In vitro biogeneration of pure thiosulfinates and propanethial-S-oxide

Article abstract of DOI:10.1021/jf000711g

A model reaction system was developed for generating, pure thiosulfinates and propanethial-S-oxide (PTSO) using an isolated alliinase (EC 4.4.1.4) and isolated or synthetic alk(en)yl-L-cysteine sulfoxides (ACSO). Reaction yields ranged from 30 to 60% after 3 h at 21-23°C, and organosulfur reaction products were extracted into CHCl₃ to yield product preparations of controlled composition. A pure thiosulfinate or PTSO was derived from a single ACSO, and a preparation containing a mixture of four thiosulfinate species was derived from reaction mixtures employing binary ACSO substrate systems. Identities of homologous thiosulfinates and PTSO were confirmed by ¹H NMR. This approach has the potential to be used as a preparative tool for yielding pure thiosulfinates and PTSO to facilitate the study of chemical and biological properties of this group of compounds or as a means to study the dynamics of organosulfur chemistry in preparations from Allium spp.

Full text of DOI:10.1021/jf000711g

6254
J. Agric. Food Chem. 2000, 48, 6254−6260
In Vitr o Biogen er a tion of P u r e Th iosu lfin a tes a n d
P r op a n eth ia l-S-oxid e
Cunxi Shen and Kirk L. Parkin*
Department of Food Science, University of Wisconsin, Babcock Hall, 1605 Linden Drive,
Madison, Wisconsin 53706
A model reaction system was developed for generating pure thiosulfinates and propanethial-S-
oxide (PTSO) using an isolated alliinase (EC 4.4.1.4) and isolated or synthetic alk(en)yl-L-cysteine
sulfoxides (ACSO). Reaction yields ranged from 30 to 60% after 3 h at 21-23 °C, and organosulfur
reaction products were extracted into CHCl3 to yield product preparations of controlled composition.
A pure thiosulfinate or PTSO was derived from a single ACSO, and a preparation containing a
mixture of four thiosulfinate species was derived from reaction mixtures employing binary ACSO
1
substrate systems. Identities of homologous thiosulfinates and PTSO were confirmed by H NMR.
This approach has the potential to be used as a preparative tool for yielding pure thiosulfinates
and PTSO to facilitate the study of chemical and biological properties of this group of compounds
or as a means to study the dynamics of organosulfur chemistry in preparations from Allium spp.
Keyw or d s: Allium; thiosulfinates; propanethial-S-oxide; biogeneration; alliinase; alk(en)yl-L-cysteine
sulfoxides
INTRODUCTION
Sch em e 1. Tr a n sfor m a tion of ACSO to
Th iosu lfin a tes a n d P TSO
Organosulfur components from tissues of Allium
species have been the focus of decades of study because
of their sensory properties (Stoll and Seebeck, 1951) and
prospects of possessing therapeutic activities (Kendler,
1987; Lau et al., 1990). In many instances, biological
effects have been demonstrated with pastes, distilled
oils, or extracts of Allium tissues, and these effects have
often been correlated with, or attributed to, the orga-
nosulfur components intrinsic to the preparation (e.g.,
Sendl et al., 1992). In some cases, subsequent studies
with pure compounds have allowed direct assignments
of specific biological activities to specific organosulfur
components. However, in other reports, the assignment
of biological effects to specific organosulfur compounds
remains equivocal. A case in point is the assignment of
antioxidant activity in a garlic extract to allicin (Prasad
et al., 1995), the principal thiosulfinate in macerated
garlic tissues (Lawson et al., 1991; Lawson, 1996; Block
et al., 1992), despite the prospect that several other
compounds in the extract may be more capable than
allicin at inhibiting oxidative processes (Yin and Cheng,
1998).
Great strides have been made in understanding the
chemistry of the organosulfur components following the
initial enzymic transformation of endogenous S-alk(en)-
yl-L-cysteine sulfoxides (ACSO) to the corresponding and
transient alk(en)yl sulfenic acids (Block, 1992). The
intial steps in ACSO transformation yields the corre-
sponding sulfenic acid (RSOH) which can rapidly rear-
range (for the 1-propenyl species to form and propaneth-
ial-S-oxide, PTSO) and/or condense to form homologous
(R1 ) R2) or heterologous (R1 * R2) thiosulfinates (R1S-
(O)S R2) as depicted in Scheme 1.
Perhaps the most important of the organosulfur
components in freshly macerated Allium preparations
are the thiosulfinates and PTSO, since virtually all
other organosulfur compounds arising in processed
Allium tissues are derived from these compounds.
However, understanding the chemistry and fate of these
organosulfur compounds in tissue preparations is com-
plicated by the complexity of the tissue milieu and how
this impacts further transformation of these compounds.
It became evident to us that having a source of pure
thiosulfinates and related organosulfur compounds
would allow for an examination of their intrinsic chemi-
cal properties and biological activities. Thiosulfinates
and related organosulfur compounds have been chemi-
cally synthesized or isolated from extracts of Allium
tissues (Lawson et al., 1991; Block, 1992). However, with
few exceptions (Block et al., 1986; Lawson et al., 1991;
Morimitsu et al., 1992; Block et al., 1997), there has not
been a systematic evaluation of structure-function
relationships, in terms of chemical and biological prop-
erties, of thiosulfinates and related compounds. Fur-
thermore, the dynamics of the fate of thiosulfinates and
PTSO in situ remains enigmatic. With these issues in
mind, we sought to develop a model reaction system to
facilitate the preparation and study of the full range of
thiosulfinate species and PTSO that can be found in
various, freshly prepared Allium extracts. Our approach
was based on using isolated or synthetic ACSO and an
* Author to whom correspondence should be addressed
(telephone (608) 263-2011; fax (608) 262-6872; e-mail klparkin@
facstaff.wisc.edu).
