Cyclodextrin-Derived Diorganyl Tellurides as Glutathione Peroxidase Mimics and Inhibitors of Thioredoxin Reductase and Cancer Cell Growth

Article abstract of DOI:10.1021/jm030916r

Water-soluble diorganyl tellurides of the alkyl aryl or dialkyl type were prepared by treatment of mono-6-tosyl-β-cyclodextrin with sodium alkanetellurolates or arenetellurolates or sodium telluride. The novel cyclodextrin-derived organotelluriums were ev

Full text of DOI:10.1021/jm030916r

J . Med. Chem. 2004, 47, 233-239
233
Cyclod extr in -Der ived Dior ga n yl Tellu r id es a s Glu ta th ion e P er oxid a se Mim ics
a n d In h ibitor s of Th ior ed oxin Red u cta se a n d Ca n cer Cell Gr ow th
Michael McNaughton,† Lars Engman,*,† Anne Birmingham,‡ Garth Powis,‡ and Ian A. Cotgreave§
Department of Organic Chemistry, Institute of Chemistry, Uppsala University, Box 599, S-751 24 Uppsala, Sweden,
Arizona Cancer Center, University of Arizona, Tucson, Arizona 85724-5024, and Division of Biochemical Toxicology,
Institute of Environmental Medicine, Karolinska Institute, Box 210, S-171 77 Stockholm, Sweden
Received J une 4, 2003
Water-soluble diorganyl tellurides of the alkyl aryl or dialkyl type were prepared by treatment
of mono-6-tosyl-â-cyclodextrin with sodium alkanetellurolates or arenetellurolates or sodium
telluride. The novel cyclodextrin-derived organotelluriums were evaluated for their capacity
to catalyze the reduction of hydrogen peroxide, tert-butyl hydroperoxide, and cumene hydro-
peroxide in the presence of glutathione, NADPH, and GSSG-reductase (coupled reductase
assay). Cyclodextrins 4d and 4e, carrying 4-(N,N-dimethylamino)phenyltelluro and n-butyl-
telluro groups, respectively, were the most efficient glutathione peroxidase mimics. Reduction
of lipophilic cumene hydroperoxide often proceeded 10-20 times faster than reduction of the
more hydrophilic hydroperoxides, which cannot bind into the hydrophobic interior of the
cyclodextrin. Thus, it seems that the carbohydrate moiety acts as a binding site for the
hydroperoxide substrate. The cyclodextrin derivatives were also evaluated for their capacity
to inhibit thioredoxin reductase/thioredoxin and cancer cell growth in culture. IC50 values for
inhibition of thioredoxin or thioredoxin/thioredoxin reductase were in the submicromolar range
for the best inhibitors (compounds 4d and 5). Two of the compounds (4c and 5) were found to
inhibit the growth of MCF-7 cells in culture with IC50 values in the low micromolar range.
Sch em e 1a
In tr od u ction
Oxidants, many of which are free radicals, are present
in biological systems under normal physiological condi-
tions. To prevent undesired radical-induced damage, the
organism is equipped with several lines of antioxidant
defense. These act either by interception of free radicals
(superoxide dismutases, vitamin E, ascorbate, glu-
tathione, and uric acid) or by destruction of precursors
to free radicals (catalase and glutathione peroxidases).
The selenium-containing glutathione peroxidases1 are
selenoproteins, a class of enzymes of which about a
dozen are known with respect to structure and function
in mammals.2 These enzymes catalyze the reduction of
hydrogen peroxide, fatty acid hydroperoxides, and phos-
pholipid and cholesterol hydroperoxides using glu-
tathione (GSH) and other thiols as stoichiometric re-
ducing agents. The enzyme catalytic site contains a
selenocysteine residue where selenium undergoes redox
cycling (Scheme 1). Enzyme-bound selenol (Enz-SeH)
reduces hydrogen peroxide and other hydroperoxides to
produce the corresponding selenenic acid (Enz-SeOH).
This reacts with 2 equiv of reduced glutathione (GSH),
first to produce a selenosulfide (Enz-SeSG) and finally
to regenerate the active selenol form of the enzyme. The
five varieties of the enzyme presently known serve their
protective function in cells (GSH-Px-1),3 plasma (GSH-
Px-P),4 membranes (PHGPX),5 the gastrointestinal tract
(GSHPx-GI),6 and in the sperm nuclei.7 Considerable
a
GSH ) γ-Glutamyl cysteinyl glycine.
efforts have been made recently to find compounds that
could mimic the properties of the glutathione peroxidase
enzymes. Ebselen (2-phenyl-1,2-benzisoselenazol-3(2H)-
one), ebselen homologues, selenenamides, diselenides,
R-phenylselenoketones, and selenium-containing en-
zymes, antibodies, and cyclodextrins have all been
demonstrated to catalyze the reduction of hydroperox-
ides in the presence of thiols.8-10
These catalysts serve as precursors of selenols, selene-
nic acids, or selenosulfides, and the catalytic mechanism
probably resembles that proposed for the glutathione
peroxidase enzymes (Scheme 1).11 Some time ago,
we12-15 and others16 found that diorganyl tellurides
could also mimic the glutathione peroxidases. However,
the catalytic mechanism is distinctly different from the
three-step cycle proposed for the enzymatic reaction.
Thus, catalysis relies on redox cycling of the heteroatom
between the two oxidation states II and IV (Scheme 2;
oxidation with a hydroperoxide is shown). In principle,
all mild oxidants (such as hydrogen peroxide, organic
* Corresponding author. Tel: +46 18 4713784. Fax: +46 18 471
3818. E-mail: Lars.Engman@kemi.uu.se.
† Uppsala University.
‡ University of Arizona.
§ Karolinska Institute.
10.1021/jm030916r CCC: $27.50 © 2004 American Chemical Society
Published on Web 12/02/2003

234 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
McNaughton et al.
