Actinidia arguta: Volatile compounds in fruit and flowers

Article abstract of DOI:10.1016/S0031-9422(03)00142-0

More than 240 compounds were detected when the volatile components of the flowers and the fruit from several Actinidia arguta genotypes were investigated. Around 60-70 different compounds were extracted from individual tissues of each genotype. Two different methods of volatile sampling (headspace and solvent) favoured different classes of compounds, dependent upon their volatilities and solubilities in the flower or fruit matrices. The compounds extracted from flowers largely comprised linalool derivatives including the lilac aldehydes (12a-d) and alcohols (13a-d), 2,6-dimethyl-6-hydroxyocta-2,7-dienal (8), 8-hydroxylinalool (9), sesquiterpenes, and benzene compounds that are presumed metabolites of phenylalanine and tyrosine. Extracts of fruit samples contained some monoterpenes, but were dominated by esters such as ethyl butanoate, hcxanoate, 2-methylbutanoate and 2-methylpropanoate, and by the aldehydes hexanal and hex-E2-enal. A number of unidentified compounds were also detected, including 8 from flowers that are so closely related that they are either isomers of one compound or two or more closely related compounds. This is the first report of the presence of a range of linalool derivatives in Actinida.

Full text of DOI:10.1016/S0031-9422(03)00142-0

Phytochemistry 63 (2003) 285–301
www.elsevier.com/locate/phytochem
Actinidia arguta: volatile compounds in fruit and flowers
Adam J. Maticha,*, Harry Youngb, John M. Allena, Mindy Y. Wangb, Simon Fieldera,
Mark A. McNeilageb, Elspeth A. MacRaeb
aThe Horticultural and Food Research Institute of New Zealand Ltd, Private Bag 11030 Palmerston North, New Zealand
bThe Horticultural and Food Research Institute of New Zealand Ltd, Private Bag 92169 Auckland, New Zealand
Received 6 November 2002; received in revised form 21 February 2003
Abstract
More than 240 compounds were detected when the volatile components of the flowers and the fruit from several Actinidia arguta
genotypes were investigated. Around 60–70 different compounds were extracted from individual tissues of each genotype. Two
different methods of volatile sampling (headspace and solvent) favoured different classes of compounds, dependent upon their
volatilities and solubilities in the flower or fruit matrices. The compounds extracted from flowers largely comprised linalool deri-
vatives including the lilac aldehydes (12a–d) and alcohols (13a–d), 2,6-dimethyl-6-hydroxyocta-2,7-dienal (8), 8-hydroxylinalool (9),
sesquiterpenes, and benzene compounds that are presumed metabolites of phenylalanine and tyrosine. Extracts of fruit samples
contained some monoterpenes, but were dominated by esters such as ethyl butanoate, hexanoate, 2-methylbutanoate and
2-methylpropanoate, and by the aldehydes hexanal and hex-E2-enal. A number of unidentified compounds were also detected,
including 8 from flowers that are so closely related that they are either isomers of one compound or two or more closely related
compounds. This is the first report of the presence of a range of linalool derivatives in Actinidia.
# 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Actinidia arguta; Actinidiaceae; Baby kiwifruit; Headspace; Solvent; GC–MS; Flowers; Fruit; Aroma compounds; Lilac alcohols; Lilac
aldehydes; Terpenes; Esters; Aldehydes
1. Introduction
shown to be important (Perera et al., 1998; Young and
Paterson, 1990, 1995). In artificial systems, increased
levels of hex-E2-enal and hexanal enhanced the per-
ceived intensity of kiwifruit flavour, while increased
levels of ethyl butanoate enhanced the perception of
characteristic kiwifruit flavour (Gilbert et al., 1996).
Terpenes are present at very low levels, if at all in
headspace analyses, although some are present as gly-
cosides which can be released by enzymatic hydrolysis
(Young and Paterson, 1995).
A. arguta fruit have been described as having banana,
floral, fruit candy, grassy, green, and melon odours, and
blackcurrant, fruit candy, grassy, green, melon, stalky/
woody and tropical flavours (unpublished data). Some
A. arguta flowers sampled in the orchard were described
as having sweet, aromatic aromas with magnolia,
vanilla, tea-rose and balsam/sandalwood notes (unpub-
lished observations), while others were described as
containing strong carnation, honey or citrus-lime aro-
mas. Fresh ‘Hayward’ flowers were described as tea-
rose, sweet, cider, and hawthorn-like.
Actinidia arguta (Sieb. Et Zucc.) Planch. ex Miq. var.
arguta is a smooth-skinned grape-sized kiwifruit native
to northern China, Korea, Siberia and Japan. It is
commercially available in New Zealand as well as sev-
eral other fruit-producing countries. The fruit and
flowers of A. arguta have aroma notes distinct from
those of the major commercial cultivar (A. deliciosa [A.
Chev] C.F. Liang et A.R. Ferguson var. deliciosa ‘Hay-
ward’). Several studies have been performed on the
volatile aroma and flavour compounds of kiwifruit fruit
(Bartley and Schwede, 1989; Paterson et al., 1991;
Young and Paterson, 1990, 1995; Young et al., 1983,
1992), and one on the flowers (Tatsuka et al., 1990).
There is a paucity of studies on A. arguta. In ‘Hayward’
fruit, 80–90 compounds have been identified, with ca. 15
* Corresponding author. Tel.: +646-3568080x7778; fax: +646-
3517004.
E-mail address: amatich@hortresearch.co.nz (A.J. Matich).
0031-9422/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0031-9422(03)00142-0

286
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
This work is an investigation of the volatile com-
eight unidentified (possibly terpenoid) compounds were
also isolated using solvent extraction only. Their mass
spectra indicate they may comprise one or more groups
of diastereoisomers.
pounds present in the flowers and fruit of some A.
arguta genotypes to determine what makes them distinct
from ‘Hayward’ kiwifruit. The data also suggest tissue
specific terpene biosynthetic pathways are operating.
2.2.1. Linalool derivatives
The main groups of compounds in the flowers were
four diastereoisomers each of lilac aldehyde (12a–d)
and lilac alcohol (13a–d), which were present in nearly
all of the solvent and headspace extracts (Table 1).
There are eight possible isomers (four pairs of enan-
tiomers) of each of these compounds, but only four
can be resolved on the GC columns used in this study.
The amount of lilac alcohol in each sample ranged
from 11% (A1) to 57.5% (A7) of the compounds
extracted by solvent, with an average value of 32%.
The lilac aldehyde content ranged from 0.16% (A1) to
4.7% (A3) of the compounds present, with an average
content of 1.4%. Alcohol levels were generally lower in
the headspace (12–16%) than in the solvent extracts, while
aldehyde levels were higher in the headspace (4–14%).
The lilac compounds have not been reported in pre-
vious headspace or solvent extraction studies of ‘Hay-
ward’ flowers (Tatsuka et al., 1990) or fruit (Bartley and
Schwede, 1989; Jordan et al., 2002; Takeoka et al., 1986;
Young and Paterson, 1990), and we have also not found
them in a range of A. deliciosa flowers or fruit, including
‘Hayward’ (unpublished data). Lilac alcohol (13) is
proposed to arise via reduction of lilac aldehyde (12),
which may be biosynthetic derivatives (Fig. 1) of ger-
anyl diphosphate (1) via linalyl diphosphate (2) and
linalool (3) (Pichersky et al., 1994; Tollsten and Berg-
strom, 1993). Only small concentrations (<1%) of
linalool were found in six of the solvent extracts, but
linalool was a major component in the headspace of
three out of the four headspace extracts (up to 27%,
Table 1). Linalool was a minor constituent of an SDE
extract of ‘Hayward’ flowers, but was absent from the
headspace extract (Tatsuka et al., 1990).
2. Results and discussion
2.1. Different compounds are identified with solvent or
headspace extraction depending on the tissue being
extracted
A feature of the data in Tables 1 and 2 is the differ-
ence between the relative amounts of compounds iden-
tified in the headspace and the solvent extracts. The
more volatile, lower molecular weight, compounds were
more poorly represented in the solvent extracts. This
occurred even for compounds that constituted a large
proportion of the aroma profile. 3-methylbutanal was
present in all headspace extracts (at 20% of total in A4),
but was not detected in the solvent extracts of petals of
the same flowers.
The solvent extracts displayed a bias towards the
higher molecular weight, non-volatile, and wax soluble
fatty acid esters (oleates and linoleates) and hydrocarbons
(C15–C20). Compounds such as 8-hydroxylinalool, 2-(4-
methoxyphenyl)ethanol,
2-(4-hydroxyphenyl)ethanol
and squalene were all substantial components of the
solvent extract of petals but were not observed in the
headspace extracts (Table 1).
The compounds collected by the two sampling meth-
ods also reflected their relative solubilities in the sample
matrix. While 2- and 3-methylbutanal were detected in
the headspace of many of the flower samples and not in
the solvent extract, these compounds were not seen in
the headspace of the fruit samples but were detected in
the solvent extracts (Tables 1 and 2). Because of the
polar nature of these aldehydes their air/water (fruit)
partition coefficients might be lower than their air/wax
(flowers) partition coefficients. Headspace sampling
identified which compounds contributed to the aroma
of these flowers and fruits. Solvent extraction identified
the less-volatile or more matrix-soluble flavour or
aroma compounds and possible biosynthetic precursors
of some of the volatile compounds.
2,6-dimethyl-6-hydroxyocta-2,7-dienal (8) is an inter-
mediate in the chemical synthesis of lilac aldehyde
[Fig. 1, (12)] from linalool (3) (Wilkins et al., 1993), and
was identified in three of the A. arguta solvent extracts
at concentrations of around 10%. It was not observed
in the headspace. The first reported natural source of
this compound was in a solvent extract of apricot flow-
ers (Watanabe et al., 1974). Two other compounds in
apricot flowers that were identified in the A. arguta
extracts were the Z- and E- isomers of 8-hydroxylinalool1
2.2. Flowers
Sixty to seventy compounds were identified from
flowers sampled during petal expansion (Table 1). These
could be separated into a number of major groups; the
major contributors being terpenes and benzene com-
pounds. A number of linalool derivatives have been
recorded for the first time in Actinidia, and they were
significant components of some extracts. A group of
1
The mass spectral library identifies this compound as 1-hydroxy-
linalool. However, this is an unrealistic structure and is most likely a
mis-naming of 8-hydroxylinalool (9). In the present work, the LiAlH4
reduction of 2,6-dimethyl-6-hydroxy-2,7-octadienal (8) produced two
compounds that the mass spectral library identified as 1-hydroxy-
linalool. Reduction of this aldehyde would be expected to produce
8-hydroxylinalool (2,6-dimethyl-2,7-octadien-1,6-diol).

