A simple and efficient strategy to enhance the antioxidant activities of amino-substituted glutathione peroxidase mimics

Article abstract of DOI:10.1002/chem.200800963

The glutathione peroxidase (GPx) activities of some diaryl diselenides incorporating tertiary amino groups were studied with H₂O ₂, Cum-OOH, and tBuOOH as substrates and with PhSH as thiol co-substrate. Simple replacement of a hydrogen atom with a methoxy group dramatically enhances the GPx activity. The introduction of methoxy substituents ortho to selenium in N,N-dialkylbenzylamine-based compounds makes the basicity of the amino groups perfect for the catalysis. The presence of 6-OMe groups prevents possible Se-N interactions in the selenols, increasing their zwitterionic characters. The methoxy substituents also protect the selenium in the selenenic acid intermediates from overoxidation to seleninic acids or irreversible inactivation to selenonic acid derivatives. The additional substituents also play a crucial role in the selenenyl sulfide intermediates, by preventing thiol exchange reactions-which would normally lead to an inactivation pathway-at the selenium centers. The strengths of Se...N interactions in the selenenyl sulfide intermediates are dra-matically reduced upon introduction of the methoxy substituents, which not only reduce the thiol exchange reactions at selenium but also enhance the nucleophilic attack of the incoming thiols at sulfur. The facile attack of thiols at sulfur in the selenenyl sulfides also prevents the reactions between the selenenyl sulfides and H₂O₂ that can regenerate the selenenic acids (reverseGPx cycle). These studies reveal that the simple 6-OMe groups play multiple roles in each of the catalytically active intermediates by introducing steric and electronic effects that are required for efficient catalysis.

Full text of DOI:10.1002/chem.200800963

DOI: 10.1002/chem.200800963
A Simple and Efficient Strategy To Enhance the Antioxidant Activities of
Amino-Substituted Glutathione Peroxidase Mimics
Krishna P. Bhabak and Govindasamy Mugesh*[a]
Abstract: The glutathione peroxidase
(GPx) activities of some diaryl disele-
nides incorporating tertiary amino
groups were studied with H2O2, Cum-
OOH, and tBuOOH as substrates and
with PhSH as thiol co-substrate. Simple
replacement of a hydrogen atom with a
methoxy group dramatically enhances
the GPx activity. The introduction of
methoxy substituents ortho to selenium
in N,N-dialkylbenzylamine-based com-
pounds makes the basicity of the
amino groups perfect for the catalysis.
The presence of 6-OMe groups pre-
vents possible Se···N interactions in the
selenols, increasing their zwitterionic
characters. The methoxy substituents
also protect the selenium in the sele-
nenic acid intermediates from overoxi-
dation to seleninic acids or irreversible
inactivation to selenonic acid deriva-
tives. The additional substituents also
play a crucial role in the selenenyl sul-
fide intermediates, by preventing thiol
exchange reactions—which would nor-
mally lead to an inactivation path-
way—at the selenium centers. The
strengths of Se···N interactions in the
selenenyl sulfide intermediates are dra-
matically reduced upon introduction of
the methoxy substituents, which not
only reduce the thiol exchange reac-
tions at selenium but also enhance the
nucleophilic attack of the incoming
thiols at sulfur. The facile attack of
thiols at sulfur in the selenenyl sulfides
also prevents the reactions between the
selenenyl sulfides and H2O2 that can
regenerate the selenenic acids (reverse-
GPx cycle). These studies reveal that
the simple 6-OMe groups play multiple
roles in each of the catalytically active
intermediates by introducing steric and
electronic effects that are required for
efficient catalysis.
Keywords: antioxidant activity
·
GPx mimics · noncovalent interac-
tions · selenium · selenoenzymes
Introduction
pleted in the reaction mixture, the selenenic acid produced
in response to GPx oxidation may undergo further oxidation
to seleninic (E-SeO2H) or selenonic (E-SeO3H) acids that
disturb the main catalytic pathway. In the catalytic cycle, the
rapid reactions of the selenenic acid with GSH and of the
resulting selenenyl sulfide (E-SeSG) with a second GSH to
produce the selenol appear to be very important, because
these reactions ensure that the selenium moiety in the
enzyme is not irreversibly inactivated.
Because of the importance of GPx as a natural antioxi-
dant enzyme that plays a crucial role in oxidative stress,[2]
several groups have been working toward the design and
synthesis of small-molecule organoselenium compounds that
mimic the peroxide-reducing ability of GPx in the presence
of thiols. After the successful identification of ebselen (1) as
a clinically useful antioxidant and anti-inflammatory drug,[3]
several other selenium compounds displaying GPx-like ac-
tivities were reported in the literature (Figure 1.). These
compounds can be classified into two categories depending
upon their reaction behavior with thiols and peroxides. In
the first category, the compounds (e.g., 1, 2, 4–8, 11, and 12)
undergo one- or two-step reactions with excess amounts of
Glutathione peroxidase (GPx) is a selenocysteine-containing
mammalian antioxidant enzyme that protects various organ-
isms from oxidative damage by catalyzing the reduction of
harmful peroxides in the presence of glutathione (GSH).[1]
The GPx redox cycle involves the oxidation of the catalyti-
cally active selenol (E-SeH) moiety by the peroxides to pro-
duce the corresponding selenenic acid (E-SeOH), which
upon reaction with the thiol cofactor GSH generates the
key intermediate selenenyl sulfide (E-SeSG). The attack of
a second GSH moiety at the -Se-S- bond regenerates the
active site selenol with release of the cofactor in its oxidized
form, GSSG (Scheme 1).[1e–g] When the GSH cofactor is de-
[a] K. P. Bhabak, Prof. G. Mugesh
Department of Inorganic and Physical Chemistry
Indian Institute of Science, Bangalore 560 012 (India)
Fax : (+91)80-2360-1552/2360-0683
E-mail: mugesh@ipc.iisc.ernet.in
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800963.or from the author.
8640
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8640 – 8651

FULL PAPER
lead to a thiol exchange reaction at the selenium center
rather than a nucleophilic attack of the incoming thiol at the
sulfur center. Recent model studies on low-molecular-
weight selenium compounds show that reductions of sele-
nenyl sulfides to selenols need to overcome large energy
barriers, and so the nucleophilic attack of thiols (or thio-
ꢀ
lates) at the selenium centers of the Se S moieties are
therefore more favorable than at sulfur.[6] The thiol attack at
the selenium center in a selenenyl sulfide is further en-
hanced by nonbonding interactions between selenium and
other heteroatoms such as O and N.[7,8b–d] The effect of
thiols on the GPx activity of ebselen is one of the interesting
examples in which these nonbonding interactions play an
important role. These thiol exchange reactions, hampering
the regeneration of the catalytically active selenol species,
may account for the relatively low catalytic activities of syn-
thetic selenium compounds with certain thiol cofactors. Eb-
selen, for example, has been found to be an inefficient cata-
lyst in the reduction of hydroperoxides with aryl and benzyl
thiols (such as PhSH and BnSH) as cofactors,[5c,d,8] due to
the thiol exchange that takes place at the selenium center in
the selenenyl sulfide intermediate (14, Scheme 2).[8b]
Scheme 1. Proposed catalytic mechanism of GPx, involving selenol, sele-
nenyl sulfide, and selenenic acid intermediates. In the absence of thiols,
the selenenic acid undergoes overoxidation to produce the seleninic acid
species.
Scheme 2. The two possible reactions of the selenenyl sulfides derived
from ebselen (1) and of diaryl diselenides incorporating tertiary amino
groups.
The amino-substituted diselenide 4 exhibits much higher
GPx activity than ebselen in the presence of aromatic thio-
ls.[4b,e] This is probably due to the relatively weak Se···N in-
teractions in compound 17 relative to the Se···O interactions
in compound 14. However, the selenenyl sulfide 17 does un-
dergo a thiol exchange reaction in addition to the reaction
that leads to the formation of selenol 18. Furthermore, the
strength of Se···N interactions in 17 is sufficient to perform a
reverse-GPx cycle (i.e., the reaction of selenenyl sulfides
with peroxides to regenerate the corresponding selenenic
acids), which considerably reduces the GPx activity.[4e] These
observations suggest that the prevention of strong Se···N in-
teractions in amino-substituted diaryl diselenides might lead
to the development of better GPx mimics.
Unfortunately, it has not been possible to design and syn-
thesize diaryl diselenides that have basic amino groups in
close proximity to selenium but do not exhibit any strong
Se···N interactions in the selenenyl sulfide intermediates. In
this paper we report that the replacement of an aryl proton
in compound 4 by a methoxy group prevents the Se···N in-
Figure 1. Ebselen and some other important synthetic GPx mimics re-
ported in the literature.
thiols to generate reactive selenols, which undergo reactions
with the peroxides to produce the corresponding selenenic
acids.[3a,4] The selenenic acids thus produced undergo further
reactions with thiols to regenerate the selenols through the
formation of selenenyl sulfides as the key intermediates.
The second class of compounds (e.g., 3, 9, 10, 13), on the
other hand, do not produce any selenols, but they effectively
reduce the peroxides in the presence of thiols through SeII–
SeIV redox cycles.[5]
If we consider the GPx catalytic cycle, the reaction of E-
SeOH with GSH would be expected to be facile, due to the
electrophilic nature of selenium. On the other hand, the re-
action of E-SeSG with GSH to produce the selenol form of
the enzyme is an unusual and unexpected reaction, because
the selenium center in the E-SeSG intermediate would also
be expected to display electrophilic reactivity.[1e] This would
Chem. Eur. J. 2008, 14, 8640 – 8651ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8641

