Metabolism of geraniol in grape berry mesocarp of Vitis vinifera L. cv. Scheurebe: Demonstration of stereoselective reduction, E/Z-isomerization, oxidation and glycosylation

Article abstract of DOI:10.1016/j.phytochem.2004.12.017

The metabolism of deuterium labeled geraniol in grape mesocarp of Vitis vinifera L. cv. Scheurebe was studied by in vivo-feeding experiments. Stereoselective reduction to (S)-citronellol, E/Z-isomerization to nerol, oxidation to neral/geranial and glycosylation of the corresponding monoterpene alcohols could be demonstrated. Time course studies including the determination of conversion rates revealed that the activity of these secondary transformations of monoterpenes is dependent on the ripening stage and can be distinguished from the development of the primary monoterpene synthase activities by the sharp increase at the end of the ripening period. The stereoselective biosynthesis of the potent odorant cis-(2S,4R)-rose oxide from labeled geraniol in grape berry mesocarp is demonstrated as well. Since (S)-citronellol is the precursor of cis-(2S,4R)-rose oxide it can be concluded that especially the last part of the ripening period is important for the generation of this potent odorant. This finding confirms the conclusion that a higher concentration of flavor compounds could be established in the berries by leaving the fruit on the vine for extended periods.

Full text of DOI:10.1016/j.phytochem.2004.12.017

PHYTOCHEMISTRY
Phytochemistry 66 (2005) 295–303
www.elsevier.com/locate/phytochem
Metabolism of geraniol in grape berry mesocarp of Vitis vinifera L.
cv. Scheurebe: demonstration of stereoselective reduction,
E/Z-isomerization, oxidation and glycosylation
a
Fang Luan , Armin Mosandl , Andreas M u¨ nch , Matthias W u¨ st
a
a
b,
*
a
Institut f u¨ r Lebensmittelchemie, Johann Wolfgang Goethe-Universit a¨ t, Marie-Curie-Strasse 9, D-60439 Frankfurt (Main), Germany
b
University of Applied Sciences Valais, Department of Life Technologies, Route du Rawyl 47, CH-1950 Sion 2, Switzerland
Received 24 September 2004; received in revised form 9 December 2004
Abstract
The metabolism of deuterium labeled geraniol in grape mesocarp of Vitis vinifera L. cv. Scheurebe was studied by in vivo-feeding
experiments. Stereoselective reduction to (S)-citronellol, E/Z-isomerization to nerol, oxidation to neral/geranial and glycosylation of
the corresponding monoterpene alcohols could be demonstrated. Time course studies including the determination of conversion
rates revealed that the activity of these secondary transformations of monoterpenes is dependent on the ripening stage and can
be distinguished from the development of the primary monoterpene synthase activities by the sharp increase at the end of the rip-
ening period. The stereoselective biosynthesis of the potent odorant cis-(2S,4R)-rose oxide from labeled geraniol in grape berry
mesocarp is demonstrated as well. Since (S)-citronellol is the precursor of cis-(2S,4R)-rose oxide it can be concluded that especially
the last part of the ripening period is important for the generation of this potent odorant. This finding confirms the conclusion that a
higher concentration of flavor compounds could be established in the berries by leaving the fruit on the vine for extended periods.
ꢀ
2004 Elsevier Ltd. All rights reserved.
Keywords: Vitis vinifera; Monoterpenes; Rose oxide; Citronellol; Biosynthesis; Enantioselective multidimensional gas chromatography–mass spec-
trometry (enantio-MDGC–MS); Stir bar sorptive extraction (SBSE)
1
. Introduction
identified in grape must and wine (Rapp, 1992; Williams
et al., 1980, 1985; Luan et al., 2004). In addition to the
free odour-producing forms of monoterpenes, several
glycosidically bound forms of monoterpenes have been
described (Strauss et al., 1986).
Monoterpenes contribute substantially to the charac-
teristic floral aroma and flavor of the so called muscat
and non-muscat aromatic grape varieties (Vitis vinifera).
Well known representatives are Muscat of Alexandria
and Gew u¨ rztraminer, respectively. Beside monoterpene
alcohols (linalool, geraniol 6, nerol 8, citronellol 10
and a-terpineol) and monoterpene ethers (rose oxide
Among the odiferous monoterpenes the cyclic ether
0
rose oxide 12 (cis- and trans-4-methyl-2-(2 -methyl-1-
propenyl)-tetrahydropyran) is a potent odorant in
Scheurebe and Gew u¨ rztraminer wines, as could be
shown by Guth (1997a,b) using gas chromatography/
olfactometry. 12 has been previously detected in many
fruits and essential oil producing plants (Naves et al.,
12, nerol oxide) numerous other monoterpene com-
pounds, such as polyhydroxylated derivatives, have been
*
Corresponding author. Tel.: +41 27 606 86 54; fax: +41 27 606 85
1961; von Sydow and Karlsson, 1971; Kreis and Mo-
sandl, 1994).
15.
E-mail address: matthias.wuest@hevs.ch (M. W u¨ st).
0
031-9422/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2004.12.017

296
F. Luan et al. / Phytochemistry 66 (2005) 295–303
Despite of their importance to wine flavor and aroma
The enantioselective analysis of the fractions was car-
ried out with an enantio-MDGC–MS system equipped
with an achiral pre column and a chiral main column.