Sch em e 2
modified (eq 1). Thus, tosyl chloride and â-cyclodextrin
were stirred in water for 2 h at ambient temperature,
after which aqueous NaOH was added to complete the
reaction. Ammonium chloride was then added to adjust
the pH to ca. 8. The white precipitate of compound 3
that formed by cooling of the solution (32% isolated
yield) could be used in the following steps without any
further purification.
The desired cyclodextrin-derived diorganyl tellurides
were obtained by addition of preformed (from sodium
borohydride and the appropriate diorganyl ditelluride)
ethanolic solutions of sodium organyltellurolates to
aqueous suspensions of 6-tosyl-â-cyclodextrin (eq 2).
hydroperoxides, hypochlorite, halogens) will convert
diorganyl tellurides to the corresponding tellurium(IV)
species. Reduction is then brought about by mild reduc-
ing agents such as thiols or ascorbate. For studies and
applications of glutathione peroxidase mimics in hydro-
philic environments, water-soluble compounds17 would
be required. In this respect, cyclodextrins are an attrac-
tive class of materials. Carbohydrate-based compounds
of this type are readily available, bind hydrophobic
substrates to their cavities in water solution, and have
two rims of hydroxyl groups that can be used for
anchoring them.18 In the most common pharmaceutical
applications, their role is to enhance the solubility,
stability, and bioavailability of various drug molecules.
However, they may also bind substrates and then either
catalyze their reactions or mimic a step in an enzymatic
catalytic sequence.19
After heating at 60 °C for 24 h under an inert atmos-
phere, addition of water caused precipitation of the
diorganyl tellurides 4a -e (23-62% isolated yields).
Traces of diorganyl ditelluride present in the products
due to air oxidation of the sensitive sodium organyltel-
lurolates was easily removed by washing the product
with dichloromethane. Phenylseleno derivative 4f was
similarly prepared in 40% yield from sodium benzense-
lenolate. This and other â-cyclodextrin-derived organo-
seleniums have previously been prepared in DMF using
a similar protocol.29
We also thought it would be interesting to incorporate
more than one â-cyclodextrin moiety into the diorganyl
telluride scaffold. Sodium telluride was therefore pre-
pared by sodium borohydride reduction of elemental
tellurium.30 When added to a suspension of mono-6-
tosyl-â-cyclodextrin, the water-soluble diorganyl tel-
luride 5 was formed (eq 3). After filtration through
Among chalcogen-containing cyclodextrins, 2,2′-dis-
elenobis(2-deoxy-â-cyclodextrin) (1a ),20 its tellurium
analogue 1b,21 and 6,6′-diselenobis(6-deoxy-â-cyclodex-
trin) (2)22 were recently shown to act as glutathione
peroxidase mimics. Since these materials are dise-
lenides9 or ditellurides,23 they will enter the catalytic
cycle that has been proposed for the enzyme (Scheme
1). We thought it would be interesting to prepare some
tellurium-based cyclodextrins that could mimic the
glutathione peroxidases according to the alternative
mechanism (Scheme 2) and evaluate their antioxidative
properties and some of their biological effects.
Syn th esis. For introducing functional diversity into
â-cyclodextrins, mono-6-tosyl-â-cyclodextrin (3) is one
of the commonly used starting materials.24 Several
methods for its preparation have been published with
everything from poor to moderate reported yields.25-28
In one recent procedure, p-toluenesulfonic anhydride
was used to introduce a tosyl group with a reported yield
of 61%.25 However, in our hands this method was not
very useful for obtaining the desired product. Instead,
the procedure by Petter,26 involving treatment of â-cy-
clodextrin with p-toluenesulfonyl chloride, was slightly
Celite to remove some elemental tellurium, addition of
ethanol caused precipitation of the crude product. The
isolated yield of recrystallized compound 5 (from water/
ethanol) was 54%.
Resu lts a n d Discu ssion
Glu ta th ion e P er oxid a se-lik e P r op er ties. The glu-
tathione peroxidase-like activity of the cyclodextrin
derived diorganyl tellurides 4 and 5 was assessed by
their ability to catalyze the reaction between hydro-

Cyclodextrin-Derived Diorganyl Tellurides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 235
Sch em e 3
3680, and 6240% catalysis, respectively, with hydrogen
peroxide, tert-butyl hydroperoxide and cumene hydro-
peroxide). This result is in line with previous findings
that compounds carrying alkyltelluro moieties are better
catalyst than those carrying aryltelluro groups.10 Among
cyclodextrins 4 carrying aryltelluro moieties, there is a
trend that catalytic efficiency decreases with decreasing
electron density at tellurium. Thus, the following order
of catalytic efficiency was found with hydrogen peroxide
and tert-butyl hydroperoxide as substrates: 4d > 4c ≈
4b > 4a . With cumene hydroperoxide, the phenyltelluro
derivative 4a was slightly more active than the 4-hy-
droxyphenyl derivative 4c. Apparently, nucleophilic
attack of tellurium on the hydroperoxide is facilitated
as the electron density at tellurium increases. The most
notable result from the coupled reductase assay is the
specificity for reduction of cumene hydroperoxide. With
many of the catalysts (4a ,b,e), cumene hydroperoxide
is reduced 10-12 times faster than hydrogen peroxide.
tert-Butyl hydroperoxide is also reduced more rapidly
than hydrogen peroxide, but often only by a factor of
3-4 (catalysts 4a -d ). Some previously studied organo-
tellurium glutathione peroxidase mimics not anchored
to cyclodextrins, bis(4-aminophenyl)telluride (6)12 and
Ta ble 1. Hydroperoxide Decomposing Capacity of
Cyclodextrin-Derived Organotellurium Compounds As
Determined by the Coupled Reductase Assay
5-hydroxy-2,3-dihydrobenzo[b]tellurophene (7),15 were
found to reduce cumene hydroperoxide as well as tert-
butyl hydroperoxide only 3-4 times faster than hydro-
gen peroxide. A similarly modest selectivity for cumene
hydroperoxide was observed for the cyclodextrin-derived
diselenide20 and ditelluride21 glutathione peroxidase
mimics prepared by Luo. Thus, it seems that the
cyclodextrin part of compounds 4 acts as a binding site
for the lipophilic hydroperoxide. The effect was even
more pronounced with catalyst 5, which carries two
cyclodextrin moieties. In the presence of this compound,
cumene hydroperoxide was reduced almost 20 times
faster than hydrogen peroxide (Table 1). It is expected
that cumene hydroperoxide interacts with the cyclodex-
trin in such a way as to bring the HOO group into close
proximity to tellurium at the primary face of the
cyclodextrin.
