Expanding the biocatalytic toolbox of flavoprotein monooxygenases from Rhodococcus jostii RHA1

Article abstract of DOI:10.1016/j.molcatb.2012.11.009

With the aim to enlarge the set of available flavoprotein monooxygenases, we have cloned 8 unexplored genes from Rhodococcus jostii RHA1 that were predicted to encode class B flavoprotein monooxygenases. Each monooxygenase can be expressed as soluble protein and has been tested for conversion of sulfides and ketones. Not only enantioselective sulfoxidations, but also enantioselective Baeyer-Villiger oxidations could be performed with this set of monooxygenases. Interestingly, in contrast to known class B flavoprotein monooxygenases, all studied biocatalysts showed no nicotinamide coenzyme preference. This feature coincides with the fact that the respective sequences appear to form a discrete group of sequence related proteins, distinct from the known class B flavoprotein monooxygenases subclasses: the so-called flavin-containing monooxygenases (FMOs), N-hydroxylating monooxygenases (NMOs) and Type I Baeyer-Villiger monooxygenases (BVMOs). Taken together, these data reveal the existence of a new subclass of class B flavoprotein monooxygenases, which we coined as Type II FMOs, that can perform Baeyer-Villiger oxidations and accept both NADPH and NADH as coenzyme. The uncovered biocatalytic properties of the studied Type II FMOs make this newly recognized subclass of monooxygenases of potential interest for biocatalytic applications.

Full text of DOI:10.1016/j.molcatb.2012.11.009

Accepted Manuscript
Title: Expanding the biocatalytic toolbox of flavoprotein
monooxygenases from Rhodococcus jostii
Authors: Anette Riebel, Gonzalo de Gonzalo, Marco W.
Fraaije
PII:
DOI:
Reference:
S1381-1177(12)00310-4
doi:10.1016/j.molcatb.2012.11.009
MOLCAB 2579
To appear in:
Journal of Molecular Catalysis B: Enzymatic
Received date:
Revised date:
Accepted date:
10-9-2012
11-11-2012
13-11-2012
Please cite this article as: A. Riebel, G. de Gonzalo, M.W. Fraaije, Expanding the
biocatalytic toolbox of flavoprotein monooxygenases from Rhodococcus jostii, Journal
of Molecular Catalysis B: Enzymatic (2010), doi:10.1016/j.molcatb.2012.11.009
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

