Biocatalytic Enantioselective Oxidation of Sec-Allylic Alcohols with Flavin-Dependent Oxidases

Article abstract of DOI:10.1002/adsc.201900921

The oxidation of allylic alcohols is challenging to perform in a chemo- as well as stereo-selective fashion at the expense of molecular oxygen using conventional chemical protocols. Here, we report the identification of a library of flavin-dependent oxidases including variants of the berberine bridge enzyme (BBE) analogue from Arabidopsis thaliana (AtBBE15) and the 5-(hydroxymethyl)furfural oxidase (HMFO) and its variants (V465T, V465S, V465T/W466H and V367R/W466F) for the enantioselective oxidation of sec-allylic alcohols. While primary and benzylic alcohols as well as certain sugars are well known to be transformed by flavin-dependent oxidases, sec-allylic alcohols have not been studied yet except in a single report. The model substrates investigated were oxidized enantioselectively in a kinetic resolution with an E-value of up to >200. For instance HMFO V465S/T oxidized the (S)-enantiomer of (E)-oct-3-en-2-ol (1 a) and (E)-4-phenylbut-3-en-2-ol with E>200 giving the remaining (R)-alcohol with ee>99% at 50% conversion. The enantioselectivity could be decreased if required by medium engineering by the addition of cosolvents (e. g. dimethyl sulfoxide).

Full text of DOI:10.1002/adsc.201900921

UPDATES
DOI: 10.1002/adsc.201900921
Biocatalytic Enantioselective Oxidation of Sec-Allylic Alcohols
with Flavin-Dependent Oxidases
Somayyeh Gandomkar,a Etta Jost,a Doris Loidolt,a Alexander Swoboda,a
Mathias Pickl,a Wael Elaily,b, c Bastian Daniel,b, e Marco W. Fraaije,d
Peter Macheroux,b and Wolfgang Kroutila,*
a
Institute of Chemistry, NAWI Graz, BioTechMed Graz, University of Graz, Heinrichstr. 28, 8010 Graz, Austria
Tel: +43-316-380-5350
Fax: +43-316-380-9840
E-mail: wolfgang.kroutil@uni-graz.at
Institute of Biochemistry, Graz University of Technology, Petersgasse 12/II, 8010 Graz, Austria
Chemistry of Natural & Microbial Products Department, National Research Centre, 33 El Buhouth St, 12622 Cairo, Egypt
Molecular Enzymology Group, University of Groningen, Nijenborgh 4, 9747AG, Groningen, The Netherlands
b
c
d
e
Austrian Centre of Industrial Biotechnology, c/o Institute of Molecular Biosciences, University of Graz, Humboldtstraße 50,
8010 Graz, Austria
Manuscript received: July 25, 2019; Revised manuscript received: September 26, 2019;
Version of record online: ■■■, ■■■■
Supporting information for this article is available on the WWW under https://doi.org/10.1002/adsc.201900921
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited.
Keywords: Biocatalysis; Biotransformation; sec-Al-
lylic alcohol; Asymmetric catalysis; Aerobic Oxida-
Abstract: The oxidation of allylic alcohols is
challenging to perform in a chemo- as well as
tion
stereo-selective fashion at the expense of molecular
oxygen using conventional chemical protocols.
Here, we report the identification of a library of
flavin-dependent oxidases including variants of the
berberine bridge enzyme (BBE) analogue from
Introduction
Arabidopsis thaliana (AtBBE15) and the 5-
The oxidation of alcohols to the corresponding
(hydroxymethyl)furfural oxidase (HMFO) and its
carbonyl compounds at the expense of molecular
variants (V465T, V465S, V465T/W466H and
oxygen still belongs to the challenges in chemistry, as
V367R/W466F) for the enantioselective oxidation
discussed in various recent reports and reviews using
of sec-allylic alcohols. While primary and benzylic
alcohols as well as certain sugars are well known to
be transformed by flavin-dependent oxidases, sec-
e.g. Ru-catalysts,[1] oxovanadium complexes,[2]
colloidal[3] or metallic gold.[4] Additionally to the
challenge of activating molecular oxygen as oxidant,
allylic alcohols have not been studied yet except in
the chemoselectivity is still poorly addressed. Espe-
a single report. The model substrates investigated
cially allylic alcohols are prone to various side
were oxidized enantioselectively in a kinetic reso-
reactions such as epoxidation, 1,3H-shifts followed by
lution with an E-value of up to >200. For instance
HMFO V465S/T oxidized the (S)-enantiomer of
(E)-oct-3-en-2-ol (1a) and (E)-4-phenylbut-3-en-2-
ol with E>200 giving the remaining (R)-alcohol
with ee>99% at 50% conversion. The enantiose-
lectivity could be decreased if required by medium
tautomerization or polymerization.[5] An alternative to
the metal-based oxidation, may be the biocatalytic
oxidation of alcohols, including the use of alcohol
dehydrogenases and oxidases.[6] Since alcohol dehy-
drogenases require another enzyme for cofactor [NAD
(P)+] recycling, oxidases using molecular oxygen as
engineering by the addition of cosolvents (e.g.
the direct oxidant would be preferred from a practical
dimethyl sulfoxide).