G. Mugesh and K. P. Bhabak
teractions in the key intermediates and dramatically enhan-
ces the GPx activities of 4 and of some related compounds.
We also show that the strong Se···N interactions reduce the
basicity of the tertiary amino group and that the introduc-
tion of a methoxy substituent helps the tertiary amino group
to act as a general base during the catalytic cycles of benzyl-
amine-based compounds.
metric selenenylation reactions.[9] They showed the in-
creased transfers of chirality to be due to the presence of
the methoxy substituents, which exhibit Se···O noncovalent
interactions. The additional substituents may also provide
the necessary steric environments around the selenium moi-
eties.[10]
From Table 1and Figure 2, it is clear that the methoxy-
substituted diselenides (22–24) are much better catalysts
than the diselenides 4, 20, and 21 and that compounds 22–24
show very high activities in all three peroxide (H2O2, Cum-
OOH, tBuOOH) systems. In particular, the diselenide 22
Results and Discussion
1
The amino-substituted diselenides (4, 20–24) required for
this study were synthesized from N,N-dialkylbenzylamines,
2-bromo-N,N-dialkylbenzylamines, or 3-methoxy-N,N-dia-
was found to be a remarkably active catalyst, with the t =
2
value (3.8 min) obtained for this compound at 5 mm concen-
tration being much lower than that found for 4 (19.2 min) at
10 mm concentration. This indicates that the catalytic activity
of the methoxy-substituted compound 22 is almost one
order of magnitude higher than that of 4. Furthermore,
quantitative conversion of PhSH into PhSSPh over the
given time period was observed only with compounds 22–24.
The times required for complete conversion were found to
be 12.5 min for 22, 23 min for 23, and 30 min for 24
(Figure 2). In contrast, only 90%, 77%, and 86% conver-
sions were observed after 58 min, 44 min, and 69 min, re-
spectively, when compounds 4, 20, and 21 were used as cata-
lysts. In the presence of ebselen, only 11.5% conversion was
observed after a reaction time of 100 min (Figure 2). This is
lkylbenzylamines by the well-established heteroatom-direct-
ed ortho-lithiation methodology. Metallation of substituted
benzylamines with nBuLi in diethyl ether afforded the cor-
responding ortho-lithiated compounds. Subsequent treat-
ment with selenium powder and oxidative workup afforded
diselenides in moderate yields. The GPx-like activities of
these compounds were studied with hydrogen peroxide
(H2O2), cumene hydroperoxide (Cum-OOH), and tert-butyl
hydroperoxide (tBuOOH) as the substrates and with PhSH
as the thiol co-substrate. The catalytic reduction of perox-
ides by PhSH in the presence of various selenium com-
pounds was studied by a method similar to that of Back and
co-workers.[4c,5c,d] The formation of PhSSPh in the reactions
was studied by a reversed-phase HPLC method, and the
times required for 50% conversion of PhSH into PhSSPh
1
(t = ) were calculated by determining the peak areas at differ-
2
ent time intervals. A calibration plot was used to calculate
the amounts of PhSSPh formed during the reactions.
As expected, the diselenides based on N,N-dialkylbenzyla-
mine moieties (4, 20, 21) exhibited much higher GPx activi-
ties than ebselen in all three peroxide systems used. It has
previously been shown that the high GPx activity of 4 is
probably due to the presence of the dimethylamino group,
which can deprotonate the selenol 18 to generate a more re-
active selenolate.[4b,e] Similar activation may occur in com-
pounds 20 and 21, containing a diethyl- and a dipropylamino
group, respectively.
Interestingly, remarkable enhancements in the activity
were observed when the protons at the 6-positions in com-
pounds 4, 20, and 21 were replaced by simple methoxy sub-
stituents. Wirth and co-workers had previously reported an
interesting observation that the substitution of aryl protons
by methoxy groups in certain chiral diselenides incorporat-
ing alcohol moieties improved the stereoselectivity in asym-
Figure 2. Catalytic reduction of Cum-OOH by PhSH in the presence of
various selenium compounds. The formation of PhSSPh was followed by
reversed-phase HPLC, and the % conversions were calculated from cali-
bration plots: a) 22, b) 23, c) 24, d) 4, e) 20, f) 21, g) 1, and h) control.
Assay conditions: catalyst (10.0 mm, except compound 22; [catalyst 22]=
5 mm), PhSH (1.0 mm), Cum-OOH (2.0 mm) in MeOH at 228C.
due to the thiol exchange reactions that take place at the se-
lenium centers in the selenenyl sulfides derived from com-
pounds 4, 20, 21, and ebselen, which lead to the accumula-
tion of the corresponding selenenyl sulfide species in the
solutions.[8b] Our HPLC experiments indicated that the sele-
nenyl sulfides were the predominant species in the reaction
mixtures, particularly when higher concentrations of the se-
lenium compounds (4, 20, 21, and ebselen) were used. In the
case of compound 4, the accumulation of the selenenyl sul-
fide species was observed even at lower concentrations of
8642
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8640 – 8651

Glutathione Peroxidase Mimics
FULL PAPER
catalyst (10 mm). In contrast, no such species was detected
for compound 22 during the entire catalytic cycle (Figure S2,
Supporting Information).
teractions in compounds 22–24 are sufficient for facile re-
duction of the diselenide bonds by thiols (Scheme 3).
Treatment of the diselenides 4, 20, and 21 with PhSH
(1equiv in each case) produced the corresponding selenols
1
Table 1. Values of t = for the reduction of peroxides by PhSH in the pres-
2
ence of compounds 1, 4, and 20–24 at 228C.
t = values [min][a]
1
Compound
2
H2O2
Cum-OOH
tBuOOH
control
1 (ebselen)
4
20
21
22[b]
23
24
1460.0
821.0
19.2
19.9
22.7
3.8
1053.0
522.0
16.5
17.6
24.3
2.6
780.0
744.0
24.4
35.2
49.7
6.3
12.5
13.2
9.1
12.1
29.0
35.2
[a] Assay conditions: the reactions were carried out in MeOH at 228C
[catalyst (10.0 mm, except compound 22), PhSH (1.0 mm), peroxide
(2.0 mm)]. [b]The conversion was too fast to be measured at 10 mm con-
centration, and so a 5 mm concentration of the catalyst was used.
Scheme 3. Proposed catalytic mechanism for the reduction of H2O2 by
PhSH in the presence of compounds 22–24.
25, 27, and 29, together with the selenenyl sufides 31, 33,
and 35, respectively, in nearly 1:1 ratios. The addition of fur-
ther PhSH (a second equivalent in each case) does not lead
To understand the higher GPx activities of compounds
22–24 in relation to those of the parent diselenides (4, 20–
21), we have undertaken a detailed study to probe the role
of amino and methoxy substituents. We have studied the
catalytic cycles of compounds 4, 20, and 21 individually and
have compared them with those of 22–24 (Scheme 3) to un-
derstand the role of amino substituents in each catalytically
active intermediate. Although the catalytic mechanisms of
diselenides 22–24 were found to be identical with those of 4,
20, and 21, these studies revealed that the introduction of
the methoxy substituents leads to dramatic changes in the
reactivity of selenium. Our DFT calculations and experi-
mentally measured 77Se NMR chemical shifts (Table 2) sug-
gest that the strengths of the Se···N interactions in com-
pounds 22–24 are significantly less than those in 4, 20, and
21, which is consistent with the report by Wirth et al. that
the presence of a methoxy substituent in the 6-position pre-
vents interaction between selenium and the alcohol side-
chain.[9] Although the strong Se···N interactions in com-
pound 4 and related diselenides have been shown to be im-
portant for the reductive cleavage of the -Se-Se- bond by
thiols,[4b,e] we have found that the relatively weak Se···N in-
to complete conversion of the selenenyl sufides into the se-
lenols. This is due to the presence of Se···N interactions in
the selenenyl sulfides 31, 33, and 35, which lead to thiol ex-
change reactions rather than to the formation of the corre-
sponding selenols. The 77Se NMR chemical shifts (Table 2)
for compounds 31 (564 ppm), 33 (558 ppm), and 35
(555 ppm) show significant downfield shifts relative to that
of PhSeSPh (526 ppm). Large excesses of thiol are therefore
required for the conversion of the selenenyl sulfides into the
selenols. However, the selenenyl sulfides 31, 33, and 35 were
found to be the major species during the catalytic cycle. The
addition of 4-Me-C6H4SH to solutions containing 31, 33, or
35 produced new selenenyl sulfides through thiol exchange
reactions (Figure S3, Supporting Information). Interestingly,
the reactions of compounds 22–24 with PhSH (1equiv in
each case) readily produced the corresponding selenols (26,
28, and 30) with the formation of only trace amounts of the
corresponding selenenyl sulfides (32, 34, and 36). As the
amounts of thiol were not sufficient for quantitative conver-
sion of 22–24 into the selenols, we detected some unreacted
diselenides in the reaction mixtures. However, the addition
of further PhSH (a second equivalent in each case) to the
reaction mixtures converted the diselenides and selenenyl
Table 2. The theoretical data for 14 and 31–36 obtained by DFT calcula-
tions at the B3LYP/6-31+G(d)//B3LYP/6-311+G(d,p) levels, along with
G
the experimentally measured and theoretically calculated 77Se NMR
chemical shifts.
Compd
rSe···O/N
[Š]
qO/N-Se-S
77Se [ppm]
(calcd)[a]
77Se [ppm]
(exptl)
[8]
N
ACHTREUNG
14 (R=Ph)
2.456
2.680
2.869
2.690
2.819
2.663
2.846
176.2
175.5
158.5
168.6
158.3
176.2
158.7
660
556
419
494
414
522
423
588
564
470
558
461
555
451
31
32
33
34
35
36
[a] The values are cited with respect to Me2Se.
Chem. Eur. J. 2008, 14, 8640 – 8651ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8643

