Biosynthesis of estragole and methyl-eugenol in sweet basil ...

Plant Science 160 (2000) 27–35
Biosynthesis of estragole and methyl-eugenol in sweet basil
(Ocimum basilicum L). Developmental and chemotypic association
of allylphenol O-methyltransferase activities
Efraim Lewinsohn a, *, Iris Ziv-Raz a,b , Nativ Dudai a , Yaacov Tadmor a ,
Elena Lastochkin a , Olga Larkov a , David Chaimovitsh a , Uzi Ravid a , Eli Putievsky a ,
Eran Pichersky c , Yuval Shoham b
Abstract
a Newe Ya’ar Research Center, Agricultural Research Organization, P.O. Box 1021, Ramat Yishay 30095, Israel
b Department of Food Engineering and Biotechnology, Technion-Israel Institute of Technology, Haifa 32000, Israel
c Department of Biology, Uni�ersity of Michigan, Ann Arbor, MI 48109-1048, USA
Received 19 May 2000; received in revised form 28 July 2000; accepted 31 July 2000
Sweet basil (Ocimum basilicum L., Lamiaceae) is a common herb, used for culinary and medicinal purposes. The essential oils
of different sweet basil chemotypes contain various proportions of the allyl phenol derivatives estragole (methyl chavicol), eugenol,
and methyl eugenol, as well as the monoterpene alcohol linalool. To monitor the developmental regulation of estragole
biosynthesis in sweet basil, an enzymatic assay for S-adenosyl-L-methionine (SAM):chavicol O-methyltransferase activity was
developed. Young leaves display high levels of chavicol O-methyltransferase activity, but the activity was negligible in older leaves,
indicating that the O-methylation of chavicol primarily occurs early during leaf development. The O-methyltransferase activities
detected in different sweet basil genotypes differed in their substrate specificities towards the methyl acceptor substrate. In the
high-estragole-containing chemotype R3, the O-methyltransferase activity was highly specific for chavicol, while eugenol was
virtually not O-methylated. In contrast, chemotype 147/97, that contains equal levels of estragole and methyl eugenol, displayed
O-methyltransferase activities that accepted both chavicol and eugenol as substrates, generating estragole and methyl eugenol,
respectively. Chemotype SW that contains high levels of eugenol, but lacks both estragole and methyl eugenol, had apparently no
allylphenol dependent O-methyltransferase activities. These results indicate the presence of at least two types of allylphenol-specific
O-methyltransferase activities in sweet basil chemotypes, one highly specific for chavicol; and a different one that can accept
eugenol as a substrate. The relative availability and substrate specificities of these O-methyltransferase activities biochemically
rationalizes the variation in the composition of the essential oils of these chemotypes. © 2000 Elsevier Science Ireland Ltd. All
rights reserved.
Keywords: Ocimum basilicum L.; Lamiaceae; Sweet basil; Essential oils; Estragole; Methyl eugenol; O-methyltransferase
1. Introduction
The genus Ocimum includes about a dozen species
and subspecies, native to the tropical and
sub-tropical regions of the world. Sweet basil
(Ocimum basilicum L., Lamiaceae) is an important
* Corresponding author. Tel.: +972-4-9539552; fax: +972-4-
9836936.
E-mail address: twefraim@netvision.net.il (E. Lewinsohn).
0168-9452/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved.
PII: S0168-9452(00)00357-5
www.elsevier.com/locate/plantsci
medicinal plant and culinary herb [1]. Sweet basil
is widely cultivated for the production of essential
oils, and is also marketed as an herb, either fresh,
dried, or frozen [1]. The essential oil of sweet basil
possesses antifungal, insect-repelling and toxic activities
[2–4]. Despite the wide use and the importance
of sweet basil and its essential oils, little is
known about the biosynthesis and developmental
regulation of the compounds responsible for the
flavor quality of the fresh and dried herbs and that
constitute its essential oil.

28
Many members of the Lamiaceae such as Mentha,
Sal�ia, Origanum, and Thyme spp., are also
cultivated to be used as herbs and as a source of
essential oils. The essential oils of these species
and of many other species of the Lamiaceae are
mostly composed of mono- and sesquiterpenes [5].
Similarly, many species of the genus Ocimum,
contain essential oils based primarily on monoterpene
derivatives such as camphor, limonene, thymol,
citral, geraniol and linalool [5–8].
