Structural and catalytic characterization of a fungal baeyer-villiger monooxygenase

Article abstract of DOI:10.1371/journal.pone.0160186

Baeyer-Villiger monooxygenases (BVMOs) are biocatalysts that convert ketones to esters. Due to their high regio-, stereo- and enantioselectivity and ability to catalyse these reactions under mild conditions, they have gained interest as alternatives to chemical Baeyer-Villiger catalysts. Despite their widespread occurrence within the fungal kingdom, most of the currently characterized BVMOs are from bacterial origin. Here we report the catalytic and structural characterization of BVMO_AFL838 from Aspergillus flavus. BVMO_AFL838 converts linear and aryl ketones with high regioselectivity. Steady-state kinetics revealed BVMO_AFL838 to show significant substrate inhibition with phenylacetone, which was more pronounced at low pH, enzyme and buffer concentrations. Para substitutions on the phenyl group significantly improved substrate affinity and increased turnover frequencies. Steady-state kinetics revealed BVMO_AFL838 to preferentially oxidize aliphatic ketones and aryl ketones when the phenyl group are separated by at least two carbons from the carbonyl group. The X-ray crystal structure, the first of a fungal BVMO, was determined at 1.9 A and revealed the typical overall fold seen in type I bacterial BVMOs. The active site Arg and Asp are conserved, with the Arg found in the ginh position. Similar to phenylacetone monooxygenase (PAMO), a two residue insert relative to cyclohexanone monooxygenase (CHMO) forms a bulge within the active site. Approximately half of the gvariableh loop is folded into a short ?-helix and covers part of the active site entry channel in the non-NADPH bound structure. This study adds to the current efforts to rationalize the substrate scope of BVMOs through comparative catalytic and structural investigation of different BVMOs.

Full text of DOI:10.1371/journal.pone.0160186

RESEARCH ARTICLE
Structural and Catalytic Characterization of a
Fungal Baeyer-Villiger Monooxygenase
Felix Martin Ferroni, Carmien Tolmie, Martha Sophia Smit, Diederik Johannes Opperman*
Department of Biotechnology, University of the Free State, Bloemfontein, South Africa
* opperdj@ufs.ac.za
Abstract
Baeyer-Villiger monooxygenases (BVMOs) are biocatalysts that convert ketones to esters.
Due to their high regio-, stereo- and enantioselectivity and ability to catalyse these reactions
under mild conditions, they have gained interest as alternatives to chemical Baeyer-Villiger
catalysts. Despite their widespread occurrence within the fungal kingdom, most of the cur-
rently characterized BVMOs are from bacterial origin. Here we report the catalytic and struc-
tural characterization of BVMOAFL838 from Aspergillus flavus. BVMOAFL838 converts linear
and aryl ketones with high regioselectivity. Steady-state kinetics revealed BVMOAFL838 to
show significant substrate inhibition with phenylacetone, which was more pronounced at
low pH, enzyme and buffer concentrations. Para substitutions on the phenyl group signifi-
cantly improved substrate affinity and increased turnover frequencies. Steady-state kinetics
revealed BVMOAFL838 to preferentially oxidize aliphatic ketones and aryl ketones when the
phenyl group are separated by at least two carbons from the carbonyl group. The X-ray
crystal structure, the first of a fungal BVMO, was determined at 1.9 Å and revealed the typi-
cal overall fold seen in type I bacterial BVMOs. The active site Arg and Asp are conserved,
with the Arg found in the “in” position. Similar to phenylacetone monooxygenase (PAMO), a
two residue insert relative to cyclohexanone monooxygenase (CHMO) forms a bulge within
the active site. Approximately half of the “variable” loop is folded into a short α-helix and cov-
ers part of the active site entry channel in the non-NADPH bound structure. This study adds
to the current efforts to rationalize the substrate scope of BVMOs through comparative cata-
lytic and structural investigation of different BVMOs.
a11111
OPEN ACCESS
Citation: Ferroni FM, Tolmie C, Smit MS, Opperman
DJ (2016) Structural and Catalytic Characterization of
a Fungal Baeyer-Villiger Monooxygenase. PLoS ONE
11(7): e0160186. doi:10.1371/journal.pone.0160186
Editor: Petri Kursula, Universitetet i Bergen,
NORWAY
Received: April 8, 2016
Accepted: July 14, 2016
Published: July 29, 2016
Copyright: © 2016 Ferroni et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited.
Data Availability Statement: All coordinate and
structure factor files are available from the Protein
Data Bank (PDB) database (accession number
5J7X).
Introduction
Baeyer-Villiger monooxygenases (BVMOs) are flavin-dependent enzymes that catalyze the
conversion of ketones to esters using NAD(P)H and molecular oxygen [1–4]. In addition to
this typical reaction, they can also catalyze heteroatom oxidation, including sulfoxidation and
N-oxidation, as well as epoxidation reactions. The substrate scope of the collective BVMO fam-
ily of enzymes has grown to include a variety of substrates ranging from acetone to larger
ketones such as steroids. The mild reaction conditions and often high regio-, stereo- and enan-
tioselectivity have made them very attractive as an alternative to chemical Baeyer-Villliger cata-
lysts. Indeed, many directed evolution studies have been performed to increase or alter the
substrate scope as well as improve the selectivity and specificity of these enzymes [5,6].
Funding: Support was provided by DJO—National
Research Foundation, South Africa (NRF Grant No
87889). The funder had no role in study design, data
collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: The authors have declared
that no competing interests exist.