In h ibition of Th ior ed oxin Red u cta se a n d Ca n -
cer Cell Gr ow th . The thioredoxin systemsa collective
name for thioredoxin (TrX), thioredoxin reductase (TrxR),
and NADPHsis a powerful cellular protein disulfide
reductase system that is present in all living organisms.
TrxR is a selenocysteine-containing dimeric FAD-
containing enzyme that catalyzes the NADPH-depend-
ent reduction of the active site disulfide in oxidized Trx.
The reduced Trx is a hydrogen donor for ribonucleotide
reductasesthe essential enzyme for DNA synthesiss
and a powerful protein reductase with many functions
in cell growth and redox regulation. There are two
mammalian thioredoxin reductases, Trx-1, found in the
cytoplasm and nucleus of cancer cells, and Trx-2, found
in the mitochondria.32,33 Trx-1 protein levels have been
a
The catalyst’s percentage increase of the basal reaction rate
between GSH and hydroperoxide was calculated as rate of NADPH
consumption + 5 mol % catalyst (µM/min)/rate of NADPH con-
sumption + vehicle (µM/min) × 100.
peroxides and glutathione in an aqueous buffer at
physiological pH. The oxidation of glutathione (γ-
glutamyl cysteinyl glycine; GSH) to the corresponding
disulfide (GSSG) was measured indirectly by spectro-
photometrically assessing the oxidation of NADPH in
the presence of glutathione reductase at 340 nm (coupled
reductase assay;31 Scheme 3).
Data (Table 1) for hydrogen peroxide, tert-butyl hy-
droperoxide. and cumene hydroperoxide are expressed
as a catalytic efficiency (percent catalysis), determined
by reference to the respective control rates in the
absence of catalyst, for each hydroperoxide. As expected,
the phenylseleno derivative 4f of â-cyclodextrin did not
show any catalytic activity with the three hydroperox-
ides tested. The problem is that the oxidation of
selenium is too slow. Among organotelluriums 4, the
cyclodextrin 4e carrying a butyltelluro group was the
best catalyst with all three hydroperoxides tried (520,

236 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
McNaughton et al.
Ta ble 2. Thioredoxin Reductase and Cancer Cell Growth
Inhibiting Capacity of Cyclodextrin-Derived Organotellurium
Compounds
exceptions (vide infra; compounds 4c and 4e), the IC50
values (concentrations required to inhibit thionitroben-
zoate formation by 50%) obtained using the two methods
were rather similar. The effect of the compounds on
growth of MCF-7 breast cancer cells in culture was also
studied. This cell line was chosen for its relatively high
level of thioredoxin gene expression. Inhibition data
expressed as IC50 values (concentration required to
inhibit cell growth by 50% of that observed in a control
incubation) are shown in Table 2.
All compounds except for selenium derivative 4f
inhibited TrxR/Trx with IC50 values in the interval 0.5-
10 µM. These results are in line with those from a
previous, comparative study of diaryl chalcogenides
inhibitors that showed that organotellurium compounds
are much better inhibitors than their organoselenium
counterparts.34 Compounds shown to possess good per-
oxide decomposing activity (Table 1) also turned out to
be among the best inhibitors of TrxR/Trx [cyclodextrins
carrying butyltelluro (4e), 4-(dimethylamino)phenyltel-
luro (4d ) and 4-hydroxyphenyltelluro (4c) groups; bis-
cyclodextrin compound 5]. However, some of these
materials (compounds 4c and 4e) were significantly
poorer inhibitors in the TrxR assay. This could indicate
that they are inhibitors of Trx rather than TrxR in the
combined assay.
In the MCF-7 cell culture assay, most of the hydro-
philic cyclodextrin compounds showed only moderate
activity with IC50 values in the range of 5-10 µM. The
failure of several other good, water-soluble (containing
sulfopropyl groups) organotellurium inhibitors of TrxR
to kill MCF-7 cells in culture was also recently observed.
However, two of the compoundss4-hydroxyphenyltel-
luro-derived cyclodextrin 4c and biscyclodextrin 5swere
more potent (IC50 values were 1.6 and 1.4 µM, respec-
tively). Except for compound 5 (which is also a good
inhibitor in the TrxR/Trx and TrxR assays) there is only
a weak correlation between inhibition of the thioredoxin
system and inhibition of cancer cell growth in the
MCF-7 assay. This could be due to differences in the
cellular uptake of the compounds.
In this and previous work we have found that a
variety of organotellurium compounds inhibit the thiore-
doxin system at low micromolar concentrations. How-
ever, the mechanisms of inhibition are, as yet, unclear.
According to a tentative hypothesis previously for-
warded by us, the oxidized organotellurium compound
(presumably a tellurium(IV) oxide) could serve to cross-
link (oxidize) active site selenocysteine residues or thiols
crucial for enzyme activity.35 However, it turned out that
an in situ prepared telluroxide was a poorer inhibitor
of the thioredoxin system than the corresponding tel-
luride. This suggests that tellurium(IV) compounds are
not crucial for the events responsible for enzyme inhibi-
tion by organotelluriums. In addition to two-electron
oxidation, organotelluriums are also known to undergo
one-electron oxidation by various oxidizing agents.
