Discovery of Baeyer-Villiger monooxygenases from photosynthetic eukaryotes

Article abstract of DOI:10.1016/j.molcatb.2013.10.006

Baeyer-Villiger monooxygenases are attractive "green" catalysts able to produce chiral esters or lactones starting from ketones. They can act as natural equivalents of peroxyacids that are the catalysts classically used in the organic synthesis reactions, consisting in the cleavage of CC bonds with the concomitant insertion of an oxygen atom. In this study, two type I BVMOs have been identified for the first time in photosynthetic eukaryotic organisms, the red alga Cyanidioschyzon merolae (Cm) and the moss Physcomitrella patens (Pp). A biocatalytic characterization of these newly discovered enzymes, expressed in recombinant forms, was carried out. Both enzymes could be purified as holo enzymes containing a FAD cofactor. Their thermostability was investigated and revealed that the Cm-BVMO is the most thermostable type I BVMO with an apparent melting temperature of 56 C. Substrate profiling revealed that both eukaryotic BVMOs accept a wide range of ketones which include aromatic, aliphatic, aryl aliphatic and bicyclic ketones. In particular, linear aliphatic ketones (C9 and C12), carrying the keto functionality in different positions, resulted to be the best substrates in steady state kinetic analyses. In order to restore the BVMO-typifying sequence motif in the Pp-BVMO, a mutant was prepared (Y160H). Intriguingly, this mutation resulted in higher activities on most tested substrates. The recombinant enzymes displayed k_cat values in the 0.1-0.2 s^-1 range, which is relatively low when compared with other known type I BVMOs. This may hint to a role in secondary metabolism in these photosynthetic organisms, though their exact function remains to be established.

Full text of DOI:10.1016/j.molcatb.2013.10.006

Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Discovery of Baeyer–Villiger monooxygenases from
photosynthetic eukaryotes
Elisa Beneventia, Mattia Nieroa, Riccardo Motterleb,
Marco Fraaijec, Elisabetta Bergantinoa,∗
a Dipartimento di Biologia, Università di Padova, Viale G.Colombo 3, 35131 Padova, Italy
b F.I.S.-Fabbrica Italiana Sintetici S.p.A., Viale Milano 26, 36075 Alte di Montecchio Maggiore, Vicenza, Italy
c Biomolecular Sciences and Biotechnology Institute, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Baeyer–Villiger monooxygenases are attractive “green” catalysts able to produce chiral esters or lactones
starting from ketones. They can act as natural equivalents of peroxyacids that are the catalysts classically
used in the organic synthesis reactions, consisting in the cleavage of C C bonds with the concomitant
insertion of an oxygen atom.
Received 31 July 2013
Received in revised form
23 September 2013
Accepted 5 October 2013
Available online 21 October 2013
In this study, two type I BVMOs have been identified for the first time in photosynthetic eukaryotic
organisms, the red alga Cyanidioschyzon merolae (Cm) and the moss Physcomitrella patens (Pp). A biocat-
alytic characterization of these newly discovered enzymes, expressed in recombinant forms, was carried
out. Both enzymes could be purified as holo enzymes containing a FAD cofactor. Their thermostability
was investigated and revealed that the Cm-BVMO is the most thermostable type I BVMO with an appar-
ent melting temperature of 56 ◦C. Substrate profiling revealed that both eukaryotic BVMOs accept a wide
range of ketones which include aromatic, aliphatic, aryl aliphatic and bicyclic ketones. In particular, linear
aliphatic ketones (C9 and C12), carrying the keto functionality in different positions, resulted to be the
best substrates in steady state kinetic analyses. In order to restore the BVMO-typifying sequence motif in
the Pp-BVMO, a mutant was prepared (Y160H). Intriguingly, this mutation resulted in higher activities
on most tested substrates. The recombinant enzymes displayed kcat values in the 0.1–0.2 s−1 range, which
is relatively low when compared with other known type I BVMOs. This may hint to a role in secondary
metabolism in these photosynthetic organisms, though their exact function remains to be established.
© 2013 Elsevier B.V. All rights reserved.
Keywords:
Baeyer–Villiger monooxygenases
Photosynthetic eukaryotes
Recombinant enzymes
Biocatalysis
1. Introduction
the corresponding ester or lactone; one of the group attached to
the carbonyl carbon migrates onto the electron deficient oxygen
The Baeyer–Villiger reaction is an important oxidative reaction
in organic synthesis firstly reported by Adolf von Baeyer and
Victor Villiger in 1899 [1]. In this reaction, ketones are oxidized
into the corresponding esters or lactones by peroxyacids resulting
in oxygen insertion next to the carbonyl group. Reagents able
to carry out the Baeyer–Villiger reaction are peroxyacids such
as meta-chloroperoxybenzoic acid, trifluoroperoxyacetic acid,
peroxyacetic acid, hydrogen peroxide and others. Such molecules
are expensive, toxic and hazardous; they are strong oxidants and
react with other functional groups [2]. The reaction mechanism
consists in the nucleofilic attack of peroxyacid to the carbonyl
function of the substrate forming a tetrahedral intermediate called
the “Criegee intermediate”, which undergoes rearrangement to
atom with the simultaneous dissociation of the O O bond. The
regiochemistry of the reaction depends on the relative migratory
ability of the substituents attached to the carbonyl group and on
the stereoelectronic features of the substrate [3]. Moreover, if the
migrating carbon is chiral the stereochemistry is retained [4].
Baeyer–Villiger monooxygenases (BVMOs) are flavin-
containing monooxygenases able to catalyze a remarkable wide
variety of oxidative reactions [5], which are difficult to obtain
chemically: the nucleophilic oxidation of carbonyl groups but also
the electrophilic oxidation of heteroatoms such as sulphur [6,7],
nitrogen [8] and boron [9], with good enantioselectivity. Complete
enantioselectivity has been reported for the epoxidation of few
selected olefin substrates too [10]. Indeed, the great potential of
these enzymes is the ability to produce chiral lactones and sulfox-
ides by asymmetric synthesis and this makes them attractive for
application in industrial organic synthesis [11,12].
∗
Corresponding author. Tel.: +39 0498276342; fax: +39 0498276300.
E-mail addresses: elisa.beneventi@gmail.com (E. Beneventi),
BVMOs contain a flavin cofactor, FAD or FMN, not covalently
but tightly bound to the enzyme. This flavin cofactor has to be acti-
vated by electron donors, NADH or NADPH, which give reduction
mattia.niero87@gmail.com (M. Niero), riccardo.motterle@fisvi.com (R. Motterle),
m.w.fraaije@rug.nl (M. Fraaije), elisabetta.bergantino@unipd.it (E. Bergantino).
1381-1177/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcatb.2013.10.006

146
E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
equivalents allowing molecular oxygen binding. Depending on
cofactor usage, BVMOs can be classified into two types [13].
Type I BVMOs are FAD and NADPH dependent and contain two-
dinucleotide binding domains (␤␣␤-folds) known as Rossman
motifs, one for the FAD binding and the other one for the NADPH
binding. They are composed of only one polypeptide. Type II BVMOs
are FMN and NADH dependent and consist of two distinct subunits,
a dehydrogenase using NADH to reduce FMN and a second sub-
unit able to perform the Baeyer–Villiger reaction using the reduced
flavin.
