First insights into the mode of action of a "lachrymatory factor synthase" - Implications for the mechanism of lachrymator formation in Petiveria alliacea, Allium cepa and Nectaroscordum species

Article abstract of DOI:10.1016/j.phytochem.2011.07.013

A study of an enzyme that reacts with the sulfenic acid produced by the alliinase in Petiveria alliacea L. (Phytolaccaceae) to yield the P. alliacea lachrymator (phenylmethanethial S-oxide) showed the protein to be a dehydrogenase. It functions by abstracting hydride from sulfenic acids of appropriate structure to form their corresponding sulfines. Successful hydride abstraction is dependent upon the presence of a benzyl group on the sulfur to stabilize the intermediate formed on abstraction of hydride. This dehydrogenase activity contrasts with that of the lachrymatory factor synthase (LFS) found in onion, which catalyzes the rearrangement of 1-propenesulfenic acid to (Z)-propanethial S-oxide, the onion lachrymator. Based on the type of reaction it catalyzes, the onion LFS should be classified as an isomerase and would be called a "sulfenic acid isomerase", whereas the P. alliacea LFS would be termed a "sulfenic acid dehydrogenase".

Full text of DOI:10.1016/j.phytochem.2011.07.013

Phytochemistry 72 (2011) 1939–1946
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
First insights into the mode of action of a ‘‘lachrymatory factor synthase’’ –
Implications for the mechanism of lachrymator formation in Petiveria alliacea,
Allium cepa and Nectaroscordum species
Quan He a, Roman Kubec b, Abhijit P. Jadhav a, Rabi A. Musah a,
⇑
a Department of Chemistry, University at Albany, State University of New York, Albany, NY 12222, USA
b
ˇ
ˇ
Department of Applied Chemistry, University of South Bohemia, Branišovská 31, 370 05 Ceské Budejovice, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
A study of an enzyme that reacts with the sulfenic acid produced by the alliinase in Petiveria alliacea L.
(Phytolaccaceae) to yield the P. alliacea lachrymator (phenylmethanethial S-oxide) showed the protein
to be a dehydrogenase. It functions by abstracting hydride from sulfenic acids of appropriate structure
to form their corresponding sulfines. Successful hydride abstraction is dependent upon the presence of
a benzyl group on the sulfur to stabilize the intermediate formed on abstraction of hydride. This dehy-
drogenase activity contrasts with that of the lachrymatory factor synthase (LFS) found in onion, which
catalyzes the rearrangement of 1-propenesulfenic acid to (Z)-propanethial S-oxide, the onion lachryma-
tor. Based on the type of reaction it catalyzes, the onion LFS should be classified as an isomerase and
would be called a ‘‘sulfenic acid isomerase’’, whereas the P. alliacea LFS would be termed a ‘‘sulfenic acid
dehydrogenase’’.
Received 16 December 2010
Received in revised form 11 July 2011
Available online 15 August 2011
Keywords:
Petiveria alliacea
Phytolaccaceae
Lachrymatory factor synthase
Sulfenic acid
Sulfenic acid dehydrogenase
Phenylmethanesulfenic acid dehydrogenase
Sulfenic acid isomerase
1-Propenesulfenic acid isomerase
1-Butenesulfenic acid isomerase
Onion
Ó 2011 Elsevier Ltd. All rights reserved.
Petiveriin
Isoalliin
Homoisoalliin
Sulfine
1. Introduction
and catalyzes its breakdown into two cleavage products. These are
the highly reactive 1-propenesulfenic acid [PSA, (3) Fig. 1], and
The familiar lachrymatory effect elicited when onions are cut is
caused by a volatile small molecule sulfine, (Z)-propanethial S-oxide
[PTSO, (1), Fig. 1] (Brodnitz and Pascale, 1971). The formation of this
molecule has been shown to be catalyzed by a novel enzyme, termed
lachrymatory factor synthase (LFS, Fig. 1) (Imai et al., 2002). The
a-aminoacrylic acid (not shown), which undergoes rapid hydrolysis
toyieldammoniaandpyruvate. PSA(3)is thenacteduponbytheLFS,
and the highly volatile and lachrymatory PTSO (1) is formed.
Prior to the discovery of the LFS, it was believed that PTSO was
formed via a [1,4]-sigmatropic rearrangement of PSA (3) at the
onion alliinase active site, by analogy to non-enzymatic conversion
of PSA (3) to PTSO (1) at elevated temperatures in the gas phase
(Block et al., 1996). It has since been shown however, that in the
absence of the onion LFS, onion alliinase is incapable of catalyzing
the formation of PTSO (1). Thus, when the alliinase alone interacts
with isoalliin (2), the PSA (3) product undergoes a variety of reac-
tions that yield a plethora of downstream organosulfur compounds
(Block, 1992, 2010), but no PTSO (1) is formed (Imai et al., 2002;
Eady et al., 2008). Although the onion LFS has been cDNA cloned,
little is known of its catalytic mechanism.
