Influence of Substrate Binding Residues on the Substrate Scope and Regioselectivity of a Plant O-Methyltransferase against Flavonoids

Article abstract of DOI:10.1002/cctc.202000471

Methylation of free hydroxyl groups is an important modification for flavonoids. It not only greatly increases absorption and oral bioavailability of flavonoids, but also brings new biological activities. Flavonoid methylation is usually achieved by a specific group of plant O-methyltransferases (OMTs) which typically exhibit high substrate specificity. Here we investigated the effect of several residues in the binding pocket of the Clarkia breweri isoeugenol OMT on the substrate scope and regioselectivity against flavonoids. The mutation T133M, identified as reported in our previous publication, increased the activity of the enzyme against several flavonoids, namely eriodictyol, naringenin, luteolin, quercetin and even the isoflavonoid genistein, while a reduced set of amino acids at positions 322 and 326 affected both, the activity and the regioselectivity of the methyltranferase. On the basis of this work, methylated flavonoids that are rare in nature were produced in high purity.

Full text of DOI:10.1002/cctc.202000471

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Influence of substrate binding residues on the substrate scope
and regioselectivity of a plant O-methyltransferase against
flavonoids
Authors: Qingyun Tang, Yoanes Maria Vianney, Klaus Weisz,
Christoph W. Grathwol, Andreas Link, Uwe T. Bornscheuer,
and Ioannis V Pavlidis
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202000471
Link to VoR: https://doi.org/10.1002/cctc.202000471
01/2020

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
Influence of substrate binding residues on the substrate scope
and regioselectivity of a plant O-methyltransferase against
flavonoids
Qingyun Tang,[a] Yoanes Maria Vianney,[a] Klaus Weisz,[a] Christoph W. Grathwol,[b] Andreas Link,[b]
Uwe T. Bornscheuer*[a] and Ioannis V. Pavlidis*[c]
Dedication: to Prof. Marko Mihovilovic on the occasion of his 50th birthday
Abstract: Methylation of free hydroxyl groups is an important
modification for flavonoids. It not only greatly increases absorption
and oral bioavailability of flavonoids, but also brings new biological
activities. Flavonoid methylation is usually achieved by a specific
group of plant O-methyltransferases (OMTs) which typically exhibit
high substrate specificity. Here we investigated the effect of several
residues in the binding pocket of the Clarkia breweri isoeugenol OMT
on the substrate scope and regioselectivity against flavonoids. The
mutation T133M, identified as reported in our previous publication,
increased the activity of the enzyme against several flavonoids,
namely eriodictyol, naringenin, luteolin, quercetin and even the
isoflavonoid genistein, while a reduced set of amino acids at positions
322 and 326 affected both, the activity and the regioselectivity of the
methyltranferases. On the basis of this work, methylated flavonoids
that are rare in nature were produced in high purity.
on dietary flavonoids greatly increases their absorption and oral
bioavailability through the improvement of their metabolic stability
and better membrane transportation.[4] Methylation of flavonoids
can also bring new biological activities. For example, chrysoeriol
(4′,5,7-trihydroxy-3′-methoxyflavone,
3′-methylluteolin)
and
isohamnetin (3,4′,5,7-tetrahydroxy-3′-methoxyflavone,
3′-methylquercetin) show strong and selective inhibition of the
formation of a carcinogenic estrogen metabolite related to breast
cancer.[5] Besides, eriodictyol (3′,4′,5,7-tetrahydroxyflavanone)
and homoeriodictyol (4′,5,7-trihydroxy-3′-methoxyflavanone, 3′-
methyleriodictyol) are known by their remarkable bitter masking
effect.[6]
Methylation of hydroxyl moieties can be achieved by
S-adenosyl-L-methionine (SAM)-dependent
O-methyltrans-
ferases (OMTs). The methyl group provided by SAM is transferred
to the free hydroxyl groups and methyl ether derivatives are
produced. According to Noel et al., OMTs isolated from plants are
categorized into three classes based on sequence alignments
and structural studies.[7] Type I OMTs accept phenylpropanoids,
flavonoids, isoflavonoids or chalcones as substrates with
molecular weights ranging from 40 to 43 kDa and are not bivalent
cation-dependent. Type II and type III OMTs methylate the
phenylpropanoid-CoA derivatives and the carboxylic acid,
respectively. Although phenylpropanoid OMTs (POMTs),
flavonoid OMTs (FOMTs) and isoflavonoid OMTs (IOMTs) which
all belong to type I OMTs are highly similar regarding their
sequences and protein structures, they are divided into different
OMT groups based on their substrate specificity. POMTs catalyze
methylation of caffeic alcohol/aldehyde/acid which products then
serve for lignin and flavonoid biosynthesis.[8] FOMTs accept a
wide range of flavonoid substrates and show strict regioselectivity.