1
2
3
4
5
6
7
8
9
Expanding the biocatalytic toolbox of flavoprotein
monooxygenases from Rhodococcus jostii
Anette Riebel, Gonzalo de Gonzalo, Marco W. Fraaije*
Laboratory of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute,
University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands.
* To whom correspondence should be addressed
1
1
0
1
Tel: +31503634345; Fax: +31503634165; E-mail: m.w.fraaije@rug.nl
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Abstract
With the aim to enlarge the set of available flavoprotein monooxygenases, we have cloned 8
unexplored genes from Rhodococcus jostii RHA1 that were predicted to encode class B
flavoprotein monooxygenases. Each monooxygenase can be expressed as soluble protein
and has been tested for conversion of sulfides and ketones. Not only enantioselective
sulfoxidations, but also enantioselective Baeyer-Villiger oxidations could be performed with
this set of monooxygenases. Interestingly, in contrast to known class B flavoprotein mono-
oxygenases, all studied biocatalysts showed no nicotinamide coenzyme preference. This
feature coincides with the fact that the respective sequences appear to form a discrete group
of sequence related proteins, distinct from the known class B flavoprotein monooxygenases
subclasses: the so-called flavin-containing monooxygenases (FMOs), N-hydroxylating
monooxygenases (NMOs) and Type I Baeyer-Villiger monooxygenases (BVMOs). Taken
together, these data reveal the existence of a new subclass of class B flavoprotein
monooxygenases, which we coined as Type II FMOs, that can perform Baeyer-Villiger
oxidations and accept both NADPH and NADH as coenzyme. The uncovered biocatalytic
properties of the studied Type II FMOs make this newly recognized subclass of
monooxygenases of potential interest for biocatalytic applications.
Keywords: Enzyme screening, flavoprotein monooxygenases, sulfoxidation, Baeyer-
Villiger reaction, enantioselectivity
Abbreviations: CE: cell extract; CCE: cleared cell extract; FMOs: flavin-containing
monooxygenases; NMOs: N-hydroxylating monooxygenases; BVMOs: Baeyer-Villiger
monooxygenases
1
Page 1 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
1. Introduction
Flavoprotein monooxygenases are attracting attention as selective and oxidative
biocatalysts that can be used for the production of high-value chemical building blocks or
pharmaceuticals [1]. These biocatalysts efficiently catalyze chemo-, regio-, and/or
enantioselective oxygenations using dioxygen as mild oxidant while using NAD(P)H as
reductant. Flavoprotein monooxygenases can be divided into 6 distinct classes, with each
class containing sequence- and structure-related monooxygenases. Two of these classes
(class A and B) are especially appealing when considering biocatalysis. These two classes
are typified by being single component enzymes that contain a tightly bound flavin cofactor,
while the other monooxygenases rely on multiple protomers and often employ a loosely
bound flavin cofactor. The class A of flavoprotein monooxygenases seems to have evolved
to catalyse aromatic hydroxylations, as most of the characterized monooxygenases of this
class represent hydroxylases, typically acting on a very restricted number of substrates [2].
Class B flavoprotein monooxygenases do not catalyse hydroxylations but perform Baeyer-
Villiger oxidations and/or oxygenations of heteroatom-containing compounds. In fact, three
class B flavoprotein monooxygenase subclasses have been identified based on specific
sequence motifs, which coincide with a preference for specific oxygenation types for each
subclass [3]:
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
1 - Baeyer-Villiger monooxygenases (BVMOs) contain the sequence motif
(FxGxxxHxxxWP/D) and primarily catalyse Baeyer-Villiger oxidations, while they are also
able to oxygenate heteroatom-containing compounds (N, S, B or Se containing compounds);
2 - the so-called Flavin-containing monooxygenases (FMOs) contain a slightly
different sequence motif (FxGxxxHxxxYK/R), and are specialized in oxidizing heteroatom-
containing compounds while they are inefficient in catalyzing Baeyer-Villiger oxidations.
The FMOs have mainly been studied as xenobiotic degrading enzymes that help the human
body to dispose toxic compounds [4, 5]. The human proteome encompasses six FMO
isoforms [6], which are able to activate or degrade many drugs. Only very recently FMOs
have been considered for their use as biocatalysts, due to the identification and production of
a microbial FMO which, in contrast to the human homologs, can be easily expressed as a
soluble protein [7].
3 - N-hydroxylating monooxygenases (NMOs) share sequence homology with the
above-mentioned class B flavoprotein monooxygenases but lack a typifying sequence motif.
Only a conserved histidine can be identified in the region of the BVMO/FMO sequence
2
Page 2 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
motif. So far only a few NMOs from bacteria and fungi have been reported. They typically
convert long-chain primary amines by N-hydroxylation [8].
The overall sequence homology among all class B flavoprotein monooxygenases
reflects the fact that they are all single-component FAD-containing monooxygenases
composed of two dinucleotide binding domains (Rossman folds to bind both FAD and
NADPH), that allow them to combine flavin reduction and monooxygenation in one
polypeptide chain. They often prefer the use of NADPH as electron donor, keeping the
+
NADP tightly bound throughout the catalytic cycle [9, 10]. From the three subclasses,
mainly the BVMOs have been extensively studied as biocatalysts.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
To tap the natural diversity for the discovery of novel oxygenating enzymes
Rhodococcus jostii RHA1 is a very promising candidate. The proteome of this bacterium is
predicted to contain an unusually high variety of oxidative enzymes [11, 12]. Inspired by
this observation, we and the Grogan group have recently cloned 22 putative BVMO-
encoding genes, and succeeded in producing and exploring the biocatalytic properties of
these enzymes [13, 14]. This research has confirmed that all these enzymes indeed act as
BVMOs which can be used for a large number of oxygenations. When screening the
predicted proteome of R. jostii RHA1 for monooxygenases, we also identified a relatively
large number of other putative class B monooxygenases that seem to be more closely related
to NMOs and FMOs. While BVMOs are relatively rare enzymes that are only found in
bacteria and fungi with an average distribution of only one or two BVMO-encoding genes
per microbial genome [15], FMOs and NMOs are even more scarce in microbes. FMOs are
quite abundantly present in higher eukaryotes (e.g. the 6 isoforms in the human genome and
often more than 10 in plant genomes [16]), but these FMOs are often difficult to produce
due to their membrane association.
In this paper, we report on the exploration of 8 novel class B flavoprotein
monooxygenases obtained from a single microorganism, R. jostii RHA1, that share sequence
homology with FMOs and NMOs. By optimizing expression vectors and expression
conditions, all enzymes were obtained in soluble and holo form. A set of ketones and
aromatic sulfides was tested for all enzymes to explore their biocatalytic potential. Also their
coenzyme specificity and enantioselective properties were analysed.
2. Experimental
2.1. General materials and methods
3
Page 3 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
Oligonucleotide primers were purchased from Sigma, dNTPs and In-Fusion™ 2.0
CF Dry-Down PCR Cloning Kit from Clontech, Phusion polymerase from Finnzymes. All
other chemicals were obtained from Acros Organics, ABCR, Sigma-Aldrich, TCI Europe,
and Roche Diagnostics GmbH. The nicotinamide coenzymes were purchased from Codexis.
The NCBI server (http://blast.ncbi.nlm.nih.gov/Blast.cgi) was used for DNA sequence
