Diallyl Trisulfide Is a Fast H2S Donor, but Diallyl Disulfide Is a Slow One: The Reaction Pathways and Intermediates of Glutathione with Polysulfides

Article abstract of DOI:10.1021/acs.orglett.5b01962

Diallyl trisulfide (DATS) reacts rapidly with glutathione (GSH) to release H₂S through thiol-disulfide exchange followed by allyl perthiol reduction by GSH. Yet diallyl disulfide (DADS) only releases a minute amount of H₂S via a sluggish reaction with GSH through an α-carbon nucleophilic substitution pathway. The results clarify the misunderstanding of DADS as a rapid H₂S donor, which is attributed to its DATS impurity.

Full text of DOI:10.1021/acs.orglett.5b01962

Letter
pubs.acs.org/OrgLett
Diallyl Trisulfide Is a Fast H2S Donor, but Diallyl Disulfide Is a Slow
One: The Reaction Pathways and Intermediates of Glutathione with
Polysulfides
Dong Liang,† Haixia Wu,† Ming Wah Wong,‡ and Dejian Huang*,†,§
†Food Science and Technology Program, Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore
117543, Singapore
‡Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore
§National University of Singapore (Suzhou) Research Institute, 377 Lin Quan Street, Suzhou Industrial Park, Jiangsu 215123, China
S
* Supporting Information
ABSTRACT: Diallyl trisulfide (DATS) reacts rapidly with glutathione
(GSH) to release H2S through thiol−disulfide exchange followed by allyl
perthiol reduction by GSH. Yet diallyl disulfide (DADS) only releases a
minute amount of H2S via a sluggish reaction with GSH through an α-
carbon nucleophilic substitution pathway. The results clarify the
misunderstanding of DADS as a rapid H2S donor, which is attributed to
its DATS impurity.
a
Scheme 1
ydrogen sulfide (H2S) is a gaseous signaling molecule
Hthat exerts important regulatory functions in cardiovas-
cular, immune, gastrointestinal, endocrine, and nervous
systems.1 In mammalian tissues, H2S is produced endogenously
by four enzymes including cystathionine γ-lyase (CSE, EC
4.4.1.1), cystathionine β-synthase (CBS, EC 4.21.22), 3-
mercapto-pyruvate sulfurtransferase (3-MST, EC 2.8.1.2), and
cysteine aminotransferase (CAT, EC 2.6.1.3).2 Because of their
therapeutic potentials, a large number of H2S donors have been
reported.3
a
A = allyl.
Addition of DATS (50 μM) to GSH (10-fold excess)
solution led to an instant production of H2S (Figure 1A). In
contrast, addition of DADS (purified, 100 μM) to GSH (500
Diallyl trisulfide (DATS) and diallyl disulfide (DADS) are
the two major organosulfides in garlic oil.4 They are produced
from the decomposition of allicin and have been studied
extensively for their health benefits.5 In 2007 Benavides and co-
workers reported that DATS and DADS are donors of H2S.6
They showed that DATS and DADS could be converted into
H2S by human red blood cells or by rat aorta through a thiol,
mainly glutathione (GSH), dependent mechanism. Moreover,
the H2S produced was able to relax rat aorta rings. Based on
their observation that both DADS and DATS rapidly released
H2S when mixed with GSH, an α-carbon nucleophilic
substitution mechanism was proposed to explain the rapid
H2S generation from DADS (Scheme 1). First, DADS reacts
with GSH to generate S-allyl glutathione (GSA) and the key
intermediate allyl perthiol (ASSH), which quickly reacts with
GSH to produce S-allyl glutathione disulfide (GSSA) and H2S.
Furthermore, it was purposed that GSSA also could undergo α-
carbon nucleophilic substitution and generate H2S rapidly.6
This mechanism has been widely accepted.7 However, we have
found and reported herein that DADS is not a rapid H2S donor.
Figure 1. (A) H2S releasing dynamics of the reaction between
allylsulfides and GSH in PBS (10 mM, pH 7.4). Samples were added at
the time shown by the arrow. Initial concentrations for each
compound were DATS (50 μM), DADS (100 μM), commercial
DADS (100 μM), and GSH (500 μM). (B) HPLC traces of DATS,
commercial DADS, and purified DADS.