Maybe it is this capacity that makes them interfere with
the thioredoxin system.
a
Inhibition of thioredoxin reductase in the presence of thiore-
doxin and insulin. Inhibition of thioredoxin reductase. c Inhibi-
tion of growth of MCF-7 breast cancer cells in culture.
b
observed to be significantly elevated in several human
primary cancers. Therefore, the TrxR/Trx-1 redox sys-
tem provides an attractive target for cancer drug
development. Studies of organotellurium inhibitors34 of
TrxR indicated that compounds with good antioxidative
capacity, especially hydroperoxide-decomposing ability,
were very efficient.35 The organotellurium compounds
investigated also inhibited the growth of human cancer
cells in culture, but the poor water solubility of these
materials was a restriction during administration to
animals. We therefore thought it would be worthwhile
to test the cyclodextrin-derived diorganyl tellurides 4
and 5 for their capacity to inhibit TrxR and cancer cell
growth in culture.
TrxR activity was measured as the increase in ab-
sorbance at 405 nm that occurs as added dithioni-
trobenzoic acid (DTNB) is reduced by thiols produced
in an incubation containing Trx, TrxR, NADPH, insulin,
and the inhibitor (Table 2). This assay reflects the
combined effects of the inhibition of Trx and TrxR. TrxR
activity was also assessed in incubations of DTNB,
TrxR, NADPH, and inhibitor and measured as the
initial increase in absorbance at 405 nm (Table 2). This
assay reflects the inhibition of TrxR only. With some
Con clu sion s a n d Ou tlook
Novel cyclodextrin-derived diorganyl tellurides of the
alkyl aryl or dialkyl type were found to catalyze the
reduction of hydrogen peroxide, tert-butyl hydroperox-

Cyclodextrin-Derived Diorganyl Tellurides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 237
Aqueous NaOH (2.5 M, 40 mL) was added and the solution
was stirred for 10 min before unreacted p-toluenesulfonyl
chloride was filtered off. Ammonium chloride (11.6 g) was
added to lower the pH to approximately 8. The solution was
cooled and the resultant white precipitate was collected by
filtration. The white powder was washed with acetone and
dried under vacuum to afford 3.55 g (32%) of the title
compound. Spectral data matched that reported in the litera-
ture.26
Typ ica l P r oced u r e. 6-(P h en yltellu r o)-6-d eoxy-â-cyclo-
d extr in (4a ). To a suspension of mono-6-tosyl-â-cyclodextrin
(505 mg, 0.392 mmol) in H2O (3 mL) under an inert atmo-
sphere was added a solution of sodium benzenetellurolate
[generated from diphenylditelluride (193 mg, 0.471 mmol) and
sodium borohydride (40 mg, 1.06 mmol) in EtOH (5 mL)]. The
reaction mixture was heated at 60 °C for 16 h. The mixture
was cooled to room temperature, and H2O (20 mL) and DCM
(40 mL) were added. The aqueous layer, which contained the
product as a suspension, was separated and washed with DCM
(2 × 40 mL). Filtration afforded a white solid which was made
into a slurry with H2O (10 mL). The slurry was filtered,
washed with DCM (40 mL), and dried to afford the title
compound (270 mg, 52%) as a white powder: mp 263-264 °C
(dec); 1H ((CD3)2SO) δ 3.12-3.70 (m, 41H), 4.50 (m, 7H), 4.82
(m, 7H), 5.73 (m, 14H),7.16 (m, 3H), 7.69 (d, 2H, J ) 6.8 Hz);
13C ((CD3)2SO) δ 59.9, 72.0, 72.5, 73.0, 81.5, 102.0, 112.1, 127.3,
129.1, 137.9; MS calcd for (C48H73O34Te)+ m/z 1323.3061, found
m/z 1323.3077.
ide, and cumene hydroperoxide in the presence of
glutathione. Thus, these compounds act as mimics of
the glutathione peroxidase antioxidant enzymes. Reduc-
tion of the most lipophilic hydroperoxide (cumene hy-
droperoxide) proceeded 10-20 times faster than reduc-
tion of the two more hydrophilic ones. It therefore seems
that the hydrophilic cavity provided by the carbohydrate
moiety of the catalyst acts as a binding site for the
hydroperoxide substrate. We feel that compounds of this
type could find use as various types of “antioxidant
pharmacotherapy”, i.e., remedies for pathological condi-
tions characterized by an elevation in the cellular
steady-state concentration of reactive oxygen derived
species (“oxidative stress”). For example, oxidative stress
has been implicated in chronic inflammatory disorders,
adult respiratory distress syndrome, atherosclerosis,
ischemia/reperfusion injury, and cataract.
Like other organotelluriums, the novel cyclodextrin
compounds were also found to interfere with the thiore-
doxin system and to inhibit cancer cell growth. IC50
values for inhibition of thioredoxin or thioredoxin/
thioredoxin reductase were in the submicromolar range
for the most active compounds, and two of them were
found to inhibit the growth of MCF-7 cells in culture
with IC50 values in the low micromolar range. We plan
to test these compounds in other cell lines and study
their effects in vivo.
The following compounds were similarly prepared.
6-(4-Met h oxyp h en ylt ellu r o)-6-d eoxy-â-cyclod ext r in
1
(4b): white powder (37%); mp 268-269 °C (dec); H ((CD3)2-
SO) δ 2.98-3.83 (m, 44H), 4.36-4.55 (m, 7H), 4.80-4.86 (m,
7H), 5.59-5.82 (m, 14H), 6.76 (d, 2H, J ) 8.3 Hz), 7.61 (d,
2H, J ) 8.3 Hz); 13C ((CD3)2SO) δ 55.1, 59.9, 71.9, 72.4, 73.0,
Exp er im en ta l Section
Melting points were recorded on a Stuart Scientific heated
81.5, 101.5, 101.9, 115.1, 140.1, 159.2; MS calcd for (C49H75O35
-
block and are uncorrected. H and 13C spectra were recorded
1
Te)+ m/z 1353.3167, found m/z 1353.3210. Anal. Calcd for
C49H76O35Te‚7H2O: C, 39.79; H, 6.15. Found: C, 39.53; H, 5.89.