retrieval and BLAST searches. The EBI server (http://www.ebi.ac.uk/Tools/msa/clustalw2/)
was used for multiple sequence alignment by CLUSTALW [19]. Treeview software
(http://taxonomy.zoology.gla.ac.uk/rod/treeview.html) was used for visualization of the
sequence relationships.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
2.2. Cloning and expression
Escherichia coli TOP10 from Invitrogen was used as a host for DNA manipulations
and protein expression. Two expression vectors have been used: (i) a modified pBAD vector
(pBADN) in which the NdeI site was replaced by the original NcoI site [17], and (ii) the
recently engineered pBAD-based pCRE2 vector which harbors a codon-optimized gene
encoding a thermostable mutant of phosphite dehydrogenase (PTDH) with an N-terminal
His-tag [18].
The target genes were amplified by PCR using genomic DNA of R. jostii RHA1 as
template and subsequently cloned into pBADN using the In-Fusion PCR Cloning kit from
Clontech, following the recommendations of the manufacturer. Expression of all generated
expression constructs was tested using 24-multiwell microtiter plates in the sandwich cover
system from Enzyscreen®. Cell cultures of 2.5 ml in LB medium supplemented with 50 μg
-1
mL ampicillin were grown at four different temperatures (17°C for 48 h; 24°C for 32 h;
30°C and 37°C for 16 h) with four different arabinose concentrations (0.002%, 0.02%,
0.2%, and none) each, with shaking at 200 rpm. Cell extracts (CEs) were obtained using
DNase/lysozyme in combination with freezing in liquid nitrogen and thawing at 30°C. CEs
as well as cell cleared extracts (CCEs) were analyzed by SDS-PAGE for (soluble)
expression of the monooxygenases. For the genes that did not yield expressed soluble
protein when cloned into pBADN, the in-house developed pCRE2 expression vector was
used.
2.3. General procedure for the bioconversions employing the novel FMOs
Conversions were performed essentially as described before [14]. For GC analysis,
500 µl incubations of 50 mM Tris/HCl pH 7.5, 10% glycerol, 1 mM DTT, 1 mM EDTA, 10
4
Page 4 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
µM FAD, 5 mM substrate, 5% cosolvent (1,4-dioxane), 100 μM NADPH, 3.1 μM PTDH,
10 mM phosphite and 5 μM of the corresponding monooxygenase in CCE form (total
volume of 0.5 mL), were shaken in glass vials at 24°C for 24 hours. For determining the
exact concentration of each enzyme in the respective extract, a recently developed method
was used which relies on the decrease in absorbance at 450 nm upon NADPH-mediated
reduction of the flavin cofactor [14]. Each conversion was stopped by extracting with ethyl
acetate (2 x 0.5 mL containing 0.1% mesitylene as standard), dried with magnesium sulfate
and analyzed directly by GC to determine the degree of conversion and the
enantioselectivity. The details concerning the (chiral) GC analysis can be found in the
Supplemental Information (Table S1). For every tested reaction, control experiments in the
absence of enzyme resulted in no conversion.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3. Results
3.1. Identification of putative FMO/NMO-encoding genes
By a PBLAST search of the proteome of R. jostii RHA1 [11], not only 23 Type I
BVMO-encoding genes could be found [14], but also another 8 genes putatively encoding
class B flavoprotein monooxygenases could be identified. All corresponding proteins
contain two typical Rossman fold motifs (GxGxxG), clearly distinguishing them from other
flavoprotein monooxygenase classes. As for all class B flavoprotein monooxygenases, one
Rossman fold motif is close to the N-terminus while the other is in the middle of the
sequences. In our previous work [14], we proposed a new conserved sequence motif typical
for the Type I BVMO family: [A/G]GxWxxxx[F/Y]P[G/M]xxxD located between the two
Rossman fold motifs. This motif entails the conserved active site aspartate and therefore
appears a better fingerprint for Type I BVMO sequences. The motif is absent in all
sequences used in this study and confirms that they do not represent classical BVMOs. The
sequences do also not contain the previously described BVMO-typifying motif. In fact, most
of them (all except for monooxygenase H) appear more closely related to FMOs as the
FMO-typifying sequence motif can be recognized with only one or two amino acid
substitutions (see Table 1). This is also reflected by the fact that most of the sequences have
been annotated in the sequence database as putative flavin-containing monooxygenases.
When performing a multiple sequence alignment, the clustering becomes apparent.
Monooxygenase H is most related to known NMOs, while the other proteins form another
distinct group of sequence-related proteins (Figure 1). It also shows that all studied
sequences are quite distant from Type I BVMOs. Based on these observations, we annotate
5
Page 5 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
monooxygenase H as an NMO (NMO-H). The other proteins seem to form another isolated
cluster of sequence-related class B flavoprotein monooxygenases and therefore we classify
them as Type II FMOs (while the known and well-established FMO class can be considered
as Type I FMOs). Recently another class B flavoprotein monooxygenase has been described
that shows relatively high sequence similarity with these Type II FMOs (FMO-X in Figure
1). The Grogan group has discovered this bacterial monooxygenase that can perform
sulfoxidations and can act as BVMO [20]. However, conversion rates described for this
monooxygenase seem rather modest (100-1000 fold lower when compared with conversion
rates of Type I BVMO).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
This sequence analysis hints to a new subclass of class B flavoprotein
monooxygenases, Type II FMOs that can be employed as biocatalysts. By cloning,
expressing, and testing biocatalytic oxidations, we set out to explore the above-mentioned 8
putative class B flavoprotein monooxygenases.
3.2. Expression of the putative class B flavoprotein monooxygenases
The targeted genes were cloned in a pBAD-based expression vector (pBADN, [17])
by ligation free cloning. By varying the temperature and arabinose concentration, optimal
expression conditions were determined. By this approach, 6 proteins could be obtained in
soluble form with good expression levels (Table 1). Only for 2 genes, expression did not
yield satisfying amounts of soluble protein. We solved this expression problem by cloning
the respective genes into the recently developed pCRE2 expression vector [18]. With these
constructs we were able to achieve soluble expression of all targeted proteins that were
predicted to represent class B flavoprotein monooxygenases.
Interestingly, in two cases, FMO-E and FMO-G, it was found that during expression
the growth medium developed a bluish colour, as shown in Table 1. We have previously
observed this phenomenon when expressing a phenylacetone monooxygenase mutant
capable of converting indole into indigo blue [21]. Also the expression of several type I
BVMOs from R. jostii RHA1 led to blue pigment formation [14]. The same phenomenon
has been observed when expressing a bacterial FMO in E. coli [22]. Apparently, the
respective monooxygenases are able to form indigo blue by oxidizing the endogenous
indole. The fact that indigo blue formation is only observed upon expression of distinct
enzymes, confirms the functional expression of these monooxygenases, in our case of FMO-
E and FMO-G.
6
Page 6 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
3.3. Substrate profile of the overexpressed monooxygenases
Even though using CCE for the reactivity tests, we used the same concentration of
recombinant enzyme in all assays to allow a direct comparison of the results. For this, the
amount of active enzyme in the CCEs was first quantified with the NADPH-mediated flavin
reduction method established and described before [14]. This revealed that all extracts
