition of
DMSO as cosolvent increased the enzymatic activity in most of
the catalytic processes. EUGO could also be employed in the
kinetic resolution of a racemic secondary alcohol with moder-
ate to good selectivity. Structural analysis of the EUGO catalytic
center showed that the enzyme shares a generally conserved
substrate binding mode with VAO but with key different resi-
dues finely controlling the substrate specificity. With an effec-
tive expression system, detailed structural information, and
generic oxidase screening methods to hand, EUGO appears a
perfect candidate for engineering tailored oxidase variants.
Figure 4. Comparison between EUGO and VAO active sites. A) The substrate-
binding site of EUGO in complex with isoeugenol (lawn green) is super-
posed onto that of VAO in complex with the same ligand (light green; PDB
ID: 2VAO). Unconserved residues in VAO are drawn with carbon atoms in
white. The orientation and atom coloring is the same as that of the structure
shown in Figure 2. For the sake of clarity, Tyr168, Arg278, and Glu378 were
omitted in this figure. The g carbon atom of isoeugenol in the VAO structure
was disordered and was not included in the refined model.[11] B) The super-
position of the EUGO·isoeugenol structure onto VAO in complex with 4-
(hept-1-enyl)phenol (light green; PDB ID: 1AHZ) shows that EUGO features
a slightly narrower cavity due to the presence of Gln425 (Thr457 in VAO),
which would clash with ligands with longer alkyl chains. In this figure the
structure is rotated by about 458 with respect to (A). For the sake of clarity,
the side chains of I427/T459 as well as the labels of L381/W413 and H390/
H422 were omitted.
Experimental Section
General materials and methods: Recombinant eugenol oxidase
from R. jostii RHA1 was overexpressed and purified according to
previously described methods.[20] All other chemicals and solvents
were obtained from Acros Organics, Alfa Aesar, ABCR, Sigma–Al-
drich, TCI Europe, and Roche Diagnostics. Flash chromatography
1
was performed with Merck silica gel 60 (230–400 mesh). H NMR,
13C NMR, and DEPT spectra were recorded with TMS (tetramethyl-
silane) as the internal standard, with a Bruker AC-300-DPX (1H:
300.13 MHz and 13C: 75.4 MHz) spectrometer. Mass spectra were re-
corded by ESI+ with a HP1100 chromatograph mass detector or by
EI with a Finnigan MAT 95 spectrometer. GC/MS analysis for com-
pounds 1–11 was performed with a GC Hewlett Packard 6890
Series II and a Hewlett Packard 5973 chromatograph MS (Agilent
Technologies) equipped with a HP-1 cross-linked methyl siloxane
column (30 m”0.32 mm”0.25 mm, 1.0 bar N2). The injector tem-
perature was 2258C, and the flame ionization detector (FID) tem-
perature was 2508C. The following temperature program was em-
ployed: 1008C (5 min), 108CminÀ1 to 2508C (2 min). tR 1: 11.0 min.
tR 2: 15.6 min. tR 3: 12.2 min. tR 4: 11.3 min, tR 5: 14.5 min. tR 6:
16.4 min. tR 7: 11.8 min. tR 8: 12.9 min, tR 9: 10.6 min. tR 10:
10.8 min. tR 11: 9.6 min. GC analysis for measuring the conversion
versely, the flipped isoeugenol bound to VAO positions the
methoxy group in direct contact with a valine residue that is
conserved in EUGO (Val166, Figure 4A).