6-(4-H yd r oxyp h en ylt ellu r o)-6-d eoxy-â-cyclod ext r in
(4c): white powder (23%); mp 249-250 °C (dec); 1H (CD3OD)
δ 3.47-3.55 (m, 7H), 3.71-3.94 (m, 14H), 4.64 (m, 28 H), 6.66
(d, 2H, J ) 8.4 Hz), 7.61 (d, 2H, J ) 8.4 Hz); 13C (CD3OD) δ
61.9, 73.7, 74.2, 74.9, 82.9, 99.8, 103.8, 117.7, 142.8, 159.2; MS
calcd for (C48H73O35Te)+ m/z 1339.3010, found m/z 1339.3015.
6-(4-N,N-Dim eth yla m in op h en yltellu r o)-6-d eoxy-â-cy-
clod extr in (4d ): white powder (39%); mp 267-268 °C (dec);
1H ((CD3)2SO) δ 2.84 (s, 6H), 3.16-3.62 (m, 41H), 4.31-4.51
(m, 7H), 4.75-4.83 (m, 7H), 5.59-5.82 (m, 14H), 6.53 (d, 2H,
J ) 8.4 Hz), 7.49 (d, 2H, J ) 8.4 Hz); 13C ((CD3)2SO) δ 59.9,
72.0, 72.4, 73.0, 81.5, 101.9, 113.3, 128.8, 140.1, 150.0; MS
calcd for (C50H78O34NTe)+ m/z 1367.3562, found m/z 1367.3591.
6-(Bu tyltellu r o)-6-d eoxy-â-cyclod extr in (4e): white pow-
at 400 and 100 MHz, respectively, on a Varian 400 MHz Unity
instrument. For proton spectra, the residual peak of CHCl3
was used as the internal reference (7.26 ppm), while the
central peak of CDCl3 (77.0 ppm) was used as the reference
for carbon spectra. Mass spectra were recorded using a Bruker
Reflex III MALDI-TOF. Molecular peaks are given for 130Te.
Elemental analyses were carried out by Analytical Laborato-
ries, Lindlar, Germany. Since the cyclodextrin derivatives were
extensively hydrated (compound 4b was analyzed as a hep-
tahydrate), elemental analysis was not a satisfactory proof of
purity. â-Cyclodextrin hydrate, tosyl chloride, and diphenyl
diselenide were purchased from Aldrich and used as supplied.
Diphenylditelluride,37 bis(4-methoxyphenyl)ditelluride,37 bis-
[4-(dimethylamino)telluride,37 and di-n-butylditelluride38 were
prepared according to the literature methods. Spectral data
for 6-(phenylseleno)-6-deoxy-â-cyclodextrin matched those re-
ported in the literature.29
1
der (62%); mp 248-249 °C (dec); H ((CD3)2SO) δ 0.84 (t, 3H,
J ) 8.0 Hz), 1.30 (m, 2H), 1.62 (m, 2H), 2.58 (t, 2H, J ) 7.6
Hz), 3.21-3.61 (m, 41H), 4.40-4.48 (m, 7H), 4.76-4.87 (m,
7H), 5.51-5.79 (m, 14H); 13C ((CD3)2SO) δ 2.1, 12.8, 24.0, 33.4,
59.4, 71.5, 71.9, 72.5, 81.0, 101.4; MS calcd for (C46H77O34Te)+
m/ z 1303.3373, found m/z 1303.3359.
Bis(4-Hyd r oxyp h en yl)d itellu r id e. Under an inert atmo-
sphere at -78 °C, BuLi (9.5 mL, 1.5 M) was added dropwise
t
to 4-bromophenol (1.02 g, 5.90 mmol) in THF (25 mL). The
solution was stirred at -78 °C for 30 min. The solution was
allowed to warm to room temperature over 20 min and then
added to a separate flask containing tellurium (737 mg, 5.78
mmol). The solution was stirred for a further 2 h, after which
all the tellurium had reacted, and the mixture was opened to
air and diethyl ether (100 mL) was added. Saturated aqueous
NH4Cl (80 mL) was added, the organic layer was separated,
and the aqueous layer was extracted with DCM (2 × 50 mL).
The combined organic extracts were dried (MgSO4), filtered
through Celite, and evaporated in vacuo. The product was
isolated as a 1/1.2 mixture of bis(4-hydroxyphenyl)telluride and
the title ditelluride (925 mg), and due to the unstable nature
of the matrial (detelluration), it was used without further
purification: 1H NMR (CDCl3) δ 6.84 (d, 2H, J ) 10.8 Hz),
7.24 (d, 2H, J ) 10.8 Hz).
6,6′-Tellu r o-bis(6-d eoxy-â-cyclod extr in ) (5). To a sus-
pension of mono-6-tosyl-â-cyclodextrin (997 mg, 0.774 mmol)
in H2O (6 mL) was added sodium telluride [generated from
49 mg (0.384 mmol) of tellurium in H2O (1 mL) and 70 mg
(1.85 mmol) of sodium borohydride in EtOH (5 mL)].30 The
reaction mixture was stirred at 60 °C for 16 h. The solution
was filtered through Celite prior to cooling. Water (20 mL) was
added, followed by the addition of ethanol to precipitate the
product. Further recrystallization from water and ethanol
furnished the title compound (486 mg, 54%) as a white
powder: mp 265-266 °C (dec); 1H ((CD3)2SO) δ 3.28-3.70 (m,
82H), 4.31-4.52 (m, 14H), 4.81-4.87 (m, 14H), 5.60-5.81 (m,
28H); 13C ((CD3)2SO) δ 59.9, 72.0, 72.4, 73.0, 81.6, 101.9; MS
calcd for (C84H137O68Te)+ m/z 2363.6352, found m/z 2363.6392.
Cou p led Red u cta se Assa y. The glutathione peroxidase-
like activity of the compounds under study was assessed by
their ability to catalyze the reaction between hydroperoxides
6-O-Mon otosyl-â-cyclod extr in (3). A mixture of p-tolu-
enesulfonyl chloride (2.50 g, 13.1 mmol) and â-cyclodextrin
hydrate (10.0 g, 8.68 mmol) in water (220 mL) was stirred at
ambient temperature for 2 h under an inert atmosphere.