Interestingly, other members of the genus, including
sweet basil (O. basilicum L.), contain an essential
oil based primarily on high proportions of
phenolic derivatives, such as eugenol, methyl
chavicol (estragole) and methyl cinnamate, often
combined with various proportions of linalool, a
monoterpenol [5,6,9–11]. The allylphenolic derivative
eugenol (Fig. 1) has a sharp spicy odor reminiscent
of cloves, and is used as a dental analgesic
and disinfectant [12]. Eugenol is a major component
of the essential oil of cloves (Eugenia
caryophyllus=Syzygium aromaticum, Myrtaceae),
and cinnamon leaf (Cinnamomum zeylanicum,
Lauraceae) [12]. Methyl chavicol, also known as
estragole (Fig. 1), is used in the perfume industry
and has an odor resembling that of fennel and
anise. Interestingly, estragole is also a key component
of the essential oils of aromatic plants belonging
to different families, such as aniseed
(Pimpinella anisum L., Apiaceae), star anise (Illicium
�erum Hook. f., Magnoliaceae), bitter fennel
(Foeniculum �ulgare �ar �ulgare, Apiaceae), and
tarragon (Artemisia dracunculus L., Asteraceae)
[12–14].
Information on the biosynthetic pathways to
eugenol and methyl chavicol is limited. Pulse-chase
radiotracer experiments have indicated that both
allyl phenolic derivatives are formed from pheny-
Fig. 1. SAM:allylphenol O-methyltransferase activities from
O. basilicum. A methyl group from SAM is transferred to the
p-hydroxyl group of either chavicol or eugenol to generate
either methyl chavicol (estragole) or methyl eugenol, respectively,
and S-adenosyl homocysteine (SAH).
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35
lalanine and cinnamic acid in several plants including
sweet basil [15–18]. High levels of
phenylalanine ammonia-lyase activity, a key early
enzyme in this pathway, have been detected in
sweet basil leaves, and the enzyme has been
purified and characterized from many sources
[19–21], including basil [22]. We hypothesized that
the last biosynthetic step involved in the formation
of estragole, is catalyzed by an S-adenosyl-L-methionine
(SAM) dependent O-methyltransferase
activity, able to substitute chavicol in the para-hydroxy
position to give rise to estragole (Fig. 1).
Methyltransferases are ubiquitous enzymes that
catalyze the transfer of a methyl group from SAM
to an acceptor substrate, generating either O-, N-,
S- and C-methyl derivatives and S-adenosyl-homocysteine
(Fig. 1) [23,24]. Although methyltransferases
are involved in the formation of many
natural products, they normally display strict specificities
towards their acceptor substrate, and to
the position of the methylation of the substrate
[23,24]. O-methyltransferases are involved in the
methylation of caffeic acid to generate ferulic acid,
a central precursor of lignin and possibly eugenol.
Some of the genes that code for these enzymes
have been isolated from several plants [24–26],
and also from sweet basil [27]. A methyltransferase
activity that converts t-anol, (the 1-propenyl
analog of chavicol), to generate t-anethole has
been reported in cultures of Pimpinella anisum
(Apiaceae) [28]. Similarly, an enzyme activity that
catalyzes the para-O-methylation of t-isoeugenol,
(the 1-propenyl analog of eugenol), forming
methyl isoeugenol has been detected and purified
from Clarkia breweri flowers, (Onagraceae)
[29,30]. The corresponding gene has been isolated
and expressed in Escherichia coli. Interestingly, the
enzyme from C. breweri also accepts eugenol as
substrate, at slightly lower rates, to generate
methyl eugenol [29].
Most of the common cultivated members of the
Lamiaceae accumulate essential oils rich in terpenoid
compounds [5]. Although many studies
have been conducted to monitor the developmental
regulation of the biosynthesis of these terpenoid
essential oil components [31–34],
non-terpenoid components have received much
less attention. One of the major components of the
essential oil of sweet basil is estragole, a non-terpenoid
allylphenol derivative. It was thus of interest
to assess the possible developmental regulation

of estragole biosynthesis in sweet basil. We have
previously shown that during the development of
a sweet basil leaf, there is a net accumulation of
essential oil and of estragole, both per leaf, and on
a leaf weight basis [11]. Herein we provide evidence
that in sweet basil leaves, chavicol and
eugenol can be enzymatically O-methylated, to
generate methyl chavicol and methyl eugenol, respectively,
by SAM dependent O-methyltransferase
activities (Fig. 1). These activities are active
primarily in young leaf tissues and their substrate
specificity varies chemotypically, partially explaining
the differences in the chemical compositions
observed among varieties.
2. Material and methods
2.1. Plant material
Seeds of the different chemotypes [10] were
derived from carefully selfed O. basilicum L. plants
grown at the Newe Ya’ar Research Center, Israel.
Five seeds were germinated in 4 cm-deep cells in
trays containing equal volumes of
peat:vermiculite:perlite, and watered and fertilized
daily with 0.2% (w/v) NPK (5:3:8). After emergence,
only one seedling per cell was left and
grown until the shoot tips were discernible (about
10 days after germination) and used for the experiments.
To generate young plants, the seedlings
were transferred to 1 l pots containing a mixture
of roughly equal volumes of volcanic tuff, vermiculite
and peat. The plants were watered daily and
fertilized twice a week with 0.2% (w/v) NPK
(5:3:8) until four leaf pairs were discernible. Mature
plants were obtaining by transferring the
young seedlings directly from the trays into 1×
1.2×0.24 m containers containing volcanic tuff
(0.8 mm). The plants (24 plants per m 2 ) were
dip-watered daily and fertilized as described in
[35]. In all cases, plants were grown in greenhouses
and maximal temperatures did not exceed 35–
40°C.
