Fe2+-catalyzed formation of nitriles and thionamides from intact glucosinolates

Article abstract of DOI:10.1021/np070438d

The ratio of isothiocyanates to nitriles formed upon the hydrolysis of glucosinolates is a key factor determining the physiological effect of glucosinolate-containing plants and materials. A micellar electrokinetic capillary chromatography (MECC) method was used to study the nonenzymatic Fe²⁺-catalyzed transformation of glucosinolates. At room temperature, pH 5, and in the presence of only 2 molar excess of Fe²⁺ all glucosinolate was degraded in 24 h. At all molar excess Fe²⁺ tested, nitriles were the major compounds formed. Thionamides were also formed from glucosinolates that contained a side chain hydroxylated at C-2; in this case, trace amounts of oxazolidine-2-thione were also detected. The presence of Fe³⁺ had no effect. The nonenzymatic Fe²⁺-catalyzed transformation of glucosinolates involves the binding of Fe²⁺ to the glucosinolate to form a complex.

Full text of DOI:10.1021/np070438d

76
J. Nat. Prod. 2008, 71, 76–80
Fe2+-Catalyzed Formation of Nitriles and Thionamides from Intact Glucosinolates
Natalia Bellostas,† Anne D. Sørensen,‡ Jens C. Sørensen,† and Hilmer Sørensen*,†
Department of Natural Sciences, Biochemistry and Natural Product Chemistry, Faculty of Life Sciences, UniVersity of Copenhagen,
ThorValdsensVej 40, DK-1871, Frederiksberg C, Denmark, and Department of Human Nutrition, Paediatric and International Nutrition,
Faculty of Life Sciences, UniVersity of Copenhagen, RolighedsVej 30, DK-1958, Frederiksberg C, Denmark
ReceiVed August 21, 2007
The ratio of isothiocyanates to nitriles formed upon the hydrolysis of glucosinolates is a key factor determining the
physiological effect of glucosinolate-containing plants and materials. A micellar electrokinetic capillary chromatography
(MECC) method was used to study the nonenzymatic Fe2+-catalyzed transformation of glucosinolates. At room
temperature, pH 5, and in the presence of only 2 molar excess of Fe2+ all glucosinolate was degraded in 24 h. At all
molar excess Fe2+ tested, nitriles were the major compounds formed. Thionamides were also formed from glucosinolates
that contained a side chain hydroxylated at C-2; in this case, trace amounts of oxazolidine-2-thione were also detected.
The presence of Fe3+ had no effect. The nonenzymatic Fe2+-catalyzed transformation of glucosinolates involves the
binding of Fe2+ to the glucosinolate to form a complex.
Glucosinolates are alkyl-N-hydroximine sulfate esters with a ꢀ-D-
thioglucopyranoside group attached to the hydroximine carbon in
Z-configuration to the sulfate group.1–3 Glucosinolates constitute a
group of more than 140 structurally different compounds with well-
defined chemotaxonomic occurrence in all plants of the order
Capparales and in a few other plants.4–6 Glucosinolates co-occur
with myrosinase isoenzymes (thioglucohydrolases; EC 3.2.1.147),7,8
which catalyze the hydrolysis of the ꢀ-D-thioglucopyranoside bond,
giving rise to a variety of compounds with different structures,
depending on the parent glucosinolate and the reaction conditions.3,9,10
Isothiocyanates are often the quantitatively dominant compounds
that are formed in the myrosinase-catalyzed hydrolysis of glucosi-
nolates at neutral pH, and they have, therefore, been the most
frequently studied glucosinolate hydrolysis products.1,3,9,11,12 How-
ever, glucosinolate hydrolysis can also occur via a nonenzymatic
mechanism catalyzed by Fe2+, and in this case a nitrile, instead of
an isothiocyanate, is yielded. In the case of glucosinolates hydroxy-
lated at C-2 in the side chain, a thionamide is formed.13–16
Glucosinolate degradation products have been studied widely
because of their biological activity, which includes antinutritional
effects on monogastric farm animals,17,18 beneficial effects on
human health,19,20 and fungicidal, herbicidal, and nematocidal
properties.21,22 Isothiocyanates are generally regarded as the most
active compounds; however, they are often volatile, hydrophobic,
and very reactive, which makes them very short-lived and limits
the time span in which they have an effect.23,24 Nitriles are more
stable products; however their biological activity is generally lower
than their equivalent isothiocyanates.6,25,26
glucosinolates and the product formation online.28 Our aim was to
determine the influence of different concentrations of Fe2+ on the
quantity and types of products formed from various glucosinolates
in the absence of myrosinase at pH 5. The effect of Fe3+ was also
tested. In order to elucidate the mechanism responsible for the
degradation of glucosinolates and the formation of transformation
compounds, we studied the oxidation of Fe2+ during the reaction.