238 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
McNaughton et al.
and glutathione in an aqueous buffer at physiological pH. The
oxidation of GSH to GSSG was measured indirectly by
spectrophotometrically assessing the stimulated oxidation of
NADPH in the presence of glutathione reductase. Incubations
were conducted at room temperature in a Shimadzu Model
160 spectrophotometer recording at 340 nm with air as a
reference. They were constructed in the following manner:
Incubations in quartz cuvettes were with 50 mM potassium
phosphate buffer, pH 7.4. Additions and recordings were made
in the following order (all final concentrations): NADPH (250
µM), GSH (1 mM), test substance (50 µM), record baseline,
GSSG reductase (1 unit), record, hydroperoxide (1 mM), record
the decline in absorbance. Rate assessments were performed
when the decline in absorbance was constant for at least 20 s.
Control experiments revealed that the observed catalytic action
of the compounds was not influenced by increasing amounts
of GSSG reductase (0.5, 1.0, and 2.0 units) in the incubation.
Controls also showed that none of the compounds directly
interacted with the reduction of GSSG (250 and 500 µM) by
the reductase. Shown below are initial rates of NADPH
consumption (SEM (µM/min) for the various combinations of
catalysts and hydroperoxides (including controls). Shown in
Table 1 are the catalytic efficiencies, determined by reference
to the respective control rates in the absence of catalyst, for
each hydroperoxide. Three independent experiments were
performed with each combination of catalyst and hydroper-
oxide.
bromide) was added and an additional 3 h, 37 °C, 5.5% CO2
incubation was performed. At the end of the incubation time,
the untransformed MTT was removed from each well by
aspiration and 150 µL per well of DMSO (dimethyl sulfoxide)
was added. The plate was shaken to ensure full solubilization
of the formazan dye followed by dual optical density readings
of 595 and 655 nm using a multiwell microplate spectropho-
tometer (Molecular Device Corp., Menlo Park, CA). Cytoviabil-
ity of control cells was considered to be 100%. For the treated
cells viability was expressed as a percentage of control cells.
All determinations were carried out in triplicate.
Ack n ow led gm en t. Financial support from Cancer-
fonden, the Swedish Research Council, and the NIH (CA
77204 and CA 17094) is gratefully acknowledged.
Refer en ces
(1) Birringer, M.; Pilawa, S.; Flohe´, L. Trends in Selenium Bio-
chemistry. Nat. Prod. Rep. 2002, 19, 693-718.
(2) Sunde, R. A. In Handbook of Nutritionally Essential Mineral
Elements; O’Dell, B., Sunde, R. A., Eds.; Marcel Dekker: New
York, 1997; Chapter 18.
(3) Rotruck, J . T.; Pope, A. L.; Ganther, H. E.; Swanson, A. B.;
Hafeman, D. G.; Hoekstra, W. G. Selenium. Biochemical Role
as a Component of Glutathione Peroxidase. Science 1973, 179,
588-590.
(4) Takahashi, K.; Avissar, N.; Whitin, J .; Cohen, H. J . Purification
and Characterization of Human Plasma Glutathione Peroxi-
dase: A Selenoglycoprotein Distinct from the Known Cellular
Enzyme. Arch. Biochem. Biophys. 1987, 256, 677-686.
consumption (µM/min)
(5) Ursini, F.; Maiorino, M.; Gregolin,
C . The Selenoenzyme
cumene
Phospholipid Hydroperoxide Glutathione Peroxidase. Biochim.
Biophys. Acta 1985, 839, 62-70.
(6) Chu, F.-F., Doroshow, J . H.; Esworthy, R. S. Expression,
Characterization, and Tissue Distribution of a New Cellular
Selenium-dependent Glutathione Peroxidase, GSHPx-GI. J . Biol.
Chem. 1993, 268, 2571-2576.
(7) Pfeifer, H.; Conrad, M.; Roethlein, D.; Kyriakopoulos, A.; Bri-
elmeier, M.; Bornkamm, G. W.; Behne, D. Identification of a
Specific Sperm Nuclei Selenoenzyme Necessary for Protamine
Thiol Cross-linking During Sperm Maturation. FASEB J . 2001,
15, 1236-1238.
(8) Mugesh, G.; Singh, H. B. Synthetic Organoselenium Compunds
as Antioxidants: Glutathione Peroxidase Activity. Chem. Soc.
Rev. 2000, 29, 347-357. Mugesh, G.; du Mont, W.-W.; Sies, H.
Chemistry of Biologically Important Synthetic Organoselenium
Compounds. Chem. Rev. 2001, 101, 2125-2179. See also: Er-
delmeier, I.; Tailhan-Lomont, C.; Yadan, J .-C. Copper(I)-Assisted
Mild and Convenient Synthesis of New Se-N Heterocycles:
Access to a Promising Class of GPx Mimics. J . Org. Chem. 2000,
65, 8152-8157.
H2O2
NADPH
t-BuOOH
NADPH
hydroperoxide
NADPH
catalyst
control
4a
4b
4c
4d
4e
4f
5
8 ( 0.4
13.6 ( 0.8
18.6 ( 0.8
17.8 ( 1.6
26.5 ( 0.7
41.6 ( 3.6
8.1 ( 0.4
2.0 ( 0.1
11.2 ( 0.3
12.8 ( 0.5
12.8 ( 0.8
24.8 ( 0.8
73.6 ( 6.2
2.4 ( 0.4
2.7 ( 0.2
56.0 ( 1.6
60.0 ( 4.0
52.0 ( 4.0
64.0 ( 3.2
168.5 ( 21.3
2.8 ( 0.2
23.2 ( 0.8
85.6 ( 4.0
150.4 ( 8.2
Tr xR/Tr x Assa y. Thioredoxin reductase/thioredoxin-de-
pendent insulin reducing activity was measured in an incuba-
tion with a final volume of 60 µL containing 100 mM HEPES
buffer, pH 7.2, 5 mM EDTA (HE buffer), 1 mM NADPH, 1.0
µM thioredoxin reductase, 0.8 µM thioredoxin, 2.5 mg/mL
bovine insulin, and inhibitor. Incubations were for 30 min at
37°C in flat-bottom 96-well microtiter plates. The reaction was
stopped by the addition of 50 µL of 6 M guanidine HCl, 50
mM Tris, pH 8.0, and 10 mM DTNB, and the absorbance
measured at 412 nM.