point of view.[6c,7]
Adv. Synth. Catal. 2019, 361, 1–9
1
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
asc.wiley-vch.de
Oxidases have been reported for the oxidation of
prim-alcohols[7a,c,8] as well as for specific sec-alcohols
such as the hydroxy group of α-hydroxy acids or
sugars.[7a] sec-Benzylic alcohols have been oxidized for
instance by an arylÀ alcohol oxidase,[9] the eugenol
oxidase,[10] the L182V variant of the berberine bridge
like enzyme from Arabidopsis thaliana (AtBBE15),[8b,c]
or the W466A/F variant of the 5-(hydroxymethyl)
furfural oxidase (HMFO) from Methylovorus sp.[11]
When it comes to allylic alcohols, most reports deal
with prim-allylic alcohols;[8b] sec-allylic alcohols have
only been reported using AtBBE15 L182V.[8b]
Since oxidases provide a chiral active site, the
biocatalytic oxidation of sec-alcohols can be expected
to display enantioselectivity, which may allow a kinetic
resolution. In case it is desired that both enantiomers
are oxidized, an enzyme with low enantioselectivity
would be preferred.
Here, we investigate the possibility to exploit
oxidases for the chemo- and stereo-selective oxidation
of racemic sec-allylic alcohols at the expense of
molecular oxygen as the only oxidant.
Figure 1. Docking of substrate 2a (in green) into the active site
of AtBBE15 (PDB 4UD8). The flavin cofactor is shown in
yellow with its bicovalent linkage to His115 and Cys179
(shown in pink). Residues selected for site-directed mutagenesis
are highlighted in blue (L178, L182, I184 and I409). The figure
was prepared using PyMol.
for replacement to the less bulky amino acid valine to
see the influence of these positions on the activity and
stereoselectivity. The exchange I184V was speculated
to improve the oxidase activity of the enzyme.
Results and Discussion
AtBBE15
A selection of rac-sec-allylic alcohols bearing aro-
Comparing the initial variant (L182V, the L182V
matic, aliphatic and cyclic moieties with and without exchange enables the use of molecular oxygen), with
an additional conjugated C=C double bond (5a, 6a) the I409V variant, I409V led, in general, to lower
were chosen as substrates (Scheme 1). The above conversion for the substrates investigated (Table 1).
mentioned AtBBE15 L182V variant, which needs to be The I184V exchange did not improve oxidase activity.
expressed in Komagataella phaffii (formerly classified The L182V as well as the other variants oxidized
as Pichia pastoris), was the starting point as catalyst preferentially the (S)-enantiomer leaving the (R)-
for our investigation.[8b,12]
enantiomer. The calculated enantioselectivity varied
In AtBBE15, the FAD cofactor is bi-covalently depending on the variant as well as the substrate, e.g.
bound to the enzyme backbone (Figure 1). The apolar the enantioselectivity E was >200 with AtBBE15
residues L178 and I409 in the active site were chosen L182V for 1a and 3a, but low for 2a and 4a (E=49
and 35, respectively), while the additional mutation
I409V led to low E-value only for 2a (E=26), but
was high for 1a, 3a and 4a (E>200).
HMFO
To create a library of oxidases for the oxidation of sec-
allylic alcohols, we extended our research to another
flavin-dependent oxidase previously described mainly
for the oxidation of selected prim-alcohols, the 5-
hydroxymethylfurfural oxidase (HMFO).[11,13]
In contrast to AtBBE15, the FAD in HMFO is not
covalently bound and the enzyme can efficiently be
produced in E. coli.[13a] In addition to its oxidation
activity to produce the polymer building block, 2,5-
furandicarboxylic acid (FDCA) from 5-(hydroxymeth-
yl)furfural (HMF),[11b] HMFO is active toward a wide
range of benzylic or allylic prim-alcohols and
aldehydes[13b] and its variants are able to transform sec-
Scheme 1. Oxidation of rac-sec-allylic alcohols by oxidases.
Adv. Synth. Catal. 2019, 361, 1–9
2
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
asc.wiley-vch.de
Table 1. Oxidation of allylic rac-sec-alcohols using AtBBE15
Table 2. Oxidation of rac-sec-allylic alcohols with variants of
L182V variants.[a]
HMFO.[a]
Substr.
Variant of
AtBBE15 L182V
conv.