10.1021/jf000711g CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/21/2000

Biogeneration of Allium Thiosulfinates
J. Agric. Food Chem., Vol. 48, No. 12, 2000 6255
kg) were cut into quarters and peeled. Batches of 500-600 g
tissue segments were placed in about 1 L boiling distilled water
for 5 min and then removed to cool. The heated tissue was
then homogenized in the heating medium plus another 0.2
volumes of distilled water. The slurry was filtered through
cheesecloth, and the filtrate was adjusted to pH 4 with acetic
acid before centrifugation (40 min × 13 000g) at 4 °C to remove
particulate matter.
isolated alliinase (EC 4.4.1.4) preparation in a model
reaction systems that simulates the generation of these
compounds in situ.
MATERIALS AND METHODS
Ma ter ia ls. All chemicals were obtained from Sigma (St.
Louis, MO) or Aldrich (Milwaukee, WI) Chemical Companies
unless otherwise noted. All solvents used were chromatogra-
phy grade. White onion bulbs were purchased from a local
retail market.
The resulting supernatant was loaded onto a column (5 ×
30 cm) of Dowex 50-X4 (H+), and the column was eluted with
0.1 M sodium acetate (pH 6.5). Eluting fractions were tested
for ninhydrin-positive material by spraying with 0.8% ninhy-
drin in ethanol after chromatography on Whatman #1 filter
paper (butanol-acetic acid-water: 63/10/27, v/v/v as develop-
ing solvent). Presumptive 1-PeCSO-containing fractions were
identified on the basis of comigration with a 1-PeCSO stan-
dard. The 1-PeCSO-containing fractions were pooled, applied
to a second column (2.5 × 50 cm) of Dowex 50-X4 (H+), and
eluted with 0.05 N NaOH at 30 mL h-1. Presumptive 1-PeCSO-
positive fractions (still in 0.05 N NaOH) were again pooled
and passed through a third column (2 × 15 cm) of Dowex 2-X8
(OH-) to obtain the acidic amino acids, including 1-PeCSO,
which are not retained by the column. At this point, ninhydrin-
positive fractions showed virtually one band on paper chro-
matography. Those fractions were finally loaded onto a column
(2.5 × 50 cm) of Dowex 50-X4 (H+), which was eluted with
0.05 N ammonium hydroxide. The 1-PeCSO-containing frac-
tions were pooled, adjusted to pH 6.5 with acetic acid, and
lyophilized. Recrystallization of an aqueous solution (1-2 mL)
of lyophilized material by dropwise addition of acetone yielded
pure 1-PeCSO, with a yield of about 1 g.
P r ep a r a tion of Im m obilized C-S Lya se (Alliin a se). A
crude alliinase preparation was prepared using the prelimi-
nary isolation steps described earlier (Thomas and Parkin,
1991) with all steps carried out at 0-4 °C. This procedure
involved initial homogenization of onion bulbs in 0.1 M
potassium phosphate buffer (pH 7.5) containing 10% glycerol,
0.5% poly(vinylpyrrolidone), 0.5 mM phenylmethylsulfonyl
fluoride, 5 mM ethylenediaminetetraacetic acid (EDTA), and
0.05% cysteine, followed by precipitation by 65% saturated
with ammonium sulfate at 0-4 °C. After collecting the
precipitate, exhaustive dialysis against 0.01 M potassium
phosphate buffer (pH 7.5) containing 10% glycerol, 5 mM
EDTA, and 0.05% cysteine yielded an alliinase preparation of
1.28 mg protein mL-1 (based on Coomassie blue assay;
Bradford, 1976) and a specific activity of 2.38 Unit µg-1 protein
(one Unit is defined as one µmol pryuvate produced min-1).
Immoblization of crude alliinase was carried out according
to Thomas and Parkin (1991) by placing equal volumes of
enzyme preparation and 4% sodium alginate solution (Protanal
5/60; a gift from Protan Inc., North Hampton, NH) in cellulose
dialysis tubing (12 000-14 000 MW cutoff, Spectrum Medical
Industries, Los Angeles, CA), followed by immersion in 0.1 M
CaCl2 solution to induce gelation. The resulting enzyme-loaded
gel was then cut into 5-10 mm thick slices and stored at 4 °C
until use. For the purpose of these experiments, the im-
mobilized alliinase was capable of in vitro biogeneration of
thiosulfinates after as long as 1 yr in storage at 4 °C.
P r ep a r a tion of Alk (en )yl-L-cystein e Su lfoxid e (ACSO)
Su b st r a t es. Synthesis of (()-S-Methyl-L-cysteine Sulfoxide
(MCSO). Diasteromeric MCSO was prepared by a modified
method of Synge and Wood (1956), by combining 23 mL of
S-methyl-L-cysteine (0.6 M) with 1.75 mL of 30% hydrogen
peroxide. After continuous stirring for 24 h at 21-23 °C,
particulate material was removed by filtration, and a white
precipitate of MCSO was recovered by adding 250 mL of cold
ethanol to the filtrate and holding at 4 °C overnight. Yields
were typically 60-80%.
Synthesis of (()-S-Ethyl-L-cysteine Sulfoxide (ECSO). Syn-
thesis of diasteromeric ECSO was essentially the same as the
synthesis of MCSO except that S-ethyl-L-cysteine was used
as the starting material. Yields were typically 60-75%.