2.2. Chemicals and radiochemicals
Chavicol was a kind gift of International Flavors
and Fragrances Inc., Bridgewater, NJ, USA.
Eugenol was from Roth Chemical Co. Estragole
was from Terpene Products, Jacksonville, FL,
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35 29
USA, and methyl eugenol was from Fluka. Sadenosyl-L-methionine
was from Sigma Chemical
Co. S-adenosyl-L-methyl 3 H-methionine (s.a. 15
Ci/mmol) and S-adenosyl-L-methyl 14 C-methionine
(s.a. 55 mCi/mmol) were from Amersham.
2.3. Enzyme extraction
Cell-free extracts were prepared as follows, fresh
plant leaves (1–2 g) were weighed; frozen in liquid
nitrogen; and placed in a chilled mortar. The
leaves were then ground with a pestle in the presence
of liquid nitrogen, �0.6 g sand and �0.1 g
polyvinylpolypyrrolidone (PVPP) to adsorb phenolic
materials. Ice cold extraction buffer (100
mM MOPS [pH 6.5], 10% glycerol [v/v], 6.6 mM
dithiothreitol, 1% [w/v] polyvinylpyrrolidone
[PVP-10, molecular weight 10 kDa], 5 mM
Na 2S 2O 5) was added (five volumes to fresh weight
tissue), and the fine powder was further extracted
for �30 s. The slurry was centrifuged twice at
20 000×g, for 10 min at 4°C, and the supernatant,
(crude extract) was used in the enzymatic
assays. The crude extracts (1 ml) were further
desalted using Sephadex G-25 columns (1×3 cm)
and eluted with 1.2 ml of buffer A (50 mM
HEPES [pH 7.5], 5% [v/v] glycerol, 3.3 mM
dithiothreitol).
2.4. Enzyme assays
Assays were preformed by mixing, 40 �l ofthe
crude or Sephadex-G25 purified extract (about 20
�g protein); 40 �l buffer A; 5 �M chavicol or other
acceptor substrate (from a 0.1 mM solution in
0.5% ethanol), 25 �M S-[methyl- 3 H]-adenosyl-Lmethionine
(6 Ci/mol in a total volume of 100 �l).
The assays were incubated at 21°C for 1.5 h, and
then 1 ml hexane was added to each tube, vigorously
vortexed and spun for 13 s at 20 000×g to
separate the phases. The upper hexane layers containing
the radioactive labeled enzyme products,
were transferred to scintillation tubes containing 3
ml scintillation fluid [4 g/l 2,5-diphenyloxazol
(PPO) and 0.05 g/l 2,2�-p-phenylen-bis (5-phenyloxazol)
(POPOP) in toluene]. The radioactivity
was quantified using a Packard Tri-carb liquid
scintillation counter. To confirm the identity of the
biosynthetic products, similar incubations were
performed, except that 14 C-labeled SAM (at the
same specific activity) was used. In this case the

30
Table 1
Developmental dependence of SAM:chavicol O-methyltransferase
activity in sweet basil chemotype R3 a
Plant analyzed Tissue Chavicol
analyzed O-methyltransferase
activity (pkat/g FW)
Young seedlings
(first leaf-pair
visible)
Shoot tips 1465�359
Cotyledons 98�31
Shoots 155�19
Roots 7�16
Young plant (four
leaf-pairs)
Shoot tips 8432�902
First pair of
leaves
3923�887
Second pair of
leaves
738�113
Third pair of
leaves
71�56
Mature plant Developing 1212�220
(blooming stage)
leaves
Fully
developed
leaves
63�18
Shoots
15�12
Inflorescences 135�72
Petals 43�23
a Enzyme activity was determined from Sephadex G-25
purified cell-free extracts derived from fresh tissues. Means
and S.E. of three independent determinations are shown.
hexane layer was evaporated to a volume of 20 �l
using a gentle stream of N 2, and analyzed by thin
layer chromatography (TLC) autoradiography using
Silica gel 60 F 254 plates developed with chloroform.
Spots were visualized by ultra voilet (UV)
light and radioactive spots detected by autoradiography
on Kodak X-OMAT paper. The identity of
the compounds was verified by comparison of
their R f with those of authentic standards (0.65 for
estragole, 0.57 for methyl eugenol). Protein was
assayed by the method of Bradford [36] using the
Bio-Rad protein reagent (Bio-Rad), and bovine
serum albumin (Sigma) as standard.