It is gratifying that these findings correlate with the non-
overlapping patterns of substrate preferences and selectivities
displayed by the two enzymes. Indeed, inspired by these struc-
tural data, we tested 4-allyl-2,6-dimethoxyphenol. This dimeth-
oxy compound is not a substrate of VAO.[15] Conversely, we
found that it is converted by EUGO with moderate efficiency
(kcat =0.49 sÀ1, Km =0.8 mm). Thus, the presence of Gly392 (in-
stead of Phe424 in VAO) is a key feature that enables EUGO to
act on molecules bearing two o-methoxy substituents. This is
in contrast with VAO, which does not accept dimethoxy-substi-
tuted molecules as substrates.[15]
of (Æ)-12 was performed with
a Restek RtbDEXse (30 m”
ChemBioChem 2016, 17, 1359 – 1366
www.chembiochem.org
1364 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
0.25 mm”0.25 mm, 1.0 bar N2) column. The following temperature
program was used: 708C (5 min), at 38CminÀ1 to 1608C (2 min),
208CminÀ1 to 1808C (7 min). tR 12: 41.4 min. tR 13: 40.8 min. To
monitor levels of conversion, substrates and products were quanti-
fied by use of calibration curves. HPLC analyses were developed
with a Hewlett Packard 1100 LC liquid chromatograph. The follow-
ing conditions were used for the determination of the enantiomer-
ic excess of alcohol 12: Chiralcel OJ-H column (0.46”25 cm), iso-
cratic eluent: n-hexane/EtOH (95:5), 408C, flow 1 mLminÀ1. tR (S)-
12: 31.4 min. tR (R)-12: 35.1 min.
Grenoble, France (ESRF). The images were integrated by
MOSFLM,[27] whereas data scaling was performed by use of pro-
grams of the CCP4 suite.[28] The detailed data processing statistics
of the collected datasets are shown in Table S3. EUGO’s structure
was solved by molecular replacement with use of MOLREP[29] and
the coordinates of VAO (PDB ID: 2VAO)[11] as the search model
devoid of all ligand and water molecules. COOT[30] and REFMAC5[31]
programs were employed to carry out alternating cycles of model
building and refinement (data shown in the Supporting Informa-
tion). Figures were created by use of CCP4mg.[32]
General procedure for the enzymatic oxidations catalyzed by
isolated EUGO: Unless stated otherwise, starting compounds 1, 3,
5, 7, and 9 (2–10 mm) were dissolved in Tris·HCl buffer (50 mm,
pH 8.0, 1.0 mL) containing, when indicated, organic cosolvent
(10%, v/v) and EUGO (0.33 mm). Reaction mixtures were shaken at
250 rpm and room temperature for the times indicated. Once reac-
tion was complete, the crude mixtures were extracted with EtOAc
(3”500 mL). The organic phases were dried onto Na2SO4 and ana-
lyzed directly by GC/MS.
Acknowledgements
This research was supported the European Union (EU) project
ROBOX (grant agreement no. 635734) under the EU’s Horizon
2020 Programme Research and Innovation actions H2020-LEIT
BIO-2014–1. G.d.G. (Ramón y Cajal Program) thanks MINECO for
personal funding. We thank the European Synchrotron Radiation
Facility (ESRF) and the Swiss Light Source (SLS) for providing
beamtime and assistance. We thank Mohamed Mohamed Habib
for his experimental support. We also thank the BioStruct-X pro-
gram (project no. 10205) for funding synchrotron trips.
Kinetic resolution of racemic 4-(1-hydroxyethyl)-2-methoxyphe-
nol catalyzed by EUGO: Unless stated otherwise, the starting race-
mic alcohol (Æ)-12 (10 mm, Scheme 2) was dissolved in Tris·HCl
buffer (50 mm, pH 7.5) containing, when stated, organic cosolvent
(5–10%, v/v, 1.0 mL), catalase (2 mL, six unitsmLÀ1), and EUGO
(1.0 mm). Reaction mixtures were shaken at 250 rpm and 308C in
a rotatory shaker for the times indicated. Once reaction was com-
plete, the crude mixtures were extracted with EtOAc (2”500 mL).
The organic phases were dried onto Na2SO4 and analyzed directly
by GC and HPLC in order to determine the levels of conversion to
13 and the enantiomeric excesses of the alcohol (S)-12.
Keywords: biocatalysis
structures · kinetic resolution · oxidases
·
dehydrogenation
·
enzyme
[1] S. Schulz, M. Girhard, V. B. Urlacher, ChemCatChem 2012, 4, 1889–1895.
