Antioxidant generation and regeneration in lipid bilayers: The amazing case of lipophilic thiosulfinates and hydrophilic thiols

Article abstract of DOI:10.1039/c3cc44401e

We demonstrate that the garlic-derived chemopreventive agent allicin and the related anamu-derived petivericin are poor radical-trapping antioxidants in lipid bilayers, but that the in situ reaction of a lipophilic analog and a hydrophilic thiol yields an extremely potent radical-trapping antioxidant that can be recycled in the presence of excess thiol.

Full text of DOI:10.1039/c3cc44401e

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Antioxidant generation and regeneration in lipid
bilayers: the amazing case of lipophilic thiosulfinates
and hydrophilic thiols†
Cite this: Chem. Commun., 2013,
49, 8181
Received 11th June 2013,
Accepted 1st August 2013
Feng Zheng and Derek A. Pratt*
DOI: 10.1039/c3cc44401e
www.rsc.org/chemcomm
We demonstrate that the garlic-derived chemopreventive agent allicin constant for this reaction was determined to be kinh = 3 ꢂ
and the related anamu-derived petivericin are poor radical-trapping 106 Mꢀ1 sꢀ1 6b making the reactions of sulfenic acids with ROOꢁ
,
antioxidants in lipid bilayers, but that the in situ reaction of a lipophilic among the fastest known.
analog and a hydrophilic thiol yields an extremely potent radical-
trapping antioxidant that can be recycled in the presence of excess thiol.
Garlic has been used to treat a variety of ailments in Asia for
millennia, many of which have also been treated with anamu in
the South and Central Americas.1 The garlic-derived thiosulfinate
allicin (1),2 and the related anamu-derived thiosulfinate petivericin
(2),3 are purported to be potent radical-trapping antioxidants,4
reactivity to which many of the biological activities of these plants thiosulfinates are effective radical-trapping antioxidants in lipid
a
However, it remains unclear if (and how) these plant-derived
have been ascribed.1
bilayers; most relevant given the implication of lipid peroxidation
products in degenerative disease pathogenesis.7
To provide insight on this issue, the radical-trapping abilities of 1, 2
and 5 were studied in unilamellar phosphatidylcholine liposomes –
convenient models for the lipid bilayers that make up biological
membranes. The relative reactivities of the compounds were deter-
mined using H2B–PMHC (6),8 a recently developed analog of
a-tocopherol (a-TOH, the most biologically-active form of Vitamin E
and Nature’s premier lipophilic radical-trapping antioxidant),9 which
undergoes fluorescence enhancement upon reaction with ROOꢁ.8,10 In
the presence of a good radical-trapping antioxidant, the fluorescence
onset is suppressed as the radicals are trapped by the added antiox-
idant in lieu of reacting with 6. Initial experiments focused on the
ability of either 1 or 2 to inhibit the oxidation of liposome-embedded 6
by peroxyl radicals derived from the lipid- and water-soluble azo
initiators MeOAMVN11 and ABAP,12 respectively. Representative results
are shown in Fig. 1. In fact, neither 1 nor 2 were able to slow the
oxidation of 6 up to concentrations of 15 mM (100-fold higher than 6,
Fig. 1A and B). Since the reactivity of 6 is essentially the same as that of
a-TOH,8 1 and 2 must be at least 100-fold less reactive than a-TOH in
the lipid bilayer. For comparison, oxidations of liposomes supplemen-
ted with the persistent sulfenic acid 5 were also carried out, revealing
excellent radical-trapping activity (Fig. 1C). In the presence of 5, the rate
of oxidation of 6 is reduced dramatically, and a kinetic analysis of the
data (see ESI†) reveals that 5 reacts with ROOꢁ with a rate constant that
is at least 10-fold larger than that of PMHC (7, the truncated model of
a-TOH upon which 6 is based – Fig. 1D), demonstrating that sulfenic
(1)
(2)
In organic solution, we have shown that 1 and 2 are not radical-
trapping antioxidants per se, and must first undergo Cope elimination
to form the sulfenic acids 3 and 4, respectively (eqn (1) and (2)), which
react rapidly with radicals.5 Since the reactivities of 3 and 4 could not be
determined directly due to their short lifetimes in solution (they self-
condense to reform the parent thiosulfinates or undergo other reac-
tions),1 we carried out studies of the persistent 9-triptycenesulfenic acid
(5) to provide direct thermodynamic and kinetic support for this
hypothesis.6 The O–H bond strength of 5 was determined to be
72 kcal mol ,
ꢀ1 6a making the reaction of sulfenic acids with peroxyl
radicals (ROOꢁ) exothermic by B16 kcal molꢀ1, and the rate
Department of Chemistry, University of Ottawa, 10 Marie Curie Pvt., Ottawa,
Ontario, Canada. E-mail: dpratt@uottawa.ca; Fax: +1-613-562-5170;
Tel: +1-613-562-5800 ext. 2119
† Electronic supplementary information (ESI) available: All experimental proce-
dures and raw data. See DOI: 10.1039/c3cc44401e
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 8181--8183 8181

View Article Online
Communication
ChemComm
Fig. 1 Fluorescence (at 520 nm) intensity–time profiles from MeOAMVN-
mediated (0.2 mM) oxidations of egg phosphatidylcholine liposomes (1 mM)
containing 0.15 mM 6 and increasing concentrations of 1 (A), 2 (B), 5 (C) or 7 (D).