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
287
Table 1
A. arguta flowers—percentage of each compound as a fraction of the total compounds detected in each sample. Headspace extracts were not carried
out for A2, 3, 5, 6, and 7 (all female)
Solvent extract (headspace extract)
Male
A8
Female
A1
Component
A9
A4
A2
A3
A5
A6
A7
Terpenes
camphor
–
–
–
–
0(1.56)
0(0.02)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
b-caryophyllene
2,6-dimethyl-6-hydroxyocta-2,7-dienal
2,6-dimethylocta-3,7-diene-2,6-diol
eucalyptol
–
9.84(0)
9.42(0)
–
–
8.72(0)
–
–
–
–
–
–
–
0(0.31)
0.3
–
–
0(0.10)
–
E,E-a-farnesene
Z,E-farnesol
–
–
0.89
–
–
0.15(0)
–
–
0.30(0)
–
–
E,E-farnesyl acetate
geranylacetone
germacrene D
–
0(0.37)
0(0.48)
1.41
0.65
–
–
0.22(0.55)
–
0(0.65)
–
0(0.16)
–
1.23
0.76
0.74
0.4
–
–
–
–
–
–
–
hexahydrofarnesylacetone
E-8-hydroxylinalool
Z-8-hydroxylinalool
lilac alcohol a (13a)b
lilac alcohol b (13b)
lilac alcohol c (13c)
lilac alcohol d (13d)
lilac alcohol formate 1
lilac alcohol formate 2
lilac aldehyde 1 (12a)b
lilac aldehyde 2 (12b)
lilac aldehyde 3 (12c)
lilac aldehyde 4 (12d)
limonene
–
–
–
–
–
–
0.87
–
–
21.87(0)
0.91(0)
15.85(7.6)
2.55(2.03)
3.21(0.25)
3.10(2.02)
0(0.78)
0(0.10)
0.59(2.88)
0.11(2.69)
0.10(2.15)
0.19(1.02)
–
8.65(0)
–
7.22(0)
0.30(0)
21.4(1.51)
5.72(0.72)
4.5(0.06)
4.81(0.62)
–
36.2
13.24 15.2
0.7
–
2.51
–
0.65
16.24
4.52
6.17
3.48
–
1.7
20.9(10.4)
3.28(3.18)
2.39(0.66)
0.89(1.97)
–
8.3(11.3)
10.64 23.8
3.46 9.1
6.32 10.63
7.02 26
6.68 7.6
7.84 10.4
1.04(0)
1.21(0.97)
0.52(4.0)
–
3.51
–
4.08
–
4.29
–
5.4
–
–
–
0.06(1.69)
–
–
–
–
–
–
1.09(7.31)
0.18(0)
0.15(5.19)
0.27(1.62)
–
0.68(1.65)
0.11(1.34)
0.13(1.29)
0.27(0.77)
–
1.91
0.62
1.21
0.96
–
0.24
0.08
0.08
0.08
0.01
–
0.57
0.19
0.16
0.25
–
0.9
0.15
0.3
0.4
–
0.3
0.1
0.1
0.1
–
0(0.65)
0.09(1.18)
0(0.47)
–
0(0.93)
linalool
0.33(24.1) 0(27.14)
0.08(10.6)
0(0.07)
–
1.05
–
0.08
–
0.16
–
0.07
–
cis-linalool oxide
trans-linalool oxide
6-methylhept-5-en-2-one
b-myrcene
ocimene
E-b-ocimene
phytol
0(0.07)
0(0.25)
0.15(0.86)
–
–
–
–
0.20(0.61)
–
0(0.17)
–
–
–
–
–
0.2(0.15)
0(0.14)
0(0.11)
0(1.68)
–
0.36
–
0.17
–
0.16
–
0.16
–
0.1
–
0(0.19)
0(0.18)
0(0.63)
0(0.25)
0(0.30)
0(2.53)
–
0(0.26)
–
–
–
–
–
–
– –
3.43 13.69
–
–
–
0.15(0)
–
–
–
–
–
–
6.74
–
7.01
–
–
a-pinene
b-pinene
sabinene
0(0.28)
0(0.14)
0(0.61)
11.80(0)
–
0(0.04)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
5.62(0)
–
–
–
–
14.8
–
squalene
terpinolene
3,7,11,15-tetramethyl hexadeca-6,10,14-trienol
6.56(0)
14.59(0)
–
4.92
–
9.8
–
8.5
–
–
–
0(0.28)
–
–
–
–
–
1.27
–
–
Benzenoid compounds
benzaldehyde
benzene
–
0(0.27)
0(0.11)
0(0.84)
–
0(1.91)
0(0.08)
0(0.80)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
benzyl alcohol
8.30(0.10)
–
–
–
–
–
–
–
0.16
–
1.11
1.74
–
–
benzyl benzoate
ethylbenzaldehyde
2-(4-hydroxyphenyl)ethanol
methoxybenzene
2-(4-methoxyphenyl)ethanol
methyl 4-methoxybenzoate
methyl salicylate
naphthalene
–
–
–
–
0(0.26)
–
–
2.71(0)
–
–
–
1.42(0)
–
13.40(0)
4.79(0)
–
12.56(0)
11.37(0)
0(0.12)
31.09(0)
–
–
–
2.85
–
–
0(0.08)
–
–
–
2.62(0)
–
5.47
–
–
18.27 13.2
0.12
0.2
0.16
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0.1
–
–
–
–
–
0(2.91)
0(0.11)
–
–
–
–
phenol
–
–
–
–
–
–
–
2-phenylethanal
2-phenylethanol
2-phenylethyl acetate
trimethylbenzene
–
–
8.18(10.5)
–
– –
3.81 10.5
0.13
4.48
–
–
3.26(13.4)
0(0.30)
–
8.58(17.2)
–
11.5(6.25)
0(0.10)
–
5.42 15.3
–
–
–
–
–
–
–
–
–
–
0(0.33)
–
(continued on next page)

288
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
Table 1 (continued)
Solvent extract (headspace extract)
Male
A8
Female
A1
Component
A9
A4
A2
–
A3
A5
A6
A7
Esters
ethyl acetate
0(0.05)
–
–
–
–
–
–
–
–
–
–
0(0.04)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
ethyl hexanoate
hex-Z3-enyl acetate
methyl acetate
3-methylbutyl acetate
–
–
–
–
0.03
0(0.32)
0(1.20)
0(0.18)
–
–
–
Aldehydes
acetaldehyde
decanal
heptanal
–
0(0.87)
0(0.64)
0(1.77)
–
0(0.13)
0(0.28)
0(1.74)
–
–
–
–
0.06
–
–
–
–
–
0.15(0)
–
0.40(0)
1.16(0)
0.8(0.21)
–
0.65
0.03
–
0.48
0.23
0.04
–
0.32
0.05
0.4
hexanal
–
0(1.11)
0(2.65)
0(0.32)
0(1.15)
0(5.83)
–
0.03
–
0.04
–
2-methylbutanal
3-methylbutanal
3-methylbut-2-enal
2-methylpropanal
nonanal
0(0.52)
0(4.25)
0(0.01)
–
0(1.83)
0(20.49)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0(0.18)
10.1(2.1)
0(0.18)
–
–
–
–
–
–
0(1.07)
–
6.18(1.4)
15.5(0.8)
0(0.05)
–
2.18
–
2.72
0.04
–
2.4
–
1.52
–
4.7
0.04
0.08
octanal
undecanal
0(0.45)
–
–
–
–
–
Ketones
acetone
butan-2-one
0(0.41)
0(1.83)
–
0(1.40)
0(2.47)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0(1.19)
–
0(2.20)
–
butane-2,3-dione
3-hydroxybutan-2-one
7,8-dihydro-b-ionone
b-ionone
0(0.10)
0(0.61)
–
–
–
–
–
–
–
–
–
–
0(0.99)
0(0.04)
0(0.08)
–
–
–
–
–
–
–
–
–
–
jasmone
0(0.33)
0(0.19)
0(0.09)
0(0.56)
0(0.07)
2-methylpentan-3-one
4-methylpentan-2-one
octan-3-one
0(0.07)
0(0.05)
0(0.15)
–
pentadecan-2-one
Alcohols
ethanol
0(0.45)
0(0.56)
0(0.54)
–
0(1.04)
0(0.65)
–
0(13.78)
0(0.12)
–
–
–
–
–
butanol
butan-2-ol
–
–
–
–
–
–
–
0(0.70)
0(0.08)
–
–
–
–
–
2-ethylhexanol
hexadecanol
hexanol
–
–
–
0.52(0)
–
–
–
–
–
–
0(0.12)
–
–
–
–
–
–
–
–
0(0.27)
0(0.06)
–
–
–
–
–
–
hex-Z3-enol
methanol
–
0(0.25)
–
0(0.39)
–
0(0.41)
–
–
–
–
–
0(0.62)
0(2.36)
–
–
–
–
–
1-methoxypropan-2-ol
2-methylbutanol
3-methylbutanol
3-methylbut-2-enol
3-methylbut-3-enol
2-methylbut-3-en-2-ol
2-methylpropanol
nonanol
–
–
–
0(4.7)
0(0.12)
0(20.2)
–
–
–
–
–
0(11.4)
–
–
–
0.21
–
–
–
0.04
0.44(20.8)
–
–
0.1
–
0.12
–
0.16
–
0(0.41)
0(0.54)
0(0.23)
0(0.53)
0(0.19)
0(0.01)
–
–
–
–
–
–
–
0(0.48)
–
0(0.50)
–
0(0.49)
–
–
–
–
–
–
–
0.04
–
–
–
0.04
0(0.17)
0(0.27)
0(0.17)
–
–
0(0.59)
0.03
–
0.03
–
0.06
–
–
–
pentanol
–
–
–
–
–
5.82
–
–
–
pentan-2-ol
pentan-3-ol
penten-3-ol
–
0(0.09)
0(0.27)
–
0(0.80)
0(0.49)
–
0.02
–
–
–
–
–
–
0(0.33)
–
–
–
–
0(0.13)
0(0.55)
–
–
–
–
–
–
propanol
octanol
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0.05
–
0.04
–
–
0.09
0.03
–
octan-4-ol
oct-1-en-3-ol
–
0(0.86)
–
0(1.24)
–
–
–
–
–
–
(continued on next page)