[2] R. D. Schmid, V. B. Urlacher, Modern Biooxidation: Enzymes, Reactions
and Applications, Wiley-VCH, Weinheim, 2007.
[3] N. J. Turner, Chem. Rev. 2011, 111, 4073–4087.
[4] M. J. H. Moonen, M. W. Fraaije, I. M. C. M. Rietjens, C. Laane, W. J. H. van
Berkel, Adv. Synth. Catal. 2002, 344, 1023–1035.
[5] W. P. Dijkman, G. de Gonzalo, A. Mattevi, M. W. Fraaije, Appl. Microbiol.
Biotechnol. 2013, 97, 5177–5188.
[6] R. T. Winter, M. W. Fraaije, Curr. Org. Chem. 2012, 16, 2542–2550.
[7] P. Chaiyen, M. W. Fraaije, A. Mattevi, Trends Biochem. Sci. 2012, 37, 373–
380.
[8] S. B. Bankar, M. V. Bule, R. S. Singhal, L. Ananthanaryan, Biotechnol. Adv.
2009, 27, 489–501.
[9] X. Ma, S. Deng, E. Su, D. Wei, Biochem. Eng. J. 2015, 95, 1–8.
[10] S. Dieth, D. Tritsch, J. Biellmann, Tetrahedron Lett. 1995, 36, 2243–2246.
[11] A. Mattevi, M. W. Fraaije, A. Mozzarelli, L. Olivi, A. Coda, W. J. van Berkel,
Structure 1997, 5, 907–920.
Measurement of initial rates: Conversion rates were determined
by monitoring the consumption of molecular oxygen. Oxygen con-
centrations were monitored by use of a REDFLASH sensor spot in
a 3 mL cuvette in combination with a Firesting O2 detector and
light source (Pyroscience, Aachen, Germany). Measurements were
performed with substrates (2.0 mm) in KPi buffer (50 mm, pH 7.5)
containing DMSO (10%, v/v).
Determination of kinetic parameters for 4-allyl-2,6-dimethoxy-
phenol: The reaction was monitored by following the absorbance
of the product at 270 nm (e=14.1 mmÀ1 cmÀ1 at pH 7.5) in Tris·HCl
(50 mm, pH 7.5) at 258C.
[12] M. W. Fraaije, W. J. van Berkel, J. A. Benen, J. Visser, A. Mattevi, Trends
Biochem. Sci. 1998, 23, 206–207.
[13] N. G. Leferink, D. P. Heuts, M. W. Fraaije, W. J. van Berkel, Arch. Biochem.
Biophys. 2008, 474, 292–301.
[14] E. de Jong, W. J. van Berkel, R. P. van der Zwan, J. A. de Bont, Eur. J. Bio-
chem. 1992, 208, 651–657.
[15] M. W. Fraaije, C. Veeger, W. J. van Berkel, Eur. J. Biochem. 1995, 234,
271–277.
Analysis of EUGO thermostability: EUGO thermostability was as-
sayed by use of the ThermoFAD protocol,[23] with a MiniOpticon
real-time PCR detection system and 48-well RT-PCR plates (Biorad
Laboratories). Samples were prepared by mixing EUGO (50 mm,
2.0 mL) in Tris·HCl (10 mm, pH 7.5) together with the buffers and/or
organic solvents to a final volume of 20 mL and in duplicate for
each condition tested.
[16] M. W. Fraaije, W. J. van Berkel, J. Biol. Chem. 1997, 272, 18111 –18116.
[17] M. W. Fraaije, R. H. van den Heuvel, J. C. Roelofs, W. J. van Berkel, Eur. J.
Biochem. 1998, 253, 712–719.
[18] R. H. H. van den Heuvel, M. W. Fraaije, W. J. H. van Berkel, FEBS Lett.
2000, 481, 109–112.
[19] M. P. McLeod, R. L. Warren, W. W. Hsiao, N. Araki, M. Myhre, C. Fer-
nandes, D. Miyazawa, W. Wong, A. L. Lillquist, D. Wang, M. Dosanjh, H.