Received: July 9, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b01962
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
μM) failed to produce any detectable H2S. We applied an H2S
selective fluorescence probe to detect the H2S in the reactions.
The Cu(II) complex functionalized fluorescent probe (BCu)
reacts with H2S and leads to large fluorescence enhancement by
20-fold.8 Yet, when DADS purchased from a commercial source
(100 μM) was added to GSH (500 μM), H2S production was
observed. We suspected that the commercial DADS was a rapid
H2S donor because of its DATS impurity. Indeed, HPLC
analysis of the sample revealed a significant amount of DATS
(around 10%) in the commercial DADS (Figure 1B).
Scheme 3
Mechanistic studies revealed that DATS and DADS reacted
with GSH mainly through a thiol−disulfide exchange reaction
instead of α-carbon nucleophilic substitution. The reaction
between DADS (Figure 2A) or DATS (Figure 2B) with 20
nucleophilic attack of GSH on allylic sulfur generating GSSA
and ASSH, with the latter possibly releasing H2S upon
reduction by GSH; pathway 2 is the nucleophilic attack of
GSH on the central sulfur atom of DATS leading to ASH and
GS3A. These together with formed AS2H and reactive thiols
such as GSH in the system may undergo further metathesis and
redox reaction to form polysulfides GSnA, and ASnA as
observed in Figure 2D. However, when an excessive amount
of GSH was available, these compounds could eventually be
reduced by GSH to give H2S.
We found that an α-carbon substitution reaction did take
place slowly between DADS and GSH. The first evidence for
this was the existence of perthiol intermediates (GSSH and
ASSH) in the reaction mixture between DADS and GSH.
Highly reactive perthiols are an emerging class of bioactive
compounds,10 as they have other reaction patterns such as
nucleophilic addition to alkyne while releasing an elemental
sulfide atom.11 A structurally characterized tritylhydrodisulfide
(TrtSSH) releases sulfide upon reduction or disproportionates
upon deprotonation to donate elemental sulfur.12
Figure 2. HPLC chromatograms of the products of the reactions of
(A) DADS (1 mM) + GSH (20 mM), (B) DATS (1 mM) + GSH (20
mM), (C) DADS (1 mM) + GSH (1 mM), and (D) DATS (1 mM) +
GSH (1 mM) at 210 nm. A = allyl.
equiv (20 mM) of GSH for 20 min gave nearly identical
products. The major products were oxidized glutathione
(GSSG), GSSA (Figure S2), allyl mercaptan (ASH), and
DADS, which could be formed due to oxidation of ASH.
However, the expected product GSA from α-carbon sub-
stitution was not observed. We also evaluated their reaction
products of equal molar ratio (Figure 2C, D). For DADS, the
products were the same as those from reaction at a 1:20 molar
ratio. In contrast, the products from DATS were remarkably
different. While ASH was not observed, we have detected
GSSG, GSSA, and a number of new peaks including S-allyl
glutathione trisulfide (GS3A, retention time at 35 min), and
series of allyl polysulfides A-(S)n-A (n = 2−6). Corresponding
G-(S)n-A (n = 3−6) were detected by LC-MS (Figures S3−S6).
