Determination of S-methyl-, S-propyl-, and S-propenyl-L-cysteine sulfoxides by gas chromatography-mass spectrometry after tert-butyldimethylsilylation.

Article abstract of DOI:10.1021/jf020166e

A gas chromatographic-mass spectrometric method for the determination of S-methyl-L-cysteine sulfoxide (1), S-propyl-L-cysteine sulfoxide (2), and S-propenyl-L-cysteine sulfoxide (3), specific marker compounds in the genus Allium, is described. The target amino acids were converted to the tert-butyldimethylsilyl derivatives. The products were silylated on the amino and carboxyl groups and on an additional oxygen atom and were separated on a nonpolar capillary column. That incorporation of three tert-butyldimethylsilyl groups had occurred was verified by mass spectrometry, which gave an m/z 302 fragment as base peak (amino acid side chain eliminated ion) and m/z 436 (1), 464 (2), or 462 (3) as major peaks (tert-butyl function eliminated ion), by electron impact ionization. The detection limits for 1 and 2 under selected ion monitoring at m/z 436 (1) and m/z 464 (2), respectively, were determined to be 0.3 and 1.8 ng per injection. To clean up the analytes from the solvent extract of onion, as a representative food material, onion, the sample solution was subjected to combined solid phase extraction. The eluate from a Sep-Pak C(18) cartridge was applied to a Bond Elut SCX cartridge (H(+) form), followed by washing with 0.1 M hydrochloric acid and elution with 0.5 M ammonia. From a simulated matrix solution containing 5% sucrose, 1 and 2 were extracted quantitatively, and the detection yield was approximately 75%. The contents of 1, 2, and 3 in commercial onion were estimated to be 0.3, 3.1, and 3.0 mg, respectively, per gram of fresh weight.

Full text of DOI:10.1021/jf020166e

J. Agric. Food Chem. 2002, 50, 4445−4451 4445
Determination of S-Methyl-, S-Propyl-, and
S-Propenyl-L-Cysteine Sulfoxides by Gas
Chromatography−Mass Spectrometry after
tert-Butyldimethylsilylation
KOUICHIRO TSUGE, MIEKO KATAOKA, AND YASUO SETO*
National Research Institute of Police Science, 6-3-1 Kashiwanoha, Kashiwa, Chiba 277-0882, Japan
A gas chromatographic-mass spectrometric method for the determination of S-methyl-L-cysteine
sulfoxide (1), S-propyl-L-cysteine sulfoxide (2), and S-propenyl-L-cysteine sulfoxide (3), specific marker
compounds in the genus Allium, is described. The target amino acids were converted to the tert-
butyldimethylsilyl derivatives. The products were silylated on the amino and carboxyl groups and on
an additional oxygen atom and were separated on a nonpolar capillary column. That incorporation of
three tert-butyldimethylsilyl groups had occurred was verified by mass spectrometry, which gave an
m/z 302 fragment as base peak (amino acid side chain eliminated ion) and m/z 436 (1), 464 (2), or
462 (3) as major peaks (tert-butyl function eliminated ion), by electron impact ionization. The detection
limits for 1 and 2 under selected ion monitoring at m/z 436 (1) and m/z 464 (2), respectively, were
determined to be 0.3 and 1.8 ng per injection. To clean up the analytes from the solvent extract of
onion, as a representative food material, onion, the sample solution was subjected to combined solid
phase extraction. The eluate from a Sep-Pak C18 cartridge was applied to a Bond Elut SCX cartridge
(H+ form), followed by washing with 0.1 M hydrochloric acid and elution with 0.5 M ammonia. From
a simulated matrix solution containing 5% sucrose, 1 and 2 were extracted quantitatively, and the
detection yield was ∼75%. The contents of 1, 2, and 3 in commercial onion were estimated to be
0.3, 3.1, and 3.0 mg, respectively, per gram of fresh weight.
KEYWORDS: S-Alk(en)yl-L-cysteine sulfoxide; onion; GC-MS; solid phase extraction; tert-butyldimethyl-
silylation
INTRODUCTION
with o-phthalaldehyde (7) or 9-fluorenylmethyl chloroformate
(8). In general, for an analysis of such hydrophilic compounds,
GC-MS after tert-butyldimethylsilylation (TBDMS) derivati-
zation appears to be the method of choice, and a number of
papers have indicated its utility in the analysis of amino acids
(9-11). Our laboratory has applied a TBDMS GC-MS method
for the analysis of nerve gas hydrolysis products (12-14). In
this paper, we report on a method for the determination of
S-methyl-L-cysteine sulfoxide (1), S-propyl-L-cysteine sulfoxide
(2), and S-propenyl-L-cysteine sulfoxide (3), the main S-
alk(en)yl-L-cysteine sulfoxide components in Allium, together
with a pretreatment method using solid phase extraction for the
cleanup of S-alk(en)yl-L-cysteine sulfoxides from the extract
of onion, one of the most popular food materials belonging to
the Allium genus.