Synthesis of (()-S-Propyl-L-cysteine Sulfoxide (PCSO). Syn-
thesis of diasteromeric PCSO was similar to the method
described by Lancaster and Kelly (1983). L-Cysteine-HCl (4
g) was dissolved in 75 mL of ethanol followed by dropwise
addition of 5.6 mL of 20 N NaOH. After 5 min, 4 mL of 1-propyl
bromide was added. The mixture was stirred overnight at 21-
23 °C and then adjusted to pH 5.5 using acetic acid. The
resultant suspension was cooled to 4 °C for 2 h. The S-propyl-
L-cysteine precipitate was collected by filtration, and the
corresponding sulfoxide was formed by oxidizing S-propyl-L-
cysteine with 30% hydrogen peroxide as described for MCSO
synthesis above, with a typical yield of 45-55%.
The structure of 1-PeCSO was confirmed by 1H NMR (model
AM-300 NMR spectrometer, Bruker Instruments, Inc., Bil-
lerica, MA) (300 MHz, D2O): δ 6.56 (qd, J ) 15, 6.5 Hz, 1H,
CCHdCS-), 6.37 (dd, J ) 15, 1.5 Hz, 1H, CCdCHS-), 3.96
(dd, J ) 8.5 Hz, 1H, RCH), 3.29 (ABdd, J ) 14.5 Hz, 1H,
-SCHaRC-), 3.10 (ABdd, J ) 14.8 Hz, 1H, -SCHbRC-), 1.79
1
(dd, J ) 6.5, 1.5 Hz, 3H, CH3CdC-). This H NMR spectrum
is consistent with the double bond of the 1-PeCSO being in
the trans (native) form (Nishimura et al., 1975; Zheng et al.,
1988; Mu¨tsch-Eckner et al., 1992).
Mod el Rea ction System . Typically, ACSO (0.05 mmol)
was dissolved in 4.0 mL of 100 mM Tris (pH 7.5), and 0.5 g of
immobilized alliinase (ground with a mortar and pestle to yield
gel fragments of about e 1 mm dimension) was added to
initiate the reaction at 21-23 °C. After various periods of
incubation, thiosulfinate (R1S(O)SR2; where R1/R2 groups are
methyl (Me), ethyl (Et), propyl (Pr), allyl (All), or 1-propenyl
(Pren) residues) products were obtained by extracting an
aqueous subsample of the reaction mixture into an equal
volume of CHCl3 (containing benzyl alcohol as internal stan-
dard). The authenticity of the S-containing compounds ob-
tained was confirmed by 1H NMR (300 MHz, CDCl3) as follows
(some spectra could be compared to previous reports: Lawson
et al., 1991; Naganathan, 1992; Block et al., 1996): MeS(O)-
SMe: δ 3.00 (3H, s, CH3S(O)-), 2.69 (3H, s, CH3SS(O)-); EtS-
(O)SEt: δ 3.05-3.24 (4H, m, -CH2S(O)SCH2-), 1.48 (3H, t, J
) 7.5 Hz, CH3CS(O)-), 1.42 (3H, t, J ) 7.5 Hz, CH3CSS(O)-
); PrS(O)SPr: δ 3.03-3.22 (4H, m, -CH2S(O)SCH2-), 1.77-
1.96 (4H, m, -CH2CS(O)SCCH2-), 1.10 (3H, t, J ) 7.5 Hz,
CH3CCS(O)-), 1.05 (3H, t, J ) 7.5 Hz, CH3CCSS(O)-); AllS-
(O)SAll: δ 5.87-6.04 (2H, m, CdCHC-), 5.20-5.49 (4H, m,
CH2dCC), 3.70-3.93 (4H, m, CdCCH2); PTSO (C2H5CHdS+-
O-): δ 8.18 (1H, t, J ) 7.8 Hz, -SdCHCC), 2.80 (2H, p, J )
7.8 Hz, -SdCCH2C), 1.16 (3H, t, J ) 7.8 Hz, -SdCCCH3).
The lack of a detectable downfield signal at δ 8.86 for PTSO
confirmed that the cis (natural) isomer was present (Block et
al., 1996).
Synthesis of (()-S-2-Propenyl-L-cysteine Sulfoxide (2-PeC-
SO). Synthesis of diasteromeric 2-PeCSO was similar to the
synthesis of PCSO described above except that 2-propenyl
bromide was substituted for propyl bromide, with yields of
about 15-20%.
Isolation of (+)-S-1-Propenyl-L-cysteine Sulfoxide (1-PeCSO)
from White Onion Bulbs. (+)-S-(1-propenyl)-L-cysteine sulfox-
ide, the natural form of 1-PeCSO in onion, was obtained by a
modified method of Carson et al. (1966) with all procedures
done at 21-23 °C except where noted. White onion bulbs (4-5
Quantification of organosulfur analytes was based on the
peak area in HPLC chromatography (model 2300 pumps, V4
detector set at 254 nm, and with peak integration by Chem-
Research software; Isco, Lincoln, NE). Normal phase chroma-
tography on a 250 mm × 4.6 mm, Microsorb 5 µm Silica
column (Rainin Instrument Co Inc., Woburn, MA) involved
gradient elution with 2-propanol:hexane from 2:98 (v/v) held

6256 J. Agric. Food Chem., Vol. 48, No. 12, 2000
Shen and Parkin
Ta ble 1. Efficien cy of Ch lor ofor m Extr a ction of
Th iosu lfin a tes a n d P TSO fr om Mod el Rea ction Mixtu r es
thiosulfinate
species
proportion
extracted
thiosulfinate
species
proportion
extracted
EtS(O)SEt
EtS(O)SMe
MeS(O)SEt
MeS(O)SMe
PrS(O)SPr
PrS(O)SMe
MeS(O)SPr
PTSO
99.7 (4)
95.3 (2)
100 (2)
87.7 (5)
100 (4)
99.9 (2)
102 (2)
100 (2)
AllS(O)SAll
AllS(O)SMe
MeS(O)SAll
AllS(O)SEt
EtS(O)SAll
PrS(O)SEt
EtS(O)SPr
100 (2)
95.4 (2)
98.0 (2)
103 (2)
100 (2)
97.0 (2)
97.5 (2)
F igu r e 1. Pyruvate generation in model reaction mixtures.