[%]
ees
E
Entry Substr. Variant
Conv. ees
[%]
E
[%][b]
[%][b]
[c]
1a
1a
2a
2a
2a
3a
3a
3a
4a
4a
4a
5a
–
50[d]
10
55[d]
<1
8
50
14
34
57
17
44
8
>99 (R)
11 (R)
>99 (R)
n.d.[f]
>200
>200
49
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1a
2a
2a
2a
2a
2a
3a
3a
3a
3a
3a
4a
4a
4a
4a
4a
5a
5a
5a
5a
5a
wt
V465T
V465S
V465T/W466H 25
V367R/W466F 18
wt
V465T
V465S
V465T/W466H 50
V367R/W466F 33
wt
V465T
V465S
V465T/W466H 50
V367R/W466F 46
wt
V465T
V465S
V465T/W466H 50
V367R/W466F 32
wt
V465T
V465S
V465T/W466H
V367R/W466F
16
50
50
19 (R)
>99 (R) >200
>99 (R) >200
33 (R)
21 (R)
25 (R)
>99 (R) >200
>99 (R) >200
99 (R)
34 (R)
10 (R)
94 (R)
99 (R)
96 (R)
83 (R)
14 (R)
92 (R)
96 (R)
98 (R)
45 (R)
2
>200
I409V[e]
[c]
–
L178V/I184V[e]
n.d.[f]
26
>200
55
5
I409V[e]
8 (R)
[c]
–
>99 (R)
16 (R)
51 (R)
>99 (R)
20 (R)
78 (R)
8 (R)
>200
135
>200
35
102
>200
26
29
50
50
L178V/I184V[e]
I409V[e]
[c]
–
9
>200
8
21
>200
>200
>200
>200
35
L178V/I184V[e]
I409V[e]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
10
48
50
I409V[e]
[a] Condition: KPi-buffer (200 mM, pH 7.0) containing the
oxidases (1.67 μM in case of L178V/I184V variant and
16.7 μM in case of I409V variant and AtBBE15 L182V, final
concentration in 500 μL reaction volume in 4 mL glass vials),
catalase from Micrococcus lysodeikticus (15 μL, 170000 U/
mL), the substrate (50 mM). The reaction mixtures and
13
48
48
>200
>200
>200
70
°
blanks were shaken for 16 hours (170 rpm, 21 C) and
extracted with ethyl acetate (2×300 μL), dried with Na2SO4
and measured by GC-FID.
4
n.d.[c]
46
32
38
4
44 (R)
58 (R)
n.d.[c]
[b] Ees values for 1a were measured by using GC on a chiral
phase. ees values for 2a–5a were measured by using HPLC
using a chiral column.
65
n.d.[c]
n.d.[c]
4
n.d.[c]
[c] Contains the L182V exchange only.
[d] This substrate has already been reported with AtBBE15
L182V.[8b]
[a] Condition: KPi-buffer (200 mM, pH 7.0) containing the
oxidases (14.2 μM final concentration in 500 μL reaction
volume in 4 mL glass vials), catalase from Micrococcus
lysodeikticus (15 μL, 170000 U/mL), substrate (50 mM). The
[e] Performed in the presence of 2 bar oxygen pressure.
[f] Not determined due to low conversion.
°
reaction mixtures were shaken for 16 hours (170 rpm, 21 C)
and extracted with ethyl acetate (2×300 μL), dried with
Na2SO4 and analyzed by GC-FID. Conversions were
measured based on area ratio of ketone to substrate.
Reactions were conducted in duplicate.
benzylic alcohols in a stereoselective fashion.[9,11a,14]
Furthermore, the oxidation activity of HMFO on prim-
and sec-thiols has been described recently,[15] but no
activity for sec-allylic alcohols has been reported for
this enzyme, yet.
When substrates 1a–5a were tested with the wild
type enzyme HMFO, only moderate conversions were
observed at the conditions employed (Table 2, en-
[b] Ees values for 1a were measured by using GC equipped with
chiral column. Ees values for 2a–5a were measured by using
HPLC equipped with a chiral column.
[c] Not determined due to low conversion.
tries 1, 6, 11, 16, 21). Nevertheless, it is worth noting, most cases 50% conversion (at 50 mM substrate, 16 h)
that exclusively oxidation of the allylic alcohol to the and up to >99% ee. The latter indicated excellent
α,β-unsaturated ketone was observed, thus side reac- enantioselectivity, thus only one enantiomer was
tions like epoxidation did not occur. Assuming that the preferentially transformed. All variants showed prefer-
low conversion was due to a slow transformation ence to oxidize the (S)-enantiomer leaving the (R)-
caused by steric hindrance in the active site of the enantiomer, which corresponds to the same stereo-
enzyme, variants V465T/S were investigated (Fig- preference as observed with AtBBE15 L182V.
ure 2), which proved already to be useful for the
oxidation of sec-thiols by reducing steric hindrance.[15b]
O2 Pressure Study
Furthermore, the double variant V465T/W466H was
tested as well as a previously published variant Since the oxidant is gaseous molecular oxygen, an
V367R/W466F.[11a]
increase in its concentration in the buffer may lead to
The two variants V465T and V465S oxidized all improved conversion. Consequently, the reactions were
five substrates 1a–5a very efficiently, reaching in tested in the presence of 2 and 4 bar of molecular
Adv. Synth. Catal. 2019, 361, 1–9
3
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
asc.wiley-vch.de
demanding in terms of equipment and also more time
consuming, the ambient reaction conditions were
preferred. Furthermore, and probably even more
importantly, it turned out that the enantioselectivity
expressed as the E-value decreased at elevated pressure
compared to the reaction performed in the absence of
pressure. For instance, HMFO V465S displayed an E-
value of 29 for the oxidation of 4a at ambient pressure
in the presence of 10% glycerol as cosolvent, while
already at 1.5 bar of molecular oxygen, the enantiose-
lectivity decreased to a value of 5 (Table 4).
Table 4. Enantioselectivity of oxidation of substrate 4a using
HMFO variants at ambient air pressure and at 1.5 bar O2.[a]
Figure 2. Docking of substrate 2a (in green) into HMFO
V465T (PDB 6F97). The FAD is shown in yellow and residues
selected for site-directed mutagenesis are highlighted in blue
(V367, T465, and W466). Docking was performed with Yasara.