Food materials contain characteristic components that are
species-specific. To verify the existence of a specific type of
food material in foodstuff, the identification of marker com-
ponents that are specific to the materials is necessary. S-
Alk(en)yl-L-cysteine sulfoxides are the specific components in
the genus Allium (1, 2). Volatile alkyl disulfides produced from
S-alk(en)yl-L-cysteine sulfoxides by CS-lyase are the main
source of the characteristic odor of Allium species. Quantifica-
tion of S-alk(en)yl-L-cysteine sulfoxides is an important issue
in plant physiology investigations. A variety of methods have
been developed for the determination of S-alk(en)yl-L-cysteine
sulfoxides. Headspace gas chromatography of thiosulfinates,
formed by reactions after the CS-lyase action, has been proposed
as an indicator of the extent of pungency (3). Thin layer
chromatography combined with thin layer electrophoresis has
been developed for the quantitation of S-alk(en)yl-L-cysteine
sulfoxides (4-6). Subsequently, HPLC has been applied to the
analysis of S-alk(en)yl-L-cysteine sulfoxides after derivatization
MATERIALS AND METHODS
Reagents. S-Methyl-L-cysteine was purchased from Sigma (St. Louis,
MO). N-Methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide was ob-
tained from Pierce (Rockford, IL). l-Cysteine was purchased from Wako
Pure Chemicals (Osaka, Japan). The Bond Elut SCX cartridge (regular
type, 500 mg/3 mL) was obtained from Varian (Harbor City, CA). The
* Corresponding author [telephone (81)-471-35-8001; fax (81)-471-33-
9159; e-mail seto@nrips.go.jp].
10.1021/jf020166e CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/28/2002

4446 J. Agric. Food Chem., Vol. 50, No. 16, 2002
Tsuge et al.
Sep-Pak C18 Plus cartridge was obtained from Waters (Milford, MA).
All other chemicals used were of analytical grade. All aqueous solutions
were prepared with distilled, deionized water.
with o-phthalaldehyde according to the method of Ziegler and Sticke
(19). A 20 µL aliquot of the onion extract was mixed with 180 µL of
a 2.8 mg/mL o-phthalaldehyde solution in 50 mM sodium borate (pH
9.5) including 10% methanol and 0.2% tert-butylthiol (20). A 50 µL
aliquot of the mixture was injected to the HPLC system, which consisted
of an HP 1050 series liquid chromatograph (Yokogawa Analytical
Systems). The HPLC column was a 25 cm × 4.6 mm i.d., 5 µm Inertsil
ODS-2 (G-L Science, Tokyo, Japan), and the isocratic elution was done
with a 32:68 mixture of 50 mM sodium phosphate (pH 7.15) and
methanol. The flow rate was 1.0 mL/min and the column was
maintained at room temperature. Absorbance at 339 nm was monitored.
(-)-S-Methyl-L-cysteine sulfoxide, (+)-S-methyl-L-cysteine sulfoxide,
(-)-S-propyl-L-cysteine sulfoxide, and (+)-S-propyl-L-cysteine sulfox-
ide were eluted at 8.8, 10.0, 15.1, and 17.7 min, respectively. The
synthetic compounds of 1 gave two peaks eluting at 8.8 and 10.0 min,
and those of 2 eluted at 15.1 and 17.7 min.
Solid Phase Extraction. A Sep-Pak C18 cartridge was activated by
washing with methanol followed by water and equilibrated with a 0.1
M HCl solution. A 1 mL aliquot of the sample was applied to the C18
cartridge, followed by elution with 0.1 M HCl. The combined fraction
of the cartridge pass-through and the 0.1 N HCl elution was applied to
the Bond Elut SCX cartridge, which was activated by washing with
ethanol and water and then equilibrated with 0.1 M HCl, and the SCX
cartridge was washed with 4 mL of 0.1 M HCl followed by 6 mL of
water. S-Alk(en)yl-L-cysteine sulfoxides were eluted with 4 mL of 500
mM ammonia water. The eluted fraction was dried by evaporation under
reduced pressure, dissolved with 1 mL of distilled water, and filtered
through a 0.45 mm cellulose acetate membrane.
tert-Butyldimethylsilylation (TBDMS) and Gas Chromatogra-
phy)Mass Spectrometry (GC-MS). The aqueous sample was dried
at 80 °C under a stream of nitrogen gas in a 1 mL volume glass vial
(Nichiden Rika Garasu, Tokyo, Japan). Thirty microliters of N-methyl-
N-(tert-butyldimethylsilyl)trifluoroacetamide and 20 µL of dried pyr-
idine were added, and the vial was then closed with a Teflon screw
cap, homogenized by sonication for 5 min, and incubated at 90 °C for
30 min. One milliliter of the mixture was applied to the GC-MS system,
which consisted of an HP 6890 gas chromatograph interfaced with an
HP 5973 quadrupole mass spectrometer (Yokogawa Analytical Sys-
tems). The stationary phase was a capillary 30 m × 0.25 mm i.d., 0.25
µm thickness, DB-5MS column (J&W Scientific, Folsom, CA). The
carrier gas (helium) flow rate and split ratio were adjusted to 1.0 mL/
min and 50:1, respectively. The injection port, transfer line, and ion
source were maintained at 250, 280, and 250 °C, respectively. The
oven temperature was controlled by a program [initial temperature, 100
°C (1 min hold), a ramp to 250 °C at 10 °C/min, then a ramp to 300
°C at 25 °C/min (3 min hold)]. Electron impact ionization (EI, ionization
energy ) 70 eV, ionization current ) 60 µA) was used as the ionization
mode for the general analysis. The acquisition mass range was m/z
50-550, and sampling was 0.8 scan/s. The acquisition was started 4
min after sample injection.
