Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two potential routes to nevadensin in sweet basil

Article abstract of DOI:10.1016/j.phytochem.2013.05.001

Regioselective 6-,7-,8-,3'-, and 4'-O-methylations underlie the structural diversity of lipophilic flavones produced in the trichomes of sweet basil (Ocimum basilicum L.). The positions 6, 7, and 4' are methylated by a recently described set of cation-independent enzymes. The roles of cation-dependent O-methyltransferases still require elucidation. Here, the basil trichome EST database was used to identify a Mg²⁺-dependent O-methyltransferase that was likely to accept flavonoids as substrates. The recombinant protein was found to be active with a wide range of o-diphenols, and methylated the 8-OH moiety of the flavone backbone with higher catalytic efficiency than the 3'-OH group of candidate substrates. To further investigate flavone 8-O-methylation, the activity of a putative cation-independent flavonoid 8-O-methyltransferase from the same EST collection was assessed with available substrate analogs. Notably, it was strongly inhibited by gardenin B, one of its expected products. The catalytic capacities of the two studied proteins suggest that two alternative routes to nevadensin, a major flavone in some basil cultivars, might exist. Correlating the expression of the underlying genes with the accumulation of 8-substituted flavones in four basil lines did not clarify which is the major operating pathway in vivo, yet the combined data suggested that the biochemical properties of flavone 7-O-demethylase could play a key role in determining the reaction order.

Full text of DOI:10.1016/j.phytochem.2013.05.001

Phytochemistry xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Characterization of two candidate flavone 8-O-methyltransferases
suggests the existence of two potential routes to nevadensin in
sweet basil
⇑
Anna Berim, David R. Gang
Institute of Biological Chemistry, Washington State University, Pullman, WA 99164, USA
a r t i c l e i n f o
a b s t r a c t
Regioselective 6-,7-,8-,30-, and 40-O-methylations underlie the structural diversity of lipophilic flavones
produced in the trichomes of sweet basil (Ocimum basilicum L.). The positions 6, 7, and 40 are methylated
by a recently described set of cation-independent enzymes. The roles of cation-dependent O-methyl-
transferases still require elucidation. Here, the basil trichome EST database was used to identify a
Mg2+-dependent O-methyltransferase that was likely to accept flavonoids as substrates. The recombinant
protein was found to be active with a wide range of o-diphenols, and methylated the 8-OH moiety of the
flavone backbone with higher catalytic efficiency than the 30-OH group of candidate substrates. To further
investigate flavone 8-O-methylation, the activity of a putative cation-independent flavonoid 8-O-methyl-
transferase from the same EST collection was assessed with available substrate analogs. Notably, it was
strongly inhibited by gardenin B, one of its expected products. The catalytic capacities of the two studied
proteins suggest that two alternative routes to nevadensin, a major flavone in some basil cultivars, might
exist. Correlating the expression of the underlying genes with the accumulation of 8-substituted flavones
in four basil lines did not clarify which is the major operating pathway in vivo, yet the combined data sug-
gested that the biochemical properties of flavone 7-O-demethylase could play a key role in determining
the reaction order.
Article history:
Received 28 December 2012
Received in revised form 7 March 2013
Accepted 8 May 2013
Available online xxxx
Keywords:
Ocimum basilicum L.
Lamiaceae
Sweet basil
Biosynthetic network elucidation
Lipophilic flavonoids
O-Methyltransferase
Nevadensin
Gardenin B
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
Ibrahim, 1982, 1985; Jay et al., 1985). However, plant Mg2+-dependent
OMTs studied at the molecular level were thought to be specific
Small molecule O-methyltransferases (OMTs) are subdivided
into three classes (Noel et al., 2003). Class 1 and 2 OMTs are in-
volved in methylation of mostly phenolic hydroxyl residues,
whereas class 3 OMTs methylate carboxyl groups to yield methyl
esters. The proteins are assigned to the different classes based on
molecular size, cation requirement, and structure: class 1 com-
prises OMTs with a subunit molecular weight of 38–45 kDa and
no cation dependency, while OMTs with smaller subunit size
(22–27 kDa) and requiring a cation, usually magnesium, for activ-
ity belong to class 2. All of the above mentioned OMTs use S-aden-
for a single substrate, caffeoyl coenzyme A thioester (hence the des-
ignation CCoAOMTs or CCOMTs) and therefore to play pivotal roles
in the biosynthesis of monolignols and related phenylpropanoids
(Weng and Chapple, 2010), until a Mg2+-dependent OMT from ice
plant (Mesembryanthemum crystallinum) was shown to catalyze
the O-methylation of the flavonol quercetagetin at positions 6 and
30 (Ibdah et al., 2003). This CCoAOMT-like enzyme, as well as several
putative homologs isolated in the same investigation, was also ac-
tive with a number of phenylpropanoids and flavonoids, and formed
a phylogenetically divergent OMT clade designated phenylpropa-
noid and flavonoid OMTs (PFOMTs) to reflect their multiple func-
tions. Subsequently, other members of the PFOMT subclass were
found to mediate the 30-O- and in some cases also 50-O-methylation
of various subtypes of flavonoids, such as anthocyanins, flavonols,
and flavones (Hugueney et al., 2009; Lee et al., 2008; Lucker et al.,
2010; Widiez et al., 2011). In addition, an OMT of this subclass con-
tributes to the formation of spermidine–phenylpropanoid conju-
gates in Arabidopsis (Fellenberg et al., 2008). Some of the classical
CCoAOMTs are capable of methylating flavonoids, albeit with lower
efficiency than caffeoyl-CoA (Kim et al., 2010; Lukacin et al., 2004).
osyl-L-methionine (SAM) as methyl group donor.
Early investigations of polymethoxylated flavonoid biosynthesis
in golden saxifrage (Chrysosplenium americanum) and birdsfoot tre-
foil (Lotus corniculatus) suggested that some of the involved OMTs,
e.g. the 6- and 8-OMT, required Mg2+ to be active (De Luca and
Abbreviations: SAM, S-adenosyl-L-methionine; (PF)OMT, (phenylpropanoid–
flavonoid) O-methyltransferase.
⇑
Corresponding author. Tel.: +1 509 335 0550; fax: +1 509 335 7643.
E-mail address: gangd@wsu.edu (D.R. Gang).
0031-9422/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.phytochem.2013.05.001
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

2
A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
Sweet basil (Ocimum basilicum L.) accumulates lipophilic flav-
In this study, the basil trichome EST database (Gang et al., 2001)
was used to identify representatives of class 2 OMTs in this cell
type. Biochemical analysis of the most abundantly expressed can-
didate PFOMT indicated that it is a promiscuous enzyme whose
in vivo role might depend on the availability of the accepted sub-
strates and/or their subcellular localizations. As one of its potential
functions is that of a flavonoid 8-O-methyltransferase (8-OMT), the
activities of a candidate class 1 flavonoid 8-OMT were also evalu-
ated. In addition, the correlation between the expression of the
genes encoding these two proteins and the accumulation of the
ones with hydroxyl groups at positions 5, 6, 7, 8, 30, and 40 (Grayer
et al., 1996, 2001) (Table 1, compounds 1–7, 13–16) whose produc-
tion appears to be confined to peltate trichomes (Berim et al.,
2012). With the exception of the 5-OH, all hydroxyl residues are
methylated in at least one of the accumulated flavones. While
the recently investigated family of cation-independent OMTs was
found to catalyze regioselective methylations at positions 6, 7,
and 40 (Berim et al., 2012), the substrate- and regiospecificities of
cation-dependent OMTs in basil have not yet been studied.