The figure was prepared using PyMol.[15b]
Entry
Variant
O2 [bar]
Conv.
[%]
ees
E
[%][b]
1
2
3
4
V465S
V465S
V465T
V465T
1.5
ambient
1.5
58
55
54
50
62
95
73
73
5
29
9
ambient
14
Table 3. Oxidation of rac-sec-allylic alcohols 1a–2a employ-
ing HMFO variants in the presence of air, 2 and 4 bar O2
pressure.[a]
[a] Condition: KPi-buffer (200 mM, pH 7.0) containing the
oxidases (2.1 μM final concentration in 1 mL reaction
volume in 4 mL glass vials), catalase from Micrococcus
lysodeikticus (30 μL, 170000 U/mL), the substrate (50 mM),
10% v/v glycerol as cosolvent. The reaction mixtures were
Entry Substr. Variant
Conv. [%]
air O2
O2
°
shaken for 16 hours (170 rpm, 21 C; additional 1.5 bar O2 for
(2 bar) (4 bar)
the mixtures with O2 pressure) and extracted with ethyl
acetate (2×500 μL), dried with Na2SO4 and analyzed by GC-
FID and HPLC.
1
2
3
4
5
6
7
8
9
1a
1a
1a
1a
1a
2a
2a
2a
2a
2a
wt
V465T
V465 S
V465T/W466H 25 18
V367R/W466F 18
wt
V465T
V465 S
V465T/W466H 50 48
V367R/
W466F
16 13
50 31
50 35
10
26
26
9
[b] Ees values were measured by using HPLC equipped with
chiral column.
9
7
29 33
50 48
50 48
35
45
44
46
46
When testing the most suitable HMFO variants
(V465S, V465T) as well as V367R/W466F to oxidize
other non-allylic secondary alcohols, it turned out that
none of the variants was able to oxidize sec-alcohols
such as 2-octanol or 1-phenyl-2-propanol. The benzylic
alcohol 1-phenylethanol was oxidized only by V465T.
Therefore, we hypothesized that the HMFO variants
(V465S, V465T) are chemoselective, differentiating
between allylic and non-allylic sec-alcohols. Further-
more, none of these variants was able to oxidize the
primary alcohol 2-phenyl-1-ethanol either, while 1-
octanol was oxidized by V465S and V465T but not
V367R/W466F.
10
33 34
[a] Condition: KPi-buffer (200 mM, pH 7.0) containing the
oxidases (14.2 μM final concentration in 500 μL reaction
volume in 4 mL glass vials), catalase from Micrococcus
lysodeikticus (15 μL, 170000 U/mL), 1a–2a (50 mM). The
°
reaction mixtures were shaken for 16 hours (170 rpm, 21 C)
and extracted with ethyl acetate (2×300 μL), dried with
Na2SO4 and analyzed by GC-FID. Conversions were
measured based on area ratio of ketone to substrate.
Reactions were done in duplicate.
Solvent Study
oxygen and compared to the reaction performed in air Furthermore, the oxidation of various substrates was
at ambient conditions (Table 3, for substrates 1a–2a; tested in the presence of various organic solvents
Table S9 for substrates 3a–5a). Interestingly the (Table 5). Interestingly, in general, the organic solvents
conversion increased at higher pressure for the wild tested were compatible with the enzyme, independent
type for substrate 2a–5a, while this was not the case whether a water miscible or immiscible organic solvent
for substrate 1a. In general, the variants could stand was used. It is worth noting that the prim-alcohols
higher pressure, however, since the handling is more ethanol and methanol could be used as cosolvents.
Adv. Synth. Catal. 2019, 361, 1–9
4
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
asc.wiley-vch.de
Table 5. Oxidation of rac-allylic alcohols with HMFO V465S
Table 5. continued
in the presence of 5% v/v organic solvents.[a]
Entry Substr. Cosolvents Conv. [%] ees
E
[%][b]
Entry Substr. Cosolvents Conv. [%] ees
[%][b]
E
55
5a
dioxane
19
19 (R)
11
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
3a
3a
3a
3a
3a
3a
3a
3a
3a
3a
3a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
5a
5a
5a
5a
5a
5a
5a
5a
5a
5a
DMSO
isooctane
glycerol
n-heptane
MeOH
EtOH
iPrOH
2-butanone 52
acetone
DMF
dioxane
DMSO
isooctane
glycerol
n-heptane
MeOH
EtOH
iPrOH
2-butanone 49
acetone
DMF
dioxane
DMSO
isooctane
glycerol
n-heptane
MeOH
EtOH
iPrOH
2-butanone 42
acetone
DMF
dioxane
DMSO
isooctane
glycerol
n-heptane
MeOH
EtOH
iPrOH
2-butanone 50
acetone
DMF
51
50
51
50
51
51
51
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) >200
[a] Condition: KPi-buffer (200 mM, pH 7.0) containing the
oxidases (14.2 μM final concentration in 1 mL reaction
volume in 4 mL glass vials), catalase from Micrococcus
lysodeikticus (30 μL, 170000 U/mL), the substrate (50 mM),
5% v/v various co-solvents. The reaction mixtures were
°
shaken for 16 hours (170 rpm, 21 C) and extracted with ethyl
acetate (2×500 μL), dried with Na2SO4 and analyzed by GC-
MS. Conversions were measured based on area ratio of
ketone to substrate.