FeSO4 7H2O, 500 mg/l NH4 tartrate, 5 g/l glucose, 1 ml KH2PO4/KOH
buffer pH 7, and 1 ml trace solution elements per liter. KH2PO4/KOH
buffer contained 25 g KH2PO4 per 100 ml of milliQ water; pH 7 was
obtained by titrating with 4 M KOH. Trace element solution con-
tained 55 mg/l CuSO4 5H2O, 55 mg/l ZnSO4·7H2O, 614 mg/l H3BO3,
389 mg/l MnCl2·4H2O, 55 mg/l CoCl2·6H2O, 28 mg/l KI, 25 mg/l
Na2MoO4·2H2O. Cultures were grown in a growth chambers at
24 ◦C, with 16 h light/8 h dark photoperiod and a light intensity of
40–50 ␮E. Nutrition media were autoclaved for 20 min at 121 ◦C
and plant agar (7 g/l) was added in order to prepare solid media.
There is another monooxygenase, MtmOIV, able to perform
the Baeyer–Villiger reaction, involved in mithramycin biosynthe-
sis. Mithramycin is a polyketide anticancer antibiotic produced
by the soil bacterium Streptomyces argillaceus (ATCC 12956) and
other streptomycetes [14]. Sequence analysis and crystal structure
revealed that MtmVIO can not be classified as a type I or type II
BVMOs but it appears to be an atypical BVMO belonging to a differ-
ent flavoprotein monooxygenase family [15].
The best-characterized BVMOs are those of type I, which display
good substrate promiscuity and are therefore attractive for syn-
thetic applications [16]. In particular, the most famous enzymes are
cyclohexanone monooxygenase (CHMO) from Acinetobacter NCIMB
9871 and phenylacetone monooxygenase (PAMO) from Thermobi-
fida fusca. The former enzyme was also the first one to be described,
in the middle seventies [17]. Up to ten years ago, however, few
BVMOs had been cloned and overexpressed. Thanks to the vast
genetic information that has become available by genome sequenc-
ing, the number of recombinant BVMOs has increased considerably.
In particular, the technique of genome mining has proved to be
an efficient approach to discover new biocatalysts. Previous stud-
ies performed by this approach showed that type I BVMOs were
present in a vast variety of bacteria and fungi, but no representative
was found in Archea, plants or human genomes [18].
In order to find novel and promising biocatalysts, we used
the genome mining approach to uncover BVMOs within specific
and unusual organisms. Using the sequence of a prototype BVMO,
phenylacetone monooxygenase (PAMO), as template and the “fin-
gerprint” motif for type I BVMOs (FxGxxxHxxxWP/D; [19]) as
discriminant, we have identified two new putative BVMO-encoding
genes in the red alga Cyanidioschyzon merolae (Cm) and in the moss
Physcomitrella patens (Pp). Photosynthetic eukaryotes appear as
very uncommon sources for BVMOs. In fact, only two BVMOs from
eukaryotic origin have been cloned and expressed to date. Both
BVMOs were derived from fungi: the ascomycetes Cylindrocarpon
radicicola ATCC 11011 [20] and Aspergillus fumigatus Af293 [21]. The
discovery of BVMOs in photosynthetic eukaryotes is novel and may
provide new biocatalytic features.
2.2. Reagents and enzymes
All chemicals were purchased from Fluka and Sigma–Aldrich
(Milan, Italy). Restriction enzymes were obtained from New
England Biolabs (NEB) or from Promega; PhusionTM High-fidelity
DNA polymerase from Finnzymes; TSAP (Thermosensitive Alkaline
Phosphatase) from Promega; T4 DNA Ligase from NEB. IMAC-Select
Affinity Gel resin was purchased from Sigma–Aldrich.
2.3. Sequence analysis and cloning
The NCBI resource was used for DNA sequence analysis; searches
and multiple alignments of BVMO sequences were respectively
produced by programs BLAST and ClustalW.
Expression vectors were produced by digestion of pET-28a(+)
with NcoI/NotI and ligation of the amplified Cm-BVMO and Pp-
BVMO sequences, cut by the same enzymes. These sequences
were obtained by two subsequent PCR reactions, the first one
producing
a preliminary “large” amplimer, which was then
used for the second amplification, by nested mutagenic primers.
The following synthetic oligonucleotides were used for PCR
amplification of the target gene from C. merolae genomic DNA:
external primer pair, forward 5ꢀ-AGTGATGCGCGTGGCCGGCA-3ꢀ
and reverse 5ꢀ-AGGTGTCTGCACCTCGCCAGCG-3ꢀ; nested primer
pair,
forward
5ꢀ-TTTGACCGGCCATGGGAGCGGAGCTCAAC-3ꢀ
and reverse 5ꢀ-GCATCCACCGCGCCGGCGTACAGCGAAG-3ꢀ. The
following synthetic oligonucleotides were used for PCR amplifi-
cation of the target gene from P. patens genomic DNA: external
primer pair, forward 5ꢀ-ACAGGCCACGGGGGTAGTTCTGTTG-
3ꢀ
and
reverse
5ꢀ-CAACCCTGGACAGCATCGGAAGCCT-3ꢀ;
nested primer pair, forward 5ꢀ-AAGTATGTCCAATTCCATGGC-
TGAGTTCGATGCTGTTATAGTCGGAG-3ꢀ and reverse 5ꢀ-AAACAAT-
GCCCGCGGCCGCCAGCTTGAATCCC-3ꢀ. The sequence variant
Y160H was constructed using the QuikChange® II Site-Directed
Mutagenesis Kit of Stratagene using plasmid pET28 Pp-BVMO
as template with the oligonucleotide 5ꢀ-GGCTCATCGTACCA-
CACGGGC-3ꢀ and its complementary one.
2. Experimental
2.4. Expression, analysis and purification of recombinant proteins
2.1. Organisms and culture conditions
The recombinant enzymes, hereafter called Cm-BVMO, Pp-
BVMO and Pp-Y160H-BVMO, were expressed in E. coli BL21 (DE3)
or E. coli ARCTIC Express® (Stratagene). Pre-cultures were carried
out in 5 ml LB medium at 37 ◦C containing 50 ␮g/ml kanamycin.
Larger cultures were carried out in 1 L LB medium with riboflavin
addition (100 ␮M). The cells were grown in a shaking incubator
at 37 ◦C to an optical density at 600 nm (OD600) of 0.4–0.6, then
induced by addition of IPTG to a final concentration of 0.2 mM and
cultivated at 18 ◦C overnight. The cells were harvested by centrifu-
gation (4 ◦C, 10 min, 4500 × g) and washed with Tris/HCl buffer (Tris
50 mM, pH 8.0). Cell disruption was obtained by French Press and
crude extract was centrifuged (4 ◦C, 20 min, 15,000 × g) to separate
soluble e insoluble fractions. FAD cofactor (at 100 ␮M final concen-
tration) was added to the crude extract before cell disruption.