For GC-MS quantification, an internal standard (IS) calibration
method was performed. A 10 µL aliquot of the solid phase extraction
fraction was dried in a 1 mL volume glass vial. Thirty microliters of
N-methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide and 20 mL of
a 1 µg/mL anthracene (IS) solution in dried pyridine were added and
incubated at 90 °C for 30 min. One microliter of the mixture was applied
to the GC-MS system under the above-mentioned analytical conditions
except for the split ratio (20:1). Selected ion monitoring (SIM) was
performed at m/z 302 and 436 for 1, m/z 302 and 464 for 2, m/z 302
and 462 for 3, and m/z 178 for IS, respectively.
Onion Sample. An onion bulb cultivated in Ibaraki prefecture was
commercially obtained from the local supermarket in Kashiwa, Chiba.
Spectroscopy and Melting Point Measurements. Melting points
were measured using an MP-500V micro melting point apparatus
1
(Yanako, Kyoto, Japan) and are uncorrected. H NMR spectra were
recorded on a JNM-ECP 600 spectrometer (600 MHz, JEOL, Tokyo,
Japan). Electrospray ionization (ESI) time-of-flight (TOF) mass spectra
were recorded on a Q-TOF2 (Micromass, Manchester, U.K.) instrument
with direct sample infusion using a solvent of 0.1% formic acid in
water. IR spectra were recorded on a JIR-WINSPEC 50 FT-IR
spectrometer (JEOL, Tokyo, Japan) using a potassium bromide disk.
Synthesis of S-Alkyl-L-cysteine Sulfoxides. 1 was prepared by the
oxidation of S-methyl-L-cysteine with hydrogen peroxide, according
to the method of Synge and Wood (15) with slight modifications.
S-Methyl-L-cysteine (2.7 g) dissolved in 2.5 mL of water was mixed
with 2.5 mL of 30% hydrogen peroxide solution and stirred at 25 °C
for 1 h. 1 was crystallized by the addition of cold ethanol and
recrystallized from water/ethanol: yield, 87%; mp, 166 °C [lit. 167-
168 °C (18)]; IR, 1640 cm-1 (carboxylic acid), 2900 (amine), 1000
1
(sulfoxide); ESI-TOF-MS, m/z 152.0386 [(M + 1)+], C4H10NO3S; H
NMR (in D2O) δ 2.69 (d, 3H, CH3SO-), 3.10 (q) and 3.35 (q) (1H,
CCHSO-), 3.27 (m, 1H, CCHCS), 4.14 (q) and 4.07 (m) (1H,
NH2CHC).
2 was prepared according to the method of Lancaster and Kathleen
(6). S-Propyl-L-cysteine was prepared by using the method of Armstrong
and Lewis (16) with slight modifications, as follows. l-Cysteine (11.2
g) dissolved in a mixture of 15 mL of 20 M sodium hydroxide and
200 mL of ethanol was mixed with 11.6 mL of 1-bromopropane, and
the resulting solution was stirred at 25 °C for 3 min. The solution was
adjusted to pH 5.25 by the addition of acetic acid and stirred for 30
min in an ice bath. The precipitate (S-propyl-L-cysteine) was collected
by centrifugation (3.3 g), dissolved in 150 mL of water containing 0.5
mL of 5 M sodium hydroxide, mixed with 2.0 mL of 30% hydrogen
peroxide, and stirred for 1 h at 25 °C. 2 was crystallized by the addition
of cold ethanol and recrystallized from water/ethanol: yield, 44%; mp,
160-161 °C [lit. 163-164 °C (17)]; IR, 1600 cm-1 (carboxylic acid),
2900 (amine), 1030 (sulfoxide); ESI-TOF-MS, m/z 180.0701 [(M +
H)+], C4H10NO3S; 1H NMR (in D2O), δ 0.93 (t, 3H, CH3CH2CH2SO-
), 1.65 (m, 2H, CH3CH2CH2SO-), 2.84 (m, 2H, CH3CH2CH2SO-),
3.07 (q) and 3.32 (q) (1H, CCHSO-), 3.27 (m, 1H, CCHCS), 4.10 (d
of t, 1H, NH2CHC).
Extraction of S-Alk(en)yl-L-cysteine Sulfoxides from Onion. The
extraction of S-alk(en)yl-L-cysteine sulfoxides from onion was carried
out according to the method of Bieleski and Turner (4) with slight
modifications (6). Fresh onion bulbs (10 g) were peeled and cut into
small pieces with a knife and then immersed in 100 mL of a methanol/
chloroform/water (12:5:3, v/v) solution at -20 °C for 24 h. The liquid
phase was removed, an additional 100 mL of the solvent was added to
the residue material, and the extraction was repeated. The solvent phase
was separated, and 100 mL of a 80% ethanol solution was added to
the resulting solid materials. After 2 h, the supernatant was removed
and used as the ethanol extract. The combined solvent fraction was
supplemented with 90 mL of chloroform and 110 mL of water, and
the mixture was vortexed. The aqueous upper phase was removed,
combined with the ethanol extract, and concentrated to ∼5 mL under
reduced pressure at 60 °C. The concentrate was diluted with 0.1 M
hydrochloric acid (HCl) to be 25 mL.
Chemical ionization (CI) was adopted using methane as the reagent
gas, and the acquisition mode and mass range were positive and m/z
80-600, respectively.