G. Mugesh and K. P. Bhabak
sulfides completely into the se-
lenols. The 77Se NMR chemical
shifts for the selenenyl sulfides
32 (470 ppm), 34 (461ppm),
and 36 (451ppm) show dra-
matic
upfield
shifts
ACHTREUNG
those of the selenenyl sulfides
31, 33, and 35, indicating the
absence of strong Se···N inter-
actions. The absence of any
strong Se···N interactions and
the presence of the methoxy
groups at the 6-positions pre-
vent the thiol exchange reac-
tions at the selenium centers.
Therefore, the addition of one
equivalent of thiol to a sele-
nenyl sulfide 32, 34, or 36 is
sufficient to generate the cor-
responding selenol.
To complement the experi-
mental findings, we have also
carried out detailed DFT stud-
ies on the selenenyl sulfide in-
termediates. The geometries
were fully optimized at the
B3LYP level with use of the 6-
31+G(d) basis set. The inter-
action energies between the
selenium and nitrogen atoms
Figure 3. Energy-optimized geometries of the selenenyl sulfide intermediates 31–36. The optimizations of the
geometries were performed at the B3LYP/6-31+G(d) level of theory.
were calculated by natural bond orbital (NBO) calcula-
tions.[11] These studies have revealed that the strengths of
Se···N interactions in compounds 31, 33, and 35 are dramati-
cally decreased upon incorporation of methoxy groups at
the 6-positions (Table 3, Figure 3). As an example, the
Se···N distance in compound 32, with a methoxy substituent
(2.869 Š), is significantly longer than that in 31 (2.680 Š).
Similarly, the Se···N interaction energy in compound 32
teractions. The NBO analysis shows that the introduction of
the methoxy substituents and the subsequent weakening of
the Se···N interactions lead to increases in the positive
charges on the sulfur atoms in the selenenyl sulfides
(Table 3), which would enhance the possibility of nucleo-
philic attack of incoming thiol/thiolate at the sulfur centers.
These changes in the electronic properties of selenium and
sulfur and the increases in the steric hindrance around sele-
nium upon introduction of methoxy substituents drive the
conversion of selenenyl sulfides into the corresponding sele-
nols.[12]
Because the reactions of selenenyl sulfides with thiols
should produce the corresponding selenols for the catalytic
activity, we have also studied the natures of the selenols by
experimental and computational methods (Figure 4). The
N,N-dialkylamino groups in the methoxy-substituted sele-
nols (26, 28, and 30) have been found to be stronger bases
than those in the selenols 25, 27, and 29. This can easily be
explained by comparing the 77Se NMR chemical shifts. The
experimentally measured 77Se NMR chemical shifts for the
methoxy-substituted selenols are shifted almost 100 ppm up-
field [26 (ꢀ58 ppm), 28 (ꢀ42 ppm), 30 (ꢀ37 ppm)] relative
to those for the unsubstituted selenols [25 (35 ppm), 27
(54 ppm), 29 (56 ppm)] (Table 4), indicating that the intro-
duction of a methoxy substituent significantly increases zwit-
terionic character.[13] The selenium centers in compounds 26,
(ESe···N: 5.78 kcalmolꢀ1) is much lower than that in 31 (ESe···N
:
11.20 kcalmolꢀ1). Consistently with our experimental data,
the calculated 77Se NMR chemical shift values for 32, 34,
and 36 are shifted upfield by ꢁ120 ppm relative to those of
31, 33, and 35. This indicates that the selenium centers in
compounds 31, 33, and 35 are more deshielded than in the
methoxy-substituted compounds, due to the strong Se···N in-
Table 3. Theoretical data for 14 and 31–36 obtained by NBO analysis at
the B3LYP/6-31+G(d)//B3LYP/6-311+G(d,p) level of theory.
G
Compd
qSe
qS
ESe···O/N [kcalmolꢀ1
]
14 (R=Ph)
0.408
0.309
0.304
0.297
0.3010.063
0.306
0.305
0.023
0.039
0.067
0.040
15.20
11.20
5.78
31
32
33
34
35
36
11.23
7.02
0.035
0.065
12.57
6.59
8644
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8640 – 8651

Glutathione Peroxidase Mimics
FULL PAPER
2.37 kcalmolꢀ1 more stable
than its counterpart with weak
Se···N interactions, and that
the zwitterionic form of 15 is
about
5.45 kcalmolꢀ1
less
stable than its counterpart with
weak Se···O interactions.[8b] In
this study, we have found that
the zwitterionic form 25b is
about
5.0 kcalmolꢀ1
more
stable than the undissociated
form (25a), which is consistent
with the experimentally mea-
sured 77Se NMR data. The
zwitterionic form of 26 (i.e.,
26b), on the other hand, is
ꢁ9.0 kcalmolꢀ1 more stable
than the undissociated 26a, in-
dicating that the prevention of
Se···N interactions through the
introduction of a methoxy sub-
stituent increases the stability
of the zwitterionic form.[13]
Having established the role
of methoxy substituents in the
selenols and selenenyl sulfides,
we have also studied the effect
of Se···N interactions in the se-
lenenic acids, one of the crucial
intermediate types in the GPx
catalytic cycle. Treatment of
compound 25 with hydrogen
peroxide produced two signals
in the 77Se NMR spectrum, at
1168 and 1347 ppm, that can
be ascribed to the selenenic
acid 37 and the seleninic acid
43, respectively.
Figure 4. Energy-optimized geometries of the selenol species 25–30. The optimizations of the geometries were
performed at the B3LYP/6-31+G(d) level of theory.
Table 4. The theoretical data for 15 and for 25–30 obtained by DFT calculations at the B3LYP/6-31+G(d)//
B3LYP/6-311+G
(d,p) levels, along with the experimentally measured and theoretically calculated 77Se NMR
A
chemical shifts.
Compd
rSe···O/N
[Š]
qSe
77Se [ppm]
(calcd)[a]
77Se [ppm]
(exptl)[a]
ESe···O/N
G
G
[kcalmolꢀ1
]
15
25
26
27
28
29
30
2.577
3.028
3.016
3.062
3.055
3.059
3.078
0.269
206
ꢀ30
ꢀ94
84
232
35
7.10
0.00
0.00
0.00
0.00
0.00
0.00
ꢀ0.289
ꢀ0.246
ꢀ0.312
ꢀ0.279
ꢀ0.314
ꢀ0.295
ꢀ58
54
ꢀ2
ꢀ42
56
ꢀ37
63
0
[a] The values are cited with respect to Me2Se.
28, and 30 are, therefore, more nucleophilic than those in
25, 27, and 29. Furthermore, the selenol moieties in 26, 28,
and 30 are significantly different from that in the ebselen-se-
lenol (15), which shows a large downfield shift in the
77Se NMR (232 ppm, Table 2), due to the strong Se···O inter-
actions (ESe···O =7.10 kcalmolꢀ1). We have previously shown
that the zwitterionic form of 25 (i.e., 25b) in water is about
Treatment of selenol 27 with H2O2 produced a mixture of
selenenic acid 39 (1171 ppm) and seleninic acid 45
(1349 ppm), although the required selenenic acid 39 was de-
tected only in trace amounts. On the other hand, compound
29 did not produce any detectable quantities of the selenen-
ic acid (41), but produced only the overoxidized seleninic
acid 47 (1349 ppm).[14,15] The selenenic acid 37 was convert-
ed completely into the seleninic acid 43 and the selenonic
Chem. Eur. J. 2008, 14, 8640 – 8651ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8645