Isoflavonoids are exclusively produced in leguminous plants and
function as phytoalexins. Only SOMT2 from soy bean (Glycine
max) and SaOMT2 from Streptomyces avermitilis were
discovered to show both flavonoids and isoflavonoids methylation
activities.[9] The crystal structure of an IOMT with isoformononetin
(4′-hydroxy-7-methoxyisoflavone) in the active site shows that
some residues are critical for the stabilization of isoflavonoids.[10]
However, no crystal structure of FOMT is yet available.
Introduction
Flavonoids are a large group of natural polyphenols and
secondary metabolites from plants. They attract a lot of attention
due to their nutritional, health-beneficial and pharmacological
properties including free radical-scavenging antioxidative
activities, anti-inflammatory, antimicrobial and anticancer
activities.[1] Flavonoids are classified into flavonoids
(2-phenylbenzopyrans), isoflavonoids (3-phenylbenzopyrans)
and neoflavonoids (4-phenylbenzopyrans) based on the position
of the phenyl ring. Flavonoids are further divided into flavanes,
flavanones, flavonols, flavones and anthocyanins (Figure 1).[2]
Various modification reactions such as oxidation, hydroxylation,
glycosylation and methylation lead to
a huge variety of
flavonoids.[3] In particular, O-methylation of free hydroxyl groups
[a]
Q. Tang, Y. Vianney, Prof. Dr. K. Weisz, Prof. Dr. U. T. Bornscheuer
Institute of Biochemistry, University of Greifswald
Felix-Hausdorff-Str. 4, 17489 Greifswald, Germany
E-mail: uwe.bornscheuer@uni-greifswald.de
C. W. Grathwol, Prof. Dr. A. Link
Institute of Pharmacy, University of Greifswald
Friedrich-Ludwig-Jahn-Str. 17, 17489 Greifswald, Germany
Assist. Prof. Dr. I. V. Pavlidis
[b]
[c]
In order to study the substrate discrimination between the plant
OMTs, we compared the substrate binding residues of POMT,
FOMT and IOMT with different substrate preferences and
regioselectivities. Based on the differences of these residues, we
have designed and constructed mutants based on the isoeugenol
POMT from Clarkia breweri including the mutation T133M
(IeOMT_T133M).
Dept. of Chemistry, University of Crete
Voutes University Campus, 70013 Heraklion, Greece
E-mail: ipavlidis@uoc.gr
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
Figure 1. Structures of the flavanones eriodictyol and naringenin, the flavone luteolin, the flavonol quercetin and the isoflavone genistein. Ring assignment and
backbone atoms numbering are shown in the structure of eriodictyol.
In a previous work of ours,[11] we identified this variant to have
expanded substrate scope and altered regioselectivity against
phenolic compounds. Thus, it was selected as the starting point
for this study. We discovered that some mutants specifically
methylated the 3′-OH of the tested flavonoids and others further
methylated the 4′-OH, producing dimethylated flavonoids. The
tested mutants also brought activities and different
regioselectivities to the isoflavonoid genistein.
and Val134. Although these two residues in other plant OMTs are
quite diverse, they mostly have Gly/Leu/Met and Ala/Val/Asn at
these two positions. These substrate binding residues bring
proper binding patterns to their preferred substrates and thus
determine substrate specificity and regioselectivity to different
plant OMTs. In order to investigate the influence of these residues,
we constructed the mutants L322H/N/M and Y326H/R/L using the
IeOMT-T133M variant as template. IeOMT is a phenylpropanoid
4-OMT isolated from Clarkia breweri and the variant T133M has
been proved to expand the substrate scope and to enhance the
regioselectivity.[11]
Results and Discussion
Substrate binding residues of different plant OMTs
Methylation reactions mostly take place at the 7-, 3′- and
4′-hydroxyl groups of flavonoids, the 7- and 4′-hydroxyl groups of
isoflavonoids and the 3- and 4-hydroxyl groups of
phenylpropanoids. In order to discover the factors determining the
substrate discrimination of plant OMTs, we chose 21 plant OMTs
from different plant species with different substrate preferences
and regioselectivities for comparison. Sequence alignment shows
that the sequences are extremely diverse between different plant
OMTs, with only 4.6% identity (Figure 2, Figure S1). Since several
IOMT crystal structures has been resolved but no FOMT structure
is available, we chose MeSa-7/4′-IOMT which has been obtained
in a catalytic conformation (PDB code: 1FP2) and investigated
based on it the substrate-enzyme interactions.[10] The ligand
isoformononetin, the 7-methylated product of the isoflavonone
daidzein, is situated in the enzyme active site and well-stabilized
by multiple interactions. Residues critical for substrate binding are
highlighted in Figure 1. Met183 and Met323 constrain the
aromatic A-ring and help positioning the 7-hydroxyl group to the
catalytic residue His272 and SAM. These two residues are
conserved throughout the plant OMT superfamily. Zubieta et al.
