Antitumor quinol PMX464 is a cytocidal anti-trypanosomal inhibitor targeting trypanothione metabolism

Article abstract of DOI:10.1074/jbc.M110.214833

Better drugs are urgently needed for the treatment of African sleeping sickness. We tested a series of promising anticancer agents belonging to the 4-substituted 4-hydroxycyclohexa-2,5-dienones class ("quinols") and identified several with potent trypanocidal activity (EC₅₀ < 100 nM). In mammalian cells, quinols are proposed to inhibit the thioredoxin/thioredoxin reductase system, which is absent from trypanosomes. Studies with the prototypical 4-benzothiazole-substituted quinol, PMX464, established that PMX464 is rapidly cytocidal, similar to the arsenical drug, melarsen oxide. Cell lysis by PMX464 was accelerated by addition of sublethal concentrations of glucose oxidase implicating oxidant defenses in the mechanism of action. Whole cells treated with PMX464 showed a loss of trypanothione (T(SH)₂), a unique dithiol in trypanosomes, and tryparedoxin peroxidase (TryP), a 2-Cys peroxiredoxin similar to mammalian thioredoxin peroxidase. Enzyme assays revealed that T(SH)₂, TryP, and a glutathione peroxidase-like tryparedoxin-dependent peroxidase were inhibited in time- and concentration-dependent manners. The inhibitory activities of various quinol analogues against these targets showed a good correlation with growth inhibition of Trypanosoma brucei. The monothiols glutathione and L-cysteine bound in a 2:1 ratio with PMX464 with K_d values of 6 and 27 μM, respectively, whereas T(SH)₂ bound more tightly in a 1:1 ratio with a K_d value of 430 nM. Overexpression of trypanothione synthetase in T. brucei decreased sensitivity to PMX464 indicating that the key metabolite T(SH)₂ is a target for quinols. Thus, the quinol pharmacophore represents a novel lead structure for the development of a new drug against African sleeping sickness.

Full text of DOI:10.1074/jbc.M110.214833

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 286, NO. 10, pp. 8523–8533, March 11, 2011
© 2011 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.
Author’s Choice
Antitumor Quinol PMX464 Is a Cytocidal Anti-trypanosomal
Inhibitor Targeting Trypanothione Metabolism*
Received for publication, December 21, 2010 Published, JBC Papers in Press, January 6, 2011, DOI 10.1074/jbc.M110.214833
Janine Ko¨nig‡1, Susan Wyllie‡, Geoffrey Wells§2, Malcolm F. Stevens§, Paul G. Wyatt‡, and Alan H. Fairlamb‡3
From the ‡Division of Biological Chemistry and Drug Discovery, College of Life Sciences, University of Dundee, Dundee DD1 5EH
and the §Centre for Biomolecular Sciences, School of Pharmacy, University of Nottingham, Nottingham NG7 2RD, United Kingdom
Better drugs are urgently needed for the treatment of African
sleeping sickness. We tested a series of promising anticancer
agents belonging to the 4-substituted 4-hydroxycyclohexa-2,5-
dienones class (“quinols”) and identified several with potent
trypanocidal activity (EC50 < 100 nM). In mammalian cells,
quinols are proposed to inhibit the thioredoxin/thioredoxin
reductase system, which is absent from trypanosomes. Studies
with the prototypical 4-benzothiazole-substituted quinol,
PMX464, established that PMX464 is rapidly cytocidal, similar
to the arsenical drug, melarsen oxide. Cell lysis by PMX464 was
accelerated by addition of sublethal concentrations of glucose
oxidase implicating oxidant defenses in the mechanism of
action. Whole cells treated with PMX464 showed a loss of
trypanothione (T(SH)2), a unique dithiol in trypanosomes,
and tryparedoxin peroxidase (TryP), a 2-Cys peroxiredoxin
similar to mammalian thioredoxin peroxidase. Enzyme
assays revealed that T(SH)2, TryP, and a glutathione peroxi-
dase-like tryparedoxin-dependent peroxidase were inhibited
in time- and concentration-dependent manners. The inhibitory
activities of various quinol analogues against these targets
showed a good correlation with growth inhibition of Trypano-
soma brucei. The monothiols glutathione and L-cysteine bound
in a 2:1 ratio with PMX464 with Kd values of 6 and 27 ␮M, respec-
tively, whereas T(SH)2 bound more tightly in a 1:1 ratio with a Kd
value of 430 nM. Overexpression of trypanothione synthetase in
T. brucei decreased sensitivity to PMX464 indicating that the
key metabolite T(SH)2 is a target for quinols. Thus, the quinol
pharmacophore represents a novel lead structure for the devel-
opment of a new drug against African sleeping sickness.
eflornithine) are toxic or difficult to administer (2) and treat-
ment failures, possibly due to parasite resistance, are increasing
(3, 4). New drugs are urgently needed.
One unique metabolic feature of these parasites involves
trypanothione (N1,N8-bis(glutathionyl)spermidine), a dithiol
that plays a pivotal role in thiol-redox metabolism (5–7).
Trypanothione (T(SH)2)4 has replaced many of the functions of
glutathione in other organisms and, together with trypano-
thione reductase (TryR) and tryparedoxin (TryX), subsumes
the role of thioredoxin reductase and thioredoxin in mamma-
lian systems in providing reducing equivalents for ribonucle-
otide reductase (8). Trypanothione is the physiologically rele-
vant source of reducing equivalents for the antioxidant defense
systems of the parasite (9–11). All of its components are well
characterized and involve the following electron cascade:
NADPH, TryR (12, 13), T(SH)2 (6), TryX (14, 15), and two types
of tryparedoxin-dependent peroxidases: the 2-Cys peroxire-
doxins (TryP) (15–18) and the glutathione peroxidase-like pro-
teins (TDPX) (19–22). These pathways are crucial for the sur-
vival of the parasites, as shown by the essentiality of all the
above components (13, 22–24). Furthermore, T(SH)2 and TryR
are possible targets for arsenical (25, 26) and antimonial drugs
(27, 28) currently used to treat human African trypanosomiasis
and leishmaniasis, respectively.
Recently, a new class of aromatic 4-substituted 4-hydroxycy-
clohexa-2,5-dienones (“quinols”), including the 4-benzothia-
zole-substituted quinol (PMX464), were found to have potent
growth inhibitory activity against colon, renal, and certain
breast cancer cell lines (29, 30), as well as anti-proliferative
activity against human umbilical vein endothelial cells (31).
Although the precise molecular mechanism of these promising
antitumor agents remains elusive, thioredoxin (29) and thiore-
doxin reductase (32) have been proposed as possible targets.