Reaction mixtures contained 0.05 mmol of either methyl, ethyl,
propyl, 2-propenyl or 0.025 mmol 1-propenyl cysteine sulfoxide
(MCSO, ECSO, PCSO, 2-PeCSO, 1-PeCSO, respectively), 0.5
g of immobilized alliinase in 4.0 mL of Tris (7.5) buffer. Right
ordinate denotes % reaction yield (in parentheses for 1-PeCSO)
based on the amount of pyruvate formed relative to the initial
level of ACSO. CV for duplicate experiments was 6.8%.
for 6 min to 10:90 (v/v) for the next 10 min followed by a 7
min hold, with an elution rate of 1.4 mL min-1 (adapted from
Block et al., 1992).
Concentration of each of these organosulfur components in
1
a stock solution was determined on the basis of their H NMR
signal and comparison to the signal of the internal standard
(tert-butyl alcohol). This allowed the calibration of the HPLC
detector response to external standard curves for each analyte.
The relationship between concentration and area units (au)
using 254 nm detection were as follows: MeS(O)SMe, 68.1 nM
au-1; EtS(O)SEt, 87.4 nM au-1; PrS(O)SPr, 85.6 nM au-1; AllS-
(O)SAll, 80.9 nM au-1; PSTO, 20.9 nM au-1. These values are
consistent with the extinction coefficients reported for thio-
sulfinates in a 2-propanol:hexane mixture (Lawson et al.,
1991). Heterologous thiosulfinates were presumptively identi-
fied on the basis of the elution patterns observed in Block et
al. (1992) and using response factors calculated as the average
values of the corresponding homologous thiosulfinates.
The efficiency of CHCl3 extraction of thiosulfinates from
reaction mixtures was determined on the basis of back-
extraction of standardized thiosulfinate solutions in CHCl3 into
an equal volume of water by vortexing for 1 min at 21-23 °C.
Results, with a coefficient of variation of about ( 3%, appear
in Table 1. For the purposes of this study, corrections for
fact that the (+) isomer of 1-PeCSO and the (()
diasteromeric mixtures of the other ACSOs were used
is of some, but limited, impact. Although the (+) isomer
is favored by alliinase over the (-) isomer, when
equimolar mixtures of the (() diastereomers are present,
the kinetics of the system reflect those of the faster
reacting isomer (Schwimmer et al., 1964).
The data in Figure 1 were also transformed to linear
(log-log) plots (not shown) representing the time-
dependent fractional conversion of each substrate rela-
tive to each other (theory in Deleuze et al., 1987). The
slopes of these plots represent the relative values of Vm/
Km (the selectivity constant; Fersht, 1985) for each
substrate. Fits for these linear plots (r2 g 0.94) yielded
respective and relative values for Vm/Km for the MCSO,
ECSO, PCSO, 2-PeCSO, and 1-PeCSO as 1.00:1.59:2.01:
4.38:10.7. These value are consistent with respective Vm/
Km values of 1:3:18 for MCSO:PCSO:1-PeCSO (calcu-
lated from the kinetic data provided by Freeman and
Whenham, 1975).
The reaction progress curves were typical of enzyme
reactions except for a rather abrupt decline of activity
with 1-PeCSO after 30 min, a point where only about
60% 1-PeCSO was consumed based on the evolution of
pyruvate. This phenomenon has been witnessed before
for onion alliinase in vitro using 1-PeCSO (Schwimmer,
1969) and in situ toward MCSO and PCSO (Lancaster
et al., 1998).
incomplete extraction were applied for MeS(O)SMe (× 0.88-1
)
and EtS(O)SMe and AllS(O)SMe of (× 0.95-1), whereas all
other thiosulfinates and PTSO were considered to be quanti-
tatively extracted within the limits of the ( 3% variance.
The progress of model reactions was also followed by
pyruvate formation, based on the coupling assay involving
oxidation of NADH to NAD by lactate dehydrogenase (Schwim-
mer and Weston, 1961). In this case, a 0.3 mL subsample of
the reaction mixture was quenched by the addition of 1.8 mL
of 2.2% trichloroacetic acid and then centrifuged for 3 min ×
13 600g to clarify the extract. A 0.5 mL portion of the cleared
extract (yielding up to 2 mM pyruvate) was assayed for
pyruvate.
For all these studies, at least two duplicate experiments
were conducted, and the mean values are reported along with
CV for the experimental data set.
The corresponding progress curves for the formation
of thiosulfinates in alliinase/ACSO reaction mixtures
(Figure 2; corrected for extraction efficiency) were
generally corroborative with the evolution of pyruvate
(Figure 1), in that the same order of preferences of
ACSO (except for 1-PeCSO, discussed later) reactivity
as indicated by pyruvate formation also held for thio-
sulfinate formation. However, based on the reaction
stoichiometry of 2:1 molar ratio of pyruvate:thiosulfinate
formed (Scheme 1), less than stoichiometric levels of
pyruvate were found for all but one case (Figure 3).
Pyruvate levels ranged from 36 to 104% of those
expected based on thiosulfinate yields, and the greatest
substoichiometric levels of pyruvate were observed early
in the reactions for all ACSO substrate systems. Al-
though pyruvate is widely used as an indicator of
pungency of Allium tissue preparations (Randle and
Bussard, 1993), less than stoichiometric yields of pyru-
vate in alliinase-mediated reactions in situ have been
reported previously (Lancaster et al., 1998).