Hara, A. Petrescu, R. D. Morin, G. Yang, J. M. Stott, J. E. Schein, H. Shin,
D. Smailus, A. S. Siddiqui, M. A. Marra, S. J. Jones, R. Holt, F. S. Brinkman,
K. Miyauchi, M. Fukuda, J. E. Davies, W. W. Mohn, L. D. Eltis, Proc. Natl.
Acad. Sci. USA 2006, 103, 15582–15587.
Protein crystallization, X-ray data collection, and structure de-
termination: Native EUGO was crystallized by use of the sitting-
drop vapor diffusion technique at 208C by mixing equal volumes
of EUGO (15 mgmLÀ1) in Tris·HCl (10 mm, pH 7.5) and mother
liquor containing PEG6000 (24%, w/v) and Tris·HCl (0.1m, pH 8.0).
All EUGO·ligand complexes were prepared by soaking (0.5–2 h) the
crystals in cryoprotectant solutions consisting of PEG6000 (26%),
Tris·HCl (0.1m, pH 8.0), glycerol (20%), and the compound of inter-
est (5.0 mm), followed by flash-freezing in liquid nitrogen. X-ray dif-
fraction data were collected at the PXI and PXIII beamlines of the
Swiss Light Synchrotron in Villigen, Switzerland (SLS) and at the
ID23–1 beamline of the European Synchrotron Radiation Facility in
[20] J. Jin, H. Mazon, R. H. van den Heuvel, D. B. Janssen, M. W. Fraaije, FEBS
J. 2007, 274, 2311–2321.
[21] Y. Shukla, M. Singh, Food Chem. Toxicol. 2007, 45, 683–690.
ChemBioChem 2016, 17, 1359 – 1366
www.chembiochem.org
1365 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[22] S. Yogosawa, Y. Yamada, S. Yasuda, Q. Sun, K. Takizawa, T. Sakai, J. Nat.
Prod. 2012, 75, 2088–2093.
Read, A. Vagin, K. S. Wilson, Acta Crystallogr. Sect. D Biol. Crystallogr.
2011, 67, 235–242.
[29] A. Vagin, A. Teplyakov, Acta Crystallogr. Sect. D Biol. Crystallogr. 2010, 66,
22–25.
[23] F. Forneris, R. Orru, D. Bonivento, L. R. Chiarelli, A. Mattevi, FEBS J. 2009,
276, 2833–2840.
[30] P. Emsley, K. Cowtan, Acta Crystallogr. Sect. D Biol. Crystallogr. 2004, 60,
2126–2132.
[24] J. L. L. Rakels, A. J. J. Straathof, J. J. Heijnen, Enzyme Microb. Technol.
1993, 15, 1051–1056.
[31] G. N. Murshudov, A. A. Vagin, E. J. Dodson, Acta Crystallogr. Sect. D Biol.
Crystallogr. 1997, 53, 240–255.
[32] S. McNicholas, E. Potterton, K. S. Wilson, M. E. Noble, Acta Crystallogr.
Sect. D Biol. Crystallogr. 2011, 67, 386–394.
[25] E. Krissinel, K. Henrick, J. Mol. Biol. 2007, 372, 774–797.
[26] R. H. H. van den Heuvel, M. W. Fraaije, C. Laane, W. J. H. van Berkel, J.
Bacteriol. 1998, 180, 5646–5651.
[27] T. G. Battye, L. Kontogiannis, O. Johnson, H. R. Powell, A. G. Leslie, Acta
Crystallogr. Sect. D Biol. Crystallogr. 2011, 67, 271–281.
[28] M. D. Winn, C. C. Ballard, K. D. Cowtan, E. J. Dodson, P. Emsley, P. R.
Evans, R. M. Keegan, E. B. Krissinel, A. G. Leslie, A. McCoy, S. J. McNicho-
las, G. N. Murshudov, N. S. Pannu, E. A. Potterton, H. R. Powell, R. J.
Manuscript received: March 9, 2016
Accepted article published: April 28, 2016
Final article published: June 7, 2016
ChemBioChem 2016, 17, 1359 – 1366
www.chembiochem.org
1366 ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim