A Flavoprotein Monooxygenase that Catalyses a Baeyer-Villiger Reaction and Thioether Oxidation Using NADH as the Nicotinamide Cofactor

Article abstract of DOI:10.1002/cbic.201200006

A gene from the marine bacterium Stenotrophomonas maltophilia encodes a 38.6 kDa FAD-containing flavoprotein (Uniprot B2FLR2) named S. maltophilia flavin-containing monooxygenase (SMFMO), which catalyses the oxidation of thioethers and also the regioselective Baeyer-Villiger oxidation of the model substrate bicyclo[3.2.0]hept-2-en-6-one. The enzyme was unusual in its ability to employ either NADH or NADPH as nicotinamide cofactor. The K_M and k_cat values for NADH were 23.7±9.1 μM and 0.029 s^-1 and 27.3±5.3 μM and 0.022 s^-1 for NADPH. However, k_cat/K_M value for the ketone substrate in the presence of 100 μM cofactor was 17 times greater for NADH than for NADPH. SMFMO catalysed the quantitative conversion of 5 mM ketone in the presence of substoichiometric concentrations of NADH with the formate dehydrogenase cofactor recycling system, to give the 2-oxa and 3-oxa lactone products of Baeyer-Villiger reaction in a ratio of 5:1, albeit with poor enantioselectivity. The conversion with NADPH was 15%. SMFMO also catalysed the NADH-dependent transformation of prochiral aromatic thioethers, giving in the best case, 80% ee for the transformation of p-chlorophenyl methyl sulfide to its R enantiomer. The structure of SMFMO reveals that the relaxation in cofactor specificity appears to be accomplished by the substitution of an arginine residue, responsible for recognition of the 2′-phosphate on the NADPH ribose in related NADPH-dependent FMOs, with a glutamine residue in SMFMO. SMFMO is thus representative of a separate class of single-component, flavoprotein monooxygenases that catalyse NADH-dependent oxidations from which possible sequences and strategies for developing NADH-dependent biocatalysts for asymmetric oxygenation reactions might be identified.

Full text of DOI:10.1002/cbic.201200006

DOI: 10.1002/cbic.201200006
A Flavoprotein Monooxygenase that Catalyses a Baeyer–
Villiger Reaction and Thioether Oxidation Using NADH as
the Nicotinamide Cofactor
Chantel N. Jensen,[a] Jared Cartwright,[b] Jonathan Ward,[a] Sam Hart,[a] Johan P. Turkenburg,[a]
Sohail T. Ali,[c] Michael J. Allen,[c] and Gideon Grogan*[a]
A gene from the marine bacterium Stenotrophomonas malto-
philia encodes a 38.6 kDa FAD-containing flavoprotein (Uniprot
B2FLR2) named S. maltophilia flavin-containing monooxyge-
nase (SMFMO), which catalyses the oxidation of thioethers and
also the regioselective Baeyer–Villiger oxidation of the model
substrate bicyclo[3.2.0]hept-2-en-6-one. The enzyme was un-
usual in its ability to employ either NADH or NADPH as nicoti-
namide cofactor. The KM and kcat values for NADH were 23.7Æ
9.1 mm and 0.029 sÀ1 and 27.3Æ5.3 mm and 0.022 sÀ1 for
NADPH. However, kcat/KM value for the ketone substrate in the
presence of 100 mm cofactor was 17 times greater for NADH
than for NADPH. SMFMO catalysed the quantitative conversion
of 5 mm ketone in the presence of substoichiometric concen-
trations of NADH with the formate dehydrogenase cofactor re-
cycling system, to give the 2-oxa and 3-oxa lactone products
of Baeyer–Villiger reaction in a ratio of 5:1, albeit with poor
enantioselectivity. The conversion with NADPH was 15%.
SMFMO also catalysed the NADH-dependent transformation of
prochiral aromatic thioethers, giving in the best case, 80% ee
for the transformation of p-chlorophenyl methyl sulfide to its R
enantiomer. The structure of SMFMO reveals that the relaxa-
tion in cofactor specificity appears to be accomplished by the
substitution of an arginine residue, responsible for recognition
of the 2’-phosphate on the NADPH ribose in related NADPH-
dependent FMOs, with a glutamine residue in SMFMO. SMFMO
is thus representative of a separate class of single-component,
flavoprotein monooxygenases that catalyse NADH-dependent
oxidations from which possible sequences and strategies for
developing NADH-dependent biocatalysts for asymmetric oxy-
genation reactions might be identified.
Introduction
Flavoprotein monooxygenases (FPMOs) catalyse many reac-
tions of potential interest for application in catalytic asymmet-
ric synthesis. A review by van Berkel and Fraaije in 2006 has
grouped these enzymes into six classes, A–F.[1] Of these, the
Baeyer–Villiger monooxygenases (BVMOs) and flavin-contain-
ing monooxygenases (FMOs), each from class B, have proved
to be particularly interesting in respect of their ability to cata-
lyse enantioselective oxidation reactions (Scheme 1). BVMOs[2]
catalyse the insertion of oxygen adjacent to a carbonyl group,
an important transformation in synthetic organic chemistry for
the peracid-mediated conversion of ketones to esters or lac-
tones. They utilise a nicotinamide cofactor (commonly NADPH)
to reduce a flavin coenzyme FAD to FADH2, which then reacts
with molecular oxygen to form a flavin hydroperoxidate, which
is thought to be the active catalyst in the oxygen-insertion re-
action.[3] The most common class B BVMOs, typified by the cy-
clohexanone monooxygenase (CHMO) from Acinetobacter,[4] are
single polypeptides of approximately 55–65 kDa and have the
ability to catalyse both NADPH-dependent reduction of FAD
and also the substrate oxygenation reaction within their active
site. BVMOs have also been shown to catalyse the asymmetric
electrophilic oxidation of prochiral thioethers to produce chiral
sulfoxide products with high enantioselectivity.[5]
use NADPH to reduce FAD, and use the flavin hydroperoxide
to oxidise heteroatoms such as sulfur and nitrogen.[6] FMOs do
not usually catalyse Baeyer–Villiger oxidations, perhaps be-
cause they lack a conserved arginine residue thought to be
necessary for stabilisation of the peroxidate anion required for
oxygen insertion reactions.[1] There are some reports of FMO-
catalysed Baeyer–Villiger reactions however,[7] although the
basis of peroxidate stabilisation in those processes is not yet
fully understood. The profile of FMOs is increasing with respect
to their application as biocatalysts, particularly in the oxidation
of thioethers to optically active sulfoxides.[8]
In the cases of both BVMOs and FMOs, one of the draw-
backs from the perspective of industrial application is the de-
pendence of the enzymes on the phosphorylated nicotinamide
[a] C. N. Jensen, J. Ward, S. Hart, Dr. J. P. Turkenburg, Dr. G. Grogan
Department of Chemistry, University of York
Heslington, York, YO10 5DD (UK)
E-mail: gideon.grogan@york.ac.uk
[b] Dr. J. Cartwright
Technology Facility, Department of Biology, University of York,
YO10 5DD York (UK)
[c] S. T. Ali, Dr. M. J. Allen
Plymouth Marine Laboratory
Prospect Place, Plymouth PL1 3DH (UK)
Class B FMOs also typically consist of a single polypeptide
that has the ability to both reduce flavin and oxidise their sub-
strates using, in this case, a flavin hydroperoxide.[1] They again
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201200006.
872
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2012, 13, 872 – 878

SMFMO: A Flavoprotein Monooxygenase
prot Code B2FLR2) with an amino acid motif typical of an FMO
(FAGIQLHSAHY)[1]), and two Rossman motifs (GXGXXG) for the
binding of ADP moiety of the FAD and NADPH, as typically ob-
served in class B FMOs (see Figure S1 in the Supporting Infor-
mation). The putative molecular weight of the encoded protein
was 38558 kDa, significantly smaller than class B bacterial
FMOs that have been fully characterised such as mFMO from
Methylophaga aminisulfidovorans (2XLT, 2XVI, 53.0 kDa[15,16]). A
BLAST search of the Swissprot database using the SMFMO
translated gene sequence revealed closest sequence homology
to shorter “putative flavin monooxygenases” from strains such
as Pseudomonas, but also pyridine nucleotide disulfide oxidore-
ductase family enzymes including thioredoxin reductases
(TrxRs).
Scheme 1. Baeyer–Villiger oxidation and sulfoxidation reactions catalysed by
flavin-dependent Baeyer–Villiger monooxygenases (BVMOs) and flavin mon-
ooxygenases (FMOs). NAD(P)H is used to reduce the flavin prior to reaction
with molecular oxygen, yielding a flavin hydroperoxidate, or hydroperoxide,
which is the oxidant in the reaction(s).
The sequence homology with TrxRs is instructive, as a se-
quence-based search of the RCSB Protein Structure database
also revealed that the closest sequence homology to proteins
for which the structure had been solved also includes TrxRs,
such as those from Thermus thermophilus (2ZBW) and Helico-
bacter pylori (2Q0K), as well as BVMOs [PAMO (1W4X) and
CHMO from Rhodococcus (3GWD)] and FMOs from S. pombe
(2V8) and Methylophaga spp (2XVI, 2XLT). A phylogenetic tree
(Figure 1) featuring representative sequences of BVMOs, FMOs
and TrxRs shows that SMFMO features on a branch as closely
related to TrxRs, with which it has most similarity in overall
length and sequence similarity, as to BVMOs or FMOs. A se-
quence alignment of some representative enzymes from
Figure 1 is shown in Figure S2.
cofactor NADPH. NADPH is more expensive than its nonphos-
phorylated analogue NADH, and hence attempts have been
made to investigate or engineer the cofactor preference of
BVMOs, based on structural analysis of the active site interac-
tions with the phosphate of NADPH.[9,10] However, success in
this regard has been limited, suggesting that the simple engi-
neering of NADH specificity in NADPH-dependent enzymes is
not a trivial undertaking. NADH-dependent BVMO activity has
been reported,[11,12] but in those enzymes (class C FPMOs[1]) the
flavin reduction and substrate oxygenation domains are sepa-
rate polypeptides, and the transfer of reduced flavin (FMNH2 in
those cases) between proteins for catalysis presents additional
complications with respect to application.
Cloning, expression and purification
A version of the gene encoding SMFMO was synthesised using
a sequence optimised for codon usage in E. coli, amplified
using PCR and cloned into an expression vector for production
of the enzyme in E. coli. The enzyme was then purified using
nickel-affinity chromatography and gel filtration (Figure S3) to
yield an enzyme of bright yellow colour. Calibrated size-exclu-
sion chromatography suggested the enzyme to be a dimer in
solution and the presence of FAD was confirmed by LCMS
analysis of boiled enzyme extracts against boiled, standard
samples of FAD and FMN using an established procedure.[17]
In this report, we describe the cloning and expression of
a gene encoding a single-component flavoprotein monooxyge-
nase,
S. maltophilia
flavin-containing
monooxygenase
(SMFMO), capable of Baeyer–Villiger oxidation and sulfoxida-
tion reactions, which is able to use NADH as well as NADPH as
a nicotinamide cofactor. The structure of SMFMO also reveals
the molecular determinants of cofactor promiscuity when com-
pared to NADPH-dependent FMOs. Characterisation of the pu-
rified recombinant enzyme is reported, along with its applica-
tion to the oxidation of a series of commonly used substrates
for BVMOs and FMOs. Whilst the enantioselectivity of the
enzyme is only low to moderate, the study points to a new
subfamily of flavoprotein targets that might be exploited or
evolved for improved asymmetric oxidations with considerable
economic advantages over NADPH-dependent enzymes.
Enzyme assays
In preliminary experiments, pure SMFMO was assayed for
class B FPMO-type activity using GC analysis of biotransforma-
tions of the standard FMO substrate p-tolyl methyl sulfide (1,
Scheme 2) in the presence of either NADPH or NADH as nicoti-
namide cofactor. Interestingly, SMFMO was capable of oxida-
tion of a thioether substrate using either cofactor. SMFMO was
then assayed for its ability to catalyse Baeyer–Villiger oxidation
of the standard BVMO substrate bicyclo[3.2.0]hept-2-en-6-one
(3). Again, activity was observed using either reduced nictoti-
namide cofactor, despite the enzyme being structurally dissimi-
lar to known NADPH-dependent BVMOs, which are typically
enzymes of approximately 60 kDa molecular weight. Kinetic
studies of cofactor utilisation by SMFMO were performed
Results and Discussion
Identification of SMFMO gene
As part of studies directed towards characterising new FMO
and BVMO activity,[13] analysis of the genome sequence of the
marine isolate Stenotrophomonas maltophilia[14] revealed a puta-
tive open-reading frame, encoding an enzyme (SMFMO, Uni-
ChemBioChem 2012, 13, 872 – 878
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
873

G. Grogan et al.
Table 1. Kinetic constants for SMFMO using NADH or NADPH as cofactor,
with substrate 3.
Experiment
KM
Vmax
kcat
[sÀ1
kcat/KM
[mmoldmÀ3 sÀ1
]
[mmoldmÀ3] [mmoldmÀ3 sÀ1
]
]
NADH+100mm 3 23.7Æ9.1
11.2Æ1.5ꢁ10À2 0.029
1223
806
NADPH+100mm 27.3Æ5.3
8.4Æ0.6ꢁ10À2 0.022
3
3+100mm NADH 0.07Æ0.03 1.4Æ0.2ꢁ10À2 3.6ꢁ10À3
51
3
3+100mm
NADPH
3.20Æ0.90 3.3Æ0.3ꢁ10À2 8.5ꢁ10À3
using the method described for hydroxyacetophenone mono-
oxygenase (HAPMO),[10] using either cofactor in the presence of
100 mm of the ketone substrate 3 (Table 1, Figure S4).
The apparent KM and Vmax values for NADH and NADPH in
this assay were comparable, in the region of 20–30 mm, indica-
tive of an equal capability for the use of either cofactor to
reduce flavin in the enzyme, and are comparable with the
NADPH oxidase activity of, for example mFMO from Methylo-
phaga (13 mm and 0.06 sÀ1).[15] Kinetic constants for substrate 3
were then determined in the presence of 100 mm of either co-
factor (Table 1, Figure S5A and B in the Supporting Informa-
tion). In these cases, Michaelis–Menten behaviour with respect
to 3 was observed, with the KM value for the ketone substrate
in the presence of NADH approximately 42 times lower, and
kcat/KM value 17 times higher than for NADPH. The Vmax value
for the ketone with NADH (1.4ꢁ10À2 mmsÀ1) is appreciably
smaller than that usually observed for BVMOs acting on ke-
tones, but is within the range of, for example the FPMO
MtmOIV, when expressed in the same terms (5 nmolminmgÀ1
protein versus 147 nmolminmgÀ1 [18]), and in that case MtmOIV
was acting on its natural substrate.
Figure 1. Phylogenetic tree showing position of Stenotrophomonas maltophi-
la flavin-containing monooxygenase SMFMO in the context of some BVMOs,
FMOs, and related flavin-binding thioredoxin reductases (TrxRs) for which
structure and/or function has been described. BVMOs: HI-31 CHMO=Cyclo-
hexanone monooxygenase from Rhodococcus sp. HI-31 (3GWD, Uniprot
C0STX7); AcCHMO=Cyclohexanone monooxygenase from Acinetobacter cal-
coaceticus (Uniprot Q9R2F5); MO14=ro03437 from Rhodococcus jostii RHA1
(Uniprot Q0SB46); MO3=ro03247 from Rhodococcus jostii RHA1 (Uniprot
Q0SBN6); PAMO=phenylacetone monooxygenase from Thermobifida fusca
(Uniprot Q47PU3); CPMO=Cyclopentanone monooxygenase from Comamo-
nas sp. (Uniprot Q8GAW0); TrxRs: 3F8P=TrxR from Sulfolobus solfataricus
(Uniprot Q8X236); 3CTY=TrxR from Thermoplasma acidophilum (Uniprot
Q9HJI4); 2A87=TrxR from Mycobacterium tuberculosis (Uniprot P52214);
2Q7V=thioredoxin reductase from Deinococcus radiodurans (Uniprot
Q9RSY7); 2Q0K=thioredoxin reductase from Helicobacter pylori (Uniprot
P56431); 3R9U=TrxR from Campylobacter jejuni (Uniprot Q0PBZ1);
3LZX=ferredoxin-NADP+ oxidoreductase from Bacillus subtilis (Uniprot
O05268); 2ZBW=TrxR-like protein from Thermus thermophilus (Uniprot
Q5L28); FMOs: 1VQW=Flavin-containing monooxygenase from Schizosac-
charomyces pombe (Uniprot Q9HFE4); 2XVI=FMO from Methylophaga amini-
sulfidivirans (Uniprot Q83K4).
Substrate selectivity and enantioselectivity
A range of class B FPMO ketone substrates was then tested as
substrates for SMFMO, using either NADH or NADPH as nicoti-
Scheme 2. Test substrates and products used in this study.
874
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2012, 13, 872 – 878

SMFMO: A Flavoprotein Monooxygenase
namide cofactor, and the appropriate cofactor recycling sys-
tems: formate dehydrogenase and sodium formate for NADH
and glucose-6-phosphate dehydrogenase and glucose-6-phos-
phate for NADPH. The Baeyer–Villiger oxidation activity of
SMFMO was restricted to the strained, fused cyclobutanone
system of substrate 3, with many simpler aliphatic and alicylic
ketones (Scheme 2) such as cyclopentanone (6), cyclohexanone
(7), acetophenone (8) and octan-2-one (9) not transformed
with either cofactor. Conversion of 3 was appreciably higher
when NADH was employed as cofactor with 93% conversion
observed compared to 15% with NADPH over a 24 h period.
The NADH-dependent transformation of 3 was found to be re-
gioselective in the manner of peracid-mediated Baeyer–Villiger
oxidations of this substrate, giving a yield of regioisomers in
the ratio 5:1 in favour of the “normal” 2-oxa lactone over the
“abnormal” 3-oxa product. Enantioselectivity was poor, but
measurable, with 8% ee for (1R,5S)-(+)-4 and 36% ee for (1S,
5R)-5. Chiral GC analysis of the residual ketone showed it to be
racemic.
Figure 2. Quaternary structure of dimeric SMFMO showing the two subunits
A and B. Protein backbone is shown in ribbon format. FAD molecules (one
per subunit) are shown in cylinder format. The N and C termini of subunit B
are also indicated.
More promiscuous substrate selectivity was observed in the
transformation of prochiral thioether substrates by SMFMO
(Table 2). Conversions were observed for phenyl methyl sulfide
(10), p-tolyl methyl sulfide (1), o- and p-chlorophenyl methyl
sulfide (12 and 14 respectively), benzyl methyl sulfide (17) and
phenyl ethyl sulfide (19). p-Nitrophenyl methyl sulfide (16) was
not transformed. In each case, greater conversions were ach-
ieved using NADH as cofactor, and enantiomeric excess of
products was dependent on the substrate structure. The pre-
ferred enantiomeric series in sulfoxidation by SMFMO [(R)-] was
in most cases the same as that encountered for chloroperoxi-
dase.[19] In most cases, and most notably for 10, the enantiose-
lectivity was poorer than for CHMO[5,20] or mFMO,[8] but it was
superior in the case of substrate 14 and in the opposite enan-
tiomeric series.
also shares some structural motifs with known FMOs such as
mFMO (2XLT, 2XIV),[15,16] to BVMOs such as 3GWD from Rhodo-
coccus[23] and also to enzymes of the thioredoxin family of pro-
teins such as that from Thermus thermophilus (2ZBW).
Each monomer of SMFMO is complexed with FAD, and is
complete apart from a loop of approximately twenty amino
acids between positions 212 and 233, which may constitute
a flexible loop that covers the active site containing the flavin,
and could not be modelled. Although the structure is not in
complex with NAD(P)H, the putative binding sites of the nicoti-
namide cofactor phosphates, including that which is on the 2’-
hydroxyl of ribose, and distinguishes NADP(H) from NAD(H),
are occupied with sulfate, which was used in 0.9m concentra-
tion in the crystallisation drop. Superimposition of NADPH-de-
pendent FMO structure 2XLT, which is complexed with both
FAD and NADPH, with SMFMO clearly reveals that the Arg234
(and Thr235) residue of NADPH-dependent enzymes of this
type, which is involved in binding the ribose phosphate, is re-
placed by a glutamine residue Gln193 (and His 194) in SMFMO
(Figure 3). Specificity for NADPH appears therefore to have
been relaxed owing to the removal of interactions between
the positively charged arginine side chain and the negatively
charged phosphate oxygen(s).
Table 2. Results of biotransformations of prochiral thioether substrates
by SMFMO.
Thioether
substrate
Conversion [%]
ee [%]
using NADH using NADPH using NADH using NADPH
1
10
12
14
16
17
19
90
8
6
40
0
32
27
33
1
0.2
9
0
10
2
25 (R)
21 (R)
15 (S)
80 (R)
–
44 (R)
n.d.
n.d
82 (R)
–
38 (R)
57 (R)
24 (R)
71 (R)
n.d.: not determined.
It has been reported[24] that NADP+ fulfils dual roles in
FPMOs: helping to stabilise the flavin oxygenating species
after flavin reduction and reaction with oxygen. The superior
activity of SMFMO towards substrates on addition of NADH
rather than NADPH suggests that this stabilisation may be
better supplied by the nonphosphorylated cofactor. However,
the molecular basis of stabilisation of a flavin hydroperoxide
for sulfoxidation by SMFMO is still unresolved. In mFMO,
Asn78 is thought to provide this stabilisation, and activity is re-
moved even by mutation to serine.[24] SMFMO features the
Structure of SMFMO
In order to investigate the molecular determinants of cofactor
promiscuity in SMFMO, the X-ray structure of SMFMO was
solved and refined to a resolution of 2.72 ꢂ. SMFMO exists as
a dimer (Figure 2) and its monomeric tertiary structure is most
similar to the structure of putative FMO 3D1C in the Protein
Databank, with which it shares only 16% sequence identity. It
ChemBioChem 2012, 13, 872 – 878
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
875

G. Grogan et al.
were submitted for sequencing to confirm the integrity of the
gene.
Enzyme purification: The pET-YSBLIC-3C vector containing the
SMFMO gene was used to transform cells of E. coli BL21 (DE3)
using kanamycin (30 mgmLÀ1) as antibiotic marker on Luria Bertani
(LB) agar. A single colony of a plate grown overnight was used to
inoculate 4ꢁ5 mL of LB broth, which were grown overnight at
378C with shaking at 180 rpm. The starter cultures were then used
to inoculate LB broth (4ꢁ500 mL cultures) in which cells were
grown until the optical density (OD600) of the culture had reached
approximately 0.6. At this point the expression of SMFMO was in-
duced by the addition of isopropyl b-d-1-thiogalactopyranoside
(IPTG, final concentration of 1 mm). The cultures were then incu-
bated at 188C in an orbital shaker overnight at 180 rpm. After ap-
proximately 18 h growth, the cells in each case were harvested by
centrifugation (4225 g, 15 min) in a Sorvall RC5B Plus centrifuge
and were then resuspended collectively in Tris·HCl buffer (100 mL,
50 mm, pH 7.5, henceforth referred to as “buffer”) containing
sodium chloride (300 mm). Cells were disrupted by ultrasonication
for 3ꢁ30 s bursts at 48C with 1 min intervals and the soluble and
insoluble material fractions separated by centrifugation (26892 g,
30 min). The supernatant, containing the soluble SMFMO, was
loaded onto a 5 mL His-Trap Chelating HP nickel column. After
washing with buffer (10 column volumes containing 30 mm imida-
zole) enzyme was eluted using an imidazole gradient (30–500 mm
imidazole over 20 column volumes of eluant). Column fractions
were analysed by SDS PAGE and those bright yellow fractions con-
taining a pure 38 kDa protein were pooled and concentrated using
a 10 kDa cut-off Centricon filter membrane. Concentrated protein
was loaded onto an S75 Superdex gel filtration column that had
been equilibrated with buffer, and eluted with buffer (120 mL).
Fractions containing pure protein were pooled and stored at
À208C.
Figure 3. NADP ribose 2’-phosphate recognition site in mFMO (2XLT: back-
bone, side chains and NADP carbon atoms in grey), superimposed with
structurally homologous region of SMFMO (white) bound to sulfate, illustrat-
ing replacement of Arg234 and Thr235 in NADPH-dependent 2XLT with
Gln193 and His 194 in SMFMO of relaxed cofactor specificity.
large, hydrophobic Phe52 in this position (Figure S6). However,
it has recently been shown that hydrophobic residues in this
region are tolerated by flavin-dependent oxidases and oxygen-
ases, and that oxygen activation depends very much on the
active site context.[25]
The ability of SMFMO to catalyse BV reactions is more diffi-
cult to explain. The hydroperoxidate held to be the oxidising
species in enzymatic Baeyer–Villiger reactions is thought to be
stabilised by an arginine residue, Arg337 in PAMO, which is
conserved in all known BVMOs, yet absent in SMFMO. The de-
terminants of Baeyer–Villiger activity in SMFMO are currently
under investigation using mutation experiments.
Enzyme assays: Steady-state kinetic constants for NADH and
NADPH in the presence of ketone substrate 3 (see main text) were
determined using the method employed for HAPMO by Kamer-
beek et al.[10] In a 1 mL quartz cuvette containing Tris·HCl buffer
(pH 7.5, 50 mmol) and ketone 3 (100 nmol), the decrease in absorb-
ance at 340 nm was monitored at concentrations of NAD(P)H (10–
100 mm) after the addition of enzyme (3.9 nmol). The results are
shown in Figure S5. Kinetic constants were calculated using
a value for e of 6220 moldmÀ3 cmÀ1. Kinetic constants for substrate
3 were determined using the method employed for MtmOIV by
Rçhr and co-workers.[18] The oxidation of NADPH or NADH
(100 mm) was monitored in the presence of a range of concentra-
tions of substrate 3 (0–20 mm) after the addition of SMFMO
(3.9 nmol). The results are shown in Figure S5A and B. All data
points represent the average of two separate runs.
Conclusion
The characterisation of SMFMO has identified a type of single-
component FPMO capable of NADH-dependent oxygenation
reactions, and determined the structural basis for relaxed nico-
tinamide cofactor promiscuity in this enzyme. This will help to
highlight other, related enzymes in the database, providing
a possible platform for developing NADH-dependent FPMOs
for asymmetric oxygenation reactions.
Experimental Section
Gene cloning and expression: The gene encoding SMFMO
(B2FLR2) was obtained using commercial DNA synthesis of
a codon-optimised sequence prepared in the plasmid pBlue-
ScriptSK from Geneart (Regensburg, Germany). The gene was then
amplified by PCR using the following primers: CCAGG GACCA
GCAAT GGATA GTGTA GACGT TGTAG TTATT GGTG (forward)/
GAGGA GAAGG CGCGT TAACG GTCCT GGTGA TCGGC GCAG (re-
verse) for cloning into the pET-YSBL-LIC-3C vector using an estab-
lished ligation-independent cloning (LIC) procedure.[26] The result-
ing construct yielded an N-terminal fusion of a hexahistidine tag
with a 3C-protease cleavage site (pET-YSBL-LIC-3C) under the con-
trol of the T7 promoter. The recombinant plasmid was used to
transform E. coli XL1-Blue cells (Novagen) and plasmids recovered
from the recombinant strains by standard miniprep procedures
Biotransformations: Biotransformations using isolated enzyme
with cofactor recycling were performed using the method of Faber
and co-workers.[27] For NADH-dependent biotransformations: To
a 10 mL round bottomed flask containing Tris·HCl buffer (pH 7.5,
5 mL) were added substrate(s) (1, 3, 6–9, 10, 12, 14, 16, 17 or 19
in ethanol (100 mL) to a final concentration of 5 mm), NADH (5 mg,
a final concentration of 0.7 mm), formate dehydrogenase (5 mg),
sodium formate (6.8 mg, 0.1 mmol) and SMFMO (1 mL of
a 5 mgmLÀ1 solution, 0.13 mmol). The reactions were then stirred
for 24 h at room temperature. Aliquots (500 mL) were taken at in-
tervals and extracted with ethyl acetate (500 mL). The organic layer
was transferred to a GC vial and analysed by GC. For NADPH-de-
pendent biotransformations: To a 10 mL round bottomed flask
containing Tris·HCl buffer (pH 7.5, 5 mL) were added substrate(s)
876
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2012, 13, 872 – 878

SMFMO: A Flavoprotein Monooxygenase
(1, 3, 6–9, 10, 12, 14, 16, 17 or 19 in ethanol (100 mL) to a final
concentration of 5 mm), NADPH (5.7 mg, a final concentration of
0.7 mm), glucose-6-phosphate-dehydrogenase (0.14 mg), glucose-
6-phosphate (5.2 mg, 0.02 mmol) and SMFMO (1 mL of
a 5 mgmLÀ1 solution, 0.13 mmol). The reactions were then stirred
for 24 h at room temperature and organic extracts of 500 mL ali-
quots analysed as described above.
structure exhibited Rcryst and Rfree values of 18.6 and 23.9% respec-
tively and was finally validated using PROCHECK.[35] Data collection
and refinement statistics are presented in Table S2. The Ramachan-
dran plot showed 92.3% of residues to be situated in the most fav-
oured regions, 7.4% in additional allowed and only Tyr139 residues
in each subunit exhibiting unusual backbone conformations. The
coordinates and structure factors have been deposited in the Pro-
tein Databank with the accession code 4a9w.
GC analysis: For all standard GC analysis, an Agilent HP-6890 gas
chromatograph fitted with
a HP-5 column (30 mꢁ0.32 mmꢁ
0.25 mm) was used. The injector temperature was set at 2508C and
the detector temperature was 3208C. Helium was used as the carri-
er gas at a pressure of 83 kPa. Details of programmes used for
analysis of substrates 1, 3–9, 10, 12, 14, 16, 17 or 19 can be found
in Table S1. Chiral GC analysis of substrate 3 was performed using
the same GC apparatus equipped with a BGB-175 column (30 mꢁ
0.25 mmꢁ0.25 mm, BGB-Analytik) using a temperature gradient of
1008C to 1278C at 28CminÀ1. Chiral analysis of products 2 and 3
Acknowledgements
We are grateful to the Biotechnology and Biological Sciences Re-
search Council and PML Applications Ltd. for funding a CASE stu-
dentship to C.N.J., and to Karl Heaton for LC-MS experiments.
Keywords: Baeyer–Villiger
biotransformation · flavoprotein · NADPH · oxidoreductase
monooxygenase
·
was performed using
a BGB-173 column (30 mꢁ0.25 mmꢁ
0.25 mm, BGB Analytik) with a temperature gradient of 90–348C at
18CminÀ1. The enantiomeric excesses of products 2, 11, 13, 15, 18
and 20 were measured using the BGB-173 column. Racemic sulfox-
ide standards were prepared from the substrates by oxidation with
hydrogen peroxide and acetic acid.[28] Absolute configurations of
products were assigned through comparison with products of bio-
transformations using cyclohexanone monooxygenase, in conjunc-
tion with results reported for that enzyme.[5,20]
[1] W. J. H. van Berkel, N. M. Kamerbeek, M. W. Fraaije, J. Biotechnol. 2006,
124, 670–689.
[2] D. E. Torres PazmiÇo, H. M. Dudek, M. W. Fraaije, Curr. Opin. Chem. Biol.
2010, 14, 138–144.
[3] D. Sheng, D. P. Ballou, V. Massey, Biochemistry 2001, 40, 11156–11167.
[4] N. A. Donoghue, D. B. Norris, P. W. Trudgill, Eur. J. Biochem. 1976, 63,
175–192.
[5] S. Colonna, N. Gaggero, P. Pasta, G. Ottolina, Chem. Commun. 1996,
2303–2307.
[6] M. A. Hamman, B. D. Haehner-Daniels, S. A. Wrighton, A. E. Rettie, S. D.
Hall, Biochem. Pharmacol. 2000, 60, 7–17.
[7] W. G. Lai, N. Farah, G. A. Moniz, Y. N. Wong, Drug Metab. Dispos. 2011,
39, 61–70.
[8] A. Rioz-Martinez, M. Kopacz, G. de Gonzalo, D. E. Torres PazmiÇo, V.
Gotor, M. W. Fraaije, Org. Biomol. Chem. 2011, 9, 1337–1341.
[9] H. Dudek, D. Torres PazmiÇo, C. Rodrꢃguez, G. de Gonzalo, V. Gotor,
M. W. Fraaije, Appl. Microbiol. Biotechnol. 2010, 88, 1135–1143.
[10] N. M. Kamerbeek, M. W. Fraaije, D. B. Janssen, Eur. J. Biochem. 2004, 271,
2107–2116.
[11] K. H. Jones, R. T. Smith, P. W. Trudgill, J. Gen. Microbiol. 1993, 139, 797–
805.
[12] R. Villa, A. Willetts, J. Mol. Catal. B 1997, 2, 193–197.
[13] C. Szolkowy, L. D. Eltis, N. C. Bruce, G. Grogan, ChemBioChem 2009, 10,
1208–1217.
Crystallisation of SMFMO: Pure SMFMO was subjected to crystalli-
sation trials using a range of commercially available screens in 96-
well plates using 300 nL drops at a range of protein concentra-
tions. The best crystals were obtained using the Clear Strategy
Screen[29] in conditions containing 1.8m lithium sulphate (1.8m)
and protein (4 mgmLÀ1). Larger crystals for diffraction analysis
using optimised conditions were prepared using the hanging-drop
vapour diffusion method in 24-well plate Linbro dishes and using
crystallisation drops of 2 mL with protein (4 mgmLÀ1). The best
crystals were routinely obtained in crystal drops containing lithium
sulphate (0.9m) in Bis-Tris propane buffer (100 mm) at pH 5.6. Crys-
tals were tested for diffraction using a Rigaku Micromax-007HF
generator fitted with Osmic multilayer optics and a MARRESEARCH
MAR345 imaging plate detector. Those crystals that diffracted to
greater than 3 ꢂ resolution were flash-cooled in liquid nitrogen in
a cryogenic solution containing the mother liquor containing also
10% w/v glycerol and retained for data collection at the synchro-
tron.
[14] M. J. Allen, K. Tait, K. Weynberg, C. Bradley, U. Trivedi, M. Blaxter, I. Joint,
J. Nissimov, C. N. Jensen, G. Grogan, M. Muhling, S. T. Ali, manuscript
submitted.
[15] A. Alfieri, E. Malito, R. Orru, M. W. Fraaije, A. Mattevi, Proc. Natl. Acad.
Sci. USA 2008, 105, 6572–6577.