9
51
50
50
65
49
51
50
49
50
50
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
[b] Ees values for 1a were measured by GC on a chiral phase. ees
values for 2a–5a were measured by HPLC on a chiral phase.
98 (R)
93 (R)
99 (R)
95 (R)
96 (R)
92 (R)
95 (R)
85 (R)
83 (R)
98 (R)
86 (R)
74 (R)
54 (R)
99 (R)
62 (R)
51 (R)
54 (R)
70 (R)
70 (R)
70 (R)
56 (R)
44 (R)
14
>200
>200
146
>200
79
146
44
49
41
37
17
84
Glycerol possessing two prim- and a sec-alcohol
functionality could also be used, whereby it turned out
that the reactions in glycerol are faster than in DMSO
(SI, Table S4). Thus, glycerol did not inhibit the
oxidation but led to lower enantioselectivity for
substrates 3a–5a which enabled to reach higher
conversions (e.g. 72%, entry 25, Table 5). While
substrate 1a was converted with an E-value >200 in
the presence of all cosolvents, the other substrates
were transformed with a significant decrease in
enantioselectivity E in the presence of DMSO. Thus,
additionally to the (S)-enantiomer also the (R)-enan-
tiomer was oxidized. The E-value for substrate 5a with
DMSO and glycerol was found to be 5 and 4,
respectively. Consequently, 68% of conversion was
obtained in the case of substrate 5a in the presence of
glycerol as cosolvent. Similarly, substrate 3a was
transformed in glycerol with low E-value, leading to
higher conversion (72%) compared to other solvents.
Compound 6a was in general not well accepted
leading only to low conversion (3%, Table S8).
48
55
50
49
36
72
42
38
36
50
11
24
18
84
12
124
124
74
42
37
40
67
63
55
79
53
50
51
7
>99 (R) 14
>99 (R) 18
>99 (R) 49
>99 (R)
7
For obtaining semi-preparative amounts of the
products 1b–5b, experiments were performed with
12.5 mmol at 50 mM substrate concentration employ-
ing the variant V465S (Table 6). After purification, the
isolated yields were determined and NMR analysis
proved the structure and the purity of the isolated
ketones (1b–5b).
>99 (R) 80
>99 (R) >200
>99 (R) >200
>99 (R) >200
>99 (R) 80
>99 (R) 116
>99 (R) 31
53
52
58
17
20
68
20
18
11
19
dioxane
DMSO
isooctane
glycerol
n-heptane
MeOH
EtOH
iPrOH
2-butanone 17
acetone
DMF
13 (R)
20 (R)
76 (R)
19 (R)
17 (R)
10 (R)
17 (R)
16 (R)
30 (R)
22 (R)
5
11
4
9
9
10
7
10
13
10
Conclusion
The biocatalytic oxidation of sec-allylic alcohols to the
corresponding allylic ketones represents a valuable
alternative for chemical methods, which often require
harsh conditions and suffer from poor chemo- and
enantio-selectivity. In the current study, the O2-depend-
ent oxidation of sec-allylic alcohols was performed
using flavin-dependent alcohol oxidases namely 5-
27
22
Adv. Synth. Catal. 2019, 361, 1–9
5
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
asc.wiley-vch.de
methanol (30 mL), sodium borohydrate was slowly added on
ice (see the details in Table S1). The reaction mixture was
stirred for 2 hours and formation of the product was monitored
by TLC. When the reaction was completed, quenching was
done by using saturated aqueous NH4Cl (15 mL). Then the
resultant mixture was concentrated under reduced pressure and
the residue was extracted with ethyl acetate (3×20 mL). The
combined organic fractions were washed with brine, dried with
Na2SO4 and concentrated under reduced pressure. Purification
of the residue was done by flash chromatography (8:2, c-
hexane:EtOAc).
Table 6. Semi-preparative scale oxidation using HMFO
V465S.[a]
Entry
Substr.
Conv.
[%]
Isolated
yields b [%]
ees
[%]
E
1
2
3
4
5
1a
2a
3a
4a
5a
50
53
35
52
31
70[b]
33
54
64
31
>99
92
35
92
35
>200
32
7
40
11
[a] Condition: KPi-buffer (200 mM, pH 7.0) containing the
oxidase (14.2 μM final concentration in 25 mL reaction
volume), catalase from Micrococcus lysodeikticus (750 μL,
170000 U/mL) and the substrate (50 mM). The reaction
Preparation of the Biocatalysts
HMFO wt (pEG 387), HMFO V465S (pEG 392),
HMFO V465T (pEG 393), HMFO W466H (pEG 390)
and HMFO V465T/W466H (pEG 395)
°
mixtures were shaken for 48 h (170 rpm, 21 C) and extracted
with ethyl acetate (2×50 mL), dried with Na2SO4 and
analyzed by GC-MS. Conversions were measured based on
area ratio of ketone to substrate. The percentage of isolated
yield refers to the conversion achieved.
For the different variants of HMFO, the same expression and
purification method was used as it follows:
[b] The remaining substrate was isolated in quantitative yield
with respect to the observed conversion.