PTSO precursor is (E)-S-(1-propenyl)-L-cysteine S-oxide [isoalliin,
(2)], a cysteine sulfoxide derivative that is constitutively present in
the cytoplasm of onion cells. When onion tissue is disrupted, alliin-
ase, whichis aC–Slyasepresentinonioncellvacuoles(Lancasterand
Collin, 1981; Pickering et al., 2009), comes into contact with isoalliin,
Abbreviations: BAD, benzyl alcohol dehydrogenase; BSA, 1-butenesulfenic acid;
BTSO, butanethial S-oxide; LFS, lachrymatory factor synthase; PMSA, phen-
ylmethanesulfenic acid; PMTSO, phenylmethanethial S-oxide; PSA, 1-propenesulfe-
nic acid; PTSO, propanethial S-oxide; SAD, sulfenic acid dehydrogenase; TASO,
thioacrolein S-oxide.
In previous work conducted on the Amazonian medicinal plant
Petiveria alliacea L. (Phytolaccaceae), it was shown that like onion,
⇑
Corresponding author.
E-mail address: musah@albany.edu (R.A. Musah).
0031-9422/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2011.07.013

1940
Q. He et al. / Phytochemistry 72 (2011) 1939–1946
Fig. 1. Lachrymator formation in onion (A. cepa) and P. alliacea. In onion, upon tissue disruption, the onion LFS acts on PSA, which is produced by the catalytic action of the
onion alliinase on isoalliin, to afford the lachrymator PTSO. In P. alliacea, upon tissue disruption, the P. alliacea LFS acts on PMSA, which is furnished by the catalytic action of
the P. alliacea alliinase on petiveriin, to yield the lachrymator PMTSO. In Allium subg. Nectaroscordum species, the sulfenic acid BSA, which is furnished by the action of
alliinase on homoisoalliin, is acted upon by an LFS to give the lachrymator BTSO.
P. alliacea contains not only an alliinase, but also an LFS-type en-
zyme that catalyzes the formation of a sulfine lachrymator, (Z)-
phenylmethanethial S-oxide [PMTSO (4), Fig. 1] (Kubec et al.,
2003; Musah et al., 2009a,b). Similar to onion, the P. alliacea alliin-
Mechanistic studies involving determination of the kinetics of LF
synthases are complicated by the following unique challenges:
(a) their substrates are reactive intermediates (sulfenic acid) which
are created in situ by the enzyme alliinase; (b) quantification of the
substrate (which again, is a sulfenic acid intermediate) is hampered
by the fact that under the reaction conditions, two molecules of
sulfenic acid condense with loss of water to form a thiosulfinate
product; (c) the product of the action of the LFS (i.e., a sulfine) is
itself labile and is degraded under the reaction conditions to an
aldehyde; (d) because the action of the LFS requires the presence
of an alliinase (in order to produce the sulfenic acid substrate), it
is challenging to ascribe any particular kinetic effect that one
might observe, to a particular enzyme in the system; and (e) the
results of solvent isotope effect studies would be ambiguous, since
this represents a two-enzyme system, and convincingly attributing
the observed effects to one enzyme and not the other is difficult.
As such, very little is known about the mechanisms of this class
of enzymes, despite their importance in plant defense. Thus, we
sought to glean information about the mode of action of this
novel class of enzymes by alternative methods, beginning first with
the P. alliacea LFS. Herein, we report the results of our studies. Our
observations suggest that the P. alliacea LFS is actually a dehydro-
genase that acts on the sulfenic acid produced by the P. alliacea
alliinase, to form PMTSO (4) via a hydride transfer mechanism.
ase acts on a cysteine sulfoxide derivative, in this case, S-benzyl-L-
cysteine S-oxide [petiveriin (5)], to yield a sulfenic acid [phen-
ylmethanesulfenic acid – PMSA (6)] that is acted upon by an LFS
to furnish the sulfine, PMTSO (4) (Fig. 1).
Recently, another lachrymatory sulfine, (Z)-butanethial S-oxide
[BTSO (7)], was identified in two Allium subg. Nectaroscordum spe-
cies (A. siculum and A. tripedale). It was shown that this higher
homologue of the onion lachrymator is formed from (E)-S-(1-bute-
nyl)cysteine S-oxide [homoisoalliin (8)] via 1-butenesulfenic acid
[BSA (9)] [(Kubec et al., 2010; Block et al., 2010), Fig. 1]. It is reason-
able to assume that the conversion of BSA (9) into BTSO (7) is also
mediated by an LFS-type enzyme.
A cursory glance at the LFS-catalyzed reactions that yield the
sulfines in onion, subgenus Nectaroscordum species and P. alliacea
implies that they may employ similar mechanistic tactics in their
reactions with isoalliin (2), homoisoalliin (8) and petiveriin (5),
respectively (Fig. 1). However, a fundamental difference in the
structures of the sulfenic acid precursors to the sulfines [i.e., PSA
(3), PMSA (6) and BSA (9), Fig. 1] suggests otherwise. Whereas
PSA (3) and BSA (9) possess a,b-unsaturation, a vestige of the iso-
alliin/homoisoalliin from which they are derived, PMSA (6) is
devoid of this feature. A consequence of this is that the sp2-hybrid-
ized carbon that is attached to sulfur in the sulfenic acids from
which PTSO (1) and BTSO (7) originate [i.e., PSA (3) and BSA (9)
respectively] is more highly oxidized than that [i.e., PMSA (6)] from
which PMTSO (4) is derived. Therefore, it is clear that despite the
fact that the LF synthases from onion, Nectaroscordum species
and P. alliacea act on sulfenic acids to yield sulfine lachrym-
ators, they perform these reactions by different mechanisms.
