Engineering cyclohexanone monooxygenase for the production of methyl propanoate

Article abstract of DOI:10.1021/acschembio.6b00965

A previous study showed that cyclohexanone monooxygenase from Acinetobacter calcoaceticus (AcCHMO) catalyzes the Baeyer-Villiger oxidation of 2-butanone, yielding ethyl acetate and methyl propanoate as products. Methyl propanoate is of industrial interest as a precursor of acrylic plastic. Here, various residues near the substrate and NADP+ binding sites in AcCHMO were subjected to saturation mutagenesis to enhance both the activity on 2-butanone and the regioselectivity toward methyl propanoate. The resulting libraries were screened using whole cell biotransfor-mations, and headspace gas chromatography-mass spectrometry was used to identify improved AcCHMO variants. This revealed that the I491A AcCHMO mutant exhibits a significant improvement over the wild type enzyme in the desired regioselectivity using 2-butanone as a substrate (40% vs 26% methyl propanoate, respectively). Another interesting mutant is the T56S AcCHMO mutant, which exhibits a higher conversion yield (92%) and k_cat (0.5 s^-1) than wild type AcCHMO (52% and 0.3 s^-1, respectively). Interestingly, the uncoupling rate for the T56S AcCHMO mutant is also significantly lower than that for the wild type enzyme. The T56S/I491A double mutant combined the beneficial effects of both mutations leading to higher conversion and improved regioselectivity. This study shows that even for a relatively small aliphatic substrate (2-butanone), catalytic efficiency and regioselectivity can be tuned by structure-inspired enzyme engineering.

Full text of DOI:10.1021/acschembio.6b00965

Subscriber access provided by Fudan University
Article
Engineering Cyclohexanone Monooxygenase
for the Production of Methyl Propanoate
Hugo L. van Beek, Elvira Romero, and Marco W. Fraaije
ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.6b00965 • Publication Date (Web): 09 Dec 2016
Downloaded from http://pubs.acs.org on December 12, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 12
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
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
55
56
57
58
59
60
TOC figure
80x39mm (300 x 300 DPI)
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 12
Engineering Cyclohexanone Monooxygenase for the Production
of Methyl Propanoate
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
δ,ϕ,§
δ,§
δ,
Hugo L. van Beek,
Elvira Romero, and Marco W. Fraaije *
δ
Molecular Enzymology Group, University of Groningen, Nijenborgh 4, 9747AG Groningen, The Netherlands
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ABSTRACT: A previous study showed that cyclohexanone monooxygenase from Acinetobacter calcoaceticus (AcCHMO)
catalyzes the BaeyerꢀVilliger oxidation of 2ꢀbutanone, yielding ethyl acetate and methyl propanoate as products. Methyl
propanoate is of industrial interest as a precursor of acrylic plastic. Here, various residues near the substrate and NADP
binding sites in AcCHMO were subjected to saturation mutagenesis to enhance both the activity on 2ꢀbutanone and the
regioselectivity toward methyl propanoate. The resulting libraries were screened using whole cell biotransformations, and
headspace gas chromatographyꢀmass spectrometry was used to identify improved AcCHMO variants. This revealed that the
I491A AcCHMO mutant exhibits a significant improvement over the wild type enzyme in the desired regioselectivity using
+
2
ꢀbutanone as a substrate (40% vs. 26% methyl propanoate, respectively). Another interesting mutant is the T56S AcCHMO
ꢀ
1
ꢀ
1
mutant, which exhibits a higher conversion yield (92%) and kcat (0.5 s ) than wild type AcCHMO (52% and 0.3 s ,
respectively). Interestingly, the uncoupling rate for the T56S AcCHMO mutant is also significantly lower than that for the
wild type enzyme. The T56S/I491A double mutant combined the beneficial effects of both mutations leading to higher
conversion and improved regioselectivity. This study shows that even for a relatively small aliphatic substrate (2ꢀbutanone),
catalytic efficiency and regioselectivity can be tuned by structureꢀinspired enzyme engineering.
BaeyerꢀVilliger monooxygenases (BVMOs) are broadening the applicability of these enzymes in
1
flavinꢀdependent enzymes that catalyze the insertion organic synthesis. Electronic and steric effects
of an oxygen atom next to the carbonyl group of control the regioselectivity of these reactions. After
1
ketone substrates, producing esters or lactones. forming the Criegee intermediate, one of the carbon
Oxidation of aldehydes, heteroatoms, and epoxidatꢀ atoms adjacent to the substrate carbonyl group
1
ions can also be carried out by BVMOs. These migrates to the closest oxygen atom of the peroxide
reactions are of remarkable industrial interest, for group (Scheme 1, 3A). The antiperiplanar
example in the synthesis of valuable steroids, conformation of the migrating carbonꢀcarbon bond
sulfoxides, βꢀamino acids, and eneꢀ and enolꢀ and the oxygenꢀoxygen bond allows the migration to
2
,3
7,8
lactones. BVMOs generally exhibit high regioꢀ and proceed. The normal regioisomer is formed by the
stereoselectivity in contrast to the chemical oxidants migration of the more nucleophilic carbon, which in
utilized for BaeyerꢀVilliger reactions such as peracids most cases is the more substituted carbon. A different
1
or hydrogen peroxide. Implementing BVMOs instead product, the abnormal regioisomer, is formed when
of these chemicals also avoids several environmental the less nucleophilic carbon migrates during the
and safety issues.
Criegee rearrangement. Formation of the abnormal
The overall catalytic cycle of BVMOs with ketone regioisomer has been detected in the reaction of
9
substrates is well understood, based on steadyꢀstate BVMOs with benzoꢀfused 3 ketones,
cycloꢀ
4
ꢀ6
3,10,11
and rapid kinetic studies (Scheme 1). After binding alkanones,
cycloalkenones, linear aliphatic βꢀ
12
13
of NADPH to the enzyme, the 4(R)ꢀhydride is hydroxyketones, βꢀaminoketones, and long ketones
14
transferred from NADPH to the FAD N5 atom. The (8ꢀ12 carbons).
reduced FAD reacts with dioxygen to form a C4aꢀ
Recently, we reported that several cyclohexanone
peroxyflavin intermediate. In the absence of a ketone monooxygenases (CHMO; EC 1.14.13.22) catalyze
substrate, the C4aꢀperoxyflavin intermediate decays the BaeyerꢀVilliger oxidation of 2ꢀbutanone, yielding
+
yielding hydrogen peroxide, NADP , and oxidized the normal product ethyl acetate and a significant
FAD. This process is known as uncoupling. amount of the abnormal product methyl propanoate
15
Alternatively, the C4aꢀperoxyflavin intermediate (14–27% of total product). All CHMOs capable of
carries out a nucleophilic attack on the carbonyl carrying out this reaction exhibited considerably lower
1
5
carbon of the substrate resulting in the formation of a activity on 2ꢀbutanone than on cyclohexanone. In
tetrahedral Criegee intermediate. The rearrangement the present work, we aim to increase the activity of a
of the Criegee intermediate results in the formation of CHMO on 2ꢀbutanone and to shift the regioselectivity
C4aꢀhydroxyflavin and an ester or lactone. After of this reaction to favor the formation of methyl
elimination of a water molecule from the C4aꢀ propanoate. Methyl propanoate is of industrial interest
hydroxyflavin, the FAD is returned to the oxidized because it can be converted into methyl methacrylate
+
16
state. The ester or lactone and NADP are released by reaction with formaldehyde. Methyl methacrylate
from the enzyme active site to close the catalytic cycle. is polymerized to form the acrylic plastic poly(methyl
Certain BVMOꢀcatalyzed reactions yield either two methacrylate), which is used in many applications. A
regioisomeric products or only the unexpected one,
1
ACS Paragon Plus Environment

Page 3 of 12
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
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
55
56
57
58
59
60
Scheme 1. Catalytic cycle of AcCHMO with 2ꢀbutanone as a substrate. For FAD, R is ribitol adenosine diphosphate.