Results and Discussion
Glucosinolate Transformation Products Obtained by the
Nonenzymatic Degradation of Glucosinolates Catalyzed by
Fe2+. Fe2+ catalyzed the nonenzymatic degradation of glucosibarin
at pH 5, giving rise to both thionamide and nitrile. The degradation
of glucosinolates in acidic solutions that contain Fe2+ has been
reported for sinigrin,16 epi-progoitrin,13–15 progoitrin, and gluco-
barbarin.14,29 The formation of thionamide has been documented
only from glucosinolates hydroxylated at C-2 in the side chain.13–15,29
The major product from the myrosinase-catalyzed hydrolysis of
glucosibarin at pH 5 is generally (5S)-5-phenyloxazolidine-2-thione
(OZT).6,28 In our study, this compound was produced only in trace
amounts, since it was only detected when a 50-fold injection volume
was used (data not shown).
The MECC method used28 proved to be valuable for following
the hydrolysis of glucosibarin by Fe2+ in the absence of myrosinase.
Figure 1 shows an electropherogram of the reaction in which
glucosibarin (e) appears simultaneously with nitrile (b) and thiona-
mide (c). The UV spectra of these compounds are available,14,28
which permitted their immediate identification. In the course of
the reaction, an additional peak (d) appeared between the degrada-
tion products and glucosibarin. According to its electrophoretic
mobility, it was likely to be an uncharged compound of higher
molecular weight than the thionamide or the nitrile. These char-
acteristics, together with the fact that it was unstable (see below),
indicated that it could be a Fe2+-glucosinolate complex with one
hydrated acidic Fe2+ and either one or two glucosinolates.16
Furthermore, the compound had a UV spectrum similar to a glu-
cosinolate with a maximum absorbance at 225–230 nm, which
indicates that it had retained the characteristic glucosinolate core
structure (CdN conjugated with S), although slightly changed with
respect to the parent glucosinolate (e).
Understanding the factors that influence the formation of nitriles
or isothiocyanates upon hydrolysis of glucosinolates is of great
importance in controlling and predicting the biological effects of
glucosinolate-containing plants and materials. In particular, the study
of the nonenzymatic formation of nitriles by Fe2+ is relevant, since
iron is often present in machinery, agricultural soil, food, and feed,27
and intact glucosinolates can persist because of natural (e.g., low
pH in the stomach or soil) or artificial (e.g., food processing)
inactivation of myrosinase.
We studied the nonenzymatic formation of nitriles using a
recently developed micellar electrokinetic capillary chromatography
(MECC) method that allowed us to follow the hydrolysis of
* To whom correspondence should be addressed. Tel: +45 3533
2432.Fax: +45 3533 2398. E-mail: hils@life.ku.dk.
† Department of Natural Sciences, Biochemistry and Natural Product
Chemistry.
Glucosibarin degradation started immediately after adding 1 M
excess Fe2+, and the reaction proceeded during approximately 120
min (Figure 2), after which no significant changes in the concentra-
tion of glucosibarin, nitrile, and thionamide were observed. At this
‡ Department of Human Nutrition, Paediatric and International Nutrition.
