C-27 Apocarotenoids in the Flowers of Boronia megastigma
(Nees)
CHRIS M. COOPER
†
,NOEL W. DAVIES,
‡
AND ROBERT C. MENARY*
,†
School of Agricultural Science and Central Science Laboratory, University of Tasmania,
GPO Box 252-54, Hobart, Tasmania, Australia
Five C-27 apocarotenoids were detected in acetone extracts of the flowers of Boronia megastigma
(Nees) using HPLC with UV-vis photodiode array and MS detection. Comparison of methylated and
unmethylated extracts aided identification when considered with the UV-vis and MS data. The five
apocarotenoids identified here were hydroxy-apo-10′-carotenoic acid (1), methyl hydroxy-apo-10′-
carotenoate (2), apo-10′-carotenoic acid (3), apo-10′-carotenal (4), and methyl apo-10′-carotenoate
(5). The data obtained was not sufficient to allow the specific isomeric forms to be unequivocally
identified. The results further support speculation that the C-13 norisoprenoids found in boronia are
derived from C-40 carotenoids. Possible parent molecules of â-ionone, an important component of
boronia extract, were identified. An understanding of C-13 norisoprenoid biosynthesis may assist in
the selection and postharvest processing of boronia flowers for flavor and fragrance applications.
KEYWORDS: Boronia megastigma (Nees); brown boronia; essential oils; carotenoids; apocarotenoids;
â-carotene; â-ionone; C-13 norisoprenoids; biosynthesis; 3-hydroxy-â-apo-10′-carotenoic acid; methyl
3-hydroxy-â-apo-10′-carotenoate; â-apo-10′-carotenoic acid; â-apo-10′-carotenal; methyl â-apo-10′-caro-
tenoate
INTRODUCTION
Boronia megastigma (Nees) is grown commercially in
Tasmania to produce a flower extract. The extract, which is
also made into an absolute, has a complex range of components
including C-13 norisoprenoids. These are regarded as important
compounds in many natural and synthetic flavor and fragrance
products (1, 2). Extracts from boronia flowers contain more than
20 C-13 norisoprenoids out of 129 components positively
identified (3-5). â-Ionone, first identified in boronia before
1927 (6), is a high yielding norisoprenoid ingredient of that
extract. Other C-13 norisoprenoids identified in boronia include
3-hydroxy-â-ionone, 4-oxo-â-ionone, various â-ionols, and 7,8-
dihydro-â-ionone which has a “fruity, ionone-orris-like smell”
(4). While substantial progress in improving extract quality has
been made in recent years (7-9), improved understanding of
the biochemistry of norisoprenoid production may provide an
opportunity to further enhance the quality of boronia extracts
and provide opportunities for market differentiation.
It is widely assumed that various volatiles, including C-13
noriosoprenoids, are derived from C-40 carotenoids. This is
based on structural comparisons, and various authors have
published data demonstrating the combined presence in flowers
of molecules which together structurally match the C-40
carotenoid parent molecules. In quince, Lutz and Winterhalter
(10) argued that both C-13 and C-15 end group volatiles, and
corresponding C-10 and C-12 fragments from the central part
of the polyene chain, were evidence of biosynthesis from C-40
carotenoids. In starfruit (11) the finding of C-15 and C-10
molecules, matching carotenoid end groups and central chain,
respectively, was again seen as evidence of volatile derivation
from carotenoids. Similarly in saffron, crocetin (C-20) and a
variety of C-10 molecules, including the important flavor
compound safranal, have been postulated to be derived from
the C-40 molecule zeaxanthin (12, 13).
Extracts of rose flowers are known to contain, as does
boronia, a wide range of C-13 norisoprenoids. Eugster and
Marki-Fischer (14), on the basis of the identification of C-27
apocarotenoids and rosafluene (C-14), postulated a two-step
biosynthetic process in roses. The first step proposed was
cleavage in the 9,10 position to form one C-13 norisoprenoid
molecule and a C-27 apo-carotenoid. This latter molecule is
further cleaved to form the C-14 rosafluene and a second C-13
norisoprenoid molecule. While many of the C-13 norisoprenoids
in boronia correspond to the end-groups of known carotenoids,
no matching carotenoid cleavage products have been previously
identified. In particular, extensive identification of volatiles in
boronia extracts indicates the absence of rosafluene (or other
C-14 volatiles) as is found in rose extracts. However, biosyn-
thesis of â-ionone and other C-13 norisoprenoids through
cleavage of C-40 carotenoids has long been considered possible.
Previous investigation (15) yielded evidence in boronia flowers
of lutein, neoxanthin, and â-carotene, a possible precursor of
* To whom correspondence should be addressed. Telephone: ++61 3
62 262723, Fax: ++ 61 3 62 267609; E-mail: R.Menary@utas.edu.au.
†
School of Agricultural Science.
‡
Central Science Laboratory.
2384 J. Agric. Food Chem. 2003, 51, 2384−2389
10.1021/jf026007c CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/11/2003

â-ionone, along with a number of unidentified carotenoid peaks
in chromatograms. The availability of HPLC linked to photo-
diode array and atmospheric pressure chemical ionization-mass
spectrometry (APCI-MS) detection has allowed more detailed
investigation into the carotenoid profile of boronia flowers.