Either ethyl acetate (Eac) or methyl propanoate (Mpr) is produced.
renewable feedstock may be used to produce 2ꢀ mutagenesis. The resulting libraries were screened in
17,18
butanone.
In addition, methyl propanoate is used 96ꢀwell plates using headspace gas chromatographyꢀ
19,20
for flavoring and as a solvent.
mass spectrometry (GCꢀMS). Several improved
Recently, a complete reversal in the regioselectivity variants were discovered and characterized using
of the reaction of CHMO with (+)ꢀtransꢀ steadyꢀstate and rapid kinetics to elucidate the specific
dihydrocarvone has been achieved by replacing only reaction steps affected by the mutations.
1
1
three active site residues. Another study showed that
the reaction of a single mutant of cyclopentanone
monooxygenase (CPMO; 1.14.13.16) with 3ꢀ(2ꢀ
oxocyclohexyl)propanenitrile preferentially produced
the abnormal regioisomer in contrast to the wild type
1
0
(
WT) enzyme. Both works indicate that steric effects
can control the regioselectivity of a reaction by
influencing the exact position of the substrate in the
active site. Although we anticipated that controlling
the regioselectivity may be more difficult when
smaller substrates are used, we engineered the CHMO
from Acinetobacter calcoaceticus NCIMB 9871
(
AcCHMO) to increase the ratio of the desired
abnormal product over the normal product in the
reaction with 2ꢀbutanone. AcCHMO was selected for
this work because it shows better or similar activity
and regioselectivity toward methyl propanoate than
Figure 1. Active site of WT AcCHMO. A selection of
residues that were targeted in this study is highlighted in
sticks (green carbons). A superimposition of the WT
AcCHMO model and the crystal structure of RhCHMO in
complex with ɛꢀcaprolactone (ɛꢀCapr) was carried out to
infer the position of the ligand (cyan carbons). Distances in
1
5
other CHMOs. Since there is no available crystal
structure of AcCHMO, a homology model of
AcCHMO, available BVMO crystal structures, and
data from previous mutagenesis studies were used to
select 35 residues as targets for siteꢀsaturation
2
1
2
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 12
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Å between residues and either the ligand or FAD (yellow
carbons) are indicated above the broken lines.
30–100 E. coli colonies were screened per
26
randomization site using NNK codon degeneracy.
Four hits showing an improvement in the desired
regioselectivity were identified in libraryꢀA (Table
S1). Cultures expressing the L244R, F281H, L435N,
and I491A variants produced 30 to 40% methyl
propanoate (percentage of total product), while the
WT enzyme produced 26% methyl propanoate (Table
RESULTS AND DISCUSSION
Design of the Libraries. Our previous study
showed that AcCHMO catalyzes the BaeyerꢀVilliger
1). LibraryꢀB resulted in the identification of the T56S
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
oxidation of 2ꢀbutanone yielding both ethyl acetate mutant which exhibits a higher activity on 2ꢀbutanone.
15
and methyl propanoate (Scheme 1). Here, we have The T56S mutant converted 92% of 2ꢀbutanone
carried out saturation mutagenesis to increase the compared to 52% for the WT AcCHMO in 24 h
activity of AcCHMO on 2ꢀbutanone and to improve
the regioselectivity toward methyl propanoate. Three
(Table 1). Other AcCHMO mutants obtained at
positions 56, 244, 281, 435, and 491 were not better
libraries were generated named A, B, and C (Table than those described above (Table S2). Additional
S1) and, in total, 35 residues were targeted for beneficial mutations were not found in libraryꢀC.
randomization. The randomization sites were chosen
To obtain
a
mutant improved in both
using a homology model of AcCHMO that was conꢀ regioselectivity and activity on 2ꢀbutanone, the best
21
structed for a previous study. We superimposed this mutations found in libraryꢀA were combined with the
model onto the crystal structure of phenylacetone hit found in libraryꢀB (Table 1). The double mutants
monooxygenase (PAMO; EC 1.14.13.92) in complex T56S/L244R and T56S/I491A presented
a
+
with the inhibitor MES and NADP (PDB ID: 2YLT; combination of the desired properties observed in the
22
RMSD = 1.06). This task allowed inferring the corresponding single mutants. The triple mutant
position of the above ligands relative to that of the T56S/L244R/I491A did not show an increase in the
residues in the AcCHMO model. Various AcCHMO formation of either total product or methyl propanoate
+
when compared to the T56S/I491A variant. The best
residues near the substrate and NADP binding sites
(≤ 15 Å) were chosen as randomization sites. In the regioselectivity was found for the T56S/L435N/I491A
case of libraryꢀA, the equivalent residues in PAMO mutant. This triple mutant produced 49% methyl
2
3,24
were targets of mutagenesis in previous studies.
propanoate compared to 43% for the T56S/I491A
While we were generating these mutants, the crystal variant. However, the T56S/L435N/I491A mutant
structure of Rhodococcus sp. HIꢀ31 CHMO in produced less total product than the WT AcCHMO.
+
complex
with
ɛꢀcaprolactone
and
NADP This is in contrast to the T56S/I491A and the
25
(
RhCHMOTight; PDB ID: 4RG3) was solved. In the T56S/L244R mutants, which produce the most methyl
RhCHMOTight crystal structure, ɛꢀcaprolactone is propanoate when comparing absolute amounts.
bound in an analogous position to that of MES in the
WT PAMO crystal structure. Figure 1 shows the
Table 1. Conversion of 2ꢀButanone and Production of
position in the AcCHMO active site of several
residues mutated in the present work and ɛꢀ
caprolactone, based on a superimposition of the WT
AcCHMO model and the RhCHMOTight crystal
structure (RMSD = 0.48).
a
Methyl Propanoate by AcCHMO
b
c
AcCHMO
Conversion (%)
Mpr (%)
Cells
52
92
55
52
13
35
88
73
67
40
E
Cells
26
26
32
30
33
40
29
43
43
49
E
WT
48
85
53
33
8
24
24
28
27
35
34
29
39
39
51
T56S
L244R
Whole Cell Screening. The BaeyerꢀVilliger
oxidation of 2ꢀbutanone catalyzed by AcCHMO was F281H
first investigated using whole cells of Escherichia coli
L435N
expressing the WT enzyme and its mutants. Whole
I491A
25
79
59
56
34
cell biotransformations are generally preferred over
purified enzyme reactions in industrial applications. In T56S/L244R
addition, this approach greatly simplifies the
screening of large libraries in the laboratory.
T56S/I491A
T56S/L244R/I491A
Previously, we confirmed that E. coli cells do not
produce esterases or other enzymes with activity on T56S/L435N/I491A
a
either ethyl acetate or methyl propanoate under Analyses were carried out by headspace GCꢀMS after 24 h
1
5
similar conditions to those used here. Culture of E. coli culture expressing CHMO at 24 °C (cells) or after
medium containing the expression inducer and 2ꢀ reacting purified enzyme (2 ꢁM) for 5.5 h at 25 °C and pH
butanone was added to each well of a 96ꢀsquare well 7.0 (E). The standard deviation for the conversion values is
plate prior to inoculation. After 24 h of incubation of ≤17% and that for the yields of methyl propanoate is ≤6%.
b
these cultures, formation of ethyl acetate and methyl It is based on two or three replicates. Relative to the initial
c
propanoate was determined in air samples collected concentration of 2ꢀbutanone (11 mM). Yields of methyl
from the headspace of each well using GCꢀMS.
propanoate are given in percentage of total product.