suggested that the interaction of the ketone group in the C-ring is
stabilized by the amide side chain of Asn322, a residue that is
only conserved among the 7-IOMTs. Other OMTs rather have
middle size hydrophobic residues, namely Ile, Val or Met, at this
position. Leu326 also interacts with the C-ring of isoformononetin,
but it locates closer to the ether oxygen. Residues at this position
are quite different between the selective OMTs. They are either
the hydrophobic Leu, Val, Ile or Met, or the basic residues Arg or
His. In the absence of Asn322, the basic residues might play an
important role in stabilizing the C-ring ketone group. The
accommodation of the isoflavone B-ring is achieved by Cys133
Figure 2. Part of the sequence alignment of plant flavonoid OMTs (FOMTs),
isoflavonoid OMTs (IOMTs) and phenylpropanoid OMTs (POMTs) performed
with Geneious 10.0.2 (the whole sequence alignment is shown in Figure S1).
Residues involved in substrate binding are highlighted in blue. Numbering of
residues are based on the sequence of the Clarkia breweri isoeugenol OMT
(IeOMT). OMTs are named by their original organisms and regioselectivity.
MePi, Mentha piperita; MeTr, Medicago truncatula; OrSa, Oryza sativa; ArTh,
Arabidopsis thaliana; ChAm, Chrysosplenium americanum; CaRo,
Catharanthus roseus; GlMa, Glycine max; CiAr, Cicer arietinum; MeSa,
Medicago sativa; GlEc, Glycyrrhiza echinate; LoJa, Lotus japonicus; ClBr,
Clarkia breweri; LoPe, Lolium perenne. MeTr_7-FOMT7 is a putative IOMT but
it has higher preference against naringenin (flavanone) than isoflavonoids.
Substrate scope and enzyme activities
Flavonoids are classified into flavane, flavanone, flavone and
flavonol, while isoflavonoids are divided into isoflavone and
isoflavanone, depending on their structures. We have chosen
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
several commonly known compounds eriodictyol and naringenin
(flavanones), luteolin (flavone), quercetin (flavonol) and genistein
(isoflavone) as substrates (Figure 1). Since optically pure
flavanones will racemize in aqueous solution, we used racemic
eriodictyol and naringenin as substrates and obtained racemic
products.[12] Activities of the wild type IeOMT and designed
mutants were tested against these substrates. In each reaction, a
molar excess of SAM and 25% (v/v) Escherichia coli (E. coli) cell
lysate, which contains S-adenosyl-L-homocysteine (SAH)
nucleosidase, were provided in order to increase the yield.[13]
Products were confirmed by either comparing their retention times
on HPLC to commercial standards or structurally characterized
via NMR and MS. Catalytic performance of the mutants are
presented in area percentages calculated by peak area of both
substrates and products measured by HPLC (Figure 3, Figure S2).
It needs to be stated that in an initial screening performed, the
IeOMT_T133M exhibited an expanded substrate scope
compared to the wild-type (converting naringenin and genistein),
while it also exhibited altered regioselectivity. For this reason, the
variant IeOMT_T133M was selected as the template for the
designed mutations.