contained significant amounts of NADPH-dependent flavoproteins, which is in agreement
with the SDS-PAGE analysis of CCEs, and confirms that all enzymes contain a flavin
cofactor upon expression and accept NADPH as coenzyme.
To determine the biocatalytic potential of the studied monooxygenases, each
monooxygenase was tested for activity on several ketones and aromatic sulfides: 2-
indanone, cyclopentadecanone, phenylacetone, bicyclo[3.2.0]hept-2-en-6-one, thioanisole,
4-methylthioanisole, ethyl benzyl sulfide, and benzyl phenyl sulfide.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
None of the monooxygenases was found to convert 2-indanone or
cyclopentadecanone. In contrast, intriguingly, FMO-E, FMO-F, and FMO-G were found to
be able to convert phenylacetone and bicyclo[3.2.0]hept-2-en-6-one (Scheme 1). Product
analysis by GC revealed that phenylacetone was converted into the corresponding ester,
benzylacetate (Table 2). Enantiodivergent oxidation of bicyclo[3.2.0]hept-2-en-6-one was
analyzed by chiral GC. This revealed that the monooxygenases have a preference for the
formation of the normal lactone, while still a significant amount of the abnormal lactone is
formed (Table 2). While the biocatalysts are not very enantioselective in formation of the
normal lactone, the FMO-F and FMO-G yield >65% ee for the abnormal (1S,5R)
enantiomer. Taken together, these data show that FMO-E, FMOF and FMO-G act as bona
vide BVMOs.
Next, four prochiral aromatic sulfides with varying substituents (thioanisole, 4-
methylthioanisole, ethyl benzyl sulfide, and benzyl phenyl sulfide) were tested as substrates.
The relatively small substrate thioanisole was the only sulfide that was converted by most of
the monooxygenases in detectable amounts, as indicated in Table 3. Only NMO-H did not
convert significant amounts of thioanisole or any other tested sulfide, strengthening the
classification of this particular enzyme as a NMO. For the majority of the biocatalysts,
mainly the (R)-methyl phenyl was preferentially formed, with the exception of FMO-G,
which led to the (S)-enantiomer. The enantioselectivity of the thioanisole sulfoxidation
varied from low to moderate, achieving the highest optical purity (ee = 57%), when
employing FMO-G. The conversion of 4-methylthioanisole could only be detected in very
low conversions. This is in line with the observation of Grogan that FMO-X only showed a
7
Page 7 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
poor conversion rate with this aromatic sulfide [20]. The enantioselective outcome with this
substrate when using the Type II FMOs from R. jostii RHA1 was similar to that achieved
with thioanisole. The more bulky substrate benzyl ethyl sulfide was only converted by
FMO-E, FMO-F and FMO-G, but with poor conversion. Only the corresponding methyl
phenyl sulfone was detected as product, indicating that these biocatalysts are able to
oxygenate the initially formed sulfoxide to the overoxidized product. The bulky benzyl
phenyl sulfide was not converted by any of the tested monooxygenases.
FMO-X was found to be indifferent towards NADPH or NADH [20]. This is in sharp
contrast with the known Type I BVMOs which only accept NADPH as coenzyme. It also is
different from the Type I FMOs, that typically prefer NADPH as a coenzyme. Conversion of
several of the previously identified substrates where therefore tested via GC with both
nicotinamide cofactors. When we replaced NADPH by NADH, almost identical degrees of
conversion and enantioselectivities were found for the conversions for all tested Type II
FMOs. This is a highly interesting and relevant finding as it suggests that the Type II FMOs
have as general characteristic a relaxed coenzyme specificity. This is attractive for
biocatalytic applications, as NADH is a cheaper source than it phosphorylated analogue
(when considering use as isolated enzyme) while NADH is also present at higher level in
cells (when considering use of whole cells). This shows that the Type II FMOs described in
this paper, together with the previously described FMO-X [20], represent a newly
discovered subclass of class B flavoprotein monooxygenases that can accept both NADPH
and NADH.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
4. Discussion
Eight class B flavoprotein monooxygenases from R. jostii RHA1 have been
expressed in soluble form and studied for their biocatalytic potential. From the sequence
alignment study, one of the targeted monooxygenases (NMO-H) was found to belong to the
group of NMOs. The prototype enzymes for the subclass of N-hydroxylating flavoprotein
monooxygenases (NMOs) are L-ornithine N5-hydroxylase and L-lysine N6-hydroxylase
[25]. An L-ornithine hydroxylase had already been identified in the R. jostii RHA1 proteome
(Rmo_Orn in Figure 1, [26]). When comparing NMO-H with known NMO sequences it
shows highest sequence similarity with the lysine hydroxylase IucD from E. coli (Figure 1).
This suggests that the NMO-H described in this study probably represents a lysine
hydroxylase present in R. jostii RHA1. The restricted substrate scope of lysine hydroxylases
also is in line with the observation that NMO-H is not converting any aromatic compound
8
Page 8 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
tested in this study [27]. This confirms that, in contrast to BVMOs and FMOs, NMOs are
rather restricted in substrate acceptance and appear to be of little value as biocatalysts.
The FMOs described in this paper do not seem to belong to the typical FMO subclass
as (i) they do not contain a fully conserved FMO-typifying motif (Table 1), and (ii) they
form another cluster of sequence-related proteins distinct from the three other subclasses,
FMOs, NMOs, and BVMOs (see Figure 1). Also their catalytic properties deviate from the
other subclasses by accepting both nicotinamide coenzymes, NADH and NADPH, without
significant difference. Furthermore, several of these FMOs were shown to perform Baeyer-
Villiger oxidations, a reaction for which typical FMOs display hardly any activity. We
coined this newly recognized subclass of class B flavoprotein monooxygenases Type II
FMOs.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
This study reveals that of the tested Type II FMOs, FMO-E, FMO-F, and FMO-G
may represent valuable new oxidative biocatalysts. These three enzymes were found to be
active in both sulfoxidations and Baeyer-Villiger reactions. Multiple sequence alignment
revealed that they form a separate cluster of sequences (Figure 1). Except for showing a
relative high sequence homology with each other, these three monooxygenases also share an
N-terminal extension of about 160 residues (see Figure S1). This may suggest that this
newly identified subclass of monooxygenases has evolved as separate subclass of class B
flavoprotein monooxygenases, distinct from the Type I BVMOs, towards a new subclass of
potent Baeyer-Villiger monooxygenases. Therefore, these three FMOs or close homologs
thereof may represent interesting alternative biocatalysts (for instance, due to their NADH
acceptance) when compared with the known Type I BVMOs. With merely using the N-
terminal sequence of one of three Type II FMOs, it is possible to identify hundreds of
homologous putative Type II FMOs. It will be exciting to see whether these homologs
indeed display similar biocatalytic properties. Future research will reveal whether
monooxygenases from this newly identified subclass of class B flavoprotein
monooxygenases (Type II FMOs) are indeed valuable as biocatalysts; e.g. their operational
stability and substrate scope has still to be explored.
5. Conclusions
In the present paper, a set of eight new flavoprotein monooxygenases from R. jostii
RHA1 has been expressed and obtained in soluble form by employing two expression
vectors and optimizing expression conditions. Sequence analysis established that one of
these enzymes represents a NMO, while the other cluster as a separate class B flavoprotein
9
Page 9 of 20

