Isolation and identification of a potent radical scavenger (canolol) from roasted high erucic mustard seed oil from nepal and its formation during roasting

Article abstract of DOI:10.1021/jf301738y

Roasting of high erucic mustard (HEM) seed has been reported to give a typical flavor and increase the oxidative stability of the extracted oil. A potent radical scavenging compound was successfully isolated from roasted HEM seed oil in a single-step chromatographic separation using an amino solid-phase extraction column. Nuclear magnetic resonance and mass spectrometry spectra revealed the compound as 2,6-dimethoxy-4-vinylphenol (generally known as canolol), and its identity was fully confirmed by chemical synthesis. The formation of canolol during roasting was compared among HEM varieties (Brassica juncea, B. juncea var. oriental, Brassica nigra, and Sinapis alba) together with a low erucic rapeseed variety. HEM varieties were shown to produce less than one-third of canolol compared to rapeseed at similar roasting conditions. This observation was linked to a lower free sinapic acid content together with a lower loss of sinapic acid derivatives in the HEM varieties compared to rapeseed. Around 50% of the canolol formed in the roasted seed was shown to be extracted in the oil. Roasting of HEM seed before oil extraction was found to be a beneficial step to obtain canolol-enriched oil, which could improve the oxidative stability.

Full text of DOI:10.1021/jf301738y

Article
pubs.acs.org/JAFC
Isolation and Identification of a Potent Radical Scavenger (Canolol)
from Roasted High Erucic Mustard Seed Oil from Nepal and Its
Formation during Roasting
Kshitij Shrestha,† Christian V. Stevens,‡ and Bruno De Meulenaer*,†
‡
†NutriFOODchem Unit, Department of Food Safety and Food Quality, and Department of Sustainable Organic Chemistry and
Technology, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium
S
* Supporting Information
ABSTRACT: Roasting of high erucic mustard (HEM) seed has been reported to give a typical flavor and increase the oxidative
stability of the extracted oil. A potent radical scavenging compound was successfully isolated from roasted HEM seed oil in a
single-step chromatographic separation using an amino solid-phase extraction column. Nuclear magnetic resonance and mass
spectrometry spectra revealed the compound as 2,6-dimethoxy-4-vinylphenol (generally known as canolol), and its identity was
fully confirmed by chemical synthesis. The formation of canolol during roasting was compared among HEM varieties (Brassica
juncea, B. juncea var. oriental, Brassica nigra, and Sinapis alba) together with a low erucic rapeseed variety. HEM varieties were
shown to produce less than one-third of canolol compared to rapeseed at similar roasting conditions. This observation was linked
to a lower free sinapic acid content together with a lower loss of sinapic acid derivatives in the HEM varieties compared to
rapeseed. Around 50% of the canolol formed in the roasted seed was shown to be extracted in the oil. Roasting of HEM seed
before oil extraction was found to be a beneficial step to obtain canolol-enriched oil, which could improve the oxidative stability.
KEYWORDS: 2,6-Dimethoxy-4-vinylphenol (canolol), Brassica juncea, Brassica juncea var. oriental, Brassica nigra, Sinapis alba,
mustard seed, rapeseed, antioxidant, roasting, sinapic acid
roasting on oil stability by the formation of different radical
scavengers. Roasting of sesame seed results in the production of
sesamols, which leads to the increased oil stability.6 Canolol
(2,6-dimethoxy-4-vinylphenol) is formed by the decarboxyla-
tion of sinapic acid during rapeseed roasting.5,7,8 Canolol
formation has been shown to improve the oxidative stability of
roasted rapeseed.8 The increased oxidative stability of oil
extracted from the roasted seed of different Brassica napus and
Brassica juncea breeds (both low erucic varieties) has also been
linked to canolol formation; however, quantitative data were
not presented.9
The possible formation of such radical scavengers in roasted
HEM seed oil remains to be investigated. Therefore, the
objective of the current study was to isolate and identify a
major radical scavenging compound from roasted HEM seed oil
obtained from the Nepalese market and to compare different
varieties on its formation during roasting.