Like trivalent arsenical and antimonial drugs, quinols are thiol-
reactive compounds that have been proposed to bind to vicinal
protein dithiols, such as those found in thioredoxin, via an
essentially irreversible, double Michael-addition reaction (29,
32). Others have suggested that quinols preferentially target
mammalian thioredoxin reductase instead of thioredoxin via
addition to the C-terminal selenocysteine residue involved in
Human African trypanosomiasis, or sleeping sickness, is
caused by the protozoan parasites Trypanosoma brucei gam-
biense and T. brucei rhodesiense. This vector-born disease is
transmitted by the tsetse fly in sub-Saharan Africa with an esti-
mated incidence of 70,000–80,000 cases and ϳ30,000 deaths
per annum (1). Once the infection has spread to the central
nervous system, the disease is invariably fatal without treat-
ment. The drugs in use for late stage disease (melarsoprol and
* This work was supported by Wellcome Trust Grants WT 079838, WT 077705, reduction of thioredoxin disulfide (32).
and WT 083481.
Author’s Choice—Final version full access.
1 Present address: Biochemistry and Plant Physiology, Faculty of Biology, 4 The abbreviations used are: T(SH)2, trypanothione (N1,N8-bis(glutathionyl)-
Bielefeld University, 33615 Bielefeld, Germany.
spermidine; quinols, 4-substituted 4-hydroxycyclohexa-2,5-dienones;
TryP, type I (decameric) tryparedoxin peroxidase; TDPX, type II (mono-
meric) tryparedoxin-dependent peroxidase; TryR, trypanothione reduc-
tase; TryX, tryparedoxin; VSG, variant surface glycoprotein; DTNB, 5,5Ј-di-
thiobis(2-nitrobenzoic acid); DMSO, dimethyl sulfoxide.
2 Present address: Dept. of Pharmaceutical and Biological Chemistry, School
of Pharmacy, 29–39 Brunswick Square, London, WC1N 1AX, U.K.
3 To whom correspondence should be addressed. Fax: 44-1382-385542;
E-mail: a.h.fairlamb@dundee.ac.uk.
MARCH 11, 2011•VOLUME 286•NUMBER 10
JOURNAL OF BIOLOGICAL CHEMISTRY 8523

Quinols Inhibit the Trypanothione Peroxidase System
Given the unique and essential features of thiol metabolism specific for TbTDPX2, TbTryX, LmTryP1, TbTryR, and
in T. brucei and the proposed mode of action of quinols against TbVSG221.
cancer cells, we have found that PMX464 and analogues are
Analysis of Intracellular Thiols—To determine the effect of
also potent and moderately selective trypanocidal agents in PMX464 on intracellular thiol levels, exponentially growing
vitro. Here we have investigated the mechanism of cytotoxicity cultures containing 1 ϫ 106 cells mlϪ1 were incubated with 250
in detail and provide evidence that T(SH)2 and TryP are pri- nM PMX464 for 16 h. Cells (1 ϫ 108) were collected by centri-
mary molecular targets for these experimental compounds. fugation (800 ϫ g, 10 min, 4 °C), washed once in ice-cold PSG
The quinol pharmacophore could be a valuable lead structure buffer (phosphate-buffered saline, pH 8.0, 1.5% (w/v) glucose,
Ϫ1
for the development of a new drug against African sleeping and 0.5 mg ml BSA) and cell pellets were derivatized with
sickness.
monobromobimane (6, 37). Acid-soluble thiols were separated
by ion-paired, reverse-phase HPLC on an ion-paired Ultra-
sphere C18 column using a Dionex Ultimate 3000 instrument
EXPERIMENTAL PROCEDURES
Materials—All chemicals were of the highest grade available fitted with a Dionex RF-2000 fluorometer.
from Sigma or BDH Chemicals. Ltd. (Poole, UK). Quinol com-
Expression and Purification of Recombinant Proteins—Re-
pounds 1–7 were kindly provided by Pharminox Ltd., Biocity, combinant TbTDPX2 (19) and TbTryP (15) were expressed and
Nottingham, UK. Stock solutions (10 mM) were prepared in purified as described previously. TbTryX, cloned into pET-15b
dimethyl sulfoxide (DMSO) and stored for short term at 4 °C (15) and transformed into BL21(DE3) pLysS, was expressed
and for longer terms at Ϫ20 °C. Antiserum against T. brucei using the autoinduction system (38). Briefly, autoinduction
variant surface glycoprotein 221 (VSG221) was kindly provided medium (1 liter) was inoculated with a 10-ml overnight culture
by Lucia Guther (College of Life Sciences, Dundee, UK) and (LB medium) and grown at 37 °C until the A595 nm reached 0.6,
recombinant T. brucei TryR by Magnus Alphey (College of Life followed by a 72-h incubation at 22 °C. TbTryX was purified
Sciences, Dundee, UK).
essentially as published for LmTryX (36).
Cell Culture and Cell Potency Assays—Bloodstream T. brucei
Inhibition Assays—The effects of PMX464 on components of
S427 were cultured at 37 °C in modified HMI9 medium (56 ␮M the T. brucei antioxidant defense system were analyzed in vitro
thioglycerol was substituted for 200 ␮M 2-mercaptoethanol) using standard TDPX or TryP peroxidase assays. NADPH con-
(33). MRC-5 cells were maintained in DMEM containing 10% sumption was determined at 25 °C in a final volume of 200 ␮l
fetal calf serum. Trypanosomes overexpressing trypanothione essentially as described previously (19, 36). The standard assay
synthetase (34) were cultured in an identical manner. Effective mixture contained 100 mM Tris-HCl, pH 7.6, 5 mM EDTA, 170
concentrations of test compounds inhibiting growth by 50% ␮M NADPH, 1 unit mlϪ1 TbTryR, 20 ␮M T(S)2, 5 ␮M TryX and
(EC50) against parasites and MRC-5 cells were determined in either 0.25 ␮M TDPX or 0.25 ␮M TryP. PMX464 (100 ␮M) was
96-well plates using a modification of the Alamar Blue cell via- preincubated with the assay mixture before the reaction was
bility assay (35).
initiated with 300 ␮M cumene hydroperoxide (for TDPX) or
Viability Studies—Mid-log trypanosomes were diluted to 150 ␮M H2O2 (for TryP). The final DMSO concentration in all
1 ϫ 106 cells mlϪ1 in pre-warmed medium containing ϳ5 times assays was 1%.
the EC50 for melarsen oxide (50 nM) and PMX464 (500 nM) or
To determine the sensitivity of individual components of the
DMSO (0.1%) as a control. Aliquots were taken at different time peroxidase systems to quinols, the concentrations of T(S)2,
points and parasite density was determined using a CASY Cell TDPX, or TryP were varied in the assay to obtain conditions
Counter (Scha¨rfe Systems, Reutlingen, Germany). In addition, under which NADPH consumption was proportional to only
the number of motile parasites was determined microscopically one of these components. For analysis of TDPX as the rate-
using a Neubauer hemocytometer. Studies examining potential limiting component, the assay contained 0.25 ␮M TbTDPX2
synergistic effects between PMX464 and the H2O2 generator and a trace amount of T(S)2 (0.2 ␮M to reduce the TryX and
glucose oxidase (from Aspergillus niger, Fluka) were carried out TDPX). This mixture was preincubated for different times with
in an identical manner with mid-log trypanosomes incubated various concentrations of quinol before the reaction was initi-
with either PMX464 (500 nM), 0.1 milliunits mlϪ1 glucose oxi- ated with 300 ␮M cumene hydroperoxide plus 20 ␮M T(S)2.
dase, or a combination of both.