Figure 3. Product composition of each reaction catalyzed by wild type IeOMT and its mutants. Area percentages were calculated by peak area of both substrates
and products as determined by HPLC (Figure S2). 3′-Me-E, 3′-methyleriodictyol; 3′,4′-DiMe-E, 3′,4′-dimethyleriodictyol; 3′-Me-L, 3′-methylluteolin; 3′,4′-DiMe-L, 3′,4′-
dimethylluteolin; 3′-Me-Q, 3′-methylquercetin; 3′,4′-DiMe-Q, 3′,4′-dimethylquercetin. Structures of these products are confirmed by NMR and LC-MS (see supporting
information). 7-Me-N, 7-methylnaringenin; 7-Me-G, 7-methylgenistein; 4′-Me-G, 4′-methylgenistein. These products were confirmed by comparing their retention
times on HPLC to commercial standards as well as LC-MS. Racemic eriodictyol and naringenin were used as substrates and racemic products were obtained. All
experiments were performed in triplicates and standard deviations were provided.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
Although IeOMT is a phenylpropanoid OMT, it shows high
activities against flavonoids eriodictyol, luteolin and quercetin and
is highly regiospecific to the 3′-hydroxyl group. In the absence of
the 3′-hydroxyl group, wild type IeOMT displays very low activity,
as seen against naringenin. The variant T133M has a universal
effect of increasing the enzymatic activity against these
substrates. Moreover, this variant further methylates the
4′-hydroxyl groups, producing 3′,4′-dimethylated products.
Mutants T133M/Y326H and T133M/Y326L have further increased
production of 3′,4′-dimethylated eriodictyol, luteolin and quercetin.
On the other hand, the variant T133M/L322N retains the
regiospecificity against the 3′-hydroxyl group of all tested
flavonoids and so does the mutant T133M/Y326R towards luteolin
and quercetin. It is interesting to note that the T133M,
T133M/L322M as well as other variants of IeOMT, instead of
methylating the 4′-OH of naringenin, they methylate the 7-OH. We
speculate that the product 2 from naringenin, further produced by
variant T133M/Y326L after 48h, could be another methylated
product or further methylation on the 7-methylnaringenin has
occurred.
against the 4′-hydroxyl group, T133M/Y326H and T133M/Y326L
also produced the 7-methylated genistein. Product 3 was
produced by mutants which displayed both 7- and 4′-methylation
activities especially by variant T133M/Y326L after 48h. According
to the RP-HPLC chromatograms, it was eluted later than the
single methylated genistein (Figure S2). Therefore, we assume
that is the 7, 4′-dimethylated genistein.
Interestingly, both naringenin and genistein missing the 3′-OH
group lead to 7-methylation, but when the B-ring shifts from
position 2 to position 3 of the C-ring, the regioselectivity of some
mutants shift from 7-hydroxyl groups to 4′-hydroxyl groups.
Structural analysis
To gain an insight into the structural differences that lead to the
different regioselectivity of the variants, we performed in silico
analysis. As seen in Figure 4A, the wild type IeOMT can
accommodate eriodictyol, luteolin and quercetin in an orientation
that the methylation of the 3′-position is favored. In all three cases
the distance of the oxygen of 3′-hydroxyl group to the methyl
group to be transferred is between 3.2 and 3.5 Å. It is interesting
The wild type IeOMT displayed very minor activity towards the
isoflavone genistein and the mutants showed different
regioselectivity. While T133M, T133M/L322H (or N or M-
substitutions) and T133M/Y326R exhibited higher selectivity
Figure 4. Models of Clarkia breweri IeOMT variants with SAM (grey sticks with elemental coloring at the left side) and the substrate in the binding pocket. The three
targeted residues are given as lines. For clarity, non-polar hydrogens are removed. A) (S)-eriodictyol (green), luteolin (magenta) and quercetin (blue) docked in the
binding pocket of wild type IeOMT, exposing the 3′-hydroxyl group for methylation. B) the same substrates in the binding pocket of IeOMT_T133M/Y326L, exposing
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
the 4′-hydroxyl group for methylation. C) Genistein binding on IeOMT_T133M/Y326H provides the 7-position for methylation, D) (S)-Naringenin binding pattern in
IeOMT_T133M/L322M also exposes the 7-position, however, with a totally different binding pattern.
to note that eriodictyol and luteolin bind in a similar orientation,
however, quercetin seems to be a little tilted in comparison to
SAM, something that brings also the 4′-position in closer proximity
to the transfer group (3.3 Å to the 3′-OH and 3.4 Å to the 4′-OH
group) and thus the wild-type IeOMT can also produce some
Experimental Section
dimethylated quercetin. In the case of the double mutant
T133M/Y326L the double methylation is increased in all three
substrates. As seen in Figure 4B, the double mutation enabled
the inverse binding of these three flavonoids in the binding pocket,
bringing the 4′-hydroxyl group in the proximity of the methyl group
of the SAM and thus the double methylation is favored. The
reason for this inversed binding seems to be the more
hydrophobic character of the introduced Y326L, which cannot
accommodate the carbonyl of 5-position.