10.1021/np070438d CCC: $40.75
2008 American Chemical Society and American Society of Pharmacognosy
Published on Web 12/29/2007

Fe2+-Catalyzed Formation of Nitriles from Glucosinolates
Journal of Natural Products, 2008, Vol. 71, No. 1 77
Figure 1. Electropherogram showing the simultaneous appearance of the internal standard [TNA (a)], the products [nitrile (b), thionamide
(c)], the postulated glucosibarin-Fe2+complex (d), and glucosibarin (e). The UV spectra of the different compounds are also shown.
Figure 3. Molar yield (in %) of nitrile, thionamide, the Fe2+
-
glucosibarin complex, and glucosibarin after 24 h reaction with
different molar excesses of Fe2+. Each point is the average of three
replicates.
Figure 2. Rates of glucosibarin degradation and formation of the
nitrile, thionamide, and the glucosibarin-Fe2+complex in the pres-
ence of 1 M excess of Fe2+ with respect to glucosibarin. Each point
is the average of two runs.
solutions as pH increases,30 and in our experiments, we found that
approximately half of the total iron in the stock solutions was in
the oxidized form at pH 5. This shows the importance of deter-
mining the exact amount of Fe2+ in the reaction medium and may
explain the differences found between the present and previous
studies regarding the Fe2+ molar excess needed for total degradation
of the glucosinolate.
Fe2+ molar excess, the Fe2+-glucosinolate complex achieved max-
imum concentration at around 180 min.
The nitrile was, quantitatively, the major product formed,
accounting for nearly 90% of the total pool of degradation products
after 24 h, independent of the Fe2+:glucosinolate ratio (Figure 3).
Nitrile was reported as the major product of the nonenzymatic Fe2+
-
catalyzed degradation of progoitrin at room temperature and pH
Increasing the concentration of Fe2+ increased the velocity of
the reaction, and in the presence of an 8 M excess Fe2+, only 30%
of the initial amount of glucosinolate remained after 10 min (data
not shown), representing a 3-fold increase in the degradation rate
with respect to 0.25 M excess Fe2+. The increased velocity of
reaction by increasing concentrations of Fe2+ has been shown both
in vitro16 and in vivo.31 Increasing the molar excess of Fe2+ from
0.5 to 2 increased the concentration of the complex in the mixture,
although not proportionally to the increase of Fe2+ (data not shown).
From 2 to 8 M excess Fe2+, however, the concentration of the com-
plex in the mixture remained constant (Figure 3).
Similar results were obtained when other glucosinolates were
tested. Both gluconasturtiin and progoitrin were shown to be
degraded by Fe2+. Nonenzymatic hydrolysis of gluconasturtiin gave
rise only to nitrile, and as expected, no thionamide was formed
because of the absence of a hydroxyl group in the side chain. The
expected thionamide was observed in the progoitrin hydrolysis, as
previously shown.15,16,29
29
5.2 with 10 M excess Fe2+
.
On the other hand, Austin et al.15
reported thionamide as the major product from epi-progoitrin and
glucobarbarin at 7 M excess Fe2+, room temperature, and pH 5.3,
although the yield decreased with incubation time. They also found
that, at 95 °C and with lower concentrations of Fe2+, the proportion
of nitrile increased with respect to thionamide. We cannot provide
an explanation as to why nitrile was the major transformation
product in our experiments. However, in contrast to the methods
used in previous studies,13–16,29 the MECC method presently used
allowed us to follow the reaction online;28 thus the risk for artifact
production or product degradation associated with the extraction
of compounds from the reaction medium was eliminated.