The products observed above were consistent with the thiol−
disulfide exchange reaction.9
Our results indicate that, in the presence of GSH, DADS can
react rapidly with GSH via thiol−disulfide exchange to generate
ASH and GSSA. The latter may undergo a similar thiol−
disulfide exchange with another GSH, producing ASH and
GSSG (Scheme 2). However, these reactions do not give rise to
H2S. For DATS (Scheme 3), there are two possible thiol−
disulfide exchange reaction pathways. Pathway 1 is the
The intermediates of the reaction between DADS/DATS
and GSH were trapped by monobromobimane (mBBr).13
To preserve reactive perthiol species, excessive amounts of
DATS/DADS (6 mM) were reacted with GSH (1 mM) in PBS
(pH 7.4) at 37 °C for 15 min under anaerobic conditions, and
then monobromobimane was added. The resulting mixture was
subjected to LC-MS (ESI/APCI) analysis. In the case of DATS
(Figure 3A), GS-bimane (peak 1, Figure S12) and GSS-bimane
(peak 4, Figure S14) were detected, indicating GSH and GSSH
were present. In addition, GSSA and GS3A were detected
(Figures S13, S15). For the intermediates containing an allyl
group, we found AS-bimane (peak 11), ASS-bimane (peak 12),
AS3-bimane (peak 13), and AS4-bimane (peak 14) (Figures
S18−S21), which were from ASnH (n = 1−4). When purified
DADS (free of DATS) was used in the reaction (Figure 3B)
there was no GS3A, AS3-bimane, or AS4-bimane detected. A
compound (peak 15) appeared at a similar retention time as
ASS-bimane (peak 12), but its mass spectrum did not match
with ASS-bimane (Figure S25). The GSS-bimane (Figure S23)
and ASS-bimane (Figure S26) could be detected by LC-MS at
very low intensities. The existence of these two compounds
Scheme 2
B
DOI: 10.1021/acs.orglett.5b01962
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
gave rise to the same product. The residual oxygen in the
system may account for such a reaction.
Taken together, our results suggest besides fast thiol−
disulfide exchange reaction between GSH and DADS, α-carbon
nucleophilic substitution also occurs but slowly. When the
reaction time was short, reversible thiol−disulfide exchange was
dominant. As the reaction time prolongs, these products would
be irreversibly converted into α-carbon nucleophilic substitu-
tion products. In addition, the production of DAS in the
reaction mixture suggested that ASH also participated in an α-
carbon nucleophilic substitution with either DADS or GSSA
and generated H2S (Scheme 5).
Figure 3. LC-MS-ESI (cationic mode) spectrum of (A) DATS (3
mM) + GSH (0.5 mM) + mBBr (2.5 mM), (B) DADS (3 mM) +
GSH (0.5 mM) + mBBr (2.5 mM). Peaks were identified as (1) GS-
bimane, (2) impurity, (3) GS2A, (4) GSS-bimane, (5) GSSSA, (6) i.p.,
(7) i.p., (8) monochlorobimane, (9) monobromobimane (mBBr),
(10) unknown, (11) AS-bimane, (12) AS2-bimane, (13) AS3-bimane,
(14) AS4-bimane, and (15) unknown.
Scheme 5
Because of the reactive nature of H2S, the direct
quantification of H2S generated from DADS was problematic
under aeriated media (e.g., cell culture media) using fluorescent
probes. H2S generation from DADS via ASSH would result in
accumulation of either GSA or DAS, which was shown to be
inert toward GSH (Figure S27); therefore, the amount of H2S
produced could be estimated from the total amount of GSA
and DAS.
When DADS (1 mM) and 10 equiv of GSH were mixed in
PBS (10 mM, pH 7.4) and incubated at 37 °C, the sum of GSA
and DAS generated was around 0.08 mM after 1.5 h and
increased to 0.52 mM after 12 h (Figure 5A). The formation
confirmed that GSSH and ASSH were generated during the
reaction between DADS and GSH.
Slow accumulation of GSA in the reaction between DADS
and GSH further supports the sluggish α-carbon substitution
reaction. When DADS (1 mM) and GSH (10 mM) were mixed
for 1.5 h in PBS at 37 °C, a small GSA peak (Figure S8) was
detected at 23 min, along with a diallyl sulfide (DAS) peak at
56 min (Figure 4). The GSA and DAS peaks grew slowly over
Figure 5. (A) Accumulation of GSA and DAS in the DADS (1 mM) +
GSH (10 mM) reaction in PBS (10 mM, pH 7.4) at 37 °C. (B)
Fluorescence response of H2S probe (BCu, 20 μM) toward the
reaction mixture of DADS (100 μM) + GSH (1 mM) over time (λex
=
620 nm).