INTRODUCTION
■
High erucic mustard (HEM) seed oil is one of the most
common liquid cooking oils in Nepal and India. These
traditional varieties contain 22−60% of erucic acid and are
rich in glucosinolates.1 HEM seed is generally roasted before oil
extraction for its typical flavor.2 In most parts of the world, high
erucic traditional varieties have been replaced by new varieties,
developed by different breeding practices, containing low erucic
acid (<2%) and low amounts of glucosinolates (<30 μmol/g),
such as Canadian rapeseed variety (canola) and European low
erucic rapeseed variety.1,3 Therefore, the consumption of HEM
seed oil is very limited, only in some specific areas of the world,
particularly in Nepal and India.1
Scientific studies on roasted HEM seed oil are very rare. A
recent study has shown that roasting of HEM seed before oil
extraction could significantly improve the oxidative stability of
oil together with the stability of tocopherol and lutein during
storage.2,4 The increased stability was attributed to the
formation of antioxidative compounds during roasting via the
Maillard reaction. These authors also reported a small increase
in the tocopherol content (around 40 μg/g of oil) during
roasting. Because the reported increase in the tocopherol
content was very low and the suggested Maillard reaction effect
was not clearly demonstrated, it remains unclear which factors
or compounds are responsible for the increase in oxidative
stability of roasted HEM seed oil.2 A similar roasting condition
has been reported to have no effect on the tocopherol content
during rapeseed roasting.5 Roasting has multiple consequences,
and the increased oil stability can be due to their overall effect.
Additionally, several studies have shown a positive effect of seed
MATERIALS AND METHODS
■
Materials. Roasted HEM seed oil samples were collected from the
Nepalese market. The samples were stored under refrigerated
conditions (below −18 °C) until analysis. HEM varieties (B. juncea,
B. juncea var. oriental, Brassica nigra, and Sinapis alba) and low erucic
rapeseed were collected from the local market in Belgium. Thin-layer
chromatography (TLC) plates (Silicagel 60) were obtained from
Merck (Darmstadt, Germany). The tocopherol standard was obtained
Received: April 22, 2012
Revised: June 25, 2012
Accepted: July 2, 2012
Published: July 2, 2012
© 2012 American Chemical Society
7506
dx.doi.org/10.1021/jf301738y | J. Agric. Food Chem. 2012, 60, 7506−7512

Journal of Agricultural and Food Chemistry
Article
from DSM (Parsippany, NJ). The aminopropyl solid-phase extraction
(SPE) column (Extract Clean, 500 mg/4 mL) was purchased from
Grace (Lokeren, Belgium). Syringaldehyde and sinapic acid were
purchased from Sigma Aldrich (Steinheim, Germany). Other reagents
and solvents were of analytical grade and obtained from reliable
commercial sources.
TLC Analysis. The oil sample was dissolved in hexane and applied
on a TLC plate. A mixed tocopherol standard solution (containing α-,
β-, γ-, and δ-tocopherols) was also spotted. The plate was developed
with hexane/diethyl ether/acetic acid (80:30:1, v/v/v). Afterward, it
was sprayed uniformly with 0.1 mM 2,2-diphenyl-1-picrylhydrazyl
(DPPH) solution in hexane to give a slightly pink background and
held for 5 min. The radical scavengers appeared as white spots on the
pink background.
Isolation and Purification of the Compound of Interest.
Roasted HEM seed oil (1 g) was dissolved in 2 mL of hexane and
extracted 3 times with 2 mL of methanol. On the basis of TLC
analysis, the compound of interest (Rf = 0.15) was found to be in the
methanol fraction. Therefore, the methanol fraction was dried under
reduced pressure, and the residue was dissolved in 5 mL of 0.9%
isopropanol in hexane (v/v). This solution (0.5 mL) was brought on
an aminopropyl SPE column (preconditioned with the same solvent
mixture). The SPE column was eluted with the same solvent mixture,
and the first 6 mL was discarded. Afterward, 12 fractions of each of 4
mL were collected. Finally, the column was eluted with 5% iso-
propanol in hexane (v/v), and three more fractions (each of 4 mL)
were also collected (fractions 13−15). The compound of interest was
observed in the fractions 1−5, via TLC analysis. These fractions were
combined and dried under nitrogen. The procedure was repeated to
obtain around 5 mg of white dry compound. The purity of this
compound was confirmed by TLC and high-performance liquid
chromatography (HPLC) analysis.
Identification of the Isolated Compound by Spectroscopic
Measurements. The identification of the isolated compound was
carried out on the basis of nuclear magnetic resonance (NMR), mass
spectrometry (MS), and ultraviolet−visible (UV−vis) spectroscopic
measurements.
NMR Spectroscopy. The purified compound was dissolved in
deuterated chloroform. The proton (1H) and carbon (13C) NMR
spectra were taken using a Jeol EX300 Eclipse NMR (300 MHz)
spectrophotometer (Japan).
Liquid Chromatography−Mass Spectrometry (LC−MS) Spectros-
copy. The LC−MS analysis was carried out using UltiMate 3000
ultrahigh-pressure liquid chromatography (UHPLC, Dionex) equip-
ped with a degasser, four solvent delivery modules, an autosampler, a
column oven, and an UV detector coupled with a MicroTOF MS
instrument (Bruker). The purified compound was dissolved in an
isopropanol/water/acetic acid (90:10:0.1, v/v/v) mixture (10 μg/mL)
and then injected on a C8 Zorbax 300 SB column (Agilent, Santa
Clara, CA). Mobile phase A was a water/acetonitrile/acetic acid
mixture (90:10:0.1, v/v/v), and mobile phase B was an acetonitrile/
water/acetic acid mixture (90:10:0.1, v/v/v). The method was run
with 10% mobile phase B for 1 min, then a gradient was applied to
reach 100% of mobile phase B in 11 min, which was held for 5 min.