G. Mugesh and K. P. Bhabak
acid 49 upon addition of an
excess amount of H2O2. This is
consistent with previous re-
ports on the GPx activities of
amino-substituted
com-
pounds.[4b,e] The 77Se NMR
chemical shift for the selenonic
acid 49 (1019 ppm) is almost
identical with that of the struc-
turally characterized com-
pound 50 (1022 ppm) reported
by Iwaoka and Tomoda.[4b] In-
terestingly, treatment of sele-
nols 26, 28, and 30 with H2O2
produced the selenenic acids
38 (1170 ppm), 40 (1174 ppm),
and 42 (1172 ppm), respective-
ly, and the formation of the
overoxidized seleninic acids
(44, 46, and 48) or selenonic
acids was not observed even at
very high peroxide concentra-
tions.[14,15]
The optimized geometries
(Figure 5) and NBO analysis
(Table 5) indicate that all the
selenenic acids (37–42) exhibit
strong
Se···N
interactions,
Figure 5. Energy-optimized geometries of the selenenic acid (37–42) intermediates. The optimizations of the
geometries were performed at the B3LYP/6-31+G(d) level of theory.
which help nucleophilic attack
by incoming thiols at the sele-
nium centers. Interestingly, the
addition of PhSH (1equiv in
each case) led to clean conversions of the selenenic acids 38,
40, and 42 into the corresponding selenenyl sulfides. No spe-
cies other than the selenols, selenenyl sulfides, and selenenic
acids were observed during the entire catalytic cycles of the
diselenides 22–24, and this type of selectivity appears to be
remarkable in amino-substituted benzylic compounds. The
DFT calculations on the selenenic and seleninic acids have
revealed that the introduction of methoxy substituents at
their 6-positions prevents the overoxidation of the selenenic
acids to the corresponding seleninic acids. The energy differ-
ence between 37 and 43 is almost 4.0 kcalmolꢀ1 lower than
that between 38 and 44, indicating that the conversion of 37
into 43 is more favored than the conversion of 38 into 44.
In addition to the thiol exchange, the reactions of selenen-
yl sulfides with H2O2 to produce the corresponding selenenic
and/or seleninic acids (reverse-GPx cycle) also reduce GPx
activity. The addition of one equivalent of H2O2 to the sele-
nenyl sulfide 31 readily produced the selenenic acid 37
(Scheme 4), which underwent further reaction with H2O2 to
generate the seleninic acid 43. The presence of strong Se···N
ꢀ
interactions and subsequent stabilization of the Se S bond
by thiol exchange reactions permit the selenenyl sulfide 31
to stay longer in the solution. As a result, a facile cleavage
ꢀ
of the Se S bond by H2O2 occurs, leading to the production
of the selenenic and seleninic acids. Because of the back-
ward reaction, the reaction between selenenic acid 37 and
PhSH always produces a mix-
ture of the selenenyl sulfide 31
and the selenenic acid 37. The
reverse-GPx cycle has been ob-
served previously with com-
pounds 4, 6, and related disele-
nides.[4b,e] The introduction of
methoxy substituents in the 6-
positions makes the selenenyl
sulfides short-lived, and the re-
actions therefore proceed in
the forward direction. As an
Table 5. Theoretical data for 37–42 obtained by DFT calculations at the B3LYP/631+G(d)//B3LYP/6-311+G-
A
Compd
rSe···O/N
[Š]
qO/N-Se-O
qSe
77Se (ppm)
(calcd)[a]
ESe···O/N
[8]
N
[kcalmolꢀ1
]
37
38
39
40
41
42
2.473
2.630
2.579
2.658
2.576
2.665
172.7
167.3
172.9
165.8
172.4
165.6
0.631
0.625
0.629
0.622
0.626
0.620
1047 (1168)
1068 (1170)
1056 (1171)
1060 (1174)
1062 (n.d)
20.47
12.84
16.02
12.41
16.24
12.04
1062 (1172)
[a] The 77Se chemical shifts were calculated in the gas phase and are cited with respect to Me2Se. The experi-
mentally measured 77Se chemical shifts are given in parentheses.
8646
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8640 – 8651

Glutathione Peroxidase Mimics
FULL PAPER
the methoxy groups at the 6-positions play a crucial role in
modulating the reactivities of the key intermediates in the
GPx cycle.
To gain further insight into the effect of Se···N/O interac-
tions on the kinetic behavior of synthetic GPx mimics, we
have carried out detailed kinetic studies on compounds 4
and 22. The initial rates were measured with increasing con-
centrations of thiol and fixed concentrations of catalyst and
hydrogen peroxide. The reaction rate for compound 22 in-
creased rapidly at the beginning with an increase in the thiol
concentration, and after a certain concentration of thiol, the
rate became constant. This indicates that compound 22 fol-
lows typical saturation kinetics similar to those of the native
GPx.[16] Interestingly, a linear increase in the reaction rate
was observed for compound 4 with an increase in the con-
centration of PhSH, indicating non-saturation kinetics up to
a concentration of 3 mm. The change in the kinetics pattern
can be attributed to the various degrees of thiol exchange
reactions in the selenenyl sulfides derived from compounds
4 and 22. When there are thiol exchange reactions due to
Scheme 4. The reaction between selenenic acid 37 and PhSH produces
the selenenyl sulfide 31, which upon treatment with H2O2 regenerates 37
through a reverse-GPx cycle. Compound 38 undergoes only the forward
reaction, with the reverse reaction not being observed because of the
presence of the OMe group.
example, the 77Se NMR signal at 1170 ppm due to the sele-
nenic acid 38 disappeared completely upon addition of one
equivalent of PhSH. This reaction produced a new signal for
the selenenyl sulfide 32 at 470 ppm, which was unaffected
by the addition of an excess amount of H2O2. Similar reac-
tivity was observed with compounds 40 and 42. This clearly
indicates that the backward reaction involving the selenenyl
sulfides and H2O2 would reduce the GPx activity.
ꢀ
strong Se···N interactions, the Se S bonds become readily
exchangeable, and the selenium centers in such compounds
therefore do not get saturated with thiols. This leads to rela-
tively high KM values, as observed in the case of 4. In con-
trast with 4, compounds 20 and 21 exhibited saturation ki-
netics (Tables S27 and S28, Supporting Information), indi-
cating that the replacement of the methyl groups in 4 by
ethyl or propyl groups had considerably reduced the thiol
exchange reactions at the selenium center. This is evident
from the reactions of 31, 33, and 35 with 4-Me-C6H4SH, in
which the thiol exchange reaction rate for compound 31 is
significantly higher than those for 33 and 35 (Figure S3, Sup-
porting Information). However, the introduction of sterical-
ly more demanding substituents at the nitrogen does not
appear to be sufficient for enhancing the catalytic activity,
because the strengths of Se···N interactions in 33 and 35
were found to be almost identical with those in 31 (Table 3).
In contrast with the diselenides, ebselen exhibited com-
pletely different kinetics in the presence of various concen-
trations of PhSH. When ebselen was treated with PhSH, a
decrease in the reaction rate with increasing thiol concentra-
tion (up to ꢁ2 mm) was observed. After this, a rapid in-
crease in the reaction rate with increasing PhSH concentra-
tion was observed. For example, the difference in the rate
(1.94 mmminꢀ1) due to the
To understand the effect of thiol co-substrate on the cata-
lytic activities of these amine-based compounds, we deter-
mined the initial rates at various concentrations of PhSH
with fixed catalyst and peroxide concentrations. The double
reciprocal or Lineweaver–Burk plots obtained for com-
pounds 4 and 20–24 by plotting the reciprocals of initial
rates (1/v0) against the reciprocals of substrate concentra-
tions (1/[substrate]) were used to determine the catalytic pa-
rameters (Table 6). The KM (Michaelis constant) values ob-
tained for 22–24 are much lower than those for 4, 20, and 21
under similar experimental conditions, suggesting that the
thiol exchange reactions significantly increase the KM values.
The catalytic efficiencies (h) for the methoxy-substituted
compounds are found to be much higher than those of the
parent benzylamine-based compounds. The catalytic effi-
ciency of 22 (9.75”103 mꢀ1 minꢀ1) is almost eight times
higher than that of 4 (1.24”103 mꢀ1 minꢀ1). Similarly, the cat-
alytic efficiencies of compounds 23 and 24 are found to be
around two and around five times higher than those of 20
and 21, respectively. The higher catalytic efficiencies of com-
pounds 22–24 relative to the parent compounds suggest that
change in thiol concentration
from 7 mm to 14 mm was found
to be much higher than the dif-
Table 6. Catalytic parameters [maximum velocities (Vmax), Michaelis constants (KM), catalytic constants (kcat),
and catalytic efficiencies (h)] for the reduction of H2O2 by PhSH in the presence of diselenides 4 and 20–24.[a]
Compd
Vmax [mmminꢀ1
]
KM [mm]
kcat [minꢀ1
]
h [mꢀ1 minꢀ1
]
ference
in
the
rate
(0.62 mmminꢀ1) over the first
7 mm (Figure 6B). This indi-
cates that a very high concen-
tration of PhSH is required to
overcome the thiol exchange
reaction in the selenenyl sul-
fide derived from ebselen,
which is consistent with our
4
158.7[b]
21.9
17.9
42.9
9.7
12.76[b,c]
1.04
1.70
0.44
0.22
15.87
2.19
1.79
4.29
0.98
0.87
1.24”103
2.10”103
1.05”103
9.75”103
4.45”103
5.43”103
20
21
22
23
24
8.7
0.16
[a] Assay conditions: PhSH (0.0–4.0 mm), H2O2 (1.0 mm), and catalyst (10.0 mm) in MeOH at 228C. [b] The ex-
tremely high Vmax and KM values are due to the use of saturation kinetics method for a compound that follows
non-saturation kinetics. [c] The KM value is much higher than the largest concentration of PhSH.
Chem. Eur. J. 2008, 14, 8640 – 8651ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8647