3
ACS Paragon Plus Environment

Page 5 of 12
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Reactions Using Purified AcCHMO. Small scale by adding 0.8 M KCl to the reactions. In contrast, the
conversions of 2ꢀbutanone (11 mM) were carried out addition of KCl did only slightly affect the enzyme
using purified AcCHMO fused to the NADPH regioselectivity.
regeneration
PTDH; 2 ꢁM).
enzyme
phosphiteꢀdehydrogenase
27,28
(
Fusing PTDH to AcCHMO has
not significant adverse effects, based on previous
kinetic studies2 and regioꢀ and stereoselective
transformations.
7,28
The reactions contained the
purified WT AcCHMO or one of the mutants
identified by whole cell screening (T56S, L244R,
F281H, L435N, I491A, T56S/L244R, T56S/I491A,
T56S/L244R/I491A, and T56S/L435N/I491A). In
general, the results obtained by using purified
AcCHMOs were in perfect agreement with those
achieved using whole cells (Table 1). All variants
exhibited a better regioselectivity toward methyl
propanoate (28–51% methyl propanoate) than the WT
enzyme (24% methyl propanoate), except for the
T56S variant. Yet, the latter variant produced the
highest amount of total product.
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
55
56
57
58
59
60
To study the effect of reaction conditions on the
regioselectivity and activity of AcCHMO, purified
WT enzyme and several variants were used to convert
2ꢀbutanone at various temperatures and pH values
(Figure 2 and Table S3). In the case of the WT
enzyme, the amount of total product at all pH values
was 2ꢀ5ꢀfold higher at 30 °C than that at 4 °C. The
desired regioselectivity was slightly improved at 4 °C
compared to that at 37 °C (1.5ꢀfold). The amount of
total product for the WT enzyme was lower at pH 6.0
than that at pH 7.0 and 8.5 at all temperatures (2–4ꢀ
fold), while the regioselectivity was pH independent.
Similar trends were observed for all mutants.
Figure 2. Conversion of 2ꢀbutanone and production of
methyl propanoate by purified AcCHMO. WT (●), T56S
(■), L435N (○), I491A (□), T56S/I491A (▲),
T56S/L435N/I491A (ꢀ) AcCHMOs. Reaction contained 2
ꢁ
M AcCHMO fused to PTDH, 11 mM 2ꢀbutanone, 150 ꢁM
NADPH, 50 mM TrisꢀHCl, pH 7.0, and 20 mM phosphite.
Analyses were carried out by headspace GCꢀMS after 5.5 h
of incubation.
In all reactions of purified AcCHMO with 2ꢀ
butanone, the amount of total product obtained at
30 °C was similar or higher to that obtained at 37 °C
(Figure 2 and Table S3). This is probably caused by
an increase in enzyme inactivation at higher
temperatures. Previous studies have shown that the
relatively low stability of AcCHMO limits its use as
Besides 2ꢀbutanone, 3ꢀmethylꢀ3ꢀbutenꢀ2ꢀone was
assayed as substrate for the purified WT AcCHMO
and its single mutants T56S and I491A. The abnormal
product resulting from the BaeyerꢀVilliger oxidation
of 3ꢀmethylꢀ3ꢀbutenꢀ2ꢀone is methyl methacrylate,
which can be directly polymerized to form the acrylic
plastic poly (methyl methacrylate). However, Figure
S2 shows that exclusively the normal product
isopropenyl acetate was detected in all reactions.
2
9
an industrial biocatalyst, and great efforts are being
2
1,30
made to address this issue.
We compared the
thermostability of WT AcCHMO and the mutants
using the ThermoFAD method.
Therefore,
AcCHMO
exhibits
a
different
31
The melting
regioselectivity with 3ꢀmethylꢀ3ꢀbutenꢀ2ꢀone and 2ꢀ
butanone. This result may due to the fact that 3ꢀ
methylꢀ3ꢀbutenꢀ2ꢀone presents a larger difference in
nucleophilic strength between the two carbons
adjacent to the substrate carbonyl group than 2ꢀ
butanone. Steric effects may also play a role in
controlling the different regioselectivity of these
reactions.
temperatures (T ) values for these enzymes at pH 7.0
m
ranged from 29 °C to 40 °C, being 37 °C for WT
AcCHMO (Table S4). The mutants T56S, I491A,
T56S/I491A and T56S/L435N/I491A, which are the
best mutants based on activity and/or regioselectivity,
had even a slightly higher T
enzyme (1–3 °C higher at pH 7.0). The effect of pH
on the T values was also investigated. In all enzymes,
the T values were slightly lower at pH 6.0 than at pH
m
value than the WT
m
m
7
.0 and 8.5 (1–3 °C). This correlates with the effect of
SteadyꢀState Kinetics. To further characterize the
reaction of WT AcCHMO and its mutants with a
ketone substrate, steadyꢀstate kinetic studies were
carried out at 25 °C and pH 7.0. Reactions were run
under atmospheric dioxygen concentration (0.253
pH on 2ꢀbutanone conversion yield.
The effect of various salt concentrations (0.1–0.8 M
KCl) on the reaction of purified WT AcCHMO with
2ꢀbutanone was also investigated, since a previous
29
4
,5
study showed that salts improve AcCHMO stability.
A negative correlation was found between KCl
concentration and conversion yields (Figure S1). The
amount of total product decreased from 46% to 16%
mM), which is saturating based on previous studies.
Table S5 shows the kcat values obtained for WT
AcCHMO and its mutants with 2ꢀbutanone as a
substrate. These values were determined by following
4
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 12
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the decrease in absorbance of NADPH at 340 nm. The hydrogen peroxide production for the WT AcCHMO
ꢀ1
ꢀ1
T56S mutant presented the highest kcat value (0.5 s ), was 0.05 s in the presence of 30 mM 2ꢀbutanone,
ꢀ1
which was 1.7ꢀfold higher than that determined for the while it was 0.1 s in the absence of a ketone
ꢀ1
WT AcCHMO (0.3 s ) (Figure 3, panel a). This is in substrate. The hydrogen peroxide produced during
agreement with the conversion data; the amount of catalysis of WT AcCHMO with 2ꢀbutanone has no
total product was doubled in the reactions using the effect on the yield of ethyl acetate and methyl
T56S mutant compared to WT AcCHMO. The other propanoate, since the results did not vary by adding
ꢀ1
mutants presented kcat values of 0.2–0.4 s (Figure catalase (1 and 50 U) to reactions with 2ꢀbutanone at
S3). The Km(2ꢀbutanone) values for WT AcCHMO and the pH 7.0 and 8.5 (Table S6).
T56S mutant were 2.4 mM and 0.7 mM, respectively.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The T56S/L244R mutant presented the same Km(2ꢀ
value as that obtained for the T56S mutant. In
butanone)
the case of the other mutants, the Km(2ꢀbutanone) value
could not be determined since the kcat values with 2ꢀ
butanone are similar to the uncoupling rates (Table S5
and Figure S3). Using cyclohexanone as a substrate,
the WT AcCHMO and the T56S mutant presented the
ꢀ1
same kcat values (6 s ) and similar K
m
values (9 and 4
ꢁ
M, respectively). Clearly, AcCHMO has evolved to
become an efficient biocatalyst to convert
cyclohexanone and the T56S mutation has only a
marginal effect on the efficiency with cyclohexanone.
The effect of pH on the uncoupling rate was studied.
For effective monooxygenaseꢀcatalyzed conversions,
uncoupling should be minimized. Figure 3 (panel b)
shows that the uncoupling rate for the WT enzyme is
6
ꢀfold lower at pH 6.0 than at pH 9.0. This
observation is in line with previous studies on WT
AcCHMO which revealed that the C4aꢀperoxyflavin
is less stable at pH 9.0 than at 7.0 in the absence of the
4
substrate. The same effect of pH on the stability of
Figure 3. Steadyꢀstate kinetics with 2ꢀbutanone (panel a)
and uncoupling rate (panel b) for WT (●/○) and T56S (■/□)
AcCHMOs. Rate values were determined at 25 °C by
following either the NADPH consumption (full symbols) or
the hydrogen peroxide production (open symbols).