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
289
Table 1 (continued)
Solvent extract (headspace extract)
Male
A8
Female
A1
Component
A9
A4
A2
–
A3
A5
A6
A7
Acids
acetic acid
–
–
–
–
–
–
–
0(1.44)
0(2.87)
–
–
–
–
–
–
–
0.22
–
0.09
0.17
0.1
–
dodecanoic acid
heptanoic acid
hexanoic acid
3-methylbutanoic acid
nonanoic acid
octanoic acid
–
–
–
–
–
–
–
–
–
–
–
–
1.05
0.62
0.13
0.07
0.14
1.56
0.07
0.43
0.17
0.17
0.08
2.6
0.44
0.11
–
0.27
0.17
0.36
–
0.4
0.2
0.5
–
2.33
0.11
0.1
0.02
0.2
Hydrocarbons
eicosane
–
–
–
–
–
–
0.75(0)
–
1.05
–
0.42
–
–
0.3
–
heptacosane
heptadecane
hexacosane
–
–
–
–
–
–
–
–
0.34
2.75
–
–
–
–
–
–
1.39(0)
–
–
–
–
–
–
hexadecane
–
–
0(0.41)
–
0.09
–
–
–
–
hexa-1,4-diene
hexa-Z2,Z4-diene
3-methylcyclopentene
3-methylpenta-1,3-diene
nonacosane
nonadecane
nonane
–
–
0(6.28)
–
–
–
–
–
–
0(0.26)
0(0.24)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0.23(0)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0.48
2.09
–
0.24
–
–
–
0(0.11)
0(1.10)
–
–
–
octane
0(0.01)
–
–
–
–
pentacosane
pentadecane
tricosane
0.30(0)
0.60(0)
2.60(0)
1.65(0)
–
0.4
–
–
–
0.6
–
0(0.29)
–
–
–
–
–
–
–
–
2.59
0.06
1.37
–
1.26
–
2.4
Sulphur compounds
bis(1-methylethyl)disulphide
carbon disulphide
dimethyl disulphide
–
–
–
–
–
–
0(0.05)
0.03
–
0.05
–
–
–
–
0.07
–
0.06
–
0(0.01)
–
0(0.07)
0(0.88)
–
–
–
–
–
Nitrogen compounds
butanenitrile
methenamine
0(0.48)
–
–
0(0.80)
–
0(0.15)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0(0.64)
0(0.81)
–
0(0.32)
2-methylbutanenitrile
Furans
tetrahydrofuran
–
0(0.09)
–
–
–
–
–
–
–
Unidentified compoundsa
unknown 6
unknown 7
–
–
0(0.31)
0(0.37)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0(0.32)
–
–
–
–
–
unknown 25
unknown 27
unknown 33
unknown 34
unknown 47
unknown 59
unknown 93
unknown 95
unknown 100
unknown 106
unknown 156
unknown 157
unknown 158
unknown 159
unknown 204
–
–
–
–
–
–
–
–
0(0.65)
0(1.28)
–
–
0(11.5)
0(0.02)
–
0(0.23)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0(0.08)
–
–
0(1.78)
–
–
–
–
–
–
–
–
–
–
–
–
–
3.21(0)
–
–
–
–
–
–
–
0.30(0)
0.61(0)
2.89(0)
0.15(0)
0.91(0)
0.03(0)
0.08(0)
–
–
–
–
–
–
–
–
–
5.22(0)
2.60(0)
0.30(0)
–
–
–
–
2.09(0)
5.22(0)
0.07(0)
0.30(0)
–
0.17(0)
1.73(0)
0.09(0)
0.09(0)
–
0.08(0)
0.23(0)
0.03(0)
0.08(0)
–
0.19
1.22
1.79
–
0.2
–
1.59
4.49
0.28
–
0.2
–
0.1
–
1.8
–
0.47
– –
(continued on next page)
1.1

290
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
Table 1 (continued)
Solvent extract (headspace extract)
Male
A8
Female
A1
Component
A9
A4
A2
A3
1.51
A5
A6
A7
unknown 205
unknown 206
unknown 207
unknown 227
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
10.4
0.5
0.2
–
0.56
0.19
–
1.23
0.75
0.55
a
The numbering of these unknown compounds is based upon that in our in-house database of compounds in Actinidia species and their geno-
types. Retention Indices and mass spectral data are listed in Section 4.6.
b
The configurations of the lilac aldehydes and alcohols is discussed in the last two paragraphs of Section 4.4.
[2,6-dimethylocta-2,7-diene-1,6-diol, (9)]. This com-
pound is the reduction product of the above aldehyde
(8). It was present in the A. arguta solvent extracts at
concentrations of up to 36%, but it too was not detec-
ted in the headspace. Both (8) and (9) have not been
previously reported in ‘Hayward’ flowers (Tatsuka et
al., 1990) and we have not found them in a range of A.
deliciosa flowers or fruit (unpublished data).
Two headspace samples contained the furanoid form
of trans- and cis-linalool oxide (6) at <0.25% of the
headspace (Table 1). Linalool oxide is structurally simi-
lar to lilac alcohol (13), differing only in the placement
of the hydroxyl group on the adjacent carbon atom
(Fig. 1). Linalool oxide is thought to be formed from
linalool via 6,7-epoxylinalool (4) (Pichersky et al.,
1994). The other linalool-derived compounds were two
Fig. 1. Structures of compounds identified and of some biosynthetic intermediates. Only one enantiomer of each lilac alcohol and lilac aldehyde
diastereoisomer is shown. Assignment of the configurations of the lilac aldehydes and alcohols is discussed in the last two paragraphs of Section 4.4.

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
291
Table 2
A. arguta fruit—percentage of each compound as a component of the total compounds detected in each sample
Solvent extract (headspace extract)
Component
A1
A3
A5
A6
A7
Terpenes
camphene
–
7.30(0)
–
10.8(0)
–
8.52(0)
0(0.01)
36.6(0)
–
–
–
–
–
–
–
–
–
camphor
2-carene
0.21(0)
0.18(0)
0.28(0)
–
–
cis-carveol
carvone
p-cymene
–
–
–
–
–
–
–
0.41(0)
–
0.75(0)
–
0.93(0)
0(1.19)
2.24(0)
10.9(3.3)
1.4(1.4)
–
endo-5,5,6-trimethylnorbornan-2-one
eucalyptol
limonene
0.37(0)
0.02(0)
–
0.6(0.04)
0(0.03)
0.1(0.12)
–
(0.04)
–
(0.08)
linalool
0.25(0)
–
2.2(0.06)
–
p-mentha-1,3,8-triene
p-menth-1-en-4-ol
menthol
–
–
–
0.82(0)
–
–
1.03(0.6)
2.50(0)
–
–
–
0.09(0)
1.16(0)
–
–
(0.02)
6-methylhept-5-en-2-one
1-methyl-4-(1-methylethenyl)benzene
1-methyl-4-(1-methylethyl)-cyclohex-2-enol
b-myrcene
E-b-ocimene
Z-b-ocimene
b-phellandrene
a-pinene
b-pinene
sabinene
0.07(0)
–
–
0(0.37)
0(0.77)
0.36(0)
0(1.31)
0(0.10)
–
–
–
–
–
0(1.40)
0(0.04)
–
(0.06)
–
0(0.04)
0(0.52)
0(0.96)
–
–
–
–
–
–
–
–
0.1(0.12)
0(0.20)
0(1.42)
0(0.72)
1.61(0)
–
–
0(0.02)
–
0.27(0)
0.1(0.03)
–
0(0.72)
–
–
0.22(0)
–
–
–
squalene
a-terpinene
b-terpinene
g-terpinene
a-terpineol
terpinolene
0.29(0)
0(0.01)
0(0.07)
–
–
–
–
–
0(0.47)
0(0.88)
1.5(1.07)
0.6(0.11)
0.86(3.3)
–
–
–
–
–
1.21(0)
–
–
–
4.04(0.2)
–
0.4(1.21)
(1.29)
–
0(0.06)
Benzenoid compounds
benzyl alcohol
benzyl benzoate
0.70(0)
0.74(0)
0.88(0)
–
–
–
–
–
–
1.31(0)
–
–
–
–
–
–
–
–
–
–
dimethylbenzaldehyde
1,2-dimethylbenzene
ethylbenzaldehyde
ethyl benzoate
0(0.07)
0(0.09)
–
–
–
0(0.01)
0(0.04)
6.1(0.12)
0.88(0)
0(0.03)
–
–
–
5.4(0.04)
–
0(0.58)
–
0(0.60)
–
2-(4-hydroxyphenyl)ethanol
methyl benzoate
styrene
2.33(0)
0.16(0)
1.04(0.4)
–
0(0.19)
0.27(0)
Esters
butyl acetate
–
–
0(0.01)
–
–
dimethyl carbonate
ethyl but-E2-enoate
ethyl acetate
–
0.06(0)
0(0.02)
0(0.03)
0(1.93)
0(22.5)
1.2(0.01)
0.17(0)
–
–
–
(0.06)
–
0(12.35)
0(40.1)
–
0(0.56)
0(0.44)
(0.97)
(65.8)
ethyl butanoate
ethyl decanoate
7.6(2.25)
–
0(23.1)
0(0.04)
–
–
(0.23)
–
ethyl heptanoate
ethyl hexadecanoate
ethyl hexadec-9-enoate
ethyl hexanoate
–
4.89(0)
–
7.63(0)
–
–
3.51(0)
2.43(0)
–
–
0(0.96)
–
0(0.72)
–
(1.36)
29.8(2.0)
0(0.09)
0.2(0.05)
0(0.02)
0.74(0)
–
ethyl hexa-2,4-dienoate
ethyl hex-2-enoate
ethyl hex-3-enoate
ethyl linoleate
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
5.63(0)
7.97(0)
–
0.40(0)
ethyl linolenate
ethyl 2-methylbutanoate
ethyl 3-methylbutanoate
4.33(0)
–
–
–
0(6.52)
–
–
–
–
0(0.01)
(continued on next page)

292
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
Table 2 (continued)
Solvent extract (headspace extract)
Component
A1
A3
A5
A6
A7
ethyl 2-methylpropanoate
ethyl octanoate
–
3.80(0)
0(24.6)
–
–
–
–
–
–
–
–
–
14.5(0.04)
0.85(0)
–
–
ethyl oct-Z4-enoate
ethyl oleate
ethyl pentanoate
–
1.05(0)
–
–
1.5(0.03)
0.12(0)
–
0(0.04)
0(0.01)
–
0(0.11)
0(0.79)
–
(0.10)
(0.01)
ethyl propanoate
hexadecyl acetate
methyl acetate
0(0.05)
–
3.51(0)
–
–
(0.01)
–
0(0.02)
0(0.30)
0(0.04)
0(0.78)
–
methyl butanoate
1-methylethyl tetradecanoate
methyl hexadecanoate
methyl linoleate
–
0(0.67)
(1.60)
–
4.04(0.09)
8.39(0)
6.70(0)
0.65(0)
0.85(0)
3.99(0)
–
4.23(0)
–
3.23(0)
16.3(0)
19.9(0)
–
–
–
–
–
–
–
–
–
4.07(0)
–
–
methyl linolenate
methyl octadecanoate
methyl oleate
–
0.45(0)
3.62(0)
5.4(0)
17.3(0)
–
–
–
methyl prop-2-enoate
propyl butanoate
0(0.09)
–
–
(0.002)
–
–
Aldehydes
acetaldehyde
decanal
0(1.80)
0(0.04)
–
–
–
–
–
0(0.05)
(0.04)
–
0.50(0)
0.12(0)
–
0.57(0.01)
–
0.29(0)
1.24(0)
–
hepta-E2,E4-dienal
heptanal
–
0(0.01)
–
–
hept-Z2-enal
hexanal
0.17(0)
1.1(22.5)
1.13(58)
0.86(0)
1.9(1.30)
3.8(5.87)
–
(0.35)
0.14(0.26)
0.3(0.13)
0.96(1.1)
hex-E2-enal
hex-Z2-enal
hex-E3-enal
hex-Z3-enal
3-methylbutanal
2-methylpentenal
nona-E2,Z6-dienal
nonanal
0(2.89)
0(0.06)
(3.03)
(0.05)
0(0.76)
0(0.08)
0(0.29)
–
0(0.28)
–
0(0.02)
–
–
–
–
–
0.04(0)
–
0.06(0)
–
–
–
–
–
–
0(0.02)
–
–
–
–
–
–
–
–
–
0.13(0)
–
–
1.00(0)
0.4(0.03)
–
non-E2-enal
octanal
0.12(0)
0.19(0)
0.06(0)
–
–
–
–
–
–
–
–
–
(0.24)
oct-E2-enal
propanal
–
–
0(0.09)
Ketones
acetone
butan-2-one
0(0.06)
–
0(0.01)
0(0.35)
0(0.01)
–
0(0.02)
–
0(0.07)
(0.04)
(0.64)
–
–
–
cyclopentanone
3-hydroxybutan-2-one
4-hydroxy-4-methylpentan-2-one
4-methylpent-3-en-2-one
octan-2,3-dione
penten-3-one
–
0(0.13)
–
–
–
–
–
–
–
–
–
0.09(0)
0.34(0)
–
0.06(0)
0.95(0)
–
0.06(0)
0.25(0)
–
0(0.01)
–
–
–
–
–
0(0.01)
–
–
2.00(0)
pent-E3-en-2-one
0.13(0)
Alcohols
butanol
decanol
–
0.23(0)
0.53(0)
5.06(0)
–
0.94(0)
1.36(0)
7.51(0)
–
0.75(0)
7.84(0)
2.13(0)
8.86(0)
–
dodecanol
ethanol
heptanol
–
(22.74)
0(11.08)
–
0(34.71)
–
0(39.06)
–
0(48.96)
0.37(0)
0.76(0)
1.8(0.14)
–
hexadecanol
hexanol
0.82(0)
0.19(0)
0.07(0)
–
–
0.23(0.03)
–
0.2(0.03)
–
–
hex-E2-enol
hex-Z2-enol
hex-E3-enol
hex-Z3-enol
0(0.03)
–
0(0.03)
0(0.04)
0(0.03)
–
–
–
–
(0.02)
–
–
0.06(0)
0(0.01)
–
–
–
–
(continued on next page)