2. Results and discussion
2.1. Substrate specificity of the P. alliacea lachrymatory factor
synthase
The substrates for LF synthases have been shown to be the sul-
fenic acid intermediates furnished by the reaction of an alliinase
with a suitable S-substituted cysteine S-oxide precursor (Imai

Q. He et al. / Phytochemistry 72 (2011) 1939–1946
1941
et al., 2002; Musah et al., 2009a,b). When an alliinase catalyzes the
cleavage of S-substituted cysteine S-oxides in the absence of an
LFS, the resulting sulfenic acids condense spontaneously to form
thiosulfinates (Table 1). Therefore, thiosulfinate products serve as
direct evidence of the formation of sulfenic acids by an alliinase.
Like the onion LFS, the P. alliacea LFS produces a sulfine only in
the presence of both an S-substituted cysteine S-oxide derivative
and an alliinase. In the case of P. alliacea, it has also been shown
that when suitable S-substituted cysteine S-oxides are exposed to
its LFS/alliinase complex, the relative amounts of thiosulfinate
and sulfine formed are dependent upon the ratio of LFS to alliinase
(Musah et al., 2009a). When the LFS to alliinase molar ratio is 5:1,
Table 1
The products formed when various substrates were exposed to alliinase in the presence of the P. alliacea LFS enzyme. The analysis of the scope of reactivity of the LFS was
conducted by reacting various S-substituted cysteine sulfoxides with an LFS/alliinase complex, in which the two proteins were present in a 5:3 M ratio respectively.

1942
Q. He et al. / Phytochemistry 72 (2011) 1939–1946
Table 2
The products formed when petiveriin was exposed to the P. alliacea alliinase under various conditions.
reaction products1,2
O
S
S
Entry
no.
S
O
substrate
added enzyme and (or) cofactor
thiosulfinate (petivericin, 10)
sulfine (PMTSO, 4)
O
S
NH2
COOH
alliinase
1
2
3
4
+
+
+
+
+
+
+
+
+
+
(5)
petiveriin
alliinase + LFS
petiveriin
1.0
0.08
0.08
alliinase + NAD+
alliinase + NADP+
alliinase + FAD
petiveriin
petiveriin
petiveriin
petiveriin
5
6
alliinase + FMN
alliinase + NAD+ + LFS
alliinase + NADP + + LFS
alliinase + FAD + LFS
alliinase + FMN + LFS
2.4
2.0
1.0
1.0
petiveriin
petiveriin
petiveriin
petiveriin
7
8
9
10
a
The ‘‘+’’ and ‘‘ꢀ’’ signs indicate whether the product was formed (‘‘+’’) or not formed (‘‘ꢀ’’).
b
The values represent the ratios of sulfine formed relative to the amount formed in the system consisting of LFS, petiveriin and alliinase (entry no. 2).
all of the sulfenic acid produced by the alliinase is sequestered, so
that only the sulfine product is observed. On the other hand, when
the LFS to alliinase molar ratio is 65:2, the LFS is unable to seques-
ter all of the sulfenic acid formed by the alliinase, with the result
that some of the sulfenic acid self-condenses to form thiosulfinate
(Musah et al., 2009a). Therefore, our analysis of the scope of reac-
tivity of the LFS was conducted by reacting various S-substituted
cysteine S-oxides with a P. alliacea-derived LFS/alliinase complex,
in which the two proteins were present in a 5:3 M ratio respec-
tively. It was anticipated that the use of this LFS/alliinase ratio
would reveal whether sulfenic acids were available to the LFS
(through observance of thiosulfinates) in those cases where no sul-
fine was formed by the LFS. The group of compounds tested is
listed in Table 1, and is comprised of petiveriin and its derivatives
(compounds 5, 11 and 15, a chain elongated petiveriin homologue
by HPLC, UV–Vis, and electrospray ionization mass spectrometry
(ESI-MS).
Examination of the results of these experiments revealed
clear structural distinctions between those sulfenic acids with
which the LFS reacted to form sulfines, and those with which
it did not. When the LFS/alliinase complex was reacted with
its natural substrate petiveriin (5), as expected, PMTSO (4) was
formed (Table 1). Similarly, sulfines were also formed when
the protein complex was reacted with
a- and ring-substituted
derivatives of petiveriin (5) (Table 1, compounds 11 and 15).
However, with the presence of an additional methylene group
between the benzene ring and the
a-carbon adjacent to the sul-
fur, no sulfine formation was observed (Table 1, compound 19).
Additionally, no sulfines were formed from precursor molecules
2, 23, 27 and 31, including the precursor to the onion lachryma-
tor [isoalliin (2)]. On the other hand, formation of the corre-
sponding thiosulfinates was observed for all of the other
compounds tested.