Complete glucosinolate degradation at room temperature has been
14,29
reported to occur only at 6 to 8 M excess Fe2+
,
whereas at
higher temperatures (95 °C), equimolar amounts are sufficient.14,16
In the present study, 2 M excess Fe2+ degraded all glucosibarin in
24 h at 20 °C (Figure 3). Fe2+ readily oxidizes to Fe3+ in aqueous

78 Journal of Natural Products, 2008, Vol. 71, No. 1
Bellostas et al.
recent report showed that the Fe2+-chelating activity of broccoli
(reported as an EDTA equivalent) was the highest among a number
of vegetables tested.34
Mechanism of Reaction. Fe2+ did not degrade the desulfo form
of glucosibarin when tested identically as the other intact glucosi-
nolates in the MECC assay. This shows that the presence of the
sulfate group in the glucosinolate is decisive for the formation of
the complex, as previously stated.16 Figure 5 shows the proposed
mechanism for the nonenzymatic Fe2+-catalyzed glucosinolate
degradation. The ligands in the glucosinolate-Fe2+ complex for the
formation of the nitrile (Figure 5, left) are expected to be the S in
the thioglucose and the O in the sulfate group;16 the O at C-2 of
the glucose could also participate in the complex. For the formation
of the thionamide (Figure 5, right) the ligands are expected to be
the O of the hydroxyl group at C-2 and the N in the sulfate group.
The mechanism of reaction for thionamide formation involves the
breakage of the thioglucoside bond between the S and the glucose
(as for the myrosinase-catalyzed glucosinolate hydrolysis; Figure
5), and it requires the delivery of two redox equivalents for the
formation of the thionamide. Breakage of this bond could also
explain the trace amounts of OZT found. Nitrile formation involves
the breakage of the bond between the glucosinolate carbon C-0
and S in the thioglucoside bond (Figure 5), which also requires the
delivery of two redox equivalents liberating thioglucose.
Figure 4. Variation in the concentration of Fe2+ over time (24 h
incubation) in a control sample (no glucosinolate added) compared
to a 2:1 Fe2+:glucosinolate sample.
Fe3+ has been reported to promote nitrile formation in vivo,
although to a lesser extent than Fe2+ 31 however, the in vitro results
;
are contradictory. FeCl3 promoted the decomposition of sinigrin
to allyl cyanide,16 but did not degrade epi-progoitrin.14 In our
experiments, FeCl3 did not degrade any of the glucosinolates when
tested by the MECC method under the same conditions as the FeSO4
experiments.
Identification of the Reaction Products. Glucosinolate degra-
dation by Fe2+-mediated catalysis has been postulated to occur
through the formation of a complex.16 The additional peak that
appeared during glucosibarin hydrolysis (Figure 1, d) was repeatedly
observed at different Fe2+ concentrations as well as when glucosi-
nolates other than glucosibarin were used. At concentrations of Fe2+
e 20 mM (equal or lower than 1 M excess), the peak disappeared
after 24 h, although it remained in the mixture when 2 M excess
Fe2+, or more, was added (Figure 3).
Youngs and Perlin16 suggested that the complex of Fe2+ with 2
mol of sinigrin provides a route for a concerted electron transfer,
one phase of which involves the formation of the S-S bond
between two thioglucoses. In our experiment, thioglucose and not
the disulfide form was found in the reaction medium after 24 h.
We propose that the reducing conditions in our experiments protect
the thioglucose from oxidation to the disulfide form.
Practical Implications. Our experiments have shown that
nitriles are the quantitatively major products formed nonenzymati-
cally from intact glucosinolates at mild conditions of temperature
and pH in the presence of as little as 0.25 M excess Fe2+. Fe2+ and
intact glucosinolates are often present simultaneously during the
analysis, handling, processing, and consumption of glucosinolate-
containing materials. Thus, the possibility of nitrile formation
through Fe2+-catalyzed degradation of intact glucosinolates should
be acknowledged and its implications considered when assessing
the biological effects of glucosinolate-containing materials.
To identify this compound, high-voltage electrophoresis (HVE)10,32
was performed. After incubation, the reaction medium to which
Fe2+ was added did not contain any residue of intact glucosibarin.