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
1 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
Although the available cloned BVMOs have grown significantly over the past few years, it is
only recently that BVMOs from fungal sources have been explored [7,8] despite their wide-
spread presence in the fungal-kingdom as revealed through whole-genome sequencing [9]. To
date however, the three-dimensional crystal structures of only four distinct bacterial type I Bae-
yer-Villiger monooxygenases have been determined: phenylacetone monooxygenase (PAMO)
from Thermobifida fusca [10], cyclohexanone monooxygenase (CHMO) from Rhodococcus sp.
strain HI-31 [11], steroid monooxygenase (STMO) from Rhodococcus rhodochrous [12] and
2-oxo-Δ3–4,5,5-trimethylcyclopentenylacetyl-coenzyme A monooxygenase (OTEMO) from
Pseudomonas putida ATCC 17453 [13]. Through extensive structural investigations of these
enzymes with bound co-factors, inhibitors, substrates and products, the reaction mechanism of
BVMOs has been explained [14–16]. Catalysis of BVMOs involves extensive backbone confor-
mational changes and cofactor movement. In short, NADPH is bound to the BVMO in the
“open” conformation, where after the non-covalently bound FAD is reduced and subsequently
reacts with molecular oxygen to form the reactive peroxyflavin species. Following substrate
entry, the BVMO undergoes a domain rotation and movement of the NADP+ to stabilize the
peroxyflavin. This is accompanied/mediated by the structuring of a disordered surface loop.
The BVMO, now in a “closed/tight” conformation, reorganizes to the “rotated” conformation
through the rotation of the NADP+ to allow the substrate to move into the catalytic position.
Nucleophilic attack with formation of the Criegee intermediate occurs in this “rotated” confor-
mation. Following the production of the lactone product, the BVMO returns to a “closed/
tight”-like NADP+ conformation followed by release of the product in the “loose”
conformation.
Despite these informative studies, the basis of substrate acceptance and specificity, especially
of larger substrates, is still not completely understood. PAMO has a rather limited substrate
scope of mostly phenyl substituted linear ketones [17] while STMO can only convert both pro-
gesterone and phenylacetone [12,18]. In contrast CHMO has an extremely wide substrate
scope [3]. We have recently reported on four closely related BVMOs from Aspergillus flavus
with distinct substrate profiles [7]. Amongst the four BVMOs described, BVMOAFL838 showed
the best conversion of alkanones with chain lengths of C8-C12, but was unable to convert most
of the cyclic ketones tested. Here we describe the catalytic and structural characterization of
BVMOAFL838. This structure represents the first fungal BVMO solved and contributes to the
efforts to rationalize the substrate specificity of BVMOs.
Materials and Methods
Strains and Vectors
BVMOAFL838 was heterologously expressed from the pET-22b(+) vector (Novagen) in E. coli
BL21Gold(DE3) (Stratagene). The previously constructed plasmid [7] served as a template to
construct a C-terminally histidine (CTH) tagged variant of BVMOAFL838 by deleting the stop
codon and plasmid backbone between the gene and the plasmid’s six histidine’s codons. This
variant was prepared as previously described [19] using the primers BVMO_Histag_F (5’ CAC
CAC CAC CAC CAC CAC TGA GAT C 3’) and BVMO_Histag_R (5’ TGC TTT CGC AAA
ACC AAA GAA ATC CTC 3’). Bacillus megaterium glucose dehydrogenase (BmGDH) was
kindly provided by Dr. Dirk Holtmann (Dechema, Germany) in the pETDuet vector (Nova-
gen) cloned in the second multiple cloning site via NdeI and XhoI. The BmGDH ORF was sub-
sequently sub-cloned to pET-28b(+) to generate an N-terminally histidine tagged variant using
the same restriction sites. The BVMOAFL838_K511A mutant was prepared using site-directed
mutagenesis with the QuickChange (Stratagene) method using primers K511A_F (5’ CAA
CAT TCC GGG CGC GCC TGT TCA ATC ATT G 3’) and a complementary reverse primer.
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
2 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
E. coli strains were routinely grown in LB medium containing ampicillin (0.1 mg.mL-1) or
kanamycin (0.03 mg.mL-1) at 37°C with shaking (200 rpm).
Protein Expression and Purification
E. coli BL21Gold(DE3) harbouring the pET-22:BVMOAFL838CTH or pET-28:BmGDH plasmid
was inoculated into ZYP5052 expression medium [20] containing 0.1 mg.mL-1 ampicillin or
0.03 mg.mL-1 kanamycin respectively. Cells were grown for 24–48 h at 20°C where after they
were harvested through centrifugation (8 000 xg, 10 min) and resuspended in 50 mM Tris-HCl
(pH 7.4) buffer containing 0.5 M NaCl and 20–30 mM imidazole (binding buffer). Cells were
broken by a single passage through a continuous cell disrupter (Constant Systems) using 30
kPsi at 4°C and the crude cell-free extract was obtained through ultracentrifugation (100 000
xg, 90 min). The soluble fraction was loaded onto a 5 mL HisTrap FF Ni-affinity column (GE
Healthcare) equilibrated in binding buffer. Unbound proteins were removed by washing with
10 column volumes of the same buffer. Proteins were eluted in the same buffer using a linear
gradient of increasing imidazole concentration. Fractions containing the protein of interest
were pooled and concentrated to approximately 2 mL through ultrafiltration using a 30 kDa
NMWL Amicon Ultrafiltration unit. The concentrated BVMOAFL838CTH protein was soaked
overnight in excess FAD. Concentrated protein samples were loaded onto a Sephacryl S100HR
size exclusion column (GE Healthcare). Proteins were eluted in 10 mM Tris-HCl buffer (pH 8)
containing 100 mM NaCl. Purified proteins were evaluated on SDS-PAGE using PageRuler
Prestained Protein Ladder (ThermoScientific) and stained with Coomassie brilliant blue R-
250. Protein concentrations were determined using the Bradford protein assay (Bio-Rad). FAD
content was determined by incubation of the protein sample (typically 5 mg.mL-1) in ~ 8 M
urea for 30 min, followed by spectrophotometric quantification using the extinction coefficient
of 11.3 mM-1.cm-1 at 450 nm.