Similarly, for TryP, the assay contained 0.25 ␮M TbTryP and 0.2
Immunoblotting—Exponentially growing T. brucei (1 ϫ 106 ␮M T(S)2 and the reaction was initiated with 150 ␮M H2O2 and
cells mlϪ1) were treated with varying concentrations of 20 ␮M T(S)2. For assays that were rate-limiting for T(SH)2,
PMX464 (0 to 300 nM) and cultured for 16 h. Parasites were the final concentration of T(S)2 was 1 ␮M and the reaction
pelleted (800 ϫ g for 10 min) and washed twice by resuspension was initiated with 300 ␮M cumene hydroperoxide plus 2 ␮M
in 5 mM KCl, 80 mM NaCl, 1 mM MgSO4, 20 mM Na2HPO4, 2 TbTDPX2. The initial rate of NADPH consumption was fol-
mM NaH2PO4, 20 mM glucose, pH 7.4, and centrifugation as lowed by the decrease of absorbance at 340 nm using a UV-
above. Whole cell lysates were prepared by resuspension of cell visible Shimadzu spectrophotometer. In this way only one assay
pellets in 50 mM Tris, pH 7.4, 9 M urea, and 0.1% Triton-X-100 component (TDPX, TryP, or T(SH)2) was inhibited by the
followed by heating at 95 °C for 5 min. Protein concentrations quinol due to the slow-binding behavior of these inhibitors.
were determined by the Bradford protein assay (Bio-Rad)
The inhibitory concentration causing 50% inhibition (IC50)
using BSA as standard. Protein samples (20 ␮g/lane) were of compounds 1–7 were determined in a similar manner using
separated by SDS-PAGE (12% NuPAGE, Invitrogen) and a 96-well plate format at 27 °C monitoring NADPH consump-
analyzed by Western blotting (36) using rat antiserum tion at 340 nm using a SPECTRAmax 340PC plate reader and
8524 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 286•NUMBER 10•MARCH 11, 2011

Quinols Inhibit the Trypanothione Peroxidase System
SoftMax Pro software. The residual activities were calculated as above and the rate of complex formation (R) was calculated
a percentage of the DMSO-treated control assay. Analyses of using the ␦ absorption coefficient of 1,400 MϪ1 cmϪ1 per bound
IC50 values were performed using the standard rate-limiting thiol. The initial rate R of the reaction: mA ϩ nB 3 products, is
peroxidase assays. PMX464 was diluted in DMSO, through a given by,
10-point, doubling dilution curve with 100 ␮M as the highest
m
n
R ϭ k͓A͔ ͓B͔
(Eq. 3)
0
0
compound concentration. The components of the assays for
T(SH)2, TDPX, and TryP inhibition analysis were preincubated
with varied amounts of PMX464 for 30 or 90 min and the initial
rates were determined. Initial rates were corrected for back-
ground before fitting to an equation describing a two parameter
fit,
where m is the order with respect to A, n is the order with
respect to B, and m ϩ n is the overall order. When [A] ϭ [B] this
simplifies to Equation 4.
͑m ϩ n͒
R ϭ k͓A͔
(Eq. 4)
0
100
Dissociation kinetics for PMX464-thiol complexes were deter-
mined by measuring the initial rate of thiol release using 5,5Ј-
dithiobis(2-nitrobenzoic acid) (DTNB). Reactions were initi-
ated by addition of 500 ␮M DTNB (final concentration) to a 1:1
mixture of 40, 60, 80, 100, 120, and 140 ␮M T(SH)2, L-cysteine,
or GSH and PMX464 in equilibrium. The initial rate of release
y ϭ
(Eq. 1)
s
i
1 ϩ
ͩ ͪ
IC50
where y is the percentage inhibition for any inhibitor concen-
tration, i, and s is the Hill slope.
Ϫ1
Mass Spectrometry Analysis—Mixtures containing 50 mM
Tris-HCl, pH 7.6, 5 mM EDTA, 200 ␮M NADPH, 10 ␮M
TbTryR, 10 ␮M T(S)2, 10 ␮M TbTryX, and 10 ␮M TbTDPX were
incubated for 1 h at room temperature with 100 ␮M PMX464 or
1% DMSO as control. Adduct formation was analyzed using
matrix-assisted laser desorption/ionization time-of-flight mass
spectrometry (MALDI-TOF MS) by the Mass Fingerprinting
Service Dundee (University of Dundee, Dundee, UK).
of free thiol groups was determined using ⑀412 nm ϭ 13,600 M
cmϪ1 (39).
RESULTS
Quinols Are Potent Cytotoxic Inhibitors of T. brucei—Het-
eroaromatic 4-substituted 4-hydroxycyclohexa-2,5-dienone
compounds (“quinols”) have been reported to be promising
antitumor agents (29, 40–43). To study the effect of the quinol
pharmacophore on T. brucei we selected a series of six different
quinol compounds (1–6) (see Table 1 for structures). A deriv-
ative of compound 1 lacking the quinol pharmacophore was
chosen as a control (7). Bloodstream forms of T. brucei para-
sites and MRC5 normal human fibroblasts were treated with
varying concentrations of 1–7 and EC50 values were deter-
mined (Table 1). Quinol analogues 1–6 all displayed submicro-
molar potencies against T. brucei, with compounds 1, 4, and 5
having EC50 values Յ100 nM. Compound 7, which lacks the
4-hydroxycyclohexa-2,5-dienone pharmacophore, was inac-
tive. Analogues 1–6 were also active against MRC5 cells,
revealing a narrow window of selectivity of only 2–3-fold for the
parasites. However, due to the steeper growth inhibition effects
on T. brucei (the Hill slope), selectivity of 1–6 increases to
between 4- and 7-fold at the EC90. The least active quinols
against both cell types were 2 and 6, where in 2 the hydroxyl
group of PMX464 was replaced by an O-methyl and in 6 the
benzothiazole substituent in 1 was replaced by a phenyl group.