(compound 19), and various S-alk(en)yl-L-cysteine S-oxides (com-
pounds 2, 23, 27 and 31). The substrate specificity of the P. alliacea
LFS was determined by tracking the formation of sulfine products

Q. He et al. / Phytochemistry 72 (2011) 1939–1946
1943
2.2. P. alliacea LFS catalyzes hydride transfer from a sulfenic acid
intermediate
inase and LFS were reacted together in the presence of added NAD+
or NADP+, PMTSO was produced at relative concentrations of 2.4
and 2.0 times higher, respectively, than observed with a combina-
tion of petiveriin (5), alliinase and LFS (Table 2, reactions 7 and 8).
To further define the conditions under which sulfines are
formed, petiveriin (5) was exposed to the P. alliacea alliinase in
the presence of selected enzyme cofactors. The results of these
experiments are shown in Table 2. Although the combination of
petiveriin (5) and the alliinase in the absence of the P. alliacea
LFS produced only the corresponding thiosulfinate (10) (Table 2,
reaction 1), the addition of either of the naturally occurring cofac-
tors NAD+ or NADP+ to the reaction system resulted in the forma-
tion of PMTSO (4), the P. alliacea lachrymatory sulfine, albeit at
concentrations 13 times lower than were observed in the presence
of LFS (Table 2, reactions 3 and 4). The use of the redox cofactors
FAD or FMN in a similar experiment yielded only the thiosulfinate,
and no sulfine (Table 2, reactions 5 and 6). When petiveriin (5), alli-
2.3. P. alliacea LFS substrate specificity studies
Although onion, Nectaroscordum species and P. alliacea contain
several constitutively present S-substituted cysteine S-oxides, only
a single compound in each plant serves as a precursor for the lach-
rymator that is formed on tissue disruption. However, despite the
limited number of sulfine precursors that are available in P. allia-
cea, its LFS has an active site that can accommodate not only its
natural substrate [i.e., PMSA (6)], but other non-natural sulfenic
acids (Table 1, compounds 11 and 15). Thus, the LFS is capable of
transforming PMSA (6) analogues that bear substituents on the
Fig. 2. Proposed mechanisms for formation of the P. alliacea lachrymator PMTSO via a hydride transfer from PMSA to an electron acceptor.

1944
Q. He et al. / Phytochemistry 72 (2011) 1939–1946
carbon
a
to the benzene ring (Table 1, compound 11), and on the
readily formed even when the hydrogens at both ortho positions
were replaced with methyl groups (Table 1, compound 15). This is-
sue is circumvented in Pathway C, because the Pathway C mecha-
nism does not involve attack at the ortho carbons of the benzene
ring.
The nature of the discrimination shown by the P. alliacea LFS for
different substrates, coupled with the finding that NAD(P)+, in the
presence of the alliinase and petiveriin, furnishes PMTSO (4), pro-
vided additional clues regarding the mode of action of LFS reac-
tions with sulfenic acids. Our results show that an essential
structural requirement for the LFS-sulfenic acid substrate is that
benzene ring (Table 1, compound 15). However, this ability does
not extend to the chain-elongated derivative (Table 1, compound
19), nor to the sulfenic acids derived from propiin and isoalliin (Ta-
ble 1, compounds 2 and 27). Ethiin (Table 1, compound 23) also did
not react to form its corresponding sulfine. That sulfenic acid sub-
strates were available to the LFS for all of the compounds tested
was demonstrated by the observation of thiosulfinates, the con-
densation products of the sulfenic acids formed by the alliinase,
in all cases. It is clear from the results that the sulfenic acids corre-
sponding to compounds 5, 11 and 15 are distinguished from those
corresponding to 2, 19, 23 and 27 in that abstraction of an
a
-hydro-
it possess a free hydrogen at a fully saturated a-carbon that is adja-
gen (either as H+ or Hꢀ) would yield a resonance stabilized benzylic
ion.
cent to an unsaturated center capable of providing significant res-
onance stabilization. Petiveriin (5), as well as compounds 11 and
15, meet this requirement. Although compound 19 possesses an
It was surprising though, that the observation of sulfine forma-
tion did not require the presence of the LFS, and that the cofactors
NAD+ or NADP+ (but not FAD or FMN), could serve in its stead, al-
beit much less efficiently (Table 2, entries 3–6). The inclusion of
NAD(P)+ with petiveriin, alliinase and the LFS resulted in increased
product formation, when compared to identical reactions in which
NAD(P)+ were not present. However, FAD and FMN did not elicit
this effect.
a-hydrogen at a fully saturated carbon, the a-carbon to which it
is bonded is not adjacent to an unsaturated center. The situation
is similar for compounds 23 and 27, neither of which is converted
to a sulfine, even though both are converted to sulfenic acids by the
alliinase. That sulfenic acids are actually formed in these cases is
evidenced by the formation of thiosulfinates under the reaction
conditions. From the pattern that emerged, it was predictable that
the sulfenic acid PSA (3) derived from isoalliin (2), the precursor of
the onion lachrymator, would not react with the P. alliacea LFS, as
was observed (Table 1).