3. Results and discussion
3.1. O-methylation of allylphenols by O. basilicum
cell-free-extracts
Incubation of sweet basil cell-free extracts with
14 C- or 3 H-labeled SAM and chavicol, resulted in
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35
the accumulation of a hexane-soluble radiolabeled
products, absent when the cell-free extracts were
previously boiled for 5 min. This suggested that a
methyltransferase enzymatic activity is involved in
the conversions. Omission of chavicol resulted in a
lower but still significant rate (�60%) of formation
of hexane-soluble radioactive products, possibly
due to endogenous substrates. To remove any
endogenous substrates and other low molecular
weight compounds that interfered with the assays,
the cell-free extracts were partially purified by
Sephadex G-25 chromatography. Complete assays,
containing SAM and chavicol, but utilizing Sephadex
G-25 partially-purified-enzyme displayed
30–50% higher enzymatic activity levels as compared
with crude extracts, probably due to the
removal of inhibitors. The omission of chavicol
from the assays resulted in virtually no formation
of hexane-soluble radioactive products during the
enzymatic assays. To confirm the identity of the
radiolabeled compounds formed, the hexane extracts
were evaporated and analyzed by TLC-autoradiography.
Only one radioactive substance
originating in chavicol and 14 C-SAM fed Sephadex
G-25 extracts was detected and its R f coincided
with that of authentic estragole (not shown).
These results indicate that a methytransferase activity,
present in sweet basil leaves is able to
O-methylate chavicol and release estragole (Fig.
1).
3.2. De�elopmental regulation of the
SAM:cha�icol O-methyltransferase acti�ity
Different plant tissues were analyzed to identify
the sites potentially active in estragole biosynthesis.
Five-day-old seedlings already contained substantial
levels of chavicol-dependent
O-methyltransferase activity (Table 1). At this
stage of development the cotyledons are still turgid
and active, and the shoot tips contain discernible
leaf primordia of only a few mm length. Practically
all activity found resided in the leaf primordia.
The cotyledons and shoots displayed lower
levels of activity (about 8–10% by fresh weight) to
that displayed by leaf primordia (Table 1). The
roots were practically devoid of activity. Essentially,
a similar pattern was obtained when tissues
derived from seedlings at a four leaf-pair stage
(about 10 cm height) were examined. The shoot
tips had the highest activity levels, and young

leaves (0.2–0.7 mm length) had substantial (�
45% as compared with shoot tips on a fresh weight
basis) estragole biosynthetic potential. Activity
levels decreased gradually with leaf development,
reaching almost undetectable levels of activity in
fully-developed leaves (Table 1). In mature flowering
plants, the highest levels of activity was found
in the young leaves from side branches, while
inflorescences, but not the petals, displayed some
SAM dependent chavicol O-methyltransferase activity.
As previously observed in younger plants,
shoots and mature leaves displayed only minute
levels of activity. These findings further indicate
that young developing tissues are the primary sites
of methyl chavicol biosynthesis. The levels of estragole–biosynthetic
activity correlate well with
the presence and density of glandular trichomes
present on the tissues [11]. It has been shown that
in other members of the Lamiaceae, monoterpenebased
essential oils are biosynthesized by glandular
trichomes, and it could be that estragole is
biosynthesized in a similar regulatory fashion although
allylphenols are derived from the shikimic
acid pathway, a distinct biosynthetic pathway [37].
In all cases, the highest activity per fresh weight
was found in young tissues of the plant examined.
Some plant O-methyltransferases are developmentally
regulated [26], while others are induced
by fungal elicitors and during processes related to
plant defense responses [38–41]. In a similar fashion
to the sweet basil chavicol O-methyltransferase
activity described here, an
N-methyltransferase activity involved in caffeine
biosynthesis was also particularly active in young
emerging leaves of Coffea arabica [42].
3.3. Extraction and properties of the
SAM:cha�icol O-methyltransferase acti�ity from
sweet basil
Extraction of the enzyme without PVPP or PVP
resulted in an intense browning and formation of
precipitates, accompanied by losses in the enzymatic
activity and soluble protein. After partial
purification by Sephadex G-25 chromatography,
the enzyme was more stable, and could be kept for
more than 3 months at −20°C without apparent
loss of activity.
The activity levels linearly increased with incubation
time and were linearly dependent on
protein concentration up to 25 �g protein. A
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35 31
broad temperature optimum around 20°C was
found for activity. More than 50% of the activity
was retained when assays were performed at either
42 or 17°C. A heat treatment of the enzyme
preparation for 5 min at 100°C resulted in a
complete loss of activity, probably due to protein
denaturation.
Several buffers were tested (at 100 mM), including
sodium citrate, potassium phosphate, MOPS,
HEPES, Tris, and sodium carbonate, for allowing
optimal methyltransferase activity. A narrow pH
optimum was found at 7.5, attained either when
HEPES or potassium phosphate were used. Practically,
regardless of the buffer employed, at pH
values above 8.0 or below 7.0, activity sharply
decreased, loosing all activity at pH values below
6.0 or above 9.0, similar to the optimum pH range
of many plant methyltransferases [42,43].