In the case of genistein the T133M increases the activity to a
detectable level, but the mutations in positions 322 and 326 do
not further increase the activity of the enzyme. However, the
double mutant T133M/Y326H has a shift of its regioselectivity to
position 7. As seen in Figure 4C, the substrate is bound with ring
A facing SAM, and the position 7 is closer to the methyl group for
the transfer (3.7 Å). It seems that the histidine at position 326 can
interact with the hydroxyl group at position 5 of genistein (3.0 Å
distance) and thus stabilize the substrate in this orientation to
complete the catalysis.
Materials
DpnI was purchased from New England Biolabs, Inc.. All primers were
purchased from Thermo Fisher Scientific, Inc.. Competent cells were
self-prepared. S-Adenosyl-L-methionine was purchased from J&K
Chemical Ltd.. Acetic acid, trifluoroacetic acid (TFA) and acetonitrile were
purchased from Carl Roth GmbH and VWR chemicals and were of HPLC
grade. (±)-Eriodictyol was purchased from Carl Roth GmbH, (±)-naringenin
and quercetin were purchased from Sigma, luteolin, genistein and
biochanin A (4′-methoxygenistein) were purchased from TCI GmbH,
prunetin (7-methoxygenistein) and sakuranetin (7-methoxynaringenin)
were purchased from Extrasynthese. All these chemicals have 95% or
higher purity.
Mutagenesis, expression and purification
IeOMT from Clarkia breweri (accession number of protein: O04385.2) and
the mutant T133M in pET-21a(+) with His6-tag at the C-terminal were
constructed as described in our previous work.[11] Further mutagenesis
were performed following the QuikChange (Stratagene) protocol using
pET-21a(+)-T133M as template. Primers are as below:
Naringenin differs from eriodictyol only by the lack of the
3′-hydroxyl group. Thus, although it can bind the same way in the
active site of the wild type, the B-ring cannot be methylated to a
3′-methoxy derivative and thus the wild type is almost inactive.
However, the mutation T133M (and the T133M/L322M) enabled
a different binding pattern, where the position 7 of the ring A is
accessible to the SAM and the 7-hydroxyl group at catalytic
distance (3.4 Å). The binding pattern differs for genistein, for
which the enzyme exhibited the same regioselectivity. In the case
of variant T133M/L322M (Figure 4D), the reason for this can be
the lack of the histidine at position 326 that could interact with the
5-OH, in combination with the methionine in position 322 that
pushes away the substrate. As naringenin is not planar, the ring
B may cause steric clashes with 322 and thus the substrate binds
in a different orientation.
L322H fw: 5′-CACACCGATGCCCATATGCTGG-3′;
L322H rv: 5′-CCAGCATATGGGCATCGGTGTG-3′;
L322M fw: 5′-CACACCGATGCGATGATGCTGG-3′;
L322M rv: 5′-CCAGCATCATCGCATCGGTGTG-3′;
L322N fw: 5′-CACCGATGCGAACATGCTGGCGTATAAC-3′;
L322N rv: 5′-GTTATACGCCAGCATGTTCGCATCGGTG-3′;
Y326H fw: 5′-TGCTGGCGCATAACCCGGGCGGTAAAG-3′;
Y326H rv: 5′-CTTTACCGCCCGGGTTATGCGCCAGCA-3′;
Y326L fw: 5′-TGCTGGCGCTGAACCCGGGCGGTAAAG-3′;
Y326L rv: 5′-CTTTACCGCCCGGGTTCAGCGCCAGCA-3′;
Y326R fw: 5′-TGCTGGCGCGTAACCCGGGCGGTAAAG-3′;
Y326R rv: 5′-CTTTACCGCCCGGGTTACGCGCCAGCA-3′;
After PCR, the pET-21a(+)-T133M templates were digested by DpnI
before transforming into Escherichia coli (E. coli) Top 10 chemically
competent cells for plasmid amplification. After sequence confirmation
(Eurofins, Germany), the newly constructed mutant plasmids were
transformed into E. coli Bl21 (DE3) chemically competent cells for protein
expression. Expression and purification of IeOMT and all mutants were
carried out following the protocols given in our previous work.[11]
Conclusions
In this study we highlighted the potential of IeOMT,
a
Enzyme assays and HPLC analysis
phenylpropanoid OMT to be engineered to catalyze the
methylation of a range of flavonoids and isoflavonoids. Three
positions, namely 133, 322 and 326 have a significant impact on
the binding of the substrate and variants that were produced with
semi-rational design in these positions had altered catalytic
activity and/or regioselectivity compared to the wild-type enzyme.