G. Mugesh and K. P. Bhabak
Conclusion
In this study we have shown that simple replacement of hy-
drogen atoms by methoxy substituents in N,N-dialkylbenzyl-
amine-based diselenides can lead to dramatic increases in
the catalytic activity. The methoxy substituents enhance
GPx-like activity by altering the steric and electronic envi-
ronments around selenium and sulfur atoms in the key inter-
mediates. Protection of the selenium moieties from overoxi-
dation by peroxides and the prevention of thiol exchange re-
actions at the selenium atoms in the selenenyl sulfide inter-
mediates upon introduction of the methoxy substituents
have been found to be the crucial factors for the enhance-
ment of catalytic activity. These studies have revealed that
the basic amino groups in close proximity to selenium in dis-
elenides possessing tertiary amino groups play more positive
roles when methoxy groups are present at the 6-positions.
From our present study and the literature data, we propose
the following revised roles for the basic amino groups in
GPx mimics. The tertiary amino substituents: i) should not
be involved in any Se···N interactions in the selenols, but
should be sufficiently basic to deprotonate the selenols to
produce more reactive selenolates, ii) should not participate
in strong interactions with selenium in the selenenyl sulfide
intermediates, and iii) should exhibit some noncovalent in-
teractions with selenium in the selenenic acid intermediates
to increase the electrophilic reactivity of selenium.
Figure 6. Effect of thiol co-substrate on initial rates for the reduction of
H2O2 (1.0 mm) by PhSH in the presence of 4 and 22 (10 mm) or of ebselen
(1, 20 mm). The reactions were carried out in methanol at 228C.
A) Line a) 22; line b) 4. B) 1 (ebselen).
previous observations.[8b] As a result of this unusual behav-
ior, the kinetic parameters for the reduction of H2O2 by
PhSH could not be determined for ebselen. This is consis-
tent with the report by Shi et al. that the effect of ebselen in
cells is beneficial only when thiols are present in sufficiently
high concentrations, and that the detrimental effects of ebse-
len may dominate in a system with thiols in low concentra-
tions.[17]
Although compounds 4 and 22 exert their catalytic cycle
through the formation of the selenols, selenenyl sulfides,
and selenenic acids, comparison of the initial rates for differ-
ent selenium compounds can lead to unreliable results when
the initial rates are measured at only one thiol concentra-
tion. As an example, the initial rates for 10 mm concentra-
tions of compounds 4 and 22 are identical at a 3.4 mm con-
centration of PhSH (Figure 6A). This is due to the differ-
ence in the kinetics behavior (saturation vs. non-saturation
kinetics), which does not allow a reliable comparison. The
activity of 22 must therefore be compared with that of 4 at
PhSH concentrations of 1m m or less, at which both com-
pounds exhibit linear increases in their activities with in-
creasing thiol concentrations (Figure 6A). These observa-
tions suggest that the determination of initial rates at only
one thiol concentration can be erroneous even for com-
pounds with similar structures if the reactivities of the sele-
nium centers are altered by Se···N or other noncovalent in-
teractions.
Experimental Section
General procedure: n-Butyllithium (nBuLi) was purchased from Acros
Chemical Co. (Belgium). Methanol was obtained from Merck and dried
before use. All other chemicals were of the highest purity available. All
the reactions were carried out under nitrogen with use of standard
vacuum-line techniques. Because of the unpleasant odors and toxic
nature of several of the reaction mixtures involved, most manipulations
were carried out in a well-ventilated fume hood. Et2O was dried over
sodium metal with benzophenone. Thin-layer chromatography analyses
were carried out on pre-coated silica gel plates (Merck), and spots were
visualized with UV irradiation. Column chromatography was performed
on glass columns loaded with silica gel or on automated flash chromatog-
raphy systems (Biotage) with use of preloaded silica cartridges. 1H
(400 MHz), 13C (100.56 MHz), and 77Se (76.29 MHz) NMR spectra were
obtained on a Bruker 400 MHz NMR spectrometer. Chemical shifts are
cited with respect to SiMe4 as internal (1H and 13C) and Me2Se as exter-
nal (77Se) standard. Mass spectral studies were carried out on a Q-TOF
micro mass spectrometer with ESI MS mode analysis. The synthetic pro-
cedures for the ligands are described in the Supporting Information.
Compound 4 was synthesized by the literature method.[18]
Synthesis of 20: nBuLi (1.4 mL of a 1.6m hexane solution) was added
dropwise with stirring to a cooled (ꢀ788C) solution of 2-bromo-N,N-di-
ethylbenzylamine (0.50 g, 2.06 mmol) in dry Et2O (15 mL), which was
then allowed slowly to attain room temperature. After 1.5 h, the solvent
was removed completely under reduced pressure, to remove butyl bro-
mide produced in the reaction. Freshly distilled Et2O (15 mL) was added,
followed by the addition at 08C of finely ground selenium powder
(0.16 g, 2.06 mmol). After the addition of selenium powder the color
turned brownish. After the reaction mixture had been stirred for another
3 h, the mixture was poured into an ice-cooled saturated sodium bicar-
bonate solution. Oxygen was passed through the solution at a moderate
rate for 15 min. The compound was extracted with ether and dried over
8648
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8640 – 8651