Afterward, the initial conditions were reached in 0.5 min, and the
column was allowed to equilibrate for 4.5 min before a subsequent
analytical run. The solvent flow rate was 0.2 mL/min. Electrospray
ionization (ESI) in positive-ion mode was used, and m/z values were
scanned from 50 to 1000. The capillary voltage was set at 4500 V, and
the end plate offset was at −500 V. The nebulizer pressure was 0.5 bar
and was heated to 190 °C with dry nitrogen at a flow rate of 4 mL/
min.
compound was purified on a silica gel column using the procedure
described in the same method.10 Crystallization was applied as an
additional step to increase the purity. A saturated solution of the
synthesized compound in hexane was prepared at ambient temper-
ature, and crystallization was induced by storing it inside the freezer
(−28 °C). Crystals (white) were separated from the mother liquor and
dried under nitrogen. The purity of the compound was confirmed by
NMR spectroscopy.
HPLC Analysis of Tocopherols and Canolol in Oil. Tocopherol
and canolol contents of the oil were analyzed on Agilent 1100 series
HPLC equipped with a degasser, four solvent delivery modules, an
autosampler, a column oven, and a fluorescence detector. A mobile
phase containing 0.9% isopropanol in hexane (v/v) was used in
isocratic conditions at a flow rate of 1 mL/min. Separation was carried
out on a LiChroCART 250-4,6 Purospher STAR Si (5 μm) column
(Merck, Darmstadt, Germany) with a precolumn containing the same
phase. The temperature of the column was maintained at 35 °C, and
the chromatogram was obtained with a fluorescence detector
(excitation at 285 nm and emission at 325 nm). The fluorescence
emission spectra (310−400 nm) were also obtained at a 285 nm
excitation wavelength. Analysis of each sample was carried out in
triplicate. Because the reference canolol compound was not
commercially available, the synthesized canolol was used as the
standard for quantification.
Roasting of Seed and Powder Samples of HEM and
Rapeseed and Oil Extraction. A seed sample (80 g) was added
to a heated beaker (250 mL) placed on the oil bath maintained at 180
°C. Both the heating oil and seed sample were continuously mixed
with electric mechanical stirrers. The temperatures of the heating oil
and seed sample were continuously monitored using Testo thermostat
probes. The oil temperature was maintained at 180 °C, and roasting
was carried out for 10 min. The seed temperature reached 162 3.5
°C in 10 min. The seed temperature profile with time followed the
equation T = 32.796 ln(t) + 89.422 (R2 = 0.99), where T is the seed
temperature (°C) and t is the time (min).
The same setup could not be used for homogeneous mixing of seed
powder. Therefore, 4 g of seed powder was taken in a test tube and
roasted in an oil bath maintained at 180 °C. The thermostat probe was
kept in the center of the test tube containing the seed powder, and the
temperature was monitored. After heating for 6 min (holding the oil
bath temperature constant at 180 °C), the seed powder temperature
reached 160 °C with the temperature profile represented by equation
T = 36.578 ln(t) + 96.097 (R2 = 0.99), where T is the seed powder
temperature in the center (°C) and t is the time (min). During 6−10
min of heating time, the oil bath temperature was gradually lowered to
172 °C and the seed powder temperature profile followed the equation
T = 15.569 ln(t) + 134 (R2 = 0.97). Roasting of both seed and powder
were carried out in three batches for each variety, and further analyses
were performed independently on each batch.
The oil was extracted 3 times from 30 g of ground sample using 80
mL of petroleum ether. The sample−solvent mixture was kept in a
shaker for 15 min and then centrifuged at 9000g for 10 min. The
supernatant was filtered through a filter paper and evaporated under
reduced pressure at 35 °C. The oil sample was dried overnight under
nitrogen and stored in the freezer (−28 °C) until further analysis.
Analysis of the Free Sinapic Acid (FSA) Content. Seed samples
were finely ground using a coffee grinder. The seed powder (4 g) was
mixed with 25 mL of methanol/water/acetic acid (70:30:0.2, v/v/v)
using an ultra turrax at 10 000 rpm for 2 min. The ultra turrax probe
was washed with 20 mL of the same solvent, which was added to the
mixture, and the volume was topped up to 50 mL. After centrifugation
at 2800g for 10 min, the supernatant was filtered through a Millex-
LCR filter [0.45 μm polytetrafluoroethylene (PTFE) membrane,
Millipore, Ireland]. The filtrate was diluted 4 times with 0.2% acetic
acid in water (v/v) before injecting on an UHPLC system.