Cyanidioschyzon merolae strain 10D (NIES-1332) was obtained
from the Microbial Culture Collection of the National Institute
for Environmental Studies, Tsukuba, Japan. Cultures of C. merolae
were grown at 30 ◦C in M-Allen medium (http://mcc.nies.go.jp), in
500 ml flasks on a rotatory platform shaker at 70 rpm. Light condi-
tions used were 25–27 ␮mol photons m−2 sec−1
.
E. coli strains (XL1-blue and BL21 (DE3)) were routinely cul-
tured in LB media and supplemented with the antibiotic kanamycin
(50 ␮g/ml).
The starting culture of Physcomitrella patens was kindly pro-
vided by Dr. Tomas Morosinotto (Department of Biology, University
of Padova). It was cultivated by micro propagation in PpNH4 rich
medium: 0.8 g/l CaNO3 4H2O, 0.25 g/l MgSO4 7H2O, 0.0125 g/l

E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
147
Expression of recombinant protein was checked by SDS-PAGE
analysis. The identity of the proteins was verified by immunoblot-
ting using anti His-tag antibodies. Overexpressed proteins were
purified by immobilized-metal affinity chromatography (IMAC).
Soluble fractions obtained from 1 l culture were incubated with the
resin for 1 h at 4 ◦C and then loaded on a 5 ml column (Bio-Rad).
The column was washed by gravitational flow with three column
volumes of 50 mM Tris/HCl, 300 mM NaCl buffer, then with two
column volumes of the same buffer with added 5 mM imidazole.
Elution was performed by five column volumes of 50 mM Tris/HCl,
300 mM imidazole solution. Enzyme concentration was measured
by spectrophotometry, measuring concentration of free flavin in a
solution of denatured protein and calculated as previously reported
[22].
Physcomitrella patens (a moss). Both identified BVMO sequences
also contained other conserved regions, such as two Rossmann
fold motifs (GXGXX(G/A)) that are known to be involved in dinu-
cleotide cofactor binding [25] and the recently annotated motif
[A/G]GxWxxxx[F/Y]P[G/M]xxxD, which includes residues of the
active site and therefore represents a BVMO-typifying consensus
sequence [24]. A ClustalW alignment of the two proteins with
PAMO and cyclohexanone monooxygenase (CHMO) from Acineto-
bacter NCIMB 9871 is shown in Fig. 1.
With the aim of considering the evolutionary relationships of
the two proteins with other known BVMOs, we selected a set of
24 prototype enzymes among those that are well characterized
as recombinant products and representative for substrate speci-
ficities. We aligned thus 26 protein sequences and produced the
branching diagram shown in Fig. 2. It resulted that the sequence
from C. merolae most likely shares a common ancestor with acetone
monooxygenase (ACMO) from the actinobacterium Gordonia sp.
TY-5 and methyl ketone monooxygenase (MEKMO) from the pro-
teobacterium Pseudomonas veronii, strain MEK700. The protein of
P. patens forms a cluster with two other enzymes of actinobacterial
origin: PAMO and steroid monooxygenase (STMO) from Rhodococ-
cus rhodochrous. The observed homologies were considered for the
discussion below.
The intron-less coding gene for the putative BVMO from C. mero-
lae is present in chromosome 12 of the organism. The primary
structure of the translated protein, which is annotated in GenBank
as a steroid monooxygenase (BAM80902.1), shares 43% identity
with the sequence of PAMO, 37% with cyclohexanone monooxyge-
nase (CHMO) from Acinetobacter NCIMB 9871 and 50% with CHMO
from Rhodococcus Phi1.
The 536 amino acids long putative enzyme from P. patens (pre-
dicted protein XP 001758613.1 in the NCBI data bank) is also coded
by a continuous open reading frame of the nuclear genome; it
shows 54% sequence identity with PAMO and 44% with CHMO from
Acinetobacter NCIMB 9871. One of the two fingerprint motifs for
type I BVMO is, however, not strictly conserved but differs in one
amino acid: FxGxxxYxxxWP instead of FxGxxxHxxxWP (Fig. 1).
This is noteworthy, since it is known that the central histidine of
this consensus sequence (His173 in PAMO) is involved in cataly-
sis and FAD binding [26]. For this reason we planned to study the
wild-type form of the putative flavoenzyme in comparison with
a mutant form in which the consensus motif would be artificially
repaired.
The identified coding sequences together with the variant
obtained by site-directed mutagenesis were cloned into the
pET28a(+) expression vector, permitting their translation in the
form of fusion proteins with carboxy-terminal hexahistidine tags.
The recombinant enzymes were designated Cm-BVMO (from C.
merolae), and Pp-BVMO (from P. patens) and Pp-Y160H-BVMO.
Overexpression tests were then performed in E. coli BL21(DE3)
strain and parameters such as temperature and IPTG concentra-
tion were optimized. In all cases the proteins were expressed at
very high level and were mostly present in the cell debris after cell
homogenization, indicating insolubility. To overcome this problem,
different known strategies were unsuccessfully tried: expression
at low temperature in a bacterial strain containing cold-adapted
chaperones (E.coli ARCTIC Express®) or fusion with partners favor-
ing solubility (bacterial MBP and yeast SUMO) (data not shown).
Only upon addition to the culture medium of an excess of riboflavin,
the precursor of FAD, bright yellow colored recombinant enzymes
could be recovered in the soluble fraction of the homogenate. This
result underlines the requirement of the flavin cofactor for correct
folding of BVMOs.
2.5. Activity assay, kinetics and ThermoFAD measurements
BVMO activity is usually determined spectrophotometri-
cally by monitoring the consumption of NADPH at 340 nm
(ε = 6.22 mM−1 cm−1). Reaction mixtures (1 ml) contained 50 mM
Tris–HCl buffer pH 8.0, 100 ␮M NADPH, 0.4 ␮M pure enzyme, and
10 ␮l of 1 mM of rac-bicyclo[3.2.0]hept-2-en-6-one in dioxane.
Adding the enzyme to the mixture started the reaction. One unit
of BVMO is defined as the amount of protein that oxidizes 1 ␮mol
NADPH per minute.
Steady-state kinetic parameters of the different substrates
were determined using purified enzymes and substrate concen-
trations ranging from 0 to 5 mM. Data were fitted using the
Michaelis–Menten equation by the program SigmaPlot.
The apparent unfolding temperatures of the recombinant
enzymes, Tms, were determined using the ThermoFAD method
[23]. 30 ␮l of 1 mg/mL protein in Tris-HCl buffer (50 mM, pH 7.5)
were loaded in a Real Time PCR machine (Eppendorf) fitted with
a 470–543 nm excitation filter and a SYBR Green emission filter
(523–543 nm). A temperature gradient from 20 to 90 ◦C was applied
(1 ◦C/min), and fluorescence data were recorded. A sigmoidal curve
was obtained after plotting the fluorescence amount against the
temperature. The Tm values were determined as the maximum of
the derivative of the obtained sigmoidal curve
2.6. Substrate screening
The phosphate activity assay was performed according to the
protocol reported by Riebel [24]. All substrates were dissolved in
dioxane (maximum 5%) and tested in 96 wells plates.