Previous studies have reported that various quinol analogues
exhibit similar effects on growth of mammalian cells and on
inhibition of thioredoxin reductase activity, independent of
the substituent (29, 41, 42). Therefore, we chose to evaluate the
mechanism of anti-trypanosomal activity in detail using the
well studied benzothiazole-substituted quinol (1), named
AW464 or PMX464 in previous publications (44). First, we ana-
lyzed the growth of T. brucei parasites as a function of time
Complex Formation of Low Molecular Mass Thiols with
PMX464—The reaction of PMX464 with thiols was deter-
mined spectrophotometrically (UV-visible Shimadzu spectro-
photometer) by monitoring the decrease in absorbance at 302.5
nm relative to background absorbance at 400 nm. Sequential
spectra were obtained to establish the time taken to reach equi-
librium for each thiol. PMX464 (100 ␮M) was then incubated
with varying concentrations of T(SH)2 (0–250 ␮M), L-cysteine
(0–1500 ␮M), or GSH (0–1000 ␮M) in a volume of 200 ␮l of 100
mM Tris-HCl, pH 7.6, 5 mM EDTA at 25 °C for 1.5, 2, or 4 h,
respectively, and the net differential absorbance relative to zero
time was calculated. These corrected data were plotted against
the total thiol concentration and linear regression of the lower
and higher data used to provide the tangent and asymptote,
respectively. The ratio of thiol binding to PMX464 and thiol
capacity was determined from the intersection point of these
two linear fits. Bound thiol concentration ([L]b) was calculated
Ϫ1
using the ␦ extinction coefficients (2.68, 1.32, and 1.45 mM
cmϪ1 for T(SH)2, L-cysteine, and GSH, respectively) and data
were fitted by non-linear regression using GraFit 5.0.13 to the
following ligand binding equation,
2
ͱ
͑Kd ϩ ͓L͔t ϩ ͓R͔͒ Ϫ ͑Kd ϩ ͓L͔t ϩ ͓R͔͒ Ϫ 4͓L͔ ͓R͔
t
͓L͔b ϭ
2
(Eq. 2)
where [L]t is the total ligand (RSH) concentration and [R] is the following treatment with a concentration of PMX464 equiva-
receptor (PMX464) concentration.
lent to ϳ5 ϫ EC50 (Fig. 1A). After only 4 h exposure, the para-
For kinetic analyses of complex formation, 40, 60, 80, 100, sites ceased growing and started to die thereafter, such that no
120, and 140 ␮M T(SH)2, L-cysteine, or GSH were mixed in an motile parasites could be detected after 10 h. Thiol analysis
equimolar ratio with PMX464 in a volume of 200 ␮l at 25 °C. after the 4-h exposure to PMX464 showed a marked reduction
The initial rates of differential absorbance were determined as in the T(SH)2 content and no significant change in GSH con-
MARCH 11, 2011•VOLUME 286•NUMBER 10
JOURNAL OF BIOLOGICAL CHEMISTRY 8525

Quinols Inhibit the Trypanothione Peroxidase System
TABLE 1
Cell viability assays and the effects of quinol analogues on trypanothione-dependent peroxidase systems
a Trypanosomes and MRC-5 cells were incubated for 69 h with varied amounts of 1–7 followed by addition of resazurin for 4 h as described under “Experimental Proce-
dures.” The EC50 values are the mean of n independent determination with 95% confidence intervals given in parentheses.
b Details of the rate-limiting peroxidase assays for T(SH)2, TDPX, or TRYP are described under “Experimental Procedures.” Mixtures were preincubated for 30 min with 30
␮M quinol compounds 1–7 before the reactions were started with the appropriate substrate. All values are the mean Ϯ S.D. of triplicate measurements.
c The selectivity index (SI) is the ratio of EC50 values for MRC5 and T. brucei cells.
106
105
104
106
105
104
0
4
8
12
0
2
4
6
8
FIGURE 1. Growth curve analysis of PMX464-treated T. brucei. A, the cytocidal effects of PMX464 and melarsen oxide on bloodstream trypanosomes.
Compounds were added at the defined concentrations to T. brucei parasites at a cell density of ϳ1 ϫ 106 mlϪ1. Open circle, control cells with no inhibitor; filled
circle, melarsen oxide, 50 nM; open square, PMX464, 500 nM. At different time points the cell number was determined, as described under “Experimental
Procedures.” Data are the mean Ϯ S.D. of three experiments. B, synergistic effects of PMX464 and the H2O2-producer glucose oxidase. As described above,
compounds were added to bloodstream trypanosomes at a cell density of ϳ1 ϫ 106 mlϪ1 and cell numbers were determined at defined intervals. Open circle,
control cells with no inhibitor; filled circle, glucose oxidase, 0.1 milliunits mlϪ1; open square, PMX464, 500 nM; closed square, PMX464 (500 nM) and glucose
oxidase (0.1 milliunits mlϪ1). Data are representative of two independent experiments.
tent (Fig. 2B). Trypanosomes, incubated with ϳ5 times the gests that this quinol may specifically target the antioxidant
EC50 values of melarsen oxide, started to die after 6 h in a similar defenses of these parasites.
manner to PMX464-treated parasites. Therefore, PMX464
Effect of PMX464 on Thiols and TryP in Whole Trypano-
shows the same rapid trypanocidal activity as this arsenical somes—Various studies have established that quinols inhibit
drug (25, 26). Significantly, the addition of the H2O2 generator the thioredoxin reductase/thioredoxin system in mammalian
glucose oxidase at non-trypanocidal levels was found to accel- cells (29, 31, 32) and increase cellular glutathione content (41).
erate the lytic effect of PMX464 (Fig. 1B). The apparent synergy This prompted us to analyze the effects of PMX464 on T. brucei
of PMX464 and H2O2 against bloodstream trypanosomes sug- thiol metabolism, particularly with respect to its role in metab-
8526 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 286•NUMBER 10•MARCH 11, 2011

Quinols Inhibit the Trypanothione Peroxidase System
FIGURE2. EffectofPMX464onthethiol-dependentantioxidantsysteminT. brucei. A, effectonantioxidantproteins. Mid-logtrypansomeswereincubated
for 16 h with varying concentrations of PMX464 and samples were separated by SDS-PAGE (20 ␮g of protein/lane). TryR, TryX, TDPX, and TryP and the internal
loading control, VSG, were detected by immunoblotting using specific antiserum against each protein. B, effect on thiol levels. T. brucei parasites were
incubated with either 500 or 250 nM PMX464 for 4 or 16 h, respectively, and the concentration of GSH and T(SH)2 determined by HPLC. White bars, GSH; gray
bars, T(SH)2. The results are the mean Ϯ S.D. of three experiments, expressed as a percentage of the thiol content of untreated cells. An asterisk indicates a
significant difference compared with untreated cells (p Ͻ 0.01, Student’s t test).
viously reported in these cells (34). These results demonstrate
that PMX464 perturbs T(SH)2 metabolism prior to cell death.