UV−Vis Spectroscopy. The UV and visible absorption spectra were
taken in hexane using a Cary 50 UV−vis spectrophotometer (Varian)
in a quartz cuvette.
Synthesis of 2,6-Dimethoxy-4-vinylphenol (Canolol). The
method described for the synthesis of 4-vinylphenols from 4-hydroxy-
substituted benzaldehydes under microwave irradiation was fol-
lowed.10 Syringaldehyde was chosen as an appropriate benzaldehyde
for the synthesis of 2,6-dimethoxy-4-vinylphenol. The synthesized
Analysis of the Total Sinapic Acid Content after Basic
Hydrolysis (TSAH). Alkaline hydrolysis of the esterified phenolic
compounds were carried out using a previously described method,
with slight modifications.11 The supernatant phenolic extract (3 mL)
was mixed with 3 mL of distilled water and 1.5 mL of 10 M NaOH.
7507
dx.doi.org/10.1021/jf301738y | J. Agric. Food Chem. 2012, 60, 7506−7512

Journal of Agricultural and Food Chemistry
Article
The sample was flushed with nitrogen, covered with aluminum foil,
and kept on a shaker for 4 h at room temperature. Afterward, the pH
of the solution was adjusted to 2 using 8 M HCl, and the phenolic
compounds were extracted 3 times each with 3 mL of ethyl acetate.
The extracts were combined and dried under nitrogen. The residue
was redissolved in 5 mL of methanol, and the volume was topped up
to 15 mL with 0.2% acetic acid in water. The solution was diluted 5
times with methanol/water/acetic acid (20:80:0.2, v/v/v) solvent
mixture before injecting on an UHPLC system.
UHPLC−Diode Array Detector (DAD) Analysis for the
Quantification of Sinapic Acid and Canolol. The analysis was
carried out using UltiMate 3000 UHPLC (Dionex), equipped with a
degasser, four solvent delivery modules, an autosampler, a column
oven, and a DAD. The separation was carried out using a 2.1 × 150
mm, 1.8 μm, Zorbax Eclipse Plus C18 column (Agilent, Santa Clara,
CA) maintained at 30 °C. The injection volume was 10 μL, and the
compounds were eluted with 0.2% acetic acid in water (v/v) (mobile
phase A) and 100% methanol (mobile phase B) at a flow rate of 0.3
mL/min. The gradient program was as follows: 20−40% B (5 min),
40% B (1 min), 40−50% B (1 min), 50% B (1 min), 50−60% B (5
min), 60−70% B (2 min), 70% B (1 min), 70−20% B (1 min), and
20% B (5 min). The chromatograms were recorded at 330 and 280 nm
absorbances. The sinapic acid and canolol contents were quantified on
the basis of the calibration curves using a chromatogram at 330 and
280 nm, respectively. All of the analyses were carried out in triplicates.
Statistical Analysis. All data are presented as mean values 95%
confidence interval of the mean based on three independent
experiments. Data were transformed into a logarithmic scale before
statistical analyses to have homoscedasticity. Analysis of variance
(ANOVA) and post-hoc Tukey test were performed using TIBCO
Spotfire S+ 8.1 software. The significance level is p < 0.05, unless
otherwise indicated.
Figure 1. Normal-phase HPLC chromatogram of (A) roasted HEM
seed oil, (B) flax seed oil, and (C) isolated compound of interest, using
a fluorescence detector (excitation at 285 nm and emission at 325
nm). The time axis is in minutes.
The mass spectrum of the compound of interest showed the
ions (with relative intensity) with m/z values of 181.08 (100),
149.06 (32), 203.06 (15), 121.06 (11), 166.06 (3), and 103.05
(2). The NMR chemical shifts of the compound of interest are
given in Table 1. The compound of interest was identified as
2,6-dimethoxy-4-vinylphenol, and the fragmentation pattern
completely supported the proposed structure (Figure 2).
RESULTS AND DISCUSSION
■
Isolation and Identification of a Potent Radical
Scavenger from Roasted HEM Seed Oil Obtained from
the Nepalese Market. The qualitative identification of the
different tocopherols present in roasted HEM seed oil was
carried out using normal-phase HPLC analysis. The α-, γ-, and
δ-tocopherols were identified comparing the retention time and
fluorescence spectra to the tocopherol standard. One small
peak eluting before β-tocopherol and another peak eluting after
δ-tocopherol were additionally observed in the chromatogram
(Figure 1). Both peaks showed similar fluorescence spectra as
tocopherol and did not show any tocotrienol using known
standards. Both of these peaks were observed in a number of
HEM seed oil samples analyzed on HPLC. After linseed oil was
injected, which contains plastochromanol-8,1 the small peak
before the β-tocopherol position was identified to be
plastochromanol-8. The major peak was more polar than all
of the tocopherols and also showed a slight tailing (Figure 1).