Table 1
Structures of compounds used as substrates in this study and mentioned in the text, and relative activities of ObPFOMT-1 and ObF8OMT-1 with relevant compounds.
R
1
3´
R
2
R
4´
5´
8
B
8
R
7
6
O
C
2
3
7
R
3
A
5
R
R
4
4
6
R
O
5
Name
R1
R2
R3
R4
R5
R6
R7
R8
Rel. activity (%) with
ObPFOMT-1
ObF8OMT-1
1
2
3
4
5
6
7
8
Gardenin B
Nevadensin
8-Hydroxy-salvigenin
Pilosin
Salvigenin
Cirsimaritin
Ladanein
Pectolinarigenin
Scutellarein-7-methyl ether
Scutellarein-40-methyl ether
Scutellarein
Scutellarin
Eupatorin
Cirsilineol
Cirsiliol
Genkwanin
Nepetin
Diosmetin
Chrysoeriol
H
H
H
H
H
H
H
H
H
H
H
H
OH
OCH3
OH
H
OH
OH
OCH3
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
H
OCH3
OCH3
OCH3
OCH3
OCH3
OH
OCH3
OCH3
OH
OCH3
OH
OH
OCH3
OH
OH
OH
OH
OCH3
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OH
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OH
OH
OH
H
H
H
H
H
H
H
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
H
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OH
OCH3
OH
OH
OH
OH
OCH3
OCH3
OCH3
H
OCH3
H
H
H
H
H
OH
H
H
H
OCH3
OH
OCH3
OH
OCH3
OCH3
OCH3
OH
OCH3
OH
OCH3
OCH3
OH
OH
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OH
OH
OH
H
H
H
10.5 0.4
9
11.4 2.1
122.5 20
65.9 0.5
10.0 1.3
10
11
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
37
OH
OGlcA
OCH3
OCH3
OCH3
OCH3
OH
OH
OH
OH
OCH3
OGlc
OH
OH
OCH3
OH
H
H
OH
OH
64.5 0.4
120.0 3.4
Luteolin
100.0 4.5
38.1 0.3
57.3 2.4
13.8 1.0
10.1 0.4
11.2 0.6
49.4 1.1
104.8 2.8
64.3 4.6
167.3 4.4
121.6 0.9
222.2 3.9
147.9 7.3
54.2 1.9
4.0 1.0
Luteolin-7-methyl ether
Luteolin-7-glucoside
Quercetagetin
Quercetin
Quercetin-7-methyl ether
Tricetin
30,40-OH-flavone
5,30,40-OH-flavone
7,30,40-OH-flavone
7,8,30,40-OH-flavone
7,8,40-OH-flavone
7,8-OH-flavone
6,7-OH-flavone
5,6-OH-flavone
5,6-OH-7-OCH3-flavone
8-OH-7-OCH3-flavone
Eriodictyol (C2-C3 bond saturated)
H
H
H
H
H
H
OH
OH
OH
H
OH
H
H
H
H
12.6 1.7
100.0 1.4
32.1 1.4
OH
OH
OH
H
OCH3
OCH3
OH
H
H
H
H
H
OH
H
H
H
H
H
OH
OH
H
2.9 0.3
OH
H
62.2 0.6
OH
H
OH
H
4.6 0.2
OH
R
OH
Name
R
Rel. activity (%) with
ObPFOMT-1
ObF8OMT-1
38
39
40
Caffeic acid
3,4-OH-benzaldehyde
Catechol
CH2–CH@CH–COOH
CHO
H
78.0 2.5
16.4 0.2
3.3 0.5
Flavonoid ring nomenclature and carbon numbering are indicated on the generic backbone structure. OGlcA, O-glucuronide; OGlc, O-glucoside. The primary methyl acceptor
residue is shown in bold. Turnover rates of ObPFOMT-1 are presented relative to that with luteolin (100% = 1.54 nkat mgꢀ1 protein), all substrates were supplied at 50
lM.
Turnover rates of ObF8OMT-1 (B) are presented relative to that with 7,8,40-OH-flavone (100% = 0.37 nkat mgꢀ1protein), all substrates were supplied at 100
means S.E. (n = 3).
lM. Results are
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
3
fragments (Supplemental Fig. S1). This high identity level strongly
suggests functional redundancy. The most abundantly expressed
contig (basil_0053) comprises 0.18% of the total ESTs, which
amounts to about 10-fold of the abundance of ESTs in the other
full-length PFOMT contig (basil_0750, Fig. 1). Basil PFOMTs are also
ca. 60% identical to PFOMT from ice plant (Ibdah et al., 2003) and
other known PFOMTs. Phylogenetic analysis confirmed that they
belong to the PFOMT subfamily of the Mg2+-dependent plant
OMTs, and indicated that they cluster best with uncharacterized
putative PFOMTs from castor bean (Ricinus communis) and grape
(Vitis vinifera).
2.2. Biochemical characterization of ObPFOMT-1
This work focused on the protein encoded by the major contig
basil_0053, which was provisionally designated ObPFOMT-1. The
recombinant protein was produced in E. coli and purified using
an N-terminal hexahistidine-tag (Supplemental Fig. S2). First anal-
yses showed that ObPFOMT-1 methylates the 30-OH moiety of lute-
olin (20) and cirsioliol (15), two potential natural substrates (Fig. 2,
traces 1, 2, Table 1). Its activity is highest in mildly alkaline solu-
tion at 30 °C (Supplemental Table S1). The optimal concentration
of Mg2+ ions is different for different substrates (Supplemental
Fig. S3). The cation concentration may therefore impact the reac-
tion rates with competing flavones at the cellular level.
To find out what structural features are necessary for substrate
recognition, ObPFOMT-1 was incubated with luteolin derivatives
lacking individual hydroxyl moieties, as well as a series of ortho-
diphenols of varying structure. In summary, ObPFOMT-1 appears to
require vicinal hydroxyl groups and an electron-withdrawing moiety
positioned meta to the methyl acceptor OH-group, as it was active
with caffeic acid (38), 3,4-dihydroxylbenzaldehyde (39), and 30,40-
OH-flavone (27), butonlyverylowactivitycouldbedetected with eri-
odictyol (37), a flavanone derivative of luteolin (20) where the conju-
gation to the carbonyl moiety at C4 is interrupted, and with catechol
(40), the simplest ortho-diphenol (Table 1, Supplemental Fig. S5). The
majority of other characterized PFOMTs were found to display simi-
larly broad substrate specificity (Fellenberg et al., 2008; Ibdah et al.,
2003; Lee et al., 2008; Widiez et al., 2011).