Some methyltransferases require the presence of
metal cofactors for activity, while activities from
other sources do not require the presence of metal
cofactors for activity [44]. To test whether the
chavicol-O-methyltransferase from sweet basil requires
any metal cofactor for activity, 1 or 10 mM
of either CaCl 2, MgCl 2, MnCl 2, CoCl 2, ZnSO 4,or
FeSO 4 were added to the assays. All the additions
at 1 mM caused diminution of the activity by
50–80%, as compared with controls without any
further addition. Moreover the inclusion of 10
mM CoCl 2, ZnSO 4, or FeSO 4 in the assays resulted
in a complete loss of the enzymatic activity.
This indicated that the chavicol dependent Omethyltransferase
activity from sweet basil apparently
does not require a metal cofactor to be
active.
A constant 25 �M SAM level was used to
calculate an apparent K m for chavicol of 3.0 �M
utilizing both Lineweaver–Burk and Eadie–Hofstee
equations. Conversely, keeping a constant 5
�M of the acceptor substrate chavicol, a K m for
SAM was found to be 10 �M. This is whithin the
ranges of K m’s obtained for substrates of Omethyltransferases
from other sources [44].
The experiments described were performed on a
sweet basil chemotype (R3) that contains a high
(56%) level of estragole [10]. Although this chemotype
also contains high levels of the monoterpene
linalool (30%), and lower levels of other monoand
sesquiterpenes, it completely lacks both
eugenol and methyl eugenol [10]. It was therefore
of interest to study the substrate specificity of the

32
chavicol O-methyltransferase activity in this
chemotype. Although eugenol differs from chavicol
only in the addition of a m-methoxy group
(Fig. 1), the enzyme from the estragole-rich
chemotype R3 was almost completely inactive
when eugenol was offered as a substrate in lieu of
chavicol (Fig. 2). It could be that the additional
methoxy group in eugenol causes a substantial
steric disturbance in the active site of the enzyme,
rendering eugenol as an inefficient substrate for
methylation. We also found that this enzyme had
a narrow substrate specificity towards the acceptor
substrate, being active only towards p-hydroxylated,
but further unsubstituted phenols, such as
p-propyl phenol, p-ethyl phenol, and p-cresol at
�50% slower O-methylation rates, as compared
with the rate for chavicol, and completely inactive
towards p-phenyl phenol, o-allylphenol and phenol.
This indicates that only discrete alkyl additions
together with the availability of a p-hydroxyl
group, are crucial for the O-methylating ability of
this enzyme.
Fig. 2. Linkage between the essential oil composition and the
specificity and levels of SAM:dependent O-methyltransferase
activity in sweet basil chemotypes. Top panels, essential oil
composition of the different chemotypes data from reference
[10], linalool (L, white bars), estragole (Es, black bars),
methyl eugenol (ME, light gray) and eugenol (Eu, dark grey).
Bottom panels, SAM:allylphenol O-methyltransferase activities.
Substrate used, chavicol (Ch, black bars); substrate,
eugenol (Eu, gray bars). The products formed are estragole
(=methyl-chavicol) and methyl eugenol, respectively (Fig. 1).
Means and S.E. of three independent determinations are
shown.
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35
Interesting substrate specificities have been reported
in O-methyltransferases acting on phenolic
derivatives from other sources. The t-anol Omethyltransferase
from Pimpinella anisum accepted
the 3� methoxylated derivative eugenol at
almost the same rate as the non-3� methoxylated
1-propenyl analog t-anol, but chavicol and dihydroanol
(both non-3� methoxylated) were Omethylated
at ca. 50% of the rate of t-anol [28].
The isoeugenol O-methyltransferase from C. breweri,
was able to methylate eugenol at about 70%
the rate found for the 1-propenyl analog
isoeugenol [29]. Interestingly, changes in only a
few amino acids can dramatically modify the substrate
specificities of O-methyltransferases of C.
breweri [45].
3.4. Association between O-methyltransferase
acti�ity le�els, their substrate specificities and the
essential oil composition of sweet basil chemotypes
Many sweet basil chemotypes differ in their
allylphenol content and composition, and provide
a unique experimental system to study the biochemical
regulation of allyphenol biosynthesis.
Chemotype 145 lacks linalool, and contains an
essential oil that is largely constituted (91%) by
methyl chavicol [10]. As expected, the presence of
linalool did not have a pronounced effect on the
methyl chavicol biosynthetic capability, as evidenced
by the similar levels of chavicol specific
O-methyltransferase activity found in this chemotype.
In a similar fashion to the O-methyltransferase
activity displayed by chemotype R3, the
activity extracted from chemotype 145 does not
accept eugenol as a substrate (Fig. 2).