Reactions contained 4 ꢁM AcCHMO fused to PTDH, 150
the C4aꢀperoxyflavin intermediate was observed fo3r2
liver microsomal flavinꢀcontaining monooxygenase,
the monooxygenase component 3 (C2) of pꢀ
3
hydroxyphenylacetateꢀ3ꢀhydroxylase,
siderophoreꢀassociated
and
a
flavinꢀcontaining
3
4
monooxygenase. Experimental and computational
studies have shown that the active site environment
controls the stability of the C4aꢀperoxyflavin
ꢁ
M NADPH, and 0.253 mM dioxygen. Buffers at pH 7.0
(panel a) or pH 6.0–9.0 (panel b) were used.
3
5
intermediate. Its decay rate is influenced by the
protonation state of active site residues or ligands,
which is modulated by pH changes. Figure 3 (panel
b) shows that the T56S mutant presents the same or
lower uncoupling rate than that for the WT enzyme at
the assayed pH values. Intriguingly, the pH had no
significant effect on the uncoupling rate for this
mutant, suggesting that its active site environment is
different to that of WT AcCHMO.
Enzymeꢀmonitored Turnover. To investigate
which reaction step limits the AcCHMO turnover rate,
enzymeꢀmonitored turnover experiments were carried
out using a stoppedꢀflow spectrophotometer (pH 7.0,
25 °C, and 0.253 mM dioxygen). In the presence of
WT enzyme (15 µM) and NADPH (200 ꢁM), a low
absorbance at 440 nm was measured during the
steadyꢀstate period (A440 = 0.04; 17% of the fully
oxidized FAD absorbance; Figure S4, panel a).
When all NADPH was consumed, as evidenced by a
low and stable absorbance at 340 nm, full enzyme
reoxidation was observed. In the enzyme reaction
containing 2ꢀbutanone (50 mM) or cyclohexanone
Flavinꢀcontaining monooxygenases produce little
or no hydrogen peroxide during catalysis with
adequate substrates. Yet, significant production of
hydrogen peroxide in the presence of nonꢀoptimal
3
4
substrates has been observed. We measured the rate
of hydrogen peroxide formation for WT AcCHMO
and the T56S mutant in the presence of various
concentrations of 2ꢀbutanone using a coupled assay
with horseradish peroxidase. Figure 3 (panel a) shows
the rates obtained using this method in comparison to
those determined by following the consumption of
NADPH. In both WT AcCHMO and the T56S mutant,
the rate of hydrogen peroxide production decreased in
the presence of 2ꢀbutanone. For example, the rate of
(
0.25 mM), together with NADPH (200 ꢁM), the
absorbance was found to be 35% and 46% of the fully
oxidized flavin absorbance, respectively (Figure S4).
These observations indicate that the overall turnover
is mainly governed by one of the steps in the catalytic
cycle occurring after enzyme reduction in the absence
and the presence of the ketone substrate (Scheme 1).
This was further investigated as described below.
5
ACS Paragon Plus Environment

Page 7 of 12
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Reductive Halfꢀreaction. The T56S and I491A AcCHMO fused to PTDH, 50 mM TrisꢀHCl, pH 7.0, at
AcCHMO mutants exhibit an improvement in activity 25 °C under anaerobic conditions. In panel a, stoppedꢀflow
and the desired regioselectivity, respectively, as traces with 62 ꢁM NADPH are shown.
compared to those for the WT enzyme. To identify the
specific reaction steps affected by the mutations, we
carried out preꢀsteady state kinetic studies using a
stoppedꢀflow spectrophotometer.
Oxidative Halfꢀreaction. The oxidativeꢀhalf
reaction for AcCHMO was investigated using the
The reductive halfꢀreaction for WT AcCHMO and stoppedꢀflow spectrophotometer in doubleꢀmixing
the T56S and I491A mutants was monitored by mode. The enzyme was first anaerobically reduced
following the decrease in absorbance at 440 nm due to with 1.2 equivalents of NADPH in the aging loop of
the reduction of enzymeꢀbound FAD under anaerobic the stoppedꢀflow apparatus. The reduction time was 5
conditions. The stoppedꢀflow traces at various s for the WT enzyme and 10 s for the mutants. The
NADPH concentrations (62–150 ꢁM) were fit with a anaerobically reduced enzyme was subsequently
double exponential function, with a fast phase mixed with airꢀsaturated buffer in the absence or the
accounting for 64%, 82%, and 90% of the absorption presence of 2ꢀbutanone.
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
55
56
57
58
59
60
change for the WT AcCHMO, T56S mutant, and
The first spectrum after mixing the anaerobically
I491A mutant, respectively (Figure 4, panel a). reduced WT AcCHMO with airꢀsaturated buffer
Previously, similar stoppedꢀflow traces were observed shows a maximum at 366 nm (Figure 5, panel a).
6
for the anaerobic reaction of PAMO with NADPH. It Subsequently, a fast increase in the absorbance at 366
was suggested that the fast phase is due to the PAMO nm was observed during 0.01 s.
A
second
reduction, while the slow phase may be due to the intermediate with a maximum at 379 nm was also
ꢀ1
formation of a small amount of C4aꢀperoxyflavin observed, with a formation rate of 36 s and decay
ꢀ1
intermediate in the presence of residual traces of rate of 0.05 s . Finally, an absorbance increase at 440
6
dioxygen. In WT AcCHMO and the mutants, the nm consistent with the hydrogen peroxide elimination
ꢀ1
observed rate for each phase (kobs1 and kobs2) was to form the reoxidized enzyme (k = 0.08 s ) was
constant at all assayed NADPH concentrations. observed. The enzyme reoxidation rate was similar to
ꢀ1
Accordingly, the Kd(NADPH) value for WT AcCHMO the uncoupling rate (0.11 s ). Similar results were
4
was previously estimated to be 7.0 µM. Both k and obtained when this experiment was carried out with
obs1
k
obs2 values for the WT AcCHMO were significantly the AcCHMOs mutants (Figure 5, panel b and c).
higher than those for the T56S and I491A mutants. However, the second intermediate of the mutants has
The calculated kred1 values for WT, T56S, and I491A an absorbance maximum at 370 nm. It was previously
ꢀ1
AcCHMOs are 169, 69, and 25 s , respectively suggested for WT AcCHMO that the first enzyme
(
Figure 4, panel b). The reduction of the FAD bound intermediate contains C4aꢀperoxyflavin, which is
4
to AcCHMO is not the main rateꢀlimiting step in the protonated
to
form
C4aꢀhydroperoxyflavin.
reaction with 2ꢀbutanone since the kred1 value is Alternatively, the protonation of an enzyme residue
significantly higher than the kcat value in all cases may influence the spectral properties of the C4aꢀ
4
(Table S5). These results are in accordance with those peroxyflavin intermediate.
described above for the enzymeꢀmonitored turnover
experiment with WT AcCHMO.
After mixing the anaerobically reduced AcCHMO
with 60 mM 2ꢀbutanone in airꢀsaturated buffer, a
rapid increase in absorbance at 366 nm due to the
formation of the C4aꢀperoxyꢀFAD intermediate was
observed (Figure 5, panel dꢀf). Subsequently, a
decrease in absorbance at 366 nm together with an
increase in absorbance at 440 nm was observed. The
stoppedꢀflow traces at 440 nm, obtained in the
presence of 2ꢀbutanone, were fit to a single
exponential function. The resulting rates for the WT,
ꢀ1
T56S, and I491A AcCHMOs were 0.3, 0.6, and 0.1 s ,
respectively (Figure S5). Similar values were
obtained by fitting the stoppedꢀflow traces at 366 nm
to obtain the rate of intermediate decay (Figure S5).
