Regenerable Thiophenolic Radical-Trapping Antioxidants

Article abstract of DOI:10.1021/acs.orglett.5b03169

Diphenyl disulfides carrying alkyltelluro groups in the o-, m-, and p-positions were prepared using ortho-lithiation and lithium halogen exchange reactions. The novel antioxidants showed only minimal inhibitory effect on the azo-initiated peroxidation of linoleic acid in chlorobenzene until reduced to the corresponding thiophenols by tris(2-carboxyethyl)phosphine (TCEP). The best in situ generated thiophenol (from 7c) under these conditions quenched peroxyl radicals more efficiently than α-tocopherol with an almost 3-fold increase in inhibition time.

Full text of DOI:10.1021/acs.orglett.5b03169

Letter
pubs.acs.org/OrgLett
Regenerable Thiophenolic Radical-Trapping Antioxidants
Jiajie Yan,† Jia-fei Poon,† Vijay P. Singh,† Paul Gates,‡ and Lars Engman*,†
†Uppsala University, Department of Chemistry − BMC, Box 576, SE-751 23 Uppsala, Sweden
‡University of Bristol, School of Chemistry, Bristol, BS8 1TS, United Kingdom
S
* Supporting Information
ABSTRACT: Diphenyl disulfides carrying alkyltelluro groups in the
o-, m-, and p-positions were prepared using ortho-lithiation and lithium
halogen exchange reactions. The novel antioxidants showed only
minimal inhibitory effect on the azo-initiated peroxidation of linoleic
acid in chlorobenzene until reduced to the corresponding thiophenols
by tris(2-carboxyethyl)phosphine (TCEP). The best in situ generated
thiophenol (from 7c) under these conditions quenched peroxyl
radicals more efficiently than α-tocopherol with an almost 3-fold
increase in inhibition time.
s compared with phenols, their sulfur counterparts, the
Athiophenols, would seem attractive as radical trapping
antioxidants. The S−H bond dissociation enthalpy (BDE) in
benzenethiol (79.1 kcal/mol1a) is considerably lower than the
O−H BDE in hydroxybenzene (89.9 kcal/mol1b). According to
both theory2 and experimentation,3 thiophenol is also more
reactive than phenol toward peroxyl radicals.
Some arenethiol- and heteroarenethiol derivatives have been
shown to function as antioxidants in biological systems. Due to
its anti-inflammatory properties, salicylideneamino-2-thiophenol
(1) has been used as a cure for pain, fever, and rheumatism.4,5
It is also known to stimulate bone formation.6
most stabilizer applications, but there are also chemical pro-
blems associated with the use of arenethiols as radical trapping
antioxidants. Some 50 years ago,15 Ingold and co-workers
compared the capacity of 2,4,6-tri-tert-butylphenol (5a), -aniline
(5b), and -thiophenol (5c) to inhibit autoxidation of cumene. All
compounds 5 could quench peroxyl radicals by formal H atom
transfer. However, whereas the resulting phenoxyl and phenyl-
aminyl radicals can quench a second peroxyl radical, the
phenylthiyl radicals disappear by rapid recombination (k = 5 ×
107 M−1 s−116) to form a diaryl disulfide. The stoichiometric
factor n, the number of peroxyl radicals that could be destroyed
per antioxidant molecule, was therefore only 0.95 for 5c.
We have tried for some time to find novel antioxidants that
could outperform the traditional phenols and aromatic amines
when it comes to radical trapping activity and regenerabil-
ity.17,18 In this work, we found that alkyltelluro groups have a
remarkable effect. For example, introducing an octyltelluro
group next to the OH in phenol caused a ca. 4 orders of
magnitude increase in the reactivity toward peroxyl radicals.
Furthermore, this modification of the structure also improved
the regenerability of the antioxidant in the lipid phase by
aqueous phase co-antioxidants such as N-acetylcysteine (NAC).
We were curious to see how alkyltelluro substitution in
thiophenol would affect its radical trapping activity. For
comparison, some of the corresponding alkylseleno compounds
were also prepared. Obviously, we could expect that these
compounds would be readily transformed into inactive
“dormant” diaryl disulfide antioxidants, but we were challenged
by the perspective to find a suitable co-antioxidant that, “on
demand”, could make them come alive again. We herein describe
the synthesis of the target antioxidants as well as their radical
trapping activity and regenerability in a simple two-phase
peroxidation model designed to mimic a biological membrane.