Trypanothione, TDPX, and TryP Are Molecular Targets of
PMX464—A more detailed analysis of PMX464 inhibition of
these putative molecular targets was carried out in vitro. First,
PMX464 was tested in the standard TDPX and TryP peroxidase
assays. Oxidation of NADPH by hydroperoxides was moni-
tored spectrophotometrically at 340 nm. Addition of cumene
hydroperoxide to the standard TDPX peroxidase assay con-
taining all components resulted in linear oxidation of NADPH
for the first 400 s (Fig. 4A, open circles). In contrast, addition of
PMX464 at the same time as cumene hydroperoxide to initiate
the reaction resulted in a reaction curve typical for a slow bind-
FIGURE 3. EC50 values were determined for PMX464 against WT (open
circles) and TRYS-overexpressing (closed circles) bloodstream trypano-
somes. The curves are the non-linear fits of data using a two-parameter EC50
equation provided by GraFit (see “Experimental Procedures”). EC50 values of
41.2 Ϯ 2.3 and 77.4 Ϯ 2.1 nM were determined for WT and TRYS-overexpress-
ing cell lines, respectively. Data are the mean of duplicate measurements.
ing inhibitor (open triangles). Preincubation of the reduced
assay components with PMX464 for 10 min prior to initiation
with cumene hydroperoxide abolished the peroxidase activity
completely (closed circles). This trace is identical to the one
displaying the background rate containing no TDPX (open
squares). Preincubation of PMX464 with the oxidized assay
olizing peroxides. The trypanothione peroxidase system com-
components, followed by initiation with NADPH, resulted in a
prises the following cascade: NADPH, TryR, T(SH)2, and TryX
similar reaction trace to the one where PMX464 is added
with one of the two types of tryparedoxin-dependent peroxi-
directly at the start of the reaction (closed squares). The latter
dases, TryP and TDPX, as final electron acceptor (15, 19). The
finding suggests that PMX464 only reacts with free thiols, but
effect on these proteins in PMX464-treated parasites was ana-
not with disulfides, in the assay mixture. The inhibitory behav-
lyzed by Western blotting with specific antisera (Fig. 2A).
ior of PMX464 was similar in the TryP peroxidase assay (not
shown).
Although the cellular content of TryR, TryX, and TDPX was
not significantly altered in treated cells compared with the VSG
control, the TryP band decreased in a dose-dependent manner,
indicating that TryP is depleted by PMX464 treatment. Gluta-
thione and trypanothione contents in parasites treated with 250
nM PMX464 for 16 h were analyzed by HPLC following
derivatization with monobromobimane (Fig. 2B). In compari-
son to untreated cells, the intracellular trypanothione levels
were decreased by 30%, whereas the glutathione concentration
was up-regulated by about 30%.
It should be noted that continuous cycling between dithiol
and disulfide forms of TryR, TryX, T(SH)2, TDPX, and TryP are
required for transfer of electrons from NADPH donor substrate
to peroxide acceptor substrate. To dissect which of these com-
ponents might be the molecular target of PMX464, we
exploited the fact that PMX464 is a slow binding inhibitor that
takes several minutes to abolish peroxidase activity. Thus, fur-
ther addition of any component that is inhibited in the assay
mixture by PMX464 should restore the oxidation of NADPH by
cumene hydroperoxide. The rate of peroxide reduction using
the standard TDPX peroxidase assay containing no inhibitor is
illustrated in trace a (Fig. 4B). To identify the molecular targets
in the TDPX peroxidase assay, all assay components except for
cumene hydroperoxide were preincubated with PMX464 for 30
To confirm that T(SH)2 is a target for quinols, cells overex-
pressing trypanothione synthetase and thus maintaining ele-
vated levels of intracellular T(SH)2 were tested for their sensi-
tivity to PMX464 (Fig. 3). These transgenic parasites yielded an
EC50 value of 77.2 Ϯ 3.2 nM, 1.9-fold higher than the value for
wild type cells (41.2 Ϯ 2.3 nM). This fold-shift in sensitivity
correlates well with the 2.5-fold increase in T(SH)2 levels pre- min, so that no peroxidase activity was detected after addition
MARCH 11, 2011•VOLUME 286•NUMBER 10 JOURNAL OF BIOLOGICAL CHEMISTRY 8527

Quinols Inhibit the Trypanothione Peroxidase System
FIGURE4. PMX464inhibitionofTDPX. ThestandardTDPXperoxidaseassaymixturecontainsNADPH, TryR, T(S)2, TryX, andTDPX. Theperoxidasereactionwas
initiated with cumene hydroperoxide (t ϭ 0 min) until otherwise stated. Activity was monitored by following NADPH consumption at 340 nm. PMX464 was
added at 100 ␮M as indicated. A, open circles, standard peroxidase assay mixture, no inhibitor; closed circles, assay mixture minus TDPX, no inhibitor; open
triangles, PMX464 added at zero time together with cumene hydroperoxide to the assay mixture; open squares, assay mixture preincubation with PMX464 for
10 min; closed squares, assay mixture minus NADPH was treated for 5 min with cumene hydroperoxide followed by preincubation with PMX464 for 10 min and
the reaction was started with NADPH. B, identification of the molecular targets of PMX464 in peroxidase assays containing 0.5 ␮M TDPX. Trace a, assay mixture
preincubated for 30 min without inhibitor; traces b-f, assay components preincubated for 30 min with PMX464. The last minute of the preincubation time is
shown before 300 ␮M cumene hydroperoxide was added at zero time. After 2, 4, or 6 min additional aliquots of 5 ␮M TryX, 20 ␮M T(S)2, or 0.5 ␮M TDPX were
added in different orders. C, MALDI analysis of a mixture of the electron cascade components: NADPH (200 ␮M) and TryR, T(S)2, TryX, and TDPX (10 ␮M,
respectively): left panel, control 1% DMSO; right panel, plus 100 ␮M PMX464 in 1% DMSO. MALDI analysis was performed after a 1-h incubation at room
temperature.
of the hydroperoxide at zero time (Fig. 4B, traces b-g). After 2, 4, was increased by 245 m/z in comparison to the untreated con-
and 6 min, additional T(S)2, TDPX, and TryX were added in trol with a corresponding loss of the peak at 19,153 Da. The
different orders to the inhibited assay mixtures. After addition molecular mass of PMX464 is 243, suggesting that 1 mol of
of 20 ␮M trypanothione disulfide T(S)2 (trace b) the rapid and PMX464 was bound per mole of TDPX. The signal for TryX did
stoichiometric consumption of NADPH (about 0.12 absor- not show such a shift, in agreement with the conclusion from
bance units; 19 ␮M) was observed due to the reduction of T(S)2 Fig. 4B that TryX is not inhibited by PMX464. TryR and TryP
to T(SH)2. This shows that TryR is not inhibited by PMX464. could not be detected by MALDI-TOF analysis probably due to
Moreover, the stoichiometry indicates that further reduction of their oligomeric structure in solution.