RESULTS AND DISCUSSION
P r ogr ess of Mod el Rea ction Mixtu r es. Using
reaction conditions that were developed as being “stan-
dard”, the progress of model reactions for immobilized
alliinase and ACSO is shown in Figure 1. Tris buffer at
pH 7.5 was used because of the reported optimum of
the onion alliinase as being pH 7-8 (Tobkin and
Mazelis, 1979; Nock and Mazelis, 1987; Thomas and
Parkin, 1991). Under these conditions, pyruvate forma-
tion readily took place for all ACSO substrates used,
and relative initial rates of reactions were observed in
descending order: 1-PeCSO > 2-PeCSO > PCSO >
ECSO > MCSO. This order of preference is consistent
with the relative reactivity of ACSOs and alliinase
predicted on the basis of their respective (and increas-
ing) Km values for these substrates (Schwimmer et al.,
1964; Schwimmer, 1969; Tobkin and Mazelis, 1979). The

Biogeneration of Allium Thiosulfinates
J. Agric. Food Chem., Vol. 48, No. 12, 2000 6257
experiments revealed that the fate of PTSO in these
reactions systems was dependent on the pH and choice
of buffering agent (Figure 4). Reactions taking place at
pH 7.5 were faster than those at pH 5.0, using both
pyruvate and PTSO formation as indicators, especially
considering reaction progress within the first 60 s. This
was expected based on the reported pH optimum of
onion alliinase to be in the range of 7-8 (Tobkin and
Mazelis, 1979; Nock and Mazelis, 1987; Thomas and
Parkin, 1991). However, alliinase was sufficiently reac-
tive at pH 5.0 (which is similar to the pH of macerated
onion bulb tissue) such that similar extents of ACSO
transformation were accomplished at longer incubation
times at pH 5.0 relative to shorter incubation times at
“optimum pH”. Although counterintuitive, it was clear
that model reactions at suboptimal pH were more
conducive than those at optimal pH for alliinase-
mediated generation and recovery of PTSO from 1-PeC-
SO.
F igu r e 2. Organosulfur product generation in model reaction
mixtures. Legend is the same as for Figure 1, except that
products formed were thiosulfinates, RS(O)SR, where R )
methyl (Me), ethyl (Et), propyl (Pr), 2-propenyl (All), or
propanethial-S-oxide (PTSO), from the corresponding ACSO.
CV for duplicate experiments was 2.9%.
The data in Figure 4 also indicate that the stability/
reactivity of PTSO was dependent on pH in the range
of pH 5-7, rendering PTSO more sensitive to increasing
pH than thiosulfinates (Block, 1992). Also, greater
recoveries of PTSO were observed in phosphate-buffered
compared to Tris-buffered systems. We attribute this
phenomenon to the potential for amine functional
groups (Tris) to either react directly with PTSO or serve
as a nucleophilic catalyst to facilitate the transformation
or decay of PTSO into various derivatives. The relation-
ship that exists between buffering agents and the fate
of PTSO in the alliinase/ACSO reaction system is the
subject on ongoing study in this laboratory.
Rea ction s w ith Bin a r y ACSO System s. Having
established the capability of preparing homologous
thiosulfinates from a single ACSO substrate, binary
ACSO reaction systems were evaluated for their ability
to yield mixtures of homologous and heterologous thio-
sulfinates (Table 2). The profiles of the four species of
thiosulfinates yielded by each binary ACSO system was
reflective of the relative rates of activity of the individual
ACSO. Both % yield of thiosulfinate and mole fraction
of alk(en)yl groups from the faster reacting ACSO
comprising the total thiosulfinate pool increased as the
binary ACSO system was modified to contain a faster
reacting ACSO species or a greater proportion of the
faster reacting ACSO species of the pair.
The profile of heterologous thiosulfinates formed in
the binary ACSO mixtures generally followed the pat-
tern of R1S(O)SR2 > R1SS(O)R2, where R1 and R2 were
derived from the slower and faster reacting ACSO,
respectively. A sulfenic acid (RSOH)-thiosulfinate alk-
(en)yl-exchange reaction can account for this behavior
where the RSOH reactant is derived principally from
the slower reacting ACSO and the homologous thiosul-
finate from the faster reacting ACSO (Scheme 2, adapted
from Block, 1992; Lawson, 1996). The basis of this
mechanism is kinetic, and a greater disparity in reaction
rates of different ACSO should lead to a greater
magnitude of difference between the heterologous thio-
sulfinate species. However, we found little correlation
(r2 ) 0.179, n ) 10) between the molar ratios of the
heterologous and homologous thiosulfinates in the
binary ACSO reaction mixtures (data not shown). While
the ratios of about 2:1 for the heterologous thiosulfinates
in the model reaction mixtures are consistent with what
has been found in tissue preparations (Lawson et al.,
1991; Block et al., 1992), a theoretical ratio of 3:1 of
F igu r e 3. Degree of stoichiometric yield of pyruvate in model
reaction mixtures. Yield (%) of pyruvate observed was based
on an expected product stoichiometry of 2:1 pyruvate:thiosul-
finate for the model reactions. Data and legend are from
Figures 1 and 2.
The 70-92% deficit in stoichiometric levels of pyru-
vate relative to the degree of ACSO transformation
previously reported is equivalent to ca. 16-28 mM
pyruvate in onion bulb homogenates (calculations are
based on data appearing in Table 2 in Lancaster et al.,
1998). This unmeasurable pyruvate, or “crypto-pyru-
vate”, was suggested to be underestimated or undetect-
able because of pyruvate binding to pyridoxal phosphate
at the active site (Lancaster et al., 1998). Onion bulb
tissue contains about 4 µM pyridoxal phosphate and an
upper limit of about 15 g protein kg-1 (Fenwick and
Hanley, 1985). At a subunit molecular mass of 50 kDa
(Nock and Mazelis, 1987) and assuming that alliinase
comprises up to about 10% of the protein, a calculated
in situ level of onion alliinase of 30 µM indicates that
such a mechanism would contribute very little (<0.2%)
to the “crypto-pyruvate” phenomenon. Instead, we be-
lieve that the majority of substoichiometric levels of
pyruvate must be accounted for by other means, possibly
in the slow conversion to pyruvate (Schmidt et al., 1996),
or instability and/or reactivity of the aminomethacrylic
acid intermediate (Scheme 1), or reactivity of pyruvate
with other compounds (Chu and Clydesdale 1976;
Fulcrand et al., 1998).