The heteroaromatic imidazol-4-thiol family of natural
products known as the ovothiols (represented by 2) and less
polar derivatives thereof7−9 have been found to act as biological
antioxidants. Compounds of this sort could scavenge lipid
peroxyl radicals almost as efficiently as α-tocopherol (α-T) and
vitamin C.10 The natural product ergothioneine (3), which is a
derivative of imidazol-2-thiol, has also demonstrated numerous
antioxidative and cytoprotective capabilities.11,12 6-Mercapto-α-
tocopherol (4), a sulfur analog of α-T, has been described in
the early literature,13,14 but little is known about its antioxidative
properties.
Whereas phenolic compounds are the most common radical
trapping stabilizers for natural and man-made organic materials,
aromatic thiols are rarely used. The unpleasant odor of many
compounds lacking polar substituents would of course preclude
Received: November 2, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b03169
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
For the preparation of ortho-functionalized compounds,
we decided to take advantage of the directing effect of an
arenethiolate ion in lithiation.19 Thiophenol was therefore
treated with 2.2 equiv of n-butyllithium in the presence of
tetramethylethylenediamine (TMEDA). Addition of dibutyl-,
dioctyl-, dihexadecyl ditelluride and dioctyl diselenide,
respectively, to the in situ prepared solution of 2,S-dilithiated
thiophenol afforded diphenyl disulfides 6a−d in modest yields
(19−43%) after acidic workup and chromatographic puri-
fication (Scheme 1). Obviously, the ortho-functionalized
Scheme 3. Synthesis of Compounds 11 and 13
Scheme 1. Synthesis of Compounds 6a−d
The conjugated diene formed as a result of this process was
monitored by HPLC with UV detection at 234 nm. The upper
aqueous layer contained a water-soluble co-antioxidant that
could hopefully regenerate the active antioxidant in the organic
phase. By following the increase in conjugated diene with time,
information about the radical quenching activity and regener-
ability of the antioxidant could be obtained. Usually, two stages
could be distinguished in the peroxidation traces (Figure 1).
thiophenols initially formed are oxidized to the corresponding
disulfides during workup.
For the functionalization of thiophenol in the para- and
meta-positions, we relied on lithium halogen exchange.17c
4-Bromothiophenol upon treatment with 3 equiv of tert-
butyllithium and then dibutyl-, dioctyl-, and dihexadecyl
ditelluride, respectively, returned disulfides 7a−c, again in low
(15−32%) isolated yields (Scheme 2).
Scheme 2. Synthesis of Compounds 7a−d and 9a−b
Figure 1. Peroxidation traces (conjugated diene concentration vs
time) recorded with compounds 7c, Ph2S2, and α-T (40 μM) as
antioxidants in the chlorobenzene layer in the presence of TCEP
(0.5 mM) in the aqueous phase.
During the first inhibited phase, most of the peroxyl radicals
were quenched by the antioxidant and the concentration of the
conjugated diene increased only slowly with time (Rinh). Then,
in a second phase, the diene concentration with time increased
suddenly to a higher constant value, Runinh, which was very
similar to the one recorded in the absence of any antioxidant in
the chlorobenzene layer. The cross-point of the inhibited and
uninhibited lines was calculated and reported as the inhibition
time, Tinh, of the antioxidant.
Surprisingly, when we used a similar protocol for the
preparation of the corresponding octylseleno-derivative 7d,
4-(octylseleno)thiophenol 8 was isolated in 39% yield.
Oxidation with potassium ferricyanide under basic conditions
afforded bis-4-(octylseleno)phenyl disulfide in quantitative
yield. Compounds 9, functionalized with octyltelluro (9a,
35% yield) and octylseleno groups (9b, 59% overall yield) in
the meta-position, were also prepared by lithium halogen
exchange using 3-bromothiophenol as a starting material.
For reference purposes, to see the effect of the aromatic
sulfhydryl group, we also prepared two S-isopropylated
octyltelluro-functionalized thiophenols 11 and 13. S-Isopropy-
lated thiophenol (10) was ortho-lithiated, and elemental
tellurium allowed insertion into the carbon lithium bond.
After air oxidation, the resulting crude ditelluride was reduced
with sodium borohydride and the arenetellurolate formed
alkylated with octyl bromide (Scheme 3).
S-Isopropylated 4-bromothiophenol (12) was subjected to
lithium halogen exchange and then converted to 13 using a
similar protocol as described for the preparation of compound 11.
The radical trapping capacity and regenerability of our novel
thiophenol-derived antioxidants were studied in a two-phase
model.20 In the lower chlorobenzene layer, containing
the antioxidant to be evaluated, peroxidation of linoleic acid
was ongoing, initiated by azo-bis-dimethylvaleronitrile.