A characteristic function of NAD(P)+ in biological systems is the
transfer of electrons from one redox center to another. In these
reactions, a proton and two electrons, in the form of hydride, are
transferred from a donor to NAD(P)+ to form NAD(P)H, with con-
comitant release of H+. Therefore, the observation that NAD(P)+
in the presence of petiveriin (5) and alliinase furnished PMTSO
(4), suggested not only that the cofactor served the role of abstract-
The reaction of alliin (Table 1, compound 31) with the P. alliacea
LFS/alliinase complex is an interesting case. Like the other sub-
strates that were oxidized to sulfines, it possesses an
a-hydrogen
attached to the carbon that is adjacent to an unsaturated center.
Therefore, loss of hydride from alliin (31) might be expected to oc-
cur in a fashion similar to that observed for the benzyl systems,
which would yield (Z)-thioacrolein S-oxide [TASO (34)] (Pelloux-
Léon et al., 1997). Analogous to what has been observed for the
PTSO (1) that is formed from isoalliin (2), which decomposes in
an aqueous environment to form propanal (Brodnitz and Pascale,
1971), TASO (34) could undergo a similar reaction to form acrolein
(2-propenal). Therefore, the observation of acrolein would serve as
evidence for the formation of TASO (34). Such a finding would im-
ply that the P. alliacea LFS may be similar to bacterial benzyl alco-
hol dehydrogenases (BADs), in terms of their ability to catalyze the
oxidation of both benzylic and allylic substrates. BADs have been
shown to catalyze the oxidation to aldehydes of both benzylic
and allylic alcohols (but not their non-allylic or non-benzylic coun-
terparts) through an intermediate that is partially positively
charged (Curtis et al., 1999).
ing an a-hydride from the sulfenic acid to afford the sulfine prod-
uct, but also that the LFS functions similarly.
Fig. 2 outlines three possible mechanisms that involve hydride
transfer from PMSA (6) to an electron acceptor to form PMTSO
(4). In Pathway A, a base at the active site abstracts a proton from
the benzylic carbon, and the ortho carbon of the benzene ring ab-
stracts the proton from the oxygen atom of the PMSA (6). The
resulting intermediate then looses hydride to an electron acceptor,
which yields the product sulfine. In Pathway B, an active site base
abstracts the proton that is attached to the PMSA (6) oxygen. Single
bond formation occurs between oxygen and the ortho carbon of the
benzene ring, and double bond formation occurs between the ring
and the
a-carbon, with loss of hydride from the a-carbon to an
electron pair acceptor. The resulting 1,2-oxathiole intermediate
then rearranges via a six electron exocyclic ring opening to form
the product. Pathway C is similar to Pathway B in that hydride is
lost from the benzylic carbon. However, Pathway C differs from B
in that a bicyclic intermediate is not formed, and there is consider-
able positive charge buildup in the intermediate.
However, under the reaction conditions employed, neither for-
mation of TASO (34), acrolein, nor any other product besides the
thiosulfinate allicin (33) was observed, and it is concluded that
the sulfenic acid derived from alliin (31) is not a suitable substrate
for the P. alliacea LFS. It is proposed that the reason 2-propen-
esulfenic acid does not serve as a suitable substrate for the LFS, de-
The observations made thus far favor Pathway C. For starters,
Pathways A and B appear rather unlikely due to the lack of an obvi-
ous driving force to destroy the aromatic ring system. Second,
Pathway A involves the transfer of a proton from the PMSA (6) oxy-
gen to the ortho carbon of the benzene ring. Sulfenic acid protons
are acidic and have been shown to readily exchange in D2O (Penn
et al., 1978; Davis and Billmers, 1984; Ishii et al., 1996; Goto et al.,
1997). Therefore, if the solvent system is D2O, then this exchange
would result in deuteration of the ring, regardless of whether the
ring deuteration step is rate-limiting. This would result in forma-
tion of at least some deuterium-labeled PMTSO (4). However,
when petiveriin (5), alliinase and the LFS were reacted in D2O
and the HPLC-purified product was analyzed by ESI-TOF, only unla-
beled PMTSO (4) was detected (data not shown). Third, Pathways A
and B both involve attack at the ortho carbons of the benzene ring.
It might be expected that blocking these positions with substitu-
ents would curtail formation of the corresponding sulfine for steric
reasons. However, we observed that the corresponding sulfine was
spite having an abstractable
a-hydrogen on the carbon that is
adjacent to an unsaturated center, is that the allyl substituent is
not as well accommodated by the enzyme active site when com-
pared to the benzyl substituent, and therefore, its corresponding
intermediate is not as well stabilized. This could indicate that pi-
pi stacking interactions between the benzene ring of the substrate
and active site residues are important determinants for substrate
binding. In support of this, it has been observed that the Km for
the interaction of alliin (31) with the P. alliacea alliinase was
4.32 mM, whereas that for petiveriin (5) was 0.39 mM (Musah
et al., 2009b). This implies that the binding of alliin (31) to the alli-
inase active site is ꢁ11 times less tight than that for petiveriin (5).