Data collection, structure solution, model building and refine-
ment of SMFMO structure: A complete dataset was collected on
beamline id142 at the European Synchrotron Radiation Facility,
Grenoble on 3rd March 2010. Data were processed using the
HKL2000[30] interface and the data collection statistics are given in
Table S2. The crystals were in space group P32. The structure was
solved using the program BALBES,[31] which found a solution using
separate domains of the thioredoxin reductase from Thermoplasma
acidophilum 3CTY[32] as a model. The solution contained two mole-
cules in the asymmetric unit. The structure was improved using
iterative rounds of building in Autobuild in the Phenix suite of pro-
grammes[33] and refinement using REFMAC.[34] In refinement, non-
crystallographic symmetry (NCS) restraints were applied for chain B
against chain A. After building the protein and water molecules,
FAD was modelled into clear residual density in the difference
maps, to yield the structure of the dimeric holoenzyme. The FAD
isoalloxazine ring was clearly puckered, indicative of at least a par-
tial population of the reduced state FADH2 (Figure S6). The final
[16] H. J. Cho, H. Y. Cho, K. J. Kim, M. H. Kim, S. W. Kim, B. S. Kang, J. Struct.
Biol. 2011, 175, 39–48.
[17] Q. Yu, P. Schaub, S. Ghisla, A. Al-Babili, A. Krieger-Liszkay, P. Bayer, J. Biol.
Chem. 2010, 285, 12109–12120.
[18] M. Gibson, M. Nur-e-alam, F. Lipata. M. A. Oliveira, J. Rohr, J. Am. Chem.
Soc. 2005, 127, 17594–17595.
[19] S. Colonna, N. Gaggero, A. Manfredi, L. Casella, M. Gullotti, G. Carrea, P.
Pasta, Biochemistry 1990, 29, 10465–10468.
[20] G. Carrea, B. Redigolo, S. Riva, S. Colonna, N. Gaggero, E. Battistel, D.
Bianchi, Tetrahedron: Asymmetry 1992, 3, 1063–1068.
[21] D. E. Torres PazmiÇo, R. Snajdrova, D. V. Rial, M. D. Mihovilovic, M. W.
Fraaije, Adv. Synth. Catal. 2007, 349, 1361–1368.
[22] M. T. Reetz, F. Daligault, B. Brunner, H. Hinrichs, A. Deege, Angew. Chem.
2004, 116, 4170–4173; Angew. Chem. Int. Ed. 2004, 43, 4078–4081.
[23] I. A. Mirza, B. J. Yachnin, S. Wang, S. Grosse, H. Bergeron, A. Imura, H.
Iwaki, Y. Hasegawa, P. C. K. Lau, A. M. Berghuis, J. Am. Chem. Soc. 2009,
131, 8848–8854.
[24] R. Orru, D. E. Torres-PazmiÇo, M. W. Fraaije, A. Mattevi, J. Biol. Chem.
2010, 285, 35021–35028.
ChemBioChem 2012, 13, 872 – 878
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
877

G. Grogan et al.
[25] C. A. McDonald, R. L. Fagan, F. Collard, V. M. Monnier, B. A. Palfey, J. Am.
Chem. Soc. 2011, 133, 16809–16811.
[33] P. D. Adams, P. V. Afonine, G. Bunkꢄczi, V. B. Chen, I. W. Davis, N. Echols,
J. J. Headd, L.-W. Hung, G. J. Kapral, R. W. Grosse-Kunstleve, A. J. McCoy,
N. W. Moriarty, R. Oeffner, R. J. Read, D. C. Richardson, J. S. Richardson,
T. C. Terwilliger, P. H. Zwart, Acta. Crystallogr. Sect. D. Biol. Crystallogr.
2010, 66, 213–221.
[34] G. N. Murshudov, A. A. Vagin, E. J. Dodson, Acta. Crystallogr. Sect. D. Biol.
Crystallogr. 1997, 53, 240–255.
[35] R. A. Laskowski, M. W. Macarthur, D. S. Moss, J. M. Thornton, J. Appl.
Crystallogr. 1993, 26, 283–291.
[26] K. E. Atkin, R. Reiss, N. J. Turner, A. M. Brzozowski, G. Grogan, Acta Crys-
tallogr. Sect. F Struct. Biol. Cryst. Commun. 2008, 64, 182–185.
[27] M. Hall, C. Stueckler, W. Kroutil, P. Macheroux, K. Faber, Angew. Chem.
2007, 119, 4008–4011; Angew. Chem. Int. Ed. 2007, 46, 3934–3937.
[28] H. Golchoubian, F. Hosseinspoor, Molecules 2007, 12, 304–311.
[29] A. M. Brzozowski, J. Walton, J. Appl. Crystallogr. 2001, 34, 97–101.
[30] Z. Otwinowski, W. Minor, Methods Enzymol. 1997, 276, 307–326.
[31] F. Long, A. A. Vagin, P. Young, G. N. Murshudov, Acta. Crystallogr. Sect. D.
Biol. Crystallogr. 2008, 64, 125–132.
Received: January 4, 2012
[32] H. H. Hernandez, O. A. Jaquez, M. J. Hamill, S. J. Elliott, C. L. Drennan,
Biochemistry 2008, 47, 9728–9737.
Published online on March 13, 2012
878
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2012, 13, 872 – 878