High-resolution mass spectra were obtained using an HP 6890 gas
chromatograph combined with an MSroute JMS-600W sector mass
spectrometer (JEOL, Tokyo, Japan), under the same conditions as
above.
Capillary Electrophoresis (CE). 1 and 2 were quantified electro-
phoretically using an HP 3DCE CE system (Yokogawa Analytical
Systems, Tokyo, Japan), using an HP Forensic Anion Analysis Kit
(Yokogawa Analytical Systems; 18). A 140 cm × 50 mm i.d. capillary
fused silica column was used, and the electrophoresis buffer was an
HP basic anion buffer. The voltage was set at 30 kV with a negative
power supply. Detection was by indirect ultraviolet absorption at 350
nm (reference at 200 nm), and the column temperature was maintained
at 30 °C. Samples were applied hydrodynamically at 50 mbar for 4 s.
High-Performance Liquid Chromatography (HPLC). S-Alk(en)-
yl-L-cysteine sulfoxides were quantified by HPLC after derivatization
RESULTS AND DISCUSSION
TBDMS Derivatives of Methyl-L-cysteine Sulfoxide and
Propyl-L-cysteine Sulfoxide. As shown in Figure 1, under our

GC-MS of S-Alk(en)yl-L-Cysteine Sulfoxides
J. Agric. Food Chem., Vol. 50, No. 16, 2002 4447
Figure 1. Gas chromatogram of the TBDMS derivative of S-methyl-L-cysteine sulfoxide (1): (A) total ion chromatogram; (B) extracted ion chromatogram
at m/z 302; (C) extracted ion chromatogram at m/z 436. The TBDMS derivative of 1 (200 ng) was applied to GC-MS with electron impact ionization. The
asterisk (/) indicates a peak of tri(tert-butyldimethylsilyl) derivative of 1, and number signs (#) are the other peaks derived from 1. R indicates peaks
derived from the derivatization reaction mixture.
GC conditions, the TBDMS derivative of 1 gave a major peak
eluting at a retention time of 16.4 min. The EI mass spectrum
of the peak (Figure 2A) gave an ion peak of m/z 302 as the
base peak, and the characteristic ion peak of m/z 436, as well
as a minor peak at m/z 478. The m/z 302 ion was assumed to
be the ion corresponding to the molecule with the amino acid
side chain eliminated, which is typical of TBDMS derivatives
of many amino acids (10). The m/z 436 ion was assumed to be
the des-tert-butyl form (M - 57), for which the molecular ion
was the tri(tert-butyldimethylsilyl) derivative of 1. The high-
resolution mass spectrum was measured for the ion, giving the
exact mass of m/z 436.2197. This value is consistent with the
molecular composition C18H42NO3Si3S, the calculated mass of
which is 436.2193. The methane CI mass spectrum of the
TBDMS derivative of 1 (Figure 2B) gave ion peaks at m/z 522,
494, 478, and 436. They were assumed to be the C2H5 adduct
ion (M + 29), the pseudomolecular ion (M + 1), the desmethyl
molecular ion (M - 15), and the des-tert-butyl molecular ion
(M - 57), which are commonly observed for TBDMS deriva-
tives of amino acids under methane chemical ionization (11).
Besides the peak of the tri(tert-butyldimethylsilyl) derivative
of 1, there were observed several peaks that were derived from
1 and not from the derivatization reagent (Figure 1A). They
were not structurally elucidated. The heights of these peaks
increased with the concentration of 1 in the derivatization
reaction mixture. In the extracted ion chromatogram of m/z 302
under the EI conditions, no peak corresponding to a di-TBDMS
Figure 2. Mass spectra of the TBDMS derivative of S-methyl-L-cysteine
derivative of S-methyl-L-cysteine sulfoxide was observed, which
sulfoxide (peak at 16.4 min in Figure 1): (A) mass spectrum under electron
impact ionization; (B) mass spectrum under methane chemical ionization.
is commonly seen for the TBDMS derivatives of amino acids.
As shown in Figure 3, the TBDMS derivative of 2 gave a
peak eluting at a retention time of 16.8 min. The EI mass
spectrum of the peak (Figure 4A) gave an ion at m/z 302 as
the base peak and a characteristic ion of m/z 464, along with a
minor ion of m/z 506. The m/z 464 ion was assumed to be the
des-tert-butyl form (M - 57), the molecular ion of which was
tri-tert-butyldimethylsilylated 2. The high-resolution mass spec-
trum was measured for the ion, giving m/z 464.2509. This value
is consistent with the molecular composition C20H46NO3Si3S,

4448 J. Agric. Food Chem., Vol. 50, No. 16, 2002
Tsuge et al.
reagent (Figure 3A). They were not structurally elucidated. The
heights of these peaks increased with the concentration of 2 in
the derivatization reaction mixture. In the extracted ion chro-
matogram of m/z 302 under EI conditions, no peak correspond-
ing to di-TBDMS-S-propyl-L-cysteine sulfoxide was observed.
Besides silylation at the amino and carboxyl functions as in
the case of usual amino acids, one additional TBDMS group
was introduced into 1 and 2, probably at the sulfoxide, as shown
by EI-MS, methane CI-MS, and high-resolution MS. Thus, we
have experimentally confirmed that sulfoxide can be silylated.