Biotransformations and Steady-state kinetics
Biotransformations were performed in amber glass vials (40 mL) in a total reaction volume
of 1 mL. Whole-cell (WC) and cell-free extract (CFE) biotransformations were performed as
previously described [7]. Reactions with purified BVMO were performed in 100 mM Tris-
HCl buffer (pH 8) containing 2 μM BVMO, 0.5 U BmGDH, 100 mM glucose, 0.3 mM
NADP+ and 10 mM substrate. Reactions were maintained at 20°C with shaking (200 rpm),
where after they were extracted using an equal volume (2 x 0.5 mL) ethyl acetate containing
either 2 mM 1-undecanol or 3-octanol as internal standard. GC-FID (and GC-MS for
product identification) was performed on a Finnigan Trace GC ultra (ThermoScientific)
equipped with a FactorFour VF-5ms column (60 m x 0.32 mm x 0.25 μm, Varian). Steady-
state kinetics were performed by monitoring the oxidation of NADPH spectrophotometeri-
cally at 340 nm (ε340 = 6.22 mM-1.cm-1) or 370 nm (ε370 = 2.70 mM-1.cm-1). To investigate
optimal pH, temperature, stability and effect of organic solvents, reactions typically con-
tained 2 μM BVMO, 0.3 mM NADPH, 1 mM phenylacetone, 1% (v/v) methanol (100 mM
Tris-HCl, pH 8; 25°C).
Crystallization and Structure Determination
Crystals were grown using hanging-drop vapour-diffusion in 1 μL drops consisting of equal
volumes of 6 mg.mL-1 BVMO and reservoir solution (0.1 M Tris-HCl pH 9, 1.8 M ammonium
sulphate) and 5 x molar excess of FAD and NADP+ at 289 K. Yellow crystals grew within 2
weeks. Crystals were soaked in reservoir solution containing 30% glycerol prior to cryocooling.
X-ray diffraction data (Table 1) were collected at Diamond (UK) on beamline I04-I. Data was
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
3 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
Table 1. Data Collection and Refinement Statistics.
Data Collection and Processing
Beamline
I04-I
Wavelength (Å)
Resolution (Å)
Space group
Unit cell parameters
Unique reflections
Completeness (%)
Mn(I)/sd(I)
0.9174
32.4–1.9 (2.0–1.9)
C 2 2 2
a = 109.67Å b = 177.51Å c = 76.37Å α = 90° β = 90° γ = 90°
58733 (8465)
99.6 (99.6)
15.3 (3.0)
Multiplicity
6.2 (6.3)
Rmerge
0.081 (0.669)
Refinement
a
Rwork/Rfree
0.1571/0.2046
Molecules in ASU
1
Average B-factor, all atoms (Å2)
r.m.s.d. Bond lengths (Å)
r.m.s.d. Bond angles (°)
28.0
0.031
2.71
Ramachandran distribution (%)b
96.9/2.7/0.4
Values for the highest resolution shell given in parenthesis
a Rfree calculated from 5% reflections omitted from structure refinement
b Favoured/Permitted/Disallowed
doi:10.1371/journal.pone.0160186.t001
processed using MOSFLM [21] and POINTLESS [22], with intensities scaled and merged
using SCALA [22] from the CCP4 suite of programs. Molecular replacement was performed
using Phaser [23] with PAMO (PDB:1W4X) as search model. Refinement was performed
through iterative cycles of manual model building in COOT [24] and TLS and restrained
refinement using Refmac [25]. Structures were validated using programs within the CCP4 suite
[26]. Figures were generated in PyMOL.
Accession numbers
Coordinates and structure factors for BVMOAFL838 have been deposited in the Protein Data
Bank (PDB) under accession number 5J7X.
Results and Discussion
Purification and Characterization
BVMOAFL838 was heterologously expressed in E. coli as a C-terminally 6x histidine tagged pro-
tein and purified to near homogeneity using Ni-affinity and size exclusion chromatography
(SEC). SDS-PAGE analysis confirmed the monomer molecular weight of 62.2 kDa (Fig 1A).
SEC also showed BVMOAFL838 to exist as a monomer in solution. BVMOAFL838 was recovered
with only 70–80% of bound flavin, but near full occupancy of the FAD could be achieved by
soaking with excess FAD before SEC (Fig 1B). BVMOAFL838 was optimally active at pH 9, but
retained approximately 80% activity at pH 8–8.5 and 60% at pH 7.5 and 9.5. Although the opti-
mal temperature for BVMOAFL838 was found to be 40°C, the enzyme was rapidly inactivated at
40°C, with a half-life of only 21 min (Fig 2). Of the organic solvents tested, methanol showed to
least affect activity, with nearly a 100% relative activity in up to 5% (v/v).