cumene hydroperoxide does not occur. Only following further
Specificity of Molecular Targets—The above experiments
addition of TDPX to the reaction mixture in trace b was activity indicate that T(SH)2, TDPX, and TryP are the molecular targets
restored. However, addition of TDPX alone to the inhibited of PMX464. To analyze the inhibition of each in more detail, the
assay mixture was not sufficient for the reduction of the two peroxidase assays were optimized such that TDPX, TryP,
hydroperoxide (trace c), which also required the addition of or T(SH)2 was rate-limiting over a linear concentration range
T(S)2. In contrast, further addition of TryX either alone or in (data not shown). In the TDPX- and TryP-dependent assays, a
combination with TDPX or T(S)2 did not enhance activity trace amount of T(S)2 (0.2 ␮M) was included in the assay mix-
(d–g). These results suggest that trypanothione and TDPX are ture, sufficient to allow reduction of the peroxidases before
the molecular targets for PMX464, but not TryR or TryX. TryP addition of quinol, but not enough for significant peroxidase
is inhibited in a similar manner to that of TDPX in the TryP activity. TryP- and TDPX-dependent reactions were initiated
peroxidase assay (data not shown).
with T(S)2 (20 ␮M) and hydroperoxide. For assays where
Inhibition of TDPX by PMX464 is supported by MALDI- T(SH)2 (1 ␮M) was rate-limiting, the reaction was initiated with
TOF MS studies (Fig. 4C). The TDPX peroxidase assay compo- TDPX and hydroperoxide.
nents (NADPH, TryR, T(S)2, TryX, and TDPX) were incubated
Time- and concentration-dependent inhibition by PMX464
with PMX464 (or 1% DMSO as control) for 30 min before anal- in each of these rate-limiting reactions was analyzed by prein-
ysis. Peaks corresponding to His6-TryX (theoretical mass: cubation for different times prior to determining residual activ-
18,054 Da) and TDPX (theoretical mass: 19,168 Da) were ity by addition of the appropriate substrates (Fig. 5, A–C).
detected in the MALDI-TOF spectrum of the control sample. Residual activity was fitted to a first-order rate equation for
Following treatment with PMX464, the peak mass for TDPX each concentration of PMX464 to determine the observed rate
8528 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 286•NUMBER 10•MARCH 11, 2011

Quinols Inhibit the Trypanothione Peroxidase System
FIGURE 5. Time- and concentration-dependent inhibition of T(SH)2, TryP, and TDPX by PMX464. The standard assay conditions, in which T(SH)2 (A), TryP
(B), or TDPX (C) were limiting, are described under “Experimental Procedures.” Varied amounts of PMX464 were added to the standard reaction mixtures,
respectively. Aliquotswereremovedatintervalsandresidualactivitiesweredetermined. A, inhibitionwithlimitingT(SH)2:opencircles, noinhibitor;filledcircles,
3.125 ␮M; open squares, 6.25 ␮M; filled squares, 12.5 ␮M; open triangles, 25 ␮M; filled triangles, 100 ␮M. Inhibition with limiting TryP (B) or TDPX (C): open circles, no
inhibitor; closed circles, 6.25 ␮M; open squares, 12.5 ␮M; closed squares, 25 ␮M; open triangles, 50 ␮M; closed triangles, 100 ␮M. D, net kobs as a function of PMX464
concentration: open circles, T(SH)2; open triangles, TryP; inverted closed triangles, TDPX. Data points were fitted by linear regression.
TABLE 2
30 ␮M 1–7 (Table 1). Control compound 7 did not exhibit an
Inhibitory activity of PMX464 against T(SH)2, TDPX, and TryP
inhibitory activity in any of the assays showing the quinol phar-
Residual activities were determined after 30 or 90 min preincubation with varying
macophore is specifically responsible for the inhibitory activity.
amounts of PMX464 in the rate-limiting assays for T(SH)2, TDPX, and TryP. The
data represent the weighted mean Ϯ S.E. of three independent experiments fitted by
In all cases, the trypanothione-limited assays were more sensi-
non-linear regression (Eq. 1).
tive to inhibition by quinols than the TDPX- or TryP-limited
T(SH)2
30 min 90 min
TDPX
30 min 90 min
TryP,
assays with a good correlation between cell potency and inhibi-
tion of T(SH)2-limited peroxidase.
30 min
IC50 (␮M) 5.8 Ϯ 0.3 1.8 Ϯ 0.1 15.6 Ϯ 1.3 6.0 Ϯ 0.4 14.1 Ϯ 0.8
Affinity of PMX464 for Low Molecular Mass Thiols—Gluta-
thione and trypanothione are the principal low molecular mass
thiols in trypanosomes (6) and these parasites require high con-
centrations of L-cysteine (ϳ1.5 mM) for growth in vitro. These
facts prompted us to analyze the interaction of T(SH)2, GSH,
and L-cysteine with PMX464. For this purpose we exploited the
observation that incubation with these thiols decreases the
absorbance of PMX464 at 302.5 nm in a time- and concentra-
Hill slope 1.6 Ϯ 0.1 2.1 Ϯ 0.2 1.5 Ϯ 0.1 1.9 Ϯ 0.2 1.4 Ϯ 0.1
constant of inactivation (kobs). The resulting rate constants
were then plotted against the PMX464 concentration (Fig. 5D).
For all three components, the values for kobs were directly pro-
portional to the concentration of PMX464, indicating that the
reactions were not saturable (i.e. Ki ϾϾ 100 ␮M). The enzymes
TDPX and TryP showed similar inactivation rates, whereas
substrate T(SH)2 was inactivated about 2.5-fold faster.
tion-dependent manner, with equilibrium of T(SH)2, L-cys-
1
teine, and GSH with PMX464 being attained after 1, 1 ⁄
2, and
3 h, respectively (data not shown). For quantitative analysis, 100
␮M PMX464 was incubated with varying concentrations of
each thiol and differential absorbance data were plotted against
the thiol concentration. Representative results are shown for
T(SH)2 and GSH (Fig. 6, A and B, insets). In all cases a similar
maximum absorbance plateau was achieved with 5–10-fold
excess thiol, yielding ␦ extinction coefficients of 2.68, 1.45, and
1.32 mMϪ1 cmϪ1 for T(SH)2, GSH, and cysteine, respectively. In
the case of trypanothione, one molecule of the dithiol T(SH)2
IC50 values for T(SH)2, TDPX, and TryP were determined
using the standard rate-limiting peroxidase assay conditions.