Dimethyl sulfoxide was derivatized with N-methyl-N-(tert-
butyldimethylsilyl)trifluoroacetamide and identified by GC-MS
(data not shown). In EI-MS, it gave fragment ions of m/z 105
(M - tert-butylmethyl - methyl), m/z 135 (M - tert-butyl),
m/z 73 (trimethylsilyl), and m/z 177 (M - methyl). In the case
of methane CI-MS, it gave fragment ions of m/z 135 (M -
tert-butyl), m/z 177 (M - methyl), and m/z 191 (M - H). An
active methyl hydrogen may be transferred to the sulfoxyl group,
and the resultant thiol may be silylated with N-methyl-N-(tert-
butyldimethylsilyl)trifluoroacetamide to form the silylated sulf-
oxide species. Then, the resulting silylated sulfoxide may
rearrange probably via a -S+dCH-type intermediate to form
a silylated R-hydroxysulfide. This Pummerer-type rearrangement
was also reported in the case of the TBDMS and other
derivatives of thiodiglycol sulfoxide (19).
Figure 3. Gas chromatogram of the TBDMS derivative of S-propyl-L-
cysteine sulfoxide (2): (A) total ion chromatogram; (B) extracted ion
chromatogram at m/z 302; (C) extracted ion chromatogram at m/z 464.
The TBDMS derivative of 2 (2 µg) was applied to GC-MS with electron
impact ionization. The asterisk (/) indicates a peak of tri(tert-butyldi-
methylsilyl) derivative of 2, and number signs (#) are the other peaks
derived from 2. R indicates peaks derived from the derivatization reaction
mixture.
The peak areas on the SIM chromatograms at m/z 302
(TBDMS derivatives of both 1 and 2), m/z 436 (TBDMS
derivative of 1), m/z 464 (TBDMS derivative of 2), and m/z
178 (IS) were calculated. Anthracene (IS) was eluted at 11.4
min. The calibration curves of the ratios of the 1 peak areas
monitored at m/z 302 and 436 to those at m/z 178 (IS) were
linear for GC injections ranging from 0.6 to 20 ng for 1 with
correlation coefficients >0.999. The calibration curves of the
ratios of the 2 peak areas monitored at m/z 302 and 464 to those
at m/z 178 (IS) were linear for GC injections ranging from 2.5
to 20 ng for 2 with correlation coefficients >0.999. The within-
day repeatability was determined from five successive injections
of the standards (5 ng). The relative standard deviations were
6.0% (monitored at m/z 302) and 5.5% (m/z 436) for 1 and 4.9%
(m/z 302) and 7.5% (m/z 464) for 2, respectively. The detection
limits (per injection, signal/noise ) 3) were 0.063 ng (monitored
at m/z 302) and 0.28 ng (m/z 436) for 1 and 0.43 ng (m/z 302)
and 1.84 ng (m/z 464) for 2, respectively. It is possible to detect
1 at the sub-nanogram level. The detection limit for 1 was
several times lower than that for 2.
The TBDMS derivatization mixture of synthesized 1 and 2
gave not only the tri(tert-butyldimethylsilyl) derivatives but also
several compounds appearing as peaks on GC, which were
derived from the analyte sulfoxides (Figures 1A and 3A). The
structures of the peaks could not be elucidated. The derivati-
zation reaction of S-alkyl-L-cysteine sulfoxides with N-methyl-
N-(tert-butyldimethylsilyl)trifluoroacetamide may provide many
byproducts besides the expected tri(tert-butyldimethylsilyl)
derivatives. Especially, a large peak eluting at 16.0 min was
observed in the TBDMS reaction mixture of 2 (Figure 3A).
However, the product level of the tri(tert-butyldimethylsilyl)
derivatives was proportional to the analyte concentration. The
detection sensitivity for 1 was several times higher than that
for 2. As derived from the comparison of the peak area on the
total ion chromatograms (Figures 1 and 3), the total ion amount
of 1 derivative per injection also was >10-fold higher than that
of 2. The yield of TBDMS derivative appears to be higher for
1 than for 2, and this difference in derivative yields can be
attributed to the steric hindrance to the silylation reaction at
Figure 4. Mass spectra of the TBDMS derivative of S-propyl-L-cysteine
sulfoxide (peak at 16.8 min in Figure 3): (A) mass spectrum under electron
impact ionization; (B) mass spectrum under methane chemical ionization.
the calculated mass of which is 464.2506. The methane CI mass
spectrum of TBDMS derivative of 2 (Figure 4B) gave m/z 550,
522, 506, and m/z 464, which were assumed to be the C2H5
adduct ion (M + 29), the pseudomolecular ion (M + H), the
desmethyl molecular ion (M - 15), and the des-tert-butyl
molecular ion (M - 57). Besides the peak of the tri(tert-
butyldimethylsilyl) derivative of 2, there were observed several
peaks that were derived from 2 and not from the derivatization

GC-MS of S-Alk(en)yl-L-Cysteine Sulfoxides
J. Agric. Food Chem., Vol. 50, No. 16, 2002 4449
Table 1. Recoveries of S-Alkyl-L-cysteine Sulfoxides in Solid Phase
a
Extraction from Sucrose Matrix
S-methyl-L-cysteine
S-propyl-L-cysteine
sulfoxide
sulfoxide
fraction
starting
Sep-Pak C18
Bond Elut SCX
CE
GC-MS
CE
GC-MS
100
99.5
90.5 ± 3.3
0
0
100
100
99.0
0
0
74.5 ± 4.7
74.4 ± 5.3
a S-Methyl-L-cysteine sulfoxide (1) and S-propyl-L-cysteine sulfoxide (2) were
dissolved in 5% sucrose in 0.1 M HCl solution, and 1 mL of the mixture (starting
fraction) was applied to the Sep-Pak C18 Plus cartridge. The cartridge was eluted
with 0.1 M HCl. The combined fraction of the cartridge pass-through and the 0.1
N HCl elution (Sep-Pak C18 fraction) was applied to the Bond Elut SCX cartridge,
washed with 4 mL of 0.1 M HCl followed by 6 mL of water. 1 and 2 were eluted
with 4 mL of 500 mM ammonia water. The eluted fraction was dried by evaporation
under reduced pressure (Bond Elut SCX fraction). Each fraction was examined by
CE and GC-MS after TBDMS derivatization.