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
4 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
Fig 1. (A) SDS-PAGE analysis of purified BVMOAFL838. Lane M: molecular weight marker, Lane 1: E. coli
cell-free extract expressing BVMOAFL838, Lane 2: Purified BVMOAFL838. (B) Absorbance spectra of
BVMOAFL838 before and after treatment with 8 M urea. Extinction coefficient of free FAD at 450 nm: 11.3
mM-1.cm-1, Enzyme bound FAD at 436 nm: 12.7 mM-1.cm-1 and 11.6 mM-1.cm-1 at 450 nm.
doi:10.1371/journal.pone.0160186.g001
Substrate specificity
We have previously shown BVMOAFL838 to be active towards aliphatic ketones and aryl
ketones, with very few monocyclic ketones converted during whole-cell biotransformations
[7]. Purified enzyme was tested using glucose dehydrogenase from Bacillus megaterium
(BmGDH) for cofactor regeneration to eliminate the possible uptake/transport limitations or
toxicity of substrates or products toward E. coli whole cells. Similar to the whole-cell biotrans-
formations which we previously reported, efficient conversion of the 2-alkanones was observed
with nearly complete conversion of the C8-C12 substrates (10 mM of 1a, 3-5a) within 8 h and
even within 2 h for C8 (1a), with absolute regioselectivity forming the alkyl acetate products
(Fig 3, Table 2). Turnover frequency (TOF) decreased with increasing chain-length but this
could possibly be attributed to a decrease in solubility. Complete conversion of 3-octanone
(2a) was also observed after 2 h, and observed TOF (kobs) values for 10 mM 2-octanone (1a)
and 3-octanone were 5.3 s-1 and 6.6 s-1 respectively. To better investigate the effect of the posi-
tion of the ketone group, steady-state kinetics were performed (Fig 4). A much lower KM was
observed for 2-octanone compared to 3-octanone, and although both substrates suffered from
substrate inhibition, 2-octanone’s inhibition was more pronounced (Ki = 1.5 0.4 mM) leading
to lower TOF at high (10 mM) substrate concentrations (Table 2). Conversions of aliphatic
Fig 2. Temperature stability of BVMOAFL838
.
doi:10.1371/journal.pone.0160186.g002
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
5 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
Fig 3. Baeyer-Villiger oxidation of aliphatic ketones.
doi:10.1371/journal.pone.0160186.g003
ketones with substrate concentrations of up to 25 mM have been reported for BVMO3 from
Dietzia sp. D5 [27], however no kinetic data was reported to evaluate reaction rates or possible
substrate inhibition.
Investigation into BVMOAFL838’s affinity for longer chain 2-alkanones (3-5a) revealed Km
values lower than 0.01 mM and less substrate inhibition than with 2-octanone with TOF reach-
ing 6.8 s-1 (6.6 U.mg-1) for 2-decanone at 10 mM. Wild-type PAMO has been shown to accept
aliphatic ketones [17] such as 2-dodecanone with a turnover frequency of 0.23 s-1 (Km = 0.26
mM) and a quadruple mutant (15-F5) with improved activity towards 2-octanone (kcat 2.3 s-1,
Km = 0.25 mM) has also been reported [28]. BVMOs from Pseudomonas putida KT2440 [29]
and P. veronii MEK700 (MekA) [30] have also been shown to efficiently convert various ali-
phatic ketones, with specific activities for MekA between 0.45–0.89 U.mg-1 for C8-C12
2-alkanones.
A good correlation was observed between whole-cell biotransformations and biotransfor-
mations using purified enzymes for aryl ketones, except no conversion was observed for pheny-
lacetone (8a) and benzaldehyde (12a) using purified enzyme (Fig 5, Table 3). Very low TOF
(0.3 s-1) were observed against 10 mM phenylacetone and steady-state kinetics revealed severe
substrate inhibition with phenylacetone especially at lower enzyme concentrations (Fig 6A).
Incubation of BVMOAFL838 with excess phenylacetone followed by reduction with NADPH
did not affect the initial FAD reduction (NADPH binding), suggesting inhibition occurs during
one of the intermediate steps of catalysis. Substrate inhibition by some acetophenone and benz-
aldehyde derivatives have also been observed with HAPMO [31] with a strong correlation
between substrate affinity (Km) and degree of substrate inhibition. Substrate inhibition of
BVMOAFL838 was also more pronounced at lower buffer concentrations and pH values (Fig 6B
and 6C) to an extent that no observable rate could be detected. Glucose dehydrogenase (GDH)
was used for cofactor regeneration in the biotransformations with purified enzyme and GDH
is known to rapidly acidify the reaction mixture [32]. This type of fast acidification is not found
with whole-cell biotransformation using glucose for cofactor regeneration through central
metabolism. In addition, higher enzyme concentrations in whole-cells and potentially reduced
effective substrate concentration due to diffusion/uptake limitations to the cytoplasm may also
alleviate the substrate inhibition observed when using purified enzyme. The initial rates (mea-
sured as NADPH oxidation) were not maintained even at high substrate concentrations, sug-
gesting also product inhibition.
Table 2. Biotransformations of aliphatic ketones (10 mM) by purified BVMOAFL838 (2 μM) and kinetic parameters determined from initial velocities
of NADPH oxidation.