The elution order of the enantiomers of citronellal 14a/
relatively little is known about the biochemistry of mon-
oterpenoids in V. vinifera. The de novo biosynthesis of
monoterpenes in grapes and leaves via the novel
DOXP/MEP pathway has been demonstrated (Luan
and W u¨ st, 2002) and recently a terpene synthase catalyz-
ing the cyclization of geranyl diphosphate to (ꢀ)-a-ter-
pineol has been cloned and functionally characterized
for the first time in an angiosperm (Martin and Bohl-
mann, 2004). However, the mechanisms that underlie
secondary transformations of monoterpenes like
hydroxylation, reduction or glycosylation have not yet
been characterized in vivo. Glycosylation activity in de-
tached grape berries grown in vitro have been demon-
strated (Bravdo et al., 1990) and cell suspension
cultures of the variety Muscat de Frontignan are able
to transform geraniol 6 into geranial 16 ((E)-3,7-dimeth-
ylocta-2,6-dien-1-al), neral 18 ((Z)-3,7-dimethylocta-2,6-
dien-1-al), nerol 8 and citronellol 10 (Ambid et al.,
b, citronellol 10a/b and d -linalool 19a/b was determined
6
on the chiral stationary phase heptakis(2,3-di-O-acetyl-
6-O-tert-butyldimethylsilyl)-b-cyclodextrin (diAc-b) by
co-injection of enantiopure and racemic standards. The
enantioselective analysis of the SBSE extracts were car-
ried out with an enantio-SBSE-MDGC–MS system
equipped with an achiral precolumn and a chiral main
column. The elution order of the enantiomers of cis-
and trans-rose oxide 12a–d was determined on the chiral
stationary phase heptakis(2,3-di-O-methyl-6-O-tert-
butyldimethylsilyl)-b-cyclodextrin (diMe-b) by co-injec-
tion of enantiopure and racemic standards.
All assays were performed at least three times. Con-
trol experiments with musts that had been boiled for
10 min showed no enzymatic activity. Fig. 1 shows a
standard chromatogram using the chiral main column
of the enantio-MDGC-MS system. The enantiomers of
linalool (19), citronellol (10) and citronellal (14) are well
separated with Di-Ac-b as the chiral stationary phase.
A typical main column chromatogram of free labeled
and genuine geraniol 5/6, nerol 7/8, and citronellol 9/10
obtained from grape berries after administration of
labeled geraniol is shown in Figs. 2(A)–(C). The conver-
sion of d -geraniol 5 into d -nerol 7 and d -(S)-citronel-
1
983). However, the cell suspension cultures used in
these studies did not accumulate detectable amounts of
monoterpenoids (as it is frequently observed in plant cell
cultures of essential oil accumulating plants due to an
increased activity of catabolic enzymes (Falk et al.,
1
990)) and therefore may not be comparable with the
situation in vivo. First preliminary data on the in vivo
metabolism of deuterium labeled linalool and geraniol
have been previously published (Luan et al., 2003) and
in this study we would like to present a more detailed
investigation on the metabolism of geraniol, which in-
cludes time course studies, determination of conversion
rates and subsequent enantioselective multidimensional
GC/MS analysis of the metabolites giving information
on the stereochemical aspects. The stereoselective bio-
synthesis of the potent odorant cis-(2S,4R)-rose oxide
in grape berry mesocarp is demonstrated as well.
6
6
6
lol 9b is clearly detectable as illustrated by Fig. 2(B). The
labeled terpenes can be detected selectively on mass lane
m/z = 75 because the base peaks of the corresponding
unlabeled geraniol 6, nerol 8 and citronellol 10 [m/
z = 69] are shifted by 6 mass units as illustrated by
Fig. 3. Additionally, inverse isotope effects (Matucha
et al., 1991) were observed during the enantio-
2
. Results and discussion
Almost all monoterpens are derived from geranyl
diphosphate. However, the metabolism of geraniol itself
in intact grape berries in vivo has remained unknown.
In order to investigate the metabolism of geraniol 6 in
grapes, in vivo-feeding experiments were performed with
deuterium labeled d -geraniol 5, which was prepared by
6
a 4 step synthesis using geranyl acetate 1 as starting
material. To avoid light induced oxygenation and E/Z-
isomerization of geraniol 6 the grape berries were incu-
bated with exclusion of light following the injection of
labeled geraniol into the mesocarp of ripening berries
at different stages of ripeness as characterized by their
sugar content in degrees of Oechsle (ꢁOe). The isolation
and fractionation of the target compounds was achieved
by solid phase extraction (SPE) according to the method
of Mateo et al. (1997).
Fig. 1. Chromatogram of chiral and nonchiral reference compounds,
using enantio-MDGC–MS (main column Di-Ac-b-CD): geraniol (6);
nerol (8); geranial (16); neral (18); Chiral compounds are well
differentiated into their enantiomers: (3R)-citronellol (10a); (3S)-
citronellol (10b); (3R)-citronellal (14a); (3S)-citronellal (14b); d
6
-(3R)-
linalool (19a), used as internal standard; d -(3S)-linalool (19b).
6

F. Luan et al. / Phytochemistry 66 (2005) 295–303
297
Fig. 3. Mass spectra of genuine and corresponding deuterium labelled
geraniol.
nellol (Gramatica et al., 1982). It should be, therefore,
possible to distinguish the origin of citronellol from wine
berries and yeast fermentation in wines, respectively, by
means of an enantioselective analysis.