A different composition of allylphenol derivatives
is found in chemotype 147/97, rich both in
methyl chavicol and in methyl eugenol [10]. It
would be difficult to conceive that O-methyltransferases
with a strict substrate specificity for chavicol,
and unable to accept eugenol as a substrate
(such as those found in chemotypes R3 or 145)
could mediate the formation of methyl eugenol in
chemotype 147/97. Therefore, we postulated that
chemotype 147/97 possessed distinct O-methyltransferase
activities, able to accept either chavicol
or eugenol as acceptor substrates. The substrate
specificity of the O-methyltransferase activity obtained
from chemotype 147/97 is clearly different
and less strict, as compared with the one displayed

y the activity obtained from chemotype R3. The
results are shown in Fig. 2. The O-methyltransferase
activity extracted from chemotype 147/97
readily accepts eugenol as a substrate, even at a
higher rate than chavicol (Fig. 2). The identity of
the products formed was confirmed to be methyl
eugenol (or estragole, respectively), by radio-TLC
(not shown). The substrate specificity of the Omethyltransferase
activity(ies) extracted from
chemotype 147/97 provides a good biochemical
rationalization for the presence of both estragole
and methyl eugenol in the essential oil of this
chemotype. It is possible that the O-methyltransferase
activity extracted from chemotype 147/97
resides in one enzyme with a broad substrate
specificity, accepting either eugenol or chavicol as
substrates. Nevertheless, it is also possible that
chemotype 147/97 possesses more than one Omethyltransferase
activity, each with strict substrate
specificities for the methyl acceptors.
Finally, the sweet basil chemotype SW was examined
for the presence of chavicol or eugenolspecific
O-methyltransferase activities. This
chemotype accumulates high levels of linalool and
eugenol, but completely lacks the p-methoxylated
allylphenols methyl chavicol or methyl eugenol
[10]. As expected, this chemotype lacked chavicol
and eugenol dependant O-methyltransferase activities
(Fig. 2). These observations provide a biochemical
rationalization for the lack of
p-methoxylated allylphenols in this chemotype. Interestingly,
similar correlations between the levels
of isoeugenol methyltransferase (IEMT) activity
were found in flowers of C. breweri lines that emit
methyl isoeugenol versus lines that do not emit it,
and the regulation was apparently at a pre-translational
step of the IEMT involved [29].
Although, we cannot exclude the possibility that
the p-methylation of allylphenol derivatives takes
place before the allyl chain is biosynthesized, our
results indicate that p-methylation takes place after
the formation of the allylic chain. Moreover,
our findings suggest that the chemical composition
of the essential oil in sweet basil chemotypes is (at
least partially) dictated by the levels and substrate
specificities of the O-methyltransferases present.
The bulk of the allylphenol-specific O-methyltransferase
activity seems to be restricted to young
developing tissues. At present, it is unknown
whether only one enzymatic activity with a broad
substrate specificity is able to transfer methyl
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35 33
groups to either eugenol or chavicol, or whether
two or more separable enzymatic activities are
involved in these biochemical conversions. Crosses
between the different chemotypes will enable us to
learn more about the genetics of chemotype determination
in sweet basil. Additionally, information
obtained through purification and biochemical
characterization of the sweet basil enzymes, as well
as the isolation and analysis of their respective
genes will help us to better understand the process
of accumulation of allylphenolic derivatives in
sweet basil chemotypes.
Acknowledgements
We thank Ronald Fenn and Dr William L.
Schreiber for the chavicol. This work was supported
by a grant number 255-0495 from the Chief
Scientist of the Ministry of Agriculture of the
State of Israel and The US–Israel Binational Agricultural
Research and Development grant IS-2709-
96. Additional support was provided by the Otto
Meyerhof Center for Biotechnology established by
the Minerva Foundation, Federal Republic of
Germany. Contribution from the ARO, The Volcani
Center, Bet Dagan 50250, Israel, No. 101-
2000.
References
[1] E. Putievsky, B. Galambosi, Production systems of
sweet basil, in: R. Hiltunen, Y. Holm (Eds.), Basil. The
Genus Ocimum, Harwood Academic Publishers, 1999,
pp. 39–65.
[2] R. Reuveni, A. Fleischer, E. Putievsky, Fungistatic activity
of essential oils from Ocimum basilicum chemotypes,
Phytopath. Z. 110 (1984) 20–22.
[3] S. Dube, P.D. Upadhyay, S.C. Tripathi, Antifungal,
physicochemical, and insect repelling activity of the essential
oil of Ocimum basilicum, Can. J. Bot. 67 (1989)
2085–2087.
[4] R.A. Werner, Toxicity and repellency of 4-allylanisole
and monoterpenes from white spruce and tamarack to
the spruce beetle and Eastern larch beetle (Coleoptera:
Scolytidae), Environ. Entomol. 24 (1995) 372–379.
[5] B.M. Lawrence, Labiatae oils — mother nature’s chemical
factory, in: Essential Oils, Allured Publishing, Carol
Stream, IL, 1993, pp. 188–206.
[6] J.E. Simon, J. Quinn, R.G. Murray, Basil: a source of
essential oils, in: Advances in New Crops, Timber Press,
Portland, OR, 1990, pp. 484–489.