The oil sample was subjected to a separation on silica using
TLC. Radical scavenging compounds were detected simply by
spraying a DPPH solution on the TLC plate after development.
Tocopherols eluted at the Rf value of 0.3−0.5. Another radical
scavenging spot was observed at a Rf value of 0.15, which was
isolated from the TLC plate and injected on HPLC. It was
proven that the isolated compound had the same elution
behavior as that of the unknown compound eluting at 14 min,
as shown in Figure 1.
Table 1. NMR Chemical Shifts of the Isolated Compound of
Interest
Using an aminopropyl SPE column, the compound of
interest could easily be purified by washing the column with 6
mL of 0.9% isopropanol in hexane (v/v) and then eluting the
compound with 20 mL of the same solvent mixture. The UV
absorption spectra of the isolated compound showed peaks at
222 and 272 nm in hexane.
The final confirmation of the identity of the isolated
compound of interest was performed by synthesizing 2,6-
dimethoxy-4-vinylphenol from syringaldehyde (Figure 3). The
isolated and synthesized compound had similar MS, NMR, UV,
and fluorescence spectra, along with the same retention time in
HPLC, and was also in line with the existing literature
7508
dx.doi.org/10.1021/jf301738y | J. Agric. Food Chem. 2012, 60, 7506−7512

Journal of Agricultural and Food Chemistry
Article
and separation on an aminopropyl SPE column by sequential
elution with hexane/isopropanol (98.8:1.2, v/v), followed by a
mixture in a 95:5 (v/v) ratio. Afterward, the canolol-rich
fraction was collected with isopropanol.9 This fraction contains
other polar compounds together with canolol. In a more recent
isolation method, roasted ground rapeseed was extracted with
70% methanol after washing with petroleum ether to remove
the oil. The extract was concentrated and extracted with
hexane/isopropanol (4:1, v/v). Afterward, it was further
extracted with a 0.1 M hydrochloric acid solution, and
separation was carried out using SPE (Chromabond HR-P)
by eluting with methanol. The methanol fraction was
concentrated and further extracted with hexane. Finally, canolol
was purified using TLC by developing with a hexane/ethyl
acetate (7:3, v/v) mixture.7 In comparison to all of these
previous isolation procedures, the method developed in our
laboratory could be considered as one of the simplest methods
currently available. The developed method was close to the one
previously described,9 with the additional benefit of obtaining
canolol without eluting other polar compounds together in the
same fraction. Furthermore, this method was also applicable for
the isolation of canolol from roasted rapeseed oil.
Figure 2. Possible fragmentation pattern and formation of positive
ions from canolol as detected in the MS spectra.
Comparison of Canolol Formation among Different
Varieties of HEM and Rapeseed during Roasting.
Roasting is generally practiced on either the seed form or the
flaked form. To study the effect of the physical integrity of the
seed, canolol formation was studied during both seed roasting
and seed powder roasting. The formation of canolol after
roasting could be observed in Figure 4. The amount of canolol
formed during roasting of different varieties of HEM seed and
seed powder is shown in Table 2. The data on the rapeseed are
also presented together for comparison. Canolol formation was
observed in all varieties, during both seed and seed powder
roasting; however, canolol formation was far lower in all of the
HEM varieties compared to that in rapeseed. The highest
canolol formation during seed roasting among HEM varieties
was observed in B. juncea, followed by B. nigra, B. juncea var.
oriental, and S. alba.
data.5,7,8,12 These observations fully confirmed the identification
of the isolated compound.
Canolol was previously isolated from roasted rapeseed oil
using different techniques. Most of these methods were long
and tedious. The first method consisted of the fractionation of
roasted rapeseed oil on a silica SPE column by eluting with
heptane/diethyl ether (90:10, v/v), followed by separation on a
C18 SPE column by eluting with methanol.8 The second
isolation method consisted of the extraction with methanol by
liquid−liquid partition, followed by fractionation with silica gel
chromatography by sequential elution with hexane/diethyl
ether (9:1, v/v), chloroform, acetone, and methanol. The
chloroform fraction was then separated by preparative TLC,
and purification was carried out by HPLC.5 Another method
comprised the extraction of roasted rapeseed with methanol
Figure 3. Synthesis of canolol from syringaldehyde (adapted from ref 10).
7509
dx.doi.org/10.1021/jf301738y | J. Agric. Food Chem. 2012, 60, 7506−7512

Journal of Agricultural and Food Chemistry
Article
Figure 4. Reverse-phase UHPLC chromatogram of the 70% methanol extract of the B. juncea var. oriental variety: (A) unroasted seed measured at
330 nm, (B) after basic hydrolysis measured at 330 nm, (C) 10 min roasted seed measured at 280 nm, and (D) unroasted seed measured at 280 nm.