The reaction mixtures were incubated with varying concentra-
tions of PMX464, for either 30 or 90 min, before the residual
peroxidase activity was determined. The calculated IC50 values
are summarized in Table 2. TDPX and TryP displayed similar
IC50 values of ϳ15 ␮M, whereas T(SH)2 had a lower IC50 value
of 5.8 ␮M after 30 min, decreasing to 1.8 ␮M after 90 min. The
same standard assay conditions with T(SH)2, TDPX, or TryP as
the rate-limiting component were used to examine the struc- bound to one quinol molecule (Fig. 6A, inset), whereas for the
ture-activity relationship for the various quinol analogues 1–7. monothiols GSH (Fig. 6B, inset) or L-cysteine (not shown) the
Inhibition was determined after a 30-min incubation with stoichiometry is 2:1. The concentration of bound thiol was then
MARCH 11, 2011•VOLUME 286•NUMBER 10
JOURNAL OF BIOLOGICAL CHEMISTRY 8529

Quinols Inhibit the Trypanothione Peroxidase System
FIGURE 6. Analysis of PMX464 binding to T(SH)2, GSH, and L-cysteine. PMX464 (100 ␮M) was incubated with increasing amounts of T(SH)2 (A), GSH (B), or
L-cysteine (not shown). At equilibrium the absorbance at 302.5 and 400 nm (background) were measured. The insets show the calculated differential absorb-
ance plotted againstthethiolconcentration. Bindingratiosandthiolcapacities were determined and used for calculation of bound thiol concentrations. These
data were plotted against the total thiol concentration and fitted to an equation describing tight ligand binding. Data are listed in Table 3. C, the initial rate of
binding of PMX464 to T(SH)2 (open circles), GSH (open squares), or L-cysteine (closed circles) at equimolar concentrations were determined by measuring the
absorbance at 302.5 and 400 nm. The initial rates of thiol binding were calculated and plotted against the thiol concentration. D, an equimolar mixture of
PMX464 and T(SH)2 (open circles), GSH (open squares), or L-cysteine (closed circles) was incubated until equilibrium was reached. DTNB was added and the
reaction with reactive-free thiols was monitored at 412 nm. Here, the apparent initial rate is plotted against the thiol concentration. See “Experimental
Procedures” for further details.
TABLE 3
Analysis of T(SH)2, GSH, and L-cysteine binding to PMX464
See Fig. 5 and “Experimental Procedures” for further details. Results are the weighted mean Ϯ S.D. of three independently performed experiments.
Overall reaction
Complex
Complex
Half-time
a
Kd
orderb
formation,b k1
dissociationc
for dissociation
Ϫ3
Ϫ3
␮M
␮MϪ1 minϪ1 ϫ 10
k2 , minϪ1 ϫ 10
min
GSH
6.2 Ϯ 1.9
1.11 Ϯ 0.07
1.37 Ϯ 0.16
1.08 Ϯ 0.06
3.6 Ϯ 1.1
6.6 Ϯ 4.5
45.6 Ϯ 8.9
1.21 Ϯ 0.07
8.59 Ϯ 0.20
2.50 Ϯ 0.08
573
81
L-Cysteine
T(SH)2
26.7 Ϯ 2.8
0.43 Ϯ 0.09
277
a Apparent dissociation constants (Kd) were calculated using the ligand binding equation (Eq. 2).
b The overall reaction order and the apparent rate constants, k1 were calculated from the initial reaction rates fitted to Eq. 4.
c The apparent rate constants, k2 for complex dissociation were calculated by linear regression (correlation coefficients Ͼ0.994) of the initial rates of thiol formation follow-
ing the addition of DTNB.
calculated from the above extinction coefficients (Fig. 6, A and residual concentration of free thiol with DTNB gave essentially
B) and fitted to the ligand binding equation (Equation 2) yielding identical values for the bound thiol concentration at low thiol con-
apparent dissociation constants (Kd) of 0.43, 6.2, and 27 ␮M for centrations, but was not sufficiently accurate at higher thiol con-
T(SH)2, GSH, andL-cysteine, respectively(Table3). Analysisofthe centrations to provide meaningful estimates of Kd.
8530 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 286•NUMBER 10•MARCH 11, 2011

Quinols Inhibit the Trypanothione Peroxidase System
1
2
3
4
SH
SH
TryR
TryX
lanes 1, 3
H2O2
Reduced
TDPX
Oxidised
DTT
PMX464
*
*
S
S
S
PMX
464
S
TDPX monomer
lane 2
lane 4
FIGURE 7. A, mobility shift analysis of a mixture of PMX464 incubated with TDPX by non-reducing SDS-PAGE (12% NuPAGE). TDPX2 was incubated for 30 min
with 10 mM DTT (lane 1), 100 ␮M H2O2 (lane 2), 100 ␮M NADPH, 0.1 nmol of TryR, 0.1 nmol of TryX, 0.1 nmol of T(SH)2 (lane 3), and as in lane 3 with an additional
100 ␮mol of PMX464 (lane 4). 0.1 nmol of TDPX2 was analyzed per lane. The samples for lanes 3 and 4 were not heated before SDS-PAGE analysis. B, a ribbon
diagram of the reduced monomeric TDPX is depicted. The two yellow spheres represent the two catalytic cysteine residues (19). C, the scheme illustrates the
TDPX redox reactions. The reduced monomeric TDPX forms upon oxidation of an intramolecular disulfide bridge. It is proposed that the two thiols are involved
in binding to one molecule of PMX464.
1
2
3
4
SH
SH
HS
HS
TryR
TryX
TryP
lanes 1, 3
dimer
TryP
monomer
H2O2
DTT
PMX464
PMX
S
S
S
S
S
S
S
464
TryP dimer
PMX
464
S
lane 2
lane 4
FIGURE 8. A, mobility shift analysis of a mixture of PMX464 incubated with TryP by non-reducing SDS-PAGE (12% NuPAGE). TryP was incubated for 30 min with
10 mM DTT (lane 1), 100 ␮M H2O2 (lane 2), 100 ␮M NADPH, 0.1 nmol of TryR, 0.1 nmol of TryX, 0.1 nmol of T(SH)2 (lane 3), and as in lane 3 with an additional 100
␮mol of PMX464 (lane 4). 0.1 nmol of TryP were analyzed per lane. The samples for lanes 3 and 4 were not heated before SDS-PAGE analysis. B, a ribbon diagram
of the reduced TryP homodimer, with the yellow spheres indicating the position of the catalytic cysteine residues (45). C, the scheme illustrates the TryP redox
reactions. The reduced dimer decays under the denaturing conditions of an SDS-PAGE to a monomer. TryP forms upon oxidation of an intermolecular disulfide
bridge so that the dimer is cross-linked in the oxidized form. It is proposed that the two thiols are involved in binding to one molecule of PMX464 cross-linking
the two TryP subunits.