2.7. Conversions and GC/GC–MS analysis
For GC and GC–MS analysis, samples of 500 ␮l 50 mM Tris–HCl
(pH 7.5) containing 2 mM substrate, 5% dioxane, 100 ␮M NADPH,
3.0 ␮M PTDH, 10 mM phosphite and 1 ␮M BVMO were incubated
shaking at 25 ◦C from 1 to 20 h. The reactions were stopped by
extraction with ethyl acetate (3 × 0.5 ml, including 0.1% mesity-
lene as an internal standard), dried with magnesium sulfate and
analyzed directly by GC or GC–MS [5] to determine the degree of
conversion.
3. Results
3.1. Two new BVMOs from photosynthetic eukaryote organisms
By using the protein sequence of PAMO as template for a
tBLASTn search and the consensus motif FxGxxxHxxxWP [19] as
discriminant (allowing just single conservative mutation), two new
putative type I BVMO enzymes were identified from two pho-
tosynthetic eukaryotes: Cyanidioschyzon merolae (a red alga) and
The proteins were purified to homogeneity by IMAC chromatog-
raphy (Fig. 3); the evaluated yields were 45 mg for Cm-BVMO,
9.5 mg for Pp-BVMO and 10 mg for Pp-Y160H-BVMO per liter

148
E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
Fig. 1. CLUSTALW Multiple sequence alignment of new BVMOs, PAMO and CHMO. CLUSTALW Multiple sequence alignment of phenylacetone monooxygenase from Thermob-
ifida fusca (PAMO), cyclohexanone monooxygenase from Acinetobacter NCIMB 9871 (CHMO) and novel BVMOs from Physcomitrella patens (Pp-BVMO) and Cyanidioschyzon
merolae (Cm-BVMO). The two Rossmann fold domains (GxGxx(G/A), ATG) and the two BVMO fingerprint motifs (FxGxxxHxxxW [20] and GGxWxxxxYPGxxxD [24]) are
highlighted. Number of omitted terminal residues is in brackets.
of bacterial culture. They were tested in reactions with rac-
bicyclo[3.2.0]hept-2-en-6-one as substrate, a benchmark molecule
for Baeyer–Villiger monooxygenase activity. Conversion of the
substrate, as observed by consumption of NADPH, confirmed the
nature of the newly discovered enzymes and prompted their fur-
ther characterization.
led to inactivation of the enzyme. Considering that the red alga
typically grows at 45 ◦C, this result was not unexpected. Unfol-
ding experiments using the ThermoFAD method revealed that the
melting temperature (Tm) of Cm-BVMO was 56 ( 0.5) ◦C, which
is considerably higher than that of most known BVMOs. To our
knowledge, the only other thermostable type I BVMO found so
far is PAMO, i.e. the enzyme discovered by Fraaije et al. [22] in
Thermobifida fusca, that exhibits an activity half-life of 24 h at 52 ◦C
and a Tm of 61 ◦C.
3.2. pH and temperature profiles
Optimal temperatures for Pp-BVMO and Pp-Y160H-BVMO were
probed between 17 ◦C and 40 ◦C: they resulted around 25 ◦C and
30 ◦C respectively for Pp-BVMO and Pp-Y160H-BVMO. Concern-
ing stability, after 1.5 h incubation between 17 ◦C and 25 ◦C both
enzymes retained their activity. Tm of Pp-BVMO and mutant Pp-
Y160H-BVMO were 44.0 ( 0.5) ◦C and 43.5 ( 0.5) ◦C respectively.
Apparently, the restored histidine residue in the fingerprint motif of
the mutant did not significantly influence the melting temperature
of the enzyme.
All recombinant type I BVMOs showed a bell-shaped pH pro-
file in reactions with rac-bicyclo[3.2.0]hept-2-en-6-one. Maximum
activity was measured at pH 8.5 for Cm-BVMO; at pH 8.0 for
Pp-BVMO and Pp-Y160H-BVMO. The latter optimum pH value cor-
responds to that found for PAMO, whereas pH 7.5 and 9 are the
values respectively reported for HAPMO (4-hydroxyacetophenone
monooxygenase) and CHMO (cyclohexanone 1,2-monooxygenase;
[27]).
Temperature optima were measured between 17 ◦C and 70 ◦C.
Enzymatic activity of Cm-BVMO increased with temperature up to
70 ◦C: at this temperature initial enzymatic activity was five-fold
higher compared to that measured at 30 ◦C. The enzyme from C.
merolae also demonstrated considerable thermostability, since its
activity was fully retained in incubation for 1.5 h in the temperature
range between 20 ◦C and 42 ◦C. Higher incubation temperatures
3.3. Substrate scope
To explore the biocatalytic potential of the two eukaryotic
BVMOs, a broad range of potential substrates was used in a so-called
phosphate assay [24]. In this indirect assay the Baeyer–Villiger

E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
149
CHMO_Rhodococcus_HI-31
CHMO_Xanthobacter
CHMO2_Rhodococcus
CHMO_Acinetobacter
CHMO1_Rhodococcus
CAMO
BVMO_Af1
CPMO
CHMO_Arthrobacter
CHMO_Arthrobacter_L661
CHMO2_Brevibacterium
CHMO_Brachymonas
Pp-BVMO
Cm-BVMO
MEKMO
PAMO
STMO
ACMO
OTEMO
HAPMO_P.putida
HAPMO_P.fluorescens
CHMO1_Brevibacterium
PtlE
EtaA
CPDMO
CDMO
Fig. 2. Phylogenetic analysis of Cm-BVMO, Pp-BVMO and selected BVMOs. MUSCLE multialignment of 26 type I BVMOs (selected as described in Section 3.1) was used as
input into MEGA program version 5.0 [40] in order to generate a phylogenetic tree using the neighbor-joining method [41] with an associated bootstrap analysis, based
on 1000 replications. The tree was visualized using Dendroscope 3 [42]. Abbreviations and GenBank accession numbers of protein sequences: CHMO Acinetobacter, CHMO
Acinetobacter sp. NCIMB 9871 (BAA86293); CHMO1 Brevibacterium, CHMO 1 Brevibacterium sp. HCU (AAG01289); CHMO2 Brevibacterium, CHMO 2 Brevibacterium sp. HCU
(AAG01290); CHMO1 Rhodococcus, CHMO Rhodococcus sp. Phi1 (AAN37494); CHMO2 Rhodococcus, CHMO Rhodococcus sp. Phi2 (AAN37491); CHMO Arthrobacter, CHMO
Arthrobacter sp. BP2 (AAN37479); CHMO Brachymonas, CHMO Brachymonas petroleovorans (AAR99068); CHMO Xanthobacter, CHMO Xanthobacter sp. ZL5 (CAD10801);
CPMO, cyclopentanone monooxygenase Comamonas sp. NCIMB 9872 (BAC22652); PAMO, phenylacetone monooxygenase Thermobifida fusca (PDB: 1W4X A); HAPMO P.