The time axis is in minutes.
Table 2. Changes in FSA, TSAH, and Canolol Contents during Roasting of B. juncea (BJ), B. juncea var. oriental (BJO), B. nigra
(BN), S. alba (SA), and Rapeseed (RS)
a
a
a
variety
roasting condition
FSA (μg/g of DM)
TSAH (mg/g of DM)
canolol (μg/g of DM)
unroasted
26.55 1.86 ab
56.90 1.52 f
58.52 0.92 f
46.92 2.17 e
39.68 1.23 cd
42.41 0.45 d
77.18 4.11 g
26.50 0.72 ab
35.97 1.25 c
64.97 1.74 f
29.58 1.46 b
23.10 0.64 a
294.95 4.06 j
124.55 15.71 i
106.91 11.69 h
8.41 0.58 g
6.30 0.13 ef
7.49 0.11 f
6.72 0.09 f
5.38 0.06 bc
5.55 0.13 cd
5.63 0.12 cd
4.22 0.21 a
4.94 0.27 b
7.15 0.52 f
6.04 0.08 def
6.87 0.29 f
8.48 0.12 g
5.61 0.23 cd
10.83 0.11 a
212.28 5.14 i
159.15 2.00 h
BJ
seed roasted
seed powder roasted
unroasted
b
nd
BJO
BN
SA
seed roasted
135.56 2.07 g
118.97 2.83 f
seed powder roasted
unroasted
b
nd
seed roasted
143.00 2.14 g
95.50 5.86 e
seed powder roasted
unroasted
b
nd
seed roasted
75.96 2.19 d
56.67 0.97 c
24.74 0.45 b
707.69 50.68 j
790.41 50.78 k
seed powder roasted
unroasted
RS
seed roasted
seed powder roasted
5.88 0.31 cde
a
b
Values with different letters in the same column are significantly different (p < 0.05). nd = not detected.
Canolol formation has been proposed to be formed as a
varieties. Rapeseed contained 8.48 mg of TSAH/g of DM
(13.95 mg/g of fat-free mass), which was similar to the
previously reported value of 11.00−15.33 mg of total sinapic
acid equiv/g of defatted meal.13 The FSA content in the
different HEM varieties and rapeseed was 0.32−1.37 and 3.48%
TSAH, respectively. The FSA content was reported earlier to
be 0.86−3.76% of the total phenolics in defatted canola.13
After roasting, the loss of the TSAH content of all of the
varieties was observed. This loss was significant (p < 0.05) for
all of the varieties, except for S. alba. The FSA content also
decreased significantly (p < 0.05) after roasting, except for B.
juncea. However, the loss of FSA was far below the amount of
sinapic acid converted into canolol (the molecular weight ratio
of sinapic acid/canolol is 1.24). Therefore, esterified sinapic
acid derivatives must have provided FSA by hydrolysis during
roasting. The significant increase in FSA in B. juncea during
roasting also supported this fact (p < 0.05). The possible
contribution of sinapic acid derivatives on canolol formation
during roasting was also stated previously.7,9 The presence of a
result of decarboxylation of sinapic acid during roasting.5,7−9 It
is well-known that the major quantity of sinapic acid in both
mustard and rapeseed is present in the form of different
derivatives (sinapine, sinapoyl-hexoside, disinopoyl-dihexoside,
disinopoyl-hexoside, trisinapoyl-dihexoside, etc).8,11,13,14 There-
fore, both the FSA and TSAH contents were quantified. The
phenolic compound extraction was carried out with 70%
methanol, because a better extractability with this solvent
mixture was reported previously.13 After basic hydrolysis, the
majority of the peaks were lost, giving a single major peak of
sinapic acid, indicating that most of the peaks observed before
hydrolysis were sinapic acid derivatives (Figure 4).
The FSA and TSAH contents of all of the varieties during the
different roasting conditions are presented in Table 2. The FSA
content in different HEM varieties varied from 26.55 to 77.18
μg/g of dry matter (DM), while the rapeseed contained a
higher quantity of FSA (294.95 μg/g of DM). The TSAH
content varied between 5.63 and 8.41 mg/g of DM in all HEM
7510
dx.doi.org/10.1021/jf301738y | J. Agric. Food Chem. 2012, 60, 7506−7512

Journal of Agricultural and Food Chemistry
Article
higher quantity of FSA together with a higher loss of sinapic
acid derivatives in rapeseed during roasting supported the
formation of a higher quantity of canolol compared to other
HEM varieties.
to the differences in the oil content of the different seeds. The
fat content of rapeseed, B. juncea var. oriental, B. juncea, S. alba,
and B. nigra was 45, 40, 37, 32, and 30% DW basis, respectively.