Instead, a compound with no electrophoretic mobility that still
retained the sulfur atom was present (data not shown). Preparative
HVE allowed for the isolation of the neutral compounds formed in
the reaction, but MECC of the sample prior to NMR revealed that
the Fe2+-glucosinolate complex had been degraded during purifica-
tion. Apart from the expected nitrile, the neutral fraction of the
HVE electropherogram was shown to contain thioglucose. 1H NMR
spectroscopy showed the complex pattern at δ 3.2–3.9 that is
characteristic of glucose plus a doublet at δ 4.4 for C-1′, whereas
13C NMR spectroscopy showed the following chemical shifts (δ):
C-1′ 89.7, C-2′ 77.3, C-3′ 77.5, C-4′ 71.5, C-5′ 80.6, C-6′ 62.9
(DEPT). The 13C NMR resonances found were in agreement with
those obtained with the use of a commercial sample of thioglucose
(data not shown) and with those described previously for its Na
salt.33
The presence of thioglucose instead of glucose as a product of
the Fe2+-catalyzed glucosinolate degradation has previously been
reported.16 Youngs and Perlin16 found only trace amounts of
glucose after the degradation of sinigrin by Fe2+, and half the
amount of the sugar unit was recovered as bis(ꢀ-D-glucopyranosyl)
disulfide.
Oxidation of Fe2+. The presence of glucosibarin prevented the
oxidation of Fe2+ to Fe3+ in the first hours of the reaction when a
2:1 ratio of Fe2+:glucosinolate was used (Figure 4). At this Fe2+
molar excess, the difference between the Fe2+ concentration in the
samples, with and without glucosinolate, decreased with the
incubation time. In an aqueous aerobic solution under weakly acidic
and alkaline conditions, oxygen can convert Fe2+ to Fe3+ spontane-
ously because of the redox potential of iron.30 Our results show
that the transient complex formed between the glucosinolate and
Fe2+ has the ability to shield the Fe2+ and prevent its oxidation. A
Experimental Section
General Experimental Procedures. Disodium hydrogenphosphate
was purchased from Riedel-de Häen (Seelze, Germany), ascorbic acid
from Bie & Berntsen (Rødovre, Denmark), and ferrous sulfate
heptahydrate from Merck (Darmstadt, Germany). Water was purified
using a Milli-Q system (Millipore, Bedford, MA). Sodium cholate,
taurine, trigonellinamide (TNA), ferric chloride hexahydrate, HEPES
buffer (N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid, sodium
salt), trichloroacetic acid (TCA), chlorhydric acid, hydroxylamine
monohydrochloride, and ferrozine chromogen [3-(2-pyridyl)-5,6 bis(4-
phenylsulfonic acid)-1,2,4-triazine, disodium salt] were purchased from
Sigma-Aldrich (Steinheim, Germany). The intact glucosinolates (glu-
cosibarin, progoitrin, and gluconasturtiin) and the desulfo glucosinolate
(desulfoglucosibarin) used in the experiment were from our laboratory
collection.35,36
MECC analyses were performed in a Hewlett-Packard HP3D capillary
electrophoresis (CE) system (Agilent, Waldbronn, Germany) that was
equipped with a diode array detector. For data processing, we used a
HP Vectra 5/100 mHz Pentium with HP ChemStation Rev. B. 01.03.
1H and 13C 1D NMR spectra were recorded at room temperature on
a Bruker Advance 400 NMR spectrometer using TMS as reference.
Measurement of Fe2+ Ions in the Solutions. Prior to utilization
in the MECC assay, each new Fe2+ stock solution was analyzed for
Fe2+ and total iron to determine the exact volume of solution that was
needed in the assay. The method that we used is based on the reaction
with the chromogene ferrozine,37 modified from Kapsokefalou and
Miller,38 and adapted for microcale use in 96-well microplates. Fe2+
was analyzed by adding 100 µL of nonreducing protein precipitate

Fe2+-Catalyzed Formation of Nitriles from Glucosinolates
Journal of Natural Products, 2008, Vol. 71, No. 1 79
Figure 5. Proposed mechanism for nitrile (left) and thionamide formation (right) through the Fe2+-catalyzed degradation of glucosinolates
containing a hydroxyl group at C-2 in the side chain. The figures include only one of the two glucosinolates supposedly present in the
complex.16 See text for details.