Additionally, the catalytic efficiencies of the alliinase for alliin (31)
and petiveriin (5) were found to be 42,800 and 100,000 sꢀ1 Mꢀ1
respectively, indicating that the breakdown of petiveriin (5) by

Q. He et al. / Phytochemistry 72 (2011) 1939–1946
1945
the alliinase is much more efficient than for alliin (31) (Musah
et al., 2009b). Although it is not customary to compare the catalytic
efficiencies of distinct enzymes that catalyze different reactions, it
has been found that the P. alliacea LFS shares with the P. alliacea
alliinase four of the alliinases five protein subunits (Musah et al.,
2009a,b). Thus, it may be possible that as a consequence of the
structural similarity between the P. alliacea LFS and alliinase, the
discrimination shown by the alliinase for different S-substituted
cysteine S-oxides based upon differing R group substituents, is
shared by the LFS. However, the possibility cannot be ruled out
that TASO (34) is formed, but is quickly transformed into deriva-
tives that we did not detect under the experimental conditions
used.
The loss of hydride from the sulfenic acid, a conclusion implied
by the ability of NAD(P)+ to furnish PMTSO (4) under appropriate
conditions, coupled with the aforementioned substrate structural
requirements, implicates the involvement of an intermediate pos-
sessed of significant positive charge buildup. The Pathway C mech-
anism shown in Fig. 2 is in alignment with this conclusion. Once a
sulfenic acid is formed by the alliinase, it has two possible fates. It
can condense with another sulfenic acid molecule to form a thio-
sulfinate, or be sequestered by the LFS and undergo further trans-
ent in subgenus Nectaroscordum species should be termed 1-but-
enesulfenic acid isomerase.
The fact that various redox cofactors are capable of promoting
the formation of sulfines from sulfenic acids in the absence of
SAD implies that SAD itself utilizes a cofactor as part of its catalytic
mechanism. Detailed exploration of this hypothesis is the subject
of ongoing studies.
3. Conclusions
These studies of the enzyme isolated from P. alliacea that exhib-
its LFS activity show it to be a dehydrogenase that functions by
abstracting hydride from sulfenic acids of suitable structure. A ben-
zene ring adjacent to the carbon
a to sulfur is an important deter-
minant for successful hydride abstraction because it stabilizes the
positively charged intermediate formed upon abstraction of hy-
dride. The ability of various redox cofactors to promote formation
of sulfines from sulfenic acids in the presence of this enzyme im-
plies that the enzyme itself might utilize a co-factor as part of its
catalytic mechanism. Although onion (Allium cepa), Allium subg.
Nectaroscordum species and P. alliacea contain LFSs that mediate
formation of sulfines, the mechanisms by which they function
are fundamentally different. Sulfine formation in onion and Nec-
taroscordum species from the precursor sulfenic acids is formally
a rearrangement reaction, whereas sulfine formation from the pre-
cursor sulfenic acid in P. alliacea is formally an oxidation. Thus, it
may be more appropriate to term the onion and Nectaroscordum
LFSs ‘‘sulfenic acid isomerases’’ and the P. alliacea LFS a ‘‘sulfenic
acid dehydrogenase’’.
formation. When the sulfenic acid possesses an
a-hydrogen
adjacent to a benzene ring, this hydrogen is abstracted by the
LFS to yield an intermediate in which significant positive charge
buildup can be stabilized by the adjacent aromatic ring. Either con-
comitantly, or as a second step, a bonding interaction is established
between a basic residue at the active site, and the proton that is at-
tached to the oxygen of the sulfenic acid. This proton is ultimately
abstracted by the active site base, and subsequently released to the
surroundings. Although the timing of abstraction of the proton
from the sulfenic acid is unclear, its loss, along with the loss of hy-
4. Experimental
dride from the
a-carbon of the sulfenic acid, would yield the sul-
fine. It is also possible that the divalent sulfur contributes
significantly to stabilization of the cation. In that case, a protonated
sulfine would be formed which could be expected to release its
proton faster than occurs in the sulfenic acid.
The Pathway C mechanism proposed in Fig. 2 provides a ratio-
nale for why the P. alliacea LFS does not catalyze sulfine formation
from isoalliin (2), the precursor of the onion lachrymator. As a con-
4.1. Plants and materials
Unless otherwise noted, all chemicals were obtained from the
Sigma–Aldrich Chemical Company (St. Louis, MO, USA). Whole
fresh plants of P. alliacea were obtained from Native Habitat Land-
scaping (Vero Beach, FL, USA), and stored at ꢀ30 °C until analysis. A
voucher specimen is deposited at the herbarium PIHG at the Flor-
ida Department of Agriculture and Consumer Services, Division of
Plant Industry, Gainesville, FL, USA, under accession number 7801.
sequence of its a,b-unsaturation, PSA (3) is an isomer of its corre-
sponding sulfine, PTSO (1) (i.e., the onion lachrymator). Thus, it
need simply undergo a rearrangement in order to furnish the sul-
fine. Apparently, it is this rearrangement that is catalyzed by the
onion LFS. The absence of
a,b-unsaturation in petiveriin (5) means
4.2. ESI-TOF
that the -carbon in PMSA (6) is less oxidized than that in PSA.