Compounds 6, 7, 9, 11, and 13 were first evaluated in the
absence of any co-antioxidant in the aqueous phase. Diphenyl
disulfide (Ph2S2) and α-T were also included for reference
purposes (Table 1). Except for α-T (Rinh = 28 μM/h and Tinh
=
109 min), none of the compounds tested could inhibit
peroxidation. Rather, they acted as poor retarders of
peroxidation and no inhibition time could be recorded.
Previously, we have found that the water-soluble thiol NAC
could significantly prolong the Tinh for many chalcogen-
containing antioxidants.17 The thiol probably serves to reduce
the chalcogen from the tetravalent to the divalent state. It was
our hope that NAC, via thiol exchange with diaryl disulfides
(eq 1), would also serve to increase the concentration of
aromatic thiol in the lipid phase. All compounds were therefore
B
DOI: 10.1021/acs.orglett.5b03169
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Table 1. Inhibited Rates of Conjugated Diene Formation (Rinh) and Inhibition Times (Tinh) with TCEP (0.5 mM), NAC (1.0
mM), and without Co-antioxidants
with TCEP
Rinh (μM/h)
with NAC
without co-antioxidant
a
b
c
b
c
b
c
AO
Tinh (min)
Rinh (μM/h)
Tinh (min)
Rinh (μM/h)
Tinh (min)
6a
6b
6c
6d
7a
7b
7c
7d
9a
9b
9.7
8.2
7.3
307
3.9
3.7
2.3
304
6.9
338
81
1
2
1
150
148
156
0
5
0
5
43
3
3
2
66
73
84
0
3
6
1
437
333
318
371
358
321
342
385
367
436
390
388
421
0
0
0
0
0
0
0
0
0
0
0
0
0
32
19
300
19
0
1
1
284
304
324
0
7
4
5
3
1
1
132
146
170
0
3
3
4
20
15
343
20
1
164
0
2
1
88
0
7
4
371
72
d
11
2
37
2
3
2
1
78
101
0
d
13
118
431
21
5
32
23
3
Ph2S2
0
372
25
α-T
2
123
7
1
97
5
28
2
109
2
a
b
AO = Antioxidant. 40 μM was used in the assay. Rate of peroxidation during the inhibited phase. Uninhibited rate is 453 μM/h with TCEP,
c
425 μM/h with NAC, and 479 μM/h without antioxidant and co-antioxidant. Errors correspond to
peroxidation. Reactions were monitored for 460 min. Errors correspond to SD for triplicates. 80 μM of antioxidant was used in the assay.
SD for triplicates. Inhibited phase of
d
tested again in the presence of NAC (1 mM). Considerable
improvement was seen. Some of the catalysts (6c, 7a−c, 9a)
could now match α-T when it comes to the inhibited rate of
peroxidation and inhibition time. As noted before, NAC did not
show any inhibiting effect on peroxidation when tested alone in
a control experiment and it did not serve to extend the Tinh
for α-T. In a similar vein, all catalysts were also tested in the
presence of 0.5 mM of the water-soluble dithiol dithiothreitol
(DTT), acidified with acetic acid to the same pH (3.3; Rinh and
Tinh showed some pH-dependence).
However, the Rinh and Tinh data recorded were almost a
blueprint of those obtained with NAC (Table S1 in the
Supporting Information). It may be that the thiol exchange of
NAC and DTT with the diaryl disulfides is not favorable
enough that the compounds could exert their full antioxidant
capacity. Therefore, a better water-soluble disulfide reducing
agent was sought. It occurred to us that tris(2-carboxyethyl)-
phosphine (TCEP, 14) could be a useful co-antioxidant in our
system. It is known to reduce disulfides as shown in eq 2.21
disulfide 7c, inhibited peroxidation for 324 min with Rinh as low
as 2.3 μM/h (see Figure 1). In contrast, none of the analogous
alkylseleno substituted diphenyl disulfides 6d, 7d, and 9b or
diphenyl disulfide could act as radical trapping agents in the
two-phase model. Apparently, tellurium is a vital constituent in
these antioxidants.
We postulate below that alkyltelluro thiophenols are the
active antioxidants generated in situ under the conditions of
the two-phase model. It was therefore of interest to study the
performance of some S-alkylated thiophenols. In the presence
of NAC, S-isopropylated compound 13 (at 80 μM) showed
the same antioxidant characteristics as α-T (at 40 μM). The
performance of the corresponding o-disubstituted compound
11 was not so impressive. Although an inhibited phase of
peroxidation could be distinguished for both 11 and 13, the
Rinh and Tinh values were not near those recorded for the
corresponding disulfides 6b and 7b (at 40 μM). It may be that
these compounds quench peroxyl radicals by donating
electrons from tellurium. In the absence of an aqueous-phase
reducing agent they are likely to be oxidized to the corre-
sponding inactive telluroxides by residual amounts of linoleic
acid hydroperoxide always present in linoleic acid. Our pro-
posed catalytic antioxidant mechanism is shown in Scheme 4.