Beside a stereoselective reduction, an E/Z-isomeriza-
tion of labeled geraniol to nerol could be detected dem-
onstrating the presence of an isomerase. In flower heads
of Rosa damascena such an isomerase activity has been
previously detected (Banthorpe and Poots, 1979) and
isomerization of geraniol and geranyl phosphate to ner-
ol and neryl phosphate, respectively, by cell-free enzyme
systems from carrot leaves have been reported as well
(Shine and Loomis, 1974). It is conceivable that in intact
berries a non-enzymatic E/Z isomerization is triggered
by natural light. If this is a process of significance in vivo
still needs to be verified by experiments with berries illu-
minated in a diurnal fashion. Furthermore, in the pres-
ent study the generation of geranial and neral in berry
mesocarp could be detected, which requires the presence
of an oxidase and/or a dehydrogenase. However, only
trace amounts of neral and geranial were generated
and were thus not quantified. A very low monoterpenol
oxidase and dehydrogenase activity has been previously
detected in cell-free extracts from Pelargonium graveo-
lens leaves (Banthorpe et al., 1983).
Fig. 2. Main column chromatogram (Di-Ac-b-CD) of labelled metab-
olites, generated from d -geraniol (conditions: see Fig. 1).
6
MDGC–MS analysis. The deuterium labeled monoter-
pene isotopomers show a shift of their peak maximum
of approximately 20 s when compared with the unla-
beled ones.
Generated free and glycosidically bound d -citronel-
6
lol 9 shows an enantiomeric purity better than 95% in fa-
vor of the (S)-configured enantiomer during the whole
ripening period (see below). Thus, the enzymatic reduc-
tion of the C2/C3 double bond of geraniol is a stereose-
lective process. Whether the corresponding reductase is
able to reduce nerol as well, that is generated by an
isomerase (see below) can not be answered at this stage
because this requires the separation of the reductase and
isomerase by a partial purification. However, a geraniol-
nerol reductase has been isolated from rose petals (Dun-
phy and Allcock, 1972). The formation of (S)-citronellol
via stereoselective reduction of geranial and concomi-
tant reduction of the generated citronellal can be ex-
cluded because traces of citronellal could not be
detected. Interestingly, the reduction of geraniol by Sac-
charomyces cerevisiae yields stereoselectively (R)-citro-
For analyzing non-polar metabolites of geraniol in
the ppb range an enantio-SBSE–MDGC–MS system
(Kreck et al., 2002) was employed. Thus, the conversion
of d -geraniol into cis- and trans-d -rose oxide 11a–c in
6
6
grape berry mesocarp could be detected (see Fig. 4).
The mass spectrum of the labeled rose oxide is in good

298
F. Luan et al. / Phytochemistry 66 (2005) 295–303
oxide biosynthesis has been investigated using deute-
rium-labelled (R)- and (S)-citronellol and their corre-
sponding glucosides. It could be shown that citronellol
(3,7-dimethyl-6-octen-1-ol) is hydroxylated by a rather
unspecific enzyme to yield two labile isomeric diols of
which (E)-3,7-dimethyloct-5-en-1,7-diol can easily cy-
clize in acidic media to yield cis- and trans-rose oxide.
Because of its instability in acidic media free (E)-3,7-
dimethyloct-5-en-1,7-diol has not been yet detected in
musts or wines. However, 3,7-dimethyloct-7-en-1,6-diol,
which is also generated by hydroxylation of citronellol,
could be detected in Morio-Muscat grape must and wine
together with cis- and trans-rose oxide (Rapp et al.,
1984). It is noteworthy that in less acidic plant material
like rose petals free and glucosylated (E)-3,7-dimethy-
loct-5-en-1,7-diol could be isolated and characterized
(
Watanabe et al., 1998). Due to the fact that the rose
oxide concentration is constantly rising during wine
aging (Guth, 1998), it seems reasonable to assume that
glycosylated (E)-3,7-dimethyloct-5-en-1,7-diol is present
in grape must as well. Hydrolysis of glycosylated flavor
precursors during wine ageing with subsequent forma-
tion of odiferous compounds from the liberated reactive
precursors is a well known process (Strauss et al., 1986).
The metabolism of d -geraniol 5 in grape berries is sum-
6
marized in Fig. 5. It is remarkable that the main stereo-
isomer of rose oxide that is generated in the berry is the
one with the lowest odor threshold value (cis-(2S/4R)-
rose oxide: 0.2 ppb as determined by Guth (1997a,b)).
Even a small net conversion of the precursor citronellol
has therefore a significant influence on the wine aroma.
Additionally, the quantification of the enzymatic
Fig. 4. Mass spectra of genuine (A) and deuterium labeled (B) rose
oxide; Chiral main column chromatogram of labelled rose oxide
transformation of d -geraniol 5 into free and bound
6
d -nerol 7 and d -(S/R)-citronellol 9 was carried out
6 6
6
stereoisomers, generated from d -(S)-citronellol (C).
with d -(R)-linalool 19a as internal standard. This con-
6
version rate was measured with ripening berries at dif-
ferent stages of maturity as characterized by their sugar
content in ꢁOe degrees. However, the conversion rates
presented in the following section should be regarded
as semi-quantitative given the rather imprecise method
of precursor application. Nevertheless, the data show
agreement with that of a synthesized reference com-
pound (W u¨ st et al., 1998a,b). The predominant stereo-
isomer generated is cis-d -(2S,4R)-rose oxide 11b,
6
which is in agreement with a previous study by Guth
(1997a) showing the same predominant stereoisomer in
Scheurebe and Gew u¨ rztraminer wines. However, the
that d -geraniol was quite efficiently metabolized in
6
enantiomeric purity of labeled cis-rose oxide (ratio of
the mesocarp. The total conversion rate in % (free
and bound metabolites) raised from 34% in immature
berries (ꢁOe: 14.6) to 62% in ripe berries (ꢁOe: 65.6).