fluorescens, 4-hydroxyacetophenone monooxygenase Pseudomonas fluorescens (AAK54073); STMO, steroid monooxygenase Rhodococcus rhodochrous (BAA24454); CDMO,
cyclododecanone monooxygenase Rhodococcus rubber (AAL14233); EtaA, EtaA Mycobacterium tuberculosis H37Rv (CAB06212); CPDMO, cyclopentadecanone monooxygenase
Pseudomonas sp. HI-70 (BAE93346); ACMO, acetone monooxygenase Gordania sp. strain TY-5 (BAF43791); MEKMO, methylethylketone monooxygenases Pseudomonas veronii
MEK700 (ABI15711); OTEMO, ꢀ(3)-4,5,5-trimethylcyclopentenylacetyl-coenzyme A monooxygenase Pseudomonas putida ATCC 17453 (H3JQW0); CAMO, cycloalkanone
monooxygenase Cylindrocarpon radicicola ATCC 11011 (AET80001); BVMO AF1, BVMO Aspergillus fumigatus Af293 (NCBI: XM 742067); PtlE, BVMO Streptomyces avermitilis
MA-4680 (NP 824170); CHMO Arthrobacter L661, CHMO Arthrobacter sp. L661 (ABQ10653); CHMO Rhodococcus HI-31, CHMO Rhodococcus sp. HI-31(BAH56677); HAPMO P.
putida, HAPMO Pseudomonas putida (ACJ37423).
oxidation is coupled with a regeneration reaction catalyzed by
phosphite dehydrogenase (PTDH). PTDH is able to generate one
molecule of phosphate from a molecule of phosphite in water,
with simultaneous regeneration of one molecule of NADPH. The
formed phosphate can be quantified by using a chromogenic reac-
tion that allows spectrophotometric detection of the degree of
conversion.
By using this assay, a collection of 46 potential substrates for
BVMOs was tested. They were chosen among different classes of
molecule, e.g. linear and cyclic aliphatic ketones, aryl and aromatic
ketones, steroids, aromatic amines, and aromatic sulphides. The
substrate scope obtained for each enzyme is summarized in Table 1.
In general, all the three BVMOs seem to have a broad sub-
strate scope. As expected, they exhibited significant activity against
Fig. 3. Expression and purification of His-tagged Pp- and Cm-BVMOs. 12% SDS-PAGE analysis of cell extracts from E. coli cells expressing Pp-BVMO (A) and Cm- BVMO (B).
Fractions from progressive purification fractions were loaded: protein standard marker (M), total cell extracts from non induced cells (NI) or over night induced cells (ON),
pellet fraction (p) and soluble protein fraction (sur), flow trough (Ft) and elution fractions (E).

150
E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
Table 1
Substrate scope.
3.4. Steady-state kinetics
Steady-state kinetic parameters of the three recombinant
enzymes were determined in 50 mM Tris/HCl using the optimal
conditions defined for each enzyme, that are pH 8.0, at 25 ◦C for Pp-
and Pp-H160Y BVMOs and pH 8.5, at 35 ◦C for Cm-BVMO. All the
three enzymes showed a typical Michaelis–Menten behavior. The
kinetic parameters kcat and KM were calculated for a set of iden-
tified substrates besides rac-bicyclo[3.2.0]hept-2-en-6-one, used
in the preliminary assay. Selected substrates were linear aliphatic
ketones, cyclic aliphatic ketones and aromatic ketones. In particu-
lar, for the linear ketone containing eight carbon atoms, the position
of the keto group in the molecule was investigated. Tables 2 and 3
summarize the results obtained for Cm-BVMO, and Pp- and Pp-
Y160H BVMOs, respectively.
Cm-BVMO exhibited the best catalytic efficiency against 2-
dodecanone (kcat/KM 90.8 s−1 mM−1). Octanones also appeared to
be good substrates. In particular, by comparing KM and kcat/KM val-
ues reported in Table 2, the position of the keto group at C3 seemed
to be preferred. By limiting the observation to the positions of the
carbonyl function in octanone, it can be noticed that the Michaelis
constant is the main parameter influencing the catalytic efficiency,
since the kcat values were similar for all investigated positions of
the keto groups.
Also for Pp- and Pp-Y160H BVMOs, linear ketones seemed to be
very well accepted substrates (Table 3). 2-dodecanone was the one
for which the enzymes display the highest catalytic efficiency. Dif-
ferently from the enzyme from C. merolae, the one from P. patens
preferred the keto group at position 2 of octanone. 4-octanone was
not a good substrate for either wt or mutant enzymes, as already
observed in the previous substrate-screening assay (Table 1). kcat
values, for all the four linear C8-ketones tested, were approximately
in the same range of those measured for the algal enzyme. As
discussed above in relation to the activity and substrate profile,
restoration of the consensus histidine makes the mutant variant
of Pp-BVMO more competent for catalysis and faster than the wt
form, showing higher turnover number.
All the recombinant enzymes were able to oxidize some substi-
tuted aryl ketones, like phenylacetone and 4-(4-hydroxyphenyl)-2-
butanone. The latter compound was a good substrate, particularly
for Pp- and Pp-Y160H BVMOs. Due to the limited solubility,
other substrates such as 4-dimethylaminobenzaldehyde and 3-
acetylindole could not be investigated by kinetic analysis. Finally,
the affinity for coenzyme NADPH was found to be very high, as a
KM values <5 ␮M were determined.
Substrate
Cm
Pp
PpY160H
Acetone
Methylketone
Methyl vinyl ketone
2-Octanone
3-Octanone
4-Octanone
2-Dodecanone
3-Methyl-2,4-pentanedione
Cyclobutanone
Cyclopentanone
+
+
+
++
+
+
+++
+
+
++
+
+
+
++
++
+
+++
+
++
+++
+
+
++
+
+
+++
+
+
+
+
Cyclohexanone
Cyclopentadecanone
2-Oxocyclohexanecarbonitrile
4-Methylcyclohexanone
2-Propylcyclohexanone
Dihydrocarvone
Cyclopropylmethylketone
Norcamphor
Bicycloheptenone
Progesterone
++
+
+
+
+
+
++
+++
+
+
+++
+
+
++
++
Androstenedione
4-Dimethylaminobenzaldehyde
Nicotin
+
++
Thioanisole
+
++
Benzylethyl sulfide
Benzylphenyl sulfide
Ethionamide
Diphenylmethylthioacetamide
Thiacetazone
Indole
3-Acetylindole
5-Mathylfurfural
Benzaldehyde
Acetophenone
4-Hydroxyacetophenone
2,6-Dihydroacetophenone
3-Phenylpentane-2,4-dione
Phenylacetone
4-(4-Hydroxyphenyl)-2-butanone
2-Phenylcyclohexanone
Benzoin
Phenindione
2-Indanone
1-Indanone
6-Hydroxy-1-indanone
+++
+
+++
+
+
+++
++
+
+
++
+
+++
+++
+++
+
+
+
++
+
+
+
++
+++
+++
++
++
+++
+
+
+
+
+
+
Observed activities for Cm-BVMO, Pp-BVMO and Pp-Y160H-BVMO, measured by
phosphate formation. Activities are indicated as +, ++ or +++ and reflect a 1.2-, 2- or 5-
fold increase in phosphate formation, respectively, when compared with uncoupling
reactions that lack the tested compound.