Figure 5. Electropherogram of S-methyl-L-cysteine sulfoxide (1) and
S-propyl-L-cysteine sulfoxide (2) (each 1 mg/mL). Electroosmotic flow (EOF)
appeared at 18 min.
not have been detected by TBDMS GC-MS. The cleanup of
S-alk(en)yl-L-cysteine sulfoxides, using the combination of
reversed phase and cation exchange cartridges, resulted in an
almost quantitative recovery from the simulated matrix (high
level of sucrose) solution. The reason for the GC-MS detection
yields of ∼75% (Table 1) may be that the remaining sucrose
in the final solid phase extract, even though low, might interfere
with the TBDMS derivatization of 1 and 2 to some extent.
Determination of Alk(en)yl-L-cysteine Sulfoxides in Onion.
The method developed was applied for the quantification of
alk(en)yl-L-cysteine sulfoxides in commercially available onion.
The extract (5%) of onion slices (10 g) was treated with the
combined solid phase extraction of Sep-Pak C18 and Bond Elut
SCX, and the solid phase extract was subjected to TBDMS
derivatization and GC-MS analysis. Without any pretreatment,
alk(en)yl-L-cysteine sulfoxides were not detected from the onion
extract by TBDMS GC-MS analysis. As shown in Figure 6C,
the TBDMS derivative of 1 was detected in the solid phase
extract of the onion extract, which was eluted at 16.4 min,
precisely the same as the authentic compound. The mass
spectrum was identical to that of the authentic compound (data
not shown). The TBDMS derivative of 2 was also detected
(Figure 6D), which showed the same retention time (16.8 min)
and mass spectrum as the authentic compound. 3 was assumed
to be present in the onion extract, the peak for which appeared
at the expected elution position (16.8 min) slightly earlier than
the derivative of 2 (Figure 7). The structure of 3 could be
inferred from its EI mass spectrum, which gave ions at m/z 504
(M - 15), 462 (M - 57), 360 (M - 159), and 302 (M -
propenyl) (Figure 8). Although the issue of whether the
S-alkenyl function of the detected 3 peak was 1-propenyl or
2-propenyl could not be ascertained mass spectrometrically, it
is said that onion contains only the 1-propenyl compound (2).
Synthetic S-alk(en)yl-L-cysteine sulfoxides give (()-S-dia-
stereomers. These isomers can be separated on a reversed phase
HPLC column after o-phthalaldehyde derivatization (7, 20, 21).
However, the capillary DB-5 column failed to separate the
TBDMS derivatives of the diastereomers clearly (Figures 1 and
3). The resolution of the TBDMS derivatives of 2 and 3 was
also poor, and the two peaks almost coeluted (Figure 6), in
contrast to the good resolution of o-phthalaldehyde derivatives
of 2 and 3 on a reversed phase HPLC column (20). However,
using mass spectrometric analysis, 2 and 3 could be separately
identified and quantified by GC-MS.
the sulfoxide or stability of the produced tri(tert-butyldimethyl-
silyl) derivatives.
Efficiency of Solid Phase Extraction. The yield of TBDMS
derivatization is influenced by sample matrix components. The
solvent extract of onion consists of the complex matrix, which
could interfere with the GC-MS analysis. The detection of the
TBDMS peaks could be affected, and the yields of TBDMS
derivatization would be suppressed. The majority of the
components in the onion extract are probably carbohydrates.
In this experiment, we adopted sucrose as a suspected interferent
and developed a cleanup method using solid phase extraction,
by which S-alk(en)yl-L-cysteine sulfoxides were isolated and
sucrose was removed. A Sep-Pak C18 cartridge was used for
the removal of hydrophobic interferents. Amino acids would
not be retained by the cartridge and would be recovered in the
eluted fraction. Bond Elut SCX (strong cation exchanger) was
also used for the purification of S-alk(en)yl-L-cysteine sulfox-
ides. Noncationic compounds can be eluted, and the retained
cationic compounds can then be eluted with ammonia. Ammonia
was successfully removed by evaporation from the elution
fraction. CE was used for the direct quantification of S-alkyl-
L-cysteine sulfoxides in the solid phase extracts. 1 (migration
time ) 7.3 min) and 2 (8.0 min) could be separated with high
resolution (Figure 5) and quantified in the presence of high
levels of sucrose (15 min).
The solutions containing 5% sucrose and designated 1 and 2
were treated by solid phase extraction, and the starting and
subsequent extracts were subject to CE and TBDMS GC-MS.