solution (non-RPP, 1 g of trichloroacetic acid (TCA), 1 mL of 37%
HCl adjusted to 10 mL with water) to 200 µL aliquots. Total iron was
analyzed by adding 100 µL of reducing protein precipitate solution
(RPP, 1 g of TCA, 1 mL of 37% HCl, 0.5 g of hydroxylamine
monohydrochloride adjusted to 10 mL with water) to 200 µL aliquots
of sample or to ferrous sulfate in 2-fold dilutions starting at 1 mM
FeSO4 for quantification. The aliquots, which contained either non-
RPP or RPP, were left overnight at room temperature and centrifuged
(2575g; 10 min) prior to use. Duplicates of 100 µL supernatants were
placed in microtiter wells and then mixed with 200 µL of HEPES buffer
(0.3 M, pH 9.9) and 25 µL of ferrozine chromogen solution (5 mg/mL
in water). The absorbance was measured in a microplate reader (Bio
kinetic reader EL 340 Microplate; Bio-Tek Instruments; Software: KC3;
KinetiCalc for Windows, version 1.5) at 570 nm immediately after
ferrozine addition for the quantification of Fe2+ and after 1 h for
determination of total iron content.
MECC Reaction Procedure. The formation of the nitrile, the
thionamide, and the complex from the different glucosinolates inves-
tigated was followed online with the method developed by Bellostas
et al.28 The MECC run buffer was composed of 35 mM sodium cholate,
100 mM disodium phosphate, 500 mM taurine, and 2% 1-propanol,
and the pH was kept at 8.2. A solution of 0.8 M citric acid was adjusted
to pH 5.5 ((0.1), and ferrous sulfate was added to a concentration of
0.4 M. The final pH of the ferrous sulfate stock solution ranged in all
cases between 4.5 and 5. The content of Fe2+ in this solution was
determined by the procedure described above, which allowed us to
calculate the volume that was needed to have the desired µmoles of
Fe2+ in the reaction medium. This solution was prepared freshly every
time a new determination was conducted. The reaction medium was
composed of TNA (as an internal standard; 20 µL, 100 mM), glu-
cosinolate (20 µL, 50 mM), Milli-Q water (6.5 µL), and acetate buffer
(3.5 µL, 100 mM, pH 5), to which different volumes of the above-
mentioned FeSO4 solution were added. The final pH of the reaction
medium was in no case lower than 4.5. The measurements were
conducted in duplicate. Concentrations of glucosibarin and ni-
trile (phenylacetonitrile instead of 3-hydroxyphenylproprionitrile was
used) were determined by the use of concentration–response curves of
the pure compounds at 206 nm in MECC (cholate buffer, 30 °C). The
E value for TNA at 206 nm was calculated by UV spectroscopy
(Shimadzu MPS-2000 UV–visible light spectrophotometer). Molar
response factors of the complex and the thionamide were assumed to
be identical to that of the parent glucosinolate at 206 nm.
Preparative HVE was conducted to isolate the neutral degradation
products. The reaction was conducted in a large volume by mixing
glucosibarin (3 mg) with twice the molar amount of Fe2+ at pH 5 and
allowing it to stand overnight at room temperature. The reaction mixture
was checked by MECC for the presence of the complex previous to
preparative HVE. Once HVE was run, the band where the neutral
compounds appeared was washed out with water, the wash-out
1
evaporated, and the residue redissolved in D2O prior to H and 13C
NMR spectroscopy.
Measurement of the Oxidation of Fe2+ in the Presence of
Glucosinolate. The reaction mixture consisted of glucosibarin
(40 µL, 10 mM) and acetate buffer (1 mL, 100 mM, pH 5), and the
volume was adjusted to 4 mL with MilliQ-water. The control sample
consisted of acetate buffer (1 mL, 100 mM) and Milli-Q water (3 mL).