Fig. 1. Molecular phylogenetic analysis of selected cation-dependent OMTs by
maximum likelihood method. Phylogeny was inferred using MEGA5 as described in
Experimental, and tested by bootstrapping (500 replicates). Only bootstrap values
greater than 70% are presented. Scale is in substitutions per amino acid. Basil
proteins whose activities have been assessed are in bold. FkbG OMT from
Streptomyces hygroscopicus (AAF86386Sh) was used as outgroup to root the tree.
Branches are labeled with NCBI GenBank™ accession number and genus and
species initials. Abbreviations are: Rc: Ricinus communis, Ptr: Populus trichocarpa,
Vv: Vitis vinifera, Sl: Stellaria longipes, Mc: Mesembryanthemum crystallinum, Vp:
Vanilla planifolia, At: Arabidopsis thaliana, Ppp: Physcomitrella patens subsp. patens,
Mt: Medicago truncatula, Ptm: Populus tremuloides, Eg: Eucalyptus gunnii, St:
Solanum tuberosum, Nt: Nicotiana tabacum, Pc: Petroselinum crispum, Am: Ammi
majus, Gh: Gossypium hirsutum, Pr: Pinus radiata, Pt: Pinus taeda, Pp: Pinus pinaster,
Pa: Picea abies, Osj: Oryza sativa cv. japonica, Bd: Brachypodium distachion, Syn:
Synechocystis sp. (strain PCC 6803).
2.2.1. Regioselectivity of ObPFOMT-1
Basil accumulates flavones whose backbones are additionally
substituted at positions 6 and 8. ObPFOMT-1 was therefore tested
with a wide range of natural and unnatural flavones possessing
two or three vicinal hydroxyl groups. These tests indicated that its
regioselectivity varies with the structural context. In 5,6,7-trihydr-
oxyflavones whose 7-OH moiety is methylated, such as scutella-
rein-7-methyl ether (9) and ladanein (7), the 5-OH moiety acts as
the major methyl acceptor (Fig. 2 trace 3, Supplemental Fig. S4). Both
of these substrates (7 and 9) belong to the flavone network in basil,
yet their 5-O-methylated derivatives have not been reported to oc-
cur, suggesting that this activity of ObPFOMT-1 has little or no rele-
vance in vivo. In contrast, with unmethylated scutellarein (11) as
substrate, it selectively catalyzes 6-O-methylation (Supplemental
Fig. S4, trace 1). Scutellarin (12), the 7-O-glucuronide of scutellarein,
is also predominantly methylated at position 6 despite the bulky vic-
inal sugar moiety (Supplemental Fig. S4, trace 2). The first methyla-
tion of quercetagetin (23) is also highly selective for the OH moiety
at position 6 ratherthan 30 (Supplemental Fig. S4). Remarkably, incu-
bation with tricetin (26), a flavone with 30,40,50-hydroxylated ring B
not accumulated by basil, results in three distinct products, one
monomethylated and two dimethylated (Supplemental Fig. S4, trace
5). Because of this substrate’s physiological irrelevance, unambigu-
ous identification of these products was not pursued at this time
point. However, a comparison with methylated tricetin derivatives
used in the characterization of a class 1 OMT from wheat (Zhou
relevant 8-substituted flavones in young leaves of four basil lines
was analyzed.
2. Results
2.1. Analysis of cation-dependent O-methyltransferases in the basil
EST database
The basil trichome EST database (Gang et al., 2001) contains
representatives of both CCoAOMT and PFOMT subclasses of the
cation-dependent OMT family. Basil CCoAOMT contigs are abun-
dantly represented, comprising more than 0.45% of total ESTs in
the database. The encoded proteins share more than 90% identity
with other known eudicot CCoAOMTs (Fig. 1). The protein desig-
nated ObCCoAOMT-1, encoded by a major contig basil_0029, was
found to be active with caffeoyl-CoA (not shown). It is 91% identi-
cal to another full-length CCoAOMT contig (basil_0110, ObC-
CoAOMT-2) whose activities have not yet been assessed. Basil
PFOMTs share ca. 60% identity with the basil CCoAOMTs, and are
represented by eight contigs, two of them full-length, in the data-
base. The encoded proteins form two groups differing essentially
by four amino acids, equaling 98–100% identity on available
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

4
A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
A
C
D
E
F
B
Fig. 2. Activites of ObPFOMT-1 and ObF8OMT-1 with selected substrates. A. and B: Representative traces of reaction products of ObPFOMT-1 (A, traces 1–7) and ObF8OMT-1
(B, traces 8–10). Peaks labeled S are substrates, P are products. Methylation positions are indicated for ObPFOMT-1. Substrates are indicated on the right. Product of reaction in
trace 1 was identified based on MS/MS spectrum comparison with authentic standards of chrysoeriol (C) and diosmetin (D). Based on fragmentation patterns (E, F) and
comparisons with ObF8OMT (trace 10), product 1 (P1) in trace 7 tentatively identified as 30-methyl ether (E), product 2 (P2) as 8-methyl ether (F) of 7,8,30,40-OH-flavone. C-F.
MS/MS fragmentation patterns used for product identification. Fragmentation of parent ions was achieved using negative ionization mode and 35% normalized collision
energy as described in Section 5. Identity of fragment peaks is shown in C and is the same for C-F.
et al., 2006) indicates that the third acceptor moiety is the 40-OH
group (in addition to 30- and 50-OH). Additionally, a shift in the UV
spectrum of the monomethylated product peak along the retention
time and with progressing turnover suggests that it contains two
products that are not chromatographically resolved and are methyl-
ated at 30- or 40-OH. This behavior is reminiscent of the Mg2+-depen-
dent OMT from the cyanobacterium Synechocystis (Q55813Syn,
Fig. 1), which methylates both meta and para hydroxyl residues of
di- and triphenols (Kopycki et al., 2008). Remarkably, it exhibited
such relaxed regioselectivity only with cinnamic acid derivatives,
while with tricetin as substrate it was specific for the 40-OH group.
In contrast, PFOMTs from vanilla (Widiez et al., 2011) and rice (Lee
et al., 2008) were reported to be regiospecific for 30- and 50-OH
groups of this substrate. ObPFOMT-1 gives rise to two products with
6,7-OH-flavone (33), with 6-OH acting as the major acceptor group.