[7] D.J. Charles, J.E. Simon, A new geraniol chemotype of
Ocimum gratissimum L, J. Ess. Oil Res. 4 (1992) 231–
234.

34
[8] A.P. Martins, L.R. Salgueiro, R. Vila, F. Tomi, S.
Cañigueral, J. Casanova, A. Proenca da Cunha, T.
Adzet, Composition of the essential oils of Ocimum
canum, O. gratissimum and O. minimum, Planta Med. 65
(1999) 187–189.
[9] Z. Fleisher, A. Fleisher, Volatiles of Ocimum basilicum
traditionally grown in Israel. Aromatic plants of the
Holy Land and Sinai Part VIII, J. Ess. Oil Res. 4 (1992)
97–99.
[10] R.J. Grayer, G.C. Kite, F.J. Goldstone, S.E. Bryan, A.
Paton, E. Putievsky, Infraspecific taxonomy and essential
oil chemotypes in sweet basil Ocimum basilicum,
Phytochemistry 43 (1996) 1033–1039.
[11] E. Werker, E. Putievsky, U. Ravid, N. Dudai, I. Katzir,
Glandular hairs and essential oil in developing leaves of
Ocimum basilicum L. (Lamiaceae), Ann. Bot. 71 (1993)
43–50.
[12] G. Samuelsson, Drugs of natural origin, in: A Textbook
of Pharmacognosy, Swedish Pharmaceutical Press,
Stockholm, 1991, p. 320.
[13] O. Barazani, A. Fait, Y. Cohen, S. Diminshtein, U.
Ravid, E. Putievsky, E. Lewinsohn, J. Friedman, Chemical
variation among indigenous populations of Foeniculum
�ulgare var. �ulgare in Israel, Planta Med. 65 (1999)
486–489.
[14] E. Werker, E. Putievsky, U. Ravid, N. Dudai, I. Katzir,
Glandular hairs, secretory cavities and the essential oil in
leaves of tarragon (Artemisia dracunculus L), J. Herbs
Spices Medicinal Plants 2 (1994) 19–32.
[15] P. Manitto, P. Gramatica, D. Monti, Biosynthesis of
phenylpropanoid compounds. Part II. Incorporation of
specifically labelled cinnamic acids into eugenol, J.
Chem. Soc. Perkin I (1975) 1548–1551.
[16] P. Manitto, D. Monti, P. Gramatica, Biosynthesis of
phenylpropanoid compounds. Part I. Biosynthesis of
eugenol in Ocimum basilicum, J. Chem. Soc. Perkin I
(1974) 1727–1731.
[17] U.M. Senanayake, R.B.H. Wills, T.H. Lee, Biosynthesis
of eugenol and cinnamic aldehyde in Cinnamomum zeylanicum,
Phytochemistry 16 (1977) 2032–2033.
[18] R. Tressl, F. Drawert, Biogenesis of banana volatiles, J.
Agric. Food Chem. 21 (1973) 560–565.
[19] A. da Cunha, Purification, characterization and induction
of L-phenylalanine ammonia- lyase in Phaseolus
�ulgaris, Eur. J. Biochem. 178 (1988) 243–248.
[20] Y. Tanaka, I. Uritani, Purification and properties of
phenylalanine ammonia-lyase in cut-injured sweet
potato, J. Biochem. (Tokyo) 81 (1977) 963–970.
[21] S. Zimmermann, K. Hahlbrock, Light-induced changes
of enzyme activities in parsley cell suspension cultures.
Purification and some properties of phenylalanine ammonia-lyase
(E.C.4.3.1.5), Arch. Biochem. Biophys. 166
(1975) 54–62.
[22] Z. Hao, D.J. Charles, Y. Liangli, J.E. Simon, Purification
and characterization of a phenylalanine ammonialyase
from Ocimum basilicum, Phytochemistry 43 (1996)
735–739.
[23] M. Luckner, Secondary metabolism in microorganisms,
in: Plants and Animals, third ed., Springer, Berlin, 1990,
pp. 94–95.
[24] R.K. Ibrahim, A. Bruneau, B. Bantignies, Plant Omethyltransferases:
molecular analysis, common signa-
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35
ture and classification, Plant Mol. Biol. 36 (1998) 1–10.
[25] H. Meng, W.H. Campbell, Substrate profiles and expression
of caffeoyl coenzyme A and caffeic acid O-methyltransferases
in secondary xylem of aspen during seasonal
development, Plant Mol. Biol. 38 (1998) 513–520.
[26] K. Inoue, V.J.H. Sewalt, G.M. Ballance, W. Ni, C.
Stürzer, R.A. Dixon, Developmental expression and
substrate specificities of alfalfa caffeic acid acid 3-Omethyltransferase
and caffeoyl coenzye A 3-O-methyltransferase
in relation to lignification, Plant Physiol. 117
(1998) 761–770.