These variants provided access to methylated flavonoids that are
rare in nature and may have interesting biological activities due to
their methylation pattern.
The reaction mixture consisted of 150 µM flavonoid substrates, 1 mM SAM,
5 mM DTT and 25% (v/v) E. coli Bl21 (DE3) cell lysate (not including the
MTs of interest), in 50 mM sodium phosphate buffer, pH 7.5. Because the
S-adenosyl-l-homocysteine (SAH) derived from SAM is a potent inhibitor
for MTs, E. coli cell lysate which contains SAH nucleosidase was provided
in each reaction in order to reduce the inhibition effect brought by the
SAH.[13] 1.0 mg/mL (24.4 µM) purified enzymes were added into the
reaction mixtures to initiate the enzyme reaction. Due to the instability of
SAM, same amount of SAM was supplied again at 24 h. The mixtures were
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
incubated at 28°C with 700 rpm agitation in an Eppendorf Thermomixer.
Assays were performed in triplicate and negative controls were performed
by replacing purified enzymes with buffer. Samples were taken at 2, 4, 24
and 48 h and an equal volume of acetonitrile was added to quench the
reaction. Samples were vortexed vigorously and then centrifuged at full
speed for 30 min to remove protein precipitate. 200 µL supernatant were
transferred to HPLC sample vial inserts for analysis. HPLC analysis were
performed on VWR Hitachi Elite LaChrom system equipped with the
Kinetex EVO C18 (4.6x250 mm column, 5 µm particle size, Phenomenex)
reversed-phase column. 0.1% acetic acid and acetonitrile were used for
the separation of flavonoid substrates and corresponding methylated
products, with the ratio 68:32 (v/v) for eriodictyol, luteolin and quercetin,
and 60:40 (v/v) for naringenin and genistein. Wavelengths for the
detections of flavanones (eriodictyol and naringenin), flavone and flavonol
(luteolin and quercetin) and isoflavone (genistein) were 280, 260 and 260
nm respectively. All analyses were performed at a flow rate 1 mL/min and
the column temperature was 35°C. Identification of methylated products
were confirmed by comparing their retention times on HPLC to commercial
standards. For unknown products, large-scale reactions were performed
with specific substrates and mutants and the products were isolated and
identified by NMR and MS. Area percentages of substrate and each
product were calculated to show the estimated yield for each product.
against known standards. For the double-methylated products, additional
2D NMR experiments (NOESY, HSQC, HMBC) were performed. Data
were processed and analyzed using TopSpin 4.0.7. All spectra were
recorded at 25°C. All peaks were referenced towards the DMSO-d6 peak
(1H: 2.50 ppm; 13C: 39.51 ppm).
Mass spectrometry was performed on Shimadzu LC-MS 8030 equipped
with an ESI (electrospray ionization) source and a mass spectrometer,
using the same column and mobile phase as in the HPLC analysis.
Negative ionization mode was used and automatic MRM (multiple reaction
monitoring) optimization was performed to acquire optimal fragmentation
and maximal transmission of the desired product ions.
Bioinformatic analysis
The bioinformatic analysis was performed with YASARA 19.7.20. First the
structure of 3REO was back-mutated to its wild-type sequence and the
SAH was transformed to SAM by the addition of the methyl group and this
structure was refined at pH 7.5, 25°C for 500 ps, taking a snapshot every
25 ps. The structure with the lowest energy was selected for the further
experiments. For the mutants, the respective amino acids were swapped
with subsequent energy minimization. The same was performed for the
preparation of the substrate molecules. The docking experiments were
performed with VINA method, using the force field AMBER03 at 30°C. Five
receptor molecules were prepared for each experiment and in each
receptor 25 dockings were performed, and the resulting structures were
clustered when RMSD was <5 Å. The catalytic active conformation with
the higher binding energy was selected. Figures were prepared with PyMol.
Biosynthesis, isolation and purification of unknown methylated
products
Each large-scale reaction started with 30 mg flavonoid substrate (~100
µmol), 80 mg SAM (200 µmol), 20 mg purified enzyme, 0.1% NaN3 and
crude E. coli cell lysate in 100 mL sodium phosphate buffer (50 mM, pH
7.5). Purified IeOMT_T133M/Y326R was used for the biosynthesis of the
3′-methylated products of eriodictyol, luteolin and quercetin.