The absolute conversion rate in nmol/g berry showed
a sharp increase at the end of the ripening period for
all three enzymatic activities (geraniol-nerol reductase,
geraniol isomerase, and geraniol-nerol-citronellol gly-
cosidase; see Fig. 6), especially those of reductase and
glycosidase. These results show that the activity of sec-
ondary transformations of monoterpenes like reduc-
tion, isomerization and glycosylation is dependent on
the ripening stage and can be distinguished from the
development of the primary monoterpene synthase
activity by the sharp increase at the end of the ripening
(
2R,4S)/(2S/4R) = 10/90) is somewhat higher than in
Gew u¨ rztraminer wine (ratio of (2R,4S)/(2S/4R)=
0/70). The absolute configuration at C4 of cis
3
d6-(2S,4R)-rose oxide 11b corresponds with the absolute
configuration at C3 of the generated d -(S)-citronellol 9b
6
(
see Fig. 5). It is, therefore, highly probable that rose
oxide is generated from geraniol and/or nerol by a reac-
tion sequence, that is shown in Fig. 5. After stereoselec-
tive reduction the generated (S)-citronellol is cyclized to
cis- and trans-rose oxide. It is reasonable to assume that
the cyclization of citronellol is accomplished by the same
mechanism that has been demonstrated earlier in Pelar-
gonium plants (W u¨ st et al., 1996, 1998a,b), when rose

F. Luan et al. / Phytochemistry 66 (2005) 295–303
299
6
Fig. 5. Proposed metabolization scheme of d -geraniol.
it has been demonstrated for linalool synthase (lis)
in Muscat de Frontignan berries (Ebang-Oke et al.,
2003).
3. Conclusions
The stereoselective reduction of labeled geraniol 5
with subsequent cyclization of generated d -(S)-citronel-
6
lol 9b to the potent odorant cis-(2S, 4R)-rose oxide 11b
has been demonstrated in Scheurebe grape berry meso-
carp by in vivo feeding experiments. Additionally, glyco-
sylation, E/Z-isomerization and oxidation were
observed. The determination of conversion rates showed
a sharp increase of the geraniol reductase activity to-
wards the end of the ripening period. Since citronellol
is the precursor of rose oxide it can be concluded that
especially the last part of the ripening period is impor-
tant for the generation of this potent odorant. This find-
ing confirms the conclusion that a higher concentration
of flavor compounds could be established in the berries
by leaving the fruit on the vine for extended periods
Fig. 6. Relationship between ꢁOe and enzymatic activity in nmol/g
grape (n = 3 or 4): ¤, reductase activity (conversion rate of d -geraniol
into free and bound d
-citronellol); ꢁ, isomerase activity (conversion
rate of d -geraniol into free and bound d -nerol); n, glycosidase activity
conversion rate of d -geraniol into labeled bound geraniol, nerol, and
citronellol).
6
6
(
Wilson et al., 1984).
The identification of grape mesocarp as a site of
6
6
(
6
monoterpene secondary transformations is an essential
result of this study and it sets the foundation for more
definitive studies like the isolation, partial purification
and characterization of the corresponding enzymes
which is currently in progress.
period. For example monoterpene synthase activity is
characterized by a constant increase after the beginning
of ripening, reaching its maximum at full ripeness as

300
F. Luan et al. / Phytochemistry 66 (2005) 295–303
4
. Experimental
4.4. Isolation of rose oxide by SBSE
4
.1. Plant material
Rose oxide was extracted using the novel stir bar
sorptive extraction method (SBSE; Baltussen et al.
(1999)) instead of SPE because of its low concentra-
tion. Following the in vivo feeding of labeled geraniol
grape berries were homogenized as described above
and the juice was decanted. Rose oxide was extracted
for 30 min by SBSE with stir bars obtained from
Ripening grape berries (degree Oechsle between 14.6
and 65.6 as determined using an Abb e´ -refractometer)
of V. vinifera L. cultivars Scheurebe were obtained from
the Research Center Geisenheim, Department for
Grapevine Breeding and Grafting (Geisenheim, Ger-
many) during the vintage 2003.
Gerstel (M u¨ hlheim Germany) under the trade name
ꢂ
Twister
.
4
.2. In vivo feeding experiments
4.5. Determination of conversion rate and quantification
of metabolites
d6-Geraniol 5 (10 mg/ml in DMSO), was injected into
the mesocarp of ripening grape berries (5 ll/g grape)
using a microliter syringe. A large part of the substrate
solution was squirted out of the berry mesocarp after
injection due to the pressure inside the berries and was
removed by wiping off. The exact amount of labeled
geraniol that rested in the berry was determined using
d6-(R)-linalool as internal standard as described below.
The berries were incubated for 3 d in the dark at room
temperature and then stored at ꢀ20 ꢁC until used. Blank
experiments were carried out with grape must and sub-
strate: 50 g of grape berries were homogenised and the
must was boiled for 10 min. After the must had cooled
The conversion rate of d -geraniol and the quantifica-
6
tion of the generated metabolites was determined using
2
[8,9- H ]-(3R)-linalool 19a as internal standard.