3.5. Conversions
rac-bicyclo[3.2.0]hept-2-en-6-one, the above-mentioned typical
substrate for type I BVMOs. Good activities were also observed
for linear ketones, in particular long-chain alkanones. Phenylace-
tone and some derivatives also seemed to be easily accepted as
substrates, as well as some cyclic ketones like cyclopentanone. It
is worth noting that, by directly comparing results collected for
Pp-BVMO and for Pp-Y160H-BVMO, the mutant displays higher
activity and larger substrate profile. This result is in agreement
with the fundamental role reported for the central histidine of the
BVMO consensus motif in catalysis and FAD binding. Indeed, in site-
directed mutants of 4-hydroxyacetophenone monooxygenase from
Pseudomonas fluorescens (His296Ala; [19]) and of cyclohexanone
monooxygenase from Acinetobacter NCIB 9871 (His163Gln; [28]),
it was shown that substitution of this residue drastically reduces
or nearly abolishes activity.
Biotransformations using purified enzymes and different types
of substrates were investigated using GC and GC-MS analysis in
order to detect both substrates and products. PTDH and phosphite
were added for NADPH regeneration.
As already mentioned, rac-bicyclo[3.2.0]hept-2-en-6-one is a
standard probe for testing the biocatalytic potential of type I
BVMOs. This conversion is a good example of regiodivergent
parallel kinetic resolution (PKR) because the racemic substrate
leads to two regioisomeric compounds (−)-(1S,5R) 2 and (−)-
(1R,5S) 3 (Fig. 4A). One compound originates from the favored
Baeyer–Villiger-type oxygen insertion between the more substi-
tuted carbon atom and the carbonyl group in a fast reaction; the
second lactone is more slowly produced in the chemically disfa-
vored regiochemistry. The two chiral lactones are relevant building
blocks for the chemical synthesis of prostaglandins; their pro-
duction has been therefore studied in conversion experiments by
using different strains, i.e. different BVMOs. Some examples of
such studies are those regarding CHMO from Acinetobacter [29],
Other potential substrates tested with the new BVMOs, such as
acetone, indanone, phenindione, cyclopentadecanone, benzoin and
steroids, were found to be poor or not accepted substrates.

E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
151
Table 2
Steady-state kinetic analysis of Cm-BVMO.
Substrate
KM [mM]
kcat [s−1
]
kcat/KM [s−1 mM]
O
O
2-Dodecanone
0.004 0.001
0.383 0.018
90.8
25.3
22.7
O
Octanal
0.010 0.008
0.011 0.006
0.252 0.043
0.250 0.031
2-Octanone
O
3-Octanone
4-Octanone
0.0030 0.0005
0.0040 0.0009
0.210 0.006
0.202 0.006
70
O
50.5
4-(4-Hydroxyphenyl)-2-butanone
Bicyclo[3.2.0]hept-2-en-6-one
0.0070 0.0007
0.0040 0.0007
0.079 0.002
0.085 0.002
11.2
21.3
HO
O
O
O
O
3-Phenylpentane-2,4-dione
Phenylacetone
0.099 0.026
0.097 0.071
0.015 0.001
0.032 0.006
0.2
0.3
O
cyclopentanone monooxygenase (CPMO) from Pseudomonas sp.
NCIMB 9872 [30] and 4-hydroxyacetophenone monooxygenase
(HAPMO) from Pseudomonas fluorescens ACB [5].
ee values suggested poor enantioselectivity toward this racemic
substrate. “Normal” lactones for both fast and slow reactions were
mainly produced.
In our experiments, biooxidation of the rac-bicyclo[3.2.0]hept-
2-en-6-one racemic substrate led to the production of all four
possible lactone products by the three BVMOs, and the observed
The behavior of Cm-BVMO revealed (as shown for HAPMO [5]),
a significant preference for the (+)-(1R,5S) 1 enantiomer. Regio-
selectivity was scarce for both the preferred and non-preferred
Table 3
Steady-state kinetic analysis of Pp- and Pp-Y160H BVMOs.
Substrate
Pp
PpY160H
KM [mM]
kcat [s−1
]
kcat/KM [s−1 mM] KM [mM]
kcat [s−1
]
kcat/KM [s−1 mM]
O
O
2-Dodecanone
Octanal
0.00040 0.00008
0.019 0.009
0.096 0.002 240
0.0020 0.0004 0.252 0.011 126
O
0.061 0.004
0.060 0.002
3.2
20
0.0030 0.0008 0.125 0.006
0.0030 0.0006 0.204 0.007
41.7
68
2-Octanone
0.0030 0.0008
O
3-Octanone
4-Octanone
0.010 0.005
–
0.075 0.008
–
7.5
0.016 0.007
–
0.123 0.011
–
7.6
O
–
–
4-(4-Hydroxyphenyl)-
2-butanone
0.002 0.001
0.068 0.004
34
0.0030 0.0007 0.152 0.006
50.7
HO
O
O
Bicyclo[3.2.0]hept-2-
en-6-one
0.0060 0.0001
0.038 0.016
0.058 0.0001
0.077 0.009
9.7
0.016 0.002
0.032 0.009
0.120 0.005
0.146 0.011
7.5
4.6
Phenylacetone
2
O

152
E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
Table 4
Regiodivergent biooxidation by Cm-, Pp- and Pp-Y160H BVMOs.
Substrate
Cm
Pp
PpY160H
% Conv ratio
% ee
% Conv ratio
% ee
% Conv ratio
% ee
H
O
H
Fast reaction
100 73:27
100 76:24
100 72:28
12 (+)-(1R,5S)2
37 (+)-(1S,5R)3
7 (+)-(1R,5S)2
69 (+)-(1S,5R)3
14 (+)-(1R,5S)2
68 (+)-(1S,5R)3
H
O
H
Slow reaction
100 88:12
66 62:4
100 95:5
% Conv, percentage yield of conversion; ratio, ratio between produced normal and abnormal lactones; % ee, percentage enantiomeric excess values.
substrate enantiomers, that respectively produced 73% (−)-(1S,5R)
2 versus 27% (+)-(1S,5R) 3, and 88% (+)-(1R,5S) 2 and 12% (+)-(1R,5S)
3, enantiomers. All four possible lactone products, in different
amounts, were therefore finally obtained (Fig. 4B). The ee values
of the product mixture resulting from complete conversion were
12% for the “normal” (+)-(1R,5S) 2 and 37% for the “abnormal”
(+)-(1S,5R) 3. (Table 4). After 5 h, the racemic substrate was fully
converted.
Pp- and Pp-Y160H BVMOs approximately showed the same
behavior (Fig. 4C and D) (Table 4). The favored substrate was the
(+)-(1R,5S) 1 enantiomer for both enzymes; they respectively pro-
duced 76% and 72% of the expected (−)-(1S,5R) 2 versus 24% and
28% of the (+)-(1S,5R) 3. The mutant enzyme, compared to the
wild-type one, exhibited much higher regioselectivity in the slow
conversion of the (−)-(1S,5R) 1 substrate, yielding 95% of the
expected lactone and 5% of the unexpected one. In such slow reac-
tion, the WT enzyme yielded only 62% of the expected and 4%
unexpected lactone and the conversion was not complete (66%)
after 16 h. As for the case of the conversion by Cm-BVMO, ee values
for the “normal” and “abnormal” enantiomeric products were too
low to be considered for a direct biocatalytic application of the new
enzymes.