Tr xR Assa y. Assays of TrxR were carried out in flat-bottom
96-well microtiter plates. TrxR activity was measured in a final
incubation volume of 60 µL containing HE buffer, 10 mM
DTNB, 1.0 µM TrxR, and 1 mM NADPH and inhibitor.
Compounds were diluted in HE buffer and added to the wells
as 20 µL aliquots, and TrxR was then added, also as 20 µL
aliquots in HE buffer. To ensure uniform coverage of the
bottom of the well the plate was spun briefly at 3000g. To start
the reaction, NADPH and DTNB were added as a 20 µL aliquot
in HE buffer and the plate was moved to the plate reader
which had been preheated to 37 °C. The optical density at 412
nm was measured every 10 s and initial linear reaction rates
measured.
Gr ow th In h ibition Assa y. Compound cytotoxicity was
measured using modifications of the MTT assay as described
by Mosmann39 and Carmichael.40 Human MCF-7 breast cancer
cells were seeded at 3000 cells/well into 96-well plates in
Dulbecco’s minimum essential medium (DMEM) supplemented
with 10% fetal bovine serum (FBS). After 16 h at 37 °C and
5.5% CO2, drugs were added to the wells at concentrations
ranging from 0.1 to 20 µM. The cells were further incubated
for 72 h, after which 40 µL per well of a 2.5 µg/µL solution of
MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium
(9) Mugesh, G.; Panda, A.; Singh, H. B.; Punekar, N. S.; Butcher,
R. J . Glutathione Peroxidase-like Antioxidant Activity of Diaryl
Diselenides: A Mechanistic Study J . Am. Chem. Soc. 2001, 123,
839-850.
(10) Kanda, T.; Engman, L.; Cotgreave, I. A.; Powis, G. Novel Water-
Soluble Diorganyl Tellurides with Thiol Peroxidase and Anti-
oxidant Activity. J . Org. Chem. 1999, 64, 8161-8169 and refs
cited therein.
(11) Ganther, H. E.; Kraus, R. J . Oxidation States of Glutathione
Peroxidase. Meth. Enzymol. 1984, 107, 593-602.
(12) Andersson, C.-M.; Hallberg, A.; Brattsand, R.; Cotgreave, I. A.;
Engman, L.; Persson, J . Glutathione Peroxidase-like Activity of
Diaryl Tellurides. Bioorg. Med. Chem. Lett. 1993, 3, 2553-2558.
(13) Engman, L.; Stern, D.; Pelcman, M. Andersson, C-M. Thiol
Peroxidase Activity of Diorganyl Tellurides. J . Org. Chem. 1994,
59, 1973-1979.
(14) Vessman, K.; Ekstro¨m, M.; Berglund, M.; Andersson, C.-M.,
Engman L. Catalytic Antioxidant Activity of Diaryl Tellurides
in a Two-Phase Lipid Peroxidation Model. J . Org. Chem. 1995,
60, 4461-4467.
(15) Malmstro¨m, J .; J onsson, M.; Cotgreave, I. A.; Hammarstro¨m,
L.; Sjo¨din, M.; Engman, L. The Antioxidant Profile of 2,3-
Dihydrobenzo[b]furan-5-ol and Its 1-Thio, 1-Seleno, and 1-Tel-
luro Analogues. J . Am. Chem. Soc. 2001, 123, 3434-3440.
(16) Detty, M. R.; Gibson, S. L. Tellurapyrylium Dyes as Catalysts
for Oxidations with Hydrogen Peroxide and as Scavengers of
Singlet Oxygen. Dihydroxytelluranes as Mild Oxidizing Agents.
Organometallics 1992, 11, 2147-2156. Detty, M. R.; Friedman,
A. E.; Oseroff, A. R. A Mechanism for the Oxidation of Glu-
tathione to Glutathione Disulfide with Organotellurium(IV) and
Organoselenium(IV) Compounds. A Stepwise Process with Im-
plications for Photodynamic Therapy and Other Oxidative
Chemotherapy. J . Org. Chem. 1994, 59, 8245-8250.

Cyclodextrin-Derived Diorganyl Tellurides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 239
(17) For an overview of types and classes of organotellurium com-
pounds available, see: Irgolic, K. J . In Houben-Weyl. Methods
of Organic Chemistry. Organotellurium Compounds; Klamann,
D., Ed.; Thieme: Stuttgart, 1990; Vol. E 12b.
(18) For an overview of cyclodextrin chemistry, see: Chem. Rev. 1998,
98, issue 5.
(19) Breslow, R.; Dong, S. D. Biomimetic Reactions Catalyzed by
Cyclodextrins and Their Derivatives. Chem. Rev. 1998, 98,
1997-2011.
(20) Liu, J .; Luo, G.; Ren, X.; Mu, Y.; Bai, Y.; Shen, J . A Bis-
Cyclodextrin Diselenide with Glutathione Peroxidase-like Activ-
ity. Biochim. Biophys. Acta 2000, 1481, 222-228.
(21) Ren, X.; Xue, Y.; Zhang, K.; Liu, J .; Luo, G.; Zheng, J .; Mu, Y.;
Shen, J . A Novel Dicyclodextrinyl Ditelluride Compound With
Antioxidant Activity. FEBS Lett. 2001, 507, 377-380. Es-
sentially the same paper also appeared elsewhere: Ren, X.; Xue,
Y.; Liu, J .; Zhang, K.; Zheng, J .; Luo, G.; Guo, C.; Mu, Y.; Shen,
J . A Novel Cyclodextrin-Derived Tellurium Compound with
Glutathione Peroxidase Activity. ChemBioChem 2002, 3, 356-
363.