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
293
Table 2 (continued)
Solvent extract (headspace extract)
Component
A1
A3
A5
A6
A7
3-methylbutanol
2-methylpropanol
nonanol
0.10(0)
–
–
–
–
–
–
–
–
–
–
–
0.19(0)
0.97(0)
0(0.01)
–
–
0.41(0)
1.76(0)
–
–
octanol
oct-1-en-3-ol
pentanol
1.30(0)
–
2.48(0)
0.18(0)
1.90(0)
–
–
–
–
–
–
–
–
pent-E2-enol
penten-3-ol
propanol
0(0.01)
0(0.02)
0(0.11)
–
0(0.06)
–
–
0.57(0)
(0.03)
–
0.09(0)
Acids
acetic acid
butanoic acid
–
–
0(0.32)
0.30(0)
0(0.07)
–
0(0.17)
–
–
–
Hydrocarbons
heptadecane
hexadecane
0.39(0)
0.37(0)
0.51(0)
0.60(0)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
2,6-dimethyldecane
dodecane
–
0.28(0)
0.16(0)
–
–
0.81(0)
2-methylpenta-1,3-diene
2-methoxy-2-methylpropane
nonadecane
–
0(0.18)
–
(0.53)
–
0(0.10)
–
–
–
–
–
–
0.37(0)
–
–
1.11(0)
–
–
–
–
–
octadecane
pentadecane
0.44(0)
–
0.18(0)
–
tetradecane
tridecane
0.61(0)
–
Sulphur compounds
bis(1-methylethyl)disulphide
0.22(0)
–
–
1.01(0)
–
Furans
ethyl 2-furancarboxylate
2-furancarboxaldehyde
4-methoxy-2,5-dimethyl-3(2H)-furanone
2-methylfuran
0.29(0)
–
0.11(0)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
(0.31)
–
–
–
–
0(0.58)
0(0.02)
–
5-methyl-2-furfural
methyl 2-furoate
tetrahydrofuran
–
–
–
0.22
–
0(1.51)
Unidentified compoundsa
unknown 3
unknown 178
–
–
–
–
–
1.44(0)
0(0.19)
–
–
–
a
The numbering of these unknown compounds is based upon that in our in-house database of compounds in Actinidia species and their geno-
types. Retention Indices and mass spectral data are listed in Section 4.6.
isomers of lilac alcohol formate, which were found only
in the headspace of the flower samples. We have not
found these esters, or linalool oxide in the flowers of any
A. deliciosa genotypes (unpublished data). Both the
linalool oxide isomers were reported in the SDE
(simultaneous steam distillation solvent extraction)
extract of ‘Hayward’ flowers (Tatsuka et al., 1990), but
were not detected in the flower headspace. Linalool
oxide may have been an artefact of the steam distillation
process.
compound was detected at levels of 0.5% in Artemesia
pallens (davana) (Misra et al., 1991), while four isomers
of both compounds were identified at trace levels in the
flowers of apricots (Watanabe et al., 1974). In some
Platanthera (Orchidaceae) flowers four isomers were
present at up to 97% of the volatile production (Toll-
sten and Bergstrom, 1993). Three of the aldehyde iso-
mers were the major lilac compounds in Planthera
stricta (Patt et al., 1988), and up to three isomers each
of the alcohol and aldehyde were also major compo-
nents of the scent of Gaura longiflora (Kint et al., 1993),
and of some Silene species (Caryophyllaceae) (Jurgens
et al., 2002). These compounds have also been extracted
Varying numbers of lilac aldehyde and lilac alcohol
isomers were found together in the A. arguta flower
samples. Only one unidentified diastereoisomer of each

294
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
from Haze (Rhus succedanea) honey (Shimoda et al.,
1996) and from nodding thistle (Carduus nutans) honey
(Wilkins et al., 1993). 2,6-Dimethyl-6-hydroxyocta-2,7-
dienal (8) and the furanoid linalool oxide (6) have also
been reported from apricot flowers and nodding thistle
honey, suggesting these pathways may also operate in
those species.
2-(4-Hydroxyphenyl)ethanol (or tyrosol) has anti-
oxidant properties and is the major phenol in olive oil
(Giovannini et al., 1999). Tyrosol was present in five
samples at levels between 1.4 and 11.4%. 2-(4-Meth-
oxyphenyl)ethanol was in most of the samples, ranging
from 5.5 to 31.1%. These two compounds are probably
metabolites of tyrosine, the 4-hydroxy analogue of
phenylalanine. These compounds, and structural rela-
tives, were reported as odourants in the hot water
extract of the pleasant smelling wood-rotting fungus
Gloeophyllum odoratum (Rosecke and Konig, 2000).
Despite their abundance in the solvent extracts of
A. arguta flower samples, they were not detected in the
headspace.
2.2.2. Other terpenes
6-Methylhept-5-en-2-one was present in all bar one
(A1) of the solvent extracts at <0.5% and in all of the
headspace samples at <1%. This oxidation product of
a-farnesene (Filmer and Meigh, 1971) is a suspected
causal agent in superficial scald, a storage disorder of
long-term cool-stored apples (Mir et al., 1999). Ger-
anylacetone was detected in the solvent and the head-
space extracts of most of the flower samples. It is also
thought to be a terpene oxidation product, one possible
precursor being squalene (Fruekilde et al., 1998). Ger-
anylacetone is found in fruits and essential oils (Bauer et
al., 1997), and is an important component in the aroma
of fresh tomatoes (Hayase et al., 1984).
Flowers of the related A. arguta var. purpurea were
distinct from those of the other genotypes. Terpenes
that were either absent from or were minor components
of the flowers of the A. arguta var. arguta genotypes,
made up 60% of the purpurea headspace (unpublished
data). The major terpenes in this variety were germa-
2.2.4. Esters, aldehydes, ketones, alcohols, and acids
Many of the aldehydes were present in both the fruit
and the flower samples, although they predominated in
the fruit. The most prevalent aldehydes in the flower
samples were heptanal and nonanal. Heptanal was a
minor component but nonanal was consistently present
in the solvent extracts (up to 15.5%) and in the head-
space (up to 2.1%).
Straight-chain aldehydes are metabolites of fatty acids
such as oleic acid (Schmidt and Monroe, 1976). The
branched-chain 3-methylbutyl and 2-methylbutyl com-
pounds are metabolites of l-Leucine and l-isoleucine
degradation, respectively (Mosandl, 1992; Myers et al.,
1970; Rettinger et al., 1991; Rowan et al., 1996).
3-Methylbutanal was present in several of the headspace
samples (2.6–20.5%), but not in the solvent extracts. Its
reduction product, 3-methylbutanol, was detected in
several solvent extracts (0.04–0.44%), but in only one of
the headspace samples at 21%. Akin to 3-methylbuta-
nal, 2-methylbutanal was found only in the headspace,
but at much lower concentrations (0.5–1.8%) than
3-methylbutanal. 2-methylbutanol was also only found
in the headspace, but at quite high levels (4.7–20.2%).
crene
D (25.2%), d-cadinene (10.9%), a-copaene
(7.75%), g-muurolene (4.3%), b-caryophyllene (3.6%),
epi-bicyclosesqui-phellandrene (2.5%), a-muurolene
(2.2%), a-caryophyllene (1.9%), and 1,2,4a,5,6,8a-hex-
ahydro - 4,7 - dimethyl - 1 - (1 - methylethyl) - naphthalene
(1.5%). Germacrene D is a major component of the
aroma of some rose varieties (Guterman et al., 2002;
Vainstein et al., 2001) and Cyphomandra sciadostylis
flowers (Sazima et al., 1993), and b-caryophyllene is a
major odourant in carnations (Clery et al., 1999).
2.2.3. Benzenoid compounds
2.2.5. Unknown compounds
2-Phenylethanal was a small component (0.13%) of
one flower solvent extract (A5). The dominant benzene
compound was its reduction product, 2-phenylethanol
(3-18% in solvent extracts; 1–17% in headspace
extracts). Its ester, 2-phenylethyl acetate, was found in
the headspace of two flower samples at 0.1 and 0.3%.
2-phenylethanol and 2-phenylethyl acetate have both
been reported in ‘Hayward’ flowers (Tatsuka et al., 1990).
The above 2-phenylethyl compounds are probably
metabolites of phenylalanine via cinnamic acid and
phenylacetic acid (Lapadatescu et al., 2000; Silk et al.,
2001; Yaylayan and Keyhani, 1998). Benzoic acid, ben-
zyl alcohol, and therefore benzyl benzoate are also
thought to be metabolites of phenylalanine (Lapadatescu
et al., 2000). Benzoic acid based compounds are important
components of carnation perfume (Clery et al., 1999).
Eight other compounds labelled as unknowns 156-9
and 204-7 (Table 1) were variously present in eight of
the nine solvent extracted genotypes (0.03–10.4%).
They had very similar mass spectra, and therefore are
either isomers of the one compound or of a number of
very similar compounds. To be isomers of the one
compound, they must be diastereoisomers of a molecule
with four chiral centres (optical isomers not being
resolvable by this GC column). Their mass spectra were
most similar to that of epoxy-linalooloxide (3,7-dime-
thyl-1,2,6,7-diepoxyoctan-3-ol). However, the mass
spectra and GC retention times of an authentic sample
of epoxy-linalooloxide isomers differed from those of
these unknown compounds. The retention indices and
mass spectral data for these and other ‘unknown’ com-
pounds (Tables 1 and 2) are listed in Section 4.6.