As shown in Table 1, 1 and 2 derivatives were not detected by
GC-MS in either the original sucrose solution or the Sep-Pak
C18 pass-through fraction. Nearly all of the sucrose was removed
by the combined solid phase extraction treatment. The recoveries
of 1 and 2 in the final solid phase extract were ∼90%. The
GC-MS detection yields were ∼75%.
Various matrix components would be expected to interfere
with the TBDMS derivatization reaction. In our previous work
on the TBDMS GC-MS analysis of methylphosphonic acids, a
severe suppression of TBDMS derivatization was observed in
the presence of divalent cations and sugars (13). Neutral sugars
in particular may inhibit the TBDMS derivatization by serving
as competitors in the silylation reaction. The carbohydrate
concentration in onion extract may be high, and it is reasonable
that S-alk(en)yl-L-cysteine sulfoxides in the onion extract might
The concentrations of S-alk(en)yl-L-cysteine sulfoxides in the
onion sample were determined by measuring the peak areas of

4450 J. Agric. Food Chem., Vol. 50, No. 16, 2002
Tsuge et al.
Figure 8. EI mass spectrum of the former part of the peak at 16.8 min
shown in Figure 6E.
onion. The concentration of S-methyl-L-cysteine sulfoxide
measured by GC-MS was almost equal to that by HPLC, and
also the value of S-propyl-L-cysteine sulfoxide measured by GC-
MS was similar to that by HPLC. However, the value of
S-propenyl-L-cysteine sulfoxide measured by GC-MS was
∼70% higher than that by HPLC. The discrepancy may be
ascribed to the situation that the detectability of 3 in TBDMS
GC-MS and HPLC analyses was not examined.
Onion contains various types of S-alk(en)yl-L-cysteine sulf-
oxides. Thomas et al. (8) reported that the contents of 1 and
S-1-propenyl-L-cysteine sulfoxide in white bulb onion were 23.6
and 131 mg/g, respectively, and that S-2-propenyl-L-cysteine
sulfoxide and 2 were not detected. Lancaster et al. (6) reported
that the contents of 1, 2, and S-1-propenyl-L-cysteine sulfoxide
were 0.9, 2.9, and 0.6 mg/g, respectively. The levels of
S-alk(en)yl-L-cysteine sulfoxides were variable during the
growth of onion (22). Therefore, our estimated levels of
S-alk(en)yl-L-cysteine sulfoxides in commercial onion appear
to be reasonable.
Figure 6. Gas chromatogram of the TBDMS reaction product of the solid
phase extract of onion extract: (A) total ion chromatogram; (B) extracted
ion chromatogram at m/z 302; (C) extracted ion chromatogram at m/z
436; (D) extracted ion chromatogram at m/z 464; (E) extracted ion
chromatogram at m/z 462.
The established GC-MS method is not complicated or time-
consuming. It provides for the structural confirmation and
detection of S-alk(en)yl-L-cysteine sulfoxides and its sensitive
quantification, the detection limit (nanogram order) of which
is compatible with that of the HPLC method (8). GC separation
efficiency is superior to HPLC, providing detection of S-
alk(en)yl-L-cysteine sulfoxides without interference by sample
matrix components. MS offers specific detection, compared to
ultraviolet-visible absorbance detection in HPLC and CE. Using
the combined solid phase extraction cleanup pretreatment, this
method has the potential for use in the areas of food and forensic
sciences.
Figure 7. Extracted ion chromatograms at m/z 464 (A) and m/z 462 (B)
of the TBMDS reaction product of the solid phase extract of an onion
extract.
their TBDMS derivatives on SIM chromatograms at m/z 436
(1) and m/z 464 (2) and assuming that the recoveries in the
solid phase extraction were 74.4 and 74.0%, respectively (Table
1). Although authentic compound 3 was not synthesized in this
study, the concentration of 3 was estimated, under the supposi-
tion that the detectability of the M - 57 ion (SIM at m/z 462)
and the solid phase extraction recovery are the same as those
of 2. The concentrations of 1, 2, and 3 in the onion sample
were 0.3, 3.1, and 3.0 mg/g of fresh weight of onion.
ABBREVIATIONS USED
CE, capillary electrophoresis; ESI, electrospray ionization;
IS, internal standard; SIM, selected ion monitoring; TBDMS,
tert-butyldimethylsilyl(ation); TOF, time-of-flight.
LITERATURE CITED
The concentrations of 1, 2, and 3 were also determined by
HPLC. From the onion extract, (-)-S-methyl-L-cysteine sulf-
oxide and (-)-S-propyl-L-cysteine sulfoxide were detected. In
addition, the peak eluting at 14.0 min was observed, which was
assumed to be (-)-S-propenyl-L-cysteine sulfoxide. We assumed
that the o-phthalaldehyde derivatization efficiency and molar
absorbency of (-)-S-propenyl-L-cysteine sulfoxide were the
same as those of 2. The concentrations of 1, 2, and 3 in the
onion sample were 0.3, 2.7, and 1.8 mg/g of fresh weight of
(1) Whitaker, J. R. Development of flavor, odor, and pungency in
onion and garlic. AdV. Food Res. 1976, 22, 73-133.
(2) Block, E. The organosulfur chemistry of the genus Alliums
Implications for organic sulfur chemistry. Angew. Chem., Int.
Ed. Engl. 1992, 31, 1135-1178.