Glutathione Peroxidase Mimics
FULL PAPER
sodium sulfate. The solvent was evaporated to obtain a yellow colored
liquid, which was purified by flash chromatography. The expected com-
pound was eluted with ethyl acetate in petroleum ether (5%). Yield
0.48 g (49%); 1H NMR (CDCl3): d=1.02–1.06 (m, 6H), 2.54–2.60 (m,
4H), 3.63 (s, 2H), 7.07–7.10 (m, 1H), 7.25–7.29 (t, J=8 Hz, 1H), 7.49–
Synthesis of 24: nBuLi (1.7 mL of a 1.6m solution in hexane) was added
dropwise with stirring at ꢁ58C to a solution of 3-methoxy-N,N-dipropyl-
benzylamine (0.50 g, 2.26 mmol) in dry Et2O (15 mL), and the mixture
was allowed slowly to attain room temperature. After 1.5 h, finely
ground selenium powder (0.18 g, 2.26 mmol) was added at 08C. After the
addition of selenium powder the color turned brownish, and the system
was stirred for another 2 h at room temperature. The mixture was then
poured into an ice-cooled saturated sodium bicarbonate solution. Oxygen
was passed through the solution at a moderate rate for 15 min. The com-
pound was extracted with ether and dried over sodium sulfate. The sol-
vent was evaporated to provide a yellow-colored liquid, which was puri-
fied on an active neutral alumina column with ethyl acetate and petrole-
7.51(d,
J=8 Hz, 1H), 7.54–7.56 ppm (d, J=8 Hz, 1H); 13C NMR
(CDCl3): d=12.0, 47.2, 57.2, 124.2, 127.2, 128.0, 130.6, 132.5, 139.6 ppm;
77Se NMR (CDCl3): d=424 ppm; HRMS: m/z: calcd for C22H32N2Se2
[M+H]+: 485.0974; found: 485.0977.
Synthesis of 21: nBuLi (1.4 mL of a 1.6m solution in hexane) was added
dropwise with stirring at ꢀ788C to a solution of 2-bromo-N,N-dipropyl-
benzylamine (0.50 g, 1.85 mmol) in dry Et2O (15 mL), and the mixture
was allowed to reach room temperature slowly over 1.5 h. The solvent
was removed under reduced pressure to remove the butyl bromide pro-
duced in the reaction. Freshly distilled Et2O (15 mL) was then added, fol-
lowed by the addition at 08C of finely ground selenium powder (0.15 g,
1.85 mmol). After the addition of selenium powder the color turned
brownish. After the reaction mixture had been stirred for another 3 h,
the mixture was poured into an ice-cooled saturated sodium bicarbonate
solution. Oxygen was passed through the solution at a moderate rate for
15 min. The compound was extracted with ether and dried over sodium
sulfate. The solvent was evaporated to provide a yellow colored liquid,
which was purified by flash chromatography. The expected compound
was eluted with ethyl acetate in petroleum ether (2–3%). Yield 0.52 g
1
um ether as eluents. Yield 0.62 g (46%); H NMR (CDCl3): d=0.75–0.79
(t, J=7.2 Hz, 6H), 1.36–1.41 (q, J=7.2 Hz, 4H), 2.25–2.29 (t, J=7.6 Hz,
4H), 3.43 (s, 2H), 3.70 (s, 3H), 6.66–6.68 (d, J=7.6 Hz, 1H), 6.80–6.82
(d, J=7.2 Hz, 1H), 7.08–7.12 ppm (t, J=7.2 Hz, 1H); 13C NMR (CDCl3):
d=10.9, 19.2, 54.0, 54.8, 57.6, 111.0, 112.9, 120.0, 127.9, 141.2, 158.5 ppm;
77Se NMR (CDCl3): d=371ppm; HRMS: m/z: calcd for C28H44N2O2Se2
[M+H]+: 601.1733; found: 601.0029.
Synthesis of 31: Thiophenol (20 mL, 0.17 mmol) was added at room tem-
perature to the stirred solution of 4 (50 mg, 0.12 mmol) in dichlorome-
thane (5 mL). The reaction mixture was stirred for 30 min, and the sol-
vent was then evaporated. The expected compound was purified by flash
chromatography on a silica gel column with ethyl acetate and petroleum
ether as eluents to provide a pale yellow-colored oil. Yield: 21mg
(56%); 1H NMR (CDCl3): d=2.32 (s, 6H), 3.61 (s, 2H), 7.11–7.16 (m,
3H), 7.94–7.25 (m, 3H), 7.51–7.53 (d, J=7.6 Hz, 2H), 7.96–7.98 ppm (d,
J=7.6 Hz, 1H); 13C NMR (CDCl3): d=44.3, 64.6, 126.3, 126.4, 128.1,
128.7, 129.2, 129.4, 136.2, 138.9 ppm; 77Se NMR (CDCl3): d=564 ppm;
HRMS: m/z: calcd for C15H17NSSe [M+H]+: 324.0247; found: 323.8765.
(52%); 1H NMR (CDCl3): d=0.74–0.80 (m, 6H), 1.36–1.45
(m, 4H),
A
2.32–2.39 (m, 4H), 3.54 (s, 2H), 6.97–7.02 (m, 1H), 7.17–7.21 (m, 1H),
7.41–7.43 (d, J=8 Hz, 1H), 7.50–7.52 ppm (d, J=8 Hz, 1H); 13C NMR
(CDCl3): d=12.9, 21.3, 57.2, 125.0, 128.1, 128.8, 131.4, 133.4, 140.7 ppm;
77Se NMR (CDCl3): d=420 ppm; HRMS: m/z: calcd for C26H40N2Se2
[M+H]+: 541.1600; found: 541.1597.
Synthesis of 33: Thiophenol (17 mL, 0.15 mmol) was added at room tem-
perature to the stirred solution of 20 (50 mg, 0.10 mmol) in dichlorome-
thane (5 mL). The reaction mixture was stirred for 30 min, and the sol-
vent was then evaporated. The expected compound was purified by flash
chromatography on a silica gel column with ethyl acetate and petroleum
ether as eluents to provide a pale yellow-colored oil. Yield: 18 mg
(51%); 1H NMR (CDCl3): d=0.97–1.00 (t, J=6.8 Hz, 6H), 2.55–2.60 (q,
J=6.8 Hz, 4H), 3.64 (s, 2H), 7.01–7.03 (m, 3H), 7.08–7.17 (m, 3H), 7.41–
7.43 (d, J=7.6 Hz, 2H), 7.85–7.87 ppm (d, J=8.0 Hz, 1H); 13C NMR
(CDCl3): d=9.2, 43.4, 57.8, 124.6, 124.8, 126.7, 127.0, 127.7, 134.1, 137.4,
138.0 ppm; 77Se NMR (CDCl3): d=558 ppm; HRMS: m/z: calcd for
C17H21NSSe [M+H]+: 352.0560; found: 351.9353.
Synthesis of 22: nBuLi (4.0 mL of a 1.6m solution in hexane) was added
dropwise with stirring at ꢁ58C to a solution of 3-methoxy-N,N-dimethyl-
benzylamine (0.75 g, 4.55 mmol) in dry THF (25 mL), and the mixture
was allowed slowly to attain room temperature. After 1.5 h, finely
ground selenium powder (0.43 g, 5.46 mmol) was added at 08C. After the
addition of selenium powder the color turned brownish and the reaction
mixture was stirred overnight. The mixture was then poured into an ice-
cooled saturated sodium bicarbonate solution. Oxygen was passed
through the solution at a moderate rate for 15 min. The compound was
extracted with ether and dried over sodium sulfate, and the solvent was
evaporated to provide a yellow-colored liquid, which was purified by
flash chromatography with ethyl acetate and petroleum ether. Yield
1.28 g (58%); 1H NMR (CDCl3): d=2.15 (s, 6H), 3.31
(s, 2H), 3.72 (s,
A
Synthesis of 35: Thiophenol (16 mL, 0.14 mmol) was added at room tem-
perature to the stirred solution of 21 (50 mg, 0.09 mmol) in dichlorome-
thane (5 mL), the reaction mixture was stirred for 30 min, and the solvent
was then evaporated. The expected compound was purified by flash chro-
matography on a silica gel column with ethyl acetate and petroleum
ether as eluents to provide a yellow-colored oil. Yield: 16.7 mg (49%);
1H NMR (CDCl3): d=0.75–0.79 (t, J=7.2 Hz, 6H), 1.41–1.51 (m, 4H),
2.41–2.45 (t, J=8.0 Hz, 4H), 3.65 (s, 2H), 7.02–7.05 (m, 3H), 7.09–7.17
(m, 3H), 7.40–7.42 (d, J=7.6 Hz, 2H), 7.83–7.85 ppm (d, J=7.6 Hz, 1H);
13C NMR (CDCl3): d=12.6, 18.9, 54.2, 60.6, 126.2, 126.3, 128.3, 128.6,
129.2, 135.6, 138.8, 139.6 ppm; 77Se NMR (CDCl3): d=556 ppm; HRMS:
m/z: calcd for C19H25NSSe [M+H]+: 380.0873; found: 379.9544.
3H), 6.71–6.73 (d, J=8 Hz, 1H), 6.79–6.81 (d, J=8 Hz, 2H), 7.12–
7.16 ppm (t, J=8 Hz, 1H); 13C NMR (CDCl3): d=45.4, 55.2, 64.4, 112.8,
114.3, 121.5, 129.2, 140.5, 159.7 ppm; 77Se NMR (CDCl3): d=374 ppm;
HRMS: m/z: calcd for C20H28N2O2Se2 [M+H]+: 489.0559; found:
489.0559.
Synthesis of 23: nBuLi (1.9 mL of a 1.6m solution in hexane) was added
dropwise with stirring at ꢁ58C to a solution of 3-methoxy-N,N-diethyl-
benzylamine (0.50 g, 2.59 mmol) in dry Et2O (15 mL), and the mixture
was allowed slowly to attain room temperature, by which time the color-
less solution had turned yellow. After the reaction mixture had been
stirred for 1.0 h, finely ground selenium powder (0.20 g, 2.59 mmol) was
added at 08C. Soon after the addition of selenium powder, the color
turned brownish. After the reaction mixture had been stirred for 2 h at
room temperature, the mixture was poured into an ice-cooled saturated
sodium bicarbonate solution, and oxygen was passed through the solution
at a moderate rate for 15 min. The compound was extracted with ether
and dried over sodium sulfate. The solvent was evaporated to provide a
yellow colored liquid, which was purified on an active neutral alumina
column with ethyl acetate and petroleum ether as eluents. Yield 0.74 g
(53%); 1H NMR (CDCl3): d=0.93–0.96 (t, J=7.2 Hz, 6H), 2.40–2.45 (q,
J=7.2 Hz, 4H), 3.44 (s, 2H), 3.70 (s, 3H), 6.67–6.69 (d, J=7.2 Hz, 1H),
6.81–6.83 (d, J=6.8 Hz, 1H), 7.09–7.13 ppm (t, J=8 Hz, 1H); 13C NMR
(CDCl3): d=10.7, 45.7, 54.1, 56.5, 111.1, 113.2, 120.1, 128.0, 140.7,
158.5 ppm; 77Se NMR (CDCl3): d=375 ppm; HRMS: m/z: calcd for
C24H36N2O2Se2 [M+H]+: 545.1107; found: 544.9161.
GPx activity—HPLC assay: GPx-like activity was measured by high-per-
formance liquid chromatography (HPLC) with use of a 2695 separation
module and a 2996 photodiode-array detector and a fraction collector.
The assays were performed in sample vials (1.8 mL), and a built-in auto-
sampler was used for sample injection. In this assay, we employed mix-
tures containing a 1:2 molar ratio of PhSH and peroxide in methanol at
room temperature (228C) as our model system. Runs with and without
catalyst were carried out under the same conditions. Periodically, aliquots
were injected onto the reversed-phase column (Lichrosphere 60, RP-se-
lect B, 5 mm) and eluted with methanol and water (85:15), and the con-
centrations of the diphenyl disulfide (PhSSPh) product were determined
at 254 nm with the aid of pure PhSSPh as an external standard. The
amount of disulfide formed during the course of the reaction was calcu-
lated from the calibration plot for the standard (PhSSPh). The plots for
Chem. Eur. J. 2008, 14, 8640 – 8651ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8649