IeOMT_T133M/Y326L was used to synthesize the 3′,4′-dimethylated
products of eriodictyol, luteolin and quercetin. Reactions were performed
in 500 mL round bottom flasks at 30°C at 180 rpm for a total duration of 10
days, and the reaction progress was followed every 24 h by HPLC. The
same amount of SAM and purified enzyme were resupplied at specific time
intervals, when reaction seemed to slow down. When reactions were
finished, substrates and products were extracted with 100 mL ethyl acetate
(EtOAc) for three times, water was removed with dried MgSO4 and then
filtered. EtOAc were then removed by rotary evaporator and the remaining
dry substances were dissolved in 1 mL DMSO for subsequent purification.
Separation of flavonoid substrates and products were performed firstly
with the analytical column LiChrospher® 100 RP-18 (5 µm) LiChroCART®
(250×4 mm, Merck) and then with the preparative column LiChrospher®
100 RP-18 (5 µm) Hibar® RT (250×25 mm, Merck), both are equipped with
Shimadzu devices CBM-20A, LC-20A P, SIL-20A, FRC-10A and a SPD
20A UV/Vis detector. 0.1% TFA and acetonitrile (60/40, v/v) were used as
mobile phase. Product fractions were collected, dried with rotary
evaporator and lyophilizer before submission to NMR. All products had a
purity of >99%, according to HPLC. The isolated yields (not optimized
procedures) were the following: 3′-methyleriodictyol: 18.8 % (5.9 mg), 3′,4′-
dimethyleriodictyol: 35.5% (11.7 mg), 3′-methylluteolin 24.6% (15.5 mg),
3′,4′-dimethylluteolin 26.9% (17.7 mg), 3′-methylquercetin 6.8% (4.3 mg),
3′,4′-dimethylquercetin 16.8% (11.0 mg).
Accession numbers
The accession numbers of plant OMTs chosen for sequence alignment in
Figure
1 are as follows: MePi_7-FOMT1 from Mentha piperita
(AAR09598.1), MeTr_7-FOMT7 from Medicago truncatula (ABD83946.1),
OrSa_7-FOMT from Oryza sativa (BAM13734.1), ArTh_3′-FOMT from
Arabidopsis
thaliana
(AAB96879.1),
ChAm_3′-FOMT
from
Chrysosplenium americanum (AAA80579.1), MePi_3′-FOMT3 from
Mentha piperita (AAR09601.1), OrSa_3′-FOMT from Oryza sativa
(XP_015650053.1), CaRo_4′-FOMT from Catharanthus roseus
(AAR02420.1), GlMa_4′-FOMT/IOMT from Glycine max (C6TAY1.1),
MePi_4′-FOMT4 from Mentha piperita (AAR09602.1), CiAr_7-IOMT from
Cicer arietinum (XP_004489528.1), MeTr_7-IOMT1 from Medicago
truncatula (AAY18582.1), MeSa-7/4′-IOMT from Medicago sativa
(AAC49928.1), GlEc_4′-IOMT from Glycyrrhiza echinate (BAC58011.1),
LoJa_4′-IOMT from Lotus japonicus (BAC58013.1), MeTr_4′-IOMT5 from
Medicago truncatula (AAY18581.1), CaRo_3-POMT from Catharanthus
roseus (AAK20170.1), ClBr_3-POMT from Clarkia breweri (O23760.1),
LoPe_3-POMT from Lolium perenne (AAD10253.1), MeSa_3-POMT from
Medicago sativa (AAB46623.1), ClBr_4-POMT from Clarkia breweri
(O04385.2).
Acknowledgements
NMR and mass spectrometry analysis
Q.T. would like to thank the China Scholarship Council for
financial support of her PhD thesis project (File No.:
201606150073).
NMR measurements were carried out on a Bruker Avance 600 MHz
spectrometer equipped with an inverse 1H/13C/15N/19
F quadruple
resonance cryoprobehead and z-field gradients. All compounds were
dissolved in DMSO-d6 and 1H NMR, 13C NMR and DEPT-135 experiments
were performed with all compounds dissolved in DMSO-d6. Identification
of single methylated products was done by comparing assignments
Keywords: biocatalysis • methyltransferases • flavonoid •
regioselectivity • protein engineering
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
[1]
a) C. Kandaswami, E. Middleton, in Free radicals in diagnostic
[6]
[7]
a) J. P. Ley, U.S. Patent 8,685,436, 2014; b) J. P. Ley, G. Krammer,
G. Reinders, I. L. Gatfield, H. J. Bertram, J. Agric. Food Chem. 2005,
53, 6061.
medicine (Ed.: D. Armstrong), Springer, 1994, pp. 351-376; b) D.