a
Thus, sulfine formation from the PMSA (6) intermediate requires
loss of two electrons and two protons, which is the reaction cata-
lyzed by the P. alliacea LFS. According to the Nomenclature Com-
mittee of the International Union of Biochemistry and Molecular
Biology (NC-IUBMB), the naming and classification of enzymes is
based on the reactions they catalyze. Given that PMTSO (4) forma-
tion involves oxidation of PMSA (6) via loss of an electron pair and
two protons to an acceptor, it would be more appropriate to refer
to the P. alliacea enzyme that mediates this process by the common
name ‘‘sulfenic acid dehydrogenase’’ (SAD), and more specifically
as ‘‘phenylmethanesulfenic acid dehydrogenase’’. According to this
categorization, the enzyme would belong to the enzyme classifica-
tion 1 (EC 1) group of oxidoreductases. Regardless of mechanism,
enzymes that catalyze the structural rearrangement of isomers
are termed isomerases, and thus the onion LFS can be classified
as such, even though the mechanism by which the isomerization
occurs is unknown. Thus, by the current convention, the onion
LFS would be categorized as a ‘‘sulfenic acid isomerase’’, or more
specifically as 1-propenesulfenic acid isomerase. This would place
it in the EC 5 group of enzymes. Similarly, the LFS presumably pres-
An Agilent dual ESI source ESI-MSD-TOF mass spectrometer at
the Scripps Research Institute (La Jolla, CA) was used for accurate
mass determination of compounds. A mixture of standards (Agilent
ESI-TOF TUNE mix) was used to spray two lock masses at 121 and
922 from the second sprayer for internal calibration of each mass
spectrum to get the highest mass accuracy. The samples were
introduced by flow injection analysis at 4000 V using an 8
lL sam-
ple injection at 100 L/min using an Agilent 1100 HPLC system
l
with a solvent consisting of 50% ethanol.
4.3. Reference compounds
S-Substituted-L-cysteines and the corresponding S-substituted-
L-cysteine S-oxides, as well as petivericin, were synthesized
according to the methods of Kubec and Musah (2001) and Kubec
et al. (2002). Isoalliin was isolated from white onion bulbs obtained
at a local market according to the method of Shen and Parkin
(2000). 2,6-Dimethylbenzyl bromide was synthesized according
to the method of Soloshonok et al. (2001).

1946
Q. He et al. / Phytochemistry 72 (2011) 1939–1946
the lachrymatory factor of the onion (Allium cepa). J. Am. Chem. Soc. 118, 7492–
4.4. Purification of alliinase and LFS from P. alliacea
7501.
Brodnitz, M.H., Pascale, J.V., 1971. Thiopropanal S-oxide: a lachrymatory factor in
onions. J. Agric. Food Chem. 19, 269–272.
The alliinase and LFS enzymes in P. alliacea were purified
Curtis, A.J., Shirk, M.C., Fall, R., 1999. Allylic or benzylic stabilization is essential for
catalysis by bacterial benzyl alcohol dehydrogenases. Biochem. Biophys. Res.
Commun. 259, 220–223.
Davis, F.A., Billmers, R.L., 1984. Chemistry of sulfenic acids. 6. Structure of simple
sulfenic acids generated by flash vacuum pyrolysis. J. Org. Chem. 50, 2593–
2595.
Eady, C.C., Kamoi, T., Kato, M., Porter, N.G., Davis, S., Shaw, M., Kamoi, A., Imai, S.,
2008. Silencing onion lachrymatory factor synthase causes a significant change
in the sulfur secondary metabolite profile. Plant Physiol. 147, 2096–2106.
Goto, K., Holler, M., Okazaki, R., 1997. Synthesis, structure, and reactions of a
sulfenic acid bearing a novel bowl-type substituent: the first synthesis of a
stable sulfenic acid by direct oxidation of a thiol. J. Am. Chem. Soc. 119, 1460–
1461.
according to the protocols of Musah et al. (2009a,b).
4.5. Determination of LFS substrate specificity
Sulfenic acid substrates with which the LFS could react were
generated in situ through the action of a P. alliacea alliinase/LFS
complex on cysteine sulfoxide derivatives as reported in the thesis
of He (2010). The reaction mixtures in 10 mM phosphate buffer, pH
8.0 (in a total volume of 1.0 mL), were minimally comprised of
1.5 mM substrate, 25
l
M pyridoxal 50-phosphate (PLP), 3.0
l
g of
He, Q., 2010. The alliinase and lachrymatory factor synthase systems in Petiveria
alliacea. Ph.D. Thesis, State University of New York at Albany, Albany, NY, USA.
Imai, S., Tsuge, N., Tomotake, M., Nagatome, Y., Sawada, H., Nagata, T., Kumagai, H.,
2002. An onion enzyme that makes the eyes water. Nature 419, 685.
Ishii, A., Komiya, K., Nakayama, J., 1996. Synthesis of a stable sulfenic acid by
oxidation of a sterically hindered thiol (thiophenetryptycene-8-thiol) and its
characterization. J. Am. Chem. Soc. 118, 12836–12837.