Substrate
Conversion (%)
TOF (s-1)
10 mM kobs
5.3
Km (mM)
2h
8h
1a
2a
3a
4a
5a
2-octanone
>99
>99
87
>99
>99
>99
98
0.01 0.003
0.17 0.03
<0.01
3-octanone
6.6
2-decanone
2-undecanone
2-dodecanone
6.8
42
6.1
<0.01
21
99
5.3
<0.01
doi:10.1371/journal.pone.0160186.t002
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
6 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
Fig 4. Steady-state kinetics of BVMOAFL838 with octanone.
doi:10.1371/journal.pone.0160186.g004
Similarly, no conversion was observed with benzaldehyde (12a) as substrate using purified
enzyme although more than 60% conversion was observed using whole-cell biotransforma-
tions. Steady-state kinetics gave very low TOFs but no substrate inhibition was observed up to
a concentration of 30 mM. Only slight substrate inhibition (Ki = 41 6 mM) was observed
with the para-substituted 4-methoxy benzaldehyde (13a) with more than 80% conversion
observed after 2 h. Likewise, acetophenone with a hydroxyl substitution on the para position
(7a) showed complete conversion after only 2 h, compared to 4% for acetophenone (6a). The
substitutions on the phenyl ring lowered the Km values with more than an order of magnitude,
with TOF increased to more than 1 s-1 (Table 3). Increasing the distance between the ketone
and phenyl group by a single carbon (9-11a), dramatically increased BVMOAFL838’s affinity for
the substrates, with TOF now between 6 and 8 s-1 at 0.1 mM (2.4–4.3 s-1 at 10 mM) and again
the para-substituted substrates (10a and 11a) giving higher rates. Similar observations that
substituents on the para position appear to be critical for substrate recognition have been made
for 4-hydroxyacetophenone monooxygenases (HAPMOs) from both Pseudomonas fluorescens
ACB [33,34] and Pseudomonas putida JD1 [35].
Absolute regioselectivity was also observed against all the aryl ketones tested. The ester
products (12b, 13b) from benzaldehyde and 4-methoxybenzaldehyde were rapidly hydrolyzed
to form phenol and 4-methoxyphenol (mequinol) as sole products. This regioselectivity is
again similar as to what has been observed for HAPMO with various benzaldehyde derivatives
[31], where even benzaldehydes with electron-withdrawing substituents on the aromatic ring
prefers the formation of phenols.
Overall structure of BVMOAFL838
The crystal structure of BVMOAFL838, the first of a fungal BVMO, was determined at a resolu-
tion of 1.9 Å. BVMOAFL838 shares more than 40% sequence identity with PAMO, which was
Fig 5. Baeyer-Villiger oxidation of aryl ketones and aldehydes.
doi:10.1371/journal.pone.0160186.g005
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
7 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
Table 3. Biotransformations of aryl ketones and aldehydes by E. coli whole cells expressing BVMOAFL838 (WC) and purified BVMOAFL838 (ENZ).
Kinetic parameters determined from initial velocities obtained from NADPH oxidation.
Substrate
Conversion (%)
TOF (s-1)
10 mM kobs
0.05
Km (mM)
WC_2h
19
ENZ_2h
4
ENZ_8h
13
6a
Acetophenone
10.0 1.2
0.8 0.05
2.4 1.2a
<0.01b
<0.01b
<0.01b
7a
4-Hydroxyacetophenone
Phenylacetone
84
>99
n.c.
>99
>99
85
>99
n.c.
>99
>99
85
1.4
0.28a
8a
72
9a
4-Phenyl-2-butanone
4-(4-Hydroxyphenyl)-2-butanone
4-(4-Methoxyphenyl)-2-butanone
Benzaldehyde
64
2.4
10a
11a
12a
13a
85
2.8
68
4.3
62
n.c.
81
n.c.
88
0.03
6.5 0.7
0.03 0.002
4-Methoxybenzaldehyde
63
1.0
a Determined using 2 μM of BVMO
b Substrate inhibition observed, Ki>>Km
doi:10.1371/journal.pone.0160186.t003
used as search model for molecular replacement. Despite numerous attempts to co-crystallize
BVMOAFL838 with co-factor and substrates, none of the crystals analyzed showed any electron
density for bound NADP+ or the substrates used. A similar observation was made during the
crystallization of PAMO [10], where the excess ammonium sulphate used as precipitant (crys-
tallization agent) prevents the binding of NADP+. The first eight amino acids of the N-termi-
nus were not modelled as it was not resolved in the electron density. Electron density was also
not found for residues 230–239 which typically folds into a short α-helix on the surface of
other BVMOs on the opposite side of the substrate entry (NADPH binding) channel.
BVMOAFL838, similar to the four bacterial BVMOs, displays two Rossmann fold domains,
representing the FAD and NADP binding domains, with the isoalloxazine ring of FAD posi-
tioned at the domain interface (Fig 7A). The FAD is bound with the re-face exposed to the sub-
strate binding pocket and the si-face on top of a conserved Tyr69 and Trp52 forming
conserved hydrogen bond interactions with the ribitol moiety of FAD. Except for the conserved
Rossmann fold signature motif (GXGXXG) the region binding the rest of the FAD molecule is
less conserved as most of the contacts are with main chain amide groups and via water medi-
ated interactions. Despite this, Val115 is absolutely conserved between the five known BVMO
Fig 6. Steady-state kinetics of BVMOAFL838 with phenylacetone at different enzyme concentrations (A), Tris buffer concentrations (B) and pH values (C).
doi:10.1371/journal.pone.0160186.g006
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
8 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
Fig 7. BVMOAFL838 crystal structure. (A) Overall structure of BVMOAFL838 represented as a ribbon diagram with bound FAD as a stick model.