1
2
3
4
5
6
7
8
9
0
1
2
3
4
monooxygenase subclass: Type II FMOs. The monooxygenases have been tested for their
ability to perform sulfoxidations and Baeyer-Villiger oxidations. Interestingly, three of the
Type II FMOs are able to catalyze both reaction types and accept both NADPH or NADH as
coenzyme, presenting a different behavior with respect to classical FMOs and BVMOs.
Type II FMOs might develop as an interesting alternative to BVMOs, as they are able to
catalyze the similar reaction with a relaxed nicotinamide cofactor acceptance.
Acknowledgments
This research was financially supported by the Integrated Biosynthesis Organic
Synthesis (IBOS) program of the Netherlands Organisation for Scientific Research (NWO).
G. de Gonzalo received financial support from the EU-FP7 Oxygreen project.
1
1
1
1
1
References
[1] W.J.H. van Berkel, N.M. Kamerbeek, M.W. Fraaije, J. Biotechnol. 124 (2006) 670-689.
1
1
5
6
[2] S. Montersino, D.Tischler, G.T. Gassner, W.J.H. van Berkel, Adv Synth Catal 353
(2011) 2301-2319.
1
1
7
8
[3] M.W. Fraaije, N.M Kamerbeek, W.J.H. Berkel, D.B. Janssen, D.B. FEBS Lett. 518
(2002) 43-47.
19
[4] D.M. Ziegler, Trends Pharmacol. Sci. 11 (1990) 321-324.
2
2
0
1
[5] C.T. Dolphin, A. Janmohamed, R.I..Smith, E.A. Shephard, I.R. Phillips, Nat. Genet. 17
(1997) 491-494.
2
2
2
3
[6] R.N. Hines, K.A. Hopp, J. Franco, K. Saeian, F.P. Begun, Mol. Pharmacol. 62 (2002)
320-325.
2
2
4
5
[7] A. Rioz-Martínez, M. Kopacz, G. de Gonzalo, D.E. Torres Pazmiño, V. Gotor, M.W.
Fraaije, Org. Biomol. Chem. 9 (2011) 1337-1341.
2
2
6
7
[8] M. Stehr, H. Diekmann, L. Smau, O. Seth, S. Ghisla, M. Singh, P. Macheroux, Trends
Biochem. Sci. 23 (1998) 56–57.
2
2
8
9
[9] R.H.H. van den Heuvel, N. Tahallah, N.M. Kamerbeek, M.W. Fraaije, W.J.H. van
Berkel, D.B. Janssen, A.J.R. Heck, J. Biol. Chem. 280 (2005) 32115–32121.
3
3
0
1
[10] D.E. Torres Pazmiño, J.B. Baas, D.B. Janssen, M.W. Fraaije, Biochemistry 47 (2008)
4082–4093.
1
0
Page 10 of 20