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
295
2.2.6. Organoleptic properties
This compound didn’t dominate any of the aroma
profiles, but probably contributed to the A1 and A3
genotypes in which its OAVs were about 30% of that of
the major aroma contributors, 3-methylbutanal and
geranylacetone, respectively.
The aroma impacts of volatile compounds in these
samples are discussed in terms of their odour activity
values (OAV) (Rothe and Thomas, 1963). The OAV is
the concentration/threshold where the threshold is the
lowest concentration of volatile compound that is
detectable by humans (Mayol and Acree, 2001). Other
authors have used the terms ‘‘odor unit’’ (Guadagni et
al., 1966) and ‘‘aroma value’’ (Kollmannsberger and
Berger, 1992).
Lilac aldehyde (12) and lilac alcohol (13) have fresh,
floral fragrances (Wakayama and Namba, 1974;
Wakayama et al., 1973; Watanabe et al., 1974). We do
not have their aroma threshold values (TVs) and can’t
determine their contribution to the flowers’ aroma pro-
files. Apart from uncertainty about the lilac compounds,
the fresh-floral aroma of linalool probably prevailed in
many of the flowers. The concentrations measured by
headspace (16–1900 ng per litre) and solvent (105 and
4300 ng gꢀ1) extraction were well above the threshold
values of 0.6 ng lꢀ1 and 6 ng mlꢀ1, respectively (Rychlik
et al., 1998). The maximum OAV of 720 for the A2
genotype probably made this the most fragrant of the
flowers. Linalool also had the highest OAVs for the A9
(OAV=338), A8 (OAV=192), and A6 (OAV=34)
flowers.
2.3. Fruit
Between 40 and 80 individual compounds were detec-
ted in the extracts from five different fruit genotypes
(Table 2). These were primarily a mix of esters, alde-
hydes and alcohols, with varying levels of mono-
terpenes.
2.3.1. Terpenes
The fruit contained a smaller and different range of
terpenes than the flowers. The fruit of the A6 genotype
contained the highest (57% of the solvent and 17% of
the headspace extracts) and the A1 genotype fruit had
the lowest terpene content at (11% of the solvent and
0.9% of the headspace extracts). This genotype had the
largest number of compounds in common between its
flowers and fruit, both having small amounts of a- and
b-pinene, eucalyptol, linalool, squalene, and 5-methyl-
hept-5-en-2-one. Of the other genotypes only A3 had a
terpene (limonene) in both its flowers and fruit.
Although camphor was the major terpene (7–37%) sol-
vent extracted from the fruit, it was not detected in the
headspace. Camphor was previously reported in Actini-
dia fruit as a glycoside in ‘Hayward’ fruit (Young and
Paterson, 1995).
The headspace of fruit (but not the flowers) of A6 also
contained limonene (1.7%), myrcene (1.31%), a-pinene
(1.42%), and eucalyptol (3.3%). These terpenes were
present to a lesser extent in the fruits of the other geno-
types. Linalool (and not the various floral metabolic
products) was identified in the solvent extract of two
samples and as a very minor component (0.06%) in the
headspace of A3. Increased linalool alters the flavour
profile of transgenic fruit (Lewinsohn et al., 2001)
although the contribution of this compound to the sen-
sory properties of A. arguta fruit is likely to be over-
shadowed by that of the more abundant esters and
aldehydes.
The A4 and A9 flowers would have the second most
intense aromas. The A9 aroma was from linalool and
A4 was from the malty-smelling (Rychlik et al., 1998)
3-methylbutanal (OAV=360, 1620 ng lꢀ1). The A1
flowers were also dominated by 3-methylbutanal, but at
modest levels (OAV=28, 126 ng lꢀ1).
The main A5 flower odourant was 2-phenylethanal
with its honey-like or flowery aroma and an odour
threshold (in water) of 4 ng gꢀ1 (Rychlik et al., 1998).
At a concentration of 232 ng gꢀ1 its OAV was 58. The
mild rose (Bauer et al., 1997) or spicy, honey-like aroma
(Rychlik et al., 1998) of 2-phenylethanol dominated in
A7 flowers with a modest OAV of 26. Its high concen-
tration of 26000 ng gꢀ1 was offset by its high aroma
threshold of 1000 ng gꢀ1. In both these genotypes, ger-
anylacetone and linalool each had OAVs of about half
that of the dominant 2-phenyl- compounds and were
probably secondary contributors to the aroma. The
2-phenyl- compounds were present in A1, A4, A6, and A8
flowers as non-dominant contributors to their aromas.
The A3 flower aroma was jointly dominated by ger-
anylacetone [OAV=18.5, 1110 ng gꢀ1, TV=60 ppb
(Buttery et al., 1971)] and 2-phenylethanol (OAV=17.6).
Geranylacetone is used in perfumery for its fresh-green,
slightly penetrating, rose-like odour. The geranylacetone
OAV of 86.5 was five times greater in A2 flowers but was
overshadowed by the very high linalool concentration.
6-Methylhept-5-en-2-one has a sweet, fruity, green
odour with a waxy apple note that contributes to the
aroma of pureed ‘Hayward’ fruit (Jordan et al., 2002).
2.3.2. Benzenoid compounds
2-Phenylethanol and 2-phenylethanal, which impac-
ted upon the aroma profiles of a number of flower
samples, were not detected in either the solvent or
headspace extracts of the fruit samples. Similarly, of the
tyrosine derivatives found in the flowers only a small
amount of tyrosol (0.88%) was identified in the A3
fruit. Ethyl and methyl benzoate were not present in the
flowers, but were the major benzenoid compounds in
the fruit samples, particularly in the solvent extracts of
A1 and A3 (Table 2).

296
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
2.3.3. Esters
2.3.5. Organoleptic properties
Although esters had a minimal presence in the flow-
ers, ‘‘fruity’’ esters were the major aroma contributors
in A. arguta fruit. The most prominent compounds were
the high molecular weight esters such as methyl hex-
adecanoate, the oleates, linoleates, and linolenates (sol-
vent extracts only), with considerable contributions
from ethyl butanoate and ethyl hexanoate (Table 2).
The esters constituted 9% of the A6 fruit solvent
extract and 60–70% of the solvent extracts from the
other three fruit samples analysed by this method. The
headspace extracts had large percentages of the volatile
esters and little or none of the long-chain esters. Ethyl
butanoate dominated headspace extracts, with con-
centrations between 2% (A1) and 66% (A7). It was
only detected in the solvent extract of A1 fruit as a
major component at 7.6%. Ethyl butanoate is a major
odourant from ripe ‘Hayward’ fruit (Bartley and
Schwede, 1989; Gilbert et al., 1996; Jordan et al., 2002;
Young et al., 1983), particularly in fresh fruit (Paterson
et al., 1991). Esters increase as fruit ripen (Paterson et
al., 1991; Young and Paterson, 1985), and this is possi-
bly the reason for the difference between A1 and the
other fruits in this study, especially as the aldehydes
were dominant in A1, although all fruit were ripened to
similar softness.
The aroma of A3 fruit was probably the most intense
because of the very sweet, fruity odours of ethyl 2-
methylpropanoate (OAV=25ꢁ104, 38,234 ng lꢀ1) and
ethyl 2-methylbutanoate (OAV=7ꢁ104, 10,114 ng lꢀ1
)
(Rychlik et al., 1998). The next aroma contributor in
this fruit was the ‘‘fruity’’ ethyl butanoate which had an
equivalent headspace concentration (34840 ng lꢀ1), but
an aroma threshold approximately 75 times greater than
ethyl 2-methylpropanoate (Rychlik et al., 1998), thus
producing an OAV of ca. 0.3ꢁ104. All three of the
above compounds have been identified in ‘Hayward’
fruit, but ethyl butanoate was the only ester judged to
contribute to kiwifruit flavour (Young and Paterson,
1990).
In all other fruit samples, except A1, ethyl butanoate
was the main aroma compound with OAVs of ca. 3000.
The green, leaf-like smelling hex-Z3-enal was the main
headspace compound from A1 fruit (50 ng lꢀ1). With a
low aroma threshold of 0.22 ng lꢀ1 (Rychlik et al., 1998)
it had an OAV (230) over twice that of ethyl butanoate
(100). This compound has been reported in ‘Hayward’
fruit, but did not contribute to its aroma or flavour
(Young and Paterson, 1990). The headspace also con-
tained hexanal (OAV=99, 3960 ng lꢀ1), hex-E2-enal
(OAV=41, 10160 ng lꢀ1), and in the solvent extract,
nona-E2,Z6-dienal (OAV=59, 0.59 ng gꢀ1). These three
compounds have tallowy/leaf-like, apple-like/leafy/
green/fatty/unripe-fruit (concentration dependent), and
cucumber-like notes, respectively (Rychlik et al., 1998).
They would reinforce the green notes from hex-Z3-enal.
The headspace of A6 fruit contained myrcene,
limonene, a-pinene, and eucalyptol at above threshold
levels but with modest OAVs. It is probable that only
the mint-like aroma of eucalyptol contributed sig-
nificantly to the aroma profile of the fruit. Its OAV of
ca. 1670 [5020 ng lꢀ1, TV=3 ng lꢀ1 (Rychlik et al.,
1998)] was half that of ethyl butanoate. Myrcene
and eucalyptol have not previously been reported in
‘Hayward’ fruit.
Ethyl hexanoate was a major component (29.8%) in
the A3 fruit solvent extract. The headspace of this fruit
had very high concentrations of esters that were not
present in the other four fruit. Ethyl 2-methylpropano-
ate and ethyl 2-methylbutanoate were 24.7 and 6.5%,
respectively, of the headspace in this sample.
2.3.4. Aldehydes, ketones, alcohols, and acids
Hex-E2-enal and hexanal are important contributors
to kiwifruit flavour (Jordan et al., 2002), increasing
amounts resulting in greater perceived intensity of
kiwifruit flavour (Gilbert et al., 1996). Hex-E2-enal was
not found in the flower samples but was in the fruit at
0.96–3.8% for solvent extracts and 1–58% for head-
space extracts. Hexanal was a minor component in the
flowers, but ranged from 0.13% (A5) to 22.5% (A1) of
the fruit headspace and 0.14% (A3) to 1.9% (A6) of the
fruit solvent extract. These compounds increase with
tissue maceration in kiwifruit (Paterson et al., 1991) and
inhibitors of lipoxygenase severely suppress the forma-
tion of hex-E2-enal (Bartley and Schwede, 1989). How-
ever trapping of volatiles from intact A. arguta fruit also
reveals their presence (unpublished data).
Hex-Z3-enal was present in two headspace samples
(A1 and A6) at 0.29 and 0.02%, hex-E3-enal (0.08%) in
the A1 fruit headspace, and nona-E2,Z6-dienal (0.13%)
in the A1 solvent extract. Hexanal, hexenal, and nona-
E2,Z6-dienal are all products of lipoxygenase break-
down of long-chain fatty acids such as linolenic acid
(Matheis, 1995).
3. Concluding remarks
A number of compounds different to those found in
the fruit and flowers of commercial kiwifruit (A. deli-
ciosa ‘Hayward’) were extracted from Actinidia arguta.
They are likely to contribute to both the characteristic
flavour of the fruit and to the very differently perfumed
flowers. A striking difference between fruit and flowers
is the presence of many linalool derived compounds in
the flowers, but only linalool in the fruit. Sequencing of
an EST library for A. arguta petals (A1) has revealed a
strong presence of amino acid transporters; many have
homologues in Arabidopsis (unpublished data). This
parallels the presence of a number of aroma impact