These values are also similar to the kcat values for
AcCHMO with 2ꢀbutanone (Table S5). These
experiments show that the decay of the C4aꢀperoxyꢀ
flavin intermediate to form the oxidized enzyme is
faster for the T56S AcCHMO mutant than for the WT
enzyme in the presence of 2ꢀbutanone. Before the
C4aꢀperoxyꢀflavin intermediate decay, several steps
occur in the enzyme reaction with 2ꢀbutanone
(
Scheme 1): i) 2ꢀbutanone binding; ii) formation and
rearrangement of the Criegee intermediate; iii)
formation of C4aꢀhydroxyflavin; and iv) elimination
Figure 4. Reductive halfꢀreaction of WT (●), T56S (■), and
I491A (□) AcCHMO. All reactions contained 12 ꢁM
6
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 12
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Spectral changes observed during the oxidative halfꢀreaction of WT, T56S, and I491A AcCHMOs.
Anaerobically reduced AcCHMO (1.2 equivalents NADPH/1 CHMO) was mixed with airꢀsaturated buffer with or without
2ꢀbutanone in 50 mM TrisꢀHCl, pH 7.0, at 25 °C (60 mM 2ꢀbutanone, 18/17 ꢁM CHMO fused to PTDH, final
concentrations).
3
6
of a water molecule. Additional experiments are
required to clarify which of these steps is improved in
the T56S mutant compared to the WT enzyme.
N5 atom. In the oxygenase component of pꢀ
hydroxyphenylacetateꢀ3ꢀhydroxylase,
the
C4aꢀ
hydroperoxyflavin intermediate is stabilized via a Hꢀ
bond interaction between the flavin N5 atom and the
37
S171 (equivalent to AcCHMO T56). In pyranose 2ꢀ
Importance of T56, L435, and I491 Residues in
AcCHMO. The results presented above show that the
residues located at positions 56, 435, and 491 in
AcCHMO are important for the reaction of this
enzyme with 2ꢀbutanone. In the case of WT
oxidase, a C4aꢀhydroperoxyflavin intermediate was
not detected during the oxidative halfꢀreaction when
the T169 (equivalent to AcCHMO T56) was replaced
3
8
with other residues.
Other examples of
flavoenzymes harboring a serine or threonine in a
similar position are alditol oxidase (S106; PDB ID:
AcCHMO,
a
conversion yield of 48% was
3
9
40
determined using the purified enzyme. Under the
same conditions, almost full conversion of 2ꢀ
butanone was achieved by the T56S AcCHMO
mutant. In comparison to the WT AcCHMO, the
T56S mutant enzyme has a higher turnover number
with 2ꢀbutanone due to a higher rate of flavin
reoxidation. In addition, the T56S AcCHMO mutant
has a lower uncoupling rate than the WT AcCHMO
2VFS), choline oxidase (S101; PDB ID: 4MJW),
and electronꢀtransfer flavoprotein from the
methylotrophic bacterium W3A1 (S254; PDB ID:
41
1O97). The sequenceꢀrelated PAMO harbors C65 at
the equivalent position of T56 in AcCHMO. It has
been shown that in the C65D PAMO mutant, the rate
of decay of the C4aꢀperoxyflavin intermediate is
significantly increased compared to that for the WT
4
2
(
(
Table S5). Replacement of T56 with serine reduced
PAMO. As a result, the uncoupling rate for this
ꢀ1 ꢀ1
2ꢀfold) the anaerobic rate of flavin reduction at
mutant was 5.0 s while it is only 0.02 s for the WT
42
saturating NADPH concentration. This shows that
T56 is involved in both the AcCHMO reductive and
oxidative halfꢀreactions.
PAMO. Based on available crystal structures, other
BVMOs present either a cysteine or a residue with a
nonꢀpolar side chain at the equivalent position to that
of T56 in AcCHMO.
A
multiple sequence
alignment of characterized CHMOs showed that all
other CHMOs have a serine residue at the equivalent
position to that of T56 in AcCHMO (Figure S6).
Figure 1 shows that the AcCHMO T56 is located at 5
Å from the flavin N5, on the flavin siꢀside, but the
side chain of this residue is pointing away from the
isoalloxazine ring. The same side chain position is
observed for the equivalent serine in the available
CHMOs crystal structures.
The residue adjacent to T56 in AcCHMO, D57, is
strictly conserved among all known CHMOs. Based
on available BVMO crystal structures, this is a
mobile residue that participates in the positioning of
the NADPH via a hydrogen bond interaction with the
22,43
nicotinamide base.
In AcCHMO, the replacement
of the T56 residue with serine might result in small
changes in the position of the D57 residue that
improve the reaction with 2ꢀbutanone. Alternatively,
the T56 side chain could rotate to interact with the
flavin N5 atom during the reaction. As a result of this
rotation, the T56 side chain would achieve a similar
In a similar position to that of AcCHMO T56,
most flavoenzymes have a hydrogenꢀbond donor
located within hydrogenꢀbond distance to the flavin
7
ACS Paragon Plus Environment

Page 9 of 12
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
position to that of PAMO C65 side chain, which is
regioselectivity for such relatively small substrate.
Additional engineering studies are desirable to further
improve the regioselectivity of AcCHMO with 2ꢀ
butanone with the aim of using this reaction in
industry. We anticipated that it is not an easy task and
thus the implementation of an efficient method to
located at 4.5 Å from the flavin N5 atom.
Based on the 2ꢀbutanone conversions presented
here, the I491A AcCHMO mutant presents a better
regioselectivity toward methyl propanoate than that
for the WT enzyme (Table 1). AcCHMO I491 is
located on a mobile loop formed by residues 485–502. separate methyl propanoate from ethyl acetate should
Based on available CHMOs crystal structures,
changes in the position of this loop modulate the
be also considered.
2
5
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
55
56
57
58
59
60
solvent accessibility to the substrate binding site.
Replacement of the equivalent residue in PAMO METHODS
Y502) with an alanine did not affect the kcat value
while the Km(phenylacetone) value was 20ꢀfold increased
(
2
3
and the enantioselectivity was also influenced. The
adjacent residue to AcCHMO I491, W490, is strictly
conserved among all known BVMOs. The equivalent
residue in RhCHMO, W492, interacts with the
Materials. NADPH was acquired from Jülich Chiral
Solutions GmbH. All other materials were acquired from
SigmaꢀAldrich unless otherwise specified.
+
43
NADP ribose during the catalytic cycle. In
comparison with the WT RhCHMO, the W492A
mutant presented low activity, high uncoupling rate,
Generation of Mutants. The pCRE2ꢀCHMO
2
8
construct was used as a template to generate the
AcCHMO mutants. This construct harbors a gene encoding
WT AcCHMO fused to the Cꢀterminus of PTDH. A Hisꢀ
Tag at the Nꢀterminus of the PTDH is also included in this
2
5,43
and altered selectivity.
PAMO, W501A, has a lower k , a higher
The equivalent mutant for
cat
K
m(phenylacetone )2 ,3 and a higher uncoupling rate than the
construct. Site saturation mutagenesis was carried out as
21
WT enzyme.
previously described, using NNK codons instead of
single codons. Single E. coli colonies harboring the
plasmids encoding the AcCHMO mutants were picked and
grown in 250 µL LuriaꢀBertani medium containing 50
µg/ml ampicillin (LBꢀamp) in 2 mL deep 96ꢀsquare well
plates (Waters) covered by adhesive seals (AeraSeal film,
Excel Scientific). 100 µL of this culture was mixed with 50
µL 60% (v/v) glycerol in 96ꢀwell microplates (Uꢀshape,
greiner bioꢀone) and frozen at –80 °C.
The L435N AcCHMO mutant exhibited an
improvement in the desired regioselectivity compared
to that for the WT AcCHMO (Table 1). However, the
yield of total product for the L435N mutant was 6ꢀ
fold lower than that for the WT AcCHMO. Similar
trends were observed when the T56S/L435N/I491A
mutant was compared to the T56S/I491A mutant.