The extracted canolol was observed to be more concentrated in
the samples with a lower oil content than in the samples with a
higher oil content. Moreover, only 49−56% of the canolol
formed during seed roasting was observed to be extracted in oil
in all of the varieties.
Canolol was confirmed to be formed in HEM varieties
during seed roasting. The formation of canolol during roasting
and its extraction in oil supports the increased oxidative stability
of roasted HEM seed oil.2 Canolol has been accepted to have
very good antioxidative activity.16,17 The proposed simple
procedure could be useful for the isolation of canolol from both
roasted HEM and rapeseed oils. The role of sinapic acid
derivatives in canolol formation was further illustrated. HEM
varieties were shown to produce a far lower amount of canolol
compared to rapeseed, and this difference was found to be due
to differences in the amount of the FSA content and a different
degree of hydrolysis of sinapic acid derivatives. Roasting is a
beneficial step for increasing the canolol content in the
extracted HEM seed oil and, hence, could increase oxidative
stability.
Canolol formation was highly correlated with the initial FSA
content (0.94), the residual FSA content after roasting (0.98),
and also the loss of TSAH (0.93) during seed roasting.
Similarly, canolol formation was highly correlated (0.96) with
all three of these parameters during seed powder roasting.
Hence, hydrolysis of sinapic acid derivatives into FSA favors
canolol formation during roasting. Canolol was shown to be
thermally unstable, and fast degradation during longer roasting
was previously reported.7 The degradation of canolol was
proposed to be due to its involvement in possible side reactions
with lipid peroxyl radicals formed during heat treatments,
pyrolysis, etc.7 Furthermore, it has recently been shown that
hydroxycinnamic acids (including sinapic acid) could inhibit
the Maillard reaction and color development by a radical
scavenging mechanism and reacting with intermediates of the
Maillard reaction.15 The major reaction pathway with
intermediates of the Maillard reaction was the formation of
vinylphenol by decarboxylation (e.g., canolol from sinapic acid),
which reacts with Maillard intermediates (such as 3-deoxy-2-
hexosulose) to generate phenolic Maillard adducts.15 The
possibilities of these various reactions during roasting support
the lower yield of canolol compared to the loss of TSAH. The
maximum yield (molar basis) of canolol compared to the
TSAH loss was observed for rapeseed (30.69%), while the yield
was below 12.62% for the different HEM varieties during seed
roasting. Seed powder roasting generally produced less canolol
than seed roasting of the same variety, except for rapeseed.
However, the yield (molar basis) of canolol compared to the
TSAH loss was slightly higher during seed powder roasting
than during seed roasting of the same variety. The highest yield
was again for rapeseed (37.84%), while the yield varied from
12.71 to 25.42% in different HEM varieties during seed powder
roasting. The higher conversion rate of sinapic acid to canolol
during seed powder roasting could be due to the easier loss of
carbon dioxide from the matrix in the powder form compared
to the intact seed, favoring the decarboxylation step.
ASSOCIATED CONTENT
■
S
* Supporting Information
Fatty acid composition data of the different mustard and
rapeseed varieties and the MS spectra of canolol. This material
is available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Telephone: +32-9-264-61-66. Fax: +32-9-264-62-15. E-mail:
bruno.demeulenaer@ugent.be.
Funding
This study was financially supported by the Special Research
Fund (BOF), Ghent University, which is gratefully acknowl-
edged.
Notes
The canolol content was analyzed in oil samples extracted
from roasted seeds of all of the varieties (Table 3). The highest
canolol content was observed for roasted rapeseed oil, while all
roasted HEM seed oils contained significantly less canolol (p <
0.05). Significant differences in the canolol content were also
observed between the HEM varieties (p < 0.05). The
differences in the ratio of canolol formed in the seed and
that observed in the oil between different varieties were linked
The authors declare no competing financial interest.
ABBREVIATIONS USED
■
DM, dry matter; FSA, free sinapic acid; HEM, high erucic
mustard; TSAH, total sinapic acid after basic hydrolysis
REFERENCES
■
(1) Przybylski, R.; Mag, T.; Eskin, N. A. M.; McDonald, B. E. Canola
oil. In Bailey’s Industrial Oil and Fat Products, 6th ed.; Shahidi, F., Ed.;
Wiley Interscience, a John Wiley and Sons, Inc., Publication: Hoboken,
NJ, 2005; pp 61−121.
(2) Vaidya, B.; Choe, E. Effects of seed roasting on tocopherols,
carotenoids, and oxidation in mustard seed oil during heating. J. Am.
Oil Chem. Soc. 2011, 88 (1), 83−90.
Table 3. Canolol Content in the Extracted Oil (μg/g of Oil)
from Unroasted and 10 min Roasted Seeds of B. juncea (BJ),
B. juncea var. oriental (BJO), B. nigra (BN), S. alba (SA), and
Rapeseed (RS)
a
canolol content (μg/g of oil) in the extracted oil
(3) Shahidi, F. Canola and Rapeseed: Production, Chemistry, Nutrition
and Processing Technology; Van Nostrand Reinhold: New York, 1990.