6
4.6. Instrumentation
4.6.1. Gas chromatography–mass spectrometry
The gas chromatography–mass spectrometry (GC–
MS) analysis of the synthesised compounds was carried
out with a Fisons Instruments GC 8065, coupled to a
Fisons Instruments MD800 mass spectrometer. Chro-
matography was performed with a fused silica column
(30 m · 0.25 mm i.d., 0.23 lm) coated with SE 52 and
programmed from 60 (5 min isothermal) to 260 ꢁC
down d -geraniol 5 was added (0.5 ll/g must). The mix-
6
ture was then incubated for 3 d in the dark at room tem-
perature and then stored at ꢀ20 ꢁC until used.
(
20 min isothermal) at 5 ꢁC/min with a 2 ml/min He
4.3. Isolation of free and glycosidically bound
monoterpenes by SPE
flow rate and 20 ml/min split. Mass spectra were ob-
tained at 70 eV with an ion source temperature of
2
00 ꢁC.
Seeds of about 50 g grape berries were removed and
the berries were ground with a mortar and pestle. Dur-
ing grinding the pH-value of the must was adjusted to 4
with phosphate buffer (pH 7). The mixture was com-
pletely homogenised with an Ultra Turrax blender for
4.6.2. Enantio-MDGC–MS system
The MDGC analysis was performed with a Siemens
SiChromat 2 coupled to a Finnigan MAT ITD 800.
The two capillary columns of the MDGC were coupled
with a live-T-switching device. The pre column was the
same as the column of GC–MS described above,
2
min. After centrifugation at 3000g for 5 min, the
supernatant was treated with Carrez I and II (Matissek
et al., 1992). After filtration, the filtrate was used for the
isolation of free and bound monoterpenes.
carrier gas, H ; split, 20 ml/min; injector temperature,
2
250 ꢁC; detector, FID; oven temperature, 60 ꢁC
(5 min isothermal) to 250 ꢁC at 5 ꢁC/min. The chiral
main column was a 30 m · 0.25 mm i.d., 0.25 lm film
thickness DiAc-b (heptakis-(2,3-di-O-acetyl-6-O-tert-
For the isolation and fractionation of the target com-
pounds solid phase extraction was used according to the
method of Mateo et al. (1997). The filtrate was passed
through a 2-g C18-RP column (SUPELCO, Bellefonte,
USA), previously activated with methanol. Free mono-
terpenes were eluted with dichloro methane and glyco-
sides with methanol. The methanol was removed
under reduced pressure, and the residue was taken up
in citrate–HCl buffer (pH 4.0). The resulting solution
butyldimethylsilyl)-b-cyclodextrin); carrier gas H ; oven
2
temperature, 60 ꢁC (30 min isothermal) to 200 ꢁC at
2 ꢁC/min. The main column was coupled to the transfer
line of the mass spectrometer, using an open split inter-
face with a helium sweeping flow of 1 ml/min; temper-
ature of ion trap manifold, 200 ꢁC, temperature of
transfer line: 250 ꢁC; EI, 70 eV.
ꢂ
was incubated with 2 mg of Glucanex per gram tissue
(
Novo Nordisk, Dittingen, Switzerland) for 24 h at
room temperature to release the glycosidically bound
monoterpenes. The liberated monoterpenes were ex-
tracted with methyl tert-butyl ether (MTBE).
4.6.3. Enantio-SBSE–MDGC–MS system
The system has been previously described by Kreck
et al. (2002).

F. Luan et al. / Phytochemistry 66 (2005) 295–303
301
1
.6.4. H NMR
2
and 2.5 g (19.2 mmol) of 2-bromo-[1,1,1,2,3,3,3,3- H ]-
4
7
The NMR spectra were obtained with a Bruker ARX
00, 300 MHz, in CDCl with TMS as the internal stan-
dard. The chemical shifts are given in d (ppm).
propane was heated at 150 ꢁC for 40 h in a special thick
wall tube with a sealed scew cap. Recrystallization of the
crude product with diethyl ether/ethanol (3:2 v/v)
yielded 6.53 g (86.7%) of the title compound; mp of
the resulting crystals: 239–240 ꢁC (lit. 238–239 ꢁC for
the unlabeled compound (Fagerlund and Idler, 1957)).
3
3
4
.7. Synthesis
2
.7.1. Synthesis of the precursor [8,8,8- H ]-3-methyl-7-
4
[
4
3
2
2
4.7.1.4. Synthesis of [8,8,8- H ]-1-acetoxy-3-methyl-7-
H ]methylocta-2E,6-dien-1-ol (d -geraniol) 5
3
6
3
2
[ H ]methylocta-2E,6-diene 4 (Pyun et al., 1993). A
.7.1.1. Synthesis of 1-acetoxy-6 7-epoxy-3 7-dimethyl-
3
octa-2-ene 2 (Manning, 1979). A solution containing
5.0 mmol (2.94 g) of 1-acetoxy-3,7-dimethylocta-2,6-
diene 1 in 50 ml of methylene chloride was cooled to
solution of 1.6 M n-butyllithium in hexane (3.7 ml,
5.88 mmol) was added dropwise to a suspension of
1
2
2.31 g (5.88 mmol) [1,1,1,3,3,3- H ]-[2- H ]-isopropyl
2
6
1
0
2
ꢁC. To this solution
a
suspension containing
0.2 mmol (4.37 g) of 78,7% m-chlorperoxybenzoic
triphenylphosphoniumbromide in THF in ice-bath un-
der N -atmosphere. After 1 h at room temperature
2
(
MCPBA) in 25 ml of methylene chloride was added
over 10 min. The mixture was stirred at 5 ꢁC for
0 min and then treated with small portion of suspen-
sion of MCPBA until the starting material disappeared
TLC-control). The mixture was filtered and the filtrate
1.0 g (5.88 mmol) of 6-acetoxy-4-methyl-5-hexenal 3
was added to the solution at 0 ꢁC. After 3 h stirring at
room temperature, the solution was quenched by adding
1.2 ml of MeOH. Solvent removed under reduced pres-
sure gave the crude product. The residue was extracted
three times with hexane by refluxing. The combined hex-
ane layer was concentrated and the residue was purified
by flash chromatography (pentane/diethylether 3:1 v/v)
to give 531 mg (44.6%) of the title compound 4.