[27] J. Wang, N. Dudareva, C.M. Kish, J.E. Simon, E.
Lewinsohn, E. Pichersky, Nucleotide sequences of two
cDNAs encoding caffeic acid O-methyltransferases (accession
nos. AF154917 and AF154918) from sweet basil,
Plant Physiol. 120 (1999) 1205.
[28] B. Kemmerer, J. Reichling, S-Adenosyl-L-methionine:anol-O-methyltransferase
activity in organ cultures
of Pimpinella anisum, Phytochemistry 42 (1996) 397–
403.
[29] J. Wang, N. Dudareva, S. Bhakta, R.A. Raguso, E.
Pichersky, Floral scent production in Clarkia breweri
(Onagraceae). II Localization and developmental modulation
of the enzyme S-adenosyl-L-methionine:(iso)eugenol
O-methyltransferase and
phenylpropanoid emission, Plant Physiol. 114 (1997)
213–221.
[30] J. Wang, E. Pichersky, Characterization of S-adenosyl-
L-methionine:(iso)eugenol O-methyltransferase involved
in floral scent production in Clarkia breweri, Arch.
Biochem. Biophys. 349 (1998) 153–160.
[31] R. Croteau, M. Felton, F. Karp, R. Kjonaas, Relationship
of camphor biosynthesis to leaf development in sage
(Sal�ia officinalis), Plant Physiol. 67 (1981) 820–824.
[32] J. Gershenzon, Metabolic costs of terpenoid accumulation
in higher plants, J. Chem. Ecol. 20 (1994) 1281–
1328.
[33] J. Rohloff, Monoterpene composition of essential oil
from peppermint (Mentha x piperita L.) with regard to
leaf position using solid-phase microextraction and gas
chromatography/mass spectrometry analysis, J. Agric.
Food Chem. 47 (1999) 3782–3786.
[34] J. Gershenzon, M.E. McConkey, R.B. Croteau, Regulation
of monoterpene accumulation in leaves of peppermint,
Plant Physiol. 122 (2000) 205–213.
[35] N. Dudai, E. Lewinsohn, O. Larkov, I. Katzir, U.
Ravid, D. Chaimovitsh, D. Sa’adi, E. Putievsky, Dynamics
of yield components and essential oil production
in a commercial hybrid sage (Sal�ia officinalis x Sal�ia
fruticosa cv. Newe Ya’ar No. 4), J. Agric. Food Chem.
47 (1999) 4341–4345.
[36] M.M. Bradford, A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing
the principle of protein-dye binding, Anal. Biochem. 72
(1976) 248–254.
[37] R.K. Croteau, Origin of natural odorants, in: P.M.
Muller, D. Lamparsky (Eds.), Perfumes Art, Science and
Technology, Elsevier, London, 1991, pp. 101–126.
[38] A.E. Pakusch, R.E. Kneusel, U. Matern, S-adenosyl-Lmethionine:t-caffeoyl-coenzyme
A 3-O-methyltransferase
from elicitor-treated parsley cell suspension
cultures, Arch. Biochem. Biophys. 271 (1989) 488–494.

[39] M. Ichimura, T. Furuno, T. Takahashi, R.A. Dixon,
S.I. Ayabe, Enzymic O-methylation of isoliquiritigenin
and licodione in alfalfa and licorice cultures, Phytochemistry
44 (1997) 991–995.
[40] B. Dumas, M. Legrand, P. Geoffroy, B. Fritig, Purification
of tobacco O-methyltransferases by affinity
chromatography and estimation of the rate of synthesis
of the enzymes during hypersensitive reaction to virus
infection, Planta 176 (1988) 36–41.
[41] R. Rakwal, M. Hasegawa, O. Kodama, A methyltransferase
for synthesis of the flavanone phytoalexin
sakuranetin in rice leaves, Biochem. Biophys. Res.
Commun. 222 (1996) 732–735.
[42] S.S.M. Waldhauser, J.A. Kretschmar, T.W. Baumann,
N-methyltransferase activities in caffeine biosynthesis:
E. Lewinsohn et al. / Plant Science 160 (2000) 27–35 35
.
biochemical characterization and time course during
leaf development of Coffea arabica, Phytochemistry 44
(1997) 853–859.
[43] H. Wengenmayer, J. Ebel, H. Grisebach, Purification
and properties of a S-adenosylmethionine isoflavone
4�-O-methyltransferase from cell suspension cultures of
Cicer arietinum L, Eur. J. Biochem. 50 (1974) 135–143.
[44] D. Strack, H.P. Mock, Hydroxycinnamic acids and
lignins, in: P.J. Lea (Ed.), Methods in Plant Biochemistry.
Enzymes of Secondary Metabolism, vol. 9, Academic
Press, London, 1993, pp. 45–97.
[45] J. Wang, E. Pichersky, Identification of specific
residues involved in substrate discrimination in two
plant O-methyltransferases, Arch. Biochem. Biophys.
368 (1999) 172–180.