To both samples, FeSO4 (7.2 µL, 100 mM) was added, and the pH
was adjusted to 5. Aliquots (200 µL) were collected for measurement
of Fe2+ in solution (vide supra), at times corresponding to 0, 10, 30,
60, 120, and 180 min and 24 h.
Acknowledgment. The Commission of the European Union (FP-
6-NovelQ 015710-2) is gratefully acknowledged for the financial
support of this work.
References and Notes
(1) Ettlinger, M. G.; Kjær, A. In Recent AdVances in Phytochemistry;
Mabry, T. J., Alston, R. E., Runeckles, V. C., Eds.; North Holland
Publishing Company: New York, 1968; Vol. 1, Chapter 3, pp 60–
144.
(2) Kjær, A. In Progress in the Chemistry of Organic Natural Products;
Herz, W., Grisebach, H., Kirby, G. W., Eds.; Springer-Verlag: Berlin,
1960; pp 122–176.
(3) Sørensen, H. In Canola and Rapeseed; Shahidi, F., Ed.; Van Nostrand
Reinhold: New York, 1990; Chapter 9 pp 149–172.
(4) Fahey, J. W.; Zalcmann, A. T.; Talalay, P. Phytochemistry 2001, 56,
5–51.
(5) Kjær, A. In The Biology and Biochemistry of the Cruciferae; Vaughan,
J. G., MacLeod, A. J., Jones, B. M. G., Eds.; Academic: London,
1976; pp 207–219.
(6) Bellostas, N.; Sørensen, A. D.; Sørensen, J. C.; Sørensen, H. In
Rapeseed Breeding: AdVances in Botanical Research; Gupta, S., Ed.;
Academic Press/Elsevier: San Diego, 2007; pp 369–415.
(7) Bellostas, N.; Jørgensen, A. L. W.; Lundin, N. V. F.; Petersen, I. L.;
Sørensen, H.; Sørensen, J. C.; Sørensen, R.; Tidman, K. D. Proceed-
ings of the 11th International Rapeseed Congress; Copenhagen,
Denmark, 2003; Vol. 2 pp 720–723.
(8) Bussy, A. J. Pharm. Sci. Accessoires 1840, 26, 39–48.
(9) Bones, A. M.; Rossiter, J. T. Phytochemistry 2006, 67, 1053–1067.
(10) Olsen, O.; Sørensen, H. J. Am. Oil Chem. Soc. 1981, 58, 857–865.
(11) Daxenbichler, M. E.; VanEtten, C. H.; Spencer, G. F. J. Agric. Food
Chem. 1977, 25, 121–124.
(12) Song, L.; Morrison, J. J.; Botting, N. P.; Thornalley, P. J. Anal.
Biochem. 2005, 347, 234–243.
(13) Austin, F. L.; Gent, C. A. Chem. Commun. 1967, 2, 71–72.
(14) Austin, F. L.; Gent, C. A.; Wolff, I. A. Can. J. Chem. 1968, 46, 1507–
1512.
Identification of the Degradation Compounds. HVE10,32 was
performed to determine the presence of the Fe2+ complex. Two aliquots
of the MECC reaction medium (described above), one with Fe2+ (2 M
excess) and the other one without Fe2+, were incubated overnight at
20 °C. Previous to HVE, the sample to which Fe2+ was added was
checked by MECC to ensure that the complex was present. The HVE
was carried out in Whatman 3 MM paper in a flat plate unit. The
samples were applied with a capillary to the paper in three replicates
with increasing volumes (one, three, and six applications). Sinigrin was
used as a reference compound. The run buffer was glacial acetic acid,
formic acid, and water (4:1:45) at pH 1.9, and the electrophoresis was
run for 1 h at 3 kV. The presence of glucosinolates was determined by
silver nitrate staining.10,32
(15) Austin, F. L.; Gent, C. A.; Wolff, I. A. J. Agric. Food Chem. 1968,
16, 752–755.