Assays with 5,6-OH-flavone (34) and 7,8-OH-flavone (32) yielded 5-
OH and 8-OH-methylated products, respectively (Fig. 2, Supplemen-
tal Fig. S4). Both 30- and 8-OH-residues of 7,8,30,40-OH-flavone (30)
were methylated, however, the dominant acceptor appeared to be
the hydroxyl residue at position 8, while the respective 30-mono-
methyl ether was barely detectable (Fig. 2 trace 7). Therefore, at least
for a substrate with this substitution pattern, 8-O-methylation is a
more favorable reaction than the 30-O-methylation.
flavone 7-O-methylationisprerequisite for the subsequent6-hydrox-
ylation (Berim and Gang, 2013), therefore, scutellarein-40-methyl
ether does not serve as an intermediate and is unlikely to be present
in basil in appreciable amounts at any time. The former three com-
pounds are unnatural, synthetic flavones. However, the 7,8-dihy-
doxyl substitution pattern of the ring A in substrates 31 and 32 is
reminiscent of pilosin (4), which accumulates in some basil varieties
(Grayeretal., 2001, Section2.4 of thiswork),butisnot currentlyavail-
able to us. The relative activities with basil flavones scutellarein-7-
methyl ether (9) and ladanein (7) were under 25% of that found with
luteolin (20). Cirsiliol (15) was considered an auspicious substrate as
its 30-O-methylated derivative cirsilineol (14) occurs in basil (Grayer
et al., 1996). However, therelative turnoverwith15onlyamounted to
64% of that with luteolin (20). Notably, it was observed that higher
cation concentrations (100 lM vs. 10 l
M Mg2+) reduce the relative
turnover rates of 7-O-methylated substrates 7 and 9 by ca. 12% and
30%, respectively, but significantly increase that of caffeic acid (37)
(27% vs. 78%). As this effect might extend to other substrates, the re-
sults shown in Table 1 and Supplemental Fig. S5 should be viewed as
specific for the assay conditions used. The cytosolic concentrations of
free Mg2+ vary between different cell types and subcellular compart-
ments, usually not exceeding 2 mM in cytosol (Bose et al., 2011)
wherethe OMT reactions are likelyto occur, and have never been esti-
mated for secretory cells of peltate trichomes. It is therefore impossi-
ble to predict how this will influence the reaction rates in vivo.
Studies with the PFOMT from ice plant demonstrated that its N-
terminus affected the protein’s regioselectivity (Vogt, 2004). An
analogously N-truncated (N-10) ObPFOMT-1 variant was produced
(Supplemental Fig. S2) and its activities tested with relevant sub-
strates (7, 9, 11, and 30), but no differences from the full-length
protein were detected (not shown).
2.2.3. Kinetic properties
Kinetic analyses showed that 7,8,40-OH-flavone (31), which is
methylated at position 8 by ObPFOMT-1, is converted with higher
catalytic efficiency and is therefore a better substrate than luteolin
(20) and cirsioliol (15) (Table 2). The apparent affinity constants in
low micromolar range are typical and physiological for many of the
reported O-methyltransferases (Fellenberg et al., 2008; Ibdah et al.,
2003; Lucker et al., 2010). However, all of the enzymes involved in
flavone biosynthesis in basil that have been studied to date dis-
played much higher affinity for their proposed physiological
2.2.2. Relative activities with various substrates
Of the offered substrates, the highest relative reaction rates were
measured with 7,8,40-OH-flavone (31), 7,30,40-OH-flavone (29), 7,8-
OH-flavone (32), and scutellarein-40-methyl ether (10) (Table 1,
graphical overview in Supplemental Fig. S5). All of them can be ex-
cluded as natural substrates. Our previous work has established that
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
5
Table 2
Kinetic properties of ObPFOMT-1 and ObF8OMT-1 with selected substrates.
Enzyme
Substrate
Km
(
l
M)
Vmax (nkat mgꢀ1
)
kcat (sꢀ1) ꢁ 10ꢀ3
kcat/Km (sꢀ1mMꢀ1
)
ObPFOMT-1
Luteolin (20)a
3.98 0.36
8.62 0.50
3.31 0.42
12.19 1.42
4.69 0.12
7.82 0.50
6.99 0.39
1.78 0.22
1.61 0.03
1.22 0.06
4.30 0.21
1.94 0.08
0.39 0.07
0.25 0.01
0.11 0.01
0.41 0.05
45.39 0.93
34.30 1.63
121.1 5.87
54.77 2.28
16.71 3.04
10.75 0.29
4.61 0.37
11.38
3.98
36.61
4.49
3.57
1.38
0.66
9.49
Cirsiliol (15)a
7,8,40-OH-flavone (31)a
SAMb
ObF8OMT-1
7,8,40-OH-flavone (31)a,c
7-OCH3-8-OH-flavone (36)a,c
7,8-OH-flavone (32)a,c
SAMd
17.79 0.22
a
b
c
Reactions were performed with 50
Reactions were performed with 50
Results are means S.E. (n = 4), for all other series results are means S.E. (n = 3).
Reactions were performed with 100 l
M 7,8,40-OH-flavone as substrate.
l
l
M SAM as cosubstrate.
M luteolin as substrate.
d
substrates, as reflected by ca. 10- to 100-fold lower apparent Km
values (Berim and Gang, 2013; Berim et al., 2012). As stated above,
the only naturally occurring basil flavone that could act as a sub-
strate for 8-O-methylation by ObPFOMT-1, pilosin (4), is presently
not available from any source. Considering the fact that no
uncharacterized ESTs in our current database seem to correspond
to the flavonoid 30-OMT from peppermint (Willits et al., 2004),
ObPFOMT-1 remains the most promising candidate to fulfill this
function in basil. An analysis of other steps involved in the accu-
mulation of 30-substituted flavones (e.g. the activities of flavone
6- and 30-hydroxylases with relevant substrates) is necessary to
delineate this side branch of flavone network in basil trichomes.
the product of 8-hydroxysalvigenin 8-O-methylation (Fig. 3). Plot-
ting the data according to Dixon (1953) indicated competitive inhi-
bition as the underlying mechanism and a very low Ki value of ca.
10–15 nM (Fig. 3B). Nevadensin (2), the product of pilosin (4) 8-O-
methylation, which differs from gardenin B (1) by one methyl group
at position 7, does not exert a measurable inhibitory effect, nor does
genkwanin (16), a 7-O-methylated flavone occurring in basil. None
of these flavones inhibit ObPFOMT-1 (Fig. 3A). Overall, this suggests
that the inhibition of ObF8OMT-1 by gardenin B is highly specific.
Whether it becomes physiologically important depends on the con-
centration of gardenin B in the catalytically active compartment, the
secretory cells of the trichomes, as well as on the affinity of Ob-
F8OMT-1 for its natural substrate(s) and their abundance. Our pre-
vious results implied that lipophilic flavones including gardenin B
2.3. Properties of the cation-independent putative flavonoid 8-OMT
from basil
The above data indicate that ObPFOMT-1 could be involved in
methylation of pilosin (4) to yield nevadensin (2). However, based
on our prior results (Berim and Gang, 2013; Berim et al., 2012),
gardenin B (1) originates from the methylation of 8-hydroxysalvi-
genin (3), whereas the 7-O-methylation of nevadensin (2) is a very
slow and probably negligible process. As ObPFOMT-1 does not cat-
alyze the methylation of 8-OH-7-OCH3-flavone (36) and in general
requires vicinal hydroxyl residues for activity, it is not expected to
be involved in this step.