(3) Freeman, G. G.; Whenham, R. J. A survey of volatile components
of some Allium species in terms of S-alk(en)yl-L-cysteine
sulphoxides present as flavour precursors. J. Sci. Food Agric.
1975, 26, 1869-1886.

GC-MS of S-Alk(en)yl-L-Cysteine Sulfoxides
J. Agric. Food Chem., Vol. 50, No. 16, 2002 4451
(4) Bieleski, R. L.; Turner, N. A. Separation and estimation of amino
acids in crude plant extracts by thin-layer electrophoresis and
chromatography. Anal. Biochem. 1966, 17, 278-293.
(5) Granroth, B. Separation of Allium sulfur amino acids and peptides
by thin-layer electrophoresis and thin-layer chromatography. Acta
Chem. Scand. 1968, 22, 3333-3335.
(6) Lancaster, J. E.; Kelly, K. E. Quantitative analysis of the
S-alk(en)yl-L-cysteine sulfoxide in onion (Allium cepa L.). J. Sci.
Food Agric. 1983, 34, 1229-1235.
(7) Gustine, D. L. Determination of S-methyl cysteine sulfoxide in
Brassica extracts by high-performance liquid chromatography.
J. Chromatogr. 1985, 319, 450-453.
(8) Thomas, D. J.; Parkin, K. L. Quantification of alk(en)yl-L-
cysteine sulfoxides and related amino acids in Alliums by high-
performance liquid chromatography. J. Agric. Food Chem. 1994,
42, 1632-1638.
(9) Chaves Das Neves, H. J. Capillary gas chromatography of amino
acids, including asparagine and glutamine: sensitive gas chro-
matographic-mass spectrometric and selected ion monitoring gas
chromatographic-mass spectrometric detection of the N,O(S)-
tert-butyldimethylsilyl derivatives. J. Chromatogr. 1987, 392,
249-258.
(10) Mawhinney, T. P.; Robinett, R. S. R.; Atalay, A.; Madson, M.
A. Analysis of amino acids ad their tert-butyldimethylsilyl
derivatives by gas-liquid chromatography and mass spectrom-
etry. J. Chromatogr. 1986, 358, 231-242.
(11) MacKenzie, S. L.; Tenaschuk, D.; Fortier, G. Analysis of amino
acids by gas-chromatography as tert-butyldimethylsilyl deriva-
tives. Preparation of derivatives in a single reaction. J. Chro-
matogr. 1987, 387, 241-253.
(12) Kataoka, M.; Tsunoda, N.; Ohta, H.; Tsuge, K.; Takesako, H.;
Seto, Y. Effect of cation-exchange pretreatment of aqueous soil
extracts on the gas chromatographic-mass spectrometric deter-
mination of nerve agent hydrolysis products after tert-butyldi-
methylsilylation. J. Chromatogr. A 1998, 824, 211-221.
(13) Kataoka, M.; Tsuge, K.; Seto, Y. Efficiency of pretreatment of
aqueous samples using a macroporous strong anion exchanger
resin on the determination of nerve gas hydrolysis products by
gas chromatographic-mass spectrometry after tert-butyldimeth-
ylsilylation. J. Chromatogr. A 2000, 891, 295-304.
(14) Kataoka, M.; Tsuge, K.; Seto, Y. Comparative examination of
various capillary electrophoretic analysis for nerve gas hydrolysis
products. Jpn. J. Assoc. Sci. Technol. Ident. 2000, 4, 67-75.
(15) Synge, R. L. M.; Wood, J. C. (+)-(S-Methyl-L-cysteine S-oxide)
in cabbage. Biochem. J. 1956, 64, 252-259.
(16) Armstrong, M. D.; Lewis, J. D. Thioether derivatives of cysteine
and homocysteine. J. Org. Chem. 1951, 16, 749-753.
(17) Freeman, G. G.; Whenham, R. J. Synthetic S-alk(en)yl-L-cysteine
sulphoxides-alliinase fission products: Simulation of flavour
components of Allium species. Phytochemistry 1975, 15, 521-
523.
(18) Soga, T.; Ross, G. A. Simultaneous determination of inorganic
anions, organic acids, amino acids and carbohydrates by capillary
electrophoresis. J. Chromatogr. A 1999, 837, 231-239.
(19) Black, R.; Holden, I.; Reid, M. Derivatisation reactions of
thiodiglycol and thiodiglycol sulphoxide. SeVenth International
Symposium on Protection against Chemical and Biological
Warfare Agents, Stockholm, Sweden, June 15-19, 2001;
Abstract, p 142.
(20) Ziegler, S. J.; Sticher, O. Optimization of the mobile phase for
HPLC separation of S-alk(en)yl-L-cysteine derivatives and their
corresponding sulfoxide isomers. J. Liq. Chromatogr. 1989, 12,
199-220.
(21) Ziegler, S. J.; Sticher, O. HPLC of S-alk(en)yl-L-cysteine
derivatives in garlic including quantitative determination of (+)-
S-allyl-L-cysteine sulfoxide (Alliin). Planta Med. 1989, 55, 372-
378.
(22) Lancaster, J. E.; McCallion, B. J.; Shaw, M. L. The levels of
S-alk(en)yl-L-cysteine sulfoxides during the growth of the onion
(Allium cepa L.). J. Sci. Food Agric. 1984, 35, 415-420.
Received for review February 7, 2002. Revised manuscript received
May 14, 2002. Accepted May 14, 2002.
JF020166E