rate of H2S was estimated (from the slope of the linear fit of the
blue line in Figure 5A). Since the concentration of GSH is in
large excess, the α-substitution reaction rate is roughly pseudo-
first-order to GSSA and DADS (the sum of the two
concentrations should be equal to that of the DADS in the
beginning of the reaction). Therefore, the pseudo-first-order
rate constant is estimated to be (1.24 0.04) × 10−5 s−1. The
slow formation rate of H2S during this process was confirmed
by the fluorescence assay. When the H2S probe (BCu, 20 μM)
was added to a reaction mixture of DADS (100 μM) and GSH
(1 mM) that had been mixed under argon (to prevent H2S
from oxidation) for 20 min in PBS at 37 °C, no obvious
increase in fluorescence was observed, indicating little H2S
formed. If the reaction was allowed to continue for 6 h before
the addition of the H2S probe, a 20-fold gain in fluorescence
Figure 4. HPLC chromatograms (210 nm) of the products of the
reactions of DADS (1 mM) + GSH (10 mM) in PBS (pH 7.4) at 37
°C at different time points.
time with the decreasing concentration of GSH, GSSA, ASH,
and DADS. Since thiols are fairly sensitive to oxidation, the
reaction in an anaerobic environment was carried out for 90 h
to minimize oxidation during the long period of incubation
(Figure 4). Under such conditions, the GSSA, ASH, and DADS
peaks were barely detectable, while GSSG, GSA, and DAS
peaks were the major ones. It is worth noticing that after 12 h, a
small peak at the 20 min retention time was observed, which is
likely to be a product from the oxidation of GSSG by H2O2. In
accordance, incubating GSSG with H2O2 (Figures S9, S10)
C
DOI: 10.1021/acs.orglett.5b01962
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Notes
intensity was observed (Figure 5B). Our result is in contrast to
that of Benavides and co-workers, in which around 30−40 μM
of H2S was generated from the reaction between 100 μM of
DADS and 2 mM GSH in less than 10 min.6
H2S imaging studies were conducted to evaluate the H2S
releasing activity of DADS/DATS in cell lines. The probe
(BCu) treated breast cancer MCF-7 cells were washed with
PBS thrice and then treated with 100 μM DADS, DATS, or
Na2S for another hour.8 As shown in Figure 6A, control or 100
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank the Agency for Science, Technology and
Research (A*Star) of Singapore for financial support (Grant
Number: 112 177 0036), Singapore Ministry of Education
(Grant Number: MOE2014-T2-1-134), and the support of a
Jiangsu Province Grant to NUSRI for Food Science and
Technology (platform 2).
REFERENCES
■
(1) (a) Wang, R. Antioxid. Redox Signaling 2010, 12, 1061.
(b) Hosoki, R.; Matsuki, N.; Kimura, H. Biochem. Biophys. Res.
Commun. 1997, 237, 527. (c) Abe, K.; Kimura, H. J. Neurosci. 1996, 16,
1066. (d) Elrod, J. W.; Calvert, J. W.; Morrison, J.; Doeller, J. E.;
Kraus, D. W.; Tao, L.; Jiao, X.; Scalia, R.; Kiss, L.; Szabo, C. Proc. Natl.
Acad. Sci. U. S. A. 2007, 104, 15560. (e) Yang, G.; Wu, L.; Jiang, B.;
Yang, W.; Qi, J.; Cao, K.; Meng, Q.; Mustafa, A. K.; Mu, W.; Zhang, S.
Science 2008, 322, 587.
(2) (a) Singh, S.; Padovani, D.; Leslie, R. A.; Chiku, T.; Banerjee, R. J.
J. Biol. Chem. 2009, 284, 22457. (b) Stipanuk, M. H.; Beck, P. W.
Biochem. J. 1982, 206, 267. (c) Shibuya, N.; Tanaka, M.; Yoshida, M.;
Ogasawara, Y.; Togawa, T.; Ishii, K.; Kimura, H. Antioxid. Redox
Signaling 2009, 11, 703.
(3) (a) Song, Z. J.; Ng, M. Y.; Lee, Z.-W.; Dai, W.; Hagen, T.; Moore,
P. K.; Huang, D.; Deng, L.-W.; Tan, C.-H. MedChemComm 2014, 5,
557. (b) Zhao, Y.; Biggs, T. D.; Xian, M. Chem. Commun. 2014, 50,
11788.
Figure 6. Fluorescence images of H2S generated from DATS in cell
line. MCF-7 cells were incubated with 20 μM H2S probe BCu for 2 h
and then washed and subjected to different treatments for another 1 h.