A
Prior to the analysis of the cation-dependent OMTs it seemed
likely that the 8-O-methylation is carried out by a class 1 OMT, a
homolog of F8OMT from peppermint (Willits et al., 2004). The basil
EST database contains four contigs sharing ca. 67% identity with that
enzyme. This identity level is similar to those found for flavone 7-
and 40-OMTs in the two species (Berim et al., 2012). The four contigs
are >90% identical, yet the activities of the encoded proteins may dif-
fer. The most abundantly expressed putative 8-OMT (termed Ob-
F8OMT-1) was chosen for heterologous expression and
preliminary biochemical characterization (for some basic proper-
ties, see Supplemental Table S1). ObF8OMT-1 was active with all
four available 8-substituted flavones (Fig. 2, Table 1), and displayed
themostfavorablekineticswith7,8,4-hydroxyflavone(31)(Table2).
The relative reaction rates with the offered substrates differed from
those determined for the peppermintF8OMT, whichshowed highest
turnover with 8-OH-7-OCH3-flavone (36) (Willits et al., 2004). As
found for ObPFOMT-1, the apparent affinity constants were in the
low micromolar range, yet far higher than those determined for
other enzymes involved in flavone metabolism in basil (Berim and
Gang, 2013; Berim et al., 2012). Kinetic properties suggest that the
presence of a 7-O-methyl moiety improves the catalytic efficiency,
possibly indicating that 8-hydroxysalvigenin (3) would be a better
substrate than pilosin (4) (Table 2). However, only experimental
data with 3 and 4 will allow us to make definitive conclusions
regarding the preferred natural substrate of ObF8OMT-1.
Importantly, ObF8OMT-1 is strongly inhibited by gardenin B (1),
B
Fig. 3. Inhibition of ObPFOMT-1 and ObF8OMT by selected basil flavones. A.
Turnover measured in reactions supplied with 2.5
of ObPFOMT-1 with luteolin (50 M) and of ObF8OMT-1 with 8-OH-7-OCH3-flavone
(100 M) without inhibitor addition was set as 100%. Results are means S.D.
lM of tested inhibitor. Turnover
l
l
(n = 3). Asterisk indicates data point statistically different from control according to
Student’s t-test (P = 0.0007). All other means were not statistically different from
control (P > 0.05). B. Dixon plot of ObF8OMT-1 inhibition by gardenin B. Concen-
trations of 8-OH-7-OCH3-flavone (substrate) are indicated in the legend. Intercept
served for the estimation of a Ki value of ca. 10 nM. Assays were performed in
triplicates, means are shown.
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

6
A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
(1) are enriched in the subcuticular cavities of mature trichomes
(Berim et al., 2012), but there are presently no data regarding the
time course of their translocation or the reversibility of this process.
A
2.4. Expression of ObPFOMT-1 and ObF8OMT-1 and accumulation of
relevant flavones in four basil cultivars
The EST counts for ObPFOMT-1 and ObF8OMT-1 suggest that
these genes are expressed at different levels in the four basil lines
routinely used in our lab, for example, no ESTs coding for Ob-
F8OMT-1 originate from the line MC. In order to verify this, the
transcript levels were assessed using qRT-PCR. In a preliminary
experiment, the mRNA levels in isolated trichomes were estimated.
The trends were in agreement with the EST counts, confirming
high expression of ObPFOMT-1 in all lines and significantly higher
expression of ObF8OMT-1 in the lines EMX-1 and SD as compared
to the lines SW and MC (Fig. 4). The expression of ObPFOMT-1
and ObF8OMT was then correlated with the accumulation of rele-
vant flavones in the youngest leaf pair on a stem with seven leaf
pairs. The expression profiles in whole leaves were very similar
to those found in isolated trichomes (Fig. 4). The ObPFOMT-1 tran-
scripts were most abundant in the line EMX-1, whereas ObF8OMT-
1 was barely detectable in the lines MC and SW. For the metabolite
analysis, liquid chromatography conditions were modified to sep-
arate the isobaric flavones nevadensin (2) and 8-hydroxysalvigenin
(3), which were previously quantified as one compound. Flavone
profiles were similar in lines SD and EMX-1 and differed from those
in the lines SW and MC (Fig. 5). Whereas in the former two lines
nevadensin is the major flavone with m/z 345, in the latter cultivars
it is 8-hydroxysalvigenin. Nevadensin is a minor metabolite in
these lines (SW and MC). Pilosin (4) accumulates in amounts com-
parable to those of gardenin B (1) in line SD, but not in the other
B
C
A
Fig. 5. Accumulation of selected flavones in young leaves of four basil lines. A.
Amounts of 8-hydroxysalvigenin (3) (8-OH-SALV), nevadensin (2) (NEV), salvigenin
(5) (SALV), gardenin B (1) (GARD B), and pilosin (4) (PIL) in the 7th leaf pair of four
basil lines. Results are means s.e. of 5 biological replicates. B. and C. Represen-
tative chromatograms (shown at m/z 345.0969 10 ppm) of UPLC separation of
isobaric flavones 8-OH-salvigenin (3) and nevadensin (2) as described in Section 5
and an extract from the line EMX-1 (B) and SW (C) as analyte. Insets show
respective UV-spectra.
B
three lines. Because of the large differences in absolute accumula-
tion of flavones in the different basil lines, it is important to con-
sider the relative proportions of the individual flavones, setting
the total amount of monitored compounds as 100%. The proportion
of salvigenin (5) is 34–46% in all lines. The relative abundances of
gardenin B range between 10–14% in all four lines. Nevadensin
takes up 35–40% of the total in lines SD and EMX-1, whereas 8-
hydroxysalvigenin accounts for about 50% of monitored flavones
in the lines SW and MC.
As ObPFOMT-1 may act as flavone 30-OMT, the abundance of po-
tential product, cirsilineol (14) was also monitored in the different
basil lines. As it proved difficult to separate it from its isomer,
eupatorin (15) (Grayer et al., 1996, 2001; our work), both were
co-quantified. Their amounts were below reliable quantification le-
vel in line MC, while lines EMX-1, SD, and SW accumulated 5.71,
Fig. 4. Expression of ObPFOMT-1 (A) and ObF8OMT-1 (B) in whole leaves and
isolated trichomes of four basil lines. Transcript levels of ObPFOMT-1 and
ObF8OMT-1 in the 7th leaf pair and in isolated trichomes of four basil lines were
assessed using qRT-PCR. Legend indicates which tissue was used. Expression in the
line SW was set as 1 for both genes and each tissue. Results are means S.E. of five
biological replicates for whole leaves, and means S.E. of three technical replicates
for isolated trichomes. Values are indicated on the graph where not easily legible.
Means of data points marked with the same letter are not statistically different
according to one-way ANOVA and Tukey’s HSD post hoc tests (P < 0.05).