(22) Liu, J .; Gao, S.-J .; Luo, G.; Yan, G.-L.; Shen, J . Artificial
Imitation of Glutathione Peroxidase With 6-Selenium-bridged
â-Cyclodextrin. Biochem. Biophys. Res. Commun. 1998, 247,
397-400.
(23) Engman, L.; Stern, D.; Cotgreave, I. A.; Andersson, C. M. Thiol
Peroxidase Activity of Diaryl Ditellurides As Determined by a
1H NMR Method. J . Am. Chem. Soc. 1992, 114, 9737-9743.
Mugesh, G.; Panda, A.; Kumar, S.; Apte, S. D.; Singh, H. B.;
Butcher, R. J . Intramolecularly Coordinated Diorganyl Ditellu-
rides: Thiol Peroxidase-like Antioxidants. Organometallics 2002,
21, 884-892.
(24) For example see: La Mendola, D.; Sortino, S.; Vecchio, G.;
Rizzarelli, E. Synthesis of new carnosine derivatives of â-cyclo-
dextrin and their hydroxyl radical scavenger ability. Helvetica
Chimica Acta 2002, 85, 1633-1643; Filippone, S.; Heimann, F.;
Rassat, A. A Highly Water-soluble 2:1 â-Cyclodextrin-fullerene
Conjugate. Chem. Commun. 2002, 1508-1509.
(25) Zhong, N.; Byun, H.-S.; Bittman, R. An Improved Synthesis of
6-O-Monotosyl-6-deoxy-â-cyclodextrin. Tetrahedron Lett. 1998,
39, 2919-2920.
(26) Petter, R. C.; Salek, J . S.; Sikorski, C. T.; Kumaravel, G.;
Lin, F.-T. Cooperative binding by aggregated mono-6-(alkyl-
amino)-â-cyclodextrins. J . Am. Chem. Soc. 1990, 112, 3860-
3868.
(27) Matsui, Y.; Yokoi, T.; Mochida, K. Catalytic Properties of a Cu(II)
Complex With a Modified Cyclodextrin. Chem. Lett. 1976, 1037-
1040.
(28) Baussanne, I.; Benito, J . M.; Ortiz Mellet, C.; Garc´ıa Ferna´ndez,
J . M.; Law, H.; Defaye, Synthesis and Comparative Lectin-
binding Affinity of Mannosyl-Coated â-Cyclodextrin-dendrimer
Constructs. Chem. Commun. 2000, 1489-1490.
(29) Liu, Y.; Li, B.; Wada, T.; Inoue, Y. Enantioselective Recognition
of Aliphatic Amino Acids by Organoselenium Modified â-Cyclo-
dextrins. Supramol. Chem. 1999, 10, 173-184.
(30) Bergman, J .; Engman, L. Preparation of Selena- and Tellura
Phthalic Anhydride. Org. Prep. Proc. Int. 1978, 10, 289-290.
(31) Cotgreave, I. A.; Molde´us, P.; Brattsand, R.; Hallberg, A.;
Andersson, C.-M.; Engman, L. R-(Phenylselenenyl)acetophenone
Derivatives With Glutathione Peroxidase-like Activity. Biochem.
Pharmacol. 1992, 43, 793-802 and references therein.
(32) Grogan, T. M.; Fenoglio-Prieser, C.; Zeheb, R.; Bellamy, W.;
Frutiger, Y.; Vela, E.; Stemmerman, G.; MacDonald, J .; Richter,
L.; Gallegos, A.; Powis, G. Thioredoxin, a Putative Oncogene
Product, is Overexpressed in Gastric Carcinoma and Associated
With Increased Proliferation and Increased Cell Survival. Hu-
man Pathol. 2000, 31, 475-481.
(33) Spyrou, G.; Enmark, E.; Miranda-Vizuete, A.; Gustafsson, J .
Cloning and expression of a novel mammalian thioredoxin. J .
Biol. Chem. 1997, 272, 2936-2941.
(34) Engman, L.; Cotgreave, I.; Angulo, M.; Powis, G. Diaryl Chal-
cogenides as Selective Inhibitors of Thioredoxin Reductase and
Potential Antitumor Agents. Anticancer Res. 1997, 17, 4599-4606.
(35) Engman, L.; Al-Maharik, N.; McNaughton, M.; Birmingham, A.;
Powis, G. Thioredoxin Reductase and Cancer Cell Growth
Inhibition by Organotellurium Antioxidants. Anti-Cancer Drugs
2003, 14, 153-161.
(36) Engman, L.; Kanda, T.; Gallegos, A.; Williams, R.; Powis, G.
Water-soluble Organotellurium Compounds Inhibit Thioredoxin
Reductase and the Growth of Human Cancer Cells. Anti-Cancer
Drug Des. 2001, 15, 303-306.
(37) Engman, L.; Persson, J . Improved Preparation of Diaryl Ditel-
lurides. J . Organomet. Chem. 1990, 388, 71-74.
(38) Engman, L.; Cava, M. Organotellurium Compounds 5. A Con-
venient Synthesis of Some Aliphatic Ditellurides. Synth. Com-
mun. 1982, 12, 163-165.
(39) Mossman, T. Rapid Colorimetric Assay for Cellular Growth and
Survival: Application to Proliferation and Cytotoxicity Assays
J . Immunol. Methods 1983, 65, 55-63.
(40) Carmichael, J .; DeGraff, W. G.; Gazdar, A. F.; Minna, J . D.;
Mitchell, J . B. Evaluation of a Tetrazolium-based Semiauto-
mated Colorimetric Assay: Assessment of Chemosensitivity
Testing. Cancer Res. 1987, 47, 936-942.
J M030916R