3
(
was washed sequentially with aqueous NaHCO , water
3
and brine and then dried (Na SO ). The solution was
2
4
concentrated to give an oil which gave 13.9 mmol
(2.95 g, 92.8%) of epoxide 2.
MS m/e (rel. int.) 152 (3), 137 (3), 109 (13), 94 (9), 85
(43), 81 (46), 79 (43), 71 (40), 59 (40), 43 (100), 41 (39).
+
MS m/e (rel. int.) 160 (1, M ꢀ OAc), 142 (12), 124
(12), 93 (39), 80 (20), 75 (100), 67 (28), 43 (68).
H NMR: d (ppm) 1.70 (s, 3 H, 10-H, CH C@); 2.06
1
1
H NMR: d (ppm) 1.26 (s, 3H, 8-H, (CH ) CO); 1.30
3
2
3
(
s, 3H, 9-H, (CH ) CO); 1.72 (s, 3H, 10-H, CH C@);
(s, 3H, 12-H, OAc); 2.05–2.13 (m, 4H, 4-H, 5-H, meth-
ylene); 4.57–4.58 (d, 2H, 1-H, CH COAc); 5.06–5.10
3
2
3
1.51–1.60 (m, 2H, 4-H, methylene); 2.05 (s, 3H, 12-H,
OAc); 2.10–2.23 (m, 2H, 5-H, methylene); 2.68–2.72 (t,
2
(t, 1H, 6-H, CH @C(CD ) ); 5.32–5.37 (t, 1H, 2-H,
2
3 2
1
-H, 6H, HCO), 4.57–4.60 (d, 2H, 1-H, CH OAc);
2
CH@).
5
.35-5.41 (t, 1H, 2-H, CH @).
2
2
.7.1.5. Synthesis of [8,8,8- H ]-3-methyl-7-[ H ]meth-
2
4
3
3
4
.7.1.2. Synthesis of 6-acetoxy-4-methyl-4-hexenal 3
Manning, 1979). To the solution of 1.70 g (8.0 mmol)
ylocta-2E,6-dien-1-ol (d -geraniol) 5 (Streitwieser et al.,
6
2
2
1994). [8,8,8- H ]-1-Acetoxy-3-methyl-7-[ H ]methyl-
(
3
3
of 1-acetoxy-6,7-epoxy-3,7-dimethylocta-2-ene
3
2
2
in
0 ml of anhydrous diethyl ether a solution containing
.0 g (8.8 mmol) of periodic acid in 30 ml THF was
octa-2,6-diene 4 (531 mg, 2.62 mmol) were dissolved in
15 ml of dry methanol. To the solution 400 mg of anhy-
drous K CO was added. The reaction mixture was stir-
2
3
added slowly with vigorous stirring. The resulting mix-
ture was stirred for 90 min at room temperature. The
mixture was quenched with ice water. The organic layer
was separated and washed with aqueous NaHCO3,
water, and brine. The organic layer was dried and con-
centrated under reduced pressure to yield 1.0 g
red for 4 h at room temperature. After addition of 25 ml
diethyl ether, the mixture was washed with water and
brine. After drying the organic layer was removed under
reduced pressure. The residue was purified by flash chro-
matography (pentane/diethylether 3:1 v/v) to give
310 mg (73.8%) of d -geraniol 5.
6
+
MS m/e (rel. int.) 160 (1, M ), 142 (7), 124 (9), 93
(5.9 mmol, 73.6%) of title compound 3. This oil was
used for the next step without purification.
(46), 80 (18), 75 (100), 71 (11), 67 (11), 44 (48).
1
MS m/e (rel. int.) 126 (23), 110 (34), 84 (96), 81 (49),
9 (26), 67 (44), 55 (30), 43 (100), 41 (35).
H NMR: d (ppm) 1.68 (s, 3H, 10-H, CH C@); 1.99–
3
7
2.17 (m, 4H, 4-H, 5-H, methylene); 3.21 (s, 1H, OH);
4.09–4.16 (d, 2H, 1-H, CH OH); 5.07–5.11 (t, 1H, 6-H,
1
H NMR: d (ppm) 1.72 (s, 3H, 7-H, CH C@); 2.05 (s,
3
2
3
2
H, 9-H, OAc); 2.35–2.40 (t, 2H, 2-H, methylene); 2.55–
.60 (t, 2H, 3-H, methylene); 4.56–4.58 (d, 2H, 6-H,
CH @C(CD ) ); 5.39–5.44 (t, 1H, 2-H, CH@).
2
3 2
2
CH COAc); 5.32–5.38 (t, 1H, 5-H, CH @).