G. Mugesh and K. P. Bhabak
kinetic parameters were obtained by use either of linear or of sigmoidal
curve fitting.
Moussa, M. Parvez, Angew. Chem. 2004, 116, 1288–1290; Angew.
Chem. Int. Ed. 2004, 43, 1268–1270; e) D. Kuzma, M. Parvez, T. G.
Back, Org. Biomol. Chem. 2007, 5, 3213–3217.
Computational methods: All calculations were performed by use of the
Gaussian98 suite[19] of quantum chemical programs. The hybrid Becke 3-
Lee–Yang–Parr (B3LYP) exchange correlation functional was applied
for DFT calculations.[20] Geometries were fully optimized at the B3LYP
level of theory with use of the 6-31+G(d) basis sets. The NMR calcula-
[6] a) S. Benkova, J. Kóòa, G. Gann, W. M. F. Fabian, Int. J. Quantum
Chem. 2002, 90, 555–565; b) S. M. Bachrach, D. W. Demoin, M.
Luk, J. V. Miller, Jr., J. Phys. Chem. A 2004, 108, 4040–4046, and
references therein; c) C. A. Bayse, J. Phys. Chem. A 2007, 111,
9070–9075.
tions were performed at the B3LYP/6-311+G(d,p) level on B3LYP/6-
G
31+G(d) level optimized geometries by the GIAO method.[21] Orbital in-
[7] For more details on Se···X (X=O, N, F, Cl, Br, I, etc.) interactions,
see a) M. Iwaoka, S. Tomoda, J. Am. Chem. Soc. 1996, 118, 8077–
8084; b) W. Nakanishi, S. Hayashi, S. Toyota, Chem. Commun. 1996,
371–372; c) T. Wirth, G. Fragale, M. Spichty, J. Am. Chem. Soc.
1998, 120, 3376–3381; d) W. Nakanishi, S. Hayashi, A. Sakaue, G.
Ono, Y. Kawada, J. Am. Chem. Soc. 1998, 120, 3635–3640; e) H.
Komatsu, M. Iwaoka, S. Tomoda, Chem. Commun. 1999, 205–206;
f) M. Spichty, G. Fragale, T. Wirth, J. Am. Chem. Soc. 2000, 122,
10914–10916; g) M. Iwaoka, H. Komatsu, T. Katsuda, S. Tomoda, J.
Am. Chem. Soc. 2002, 124, 1902–1909; h) M. Iwaoka, H. Komatsu,
T. Katsuda, S. Tomoda, J. Am. Chem. Soc. 2004, 126, 5309–5317;
i) M. Iwaoka, T. Katsuda, H. Komatsu, S. Tomoda, J. Org. Chem.
2005, 70, 321–327; j) C. A. Bayse, R. A. Baker, K. N. Ortwine,
Inorg. Chim. Acta 2005, 358, 3849–3854.
teractions were analyzed by the natural bond orbital (NBO) method at
the B3LYP/6-311+G(d,p) level, and charges were calculated by natural
U
population analysis (NPA).[11] To examine the effect of solvent on the ge-
ometries of the selenol intermediates, single-point energy calculations
were performed in aqueous medium on the B3LYP/6-31+G(d)-level-op-
timized gas-phase geometries by use of the isodensity polarized continu-
um model (IPCM).[22]
Acknowledgement
[8] a) T. G. Back, Z. Moussa, J. Am. Chem. Soc. 2003, 125, 13455–
13460; b) B. K. Sarma, G. Mugesh, J. Am. Chem. Soc. 2005, 127,
11477–11485; c) K. P. Bhabak, G. Mugesh, Chem. Eur. J. 2007, 13,
4594–4601; d) B. K. Sarma, G. Mugesh, Org. Biomol. Chem. 2008, 6,
965–974.
[9] a) T. Wirth, Angew. Chem. 1995, 107, 1872–1873; Angew. Chem. Int.
Ed. Engl. 1995, 34, 1726–1728; b) T. Wirth, G. Fragale, Chem. Eur.
J. 1997, 3, 1894–1902; c) T. Wirth, Liebigs Ann. 1997, 2189–2196;
d) G. Fragale, M. Neuburger, T. Wirth, Chem. Commun. 1998,
1867–1868; e) G. Fragale, T. Wirth, Eur. J. Org. Chem. 1998, 1361–
1369; f) L. Uehlin, G. Fragale, T. Wirth, Chem. Eur. J. 2002, 8,
1125–1133.
[10] It has been reported that the introduction of a nitro group at the po-
sition ortho to the selenium in ebselen can enhance the GPx activity,
mainly through electronic effects (ref. [10a]). Further theoretical in-
vestigations suggested that the increase in the catalytic activity ob-
served upon introduction of an additional group at the position
ortho to the selenium in ebselen may arise from steric factors and
not from an electronic effect (ref. [10b]). Recently, the introduction
of methoxy substituents at the positions para to selenium has been
shown to enhance the GPx-like activities of aromatic cyclic seleni-
nate esters (ref. [10c]). However, the effects of the additional sub-
stituents on the stabilities and reactivities of various catalytically
active intermediates are not clear; a) M. J. Parnham, J. Biederman,
C. Bittner, N. Dereu, S. Leyck, H. Wetzig, Agents Actions 1989, 27,
306–308; b) J. K. Pearson, R. J. Boyd, J. Phys. Chem. A 2008, 112,
1013–1017; c) D. J. Press, E. A. Mercier, D. Kuzma, T. G. Back, J.
Org. Chem. 2008, 73, 4252–4255.
[11] a) A. E. Reed, L. A. Curtiss, F. Weinhold, Chem. Rev. 1988, 88, 899–
926; b) E. D. Glendening, J. E. Reed, J. E. Carpenter, F. Weinhold,
Natural Bond Orbital (NBO), Version 3.1.
[12] The coordination of the nitrogen to selenium can also be suppressed
by using a less nucleophilic tertiary amide group instead of a tertiary
amine. This may lead to an enhancement in the catalytic activity.
However, diselenides possessing tertiary amide substituents were
found to be inactive due to their extremely poor reactivity toward
PhSH. G. Mugesh, K. P. Bhabak, unpublished results.
[13] Although electron-donating substituents such as the methoxy group
are expected to increase the positive charge on selenium, the higher
stability of the methoxy-substituted zwitterion relative to the unsub-
stituted compound is probably due to an increase in the basicity of
the tertiary amino group. Therefore, the amino groups in com-
pounds 26, 28, and 30 are better bases than the amino substituents
in compounds 25, 27, and 29.
[14] The identification of selenenic and seleninic acids is based on litera-
ture data and a number of control experiments. Treatment of the
diselenides with H2O2 (5 equiv) produced the selenenic and seleninic
This study was supported by the Department of Science and Technology
(DST), New Delhi, India. We are grateful to the Alexander von Hum-
boldt Foundation, Bonn, Germany for the donation of an automated
flash chromatography system. G.M. acknowledges the DST for the award
of a Ramanna fellowship, and K.P.B. thanks the Council of Scientific and
Industrial Research (CSIR), India for a research fellowship.
[1] a) L. FlohØ, E. A. Günzler, H. H. Schock, FEBS Lett. 1973, 32, 132–
134; b) J. T. Rotruck, A. L. Pope, H. E. Ganther, A. B. Swanson,
D. G. Hafeman, W. G. Hoekstra, Science, 1973, 179, 588–590; c) O.
Epp, R. Ladenstein, A. Wendel, Eur. J. Biochem. 1983, 133, 51 –69;
d) Selenium in Biology and Human Health (Ed.: R. F. Burk), Spring-
er, New York, 1994; e) G. Mugesh, W.-W. du Mont, Chem. Eur. J.
2001, 7, 1365–1370; f) C. Jacob, G. I. Giles, N. M. Giles, H. Sies,
Angew. Chem. 2003, 115, 4890–4907; Angew. Chem. Int. Ed. 2003,
42, 4742–4758; g) G. Roy, B. K. Sarma, P. P. Phadnis, G. Mugesh, J.
Chem. Sci. 2005, 117, 287–303.
[2] a) H. Sies, Oxidative Stress: Introductory Remarks, in Oxidative
Stress (Ed.: H. Sies), Academic Press, London, 1985, pp. 1 –8; b) H.
Sies, Angew. Chem. 1986, 98, 1061–1075; Angew. Chem. Int. Ed.
Engl. 1986, 25, 1058–1071.
[3] a) A. Müller, E. Cadenas, P. Graf, H. Sies, Biochem. Pharmacol.
1984, 33, 3235–3240; b) A. Wendel, M. Fausel, H. Safayhi, G. Tiegs,
R. Otter, Biochem. Pharmacol. 1984, 33, 3241–3246; c) L. Engman,
D. Stern, I. A. Cotgreave, C.-M. Andersson, J. Am. Chem. Soc. 1992,
114, 9737–9743; d) H. Sies, Free Radical Biol. Med. 1993, 14, 31 3–
323; e) H. Sies, H. Masumoto, Adv. Pharmacol. 1997, 38, 229–246;
f) G. Mugesh, H. B. Singh, Chem. Soc. Rev. 2000, 29, 347–357; g) G.
Mugesh, W.-W. du Mont, H. Sies, Chem. Rev. 2001, 101, 2125–2179.
[4] a) S. R. Wilson, P. A. Zucker, R.-R. C. Huang, A. Spector, J. Am.
Chem. Soc. 1989, 111, 5936–5939; b) M. Iwaoka, S. Tomoda, J. Am.
Chem. Soc. 1994, 116, 2557–2561; c) T. G. Back, B. P. Dyck, J. Am.
Chem. Soc. 1997, 119, 2079–2083; d) T. Wirth, Molecules, 1998, 3,
164–166; e) G. Mugesh, A. Panda, H. B. Singh, N. S. Punekar, R. J.
Butcher, J. Am. Chem. Soc. 2001, 123, 839–850; f) M. L. Jauslin, T.
Wirth, T. Meier, F. Schoumacher, Hum. Mol. Genet. 2002, 11, 3055–
3063; g) S. S. Zade, H. B. Singh, R. J. Butcher, Angew. Chem. 2004,
116, 4613–4615; Angew. Chem. Int. Ed. 2004, 43, 4513–4515; h) Y.
Sun, Y. Mu, S. Ma, P. Gong, G. Yan, J. Liu, J. Shen, G. Luo, Bio-
chim. Biophys. Acta 2005, 1743, 199–204; i) P. P. Phadnis, G.
Mugesh, Org. Biomol. Chem. 2005, 3, 2467–2481.
[5] a) K. Vessman, M. Ekstrçm, M. Berglund, C.-M. Andersson, L.
Engman, J. Org. Chem. 1995, 60, 4461–4467; b) L. Engman, D.
Stern, H. Frisell, K. Vessman, M. Berglund, B. Ek, C.-M. Andersson,
Bioorg. Med. Chem. 1995, 3, 1255–1262; c) T. G. Back, Z. Moussa,
J. Am. Chem. Soc. 2002, 124, 12104–12105; d) T. G. Back, Z.
8650
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8640 – 8651