15.66, and 10.47 lg per g fresh weight, respectively. Because of
their very different ionization efficiencies in negative ionization
mode in our LCQ Advantage mass spectrometer, it is permissible
to estimate that cirsilineol (14) accounts for nearly 100% of the
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
7
Fig. 6. Potential steps around flavone 8-O-methylation in sweet basil. Solid arrows indicate conversions supported by evidence (including this work), dashed arrows indicate
there is no current data as to whether or how such a step occurs. F8H, flavone 8-hydroxylase; 7-DM, flavone 7-O-demethylase.
detected joint amount in the line EMX-1, about 50% in the line SD,
and 5% in the line SW.
relative abundances of gardenin B (1) are comparable to those
found in the lines MC and SW. On the one hand, this could be
due to further conversion of gardenin B into nevadensin (2). On
the other hand, this flavone profile could result from the strong
inhibition of ObF8OMT-1 by gardenin B (1), coupled with highly ac-
tive 7-O-demethylation of 8-hydroxysalvigenin (3) to yield pilosin
(4). In line EMX-1, pilosin (4) itself would be further converted into
nevadensin (2) by the more abundantly expressed ObPFOMT-1,
whereas in line SD it would be in part stored. Optionally, the accu-
mulation of pilosin (4) in line SD may depend on a distinct 8-
hydroxysalvigenin-specific 7-O-demethylase whose expression is
low in line EMX-1. In absence of other data, all of the above inter-
pretations appear equally legitimate.
Despite the limitations posed the by lack of natural substrates,
this work provides further valuable insights into a presently only
poorly understood segment of flavone metabolism in basil by iden-
tifying an unanticipated candidate 8-OMT and highlighting the
ambiguity regarding the major direct precursor of nevadensin. Cur-
rently, direct genetic means (e.g. transgenic techniques) that could
provide final clarity for the primary operating pathway as well as
for the respective contributions of these two OMTs to the observed
flavone profiles in basil are still under development. Obviously, a
better characterization of ObPFOMT-1 and ObF8OMT-1 with the
proposed natural 8-OMT substrates (3 and 4) would be indispens-
able to estimate the efficiency and thus the likelihood of their
involvement in the respective conversions. As ObPFOMT-1 displays
relaxed substrate specificity with simple o-diphenols, its high 8-O-
methylating activity might be an in vitro artifact. Indeed, two cat-
ion-dependent OMTs from rice were also active with 7,8-OH-fla-
vone (32), and position 8 was suggested to be the methylation
site (Lee et al., 2008). It can also be expected that the characteriza-
tion of 7-O-demethylase(s) will shed more light on the actual
routes and processes underlying the formation of nevadensin in
basil. We will use the presumed low expression of the underlying
dioxygenase(s) in SW and MC as compared to SD and EMX-1 as a
criterion to help narrow down our choices of candidate genes.
Alternatively, identification of a basil cultivar/variety displaying
low expression of the ObPFOMT-1 gene but high 7-O-demethylase
and ObF8OMT-1 activity might offer insights into the respective
gene’s functions in the plant.
3. Discussion
Our recent investigation indicated that 7-O-methylation is a
prerequisite for flavone 6-hydroxylation, and that the occurrence
of 7-unmethylated flavones nevadensin (2) and pilosin (4) is due
to oxoglutarate-dependent 7-O-demethylation in basil trichomes
(Berim and Gang, 2013). Integration of the biochemical findings
presented above into the emerging scheme of this biosynthetic
network implies that there are two potential routes yielding neva-
densin (2), one of the major flavones in basil cultivars SD and EMX-
1 (Fig. 6). One of them proceeds via gardenin B (1), and involves its
7-O-demethylation as the last step. Demethylation of gardenin B
(1) has been demonstrated in crude trichome protein extracts (Ber-
im and Gang, 2013). The other route positions pilosin (4) as the
immediate precursor of nevadensin. ObF8OMT-1 would be in-
volved in both cases, while ObPFOMT-1 only plays a role in the
accumulation of nevadensin if the second scenario is the major
one. Based on metabolic analysis indicating that pectolinarigenin
(8) does not accumulate in basil leaves (Berim et al., 2012) and
on preliminary data that 7-O-demethylation of salvigenin (5) does
not readily occur in trichome protein extracts under the conditions
used to detect the demethylation of gardenin B (1) (Berim and
Gang, unpublished), the formation of pilosin (4) itself is predicted
to occur via the 7-O-demethylation of 8-hydroxysalvigenin (3).
So far, the lack of the proposed substrate prevents us from investi-
gating this hypothesis further.
Analysis of relevant metabolites and transcript levels of both
candidate 8-OMTs in the four basil lines does not resolve this
ambiguity. The high proportion of 8-hydroxylsalvigenin (3) rela-
tive to gardenin B (1), pilosin (4) and nevadensin (2) in the culti-
vars SW and MC is in line with the notion that only a small
portion of it is converted further, matching well with the low
expression of ObF8OMT-1, and suggesting that 7-O-demethylation
is a minor process in these two lines. The fact that lines SW and
MC accumulate gardenin B (1) despite the low expression of Ob-
F8OMT-1 can indicate that even this low expression level suffices
to support measurable conversion. It can also be in part due to
the presence of divergent minor contigs not amplified by the used
qPCR primers. Overall, the seemingly absent 7-O-demethylase
activity and therefore non-occurrence of the proposed substrate
pilosin (4) make it impossible to draw conclusions regarding the
actual reaction order and the role of ObPFOMT-1 in 8-O-methyla-
tion from these two basil lines. In the cultivars SD and EMX-1,
the relative abundance of 8-hydroxylsalvigenin (3) is low. Ob-
F8OMT-1 is well expressed in these two basil lines; however, the
4. Concluding remarks
The above work presents another step towards the complete
biochemical elucidation of the flavone metabolic network in basil.
This study suggests that the roles of cation-dependent OMTs in ba-
sil might not be limited to 30-O-methylation, and draws our atten-
tion to the conundrum surrounding the two potential routes to
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

8
A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
nevadensin (2). Our next objective is to procure the currently
unavailable substrates, pilosin (4) and 8-hydroxysalvigenin (3),
which should be used to study the activities and kinetic properties
of the described methyltransferases, as well as the 7-O-demethyl-
ase(s). The elucidation of the latter step, the oxoglutarate-depen-
dent demethylation of 8-hydroxysalvigenin (3) and/or gardenin B
(1), may be of critical importance for an accurate assessment of
the true reaction order leading to the distinct profiles of 8-substi-
tuted flavones in basil.
data-dependent mode. Separation of nevadensin (2) and 8-
hydroxysalvigenin (3) using a previously described UPLC system
(Berim et al., 2012) was achieved using an Acquity HSS T3 UPLCÒ
column (100 ꢁ 2.1 mm, 1.7
the same solvents as other separations, with a flow rate of
400
L minꢀ1, and ran as follows: 0 min, 5% B; 0.86 min, 5% B;
lm, Waters). The linear gradient used
l
6.00 min, 40% B; 12.00 min, 48% B; 14.00 min, 100% B; 16.00 min,
100% B; 17.00 min, 5% B; 20.00 min, 5% B. Quantification was based
on standard curves produced using serial dilutions of authentic
compounds eupatorin (13), nevadensin (2) and gardenin B (1). Un-
der the previously made assumption of ca. equal UV335 extinction
coefficients (Grayer et al., 1996), the standard curve for nevadensin
was used to estimate the amounts of 8-hydroxysalvigenin (3) and
pilosin (4).