(A) Confocal images of the cells, (λex = 633 nm, λem > 650 nm). (B)
Average fluorescence intensity per cell in different groups; data were
represented as mean SEM, n = 3.
(4) Iciek, M.; Kwiecien, I.; Wlodek, L. Environ. Mol. Mutagen. 2009,
50, 247.
μM DADS treated cells gave off weak fluorescence. Yet, cells
treated with 100 μM DATS or Na2S produce much stronger
fluorescence. The fluorescence intensity per cell showed that
DATS treatment significantly elevated the fluorescence
intensity. In contrast, 100 μM DADS nearly had no effect
(Figure 6B). These results demonstrated that DATS was a
good H2S donor in cell lines, while DADS was not.
In conclusion, we found that DATS releases H2S instantly
through thiol−disulfide exchange with GSH in both chemical
and biological systems. DADS releases H2S sluggishly through
α-carbon nucleophilic substitution and does not lead to
significant H2S formation in cell lines. The previous
misunderstanding of DADS as a H2S donor is mainly attributed
to its DATS contamination in commercial samples. To avoid
misleading results, we suggest that the purity of DADS should
be carefully checked and clearly stated in all studies concerning
its health promoting effects. Slow H2S donors may be preferred
for cardiovascular health promotion, as it may help to maintain
H2S concentrations at “healthy” low levels for a longer period
of time. Rapid bursts of H2S, on the other hand, may lead to
toxic effects that could be utilized for anticancer or antibacterial
applications.
(5) Ried, K.; Frank, O. R.; Stocks, N. P. Eur. J. Clin. Nutr. 2013, 67,
64.
(6) Benavides, G. A.; Squadrito, G. L.; Mills, R. W.; Patel, H. D.;
Isbell, T. S.; Patel, R. P.; Darley-Usmar, V. M.; Doeller, J. E.; Kraus, D.
W. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 17977.
(7) (a) Hughes, M. N.; Centelles, M. N.; Moore, K. P. Free Radical
Biol. Med. 2009, 47, 1346. (b) Jacob, C.; Anwar, A.; Burkholz, T.
Planta Med. 2008, 74, 1580. (c) Kashfi, K.; Olson, K. R. Biochem.
Pharmacol. 2013, 85, 689. (d) Li, L.; Hsu, A.; Moore, P. K. Pharmacol.
Ther. 2009, 123, 386. (e) Nicholson, C. K.; Calvert, J. W. Pharmacol.
Res. 2010, 62, 289. (f) Olson, K. R. Am. J. Physiol. Regul. Integr. Comp.
Physiol. 2011, 301, R297. (g) Predmore, B. L.; Lefer, D. J.; Gojon, G.
Antioxid. Redox Signaling 2012, 17, 119.
(8) Wu, H.; Krishnakumar, S.; Yu, J.; Liang, D.; Qi, H.; Lee, Z. W.;
Deng, L. W.; Huang, D. Chem. - Asian J. 2014, 9, 3604.
(9) (a) Myers, A. G.; Cohen, S. B.; Kwon, B. M. J. Am. Chem. Soc.
1994, 116, 1255. (b) Chatterji, T.; Gates, K. S. Bioorg. Med. Chem. Lett.
2003, 13, 1349.
(10) Artaud, I.; Galardon, E. ChemBioChem 2014, 15, 2361.
(11) Galardon, E.; Padovani, D. Bioconjugate Chem. 2015, 26, 1013.
(12) Bailey, T. S.; Zakharov, L. N.; Pluth, M. D. J. Am. Chem. Soc.
2014, 136, 10573.
(13) Ida, T.; Sawa, T.; Ihara, H.; Tsuchiya, Y.; Watanabe, Y.;
Kumagai, Y.; Suematsu, M.; Motohashi, H.; Fujii, S.; Matsunaga, T.
Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 7606.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.5b01962.
Experimental details, LC-MS spectra, and supporting
HPLC chromatogram (PDF)
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: chmhdj@nus.edu.sg.
D
DOI: 10.1021/acs.orglett.5b01962
Org. Lett. XXXX, XXX, XXX−XXX