Figure 1 shows that the L435N is located at ~5 Å
from the substrate, based on a superimposition of
AcCHMO homology model and RhCHMO crystal
structure. The structure of PAMO in complex with
MES shows that the equivalent residue to AcCHMO
Screening of Mutants. 250 µL LBꢀamp was added to
each well of a 2 mL deep 96ꢀsquare well plate prior to
inoculation of the E. coli cells containing the plasmids
encoding the AcCHMO mutants. Inoculation was carried
out using glycerol stocks frozen in a 96ꢀwell microplate
and a cryoꢀreplicator press (CR1100, EnzyScreen). After
24 h of incubation at 30 °C and 1,050 rpm (Titramax 1000
incubator, Heidolph), 20 µL from each culture were
transferred into the wells of a 2 mL deep 96ꢀsquare well
plate. Subsequently, 180 µL LBꢀamp containing 0.02%
2
2
L435 (M446) interacts with the ligand. The M446G
PAMO mutant had a different substrate specificity
4
4
and enantioselectivity compared to the WT PAMO.
Conclusions. E. coli cells expressing WT AcCHMO
convert 52% 2ꢀbutanone after 24 h at 24 °C resulting
in the formation of 26% methyl propanoate of the
total product formed. This ester is of industrial
interest as it can be used as a precursor of acrylic
plastic. As a result of the present study, the
T56S/I491A AcCHMO mutant was obtained. It
exhibits a significant improvement in both 2ꢀ
butanone conversion yield (73%) and the desired
regioselectivity (43% methyl propanoate) compared
to those for WT AcCHMO. These results were
confirmed using purified enzyme. Interestingly, the
T56S mutation not only improved the rate of catalysis
but also significantly lowered the uncoupling rate.
This shows that by enzyme engineering such futile
catalytic activity of BVMOs can be reduced. Several
other mutations were found to influence the
regioselectivity of AcCHMO when acting on the
smallest asymmetric aliphatic ketone, 2ꢀbutanone.
This demonstrates that subtle changes in the active
site environment (i.e., replacement of AcCHMO I491
with alanine) can influence the enzyme
(
w/v) Lꢀarabinose and 11 mM 2ꢀbutanone was added to
each well and the plates were covered with
a
polypropylene cap mat (Waters). After 24 h of incubation
at 17 °C or 24 °C and 1,050 rpm, analyses were carried out
by headspace GCꢀMS as described below.
Expression and Purification. WT AcCHMO and its
mutants, fused to PTDH and a HisꢀTag, were expressed
2
1
and purified as previously described with the following
modifications: i) Lꢀarabinose was added to the expression
medium before the inoculation; ii) after the inoculation, the
cultures were incubated 40 h at 24 °C; iii) 10 µM FAD was
not added to the lysis buffer; and iv) 2 mL of Niꢀsepharose
resin was used for each 50 mL culture. For all experiments
described in this work, purified enzyme concentration was
determined based on the extinction coefficient of WT
ꢀ
1
ꢀ1 45
AcCHMO at 440 nm (13.8 mM cm ).
Reactions of Purified AcCHMO with a Ketone
Substrate. Purified WT AcCHMO and its variants (2 ꢁM)
were reacted with 2ꢀbutanone (11 mM) at various
8
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 12
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
temperatures (4, 17, 25, 30, and 37 °C) and pH values (6.0,
.0, and 8.5). In the case of WT AcCHMO, the effect of
Enzymeꢀmonitored Turnover. Enzymeꢀmonitored
7
turnover experiments were carried out using a SX20
stoppedꢀflow spectrometer (Applied Photophysics, Surrey,
UK) in singleꢀmixing configuration. WT AcCHMO (15
µM) was reacted with NADPH (200 ꢁM) and 2ꢀbutanone
(50 mM) in 50 mM TrisꢀHCl at pH 7.0, 25 °C, and
atmospheric dioxygen concentration. The absorbance of
both the FAD bound to AcCHMO and the NADPH was
monitored at 440 and 340 nm, respectively. For
comparison, the same experiment was carried out in the
absence of 2ꢀbutanone or in the presence of cyclohexanone
(0.25 mM) instead of 2ꢀbutanone. A control reaction
without both NADPH and a ketone substrate was also
assayed.
various KCl concentrations (0.1, 0.3, 0.6, and 0.8 M) and
catalase (1 and 50 U) on the reaction with 2ꢀbutanone (11
mM) at 25 °C and pH values of 7.0 and 8.5 was also
studied. Similarly, 3ꢀmethylꢀ3ꢀbutenꢀ2ꢀone (10 mM) was
assayed as substrate for the purified WT AcCHMO and the
T56S and I491A mutants (10 ꢁM) at pH 7.0, 24 °C, and
135 rpm. All reactions contained 150 ꢁM NADPH, 20 mM
phosphite, and either 50 mM TrisꢀHCl (pH 7.0 and 8.5) or
5
0 mM sodium phosphate (pH 6.0). A final reaction
volume of 0.5 mL was prepared in a 20 mL vial
headspace, screw top, Agilent). The reaction time was 5.5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
h when 2ꢀbutanone was used as a substrate, while 24 h
incubations were carried out for the reactions containing 3ꢀ
methylꢀ3ꢀbutenꢀ2ꢀone. Reactions were stopped by
incubation at 60 °C for 10 min. Analyses were carried out
by headspace GCꢀMS as described below.
Rapid Kinetics. The catalytic cycle of WT AcCHMO
and its mutants was studied using a SX20 stoppedꢀflow
spectrometer. The singleꢀ and doubleꢀmixing mode of this
apparatus was used to investigate the enzyme reductive
and oxidative halfꢀreactions, respectively. A xenon lamp
and either a photomultiplier tube (PMT) detector or a
photodiode array (PDA) detector were used. The age of the
reaction of the mixed reagents in the observation cell is 1
millisecond (deadꢀtime) when the data acquisition starts.
All experiments were carried out in 50 mM TrisꢀHCl at pH
7.0 and 25 °C. All assays were run in duplicate or triplicate
by mixing equal volumes of reactants.
Headspace GCꢀMS Analysis. Reactions of AcCHMO
with the ketone substrates were analyzed by headspace
GCꢀMS using a GCMSꢀQP2010 (Shimadzu) in an airꢀ
conditioned room (21 °C). A HPꢀ1 column (30 m x 0.25
mm x 0.25 µm, Agilent) was used. 250 µL headspace
samples were injected. The injector temperature was set at
1
50 °C. The temperature programs used for the different
reactions are as follows: i) for the screening in 96ꢀsquare
well plates, isothermic at 35 °C for 1.7 min; ii) for the
reactions of purified enzyme with 2ꢀbutanone in vials,
isothermic at 35 °C for 3.3 min; and iii) for the reactions of
purified enzyme with 3ꢀmethylꢀ3ꢀbutenꢀ2ꢀone in vials,
isothermic at 35 °C for 5 min, and then from 35 °C to
The enzyme reduction was performed in the absence of
dioxygen. To make the stoppedꢀflow spectrometer
anaerobic, the flowꢀcircuit of this apparatus was repeatedly
washed with anaerobic buffer. Anaerobic solutions (2 or 4
mL) were prepared in gastight glass syringes (5 or 10 mL,
Hamilton, Nevada, USA). NADPH and buffer solutions
were made anaerobic by bubbling argon through the
solutions for 10 min. To deoxygenate the enzyme solution,
argon was blown on the surface of the solution for 10 min.
Aspergillus niger glucose oxidase (SigmaꢀAldrich) and
glucose were added to the solutions to remove any residual
traces of dioxygen (final concentrations 0.5 ꢁM and 2 mM,
respectively).
1
50 °C at 20 °C/min. Reactions prepared in vials were
stirred at 40 °C for 2.5 min before the analysis. Standards
were used to identify the substrates and products by
retention time and their characteristic fragmentation
pattern. Calibration curves of 2ꢀbutanone, ethyl acetate,
and methyl propanoate were carried out to quantify these
compounds in the reactions.