5. Experimental
5.1. General
Unless noted otherwise, chemicals were of analytical quality
and purchased from common vendors (Fisher Scientific, VWR
International, Sigma Aldrich). Molecular biology reagents were
from Invitrogen and New England Biolabs. Rosetta™ 2 (DE3)pLysS
cells (Novagen) were used for heterologous gene expression.
5.5. RNA extraction
RNA for cloning purposes and for quantification of ObPFOMT-1
and ObF8OMT-1 transcripts was isolated and its quality checked
as described previously (Berim et al., 2012).
S-adenosyl-[14C-methyl]-
L
-methionine
(48.8 mCi/mmol)
was
obtained from Perkin Elmer. Unlabeled SAM, caffeic acid (38),
3,4-OH-benzaldenyde (39), and catechol (40), were from Sigma
Aldrich. All synthetic flavones were purchased from Indofine.
Quercetagetin (23) and tricetin (26) were from Extrasynthese.
The sources of all other naturally occurring flavonoids and authen-
tic basil flavones were the same as previously described (Berim
et al., 2012). Authentic samples of gardenin B (1), nevadensin (2),
and eupatorin (13) were kindly donated by Dr. R. Grayer (Kew
RBG, London, UK).
5.6. Cloning and heterologous expression of basil OMTs
Primers shown in Supplemental Table S2 were used to obtain
the full length clones which were subcloned into pCR2.1Ò TOPOÒ
vector (Invitrogen) followed by transfer to pET15b (Novagen) for
expression. Heterologous expression and protein purification were
conducted as reported earlier (Berim et al., 2012).
5.7. Biochemical characterization
5.2. Plant material growth and culture
For quantitative analyses, reaction conditions were chosen so
that the reaction velocity was directly proportional to incubation
time and protein amount. For kinetics measurements, saturating
concentrations of the second substrate (flavone or SAM) were pro-
vided. Affinities for SAM were determined using luteolin (ObP-
FOMT-1) or 7,8,40-OH-flavone (ObF8OMT-1) as substrate.
Energy of activation and thermal stability were determined
using 50 mM K-Pi pH 8.5 (ObPFOMT-1) or 7.5 (ObF8OMT-1) buffer
and 10 mM MgCl2 (ObPFOMT-1). Dependency of reaction rates on
pH was determined using the same buffer systems as reported ear-
lier (Berim et al., 2012).
Basil seeds (O. basilicum lines EMX-1, SW, MC, and Sweet Dani
[SD]) were germinated in vermiculite and individually re-potted
into SunGro mix. Plants were grown in growth chambers under
16/8 photoperiod and 28 °C (24 h). Light intensity of ca.
300 l
mol mꢀ2 sꢀ1 was supplied by incandescent and fluorescent
lamps.
5.3. Metabolite extracts
Flavones were extracted using the protocol reported earlier
(Berim et al., 2012). Whole small leaves (1–1.5 cm length, 7th leaf
pair from a stem with seven leaf pairs) were used for extraction.
Five biological replicates were collected for each basil line.
Radioactive assays were carried out in a total volume of 100
incubated at 30 °C for an appropriate length of time (including
5 min pre-conditioning), quenched with 5 L 6 N HCl, and ex-
tracted with 200 L EtOAc. A standard assay for ObPFOMT-1 con-
tained 100 mM Tris/HCl pH 8.5, 10 M MgCl2, 50 M substrate
(luteolin), 0.4 g purified protein, and was started with 50
SAM. Assay series for substrate comparison with ObPFOMT-1 were
supplied with 100 M MgCl2. A standard assay for ObF8OMT-1
contained 100 mM Tris/HCl pH 7.5, 100
M substrate (7,8,40-OH-
flavone), 0.6 g purified protein, and was started as above. After
lL,
l
l
5.4. Metabolite and enzyme assay analyses
l
l
l
lM
The same instrumentation (ion trap system: LCQ Advantage
system with Surveyor HPLC and photodiode array detector (Ther-
mo), and Q-TOF-MS system: Synapt G2 quadrupole-ion mobility
spectrometry-time of flight mass spectrometer system (Waters)
equipped with an Acquity UPLC system with photodiode array
detector) as described previously (Berim et al., 2012) was used
for metabolite and enzyme assay product analysis. The MS settings
for positive mode electrospray ionization and the Q-TOF-MS source
conditions were also as reported previously. Negative mode ioniza-
tion (using the above mentioned LCQ Advantage system) was ap-
plied to distinguish between chrysoeriol (19) and diosmetin (18),
as well as eupatorin (13) and cirsilineol (14). In this case, tuning
was carried out with apigenin. The optimized MS settings were
as follows: capillary at 275 °C and ꢀ21 V, source at 5.0 kV and
l
l
l
centrifugation for 3 min at 21,000g, an aliquot of the upper phase
was added to Eco-Scint scintillation cocktail (5 mL) (National Diag-
nostics) and analyzed using RackBeta 1215 scintillation counter
(LKB). Relative reaction rates with a range of flavones were mea-
sured with 50
and 50 SAM. Assays with unlabeled SAM (supplied at
200 M) for product identification were carried out in a total vol-
ume of 100 L under appropriate conditions, stopped with 6 N
HCl (10 L).
lM (ObPFOMT-1) or 100 lM (ObF8OMT) substrate
lM
l
l
l
L) and the aq. phase extracted with EtOAc (2 ꢁ 200
l
After evaporating the combined organic phases, the dry residue
spray current at 30 lA, sheath/sweep gas flow at 33/20 (arbitrary
was dissolved in MeOH–LC buffer (5 mM ammonium formate/
units). For collision-induced dissociation spectra, fragmentation
0.1% HCO2H) ((50
lL) 1:1, v/v) and analyzed using the LCQ as de-
was achieved using 35% normalized collision energy and
scribed above and earlier (Berim et al., 2012).
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001

A. Berim, D.R. Gang / Phytochemistry xxx (2013) xxx–xxx
9
De Luca, V., Ibrahim, R.K., 1985. Enzymatic syntesis of polymethylated flavonols in
5.8. Phylogenetic analysis of sweet basil OMTs
Chrysosplenium americanum. I. Partial purification and some properties of S-
adenosyl-L-methionine:flavonol 3, 6, 7, and 4’-O-methyltransferases. Arch.
Biochem. Biophys. 238, 596–605.
Selected (putative) PFOMT protein sequences were retrieved
from NCBI GenBank™ and from basil EST database. Duplicates
(contigs encoding identical peptides) and peptides shorter than
1/3 of the expected OMT length were removed. Resulting se-
quences were aligned using Clustal W2 at EBI with default settings.
MEGA5 package (Tamura et al., 2011) was used for the phyloge-
netic analysis and tree construction. Evolutionary relationships
were inferred using maximum likelihood algorithm and Jones-Tay-
lor-Thornton matrix. All sites, independently of coverage, were
used for analysis. The tree was rooted using fkbG OMT from Strep-
tomyces hygroscopicus (AAF86386Sh) as outgroup.
Dixon, M., 1953. The determination of enzyme inhibitor constants. Biochem. J. 55,
170–171.