“Variable” loop region covering the active site entry channel is shown in red. (B) Active site of BVMOAFL838 showing the catalytically important
Arg (R337) and Asp (D63) and bulge-structure (T442-F444). (C) Substrate entry channel showing the “variable” loop structure in red and residue
K511 from the “variable” loop within close proximity to the anticipated 2’-phosphate of NADPH. Also shown are residues (R214, K336)
previously implicated in co-factor (NADPH) recognition and binding.
doi:10.1371/journal.pone.0160186.g007
structures, with two hydrogen bonds forming between the adenine of FAD and the main chain
amide and carbonyl group of Val. The previously identified BVMO-signature motif [36] is
absolutely conserved in position and sequence.
Active site architecture
The catalytic site Arg337 and Asp63, implicated in stabilization of the peroxyflavin intermedi-
ate and positioning and guiding the cofactor and substrate during catalysis, are conserved. The
Arg is located in the “in” position and forms two salt bridges with Asp (Fig 7B). Similar to
PAMO, a bulge (residues 442–444) protrudes into the active site of BVMOAFL838 (Fig 7B).
Mutation studies showed that elimination of this bulge increased the substrate scope of PAMO
[37]. However, STMO and OTEMO display similar bulges, but accept progesterone and small
monocyclic ketones as substrates respectively.
Variable loop structure
Partially covering the substrate entry channel of BVMOAFL838 is a loop region (residues 449 to
514) of which the first half forms an α-helix structure (Fig 7C). This mobile loop region is
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
9 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
typically disordered (CHMO, STMO) or forms an unstructured loop (PAMO) or a β-hairpin
structure (OTEMO) that is solvent exposed and positioned away from the active site channel
in the “resting” state. In BVMOAFL838 the mobile/“variable” loop adopts a position similar to
that in the CHMO “closed” structure, where part of the active site entry channel is closed off.
In this conformation however, the conserved Trp502 (W492 CHMO numbering) is positioned
away from the active site and the NADPH cofactor, suggesting a rearrangement of this loop
upon cofactor and substrate binding. Apart from Arg214 and Lys336 that have been implicated
in NADPH binding, the conformation of the mobile loop also positioned Lys511 within close
proximity to the anticipated 2’-phosphate of NADPH as suggested through superimposition
with other BVMOs. To probe whether the conformation of the variable/mobile loop observed
in the structure is physiologically important, K511 was mutated to an Ala. Although still cata-
lytically active, the K511A mutant showed reduced catalytic activity towards 3-octanone as
well as an increased uncoupling of NADPH oxidation and product formation.
Conclusions
BVMOAFL838 has a clear preference towards aliphatic ketones and phenyl substituted 2-alka-
nones (aryl ketones where the alkyl chain is at least two carbons between the phenyl and car-
bonyl group). To our knowledge, this data represents the highest specific rates obtained to date
for BVMO transformations of 2-alkanones. Moreover, biotransformations of linear and aryl
ketones using purified enzyme compared to whole-cell biotransformations highlights the
importance of kinetic studies as opposed to single-concentration conversions to determine the
substrate scope and preference of BVMOs when working with purified enzymes as these sys-
tems are more prone to substrate inhibition. BVMOAFL838 constitutes the first structural inves-
tigation of a fungal BVMO. The crystal structure of BVMOAFL838 revealed an overall fold
similar to bacterial BVMOs. More extensive crystallization trials is currently underway to find
conditions to allow solving BVMOAFL838 with NADP+ and substrates to further investigate
conformations of the “variable” loop region, as well as determinants of substrate scope and
regioselectivity. More data is currently needed to understand the discrepancy between substrate
pocket plasticity to accommodate a wide range of substrates yet high product selectivity.
Acknowledgments
The authors would like to thank Mr. Sarel Marais for GC-(MS) analysis and Prof. Trevor Sewell
and the beamline scientists of Diamond Light Source beamline I04-I for assisting with data col-
lection under proposal mx12255.
Author Contributions
Conceived and designed the experiments: DJO MSS. Performed the experiments: FMF CT.
Analyzed the data: DJO MSS CT FMF. Wrote the paper: DJO MSS CT FMF.
References
1. Kamerbeek NM, Janssen DB, van Berkel WJH, Fraaije MW. Baeyer–Villiger Monooxygenases, an
Emerging Family of Flavin-Dependent Biocatalysts. Adv Synth Catal. 2003; 345: 667–678. doi: 10.