Scheme 4. Proposed Catalytic Mechanism for Peroxyl
Radical Trapping by Bis-4-(alkyltelluro)phenyl Disulfides
in the Presence of TCEP
Interestingly, when TCEP hydrochloride (0.5 mM; pH = 3.27)
was added to the aqueous phase, Rinh was notably reduced and
Tinh was prolonged for most of the antioxidants tested. This
would indicate efficient reduction of diaryl disulfide and facile
regeneration of the antioxidant under the conditions of the
assay. Whenever the alkyl part of the alkyltelluro moiety was
varied in a systematic way (disulfides 6 and 7) the performance
increased slightly with increasing chain length/lipophilicity
(butyl < octyl < hexadecyl). Overall, the p-alkyltelluro
functionalized disulfides 7a−c showed lower Rinh and longer
Tinh than their o-substituted counterparts 6. The antioxidant
activity of the only m-alkyltelluro functionalized derivative 9a
prepared was in between those of the corresponding o- and
p-derivatives. The best antioxidant, p-disubstituted diphenyl
C
DOI: 10.1021/acs.orglett.5b03169
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
TCEP-induced reduction of disulfide will produce alkyltelluro
thiophenol which quenches peroxyl radicals by transfer of oxygen
to tellurium.18 Then, in a solvent cage, the resulting alkoxyl
radical will abstract a hydrogen atom from the thiophenol.
Recombination of thiyl radicals will follow, accompanied by
reduction of tetravalent tellurium (shown as a telluroxide) to
reform the disulfide antioxidant. In fact, it is known that
trivalent phosphorus compounds could act as reducing agents
toward telluroxides.22 No attempt was made to isolate
intermediates proposed in Scheme 4.
REFERENCES
■
(1) S−H and O−H BDEs given were estimated as described by
Bordwell and are based on a thermodynamic cycle which rely on pKa-
values for phenol/thiophenol and oxidation potentials of the
corresponding conjugate bases. These BDE-values are in excellent
agreement with experiment. (a) Bordwell, F. G.; Zhang, X.; Satish, A.
V.; Cheng, J.-P. J. Am. Chem. Soc. 1994, 116, 6605−6610. (b) Bordwell,
F. G.; Cheng, J. J. Am. Chem. Soc. 1991, 113, 1736−1743.
(2) Denisov, E.; Chatgilialoglu, C.; Shestakov, A.; Denisova, T. Int. J.
Chem. Kinet. 2009, 41, 284−293.
(3) Chenier, J. H. B.; Furimsky, E.; Howard, J. A. Can. J. Chem. 1974,
52, 3682−3688.
We were also curious to see the performance of our catalysts
in the presence of varying (0.0625, 0.125, 0.25, 0.5, and
1.0 mM) amounts of TCEP (Table S2 in the Supporting
Information). As shown in Figure 2, Tinh increased linearly with
(4) Chung, J.; Lee, H. S.; Chung, H. Y.; Yoon, T. R.; Kim, H. K.
Biotechnol. Lett. 2008, 30, 1553−1558.
(5) Park, S.; Sung, B.; Jang, E. J.; Kim, D. H.; Park, C. H.; Choi, Y. J.;
Ha, Y. M.; Kim, M. K.; Kim, N. D.; Yu, B. P.; Chung, H. Y. Arch.
Pharmacal Res. 2013, 36, 880−889.
(6) Kim, H. K.; Park, K. S.; Lee, J. S.; Kim, J. H.; Park, D. S.; Shin, J.;
Yoon, T. R. J. Cell. Biochem. 2012, 113, 1833−1841.
(7) Holler, T. P.; Hopkins, P. B. J. Am. Chem. Soc. 1988, 110, 4837−
4838.
(8) Zoete, V.; Bailly, F.; Catteau, J.; Bernier, J. J. Chem. Soc., Perkin
Trans. 1 1997, 2983−2988.
(9) Zoete, V.; Vezin, H.; Bailly, F.; Vergoten, G.; Catteau, J.; Bernier,
J. Free Radical Res. 2000, 32, 525−533.
(10) Marjanovic, B.; Simic, M. G.; Jovanovic, S. V. Free Radical Biol.
Med. 1995, 18, 679−685.