Expression: For HMFO expression, an overnight culture of E.
coli BL21(DE3) cells bearing the previously prepared SUMO-
HMFO encoding plasmid (ChampionTM pET SUMO) in
200 mL of Terrific Broth containing 50 μg/mL kanamycin and
°
grown at 37 C until it reached an OD600 of 0.8–1.0. Cells were
induced with isopropyl-β-D-thiogalactopyranoside (IPTG,
(hydroxymethyl)furfural oxidase and AtBBE15 and
variants thereof.
°
1.0 mM) and grown overnight at 20 C. Cells were harvested by
The created library of oxidases allows the chemo-
selective oxidation of allylic alcohols to the corre-
sponding α,β-unsaturated ketones without any detect-
able side reaction such as epoxidation, polymerization
or hydride shifts. From the oxidases tested possessing
a covalently bound FAD, AtBBE15 L182V turned out
to be the most suitable. From the HMFO variants
tested, the two variants V465S and V465T led to the
highest conversions (up to 50%) and excellent enantio-
selectivity (E>200) for the oxidation of 1a–4a. All
oxidases investigated preferentially oxidized the (R)-
allylic alcohol leaving the (S)-enantiomer. Especially
the HMFO V465S/T variant showed high enantiose-
lectivity (E>200) for most substrates (except 5a). The
enantioselectivity could be tuned by applying either
pressure or by the addition of cosolvents. For instance,
the addition of DMSO as cosolvent led to a decrease in
enantioselectivity, which was associated with signifi-
cantly higher conversions (up to 79%) for selected
substrates. Thus, the oxidases may be employed for
non-enantioselective oxidation as well as for enantio-
selective oxidation of allylic alcohols.
centrifugation at 3730 g for 15 min (Hettich® Rotina 420R
°
centrifuge, 4 C) and resuspended in TrisÀ HCl (35 mL,
100 mM, pH 8.0) supplemented with glycerol (10% v/v), NaCl
(150 mM), and FAD (10 μM). The cell extract was obtained by
sonication with a Branson Digital Sonifier 250 (30% amplitude,
2 min, 1 sec pulse, 4 sec pause). The lysate was cleared by
centrifugation (20000×g for 15 min).
Purification: His-Tagged HMFO was purified by immobilized
Ni-affinity chromatography (5 mL HisTrap FF column, GE
Healthcare) following standard protocols with a 5 to 500 mM
gradient of imidazole (binding buffer: TrisÀ HCl, 50 mM,
pH 8.0 containing 150 mM NaCl and 5 mM imidazole; elution
buffer: TrisÀ HCl, 50 mM, pH 8.0 containing 150 mM NaCl and
500 mM imidazole). Fractions containing HMFO were pooled
concentrated by ultrafiltration (20 mL, 50 kDa cut-off, Viva-
spin) and desalted (SephadexTM G-25M, GE Healthcare). After
desalting the fractions were shock frosted in liquid nitrogen and
°
stored at À 20 C. For activity tests the lyophilized enzyme
preparation were dissolved in potassium phosphate buffer
(100 mM, pH 7.0) without cleaving off the SUMO-tag. For
running the biotransformations, the lyophilized pure enzymes
were rehydrated in the reaction buffer just before using them,
then the concentration of each variants was measured by
Bradford assay.
Experimental Section
Site-Directed Mutagenesis: Site-directed mutagenesis of the
wild-type HMFO gene was performed using two-step whole-
plasmid PCR. For the creation of V465T/W466H, the HMFO-
W466H plasmid was used as template. The primers were
ordered at IDT (Leuven, Belgium). After three cycles of linear
PCR, the mixture containing the forward primer and the mixture
with the reverse primer were combined for additional 15 cycles.
Template DNA was cleaved with DpnI (New England Bio-
Synthesis of Allylic Alcohols from their Corre-
sponding Ketones
Substrates (E)-oct-3-en-2-ol (1a), (E)-4-phenylbut-3-en-2-ol
(2a), (E)-4-(4-chlorophenyl)but-3-en-2-ol (3a) and (E)-4-(4-
methylphenyl)but-3-en-2-ol (4a) were synthesized from their
corresponding ketones. To a solution of various ketones in
Adv. Synth. Catal. 2019, 361, 1–9
6
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
asc.wiley-vch.de
Labs, Ipswich, MA, USA). The plasmid was purified with a
PCR purification kit (Qiagen, Hilden, Germany) and trans-
formed into E. coli TOP10 cells. The introduction of the
mutations was confirmed by sequencing.
phases were dried with Na2SO4. Samples were prepared from
the dried organic phase without further treatment and measured
on GC-MS and GC-FID. When no cosolvent was used, the
reaction volume was reduced to 500 μL and 15 μL of catalase
(170000 U/mL) was used. In this case, the extraction was done
with ethyl acetate (2×300 μL). The rest of the procedure was
the same. In case of applying oxygen pressure, the glass vials
were left unscrewed and fixed in a rack in the oxygen chamber
and the oxygen pressure was applied. Conversions were
measured based on area ratio of ketone to substrate in each
biotransformation mixture by using GC without addition of any
extra compound as internal standard.