1002/adsc.200303014
2. Balke K, Kadow M, Mallin H, Saß S, Bornscheuer UT. Discovery, application and protein engineering of
Baeyer-Villiger monooxygenases for organic synthesis. Org Biomol Chem. 2012; 10: 6249–6265. doi:
10.1039/c2ob25704a PMID: 22733152
3. Leisch H, Morley K, Lau PCK. Baeyer-Villiger monooxygenases: more than just green chemistry.
Chem Rev. 2011; 111: 4165–4222. doi: 10.1021/cr1003437 PMID: 21542563
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
10 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
4. Torres Pazmiño DE, Dudek HM, Fraaije MW. Baeyer-Villiger monooxygenases: recent advances and
future challenges. Curr Opin Chem Biol. 2010; 14: 138–144. doi: 10.1016/j.cbpa.2009.11.017 PMID:
20015679
5. Zhang Z-G, Parra LP, Reetz MT. Protein engineering of stereoselective Baeyer-Villiger monooxy-
genases. Chemistry. 2012; 18: 10160–10172. doi: 10.1002/chem.201202163 PMID: 22807240
6. Clouthier CM, Kayser MM, Reetz MT. Designing new Baeyer-Villiger monooxygenases using restricted
CASTing. J Org Chem. 2006; 71: 8431–8437. doi: 10.1021/jo0613636 PMID: 17064016
7. Ferroni FM, Smit MS, Opperman DJ. Functional divergence between closely related Baeyer-Villiger
monooxygenases from Aspergillus flavus. J Mol Catal B Enzym. 2014; 107: 47–54. doi: 10.1016/j.
molcatb.2014.05.015
8. Mascotti ML, Juri Ayub M, Dudek H, Sanz MK, Fraaije MW. Cloning, overexpression and biocatalytic
exploration of a novel Baeyer-Villiger monooxygenase from Aspergillus fumigatus Af293. AMB
Express. 2013; 3: 33. doi: 10.1186/2191-0855-3-33 PMID: 23767684
9. Mascotti ML, Lapadula WJ, Juri Ayub M. The origin and evolution of Baeyer—Villiger monooxygenases
(BVMOs): An ancestral family of flavin monooxygenases. PLoS One. 2015; 10: e0132689. doi: 10.
1371/journal.pone.0132689 PMID: 26161776
10. Malito E, Alfieri A, Fraaije MW, Mattevi A. Crystal structure of a Baeyer-Villiger monooxygenase. Proc
Natl Acad Sci U S A. 2004; 101: 13157–13162. doi: 10.1073/pnas.0404538101 PMID: 15328411
11. Mirza IA, Yachnin BJ, Wang S, Grosse S, Bergeron H, Imura A, et al. Crystal structures of cyclohexa-
none monooxygenase reveal complex domain movements and a sliding cofactor. J Am Chem Soc.
2009; 131: 8848–8854. doi: 10.1021/ja9010578 PMID: 19385644
12. Franceschini S, van Beek HL, Pennetta A, Martinoli C, Fraaije MW, Mattevi A. Exploring the structural
basis of substrate preferences in Baeyer-Villiger monooxygenases: insight from steroid monooxygen-
ase. J Biol Chem. 2012; 287: 22626–22634. doi: 10.1074/jbc.M112.372177 PMID: 22605340
13. Leisch H, Shi R, Grosse S, Morley K, Bergeron H, Cygler M, et al. Cloning, Baeyer-Villiger biooxida-
tions, and structures of the camphor pathway 2-oxo-Δ(3)-4,5,5-trimethylcyclopentenylacetyl-coenzyme
A monooxygenase of Pseudomonas putida ATCC 17453. Appl Environ Microbiol. 2012; 78: 2200–
2212. doi: 10.1128/AEM.07694-11 PMID: 22267661
14. Orru R, Dudek HM, Martinoli C, Torres Pazmino DE, Royant A, Weik M, et al. Snapshots of enzymatic
Baeyer-Villiger catalysis: oxygen activation and intermediate stabilization. J Biol Chem. 2011; 286:
29284–29291. doi: 10.1074/jbc.M111.255075 PMID: 21697090
15. Yachnin BJ, Sprules T, McEvoy MB, Lau PCK, Berghuis AM. The substrate-bound crystal structure of a
Baeyer-Villiger monooxygenase exhibits a Criegee-like conformation. J Am Chem Soc. 2012; 134:
7788–7795. doi: 10.1021/ja211876p PMID: 22506764
16. Yachnin BJ, McEvoy MB, MacCuish RJD, Morley KL, Lau PCK, Berghuis AM. Lactone-bound struc-
tures of cyclohexanone monooxygenase provide insight into the stereochemistry of catalysis. ACS
Chem Biol. 2014; 9: 2843–2851. doi: 10.1021/cb500442e PMID: 25265531
17. Fraaije MW, Wu J, Heuts DPHM, van Hellemond EW, Spelberg JHL, Janssen DB. Discovery of a ther-
mostable Baeyer-Villiger monooxygenase by genome mining. Appl Microbiol Biotechnol. 2005; 66:
393–400. doi: 10.1007/s00253-004-1749-5 PMID: 15599520
18. Miyamoto M, Matsumoto J, Iwaya T, Itagaki E. Bacterial steroid monooxygenase catalyzing the Bae-
yer-Villiger oxidation of C21-ketosteroids from Rhodococcus rhodochrous: the isolation and characteri-
zation. Biochim Biophys Acta. 1995; 1251: 115–124. Available: http://www.ncbi.nlm.nih.gov/pubmed/
7669800 PMID: 7669800
19. Opperman DJ, Reetz MT. Towards practical Baeyer-Villiger-monooxygenases: design of cyclohexa-
none monooxygenase mutants with enhanced oxidative stability. Chembiochem. 2010; 11: 2589–