1
2
3
4
5
[11] M.P. McLeod, R.L. Warren, W.W.L. Hsiao, N. Araki, M. Myhre, C. Fernandes, D.
Miyazawa, W. Wong, A.L. Lillquist, D. Wang, M. Dosanjh, H. Hara, A. Petrescu, R.D.
Morin, G. Yang, J.M. Stott, J.E. Schein, H. Shin, D. Smailus, A.S. Siddiqui, M.A. Marra,
S.J.M. Jones, R. Holt, F.S.L. Brinkman, K. Miyauchi, M Fukuda, J.E. Davies, W.W. Mohn,
L.D. Eltis, Proc. Natl. Acad. Sci. USA 103 (2006) 15582-15587.
6
7
[12] S. Montersino, W.J.H. van Berkel, Biochim. Biophys. Acta. 1824 (2012) 433-442.
[13] C. Szolkowy, L.D. Eltis, N.C. Bruce, G. Grogan, ChemBioChem 10 (2009) 1208-1217.
8
9
[14] A. Riebel, H.M. Dudek, H.M., G. de Gonzalo, P. Stepniak, L. Rychlewski, M.W.
Fraaije, Appl. Microbiol. Biotechnol. 95 (2012) 1479-1489.
1
1
0
1
[15] G. de Gonzalo, M.D. Mihovilovic, M.W. Fraaije, M.W. ChemBioChem 11 (2010)
2208-2231.
1
1
2
3
[16] N.L. Schlaich, Trends. Plant. Sci. 12 (2007) 412–418.
[17] N.M. Kamerbeek, M.W. Fraaije, D.B. Janssen, FEBS Lett. 271 (2004) 2107-2116.
1
1
4
5
[18] D.E. Torres Pazmiño, A. Riebel, J. de Lange, F. Rudroff, M.D. Mihovilovic, M.W.
Fraaije, ChemBioChem 10 (2009) 2595-2598.
1
1
6
7
[19] J.D. Thompson, D.G. Higgins, T.J. Gibson. T.J. Nucleic Acids Res. 22 (1994) 4673-
4680.
1
1
8
9
[20] C.N. Jensen, J. Cartwright, J. Ward, S. Hart, J.P. Turkenburg, S.T. Ali, M.J. Allen, G.
Grogan, ChemBioChem. 13 (2012) 872-878.
2
2
0
1
[21] D.E. Torres Pazmiño, R. Snajdrova, D.V. Rial, M.D. Mihovilovic, M.W. Fraaije, Adv.
Synth. Catal. 349 (2007) 1361–1368.
2
2
2
3
[22] H.S. Choi, J.K. Kim, E.H. Cho, Y.C. Kim, J.I. Kim, S.W. Kim, Biochem. Biophys.
Res. Commun. 306 (2003) 930-936.
2
2
4
5
[23] M.W. Fraaije, J. Wu, D.P.H.M. Heuts, E.W. van Hellemond, J.H. Lutje Spelberg, D.B.
Janssen, Appl. Microbiol. Biotechnol. 66 (2005) 393–400.
2
2
6
7
[24] S. Franceschini, W.H.L van Beek, A. Pennetta, G. Martinoli, M.W. Fraaije, A. Mattevi,
J. Biol. Chem. 287 (2012) 22626-22634.
28
[25] J. Olucha, A.L. Lamb, Bioorg. Chem. 39 (2011) 171–177.
1
1
Page 11 of 20