The second inconsistency between the trends in
Figures 1 and 2 is for reactions with 1-PeCSO as
substrate, where much less PTSO was formed than
would be predicted based on the levels of pyruvate
evolved in this reaction system. Lack of recovery of
PTSO may be attributed to volatilization and/or chemi-
cal transformation to yield products not detected by our
HPLC method over the 2 h reaction period. Subsequent

6258 J. Agric. Food Chem., Vol. 48, No. 12, 2000
Shen and Parkin
Ta ble 2. Th iosu lfin a te Yield s a n d P r ofiles in Bin a r y ACSO Rea ction Mixtu r esc
molar ratio of ACSOA/ACSOB
substrates
1:1
2:1
ACSOA/B
substrates
total
TS (mM)
TS %
yielda
TS
ratio
total
TS (mM)
TS %
yielda
TS
ratio
TS products
MCSO/ETCSO
3.4
1.9
1.8
1.0
16.4
3.1
4.8
1.0
23.4
2.7
5.8
1.0
5.3
1.7
2.9
1.0
10.4
2.1
1.0
1.0
0.9
1.0
4.7
1.7
2.5
1.0
3.6
1.3
2.4
1.0
1.6
0.9
1.5
1.0
1.4
0.9
1.9
1.0
ND
EtS(O)SEt
EtS(O)SMe
MeS(O)SEt
MeS(O)SMe
PrS(O)SPr
PrS(O)SMe
MeS(O)SPr
MeS(O)SMe
AllS(O)SAll
AllS(O)SMe
MeS(O)SAll
MeS(O)SMe
PrS(O)SPr
PrS(O)SEt
EtS(O)SPr
EtS(O)SEt
AllS(O)SAll
AllS(O)SEt
EtS(O)SAll
EtS(O)SEt
2.44
2.37
2.91
2.69
19.5
19.0
23.3
21.5
2.65
2.74
2.89
3.54
14.1
14.6
15.4
18.9
18.9
MCSO/PCSO
MCSO/2-PeCSO
ECSO/PCSO
ECSO/2-PeCSO
3.48
4.04
27.9
32.3
3.55
ND
5.0
1.0
PCSO/2-PeCSO
NDb
ND
(individual species were not resolved)
a
b
TS yield was calculated based on the amount of ACSO substrate used. Not determined. c Reaction conditions were 12.5 mM each
ACSO (1:1 molar ratio) or 25.0/12.5 mM ACSO (2:1 molar ratio), 0.8 g of immobilized enzyme in 4 mL of Tris buffer at pH 7.5 and 25 °C
for 90 min.
Sch em e 2. Exch a n ge Rea ction s betw een Su lfen ic
Acid s a n d Th iosu lfin a tes
a
R1 and R2 represent alk(en)yl groups originating from
ACSO or thiosulfinate, respectively.
F igu r e 4. Reaction progress in mixtures containing 1-PeCSO
as substrate. (a) Pyruvate formation and (b) PTSO formation
as a function of pH, using either Tris or sodium phosphate
buffer where reaction conditions are the same as in Figures 1
and 2. CV for duplicate experiments was 15.5% and 9.2% for
the determination of pyruvate and PTSO, respectively.
for such a mechanism in the literature, we dismissed
this alternative mechanism as a possibility.
Further exploitation of the model reaction system was
used to study how the evolution of heterologous thio-
sulfinates could be manipulated. Reaction systems
containing an ACSO and preformed thiosulfinate were
initiated by the addition of alliinase under standard
conditions, and products were extracted into CHCl3 and
analyzed by HPLC. When 13 mM MCSO was reacted
in the presence of 6 mM AllS(O)SAll, the ratio of MeS-
(O)SAll:MeSS(O)All was 26:1 after 2 h reaction. When
the conditions were changed to 19 mM MCSO and 7 mM
AllS(O)SAll, the ratio of MeS(O)SAll:MeSS(O)All was
14:1 after 2 h reaction. These ratios can be explained
using Scheme 2, where a “slow” reacting MCSO (step
MeS(O)SAll:MeSS(O)All has been predicted in garlic
preparations (Lawson, 1996).
An alternative to the kinetic mechanism would be
founded on a nonequivalency in reactivity of different
RSOH, based on differences in nucleo-/electrophilicity
or even steric factors. This may result in a selectivity
or enrichment in certain alk(en)yl species linked to the
sulfoxide (-S(O)-) moiety during thiosulfinate forma-
tion. However, based on the thiosulfinate profile pro-
duced in the 2:1 MCSO:ECSO binary mixture, where
equimolar levels of R1 and R2 groups were evolved as
thiosulfinates (Table 2), and the lack of prior evidence

Biogeneration of Allium Thiosulfinates
J. Agric. Food Chem., Vol. 48, No. 12, 2000 6259
sulfoxide. This material is available free of charge via the
Internet at http://pubs.acs.org.
(1)) yields MeSOH which is “trapped” by reaction (step
(2)) with AllS(O)SAll. This yields MeS(O)SAll and
AllSOH (mechanism in Scheme 34 of Block, 1992). The
low steady-state levels of MeSOH favor step (3) over
step (4) for the fate of residual AllSOH, minimizing the
formation of MeSS(O)All (which occurs only by step (4))
in this system. The reaction system with the greater
initial MCSO level apparently generated sufficient
steady-state MeSOH levels to create flux through step
(4), causing the reduced MeS(O)SAll:MeSS(O)All ratio.