Kubec, R., Cody, R.B., Dane, A.J., Musah, R.A., Schraml, J., Vattekkatte, A., Block, E.,
2010. Applications of direct analysis in real time-mass spectrometry (DART-MS)
in Allium chemistry. (Z)-Butanethial S-oxide and 1-butenyl thiosulfinates and
their S-(E)-1-butenylcysteine S-oxide precursor from Allium siculum. J. Agric.
Food Chem. 58, 1121–1128.
Kubec, R., Kim, S., Musah, R.A., 2002. S-Substituted cysteine derivatives and
thiosulfinate formation in Petiveria alliacea – part II. Phytochemistry 61, 675–
680.
Kubec, R., Kim, S., Musah, R.A., 2003. The lachrymatory principle of Petiveria alliacea.
Phytochemistry 63, 37–40.
Kubec, R., Musah, R.A., 2001. Cysteine sulfoxide derivatives in Petiveria alliacea.
Phytochemistry 58, 981–985.
Lancaster, J.E., Collin, H.A., 1981. Presence of alliinase in isolated vacuoles and of
alkyl cysteine sulfoxides in the cytoplasm of bulbs of onion (Allium cepa). Plant
Sci. Lett. 22, 169–176.
Musah, R.A., He, Q., Kubec, R., 2009a. Discovery and characterization of a novel
lachrymatory factor synthase (LFS) in Petiveria alliacea and its influence on
alliinase-mediated formation of biologically active organosulfur compounds.
Plant Physiol. 151, 1294–1303.
Musah, R.A., He, Q., Kubec, R., Jadhav, A., 2009b. Studies of a novel cysteine sulfoxide
lyase from Petiveria alliacea: the first heteromeric alliinase. Plant Physiol. 151,
1304–1316.
purified alliinase (ꢁ21 nM) and 5.7
l
g of purified LFS (ꢁ34 nM).
In those cases where the effects of cofactors were determined,
0.32 mM NAD(P)+, FAD or FMN was also present. The mixtures
were incubated for 20 min at room temperature, and then
10–20
analytical RP C-18 column (Microsorb-MV 100 Å, 250 ꢂ 4.6 mm,
m, Varian, Palo Alto, CA, USA) under the following conditions:
lL of the reaction solution was analyzed by HPLC using an
5
l
flow rate: 1.0 mL minꢀ1; mobile phase: water:acetonitrile (30:70,
v/v); detection wavelength: 210 nm. Eluted products were ana-
lyzed by UV–Vis and ESI-TOF. The reaction between the alliinase/
LFS complex and petiveriin was also conducted in 10 mM phos-
phate buffer prepared with D2O, and the HPLC eluted products
were analyzed by ESI-TOF.
Acknowledgements
The authors thank Distinguished Professor Dr. Eric Block for
helpful suggestions and critical reading of the manuscript. The
financial support of the National Science Foundation (Grant
#0239755 to R.A.M.) and the Research Foundation of SUNY are
gratefully acknowledged.
Pelloux-Léon, N., Arnaud, R., Ripoll, J.L., Beslin, P., Vallée, Y., 1997. Thioacrolein S-
oxide. Tetrahedron Lett. 38, 1385–1388.
References
Penn, R.E., Block, E., Revelle, L.K., 1978. Flash vacuum pyrolysis studies. 5.
Methanesulfenic acid. J. Am. Chem. Soc. 100, 3622–3623.
Pickering, I.J., Sneeden, E.Y., Prince, R.C., Block, E., Harris, H.H., Hirsch, G., George,
G.N., 2009. Localizing the chemical forms of sulfur in vivo using X-ray
fluorescence spectroscopic imaging: application to onion (Allium cepa) tissues.
Biochemistry 48, 6846–6853.
Block, E., 1992. The organosulfur chemistry of the genus Allium – implications for
the organic chemistry of sulfur. Angew. Chem. Int. Ed. Engl. 31, 1135–1178.
Block, E., 2010. Garlic and Other Alliums: The Lore and the Science. Royal Society of
Chemistry, Cambridge, UK.
Block, E., Dane, E.J., Thomas, S., Cody, R.B., 2010. Applications of direct analysis in
real time mass spectrometry (DART-MS) in Allium chemistry. 2-Propenesulfenic
and 2-propenesulfinic acids, diallyl trisulfane S-oxide, and other reactive sulfur
compounds from crushed garlic and other Alliums. J. Agric. Food Chem. 58,
4617–4625.
Shen, C., Parkin, K.L., 2000. In vitro biogeneration of pure thiosulfinates and
propanethial-S-oxide. J. Agric. Food Chem. 48, 6254–6260.
Soloshonok, V.A., Tang, X., Hruby, V.J., 2001. Large-scale asymmetric synthesis of
novel sterically constrained 20,60-dimethyl- and
a
,20,60-trimethyltyrosine and
Block, E., Gillies, J.Z., Gillies, C.W., Bazzi, A.A., Putman, D., Revelle, L.K., Wang, D.,
Zhang, X., 1996. Allium chemistry: microwave spectroscopic identification,
mechanism of formation, synthesis, and reactions of (E,Z)-propanethial S-oxide,
phenylalanine derivatives via alkylation of chiral equivalents of nucleophilic
glycine and alanine. Tetrahedron 57, 6375–6382.