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
297
compounds that derive from phenylalanine and tyrosine.
We also discovered a group of apparently highly-related
unknown compounds present in the flowers. Establish-
ment of their identity will define their impact on floral
aroma and the potential for discovery of new pathways.
pulp for 20 min at 23 ꢂC by the same method as used for
flower volatiles.
4.2.3. Analysis of headspace samples
Volatile compounds were thermally desorbed from
the headspace traps at 175 ꢂC and were cryo-focussed at
the beginning of the GC column (Young, 1981). The col-
umn outlet was split between the GC (Hewlett Packard
5890) FID detector (for quantitation) and VG-70SE (VG-
Micromass, Manchester, UK) mass spectrometer (for
component identification) with an electron impact ioni-
sation potential of 70 eV. Separations were carried out
in a 30 mꢁ0.32 mm i.d., 0.5 mm J & WDBWax capillary
4. Experimental
4.1. Plant material
In 2000 and 2001 flowers and fruit were obtained
from 9 genotypes of Actinidia arguta. The genotypes
sampled from were, A. arguta (Sieb. et Zucc.) Planch. et
Miq.var arguta designated as A1 (Hortgem Tahi)
(Female), A2 (Female), A3 (Female), A4 (Female), A5
(Female), A6 (Female), A7 (Female), A8 (Male), and
A9 (Male), all of which were obtained from the Hor-
tResearch orchard in Te Puke, New Zealand. Of these
genotypes, A5 and A6 were siblings, and A1 was the
offspring of A4 and A9.
ꢂ
ꢂ
column starting at 30 C for 6 min, increasing by 3 C
ꢂ
ꢂ
ꢂ
minꢀ1 to 102 C, 5 C minꢀ1 to 190 C, and held for 5
min. The carrier gas was He at 30 cm sꢀ1, and the FID
and mass spectrometer transfer line were at 220 ꢂC. Quan-
tification of compounds was carried out using an aver-
age detector response based on methyl butanoate, ethyl
butanoate, hexanol and methyl benzoate. Component
identification was by comparison with spectra in the Mass
Spectral Database (1998 NIST and an in-house data-
base), retention indices (in-house database) and in some
cases direct GC-MS comparison with authentic standards.
4.2. Headspace sampling
Headspace sampling and analysis was undertaken at the
Mount Albert Research Centre, Auckland, New Zealand.
4.3. Solvent extraction
4.2.1. Flower volatiles
All of the solvent extract analysis was undertaken at
the Palmerston North Research Centre, Palmerston
North, New Zealand.
Branches containing flowers were transported with
stems in water immediately after harvest to Auckland.
Whole flowers (2.5–5 g) that were 50–75% open, and in
good quality, were picked just under the receptacle and
placed into a 250 ml Quickfit1 Erlenmeyer flask to
which was fitted a headspace adapter with an air inlet
and outlet. A volatile trapping cartridge (s.s. tubing
packed with 100 mg of Chromosorb 105TM absorbent)
was fitted to the air outlet port of the adapter. The
closed system was allowed to equilibrate for 30 min at
room temp. (ca. 23 ꢂC), after which 1.5 l of clean air was
passed through the Erlenmeyer flask at 25 ml minꢀ1 and
was vented through the trapping cartridge. Traps were
4.3.1. Flowers
Approximately 1 g of flower petals were removed
from the flowers after their headspace was sampled. The
petals were rinsed in 2 ml of 50:50 pentane/Et2O (GPR,
BDH) which had been purified by distillation and by
passing through a column of activated alumina (Perrin
and Armarego, 1988). The volume of the solvent extract
was reduced to ca. 50 ml under a gentle stream of N2.
Samples were stored at ꢀ18 ꢂC.
ꢂ
stored at ꢀ15 C before GC-FID/MS analysis (storage
4.3.2. Fruit pulp
was less than 2 weeks).
Approximately 40 g of fruit was diced and placed in a
Waring Commercial Blendor with an equivalent volume
of absolute EtOH (Analar, 99.7–100%, BDH) and
macerated for 10–15 s. The EtOH was retrieved by fil-
tration of the pulp over 30 min at room temp. on a 9 cm
Buchner funnel using a coarse filter paper (Whatman
No. 4). Fine solids and high molecular weight waxes
were removed from the EtOH extract by vacuum dis-
tillation (1 Pa) at 40 ꢂC, with condensation into a
receiving flask cooled in liquid nitrogen. This was effec-
ted using a solvent assisted flavour evaporation (SAFE)
apparatus (Engel et al., 1999). The EtOH distillate
was rapidly stirred for 1 h with and equivalent volume of
pentane (purified as described in Section 4.3.1). Pentane
4.2.2. Fruit pulp
Each of 10 fruit [mixture of ripe stages with no unripe
or overripe fruit included (Hassall et al., 1998)] was cut
longitudinally in half and one half of each fruit was used
for headspace sampling. The other halves were frozen in
liquid N2 for solvent extraction. The 10 fruit halves were
combined by forcing them through a 2 mm sieve into an
ice-cooled beaker in which the pulp was thoroughly
mixed for 10 s. Approximately 1–1.5 g of this pulp was
weighed into a 50 ml Quickfit1 test tube containing a
magnetic stirrer bar, and fitted with a headspace adap-
ter. The fruit volatiles were sampled from the stirred

298
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
extracts were reduced to ca. 5 ml by rotary evaporation
and were dried by passing through a small column of
anhydrous MgSO4 before the volume was further
reduced to ca. 50 ml under a gentle stream of nitrogen.
4.4.2. Preparation of lilac aldehydes (12)
NaH (0.19 g, 7.9 mmol) was washed with anhydrous
THF (2ꢁ5 ml) and dried under a stream of N2. A soln.
of
(E)-2,6-Dimethyl-6-hydroxyocta-2,7-dienal
(8)
(0.38g, 2.2 mmol) in MeOH (15 ml, ANALAR, BDH)
was added cautiously over 5 min and the reaction stir-
red for a further 30 min at room temp. The dis-
appearance of starting material was monitored by TLC
(hexanes/EtOAc 1:1) and products were visualised by
treatment with acidified ethanolic vanillin (Krebs et al.,
1969). The reaction mixture was diluted with Et2O (50
ml) and washed with H2O (50 ml). The aqueous layer
was separated and re-extracted with a further aliquot of
Et2O (50 ml). The combined organic phases were
washed with satd. brine (50 ml), dried over MgSO4, fil-
tered under reduced pressure and concentrated in vacuo
to give the crude product mixture as a yellow oil. Pur-
ification by short path distillation (0.1 mm Hg, Kugel-
rohr) gave the four lilac aldehydes (12) as a mobile oil
(0.33 g, 87%) that were characterised by mass spectro-
metry (Kint et al., 1993; Wakayama and Namba, 1974;
Wilkins et al., 1993), H1 NMR (270 MHz, CDCl3)
(Wilkins et al., 1993), and 13C NMR (270MHz, CDCl3):
ꢀ 9.24, 9.91, 10.31, 10.40 (b-CH3), 26.12, 26.52, 26.72,
27.04 (5-CH3), 28.60, 29.18, 29.34, 29.97 (C-3), 36.57,
36.85, 37.38, 37.59 (C-4), 50.66, 51.02, 51.68, 51.88 (C-
2), 78.26, 78.53, 79.06, 79.53 (CHCHO), 82.74, 83.03,
83.25, 83.29 (C-5), 111.35, 111.44, 111.44, 111.62
(CH¼CH2), 143.09, 143.22, 143.79, 144.00 (CH¼CH2),
204.01, 204.19, 204.25, 204.29 (CHO).
4.3.3. Analysis of solvent extracts
Chromatographic separations were in a 30 mꢁ0.25
mm i.d.ꢁ0.25 mm film thickness SP10-Wax (Supelco,
Belefonte) or ZBWax capillary column (Phenomenex,
Torrance, CA). The He head pressure was 52 kPa (7.5
ꢂ
psig), and the injection port was at 220 C. The temp.
ꢂ
ꢂ
program was 5 min at 40 C, increasing by 5 C minꢀ1
ꢂ
ꢂ
ꢂ
to 200 C, 15 C minꢀ1 to 240 C, and held for 20 min.
The GC and mass spectrometer were the same models
as used for the analyses of headspace samples.
Compounds were quantified by comparing their total
ion current intensity with that of ethyl butanoate for
esters, hexanol for alcohols, acids and aldehydes, and a-
pinene for terpenes and hydrocarbons. Percentage com-
positions were calculated excluding the solvent peaks.
Component identification was by comparison with
spectra in the Mass Spectral Database (NIST, 1992),
and in some cases direct GC (retention indices) and MS
comparison with authentic standards. A number of
unknown compounds are listed in Tables 1 and 2. The
numbering for these compounds is based upon that in
our in-house database of compounds in Actinidia spe-
cies and their genotypes.
4.4. Synthesis of lilac aldehydes and alcohols
Mixtures of four diastereoisomeric pairs of lilac alde-
hydes (12a–d) and alcohols (13a–d) were prepared by a
modification of the procedure of Wilkins et al. (1993).
4.4.3. Preparation of lilac alcohols (13)
NaBH4 (0.4 g, 10.5 mmol) was added to a soln. of
lilac aldehydes (12) (0.19 g, 1.13 mmol) in MeOH (7.5
ml) and the reaction monitored by TLC. After 30 min
the reaction mixture was quenched by the cautious
addition of satd. NH4Cl soln. (5 ml), diluted with H2O
(10 ml) and subjected to an ethereal work-up and short
path distillation, as described above, to give a mixture
of the four lilac alcohols (0.186 g, 1.09 mmol, 96%) that
were characterised by mass spectrometry (Kint et al.,
1993; Wilkins et al., 1993), H1 NMR (270 MHz, CDCl3)
(Wakayama et al., 1973), and 13C NMR (270 MHz,
CDCl3): ꢀ 12.02, 12.38, 13.41, 13.60 (b-CH3), 26.72,
26.84, 27.18, 27.33 (5-CH3), 30.63, 30.63, 31.06, 31.06 (C-
3), 36.35, 36.91, 37.24, 37.71 (C-4), 38.15, 38.72, 40.94,
41.04 (C-2), 66.02, 66.12, 67.88, 68.50 (CHCH2OH),
81.78,82.13, 85.27, 85.66 (CH2OH), 82.46, 82.82, 83.43,
83.57 (C-5), 111.28, 111.28, 111.43, 111.72(CH¼CH2),
143.20, 143.29, 143.88, 144.01 (CH¼CH2).
4.4.1. (E)-2,6-Dimethyl-6-hydroxyocta-2,7-dienal (8)
SeO2 (1.3 g, 11.7 mmol, 98%, Sigma) was added, in
one portion, to a soln. of linalool (2.0 g, 12.9 mmol,
97%, Aldrich) in dry dioxane (10 ml) under an atmo-
sphere of dry N2 at room temp. The mixture was stirred
under reflux for 90 min and the formation of aldehydic
compounds monitored by TLC (Merck 1.05554, hexanes/
Et2O 4:1) using an acidified ethanolic soln. of 2,4-DNP
(Krebs et al., 1969). The reaction mixture was cooled to
room temp., diluted with CH2Cl2 (50 ml) and washed
with H2O (50ml). The aqueous layer was separated and
re-extracted with CH2Cl2 (50 ml). The combined organic
phases were washed with satd. brine (50 ml), dried over
MgSO4, filtered under reduced pressure and concentrated
in vacuo to give a pale yellow oil that was purified by
flash chromatography on silica gel (95 g, Merck 1.09385)
eluting with hexanes/Et2O 4:1. The aldehydic fractions
were combined and subjected to short path distillation
(0.1 mm Hg, Kugelrohr) to give the title compound
(0.38 g, 19%) as a colourless oil. Data was consistent
with that reported earlier (Wilkins et al., 1993).
The retention times and mass spectra of these
authentic compounds were used to confirm their pre-
sence in the extracts of the A. arguta flowers. The con-
figurations of the aldehydes (or their respective
enantiomers) and their order of elution on polar
GC columns, are 1: (ꢁS,2S,5S), 2: (ꢁR,2S,5S), 3:

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
299
(ꢁR,2R,5S), and finally 4: (ꢁS,2R,5S) (Tollsten and
Bergstrom, 1993; Wakayama and Namba, 1974). The
alcohols eluted thus, b: (ꢁR,2S,5S), d:(ꢁS,2R,5S), a:
(ꢁS,2S,5S), and finally c:(ꢁR,2R,5S) (Tollsten and
Bergstrom, 1993; Wakayama et al., 1973). These
authors gave the configuration of one enantiomer of
each diastereoisomer, but acknowledged that they did
not know which enantiomer, or even whether both
enantiomers of each diastereoisomer were present.
The elution order of the lilac aldehydes was deter-
mined by comparing the GC retention times of the
alcohols produced by LiAlH4 reduction of both natural
and synthetic aldehydes (Wakayama and Namba,
1974). Tollsten and Bergstrom (1993) appear to have
interpreted these results to assign the configurations of
the aldehydes based upon those of their alcohols. How-
ever, according to the CIP (Cahn-Ingold-Prelog) rules
(March, 1985), reduction of the lilac aldehydes to their
alcohols will change the bS designation of the aldehyde
to the bR for the alcohol and visa versa for the bR. It is
unclear whether the stereochemical consequences of the
reduction of the aldehydes to the alcohols was con-
sidered by these workers.
Unknown 6: RI-wax=1263, m/z (rel. int.): 69 (100),
41 (55), 107.1 (13.1), 81 (12.6), 39 (11.1), 53 (9.2), 150.1
(9), 79 (7.4), 67 (7.2), 82.1 (7.1), 135.1 (6.3), 70 (5.6).
Unknown 7: RI-wax=1149, m/z (rel. int.): 55 (100),
67 (91), 43 (67), 123 (62), 68 (55), 39 (27), 41 (24), 81
(19), 53 (19), 82 (17), 32 (14.3), 96 (14.2).
Unknown 25: RI-wax=1101, m/z (rel. int.): 81 (100),
57 (76), 32 (53), 109 (44), 138 (31), 43 (27), 39 (21), 79
(20), 41 (20), 53 (19), 123 (18), 67 (16).
Unknown 27: RI-wax=1168, m/z (rel. int.): 67 (100),
123 (73.2), 68 (62.4), 96 (20.1), 81 (19.5), 95 (18), 82
(15.8), 71 (14.9), 79 (9.1), 111 (8.7), 109 (7.8), 93 (7.7).
Unknown 33: RI-wax=1317, m/z (rel. int.): 69 (100),
41 (54.3), 81 (16.8), 39 (10.2), 150 (9.7), 53 (8.4), 82
(8.3), 79 (7), 107 (6.2), 67 (6.2), 70 (5.5), 135 (3.9).
Unknown 34: RI-wax=1451, m/z (rel. int.): 119 (100),
134 (55.8), 117 (20.3), 91 (16), 133 (12), 120 (10.4), 132
(9.8), 77 (9), 115 (7.4), 105 (6.1), 39 (5.6), 103 (5.6).
Unknown 47: RI-wax=1328, m/z (rel. int.): 43 (100),
41 (44.1), 55 (31), 71 (25), 39 (19.9), 29 (14.6), 42 (12.3),
70 (11.4), 56 (5.8), 30 (5.1), 45 (4.3), 62 (4.2).
Unknown 59: RI-wax=1140, m/z (rel. int.): 67 (100),
55 (90.6), 43 (76.3), 123 (66), 68 (55.1), 39 (29.1), 41 (28),
53 (22.8), 81 (20.2), 96 (19), 82 (19), 54 (15.9).
4.5. Synthesis of epoxy-linalooloxide
Unknown 93: RI-wax=1930, m/z (rel. int.): 81 (100),
41 (87), 58 (77), 55 (73), 67 (67), 111 (67), 39 (56), 109
(55), 43 (54), 71 (51), 93 (47), 110 (33).
Linolool was epoxidised by a modification of the
method used to prepare a-farnesene epoxide (Fielder
and Rowan, 1994). Linalool (87 mg, 0.56 mmol, 97%,
Aldrich) was added, at room temp., to a stirred slurry
of 3-chloroperoxybenzoic acid (0.4 g, 2.2 mmol, 75%,
Pfaltz & Bauer Inc.) in freshly distilled CH2Cl2 (10
ml) and the disappearance of starting material mon-
itored by TLC (hexanes/EtOAc 1:1) as described
above. After 5 h the reaction mixture was filtered
under reduced pressure, washed with satd. NaHCO3
(15 ml), H2O (10 ml) and brine (10 ml), dried over
MgSO4, filtered and concentrated in vacuo to give a
mixture of products that was characterised by GC-
MS. Analysis of the product mixture identified the
major components as the cis- and trans-linaloloxides
(68%). It was assumed that these compounds were the
cis and trans isomers of 6,7-epoxylinalool (4). Other
major products were the cis and trans isomers of the
pyranoid form of linalool oxide (5) (27%), also known
as epoxylinalol, and four isomers of epoxy-linalool-
oxide (7) (3%). The mass spectra of (4), (5) and (7)
were fully consistent with spectra reported in the 1998
NIST database.
Unknown 95: RI-wax=2085, m/z (rel. int.): 81 (100),
67 (74), 82 (54), 41 (53), 55 (39), 95 (38), 109 (33), 69
(26), 68 (25), 71 (24), 39 (23), 43 (23).
Unknown 100: RI-wax=2520, m/z (rel. int.): 69 (100),
95 (72), 41 (68), 81 (68), 93 (38), 67 (37), 55 (35), 136
(25), 137 (19), 68 (18), 121 (18), 107 (16).
Unknown 106: RI-wax=2220, m/z (rel. int.): 71 (100),
43 (50), 55 (33), 41 (32), 69 (20), 56 (16), 81 (15), 31 (13),
68 (12), 72 (11), 39 (10), 67 (10).
Unknowns 204, 156, 205, 206, 157,158, 207,159: RI-
wax=2110, 2120, 2150, 2175, 2185, 2210, 2215, 2280,
m/z (rel. int.): 43 (100), 81 (38), 143 (33), 41 (28), 55 (23),
127 (18), 83 (15), 125 (14), 67 (14), 31 (13), 69 (12), 39 (12).
Unknown 178: RI-wax=1440, m/z (rel. int.): 108
(100), 41 (88), 69 (82), 109 (75), 95 (56), 67 (51), 152
(49), 137 (46), 81 (35), 93 (31), 39 (30), 55 (25).
Unknown 227: RI-wax=2287, m/z (rel. int.): 69 (100),
41 (38.7), 81 (29.5), 93 (19.2), 55 (14), 136 (14), 67 (10),
95 (9.8), 137 (8.3), 68 (8.2), 91 (7.9), 53 (7.7).
Acknowledgements
This research was funded by the NZ Foundation for
Research, Science and Technology Contract No C06X0006.
4.6. MS and RI Data for unknown compounds in
Tables 1 and 2
References
Unknown 3: RI-wax=1021, m/z (rel. int.): 93 (100),
91 (38), 77 (36), 92 (32), 41 (11), 39 (11), 136 (10), 79
(10), 94 (8), 83 (8), 53 (6), 105 (5).
Bartley, J.P., Schwede, A.M., 1989. Production of volatile compounds
in ripening kiwi fruit (Actinidia chinensis). J. Agric. Food Chem. 37,
1023–1025.