(11) Cheah, I. K.; Halliwell, B. Biochim. Biophys. Acta, Mol. Basis Dis.
2012, 1822, 784−793.
Figure 2. Inhibition time of compound 7c versus concentration of TCEP.
(12) Servillo, L.; Castaldo, D.; Casale, R.; D’Onofrio, N.; Giovane, A.;
Cautela, D.; Balestrieri, M. L. Free Radical Biol. Med. 2015, 79, 228−
236.
the concentration of the co-antioxidant. On the other hand, the
corresponding Rinh values decreased to reach a constant value
above 0.5 mM TCEP. Obviously, at the lower concentrations of
TCEP, the thiol concentration in the chlorobenzene phase is so
low that the peroxyl radicals cannot be efficiently quenched.
In conclusion, we have shown that (alkyltelluro)thiophenols
generated in situ in a lipid phase can quench peroxyl radicals
more efficiently than α-T. The tris(2-carboxyethyl)phosphine
used to bring about disulfide to thiol reduction has an
additional role: it reduces the telluroxide form of the catalyst to
the corresponding telluride and thus allows for antioxidant
regeneration and a catalytic mode of action. What makes these
antioxidants unique is the possibility to keep them turned off
under nonreducing conditions and, whenever needed, make
them come alive. It remains to be seen if this capability can be
exploited in biological systems.
(13) Hromatka, O.; Kirnig, I. Monatsh. Chem. 1954, 85, 235−240.
(14) Csallany, A.; Saari, A.; Draper, H. H. Int. J. Vitam. Nutr. Res.
1971, 41, 368−375.
(15) Gardner, D. V.; Howard, J. A.; Ingold, K. U. Can. J. Chem. 1964,
42, 2847−2851.
(16) Scott, T. W.; Liu, S. N. J. Phys. Chem. 1989, 93, 1393−1396.
(17) (a) Kumar, S.; Johansson, H.; Kanda, T.; Engman, L.; Muller,
̈
T.; Jonsson, M.; Pedulli, G. F.; Petrucci, S.; Valgimigli, L. Org. Lett.
2008, 10, 4895−4898. (b) Kumar, S.; Johansson, H.; Kanda, T.;
Engman, L.; Muller, T.; Bergenudd, H.; Jonsson, M.; Pedulli, G. F.;
̈
Amorati, R.; Valgimigli, L. J. Org. Chem. 2010, 75, 716−725.
(c) Johansson, H.; Shanks, D.; Engman, L.; Amorati, R.; Pedulli, G.
F.; Valgimigli, L. J. Org. Chem. 2010, 75, 7535−7541. (d) Singh, V. P.;
Poon, J.; Engman, L. J. Org. Chem. 2013, 78, 1478−1487. (e) Poon, J.;
Singh, V. P.; Engman, L. J. Org. Chem. 2013, 78, 6008−6015.
(f) Singh, V. P.; Poon, J.; Engman, L. Org. Lett. 2013, 15, 6274−6277.
(g) Poon, J.; Singh, V. P.; Yan, J.; Engman, L. Chem. - Eur. J. 2015, 21,
2447−2457.
ASSOCIATED CONTENT
* Supporting Information
■
S
́
(18) Amorati, R.; Valgimigli, L.; Diner, P.; Bakhtiari, K.; Saeedi, M.;
Engman, L. Chem. - Eur. J. 2013, 19, 7510−7522.
(19) Smith, K.; Lindsay, C. M.; Pritchard, G. J. J. Am. Chem. Soc.
1989, 111, 665−669.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.5b03169.
Experimental section including spectral data for all
compounds prepared (PDF)
(20) Shanks, D.; Amorati, R.; Fumo, M. G.; Pedulli, G. F.; Valgimigli,
L.; Engman, L. J. Org. Chem. 2006, 71, 1033−1038.
(21) (a) Levison, M. E.; Josephson, A. S.; Kirschenbaum, D. M.
Experientia 1969, 25, 126−127. (b) Burns, J. A.; Butler, J. C.; Moran,
J.; Whitesides, G. M. J. Org. Chem. 1991, 56, 2648−2650.
(22) Oba, M.; Okada, Y.; Nishiyama, K.; Ando, W. Org. Lett. 2009,
11, 1879−1881.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: lars.engman@kemi.uu.se.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
The Swedish Research Council (621-2011-4006) and Carl
■
Tryggers Stiftelse for Vetenskaplig Forskning (CTS:120) are
̈
gratefully acknowledged for financial support.
D
DOI: 10.1021/acs.orglett.5b03169
Org. Lett. XXXX, XXX, XXX−XXX