Glutathione Peroxidase Mimics
FULL PAPER
acids in the cases of 4, 20, and 21, and only the selenenic acids in
the case of 22–24. Treatment of selenols 25–30 with H2O2 produced
identical signals in the 77Se NMR. This precludes the possibility of
the formation of any thiol esters in the reactions. The reactions of
the oxidized species with PhSH are also helpful for assignation of
the signals for the selenenic acid and seleninic acid intermediates.
While the selenenic acids 38, 40, and 42 required only one equiva-
lent of PhSH each to produce the corresponding selenenyl sulfides
quantitatively, higher concentrations of PhSH were required for
complete conversions of the oxidized compounds derived from 4,
20, and 21 into the corresponding selenenyl sulfides. As an example,
the signal due to the selenenic acid 37 disappeared completely upon
addition of one equivalent of PhSH to a mixture containing selenen-
ic acid 37 and seleninic acid 43. The addition of an excess amount of
PhSH to the mixture converted both 37 and 43 into the selenenyl
sulfide 31.
1969, 191, 529–541; c) F. M. Maiorino, R. Brigelius-FlohØ, K. D.
Aumann, A. Roveri, D. Schomburg, L. FlohØ, Methods Enzymol.
1995, 252, 38–48.
[17] H. Shi, S. Liu, M. Miyake, K. Liu, J. Chem. Res. Toxicol. 2006, 19,
655–660.
[18] a) R. Kaur, H. B. Singh, R. P. Patel, J. Chem. Soc. Dalton Trans.
1996, 2719–2726; b) G. Mugesh, H. B. Singh, Acc. Chem. Res. 2002,
35, 226–236.
[19] Gaussian 98, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuse-
ria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Mont-
gomery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam,
A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V.
Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo,
S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K. Mo-
rokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. J. Re-
plogle, A. J. Pople, Gaussian, Inc., Pittsburgh, PA, 1998.
[15] It should be noted that selenenic acids are generally quite unstable
and may not survive in solution for long times. They tend to dispro-
portionate in solution to give the corresponding seleninic acids and
diselenides (3RSeOH!RSeO2H+RSeSeR+H2O). However, no
such disproportionation was observed in the cases of selenenic acids
38, 40, and 42, which were stable enough to be detected by
77Se NMR spectroscopy. This is due to the presence of the methoxy
substituents, which protect the selenium moieties from overoxida-
tion. For structurally characterized stable selenenic acids, see: a) T.
Saiki, K. Goto, R. Okazaki, Angew. Chem. 1997, 109, 2320–2322;
Angew. Chem. Int. Ed. Engl. 1997, 36, 2223–2224; Angew. Chem.
Int. Ed. Engl. 1997, 36, 2223–2224; b) A. Ishii, S. Matsubayashi, T.
Takahashi, J. Nakayama, J. Org. Chem. 1999, 64, 1084–1085; c) K.
Goto, M. Nagahama, T. Mizushima, K. Shimada, T. Kawashima, R.
Okazaki, Org. Lett. 2001, 3, 3569–3572.
[20] a) C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789;
b) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[21] a) M. Bühl, J. Gauss, J. F. Stanton, Chem. Phys. Lett. 1995, 241, 248–
252; b) C. C. Ballard, M. Hada, H. Keneko, H. Nakatsuji, Chem.
Phys. Lett. 1996, 254, 170–178; c) W. Nakanishi, S. Hayashi, J. Phys.
Chem. A 1999, 103, 6074–6081, and references therein; d) C. A.
Bayse, J. Chem. Theory Comput. 2005, 1, 1119–1127.
[22] J. B. Foresman, T. A. Keith, K. B. Wiberg, J. Snoonian, M. J. Frisch,
J. Phys. Chem. 1996, 100, 16098–16104.
[16] a) C. Little, R. Olinescu, K. G. Reid, P. J. OꢁBrien, J. Biol. Chem.
1970, 245, 3632–3636; b) L. FlohØ, I. Brand, Biochim. Biophys. Acta
Received: May 20, 2008
Published online: July 30, 2008
Chem. Eur. J. 2008, 14, 8640 – 8651ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8651