When 11 mM 2-PeCSO was reacted with 2.5 mM
MeS(O)SMe under standard conditions for 3 h, a molar
ratio of MeS(O)SAll:MeSS(O)All of 1:3 was found.
Referring to Scheme 2, with 2-PeCSO being considered
a “fast” reacting substrate in step (1), only the MeSS-
(O)All regioisomer is formed in step (2), and AllSOH
apparently reached steady-state levels sufficient for
making reaction step (4) more favorable than step (3)
(in contrast to the MCSO + AllS(O)SAll system dis-
cussed above). This was also indicated by the observed
product molar ratio of AllS(O)SAll:MeS(O)SAll of 1:1.24
found for this reaction system. Collectively, the results
of these experiments are consistent with a kinetic/mass
action account of the evolution of regioisomeric thiosul-
finates in dynamic systems representative of Allium
tissue preparations. (Of course, heterologous thiosulfi-
nates can also be formed by the random condensation
of simultaneously formed RSOH, where R1 * R2 (Scheme
1)).
LITERATURE CITED
Block, E. The organosulfur chemistry of the Genus Allium -
implications for the organic chemistry of sulfur. Angew.
Chem., Int. Ed. Engl. 1992, 31, 1135-1178.
Block, E.; Ahmad, S.; Catalfamo, J . L.; J ain, M. K.; Apitz-
Castro, R. Antithrombotic organosulfur compounds from
garlic: structural, mechanistic, and synthetic studies. J . Am.
Chem. Soc. 1986, 108, 7045-7055.
Block, E.; Gulati, H.; Putnam, D.; Sha, D.; You, N.; Zhao, S.-
H. Allium chemistry: synthesis of 1-[alk(en)ylsulfinyl]propyl
alk(en)yl disulfides (cepaenes), antithrombotic flavorants
from homogenates of onion (Allium cepa). J . Agric. Food
Chem. 1997, 45, 4414-4422.
Block, E.; Naganathan, S.; Putnam, D.; Zhao, S.-H. Allium
chemistry: HPLC analysis of thiosulfinates from onion,
garlic, wild garlic (ramsoms), leek, scallion, shallot, elephant
(great-headed) garlic, chive and Chinese chive. Uniquely
high allyl-to-methyl ratios in some garlic samples. J . Agric.
Food Chem. 1992, 40, 2418-2430.
Bradford, M. M. A rapid and sensitive method for the quan-
titation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 1976, 72,
248-254.
Carson, J . F.; Lundin, R. E.; Lukes, T. M. The configuration
of (+)-S-(1-propenyl)-L-cysteine S-oxide from Allium cepa.
J . Org. Chem. 1966, 31, 1634-1635.
Chu, N. T.; Clydesdale, F. M. Reactions between amino acids
and organic acids: reaction of tryptophan and pyruvic acid.
J . Food Sci. 1976, 41, 891-894.
Deleuze, H.; Langrand, G.; Millet, H.; Baratti, J . Buono, G.;
Triantaphylides, C. Lipase-catalyzed reaction in organic
media: competition and applications. Biochim. Biophys.
Acta 1987, 911, 117-120.
CONCLUSIONS
The alliinase-based system for biogeneration of thio-
sulfinates and PTSO described in this report has several
applications. It can be used as a means to prepare
sufficient quantities of thiosulfinates (or PTSO), either
in pure form or as mixtures of controlled composition,
for the systematic evaluation of structure-function
relationships. Such functional activities that remain to
be explored include biological and selected chemical
properties and reactivities. A second potential use would
be as a tool to study the dynamics of organosulfur
transformation under conditions that simulate ACSO
profiles of the various Allium species as well as condi-
tions that are used for processing Allium tissues into
various derivatives. Last, this thiosulfinate- and PTSO-
generating system provides a new tool for studying some
persistent and poorly understood problems in the pro-
cessing of Allium tissues, such as the nature of bitter-
ness and discoloration in onion and garlic (Fenwick and
Hanley, 1990).
Fenwick, G. R.; Hanley, A. B. The genus Allium - part II.
CRC Crit. Rev. Food Sci. Nutr. 1985, 22, 273-375.
Fenwick, G. R.; Hanley, A. B. Processing of alliums; use in
food manufacture. In Onions and Allied Crops, Volume III,
Biochemistry, Food Science, and Minor Crops; Brewster, J .
L., Rabinowitch, H. D., Eds.; CRC Press: Boca Raton, FL,
1990; pp 73-90.
Fersht, A. Enzyme Structure and Mechanism, 2nd ed.; W. H.
Freeman Co.: New York, 1985; pp 105-106.
Freeman, G. G.; Whenham, R. J . The use of synthetic (()-S-
1-propyl-L-cysteine sulphoxide and of alliinase preparations
in studies of flavour changes resulting from processing of
onion (Allium cepa L.). J . Sci. Food Agric. 1975, 26, 1333-
1346.
Fulcrand, H.; Benabdeljalil, C.; Rigaud, J .; Cheynier, V.;
Moutounet, M. A new class of wine pigments generated by
reaction between pyruvic acid and grape anthocyanins.
Phytochemistry 1998, 47, 1401-1407.
Kendler, B. S. Garlic (Allium sativum) and onion (Allium
cepa): a review of their relationship to cardiovascular
disease. Prev. Med. 1987, 16, 670-685.
Lancaster, J . E.; Kelly, K. E. Quantitative analysis of the
S-alk(en)yl-L-cysteine sulphoxides in onion (Allium cepa L.).
J . Sci. Food Agric. 1983, 34, 1229-1235.