acids are extremely potent peroxyl radical trapping agents in both
homogeneous organic solution5,6b and lipid bilayers.
The time required to reach maximum fluorescence in the first
phase of the plots in Fig. 1C and D (ca. 400 counts), during which 6
and the added antioxidant compete for ROOꢁ, will hereafter be
Fig. 2 Fluorescence (at 520 nm) intensity–time profiles from MeOAMVN-
referred to as the ‘inhibited period’; its length reflects the stoichio- mediated (0.2 mM, left) and ABAP-mediated (2.7 mM, right) oxidations of egg
metry of the reaction of the antioxidant and ROOꢁ. The inhibited
phosphatidylcholine liposomes (1 mM) containing 0.15 mM 6 and 8 and/or an
equivalent of NAC.
periods in Fig. 1C and D correlate nicely with the concentration
of either 5 or 7 (see ESI†) yielding slopes of 14 min mMꢀ1 and 31 min
mMꢀ1, respectively, when radicals are generated using MeOAMVN
potential for this mechanism, NAC was added to oxidations of
liposome embedded 6 supplemented with 1, 2 and 8. Representative
results are shown in Fig. 2.
and 10 min mMꢀ1 and 21 min mMꢀ1, respectively, using AAPH. Since
7 traps two ROOꢁ under these conditions,9 5 must trap only one.13
We surmised that the poor radical-trapping activity of 1 and 2
obvious from Fig. 1A and B might be explained by their partitioning
to the aqueous phase, where Cope elimination is slowed dramatically
by H-bonding1,5,14 or H-atom transfer from the sulfenic acid would be
slowed by sequestration of the labile H-atom by H-bond for-
mation.5,15 To investigate these possibilities, a lipophilic analog of 2
was prepared (8, see ESI† for details). Unfortunately, oxidations of
liposomes supplemented with 8 yielded fluorescence profiles that
were indistinguishable from those supplemented with 1 and 2 (see
ESI†). It would seem that the concentration of sulfenic acid formed
from each of 1, 2 and 8 must be too low for it to compete with 6 for
peroxyl radicals throughout the oxidations.