4.7.2. Synthesis of the internal standard [8,8,8- H ]-
3
2
2
2
3R)-methyl-7-[ H ]methylocta-1,6-dien-3-ol (d -(3R)-
(
linalool) 19a
3
6
2
2
.7.1.3. Synthesis of [1,1,1,3,3,3- H ]-[2- H ]isopropyl
4
triphenylphosphonium bromide (Pyun et al., 1993). A
6
1
The internal standard d -(3R)-linalool 19a was pre-
6
mixture of 5.54 g (21.1 mmol) of triphenylphosphine
pared as described for the synthesis of d -geraniol 5.
6

302
F. Luan et al. / Phytochemistry 66 (2005) 295–303
(
1
3R)-Linalyl acetate ((3R)-3-acetoxy-3,7-dimethylocta-
,6-diene), synthesised from (3R)-linalool according to
Kreck, M., Scharrer, A., Bilke, S., Mosandl, A., 2002. Stir bar sorptive
extraction (SBSE)-enantio-MDGC-MS – a rapid method for
enantioselective analysis of chiral flavour compounds in strawber-
ries. Eur. Food Res. Technol. 213, 389–394.
the method of Vidari et al. (1999), was used as starting
material.
Kreis, P., Mosandl, A., 1994. Chiral compounds of essential oils. Part
XVII. Simultaneous stereoananlysis of Cymbopogon oil constitu-
ents. Flavour Fragr. J. 9, 257–260.
Luan, F., W u¨ st, M., 2002. Differential incorporation of 1-deoxy-d-
xylulose into (3S)-linalool and geraniol in grape berry exocarp and
mesocarp. Phytochemistry 60, 451–459.
Acknowledgements
The authors gratefully thank Mathias Bayer, Institut
f u¨ r Lebensmittelchemie, for providing the GC-columns,
G u¨ nter Maier, Novo Nordisk for supplying Glucanex ,
Luan, F., Hampel, D., Olszewski, K., W u¨ st, M., 2003. Mevalonate-
independent biosynthesis of monoterpenoids in grape berry
exocarp and mesocarp and in grape leaves. In: Le Que
´r e´ , J.-L.,
ꢂ
Eti e´ vant, P.X. (Eds.), Flavour Research at the Dawn of the
Twenty-first Century. Proceedings of the 10th Weurman Flavour
Research Symposium. Lavoisier, Paris, pp. 332–337.
and Ernst R u¨ hl, Department for Grapevine Breeding
and Grafting, Geisenheim, for making grapewines
available.
Financial support by the Deutsche Forschungsge-
meinschaft (DFG) and the Fonds der Chemischen
Industrie is gratefully acknowledged.
Luan, F., Hampel, D., Mosandl, A., W u¨ st, M., 2004. Enantioselective
ananlysis of free and glycosidically bound monoterpene polyols in
Vitis vinifera L. cv. Morio Muscat and Muscat Ottonel – evidence
for an oxidative monoterpene metabolism in grapes. J. Agric. Food
Chem. 52, 2036–2041.
Manning Jr., P.C., 1979. Stereoselective synthesis of the proposed
American coneflower juwenile hormone mimic. Some observations
on the cyclopropylcarbinyl rearrangement in substituted systems. J.
Org. Chem. 44, 2461–2468.
References
Ambid, C., Moisseeff, M., Fallot, J., 1983. Interconversion des
ald e´ hydes et des alcohols monoterp e` niques par des cellules de
raisin Muscat cultiv e´ s in vitro. Physiol. Veg. 21, 87–92.
Martin, D.M., Bohlmann, J., 2004. Identification of Vitis vinifera (ꢀ)-
a-terpineol synthase by in silico screening of full-length cDNA
ESTs and functional characterization of recombinant terpene
synthase. Phytochemistry 65, 1223–1229.
Baltussen, E., Sandra, P., David, F., Cramers, C., 1999. Stir bar
sorptive extraction (SBSE), a novel extraction technique for
aqueous samples: theory and principles. J. Microcolumn Sep. 11,
Mateo, J.J., Gentilini, N., Huerta, T., Jim e´ nez, M., Stefano, R.D.,
1997. Fractionation of glycoside precursors of aroma in grapes and
wine. J. Chromatogr. A 778, 219–224.
7
37–747.
Banthorpe, D.V., Poots, I., 1979. Redox interconversion of geraniol
and nerol in Rosa damascena. Phytochemistry 18, 1297–1299.
Banthorpe, D.V., Long, D.R.S., Pink, C.R., 1983. Biosynthesis of
geraniol and related monoterpenes in Pelargonium graveolens.
Phytochemistry 22, 2459–2463.
Matissek, R., Schnepel, F., Steiner, G., 1992. Lebensmittelanalytik, 2.
Auflage. Springer Verlag, Berlin, p. 128.
Matucha, M., Jockisch, W., Verner, P., Anders, G., 1991. Isotope
effect in gas-liquid chromatography of labelled compounds. J.
Chromatogr. 588, 251–258.
Bravdo, B., Shoseyov, O., Ikan, R., Altman, A., 1990. Monoterpene
glycoside biosynthesis in detached grape berries grown in vitro.
Physiol. Plant. 78, 93–99.
Naves, Y.R., Lamparsky, D., Ochsner, P., 1961. E´ tudes sur les
mati e´ res v e´ g e´ tales volatiles CLXXIV. (I) Pr e´ sence de t e´ trahydro-
pyrannes dans thuile essentiele de g e´ ranium. Bull. Soc. Chim. Fr.,
645–647.