1
2
[26] M. Bosello, A. Mielcarek, T.W. Giessen, M.A. Marahiel, Biochemistry 51 (2012)
3059-3066.
3
4
[27] S. Franceschini, M. Fedkenheuer, N.J. Vogelaar, H.H. Robinson, P. Sobrado, A.
Mattevi, Biochemistry 36 (2012) 7043-7045.
5
1
2
Page 12 of 20

1
2
O
S
S
O2
H O
2
CCE-FMO
TrisHCl pH 7.5/cosolvent
O
O
O
O
O
NAD(P)+
O
NAD(P)H
+
O
O
Phosphate
Phosphite
3
4
PTDH
Scheme 1. Oxidations of thioanisole and phenylacetone catalyzed by the novel flavoprotein
5
monooxygenases using NADH or NADPH as coenzymes.
1
3
Page 13 of 20

1
2
3
Table 1. All identified class B monooxygenases with their sequence characteristics and
expression performance in E. coli.
Rossman
motif
GxGxxG
FMO
motif
FxGxxxHxxx /
Rossman
motif
GxGxxG
a
length
(aa)
soluble
monooxygenase accession number
Y
K
F
expression
/
R
b
FMO-A
FMO-B
FMO-C
FMO-D
FMO-E
FMO-F
FMO-G
NMO-H
RHA1_ro00740 GxGxxG
RHA1_ro03334 GxGxxG
RHA1_ro04494 GxGxxG
RHA1_ro05032 GxGxxG
RHA1_ro00824 GxGxxG
RHA1_ro04244 GxGxxG
RHA1_ro05696 GxGxxG
RHA1_ro08654 GxGxxN
WxGxxxHxxxYR
AxIxxxHxxxYR
PxIxxxHxxxYR
FxGxxxHxxxYS
FxGxxxHxxxYD
FxGxxxHxxxHP
FxGxxxHxxxFV
H
GxGxxG
GxGxxG
GxSxxG
GxGxxA
GxGxxA
GxNxxA
GxCxxG
GxGxxG
375
418
365
369
580
602
595
447
++
++
++
b
++
++
+++
+++
+
4
5
6
7
8
9
a
+ indicates a clear visible protein band using SDS-PAGE, while +++ indicates a great
b
overexpression was observed; indicates that the respective monooxygenase has been
expressed using the pCRE2 expression vector while all others have been expressed using the
pBADN-vector [16].
1
4
Page 14 of 20