(4) Vaidya, B.; Choe, E. Stability of tocopherols and lutein in oil
extracted from roasted or unroasted mustard seeds during the oil
oxidation in the dark. Food Sci. Biotechnol. 2011, 20 (1), 193−199.
(5) Wakamatsu, D.; Morimura, S.; Sawa, T.; Kida, K.; Nakai, C.;
Maeda, H. Isolation, identification, and structure of potent alkyl-
peroxyl radical scavenger in crude canola oil, canolol. Biosci.,
Biotechnol., Biochem. 2005, 69 (8), 1568−1574.
variety
unroasted seed
roasted seed
BJ
5.41 0.01 c
5.06 0.15 c
0.48 0.03 a
1.25 0.01 b
5.19 0.09 c
297.76 6.26 g
171.64 3.20 e
117.29 5.68 d
266.21 4.81 f
808.48 24.67 h
BJO
SA
BN
RS
a
Values with different letters are significantly different (p < 0.05).
7511
dx.doi.org/10.1021/jf301738y | J. Agric. Food Chem. 2012, 60, 7506−7512

Journal of Agricultural and Food Chemistry
Article
(6) Kumazawa, S.; Koike, M.; Usui, Y.; Nakayama, T.; Fukuda, Y.
Isolation of sesaminols as antioxidative components from roasted
sesame seed oil. J. Oleo Sci. 2003, 52 (6), 303−307.
(7) Spielmeyer, A.; Wagner, A.; Jahreis, G. Influence of thermal
treatment of rapeseed on the canolol content. Food Chem. 2009, 112
(4), 944−948.
(8) Koski, A.; Pekkarinen, S.; Hopia, A.; Wahala, K.; Heinonen, M.
̈
̈
̈
Processing of rapeseed oil: Effects on sinapic acid derivative content
and oxidative stability. Eur. Food Res. Technol. 2003, 217 (2), 110−114.
(9) Wijesundera, C.; Ceccato, C.; Fagan, P.; Shen, Z. Seed roasting
improves the oxidative stability of canola (B. napus) and mustard (B.
juncea) seed oils. Eur. J. Lipid Sci. Technol. 2008, 110 (4), 360−367.
(10) Sinha, A. K.; Sharma, A.; Joshi, B. P. One-pot two-step synthesis
of 4-vinylphenols from 4-hydroxy substituted benzaldehydes under
microwave irradiation: A new perspective on the classical Knoevena-
gel−Doebner reaction. Tetrahedron 2007, 63 (4), 960−965.
(11) Engels, C.; Schieber, A.; Ganzle, M. Sinapic acid derivatives in
̈
defatted Oriental mustard (Brassica juncea L.) seed meal extracts using
UHPLC−DAD−ESI−MSn and identification of compounds with
antibacterial activity. Eur. Food Res. Technol. 2012, 234 (3), 535−542.
(12) Harbaum-Piayda, B.; Oehlke, K.; Sonnichsen, F. D.; Zacchi, P.;
̈
Eggers, R.; Schwarz, K. New polyphenolic compounds in commercial
deodistillate and rapeseed oils. Food Chem. 2010, 123 (3), 607−615.
(13) Khattab, R.; Eskin, M.; Aliani, M.; Thiyam, U. Determination of
sinapic acid derivatives in canola extracts using high-performance
liquid chromatography. J. Am. Oil Chem. Soc. 2010, 87 (2), 147−155.
(14) Thiyam, U.; Claudia, P.; Jan, U.; Alfred, B. De-oiled rapeseed
and a protein isolate: Characterization of sinapic acid derivatives by
HPLC−DAD and LC−MS. Eur. Food Res. Technol. 2009, 229 (5),
825−831.
(15) Jiang, D.; Peterson, D. Role of hydroxycinnamic acids in food
flavor: A brief overview. Phytochem. Rev. 2010, 9 (1), 187−193.
́
(16) Galano, A.; Francisco-Marquez, M.; lvarez-Idaboy, J. R. Canolol:
A promising chemical agent against oxidative stress. J. Phys. Chem. B
2011, 115 (26), 8590−8596.
(17) Kuwahara, H.; Kanazawa, A.; Wakamatu, D.; Morimura, S.;
Kida, K.; Akaike, T.; Maeda, H. Antioxidative and antimutagenic
activities of 4-vinyl-2,6-dimethoxyphenol (canolol) isolated from
canola oil. J. Agric. Food Chem. 2004, 52 (14), 4380−4387.
7512
dx.doi.org/10.1021/jf301738y | J. Agric. Food Chem. 2012, 60, 7506−7512