Fellenberg, C., Milkowski, C., Hause, B., Lange, P.-R., Boettcher, C., Schmidt, J., Vogt,
T., 2008. Tapetum-specific location of a cation-dependent O-methyltransferase
in Arabidopsis thaliana. Plant J. 56, 132–145.
Gang, D.R., Wang, J.H., Dudareva, N., Nam, K.H., Simon, J.E., Lewinsohn, E., Pichersky,
E., 2001. An investigation of the storage and biosynthesis of phenylpropenes in
sweet basil. Plant Physiol. 125, 539–555.
Grayer, R.J., Bryan, S.E., Veitch, N.C., Goldstone, F.J., Paton, A., Wollenweber, E., 1996.
External flavones in sweet basil, Ocimum basilicum, and related taxa.
Phytochemistry 43, 1041–1047.
Grayer, R.J., Veitch, N.C., Kite, G.C., Price, A.M., Kokubun, T., 2001. Distribution of 8-
oxygenated leaf-surface flavones in the genus Ocimum. Phytochemistry 56,
559–567.
Hugueney, P., Provenzano, S., Verries, C., Ferrandino, A., Meudec, E., Batelli, G.,
Merdinoglu, D., Cheynier, V., Schubert, A., Ageorges, A., 2009. A novel cation-
dependent O-methyltransferase involved in anthocyanin methylation in
grapevine. Plant Physiol. 150, 2057–2070.
Ibdah, M., Zhang, X.H., Schmidt, J., Vogt, T., 2003. A novel Mg(2+)-dependent O-
methyltransferase in the phenylpropanoid metabolism of Mesembryanthemum
crystallinum. J. Biol. Chem. 278, 43961–43972.
Jay, M., Deluca, V., Ibrahim, R.K., 1985. Purification, properties and kinetic
mechanism of flavonol 8-O-methyltransferase from Lotus corniculatus L. Eur. J.
Biochem. 153, 321–325.
Kim, B.-G., Kim, D.H., Sung, S.H., Kim, D.-E., Chong, Y., Ahn, J.-H., 2010. Two O-
methyltransferases from Picea abies: characterization and molecular basis of
different reactivity. Planta 232, 837–844.
5.9. Quantitative real-time RT-PCR
The relative quantification method (Schmittgen and Livak,
2008) was used to assess the abundances of transcripts. Protocol
optimization, general considerations and instrumentation were
as reported previously (Berim et al., 2012). Primers are listed in
Supplemental Table S2. Elongation factor 1 gene served as reference
for normalization. For scaling, the relative expression of both genes
in the line SW was set as 1 for each tissue. Five biological replicates
of each leaf pair, and three technical replicates of one trichome
preparation of each basil line were analyzed.
Kopycki, J.G., Stubbs, M.T., Brandt, W., Hagemann, M., Porzel, A., Schmidt, J.,
Schliemann, W., Zenk, M.H., Vogt, T., 2008. Functional and structural
characterization of
a
cation-dependent O-methyltransferase from the
cyanobacterium Synechocystis sp. strain PCC 6803. J. Biol. Chem. 283, 20888–
20896.
6. Accession numbers
Lee, Y.J., Kim, B.G., Chong, Y., Lim, Y., Ahn, J.-H., 2008. Cation dependent O-
methyltransferases from rice. Planta 227, 641–647.
Coding sequences for ObPFOMT-1 and ObF8OMT-1 were sub-
mitted to NCBI GenBank™ under accession numbers KC354401
and KC354402, respectively.
Lucker, J., Martens, S., Lund, S.T., 2010. Characterization of a Vitis vinifera cv.
Cabernet Sauvignon 30,50-O-methyltransferase showing strong preference for
anthocyanins and glycosylated flavonols. Phytochemistry 71, 1474–1484.
Lukacin, R., Matern, U., Specker, S., Vogt, T., 2004. Cations modulate the substrate
specificity of bifunctional class I O-methyltransferase from Ammi majus. FEBS
Lett. 577, 367–370.
Acknowledgements
Noel, J.P., Dixon, R.A., Pichersky, E., Zubieta, C., Ferrer, J.L., 2003. Structural,
functional, and evolutionary basis for methylation of plant small molecules.
Recent Adv. Phytochem. 37, 37–58.
Schmittgen, T.D., Livak, K.J., 2008. Analyzing real-time PCR data by the comparative
C-T method. Nat. Protoc. 3, 1101–1108.
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:
molecular evolutionary genetics analysis using maximum likelihood,
evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28,
2731–2739.
We thank Dr. R. Grayer (Kew RBG, London, UK) for the authentic
samples of nevadensin, eupatorin, and gardenin B. This work was
supported by the DOE Biological and Environmental Research Pro-
gram Grant DE-SC0001728 to D.R.G.
Appendix A. Supplementary data
Vogt, T., 2004. Regiospecificity and kinetic properties of a plant natural product O-
methyltransferase are determined by its N-terminal domain. FEBS Lett. 561,
159–162.
Weng, J.-K., Chapple, C., 2010. The origin and evolution of lignin biosynthesis. New
Phytol. 187, 273–285.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.phytochem.2013.
05.001.
Widiez, T., Hartman, T.G., Dudai, N., Yan, Q., Lawton, M., Havkin-Frenkel, D.,
Belanger, F.C., 2011. Functional characterization of two new members of the
caffeoyl CoA O-methyltransferase-like gene family from Vanilla planifolia
reveals a new class of plastid-localized O-methyltransferases. Plant Mol. Biol.
76, 475–488.
Willits, M.G., Giovanni, M., Prata, R.T.N., Kramer, C.M., De Luca, V., Steffens, J.C.,
Graser, G., 2004. Bio-fermentation of modified flavonoids: an example of in vivo
diversification of secondary metabolites. Phytochemistry 65, 31–41.
Zhou, J.-M., Gold, N.D., Martin, V.J.J., Wollenweber, E., Ibrahim, R.K., 2006. Sequential
O-methylation of tricetin by a single gene product in wheat. Biochim. Biophys.
Acta 1760, 1115–1124.
References
Berim, A., Gang, D.R., 2013. The roles of
a flavone 6-hydroxylase and 7-O-
demethylation in the flavone biosynthetic network of sweet basil. J. Biol.
Chem. 288, 1795–1805.
Berim, A., Hyatt, D.C., Gang, D.R., 2012. A set of regioselective O-methyltransferases
gives rise to the complex pattern of methoxylated flavones in sweet basil. Plant
Physiol. 160, 1052–1069.
Bose, J., Babourina, O., Rengel, Z., 2011. Role of magnesium in alleviation of
aluminium toxicity in plants. J. Exp. Bot. 62, 2251–2264.
De Luca, V., Ibrahim, R.K., 1982. Characterization of three distinct flavonol O-
methyltransferases from Chrysosplenium americanum. Phytochemistry 21,
1537–1540.
Please cite this article in press as: Berim, A., Gang, D.R. Characterization of two candidate flavone 8-O-methyltransferases suggests the existence of two
potential routes to nevadensin in sweet basil. Phytochemistry (2013), http://dx.doi.org/10.1016/j.phytochem.2013.05.001