Lancaster, J . E.; Shaw, M. L.; Randle, W. M. Differential
hydrolysis of alk(en)yl-L-cysteine sulphoxides by alliinase
in onion macerates: flavour implications. J . Sci. Food Agric.
1998, 78, 367-372.
Lau, B. H. S.; Tadi, P. P.; Tosk, J . M. Allium sativum (garlic)
and cancer prevention. Nutr. Res. 1990, 10, 937-948.
Lawson, L. D. The composition and chemistry of garlic cloves
and processed garlic. In Garlic: The Science and Therapeu-
tic Application of Allium sativum L. and Related Species,
2nd ed.; Koch, H. P., Lawson, L. D., Eds.; Williams &
Wilkins: Baltimore, MD, 1996; pp 37-108.
ACKNOWLEDGMENT
This work was supported by the College of Agricul-
tural and Life Sciences of the University of Wisconsins
Madison and The United States Department of Agri-
culture (Grants 96-35500-3352, 58-3148-7-031, and 97-
36200-5189). This manuscript is dedicated to the memory
of the late Muriel Parkin, former secretary to Professor
and 1947 ACS President, W. Albert Noyes J r., of the
University of Rochester, and former ACS secretary and
member of the ACS Committee on Professional Train-
ing.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
thiosulfinates, propanethial-S-oxide, and 1-propenyl-L-cysteine

6260 J. Agric. Food Chem., Vol. 48, No. 12, 2000
Shen and Parkin
Lawson, L. D.; Wood, S. G.; Hughes, B. G. HPLC analysis of
allicin and other thiosulfinates in garlic clove homogenates.
Planta Med. 1991, 57, 263-270.
Morimitsu, Y.; Morioka, Y.; Kawakishi, S. Inhibitors of platelet
aggregation generated from mixtures of Allium species and/
or S-alk(en)yl-L-cysteine sulfoxides. J . Agric. Food Chem.
1992, 40, 368-372.
Mu¨tsch-Eckner, M.; Meier, B.; Wright, A. D.; Sticher, O.
γ-Glutamyl peptides from Allium sativum bulbs. Phy-
tochemistry 1992, 31, 2389-2391.
Naganathan, S. Organosulfur chemistry of Allium species.
Ph.D. Thesis, SUNYsAlbany, 1992.
Nishimura, H.; Miziguchi, A.; Mizutani, J . Stereoselective
synthesis of S-(trans)-prop-1-enyl)-cysteine sulphoxide. Tet-
rahedron Lett. 1975, 37, 3201-3202.
Nock, L. P.; Mazelis, M. The C-S lyases of higher plants.
Direct comparison of the physical properties of homogeneous
alliin lyase of garlic (Allium sativum) and onion (Allium
cepa). Plant Physiol. 1987, 85, 1079-1083.
Prasad, K.; Laxdal, V. A.; Yu. M.; Raney, B. L. Antioxidant
activity of allicin, an active principle in garlic. Mol. Cell.
Biochem. 1995, 148, 183-189.
Randle, W. M.; Brussard, M. L. Streamlining onion pungency
analyses. Hortscience 1993, 28, 60.
Schwimmer, S.; Ryan, C. A.; Wong, F. F. Specificity of
L-cysteine sulfoxide lyase and partially competitive inhibi-
tion by S-alkyl-L-cysteines. J . Biol. Chem. 1964, 239, 777-
782.
Schwimmer, S.; Weston, W. J . Enzymatic development of
pyruvic acid in onion as a measure of pungency. J . Agric.
Food Chem. 1961, 9, 301-304.
Sendl, A.; Elbl, G.; Steinke, B.; Redl, K.; Breu, W.; Wagner,
H. Comparative pharmacological investigations of Allium
ursinum and Allium sativum. Planta Med. 1992, 58, 1-7.
Stoll, A.; Seebeck, E. Chemical investigations on alliin, the
specific principle of garlic. Adv. Enzymol. 1951, 11, 377-
400.
Synge, R. L. M.; Wood, J . C. (+)-(S-methyl-L-cysteine S-oxide)
in cabbage. Biochem. J . 1956, 64, 252-259.
Thomas, D. J .; Parkin, K. L. Immobilization and characteriza-
tion of C-S lyase from onion (Allium cepa) bulbs. Food
Biotechnol. 1991, 5, 139-159.
Tobkin, H. E.; Mazelis, M. Alliin lyase: preparation and
characterization of the homogeneous enzyme from onion
bulbs. Arch. Biochem. Biophys. 1979, 193, 150-157.
Yin, M.-C.; Cheng, W.-S. Antioxidant activity of several Allium
members. J . Agric. Food Chem. 1998, 46, 4097-4101.
Schmidt, N. E.; Santiago, L. M.; Eason, H. D.; Dafford, K. A.;
Grooms, C. A.; Link, T. E.; Manning, D. T.; Cooper, S. D.;
Keith, R. C.; Chance, W. O.; Walla, M. D.; Cotham, W. E.
Rapid extraction method of quantitating the lachrymatory
factor of onion using gas chromatography. J . Agric. Food
Chem. 1996, 44, 2690-2693.
Zheng, S.; Sun, X.; J iang, Y.; Zhao, T.; Zhu, H. Spectral and
Chemical Analysis of Organic Compounds (in Chinese);
J iangsu Science Publishers: Nanjing, P. R. China, 1988; pp
180-181.
Schwimmer, S. Characterization of S-propenyl-L-cysteine sul-
foxide as the principal endogenous substrate of L-cysteine
sulfoxide lyase of onion. Arch. Biochem. Biophys. 1969, 130,
312-320.
Received for review J une 8, 2000. Revised manuscript received
August 29, 2000. Accepted September 5, 2000.
J F000711G