Conversions by Cm-, Pp- and Pp-Y160H BVMOs were also eval-
uated with other compounds, chosen among those previously
identified as substrates. As shown in Table 5, conversion of these
ketones resulted in the formation of the expected ester prod-
uct, obeying the stereo-chemical rule of Baeyer–Villiger reactions
where the migrating group is the most substituted one. In partic-
ular, 2-octanone, 3-octanone and 4-octanone were fully converted
by Cm-BVMO. The mutated BVMO from Physcomitrella performed
full conversion of 2-octanone and 3-octanone and poor transfor-
mation of 4-octanone, whereas the unmodified enzyme was able
to oxidize in good amount only the first two linear ketones. Finally
phenylacetone was fully converted by all the three enzymes. The
recombinant enzymes gave poor or almost no conversion of the
cyclic ketones cyclohexanone and cyclopentanone; only Cm-BVMO
was able to transform the former molecule.
microbial organisms, particularly actinobacteria and filamentous
fungi. Only two BVMOs of eukaryotic origin have been identified
and expressed in recombinant form very recently: the BVMOs from
the ascomycetes Cylindrocarpon radicicola ATCC 11011 [20] and
Aspergillus fumigatus Af293 [21]. We have identified and selected
two new putative BVMOs from the photosynthetic eukaryotes
Cyanidioschizon merolae and Physcomitrella patens.
The two organisms can be certainly considered very uncom-
mon sources for BVMOs, being eukaryotic and photosynthetic. The
primitive red alga C. merolae 10D is a unicellular organism living
in acidic environments (like hot springs), even at pH < 2 and tem-
perature of 45 ◦C. It is one of the photosynthetic eukaryotes with
the most simple cell architecture: its cell does not present a rigid
cell wall and contains a single nucleus, a single mitochondrion and
a single chloroplast. This clearly offers advantages for studies on
these organelles [29,30]. The moss P. patens is classified as a no
vascular plant, since it differs from higher plants in not having inter-
nal vessels [31], and stands in an important phylogenetic position
for studying the evolutionary transition from the aquatic environ-
ment of algae to the terrestrial one of higher plants. Noticeably,
both organisms undergo homologous recombination with a fre-
quency that allows easy targeting of genes for replacement and
elimination [32,33]. This allows handy studying of gene func-
tion and helps in predicting the physiological role of genes in
higher organisms. P. patens and C. merolae are therefore considered
model systems for studying origin and fundamental mechanisms
of eukaryotic cells and plant evolution.
We succeeded in producing the identified putative BVMOs
in recombinant form, together with a mutant variant of the
protein from Physcomitrella completely restored in the consen-
sus sequence firstly described. The second BVMO-typifying motif
([A/G]GxWxxxx[F/Y]P[G/M]xxxD), recognized and annotated after
the beginning of our work [24], is fully conserved in both proteins.
The biocatalytic characterization of the recombinant products
could demonstrate their BVMO activity on a broad range of sub-
strates as linear and cyclic aliphatic ketones, aryl and aromatic
ketones (Table 1). Among linear ketones (C8 and C12), proven to
be the best substrates, 2-dodecanone was the preferred one.
In an overview of the data from the kinetic analysis, the
three enzymes displayed different rates of catalysis, depending
on the substrate examined (Tables 2 and 3), but in all cases mea-
sured kcat values were not higher than 0.4 s−1. Since most natural
monooxygenase enzymes display kcat values in the range of
1–100 s−1 when tested on ketones among those selected for our
analysis [34], we conclude that the new BVMOs (at least as such)
4. Discussion
BVMOs are currently in the spotlight of industrial biocatalysis
for their capability to perform challenging steps in organic chemical
processes, such as oxidations of ketones to esters (Baeyer–Villiger
oxidations), but also sulfoxidations and other oxidative reac-
tions. Up to 2012, all reported type I BVMOs originated from

E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
153
Fig. 4. Biooxidation of racemic bicyclo[3.2.0]hept-2-en-6-one. Possible structures of bicyclo[3.2.0]hept-2-en-6-one isomers and lactones formed by type I BVMO activity (A).
Time course of Cm- (B), Pp- (C) and Pp-Y160H- (D) BVMOs catalyzed oxidation of racemic bicyclo[3.2.0]hept-2-en-6-one to 2-oxabicyclo[3.3.0]oct-6-en-3-one (+)-(1S,5R)
2/(−)-(1R,5S) 2 and 3-oxabicyclo[3.3.0]oct-6-en-2-one (−)-(1S,5R) 3/(+)-(1R,5S) 3.
Table 5
Conversion of some identified substrates and corresponding products.
Substrate
Product
Conversion (%)
Cm
Pp
PpY160H
27
O
O
O
–
17
O
O
O
80
1
3
O
O
100
100
100
O
O
O
O
100
100
100
90
67
–
100
100
17
O
O
O
O
O
O

154
E. Beneventi et al. / Journal of Molecular Catalysis B: Enzymatic 98 (2013) 145–154
seem rather inefficient bio-catalysts. This is also true in terms
of enantioselectivity, which was shown to be very poor when
probed on the typical substrate racemic bicyclo[3.2.0]hept-2-en-
6-one (Table 4), transformed in the four possible lactone.
suggestions on use of programs for dendrogram production. This
work has been entirely financed by F.I.S.-Fabbrica Italiana Sintetici
(Alte di Montecchio Maggiore, Italy).
References
Nevertheless, some results deserve to be highlighted. As known
from the literature, representative BVMOs can be clustered in well-
defined groups with different substrate profile. The phylogenetic
analysis that we performed (Fig. 2) places the new BVMOs in two
distinct positions. The protein from P. patens is localized in the clade
defined by PAMO and STMO, which transform aromatic (but also
linear) ketones; the one from C. merolae is clustered together with
MEKMO and ACMO, enzymes known for converting linear ketones
[35,36]. This grouping is in agreement with the substrate scope
obtained for the recombinant enzymes even though, because of
their eukaryotic origin, a different biocatalytic behavior could be
expected.
[1] A. Baeyer, V. Villiger, Ber. Dtsch. Chem. Ges. 32 (1899) 3625–3633.
[2] G. Ten Brink, I. Arends, R. Sheldon, Chem. Rev. 104 (2004) 4105–4124.
[3] S. Chandrasekhar, C.D.J. Roy, Chem. Soc. Perkin Trans. 2 (1994) 2141–2143.
[4] R.B. Turner, J. Am. Chem. Soc. 72 (1950) 878–882.
[5] N.M. Kamerbeek, A.J.J. Olsthoorn, M.W. Fraaije, D.B. Janssen, Appl. Environ.
Microbiol. 69 (2003) 419–426.
[6] S. Colonna, N. Gaggero, P. Pasta, G. Ottolina, Chem. Commun. (Camb.) 20 (1996)
2303–2307.
[7] G. Ottolina, P. Pasta, G. Carrea, S. Colonna, S. Dallavalle, H.L. Holland, Tetrahe-
dron Asymmetry 6 (1995) 1375–1386.
[8] G. Ottolina, S. Bianchi, B. Belloni, G. Carrea, B. Danieli, Tetrahedron Lett. 40
(1999) 8483–8486.
[9] P.B. Brondani, G. de Gonzalo, M.W. Fraaije, L.H. Andrade, Adv. Synth. Catal. 353
(2011) 2169–2173.