Procházková, I. Boušová, N. Wilhelmová, Fitoterapia 2011, 82, 513;
c) S. Kaur, P. Mondal, J. Microbiol. Exp. 2014, 1, 23; d) A. García-
Lafuente, E. Guillamón, A. Villares, M. A. Rostagno, J. A. Martínez,
Inflammation Res. 2009, 58, 537; e) M. H. Pan, C. S. Lai, C. T. Ho,
Food Funct. 2010, 1, 15; f) M. K. Chahar, N. Sharma, M. P. Dobhal,
Y. C. Joshi, Pharmacogn. Rev. 2011, 5, 1; g) D. Raffa, B. Maggio,
M. V. Raimondi, F. Plescia, G. Daidone, Eur. J. Med. Chem. 2017,
142, 213.
J. P. Noel, R. A. Dixon, E. Pichersky, C. Zubieta, J. L. Ferrer, in
Recent advances in phytochemistry, Vol. 37, Elsevier, 2003, pp. 37-
58.
[8]
[9]
C. Zubieta, P. Kota, J. L. Ferrer, R. A. Dixon, J. P. Noel, Plant Cell
2002, 14, 1265.
a) D. H. Kim, B. G. Kim, Y. Lee, J. Y. Ryu, Y. Lim, H. G. Hur, J. H.
Ahn, J. Biotechnol. 2005, 119, 155; b) B. G. Kim, B. R. Jung, Y. Lee,
H. G. Hur, Y. Lim, J. H. Ahn, J. Agric. Food Chem. 2006, 54, 823.
C. Zubieta, X. Z. He, R. A. Dixon, J. P. Noel, Nat. Struct. Mol. Biol.
2001, 8, 271.
[2]
[3]
A. N. Panche, A. D. Diwan, S. R. Chandra, J. Nutr. Sci. 2016, 5.
a) K. Saito, K. Yonekura-Sakakibara, R. Nakabayashi, Y. Higashi,
M. Yamazaki, T. Tohge, A. R. Fernie, Plant Physiol. Biochem. 2013,
72, 21; b) M. L. Falcone Ferreyra, S. Rius, P. Casati, Front. Plant
Sci. 2012, 3, 222.
[10]
[11]
[12]
[13]
Q. Tang, U. T. Bornscheuer, I. V. Pavlidis, ChemCatChem 2019, 11,
3227.
[4]
[5]
a) T. Walle, Mol. Pharm. 2007, 4, 826; b) S. H. Thilakarathna, H. P.
Rupasinghe, Nutrients 2013, 5, 3367.
J. A. Yáñez, N. D. Miranda, C. M. Remsberg, Y. Ohgami, N. M.
Davies, J. Pharm. Biomed. Anal. 2007, 43, 255.
N. Parveen, K. A. Cornell, Mol. Microbiol. 2011, 79, 7.
a) H. Takemura, T. Itoh, K. Yamamoto, H. Sakakibara, K. Shimoi,
Bioorg. Med. Chem. 2010, 18, 6310; b) H. Takemura, H. Uchiyama,
T. Ohura, H. Sakakibara, R. Kuruto, T. Amagai, K. Shimoi, J. Steroid
Biochem. Mol. Biol. 2010, 118, 70.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000471
ChemCatChem
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Qingyun Tang, Yoanes Maria Vianney,
Klaus Weisz, Christoph W. Grathwol,
Andreas Link, Uwe T. Bornscheuer* and
Ioannis V. Pavlidis*
CH3
CH3
OH
O
O
3'
3'
OH
OH
O
4'
CH3
HO
O
O
HO
O
O
HO
T133M/Y326L
O
O
WT-IeOMT
OH
OH
OH
OH
OH
OH
SAM
SAH
SAM
SAH
Page No. – Page No.
Flavonoid
3'-Methylated flavonoid
3',4'-Dimethylated flavonoid
Influence of substrate binding
residues on the substrate scope and
regioselectivity of a plant O-
Designed pattern: natural but rare flavonoids with unusual methoxy/hydroxy-group
substitution pattern can be made available for in depth-exploitation of molecular
properties using engineered enzymes.
methyltransferase against flavonoids
Twitter accounts:
University of Crete: @UOC_gr (https://twitter.com/uoc_gr)
University of Greifswald: @Wissen_lockt (https://twitter.com/Wissen_lockt)
This article is protected by copyright. All rights reserved.