2596. doi: 10.1002/cbic.201000464 PMID: 21080396
20. Studier FW. Protein production by auto-induction in high-density shaking cultures. Protein Expr Purif.
2005; 41: 207–234. doi: 10.1016/j.pep.2005.01.016 PMID: 15915565
21. Leslie AGW, Powell HR. Processing diffraction data with MOSFLM. Evolving methods for macromolec-
ular crystallography. 2007; 245: 45–51. doi: 10.1007/978-1-4020-6316-9_4
22. Evans P. Scaling and assessment of data quality. Acta Crystallogr Sect D Biol Crystallogr. International
Union of Crystallography; 2006; 62: 72–82. doi: 10.1107/S0907444905036693
23. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic
software. J Appl Crystallogr. International Union of Crystallography; 2007; 40: 658–674. doi: 10.1107/
S0021889807021206 PMID: 19461840
24. Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of Coot. Acta Crystallogr Sect
D Biol Crystallogr. International Union of Crystallography; 2010; 66: 486–501. doi: 10.1107/
S0907444910007493
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
11 / 12

Fungal Baeyer-Villiger Monooxygenase Characterization
25. Murshudov GN, Skubák P, Lebedev AA, Pannu NS, Steiner RA, Nicholls RA, et al. REFMAC 5 for the
refinement of macromolecular crystal structures. Acta Crystallogr Sect D Biol Crystallogr. 2011; 67:
355–367. doi: 10.1107/S0907444911001314
26. Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, Evans PR, et al. Overview of the CCP 4
suite and current developments. Acta Crystallogr Sect D Biol Crystallogr. International Union of Crystal-
lography; 2011; 67: 235–242. doi: 10.1107/S0907444910045749
27. Bisagni S, Smuś J, Chávez G, Hatti-Kaul R, Mamo G. Cloning and expression of a Baeyer–Villiger
monooxygenase oxidizing linear aliphatic ketones from Dietzia sp. D5. J Mol Catal B Enzym. 2014;
109: 161–169. doi: 10.1016/j.molcatb.2014.08.020
28. Dudek HM, Fink MJ, Shivange AV, Dennig A, Mihovilovic MD, Schwaneberg U, et al. Extending the
substrate scope of a Baeyer-Villiger monooxygenase by multiple-site mutagenesis. Appl Microbiol Bio-
technol. 2014; 98: 4009–4020. doi: 10.1007/s00253-013-5364-1 PMID: 24247989
29. Rehdorf J, Kirschner A, Bornscheuer UT. Cloning, expression and characterization of a Baeyer-Villiger
monooxygenase from Pseudomonas putida KT2440. Biotechnol Lett. 2007; 29: 1393–1398. doi: 10.
1007/s10529-007-9401-y PMID: 17530181
30. Völker A, Kirschner A, Bornscheuer UT, Altenbuchner J. Functional expression, purification, and char-
acterization of the recombinant Baeyer-Villiger monooxygenase MekA from Pseudomonas veronii
MEK700. Appl Microbiol Biotechnol. 2008; 77: 1251–1260. doi: 10.1007/s00253-007-1264-6 PMID:
18034235
31. Moonen MJH, Westphal AH, Rietjens IMCM, Van Berkel WJH. Enzymatic Baeyer-Villiger oxidation of
benzaldehydes. Adv Synth Catal. 2005; 347: 1027–1034. doi: 10.1002/adsc.200404307
32. Pham SQ, Gao P, Li Z. Engineering of recombinant E. coli cells co-expressing P450pyrTM monooxy-
genase and glucose dehydrogenase for highly regio- and stereoselective hydroxylation of alicycles
with cofactor recycling. Biotechnol Bioeng. 2013; 110: 363–373. doi: 10.1002/bit.24632 PMID:
22886996
33. Kamerbeek NM, Moonen MJ, Van Der Ven JG, Van Berkel WJ, Fraaije MW, Janssen DB. 4-Hydroxya-
cetophenone monooxygenase from Pseudomonas fluorescens ACB. A novel flavoprotein catalyzing
Baeyer-Villiger oxidation of aromatic compounds. Eur J Biochem. 2001; 268: 2547–2557. doi: 10.1046/
j.1432-1327.2001.02137.x PMID: 11322873
34. Kamerbeek NM, Olsthoorn AJJ, Fraaije MW, Janssen DB. Substrate specificity and enantioselectivity
of 4-hydroxyacetophenone monooxygenase. Appl Environ Microbiol. 2003; 69: 419–426. doi: 10.1128/
AEM.69.1.419 PMID: 12514023
35. Rehdorf J, Zimmer CL, Bornscheuer UT. Cloning, expression, characterization, and biocatalytic investi-
gation of the 4-hydroxyacetophenone monooxygenase from Pseudomonas putida JD1. Appl Environ
Microbiol. 2009; 75: 3106–3114. doi: 10.1128/AEM.02707-08 PMID: 19251889
36. Fraaije MW, Kamerbeek NM, van Berkel WJH, Janssen DB. Identification of a Baeyer–Villiger monoox-
ygenase sequence motif. FEBS Lett. 2002; 518: 43–47. doi: 10.1016/S0014-5793(02)02623-6 PMID:
11997015
37. Bocola M, Schulz F, Leca F, Vogel A, Fraaije MW, Reetz MT. Converting Phenylacetone Monooxygen-
ase into Phenylcyclohexanone Monooxygenase by Rational Design: Towards Practical Baeyer-Villiger
Monooxygenases. Adv Synth Catal. 2005; 347: 979–986. doi: 10.1002/adsc.200505069
PLOS ONE | DOI:10.1371/journal.pone.0160186 July 29, 2016
12 / 12