Dunphy, P.J., Allcock, C., 1972. Isolation and properties of a
monoterpene reductase from rose petals. Phytochemistry 11,
Pyun, H., Coates, R.M., Wagschal, K.C., McGeady, P., Croteau,
R.B., 1993. Regiospecificity and isotope effects associated with
the methyl-methylene eliminations in the enzyme-catalyzed
biosynthesis of (R)- and (S)-limonene. J. Org. Chem. 58,
3998–4009.
1887–1891.
Ebang-Oke, J.-P., de Billerbeck, G.M., Ambid, C., 2003. Temporal
expression of the Lis gene from Vitis vinifera L., cv. Muscat de
Frontignan. In: Le Qu e´ r e´ , J.-L., Eti e´ vant, P.X. (Eds.), Flavour
Research at the Dawn of the Twenty-first Century. Proceedings of
the 10th Weurman Flavour Research Symposium. Lavoisier, Paris,
pp. 321–325.
Rapp, A., Mandery, H., Ullmeyer, H., 1984. Neue Monoterpendiole in
Traubenmosten und Weinen und ihre Bedeutung f u¨ r die Genese
einiger cyclischer Monoterpenether. Vitis 23, 84–92.
Rapp, A., 1992. Aromastoffe des Weines. Chem. Unserer Zeit 26, 273–
284.
Fagerlund, U.H.M., Idler, D.R., 1957. Marine sterols. IV. 24-
Dehydrolcholesterol: Isolation from a barnacle and synthesis by
the wittig reaction. J. Am. Chem. Soc. 79, 6473–6475.
Shine, W.E., Loomis, W.D., 1974. Isomerization of geraniol and
geranyl phosphate by enzymes from carrot and peppermint.
Phytochemistry 13, 2095–2101.
Falk, K.L., Gershenzon, J., Croteau, R., 1990. Metabolism of
monoterpenes in cell cultures of common sage (Savia officinalis).
Biochemical rationale for the lack of monoterpene accumulation.
Plant Physiol. 93, 1559–1567.
Strauss, C.R., Wilson, B., Gooley, P.R., Williams, P.J., 1986. Role of
monoterpenes in grape and wine flavor. In: Parliment, T., Croteau,
R. (Eds.), Biogeneration of Aromas. American Chemical Society,
Washington, DC, pp. 222–242.
Gramatica, P., Manitto, P., Razi, B.M., Delbianco, A., Francavilla,
M., 1982. Stereospecific reduction of geraniol to (R)-(+)-citronellol
by Saccharomyces cerevisiae. Experientia 38, 775–776.
Streitwieser, A., Heathcock, C.H., Kosower, E.M., 1994. Organische
Chemie, 2. Auflage. VCH Verlagsgesellschaft mbH, Weinheim, pp.
538.
Guth, H., 1997a. Identification of character impact odorants of
different white wine varieties. J. Agric. Food Chem. 45, 3022–3026.
Guth, H., 1997b. Quantitation and sensory studies of character impact
odorants of different white wine varieties. J. Agric. Food Chem. 45,
von Sydow, E., Karlsson, G., 1971. The aroma of black currants. IV.
The influence of heat measured by instrumental methods. Lebensm.
Wiss. Technol. 4, 54–58.
3
027–3032.
Guth, H., 1998. Comparison of different white wine varieties in odor
profiles by instrumental analysis and sensory studies. ACS Symp.
Ser. 714, 39–52.
Vidari, G., Giori, A.D., Zanoni, G., Bicchi, C., 1999. Enantioselective
synthesis of each stereoisomer of the pyranoid linalool oxides: the
linalool route. Tetrahedron: Asymmetry 10, 3547–3557.

F. Luan et al. / Phytochemistry 66 (2005) 295–303
303
Watanabe, N., Washio, H., Straubinger, M., Knapp, H., Winterhalter,
P., 1998. Occurence of a glucosidic progenitor of rose oxide in rose
flowers, Rosa damascena. Mill. Nat. Prod. Lett. 12, 5–10.
Williams, P.J., Strauss, C.R., Wilson, B., Dimitriadis, E., 1985.
Origins of some volatile monoterpenes and nor-isoprenoids in
grapes and wines – biosynthetic and biogenetic considerations. In:
Berger, R., Nitz, S., Schreier, P. (Eds.), Topics in Flavour
Research. H. Eichborn Verlag, Marzling-Hangenham, Germany,
pp. 336–352.
Wilson, B., Strauss, C.R., Williams, P.J., 1984. Changes in free and
glycosidically bound monoterpenes in developing Muscat grapes. J.
Agric. Food Chem. 32, 919–924.
W u¨ st, M., Beck, T., Dietrich, A., Mosandl, A., 1996. On the biogenesis
0
of rose oxide in Pelargonium graveolens L H e´ ritier and Pelargo-
nium radens H.E. Moore. Enantiomer 1, 167–176.
W u¨ st, M., Rexroth, A., Beck, T., Mosandl, A., 1998a. Mechanistic
0
aspects of the biogenesis of rose oxide in Pelargonium graveolens L
H e´ ritier. Chirality 10, 229–237.
Williams, P.J., Strauss, C.R., Wilson, B., 1980. New linalool deriva-
tives in Muscat of Alexandria grapes and wines. Phytochemistry
W u¨ st, M., Fuchs, S., Rexroth, A., Beck, T., Mosandl, A., 1998b.
Fragmentation mechanism of rose oxide in electron impact mass
spectrometry. Eur. Mass Spectrom. 4, 163–166.
19, 1137–1139.