AtBBE15 Variants
Creation of variants: AtBBE15 L182V originates from previous
work and was used as template.[12] Using primer pair fw
GATTCACCGTTAACGGTTTGGAACCCTTAC
and
rw
GTAAGGGTTCCAAACCGTTAACGGTGAATC the AtBBE-
like 15 variant L182V/L409V was created using the Quick
Change Protocol. For the AtBBE-like 15 L178V/I182V/I184V
variant a synthetic gene was ordered from Geneart (Regensburg,
Germany) composed of sequences encoding the α-factor and the
enzyme. The gene was clone to the pPICZα vector for
expression.
Determination of Optical Purity
The enantiomeric excess of remaining alcohols was analyzed by
HPLC (in case of substrates 2a–5a, see Tables S10–S13) and
GC (in case of 1a, see Tables S14–S15) on a chiral phase.
Absolute configurations were assigned by comparison of elution
order of enantiomers on chiral HPLC and chiral GC with
published data.[16] For determination of the absolute config-
uration of 1a, by comparison of the elution order on chiral GC
with literature data, the alcohol moiety had to be acetylated.[17]
For that purpose, derivatization was performed by adding 4-(N,
N-dimethylamino)pyridine (5 mg) dissolved in acetic anhydride
(100 μL). After washing with water, drying with Na2SO4 and
measuring on chiral GC, the stereopreference of the enzyme
toward 1a substrate was confirmed.
Expression: The expressions of AtBBE15 L182V was per-
formed as described before.[12] AtBBE-like15 L182V/I409V and
AtBBE-like15 L178V/L182V/I184V were expressed in shake
flask using minimal media. The compositions of every used
media are listed in Table S2 and the components are listed in
Table S3. 50 mL of sterile BMD medium was added to 300 mL
Erlenmeyer flasks and inoculated with the respective expression
°
strains. The cells were grown for 72 hours (300 rpm, 28 C).
After 72 hours expression was induced using 5 mL of BMM10
medium. Subsequently every 12 hours, 50 μL of absolute
methanol was added for 72 hours.
Purification: The cells were removed using an Eppendorf
centrifuge 5810 R (Eppendorf AG, Hamburg, Germany) at
Docking Experiments
°
2800 g at 4 C for 15 minutes. A 5 mL NiÀ NTA fast flow
Each enzyme was prepared in PyMol, removing all water
molecules present in the enzyme structure. The substrate (R)-2a
was separately prepared in Yasara.
column (GE Healthcare, Chicago, Illinois, USA) was equili-
brated with 50 mM potassium phosphate buffer containing
10 mM imidazole pH 8. The supernatant was loaded to the
column using a peristaltic pump at a flow rate of 15 mL/min.
The column was washed with 20 mL 50 mM potassium
phosphate buffer containing 20 mM imidazole pH 8, subse-
quently with 20 mL 50 mM potassium phosphate buffer
containing 40 mM imidazole pH 8. The enzymes were eluted
using 50 mM potassium phosphate buffer containing 150 mM
imidazole pH 8 and dialysed over night against 50 mM
potassium phosphate buffer pH 8. Finally, the enzyme was
concentrated using Amicon® Ultra Centrifugal Filters (Merck
KGaA, Darmstadt Germany) and shock frozen in liquid nitro-
gen.
For docking, the adapted structure file was loaded to Yasara.
The N5-atom of FAD was chosen as the center of the simulation
cell with a 10 Å diameter defined around the selected atom.
AMBER03 was chosen as the force field. The substrate was
added to the prepared file and the energy minimization experi-
ments were run. Afterwards the docking experiments were
performed using docking parameters including Autodock
VINA, 25 docking runs, Cluster RMSD 5.00 Å. The outcome
was analyzed in PyMol.
Acknowledgements
General Procedure for Oxidation of Sec-Allylic
Alcohols by Using Oxidases
SG and WK acknowledge the Austrian Science Fund FWF for
funding (Lise Meitner program M 2271-B21). WE is thankful to
the Science and Technology Development Fund (STDF), Cairo,
Egypt. (Project ID: 25411 STF).
The oxidation reactions were run in 4 mL glass vials under the
conditions described below:
For HMFO the oxidase (2.1 to 14.2 μM final concentration in
1 mL reaction volume), catalase from Micrococcus lysodeikticus
(30 μL, 170000 U/mL), the substrate (50 mM final concentra-
tion) and 5% to 50% v/v of various cosolvents (in case of
cosolvent study) were added to the buffer (KPi, 200 mM,
pH 7.0). The biotransformation vials were incubated for
°
16 hours at 21 C (170 rpm, vertical shaking). The extraction
was done with ethyl acetate (2×500 μL). Combined organic
Adv. Synth. Catal. 2019, 361, 1–9
7
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
asc.wiley-vch.de
[10] Q. T. Nguyen, G. de Gonzalo, C. Binda, A. Rioz-
References
Martínez, A. Mattevi, M. W. Fraaije, ChemBioChem
2016, 17, 1359–1366.
[11] a) W. P. Dijkman, C. Binda, M. W. Fraaije, A. Mattevi,
ACS Catal. 2015, 5, 1833–1839; b) W. P. Dijkman,
M. W. Fraaije, Appl. Microbiol. Biotechnol. 2014, 80,
1082–1090.
[12] B. Daniel, T. Pavkov-Keller, B. Steiner, A. Dordic, A.
Gutmann, B. Nidetzky, C. W. Sensen, E. Van Der Graaff,
S. Wallner, K. Gruber, J. Biol. Chem. 2015, 290, 18770–
18781.