(3)
(4)
Under these conditions 1 and 2 continue to have no effect on
oxidations mediated by either lipophilic or hydrophilic peroxyl
radicals (see ESI†). In contrast, pronounced inhibited periods were
observed in oxidations of liposomes supplemented with 8 (Fig. 2A
and B). Since the presence of NAC alone does little in oxidations
mediated by MeOAMVN (Fig. 2C) and has only a modest retarding
effect in oxidations mediated by ABAP (Fig. 2D), these results imply
that the reaction of NAC with 8 yields 9 and a lipophilic sulfenic acid
that partitions to the lipid phase where it can react with peroxyl
radicals. In contrast, S-thiolation of 1 or 2 leads to the mixed
disulfides 10 and 11, respectively, and hydrophilic sulfenic acids that
Sulfenic acids can also arise from bimolecular reactions of are consumed by reaction with another equivalent of NAC to form 10
thiosulfinates with nucleophiles, such as thiols.1,2 Indeed, S-thiolation and 11. Indeed, the modest activity displayed by NAC in the ABAP-
of thiosulfinates to form a sulfenic acid and a mixed disulfide mediated oxidation (Fig. 2C) disappears when allicin or petivericin
(eqn (3)) is believed to be the first of two steps accounting for the are present (see ESI†) – and implies that the rate of the S-thiolation
observed stoichiometry in the reaction of a thiosulfinate with two reaction is faster than the rate of radical generation under these
equivalents of thiol to give two equivalents of mixed disulfide conditions. Interestingly, when the mixed disulfides arising from
(eqn (4)). As such, the presence of water-soluble thiols, such as the reaction of NAC with either 2 or 8 were monitored by LC/MS
N-acetylcysteine (NAC), or the ubiquitous cellular thiol glutathione, over the course of an oxidation (see ESI†), quantitative formation of
may promote sulfenic acid formation from 1 and 2. To evaluate the 11 from 2 was observed within 10 minutes, but only a small amount
c
8182 Chem. Commun., 2013, 49, 8181--8183
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
of 9 from 8 (B10%). Amorati and Pedulli have clearly demonstrated and the thiol moiety of NAC is predicted to be highly endothermic
that allylic disulfides have no significant radical-trapping activity.16
(B15 kcal molꢀ1 in the gas phase),6a we suggest this reaction must
occur by electron transfer from the thiol (EoRSSR/RSH at pH 7)19 to the
sulfinyl radical (ERoSO =RSO ¼ 0:74 V in CH3CN)6a as in eqn (6).
ꢀ
ꢁ
(5)
(6)
Since the correspondingly small amount of sulfenic acid pro-
duced from 8 is insufficient to account for the length of the
inhibited periods in Fig. 2A and B (compare to the results for 5 in
Fig. 1C), we wondered if NAC could regenerate the sulfenic acid
from the sulfinyl radical produced upon H-atom transfer to ROOꢁ.
This reaction would effectively convert a water-soluble reducing
equivalent into a lipid-soluble one. To evaluate this hypothesis, a
series of oxidations of liposome-embedded 6 were carried out in the
presence of a constant concentration of 8 and increasing [NAC].
Representative results shown in Fig. 3 for oxidations with both
hydrophilic (Fig. 3A) and lipophilic (Fig. 3B) peroxyl radicals reveal a
clear elongation of the inhibited period with increasing [NAC]. Since
the inhibited periods are far longer than those expected solely from
the reaction of a single molecule of sulfenic acid derived from
reaction of NAC with 8, and NAC itself is not effective in inhibiting
the oxidation (cf. Fig. 2C and D), the result can only be explained by
the regeneration of the sulfenic acid derived from 8 via reaction with
NAC. This is supported by analogous experiments wherein a con-
stant amount of the persistent sulfenic acid 5 was incorporated into
the liposomes, and an increasing amount of NAC was added to the
aqueous phase (Fig. 3C and D).
The cooperativity displayed by NAC and the sulfenic acid derived
from 8 (or the persistent sulfenic acid 5) is reminiscent of that
displayed by Vitamins E and C, a phenomenon believed to be key to
the biological activities of both compounds in vivo.17,18 At the
interface of the lipid bilayer and the surrounding aqueous medium,
ascorbate can transfer an electron to the a-tocopheroxyl radical
either concerted with, or followed by, proton transfer to give a-TOH,
which can react with another peroxyl radical (eqn (5)). Since
the regeneration of RSOH from H-atom transfer between RSOꢁ
In summary, the foregoing experiments demonstrate that
allicin and petiviericin are not particularly effective as radical-
trapping antioxidants in lipid bilayers. As such, any antioxidant
activity that is observed in vivo is more likely to arise from other
mechanisms (e.g. induction of antioxidant enzymes). However,
the combination of a lipophilic thiosulfinate (such as 8) and
a hydrophilic thiol (such as NAC) is an incredibly potent
co-antioxidant system in lipid bilayers. The interaction of the
two serves to generates a highly-reactive radical-trapping anti-
oxidant and provides a mechanism by which it can be regen-
erated. Whether this amazing reactivity can be exploited by
lipophilic metabolites of allicin and/or petivericin or synthetic
derivatives thereof in vivo merits further investigation.