MATERIALS AND METHODS
Materials. Boronia megastigma (Nees) clone 3 flowers developed
by the University of Tasmania and grown in southern Tasmania were
used. All chemicals and solvents were analytical grade.
Carotenoid Extraction from Boronia Flowers. Typically 10-20
g of flower material was extracted using a method adapted from that
of Jaren-Galen et al. (16). Accordingly, 20 g of boronia flowers or
buds, plus 5.0 g of calcium carbonate for acid neutralization (17), was
homogenized in 100 mL of acetone (4 °C) for 60 s using an Ultra-
Turrex (T25 basic, Ika labortechnik, setting 6), fitted with an 18 mm
head. The mixture was centrifuged for 5 min at 15000g (Beckman J2-
21 M/E, rotor 20.0). The pellet was extracted twice more in 50 mL of
acetone, and the supernatants were combined in a separating funnel
with 100 mL of diethyl ether. The mixture was shaken, and 200 mL of
10% NaCl was added to promote separation of the ether layer. The
aqueous layer was discarded and the ether layer washed several times
with 100 mL aliquots of 10% NaCl. The ether fraction containing the
pigments was dried over sodium sulfate and filtered, and the solvent
was removed by rotary evaporation. The pigments were stored at -70
°C prior to analysis. All operations were at 4 °C under reduced or pale
yellow light (18).
HPLC. A Waters Alliance 2690 HPLC and 996 Photodiode array
detector were used for chromatography and identification of carotenoids
and apocarotenoids. A Waters Nova-Pak 150 × 3.9 mm i.d. C18 column
fitted with an Alltech Econosphere C18 guard cartridge was used to
achieve separation. Three different programs were used for analytical
purposes.
Program A: Initial conditions were 87% acetonitrile/10% methanol/
3% water for 3 min programmed to 85% methanol/15% hexane at 10
min which was then isocratic for 5 min. The flow rate was 1 mL/min.
Following the completion of each run, the column was returned to
starting conditions over 1 min and equilibrated for 8 min prior to the
next run.
Program B: Initial conditions were 50% acetonitrile/50% water for
2 min followed by a linear gradient to 85% acetonitrile/15% methanol
at 15 min. A further linear gradient to 85% methanol/15% hexane at
25 min was then held for 10 min. The flow rate was 1 mL/min. The
column was reequilibrated to start conditions for 10 min between
samples.
Program C: Starting conditions were 50% acetonitrile/50% water
for 2 min. This was followed by a linear gradient to 85% acetonitrile/
15% methanol at 22 min and a further program to 85% methanol/15%
hexane at 32 min which was held for 8 min. The flow rate was 1 mL/
min. Reequilibration was to 100% methanol for 3 min and 50%
acetonitrile/50% water for 9 min.
Data Processing. Waters Millenium software was used to analyze
data and chromatograms were extracted at 430 and 451 nm. Data were
recorded from 250 to 700 nm every1sat1.2nmresolution. Lutein
was identified on the basis of published UV-vis data and retention
times.
HPLC/MS. Mass spectral data was obtained with a Finnigan LCQ
equipped with an atmospheric pressure chemical ionization (APCI) ion
source. Settings were sheath gas 60 psi, auxiliary gas 15 psi, vaporizer
temperature 450 °C, discharge current 6 µA, capillary temperature 170
°C, capillary voltage 20 V, default MS/MS collision energy 25%.
Scanning usually occurred over the m/z range 100 to 1200. Signals for
the apocarotenoids reported here were maximized by shortening the
m/z range to 100-680.
Synthesis of Diazomethane. Diazomethane was prepared in a 250
mL conical flask by adding N-methyl-N-nitrourea (3.0 g, 22.5 mmol)
in small portions to a mixture of aqueous potassium hydroxide (40%
w/v, 40 mL) and diethyl ether (50 mL) at 5 °C with constant stirring
(magnetic stirrer) for 20-30 min. The ether layer was separated and
the aqueous layer further extracted with 25 mL of ether. The combined
ether fractions were allowed to stand over some solid potassium
hydroxide (30 min minimum) and then transferred to a screw top bottle
for storage at -15 °C.
Methylation of Carotenoid Extracts. Aliquots of carotenoid extract
(10 mg) were treated with diazomethane (1 or 2 mL of the ether fraction
above) for 30 min after which the solvent was evaporated under a stream
of nitrogen. The resultant methylated extract was redissolved in 2 mL
of acetone and analyzed by HPLC. This process was conducted under
low or nil light conditions.
RESULTS AND DISCUSSION
Five C-27 apocarotenoids were identified for the first time
in the flowers of Boronia megastigma (Nees), using HPLC with
diode array detection and HPLC-MS. This identification was
Figure 1. UV−vis spectra for five components of boronia flowers obtained using HPLC program C.
Apocarotenoids in Boronia Flowers J. Agric. Food Chem., Vol. 51, No. 8, 2003 2385