[13] a) W. P. Dijkman, G. de Gonzalo, A. Mattevi, M. W.
Fraaije, Appl. Microbiol. Biotechnol. 2013, 97, 5177–
5188; b) W. P. Dijkman, D. E. Groothuis, M. W. Fraaije,
Angew. Chem. 2014, 126, 6633–6636; Angew. Chem. Int.
Ed. 2014, 53, 6515–6518; c) M. Pickl, C. Winkler, S.
Glueck, M. Fraaije, K. Faber, Molecules 2017, 22, 2205.
[14] A. Hernández-Ortega, F. Lucas, P. Ferreira, M. Medina,
V. Guallar, A. T. Martínez, J. Biol. Chem. 2011, 286,
41105–41114.
[15] a) T. A. Ewing, W. P. Dijkman, J. M. Vervoort, M. W.
Fraaije, W. J. van Berkel, Angew. Chem. 2014, 126,
13422–13425; Angew. Chem. Int. Ed. 2014, 53, 13206–
13209; b) M. Pickl, A. Swoboda, E. Romero, C. K.
Winkler, C. Binda, A. Mattevi, K. Faber, M. W. Fraaije,
Angew. Chem. 2018, 130, 2914–2918; Angew. Chem. Int.
Ed. 2018, 57, 2864–2868.
[16] a) X. Chen, H. Zhou, K. Zhang, J. Li, H. Huang, Org.
Lett. 2014, 16, 3912–3915; b) P. He, X. Liu, H. Zheng,
W. Li, L. Lin, X. Feng, Org. Lett. 2012, 14, 5134–5137;
c) F. Chen, Y. Zhang, L. Yu, S. Zhu, Angew. Chem. 2017,
129, 2054–2057; Angew. Chem. Int. Ed. 2017, 56, 2022–
2025.
[1] S. Muthusamy, N. Kumarswamyreddy, V. Kesavan, S.
Chandrasekaran, Tetrahedron Lett. 2016, 57, 5551–5559.
[2] J. A. L. da Silva, J. J. R. F. da Silva, A. J. L. Pombeiro,
Coord. Chem. Rev. 2011, 255, 2232–2248.
[3] T. Tsukuda, H. Tsunoyama, H. Sakurai, Chem. Asian J.
2011, 6, 736–748.
[4] M. L. Personick, R. J. Madix, C. M. Friend, ACS Catal.
2017, 7, 965–985.
[5] A. Abad, A. Corma, H. García, Pure Appl. Chem. 2007,
79, 1847–1854.
[6] a) J. Dong, E. Fernández-Fueyo, F. Hollmann, C. E.
Paul, M. Pesic, S. Schmidt, Y. Wang, S. Younes, W.
Zhang, Angew. Chem. 2018, 130, 9380–9404; Angew.
Chem. Int. Ed. 2018, 57, 9238–9261; b) J. Liu, S. Wu, Z.
Li, Curr. Opin. Chem. Biol. 2018, 43, 77–86; c) F.
Hollmann, I. W. Arends, K. Buehler, A. Schallmey, B.
Bühler, Green Chem. 2011, 13, 226–265.
[7] a) M. Pickl, M. Fuchs, S. M. Glueck, K. Faber, Appl.
Microbiol. Biotechnol. 2015, 99, 6617–6642; b) C.
Holec, K. Neufeld, J. Pietruszka, Adv. Synth. Catal.
2016, 358, 1810–1819; c) R. S. Heath, W. R. Birming-
ham, M. P. Thompson, A. Taglieber, L. Daviet, N. J.
Turner, ChemBioChem 2019, 20, 276–281.
[8] a) R. S. Heath, W. R. Birmingham, M. P. Thompson, A.
Taglieber, L. Daviet, N. J. Turner, ChemBioChem 2019,
20, 276–281; b) S. Pils, K. Schnabl, S. Wallner, M.
Kljajic, N. Kupresanin, R. Breinbauer, M. Fuchs, R.
Rocha, J. H. Schrittwieser, W. Kroutil, B. Daniel, P.
Macheroux, J. Mol. Catal. B-Enzym. 2016, 133, S6–S14;
c) B. Daniel, B. Konrad, M. Toplak, M. Lahham, J.
Messenlehner, A. Winkler, P. Macheroux, Arch. Biochem.
Biophys. 2017, 632, 88–103.
[17] K. Edegger, C. C. Gruber, T. M. Poessl, S. R. Wallner, I.
Lavandera, K. Faber, F. Niehaus, J. Eck, R. Oehrlein, A.
Hafner, Chem. Commun. 2006, 2402–2404.
[9] A. Hernández-Ortega, P. Ferreira, P. Merino, M. Medina,
V. Guallar, A. T. Martínez, ChemBioChem 2012, 13,
427–435.
Adv. Synth. Catal. 2019, 361, 1–9
8
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA
��
These are not the final page numbers!

UPDATES
Biocatalytic Enantioselective Oxidation of Sec-Allylic
Alcohols with Flavin-Dependent Oxidases
Adv. Synth. Catal. 2019, 361, 1–9
S. Gandomkar, E. Jost, D. Loidolt, A. Swoboda, M. Pickl, W.
Elaily, B. Daniel, M. W. Fraaije, P. Macheroux, W. Kroutil*