300
A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
Bauer, K., Garbe, D., Surburg, H., 1997. Common Fragrance and
Flavour Materials—Preparation, Properties and Uses, third ed.
Wiley-VCH, Weinheim, Germany.
March, J., 1985. Advanced Organic Chemistry, third ed. Wiley-Inter-
science, New York.
Matheis, G., 1995. Plant enzymes linked to flavor, Dragoco Report,
Flavoring information service. Dragoco Gerberding & Co GMBH.
Mayol, A.R., Acree, T.E., 2001. Advances in chromatography—
olfactometry. In: Leland, J.V., Schieberle, P., Buettner, A., Acree,
T.E. (Eds.), Gas Chromatography—Olfactometry, The State of
the Art, Vol. 782. American Chemical Society, Washington, pp.
1–10.
Buttery, B.G., Siefert, B.M., Guadagni, D.C., Ling, L.C., 1971. Char-
acterization of additional volatile components of tomato. J. Agric.
Food Chem. 19, 524–529.
Clery, R., Owen, N.E., Chambers, F.S., 1999. An investigation into
the scent of carnations. J. Essent. Oil Res. 11, 355–359.
Engel, W., Bahr, W., Schieberle, P., 1999. Solvent assisted flavour
evaporation - a new and versatile technique for careful and direct
isolation of aroma compounds from complex food matrices. Eur.
Food Res. Technol 209, 237–241.
Mir, N., Perez, R., Beaudry, R.M., 1999. A poststorage burst of 6-
methyl-5-hepten-2-one (MHO) may be related to superficial scald in
‘Cortland’ apples. J. Amer. Soc. Hort. Sci. 124 (2), 173–176.
Misra, L.N., Chandra, A., Thakur, R.S., 1991. Fragrant components
of oil from Artemisia pallens. Phytochemistry 30 (2), 549–552.
Mosandl, A., 1992. Capillary gas chromatography in quality assess-
ment of flavours and fragrances. J. Chromatogr. 624, 267–292.
Myers, M.J., Issenberg, P., Wick, E.L., 1970. L-Leucine as a Precursor
of Isoamyl alcohol and Isoamyl Acetate, Volatile Aroma Con-
stituents of Banana Fruit Discs. Phytochemistry 1693–1700.
Paterson, V.J., MacRae, E.A., Young, H., 1991. Relationships
between sensory properties and chemical composition of kiwifruit
(Actinidia deliciosa). J. Sci. Food Agric. 57, 235–251.
Fielder, S., Rowan, D.D., 1994. The synthesis of d6-a-farnesene. J.
Labelled Compd. Radiopharm 34 (11), 1075–1085.
Filmer, A.A.E., Meigh, D.F., 1971. Natural skin coating of the apple
and its influence on scald in storage. IV. oxidation products of
a-farnesene. J. Sci. Food Agric. 22 (4), 188–190.
Fruekilde, P., Hjorth, J., Jensen, N.R., Kotzias, D., Larsen, B.S.O.,
1998. Ozonolysis at vegetation surfaces: a source of acetone, 4-oxo-
pentanal, 6-methyl-5-hepten-2-one, and geranyl acetone in the tro-
posphere. Atmos. Environ. 32 (11), 1893–1902.
Gilbert, J.M., Young, H., Ball, R.D., Murray, S.H., 1996. Volatile
flavor compounds affecting consumer acceptability of kiwifruit. J.
Sens. Studies 11, 247–259.
Patt, J.M., Rhoadesa, D.F., Corkill, J.A., 1988. Analysis of the floral
fragrance of Platanthera stricta. Phytochemistry 27 (1), 91–95.
Perera, C.O., Young, H., Beever, D.J., 1998. Kiwifruit. In: Shaw, P.E.,
Chan Jr., H.T., Nagy, S. (Eds.), Tropical and subtropical fruits.
Agscience, Auburndale, pp. 336–385.
Giovannini, C., Straface, E., Modesti, D., Coni, E., Cantafora, A.,
De-Vincenzi, M., Malorni, W., Masella, R., 1999. Tyrosol, the
major olive oil biophenol, protects against oxidized-LDL-induced
injury in Caco-2 cells. J. Nutr. Bethesda 129 (7), 1269–1277.
Guadagni, D.G., Buttery, R.G., Harris, J., 1966. Odour intensities of
hop oil components. J. Sci. Food Agric. 17, 142–144.
Perrin, D.D., Armarego, W.L.F., 1988. Purification of Laboratory
Chemicals, third ed. Pergamon Press, Oxford.
Pichersky, E., Raguso, R.A., Lewinsohn, E., Croteau, R., 1994. Floral
scent production in Clarkia (Onagraceae): I. Localization and
developmental modulation of monoterpene emission and linalool
synthase activity. Plant Physiol. 106, 1533–1540.
Guterman, I., Shalit, M., Menda, N., Piestun, D., Dafny-Yelin, M.,
Shalev, G., Bar, E., Davydov, O., Ovadis, M., Emanuel, M., Wang,
J., Adam, Z., Pichersky, E., Lewinsohn, E., Zamir, D., Vainstein,
A., Weiss, D., 2002. Rose Scent: Genomics approach to discovering
novel floral fragrance-related genes. The Plant Cell 14, 2325–2338.
Hassall, A.K., Pringle, G.J., MacRae, E.A., 1998. Development,
maturation, and postharvest responses of Actinidia arguta (Sieb. et
Zucc.) Planch. ex Miq. fruit. NZ J. Crop Hort. Sci. 26, 95–108.
Hayase, F., Chung, T.Y., Kato, H., 1984. Changes of volatile compo-
nents of tomato fruits during ripening. Food Chem. 14 (2), 113–124.
Jordan, M.J., Margaria, C.A., Shaw, P.E., Goodner, K.L., 2002.
Aroma Active Compounds in Aqueous Kiwi Fruit Essence and
Kiwi Fruit Puree by GC-MS and Multidimensional GC/GC-O. J.
Agric. Food Chem. 50, 5386–5390.
Rettinger, K., Karl, V., Schmarr, H.G., Detter, F., Hener, U.,
Mosandl, A., 1991. Chirospecific analysis of 2-alkyl-branched alco-
hols, acids, and esters: chirality evaluation of 2-methylbutanoates
from apples and pineapples. Phytochem. Anal. 2, 184–188.
Rosecke, J., Konig, W.A., 2000. Odorous compounds from the fungus
Gloeophyllum odoratum. Flav. Frag. J. 15 (5), 315–319.
Rothe, M., Thomas, B., 1963. Aroma of bread. Evaluation of chemi-
cal taste analyses with the aid of threshold value. Z. Lebensm.
Unters. Forsch. 119, 302–310 (in German).
Rowan, D.D., Lane, H.P., Allen, J.M., Fielder, S., Hunt, M.B., 1996.
Biosynthesis of 2-methylbutyl, 2-methyl-2-butenyl, and 2-methylbu-
tanoate esters in Red Delicious and Granny Smith apples using deu-
terium-labeled substrates. J. Agric. Food Chem. 44 (10), 3276–3285.
Rychlik, M., Schieberle, P., Grosch, W., 1998. Compilation of Odor
Thresholds, Odor Qualities and Retention Indices of Key Food
Odorants. Deutche Forschungsanstalt fur Lebensmittelchemie and
Institut fur lebensmittelchemie der Technischen Universitat
Munchen, Garching.
Jurgens, A., Witt, T., Gottsberger, G., 2002. Flower scent composition
in night-flowering Silene species (Caryophyllaceae). Biochem. Sys-
tem. Ecol 30 (5), 383–397.
Kint, S., Teranishi, R., Lingren, P.D., Shaver, T.N., Raulston, J.R.,
1993. Chemical composition of Gaura suffulta and Gaura longiflora
flower volatiles. J. Essent. Oil Res. 5 (2), 201–203.
Kollmannsberger, H., Berger, R.G., 1992. Precursor atmosphere sto-
rage induced flavour changes in apples cv. Red Delicious. Chem.
Mikrobiol. Technol. Lebensm. 14, 81–86.
Sazima, M., Vogel, S., Cocucci, A., Hausner, G., 1993. The perfume
flowers of Cyphomandra (Solanaceae): pollination by euglossine
bees, bellows mechanism, osmophores, and volatiles. Plant Syst.
Evol. 184, 51–88.
Krebs, K.G., Heusser, D., Wimmer, H., 1969. Z. Spray reagents. In:
Stahl, E. (Ed.), Thin-Layer Chromatography. A Laboratory Hand-
book, second ed. Springer-Verlag, Berlin-Heidleberg-New York, pp.
855–905.
Schmidt, S.P., Monroe, R.E., 1976. Biosynthesis of the wax moth sex
attractants. Insect Biochem. 6 (4), 377–380.
Lapadatescu, C., Ginies, C., Le-Quere, J.L., Bonnarme, P., 2000.
Novel scheme for biosynthesis of aryl metabolites from l-phenyl-
alanine in the fungus Bjerkandera adusta. Appl. Environ. Microbiol
66 (4), 1517–1522.
Shimoda, M., Wu, Y., Osajima, Y., 1996. Aroma compounds from
aqueous solution of haze (Rhus succedanea) honey determined by
adsorptive column chromatography. J. Agric. Food Chem. 44,
3913–3918.
Lewinsohn, E., Schalechet, F., Wilkinson, J., Matsui, K., Tadmor, Y.,
Nam, K.H., Amar, O., Lastochkin, E., Larkov, O., Ravid, U.,
Hiatt, W., Gepstein, S., Pichersky, E., 2001. Enhanced levels of the
aroma and flavor compound S-linalool by metabolic engineering of
the terpenoid pathway in tomato fruits. Plant Physiol. 127, 1256–
1265.
Silk, P.J., Aubry, C., Lonergan, G.C., Macaulay, J.B., 2001. Chloro-
metabolite production by the ecologically important white rot fun-
gus Bjerkandera adusta. Chemosphere 44 (7), 1603–1616.
Takeoka, G.R., Guntert, M., Flath, R.A., Wurz, R.E., Jennings, W.,
1986. Volatile constituents of kiwifruit (Actinidia chinensis Planch.).
J. Agric. Food Chem. 34, 576–578.

A.J. Matich et al. / Phytochemistry 63 (2003) 285–301
301
Tatsuka, K., Suekane, S., Sakai, Y., Sumitani, H., 1990. Volatile con-
stituents of kiwifruit flowers: Simultaneous distillation and extrac-
tion versus headspace sampling. J. Agric. Food Chem. 38, 2176–
2180.
thistle (Carduus nutans) honey. J. Agric. Food Chem. 41 (6), 873–
878.
Yaylayan, V.A., Keyhani, A. 1998. Application of microwave-assisted
process and Py-GC-MS to the analysis of Maillard reaction pro-
ducts. Paper presented at the Flavor analysis developments in iso-
lation and characterization, Washington, DC.
Tollsten, L., Bergstrom, L.G., 1993. Fragrance chemotypes of Pla-
tanthera (Orchidaceae)—the result of adaptation to pollinating
moths? Nordic J. Bot. 13 (6), 607–613.
Young, H., 1981. Direct desorption of traps for capillary column gas
chromatography. J. Chromatogr. 214, 197–201.
Vainstein, A., Lewinsohn, E., Pichersky, E., Weiss, D., 2001. Floral
fragrance: new inroads into an old commodity. Plant Physiol. 127,
1383–1389.
Young, H., Paterson, V.J., 1985. The effects of harvest maturity,
ripeness and storage on kiwifruit aroma. J. Sci. Food Agric. 36,
352–358.
Wakayama, S., Namba, S., 1974. Lilac aldehydes. Bull. Chem. Soc.
Jap. 47 (5), 1293–1294.
Young, H., Paterson, V.J., 1990. The flavour of exotic fruit. In: Mor-
ton, I.D., Macleod, A.J. (Eds.), Food Flavours. Part C. The flavour
of fruits. Elsevier, New York, pp. 281–326.
Wakayama, S., Namba, S., Hosui, K., Ohno, M., 1973. The synthesis
and the absolute configerations of lilac alcohols, New naturally
occuring odorous ingredients of lilac flower. Bull. Chem. Soc. Jap.
46, 3183–3187.
Young, H., Paterson, V.J., 1995. Characterisation of bound flavour
components in kiwifruit. J. Sci. Food Agric. 68, 257–260.
Young, H., Paterson, V.J., Burns, D.J.W., 1983. Volatile aroma con-
stituents of kiwifruit. J. Sci. Food Agric. 34, 81–85.
Watanabe, I., Takazawa, O., Warita, Y., Awano, K. I. 1974. Volatile
compounds of apricot flowers. Paper presented at the ACS-symp-
ser., Washington, DC.
Young, H., Perera, C.O., Paterson, V.J., 1992. Identification of E-hex-
3-enal as an important contributor to the off-flavour aroma in
kiwifruit juice. J. Sci. Food Agric. 58, 519–522.
Wilkins, A.L., Lu, Y., Tan, S.T., 1993. Extractives from New Zealand
honeys. 4. Linalool derivatives and other components from nodding