Determination of Melting Temperatures. The melting
temperature values for purified WT AcCHMO and various
mutants of this enzyme, fused to PTDH and a HisꢀTag,
All data were analyzed using the software ProꢀData and
ProꢀK (Applied Photophysics, Surrey, UK) or
KaleidaGraph (Synergy Software, Reading, PA). Stoppedꢀ
flow traces were fit to the corresponding exponential
function to determine the observed rates (kobs).
were determined using the ThermoFAD method as
31
previously described. For these assays, 10 ꢁM enzyme
was prepared in 50 mM TrisꢀHCl, pH 7.0 or 8.5, or 50 mM
sodium phosphate, pH 6.0.
ASSOCIATED CONTENT
SteadyꢀState Kinetics. Steadyꢀstate kinetic parameters
for purified WT AcCHMOs and its mutants were
determined using a spectrophotometer. Enzyme (4 ꢁM),
NADPH (150 ꢁM), and various 2ꢀbutanone concentrations
were reacted in 50 mM TrisꢀHCl pH 7.0, at 25 °C, and
atmospheric dioxygen concentration. In these reactions, the
rate of NADPH consumption was determined by following
Supporting Information Available: This material is
available free of charge via the Internet.
Tables S1–S6 and Figures S1–S6.
ꢀ1
ꢀ
the decrease in absorbance at 340 nm (ɛ340 = 6.22 mM cm
1
). Alternatively, the increase in absorbance at 515 nm (ε515 AUTHOR INFORMATION
ꢀ1
ꢀ1
= 26 mM cm ) was measured to determine the rate of
hydrogen peroxide production in the presence of
horseradish peroxidase (4 U), 4ꢀaminoantipyrine (0.1 mM),
and 3,5ꢀdichloroꢀ2ꢀhydroxybenzenesulfonic acid (1.0 mM).
The NADPH consumption assay was also carried out using
cyclohexanone instead of 2ꢀbutanone as a substrate and
Corresponding Author
*Eꢀmail: m.w.fraaije@rug.nl
Notes
ϕ
Current address: Forschungszentrum Jülich GmbH, Jülich,
Germany
These authors contributed equally to this work.
The authors declare no competing financial interest.
0
.01–0.05 ꢁM AcCHMO. The initial rates of the reactions
were plotted as a function of the ketone substrate
concentrations. These data were fit to the Michaelisꢀ
Menten equation using the software KaleidaGraph
(Synergy Software, Reading, PA).
§
9
ACS Paragon Plus Environment

Page 11 of 12
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
ACKNOWLEDGEMENTS
Villiger monooxygenases for the formation of
amino acids and ‐amino alcohols, Angew. Chem.
Int. Ed. 49, 4506–4508.
14) Fraaije, M. W., Kamerbeek, N. M., Heidekamp, A. J.,
Fortin, R., and Janssen, D. B. (2004) The prodrug
activator EtaA from Mycobacterium tuberculosis is a
BaeyerꢀVilliger monooxygenase, J Biol Chem 279,
β‐
β
This work has been supported by the Methyl Esters from Biomass
grant (MEBIO; 053.24.105) from the Netherlands Organisation
for Scientific Research (NWO).
(
(
REFERENCES
3
354–3360.
15) van Beek, H. L., Winter, R. T., Eastham, G. R., and
Fraaije, M. W. (2014) Synthesis of methyl propanoate
by BaeyerꢀVilliger monooxygenases, Chem Commun
50, 13034–13036.
(
1) Leisch, H., Morley, K., and Lau, P. C. K. (2011)
BaeyerꢀVilliger monooxygenases: more than just green
chemistry, Chem. Rev. 111, 4165–4222.
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
55
56
57
58
59
60
(2) de Gonzalo, G., Mihovilovic, M. D., and Fraaije, M. W.
(16) Harris, B. (2010) Acrylics for the future, Ingenia 45,
(2010) Recent developments in the application of
18–23.
BaeyerꢀVilliger monooxygenases as biocatalysts,
ChemBioChem 11, 2208–2231.
(
17) Yoneda, H., Tantillo, D. J., and Atsumi, S. (2014)
Biological production of 2‐butanone in Escherichia
coli, ChemSusChem 7, 92–95.
(18) Srirangan, K., Liu, X., Akawi, L., Bruder, M., Mooꢀ
Young, M., and Chou, C. P. (2016) Engineering
Escherichia coli for microbial production of butanone,
Appl Environ Microbiol 82, 2574–2584.
(19) Janssen, A. J. M., and Zwanenburg, B. (1991) PPLꢀ
catalyzed resolution of 1, 2ꢀand 1, 3ꢀdiols in methyl
propionate as solvent. An application of the tandem use
of enzymes, Tetrahedron 47, 7409–7416.
(20) Poucher, W. A. (1991) Poucher’s perfumes, cosmetics
and soaps, Vol. 1, 9 ed., Chapman and Hall, London.
(21) van Beek, H. L., Wijma, H. J., Fromont, L., Janssen,
D. B., and Fraaije, M. W. (2014) Stabilization of
cyclohexanone monooxygenase by a computationally
designed disulfide bond spanning only one residue,
FEBS Open Bio 4, 168–174.
(3) Reignier, T., de Berardinis, V., Petit, J. L., Mariage, A.,
Hamzé, K., Duquesne, K., and Alphand, V. (2014)
Broadening
the
scope
of
BaeyerꢀVilliger
monooxygenase activities toward α, βꢀunsaturated
ketones: a promising route to chiral enolꢀlactones and
eneꢀlactones, ChemComm 50, 7793–7796.
(
4) Sheng, D., Ballou, D. P., and Massey, V. (2001)
Mechanistic studies of cyclohexanone monooxygenase:
chemical properties of intermediates involved in
catalysis, Biochemistry 40, 11156–11167.
(
5) Ryerson, C. C., Ballou, D. P., and Walsh, C. (1982)
Mechanistic studies on cyclohexanone oxygenase,
Biochemistry 21, 2644–2655.
(6) Torres Pazmiño, D. E., Baas, B. J., Janssen, D. B., and
Fraaije, M. W. (2008) Kinetic mechanism of
phenylacetone monooxygenase from Thermobifida
fusca, Biochemistry 47, 4082–4093.
(
(
(
7) Kelly, D. R., Knowles, C. J., Mahdi, J. G., Taylor, I. N., (22) Orru, R., Dudek, H. M., Martinoli, C., Torres Pazmiño,
and Wright, M. A. (1995) Mapping of the functional
active site of BaeyerꢀVilligerases by substrate
engineering, J Chem Soc, Chem Commun 7, 729–730.
8) Polyak, I., Reetz, M. T., and Thiel, W. (2012) Quantum
mechanical/molecular mechanical study on the
mechanism of the enzymatic BaeyerꢀVilliger reaction, J
Am Chem Soc 134, 2732–2741.
9) Rioz‐Martínez, A., de Gonzalo, G., Torres Pazmiño,
D. E., Fraaije, M. W., and Gotor, V. (2009) Enzymatic
baeyerꢀvilliger oxidation of benzo ‐ fused ketones:
formation of regiocomplementary lactones, Eur J Org
Chem 15, 2526–2532.
D. E., Royant, A., Weik, M., Fraaije, M. W., and
Mattevi, A. (2011) Snapshots of enzymatic Baeyerꢀ
Villiger catalysis. Oxygen activation and intermediate
stabilization, J. Biol. Chem. 286, 29284–29291.
(23) Dudek, H. M., Fink, M. J., Shivange, A. V., Dennig,
A., Mihovilovic, M. D., Schwaneberg, U., and Fraaije,
M. W. (2014) Extending the substrate scope of a
BaeyerꢀVilliger monooxygenase by multipleꢀsite
mutagenesis, Appl. Microbiol. Biotechnol. 98, 4009–
4020.