The second noteworthy result concerns the central histidine of
the motif FxGxxxHxxxWP: this residue, although dispensable for
catalysis and not always conserved (see [21] and [24]), has been
proposed to be involved in conformational changes occurring along
with catalysis [26]. Its fundamental role has been already shown
[19,28]. By comparing activities of wild type and mutant Pp-BVMOs,
we further confirm here that this histidine is certainly important
for enzymatic performance.
[10] S. Colonna, N. Gaggero, G. Carrea, G. Ottolina, P. Pasta, F. Zambianchi, Tetrahe-
dron Lett. 43 (2002) 1797–1799.
[11] M.W. Fraaije, D.B. Janssen, in: R.D. Schmid, V.B. Urlacher (Eds.), Modern Bioox-
idation, Wiley-VCH, Weinheim, 2007.
[12] M.D. Mihovilovic, Curr. Org. Chem. 10 (2006) 1265–1287.
[13] A. Willetts, Trends Biotechnol. 15 (1997) 55–62.
[14] M.P. Beam, M.A. Bosserman, N. Noinaj, M. Wehenkel, J. Rohr, Biochemistry 48
(2009) 4476.
[15] W. Van Berkel, N. Kamerbeek, M. Fraaije, J. Biotechnol. 124 (2006) 670–689.
[16] M.D. Mihovilovic, B. Müller, P. Stanetty, Eur. J. Org. Chem. 22 (2002) 3711–3730.
[17] N.A. Donoghue, D.B. Norris, P.W. Trudgill, Eur. J. Biochem. 63 (1976) 175–192.
[18] N.M. Kamerbeek, D.B. Janssen, W.J.H. van Berkel, M.W. Fraaije, Adv. Synth. Catal.
345 (2003) 667–678.
[19] M.W. Fraaije, N.M. Kamerbeek, W.J.H. van Berkel, D.B. Janssen, FEBS Lett. 518
(2002) 43–47.
[20] F. Leipold, R. Wardenga, U.T. Bornscheuer, Appl. Microbiol. Biotechnol. 94
(2012) 705–717.
[21] M.L. Mascotti, M.J. Ayub, H. Dudek, M.K. Sanz, M.W. Fraaije, AMB Expr. 3 (2013)
33.
[22] M.W. Fraaije, J. Wu, D.P.H.M. Heuts, E.W. van Hellemond, J.H.L. Spelberg, D.B.
Janssen, Appl. Microbiol. Biotechnol. 66 (2005) 393–400.
[23] F. Forneris, R. Orru, D. Bonivento, L.R. Chiarelli, A. Mattevi, FEBS J. 276 (2009)
2833–2840.
Finally, we could show that the enzyme from Cyanidioschizon
displays a relatively high thermostability by having an apparent
melting temperature of 56 ◦C. This compares favorably with other
well-studied BVMOs: CHMO has a Tm of 39 ◦C (near to the value
that we measured for Pp-BVMO, i.e. 44 ◦C), while PAMO has a Tm
only five degrees higher than that measured for Cm-BVMO [37]. If
protein engineering will be considered to boost the enzyme activ-
ities in view of a possible re-consideration of their utility, the
primary sequence of Cm-BVMO could be a favorable starting point.
In the same perspective, a better characterization of the reper-
toire of accepted molecules could also be useful. The displayed
very low catalytic efficiency suggests that during evolution the
two new BVMOs had been subjected to a weak selective pressure.
Hence, according to the current view, a potential role as secondary
metabolism enzymes can be envisaged [38,39]. Their substrate
preferences point toward linear alkanones, structurally similar to
side chains of chlorophylls and pheophytins, and to their possible
involvement in modifications of photosynthetic pigments. Work is
in progress to produce knockout mutants in both the alga and the
moss, by exploiting their endogenous homologous recombination.
Comparison of null mutants with wild-type strains would help in
uncovering the natural substrates of Cm- and Pp-BVMOs in their
original biological context.
[24] A. Riebel, H. Dudek, G. de Gonzalo, P. Stepniak, L. Rychlewski, M. Fraaije, Appl.
Microbiol. Biotechnol. 95 (2012) 1479–1489.
[25] O. Vallon, Proteins 38 (2000) 95–114.
[26] E. Malito, A. Alfieri, M. Fraaije, A. Mattevi, PNAS 101 (2004) 13157–13162.
[27] J. Rehdorf, C.L. Zimmer, U.T. Bornscheuer, Appl. Environ. Microbiol. 75 (2009)
3106.
[28] M.J. Cheesman, M.B. Kneller, A.E. Rettie, Chem. Biol. Interact. 146 (2003)
157–164.
[29] V. Alphand, R. Furstoss, J. Org. Chem. 57 (1992) 1306–1309.
[30] M.D. Mihovilovic, P. Kapitán, P. Kapitánová, ChemSusChem
143–148.
1 (2008)
[31] D. Lang, A.D. Zimmer, S.A. Rensing, R. Reski, Trends Plant Sci. 13 (2008) 542–549.
[32] D.G. Schaefer, Annu. Rev. Plant Biol. 53 (2002) 477–501.
[33] A. Minoda, R. Sakagami, F. Yagisawa, T. Kuroiwa, K. Tanaka, Plant Cell Physiol.
45 (2004) 667–671.
[34] A. Bar-Even, E. Noor, Y. Savir, W. Liebermeister, D. Davidi, D.S. Tawfik, R. Milo,
Biochemistry 50 (2011) 4402–4410.
[35] D.V. Rial, P. Cernuchova, J.B. van Beilen, M.D. Mihovilovic, J. Mol. Catal., B
Enzyme 50 (2008) 61–68.
[36] M.D. Mihovilovic, F. Rudroff, B. Grötzl, P. Kapitan, R. Snajdrova, J. Rydz, R. Mach,
Angew. Chem. Int. Ed. Engl. 44 (2005) 3609–3613.
Acknowledgements
[37] H.L. van Beek, G. de Gonzalo, M.W. Fraaije, Chem. Commun. (Camb.) 48 (2012)
3288–3290.
[38] S. Picaud, L. Olofsson, M. Brodelius, P.E. Brodelius, Arch. Biochem. Biophys. 436
(2005) 215–226.
[39] B.T. Greenhagen, P.E. O’Maille, J.P. Noel, J. Chappell, Proc. Natl. Acad. Sci. U.S.A.
103 (2006) 9826–9831.
[40] K. Tamura, J. Dudley, M. Nei, S. Kumar, Mol. Biol. Evol. 24 (2007) 1596–1599.
[41] N. Saitou, M. Nei, Mol. Biol. Evol. 4 (1987) 406–425.
[42] D.H. Huson, D.C. Richter, C. Rausch, T. Dezulian, M. Franz, R. Rupp, BMC Bioin-
formatics 8 (2007) 460.
We thank Dr. T. Morosinotto (Department of Biology, University
of Padova, Italy) for providing a starting culture of Physcomitrella
patens. We are grateful to Dr. P. Polverino de Laureto (CRIBI Biotech-
nology Center, University of Padova, Italy) for help in analytical
characterization of the recombinant flavoprotein Pp-BVMO and
Dr. Enrico Negrisolo (Department of Public Health, Comparative
Pathology and Veterinary Hygiene, University of Padova) for useful