(16) Youngs, C. G.; Perlin, A. S. Can. J. Chem. 1967, 45, 1801–1804.

80 Journal of Natural Products, 2008, Vol. 71, No. 1
Bellostas et al.
(17) Bille, N.; Eggum, B. O.; Jacobsen, I.; Olsen, O.; Sørensen, H. Z.
Tierphysiol. Tierernährung Futtermittel. 1983, 49, 195–210.
(18) Bjerg, B.; Eggum, B. O.; Jacobsen, I.; Otte, J.; Sørensen, H. Z.
Tierphysiol. Tierernährung Futtermittel. 1989, 61, 227–244.
(19) Bonnesen, C.; Stephensen, P. U.; Andersen, O.; Sørensen, H.; Vang,
O. Nutr. Cancer 1999, 33, 178–187.
(20) Fahey, J. W.; Zhang, Y. S.; Talalay, P. P. Natl. Acad. Sci. U.S.A.
1997, 94, 10367–10372.
(21) Brown, P. D.; Morra, M. J. AdV. Agron. 1997, 61, 167–231.
(22) Rosa, E. A. S.; Heaney, R. K.; Fenwick, R. G.; Portas, C. A. M. In
Horticultural ReViews; Janick, J., Ed.; John Wiley & Sons, Inc.: New
York, 1997; Vol. 19,Chapter 3, pp 99–215.
(23) Björkman, R. Phytochemistry 1973, 12, 1585–1590.
(24) Morra, M. J.; Kirkegaard, J. A. Soil Biol. Biochem. 2002, 34, 1683–
1690.
(25) Matusheski, N. V.; Jeffery, E. H. J. Agric. Food Chem. 2001, 49,
5743–5749.
(26) Staack, R.; Kingston, S.; Wallig, M. A.; Jeffery, E. H. Toxicol. Appl.
Pharmacol. 1998, 149, 17–23.
(28) Bellostas, N.; Sorensen, J. C.; Sorensen, H. J. Chromatogr. A 2006,
1130, 246–252.
(29) MacLeod, A. J.; Rossiter, J. T. Phytochemistry 1986, 25, 855–858.
(30) Spiro, T. G.; Saltman, P. In Iron in Biochemistry and Medicine; Jacobs,
A., Worwood, M., Eds.; Academic Press: London/New York, 1974;
pp 1–28.
(31) Hasapis, X.; MacLeod, A. J. Phytochemistry 1982, 21, 559–563.
(32) Sørensen H.; Sørensen, S.; Bjergegaard, C.; Michaelsen, S. Chroma-
tography and Capillary Electrophoresis in Food Analysis; The Royal
Society of Chemistry: Cambridge, UK, 1999; p 183.
(33) Olsen, O.; Sørensen, H. Phytochemistry 1979, 18, 1547–1552.
(34) Zhou, K.; Yu, L. LWT-Food Sci. Technol. 2006, 39, 1155–1162.
(35) Bjerg, B.; Sørensen, H. In Glucosinolates in Rapeseeds: Analytical
Aspects; Wathelet, J. P., Ed.; Martinus Nijhoff Publishers: Dordrecht,
1987; pp 59–75.
(36) Sørensen, J. C. Biorefined Oilseed Products. Ph D. Thesis,The Royal
Veterinary and Agricultural University,Copenhagen, Denmark, 2001,
p 120.
(27) Wood, R. J.; Ronnenberg, A. G. In Modern Nutrition in Health and
Disease; Shils, M. E., Shike, M., Ross, A. C., Caballero, B., Cousins,
R. J., Eds.; Lippincott Williams & Wilkins: London, 2006; pp 248–
270.
(37) Stookey, L. L. Anal. Chem. 1970, 42, 779–781.
(38) Kapsokefalou, D. D.; Miller, D. D. J. Food Sci. 1991, 56, 352–358.
NP070438D