This work was supported by grants from NSERC, Green Centre
Canada and the Canada Research Chairs program. We thank Prof.
Gonzalo Cosa for generously providing a sample of 6.
Notes and references
1 E. Block, Garlic and Other Alliums: the Lore and the Science, The Royal
Society of Chemistry, Cambridge, UK, 2010.
2 C. J. Cavallito, J. H. Bailey and J. S. Buck, J. Am. Chem. Soc., 1945,
67, 1032.
3 R. Kubec, S. Kim and R. A. Musah, Phytochemistry, 2002, 61, 675.
4 (a) Y. Okada, K. Tanaka, E. Sato and H. Okajima, Org. Biomol. Chem.,
2006, 4, 4113; (b) Y. Okada, K. Tanaka, E. Sato and H. Okajima, Org.
Biomol. Chem., 2008, 6, 1097 and literature cited.
5 (a) V. Vaidya, K. U. Ingold and D. A. Pratt, Angew. Chem., Int. Ed.,
2009, 48, 157; (b) P. T. Lynett, K. Butts, V. Vaidya, G. E. Garrett and
D. A. Pratt, Org. Biomol. Chem., 2011, 9, 3320.
6 (a) A. J. McGrath, G. E. Garrett, L. Valgimigli and D. A. Pratt, J. Am.
Chem. Soc., 2010, 132, 16759; (b) R. Amorati, P. T. Lynett,
L. Valgimigli and D. A. Pratt, Chem.–Eur. J., 2012, 18, 6370.
7 H. Y. Yin, L. B. Xu and N. A. Porter, Chem. Rev., 2011, 111, 5944.
8 K. Krumova, S. Friedland and G. Cosa, J. Am. Chem. Soc., 2012,
134, 10102, erratum: J. Am. Chem. Soc. 2013, 135, 1625.
9 G. W. Burton and K. U. Ingold, Acc. Chem. Res., 1986, 19, 194.
10 (a) P. Oleynik, Y. Ishihara and G. Cosa, J. Am. Chem. Soc., 2007,
129, 1842; (b) K. Krumova, P. Oleynik, P. Karam and G. Cosa, J. Org.
Chem., 2009, 74, 3641.
11 N. Noguchi, H. Yamashita, N. Gotoh, Y. Yamamoto, R. Numano and
E. Niki, Free Radicals Biol. Med., 1998, 24, 259.
12 L. R. C. Barclay, S. J. Locke, J. M. MacNeil, J. VanKessel, G. W. Burton
and K. U. Ingold, J. Am. Chem. Soc., 1984, 106, 2479.
13 This is consistent with our previous work in homogenous organic
solution, where we found stoichiometric numbers r15,6b
.
14 F. Freeman and Y. Kodera, J. Agric. Food Chem., 1995, 43, 2332.
15 G. Litwinienko and K. U. Ingold, Acc. Chem. Res., 2007, 40, 222.
16 R. Amorati and G. F. Pedulli, Org. Biomol. Chem., 2008, 6, 1103.
17 (a) T. Doba, G. W. Burton and K. U. Ingold, Biochim. Biophys. Acta,
1985, 835, 298; (b) B. Frei, L. England and B. N. Ames, Proc. Natl.
Acad. Sci. U. S. A., 1989, 86, 6377.
18 While NAC cannot recycle a-TOH in lipid bilayers, NAC has been
used to recycle synthetic antioxidants in biphasic systems, e.g.
S. Kumar, H. Johansson, L. Engman, L. Valgimigli, R. Amorati,
M. Grazia Fumo and G. F. Pedulli, J. Org. Chem., 2007, 72, 2583.
19 K. K. Millis, K. H. Weaver and D. L. Rabenstein, J. Org. Chem., 1993,
58, 4144.
Fig. 3 Representative fluorescence (at 520 nm) intensity–time profiles from
MeOAMVN-mediated (0.2 mM, left) and ABAP-mediated (2.7 mM, right) oxida-
tions of egg phosphatidylcholine liposomes (1 mM) containing 0.15 mM H2B–
PMHC and either 8 (top) or 5 (bottom) with varying [NAC].
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 8181--8183 8183