24) Reetz, M. T., and Wu, S. (2009) Laboratory evolution
of robust and enantioselective BaeyerꢀVilliger
monooxygenases for asymmetric catalysis, J Am Chem
Soc 131, 15424–15432.
25) Yachnin, B. J., McEvoy, M. B., MacCuish, R. J. D.,
Morley, K. L., Lau, P. C. K., and Berghuis, A. M.
(
(10) Fink, M. J., Snajdrova, R., Winninger, A., and
Mihovilovic, M. D. (2016) Regioꢀand stereoselective
synthesis of chiral nitrilolactones using BaeyerꢀVilliger
monooxygenases, Tetrahedron 72, 7241–7248.
(
(
2014) Lactoneꢀbound structures of cyclohexanone
monooxygenase provide insight into the
stereochemistry of catalysis, ACS Chem. Biol. 9, 2843–
851.
(11) Balke, K., Schmidt, S., Genz, M., and Bornscheuer, U.
T. (2016) Switching the regioselectivity of
a
cyclohexanone monooxygenase toward (+)ꢀtransꢀ
dihydrocarvone by rational protein design, ACS Chem
Biol 11, 38–43.
2
(26) Reetz, M. T., Kahakeaw, D., and Lohmer, R. (2008)
Addressing the numbers problem in directed evolution,
ChemBioChem 9, 1797–1804.
(27) Torres Pazmiño, D. E., Snajdrova, R., Baas, B. J.,
Ghobrial, M., Mihovilovic, M. D., and Fraaije, M. W.
(
12) Rehdorf, J., Lengar, A., Bornscheuer, U. T., and
Mihovilovic, M. D. (2009) Kinetic resolution of
aliphatic acyclic βꢀhydroxyketones by recombinant
wholeꢀcell BaeyerꢀVilliger monooxygenasesꢀFormation
of enantiocomplementary regioisomeric esters, Bioorg
Med Chem Lett 19, 3739–3743.
(
2008)
Self
‐
sufficient
baeyerꢀvilliger
monooxygenases: effective coenzyme regeneration for
biooxygenation by fusion engineering, Angew Chem
(13) Rehdorf, J., Mihovilovic, M. D., and Bornscheuer, U.
1
20, 2307–2310.
T. (2010) Exploiting the regioselectivity of Baeyerꢀ
1
0
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 12
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
28) Torres Pazmiño, D. E., Riebel, A., de Lange, J.,
Rudroff, F., Mihovilovic, M. D., and Fraaije, M. W.
2009) Efficient biooxidations catalyzed by a new
generation of self ‐ sufficient BaeyerꢀVilliger
monooxygenases, ChemBioChem 10, 2595–2598.
29) Zambianchi, F., Pasta, P., Carrea, G., Colonna, S.,
Gaggero, N., and Woodley, J. M. (2002) Use of
(41) Yang, K.ꢀY., and Swenson, R. P. (2007) Modulation
of the redox properties of the flavin cofactor through
hydrogenꢀbonding interactions with the N (5) atom:
role of αSer254 in the electronꢀtransfer flavoprotein
(
from
the
methylotrophic
bacterium
W3A1,
(
Biochemistry 46, 2289–2297.
(42) Brondani, P. B., Dudek, H. M., Martinoli, C., Mattevi,
A., and Fraaije, M. W. (2014) Finding the switch:
isolated
cyclohexanone
monooxygenase
from
recombinant Escherichia coli as a biocatalyst for
BaeyerꢀVilliger and sulfide oxidations, Biotechnol.
Bioeng. 78, 489–496.
turning
a baeyerꢀvilliger monooxygenase into a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NADPH oxidase, J Am Chem Soc 136, 16966–16969.
(43) Mirza, I. A., Yachnin, B. J., Wang, S., Grosse, S.,
Bergeron, H., Imura, A., Iwaki, H., Hasegawa, Y., Lau,
P. C. K., and Berghuis, A. M. (2009) Crystal structures
of cyclohexanone monooxygenase reveal complex
domain movements and a sliding cofactor, J Am Chem
Soc 131, 8848–8854.
(44) Pazmiño, D. E. T., Snajdrova, R., Rial, D. V.,
Mihovilovic, M. D., and Fraaije, M. W. (2007) Altering
the substrate specificity and enantioselectivity of
phenylacetone monooxygenase by structure‐inspired
enzyme redesign, Adv Synth Catal 349, 1361–1368.
(45) Sheng, D., Ballou, D. P., and Massey, V. ( 2001)
Mechanistic studies of cyclohexanone monooxygenase:
chemical properties of intermediates involved in
catalysis Biochemistry 40, 11156–11167.
(30) Schmidt, S., Genz, M., Balke, K., and Bornscheuer, U.
T. (2015) The effect of disulfide bond introduction and
related Cys/Ser mutations on the stability of a
cyclohexanone monooxygenase, J. Biotechnol. 214,
1
99–211.
(
31) Forneris, F., Orru, R., Bonivento, D., Chiarelli, L. R.,
and Mattevi, A. (2009) ThermoFAD, a Thermofluor®ꢀ
adapted flavin ad hoc detection system for protein
folding and ligand binding, FEBS J. 276, 2833–2840.
32) Jones, K. C., and Ballou, D. P. (1986) Reactions of
the 4aꢀhydroperoxide of liver microsomal flavinꢀ
containing monooxygenase with nucleophilic and
electrophilic substrates, J Biol Chem 261, 2553–2559.
(
(33) Ruangchan, N., Tongsook, C., Sucharitakul, J., and
Chaiyen, P. (2011) pHꢀdependent studies reveal an
efficient hydroxylation mechanism of the oxygenase
component of pꢀhydroxyphenylacetate 3ꢀhydroxylase, J
Biol Chem 286, 223–233.
(34) Frederick, R. E., Ojha, S., Lamb, A., and DuBois, J. L.
2014) How pH modulates the reactivity and selectivity
(
of a siderophoreꢀassociated flavin monooxygenase,
Biochemistry 53, 2007–2016.
(35) Bach, R. D., and Mattevi, A. (2013) Mechanistic
aspects regarding the elimination of H
2 2
O from C (4a)ꢀ
hydroperoxyflavin. The role of a proton shuttle
required for H
8593.
2 2
O elimination, J Org Chem 78, 8585–
(36) Fraaije, M. W., and Mattevi, A. (2000) Flavoenzymes:
diverse catalysts with recurrent features, TIBS 25, 126–
1
32.
(
37) Thotsaporn, K., Chenprakhon, P., Sucharitakul, J.,
Mattevi, A., and Chaiyen, P. (2011) Stabilization of
C4aꢀhydroperoxyflavin in a twoꢀcomponent flavinꢀ
dependent monooxygenase is achieved through
interactions at flavin N5 and C4a atoms, J Biol Chem
2
86, 28170–28180.
(38) Pitsawong, W., Sucharitakul, J., Prongjit, M., Tan, T.ꢀ
C., Spadiut, O., Haltrich, D., Divne, C., and Chaiyen, P.
(2010) A conserved activeꢀsite threonine is important
for both sugar and flavin oxidations of pyranose 2ꢀ
oxidase, J. Biol. Chem. 285, 9697–9705.
(39) Forneris, F., Heuts, D. P. H. M., Delvecchio, M.,
Rovida, S., Fraaije, M. W., and Mattevi, A. (2008)
Structural analysis of the catalytic mechanism and
stereoselectivity in Streptomyces coelicolor alditol
oxidase, Biochemistry 47, 978–985.
(
40) Yuan, H., and Gadda, G. (2011) Importance of a
serine proximal to the C(4a) and N(5) flavin atoms for
hydride transfer in choline oxidase, Biochemistry 50,
7
70–779.
1
1
ACS Paragon Plus Environment