1
2
Table 2. Baeyer-Villiger oxidation of phenylacetone and (±)-bicyclo[3.2.0]hept-2-en-6-one.
phenylacetone
conv.
(±)-bicyclo[3.2.0]hept-2-en-6-one
conv.
(%)
ratio
ee normal
(%)
ee abnormal
(%)
(%)
normal:abnormal
FMO-E
FMO-F
FMO-G
20
19
5
100
88
79:21
78:22
78:22
9 (1S,5R)
< 1 (1S,5R)
7 (1R,5S)
7 (1R,5S)
70 (1S,5R)
66 (1S,5R)
100
3
4
1
5
Page 15 of 20

1
2
Table 3. Sulfoxidation of thioanisole.
thioanisole
conv. (%)
ee (%)
configuration
FMO-A
FMO-B
FMO-C
FMO-D
FMO-E
FMO-F
FMO-G
5
4
11
7
R
R
R
R
R
-
4
33
25
8
16
56
9
≤5
57
21
S
3
1
6
Page 16 of 20

1
2
3
Figure Captions.
Figure 1. An unrooted phylogenetic tree of the studied monooxygenases and known class B
flavoprotein monooxygenases. The included sequences are: type I BVMOs [EtaA,
Mycobacterium tuberculosis H37Rv (NP_218371.1); HAPMO, Pseudomonas fluorescens
ACB (Q93TJ5.1); CDMO, Rhodococcus ruber SCI (AAL14233.1); CPDMO, Pseudomonas
sp. strain HI-70 (BAE93346.1); CHMO1, Acinetobacter calcoaceticus NCIMB 9871
(BAA86293.1); PAMO, Thermobifida fusca (YP_289549.1); STMO, Rhodococcus
rhodochrous (BAA24454.1); ACMO, Gordania sp. strain TY-5 (BAF43791.1); MEKMO,
Pseudomonas veronii MEK700 (ABI15711.1); CPMO, Comamonas sp. strain NCIMB 9872
(Q8GAW0)], [BVMOs 1-23 from Rhodococcus jostii RHA1 (see [13] for numbering)], type
I FMOs [Arabidopsis thaliana: FMO1Ara(Q9LMA1); FMO2Ara (Q9FKE7); Homo
sapiens: FMO1hum(Q01740); FMO2hum (Q99518); FMO3hum (P31513); FMO4hum
(P31512); FMO5hum (P49326); Drosophila melanogaster: FMO1Dro (Q9W1E9);
FMO2Dro (Q7K3U4); FMO1dog, Canis familiaris (Q95LA2); FMOMeth, Methylophaga
sp. strain SK1 (JC7986); FMOTyr, Tyria jacobaeae (D6CHF7); FMOFish, Oncorhynchus
mykiss (B2LGF9)], NMOs [NRho_Orn, Rhodococcus jostii RHA1 (YP_704660.1);
PVDA_Orn, Pseudomonas aeruginosa (NP_251076); IucD_Lys, Escherichia coli
(YP_444061.1); SidA_Orn, Aspergillus fumigatus (XP_755103.1)], type II FMOs [FMO_X,
Stenotrophomonas maltophilia (B2FLR2); the type II FMOs A-G from Rhodococcus jostii
RHA1 described in this study].
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
1
1
1
1
1
1
1
1
1
1
2
2
2
1
7
Page 17 of 20

1
Rmo
Orn
PvdA
Orn
IucD
Lys
SidA
Orn
H
X
FMO1Ara
FMO2Ara
D
G
F
FMOFish
FMO4hum
E
FMO5hum
FMO3hum
C
B
FMO2hum
FMO1dog
FMO1hum
A
EtaA
FMOMeth
16
FMOTyr
13
FMO1Dro
FMO2Dro
8
1
0
CHMO1
18
24
HAPMO
9
5
15
1
7
STMO
19
1
4
6
PAMO
2
7
4
12
3
1
CPMO
0.1
CDMO
23
11
ACMO
2
0
CPDMO 21
MEKMO
2
3
4
5
Figure 1
1
8
Page 18 of 20

1
2
3
4
5
6
7
8
9
Research highlights
8 newly identified flavoprotein monooxygenases were identified and produced in
recombinant form
Several of the discovered monooxygenases were shown to catalyse enantioselective
sulfoxidations and Baeyer-Villiger oxidations
Sequence and biocatalytic analyses have revealed a new sequence-related group of
flavoprotein monooxygenases that can catalyse Baeyer-Villiger oxidations and sulfoxidation
and that shown no coenzyme preference.
1
9
Page 19 of 20

*Graphical Abstract (for review)
Page 20 of 20