The kinetics of thiol-PMX464 complex formation was deter- against the PMX464 concentration and analyzed by linear
mined using equimolar concentrations of each reactant. Initial regression (Table 3). The rate of complex dissociation
reaction rates were calculated and plotted against the PMX464/ decreased from L-cysteine, T(SH)2, to GSH with half-times for
thiol concentration (Fig. 6C). Data were fitted to Equation 4 dissociation of 1.4, 4.6, and 9.6 h, respectively. These findings
describing a reaction of overall order (m ϩ n) (Table 3). GSH indicate that binding of quinols to low molecular mass thiols is
and T(SH)2 were essentially pseudo-first order, whereas cys- tight, but reversible.
teine yielded a complex overall order of 1.4. The dithiol T(SH)2
was the most reactive and glutathione the least reactive under part of their catalytic cycles, TDPX (Fig. 7) forms an intramo-
these assay conditions. lecular and TryP (Fig. 8) an intermolecular disulfide bridge
Mechanism of Inhibition of TDPX and TryP by PMX464—As
The stability of the different PMX464 complexes was exam- upon oxidation with hydroperoxide. In these two redox states
ined using DTNB. Addition of excess DTNB should immedi- both enzymes display different mobility by SDS-PAGE under
ately remove free thiols from the equilibrium mixture, causing non-reducing conditions (Figs. 7 and 8, lanes 1 and 2). To
further dissociation of the PMX464-thiol complex and provid- examine the effect of PMX464, TDPX and TryP were reduced
ing an assessment of the rate of dissociation of the complexes. by the TryR/T(SH)2/TryX-dependent system followed by incu-
Dissociation of the PMX464-thiol complexes was found to be bation with PMX464 or DMSO as control. Proteins were sepa-
slow, taking ϳ5 h to reach completion for cysteine and more rated by SDS-PAGE without heating the samples. PMX464
than 20 h for GSH (data not shown). The apparent initial rates treatment promoted a small mobility shift in the direction of
of dissociation were determined for each equimolar mixture of the band for oxidized TDPX (Fig. 7, compare starred bands in
PMX464 and thiol in equilibrium. The data were plotted lanes 3 and 4), which can be discerned relative to the TryX band
MARCH 11, 2011•VOLUME 286•NUMBER 10
JOURNAL OF BIOLOGICAL CHEMISTRY 8531

Quinols Inhibit the Trypanothione Peroxidase System
that has previously been shown not to react with PMX464 (Fig. whereas the other components, TryR and TryX, are not. The
4C). A more pronounced shift can be observed for TryP (Fig. 8). flexible nature of TDPX (19) and TryP (45, 50) and the conse-
In solution TryP is a homodimer that can oligomerize to form a quent accessibility of the peroxidatic and resolving cysteine res-
doughnut-shaped pentamer of dimers (45). The reduced pro- idues could favor reaction with both Michael acceptor sites on
tein migrates as a monomer of 24.6 kDa by SDS-PAGE (Fig. 7, a single quinol molecule. In contrast, in TryX, a thioredoxin-
lane 1). Following oxidation of the reduced enzyme with per- like protein, and in TryR, a homologue of glutathione reduc-
oxide, the protein migrates as a disulfide cross-linked dimer of tase, only one of the redox-active cysteine residues is readily
49.2 kDa under non-reducing conditions (Fig. 8, lane 2). In the accessible to solvent (51–54).
absence of sample heating, complete dissociation of the TryP
Given that quinols are promiscuous toward dithiol-contain-
decameric complex does not occur and only a faint band at the ing biological molecules, the following evidence supports the
size of a monomer is seen in the TryP sample reduced by TryX conclusion that trypanothione and the monomeric and deca-
(Fig. 8, lane 3). Addition of PMX464 resulted in complete dis- meric tryparedoxin peroxidases are primary targets of PMX464
appearance of the band corresponding to the size of the mono- in trypanosomes. First, exposure to PMX464 results in de-
mer, which was associated with the appearance of a band cor- creased T(SH)2 content, whereas glutathione is not decreased
responding to the size of a TryP dimer (Fig. 8, lane 4). These (Fig. 2B). Second, transgenic trypanosomes overexpressing
results suggest that PMX464 reacts with the two catalytic cys- trypanothione synthetase with elevated T(SH)2 cellular content
teine residues of TDPX and TryP, respectively.
show increased resistance to this quinol (Fig. 3). Third, non-
Brief heat treatment (95 °C, 1 min) of PMX464-inactivated lytic concentrations of H2O2 accelerate cell lysis induced by
TryP resulted in complete dissociation of the TryP dimeric PMX464 (Fig. 1B). This could be due to either a decrease in
complex. Presumably other buried SH groups in the denatured T(SH)2 to critical levels as a result of oxidative stress, or a direct
protein are then able to exchange with quinol cross-linked inhibition of TryP and/or TDPX. Importantly, a combination of
dimers to form monomer quinol adducts. However, this com- both effects could account for the higher IC50 values obtained
plex appears to be highly stable in the native state, because no in vitro against the peroxidase systems compared with EC50
peroxidase activity could be recovered even after extensive dial- values against whole cells. However, it is also possible that the
ysis for 24 h and the complex retained the shift in mobility seen IC50 values determined in Table 3 are underestimated because
in Fig. 8, lane 4.
PMX464 is a time-dependent inhibitor and instability of the
reaction components limits these determinations to 90 min. In
contrast, EC50 values are determined after a 72-h exposure to
DISCUSSION
Recently, quinols were found to have promising inhibitory PMX464.
activity against various types of cancer cells (29, 31, 40, 46).
The molecular mechanism of quinol inhibitory activity on
Here, we report that quinols are also very potent trypanocidal trypanosome thiol metabolism may give valuable insights con-
agents. Among these, PMX464 and the arylsulfonylindole-sub- cerning quinol-mediated cytotoxic effects in other organisms.
stituted quinols were the most potent with EC50 values Յ100 Homologues of TryP, the peroxiredoxins, are ubiquitous pro-
nM. PMX464 exhibits a rapid cytocidal rather than a cytostatic teins with highly conserved structures (49). Therefore, it seems
effect on trypanosomes, a distinct advantage for future drug likely that peroxiredoxins are also targets for quinols in mam-
development. All quinol analogues tested were consistently malian cells. Indeed, PrxI was recently identified as a quinol
more selective, albeit marginally, against T. brucei than against target in HCT 116 cell lysates by affinity chromatography using
MRC5 cells. Selectivity may be underestimated possibly due to an immobilized quinol (41). Moreover, peroxiredoxins have
sequestration of the free quinol by the high concentration of been reported to be overexpressed in various cancer cell lines
L-cysteine in the parasite culture medium. However, the revers- (55). Therefore, we propose that inhibition of peroxiredoxins
ible nature of the cysteine-quinol complex and the fairly rapid could contribute to the antitumor activity of quinols.
rate of dissociation may allow equilibrium exchange with other
In conclusion we have shown that the antitumor quinols are
higher affinity intracellular targets such as T(SH)2, TryP, and very potent trypanocidal agents whose effects are mediated, at
TDPX to take place over the 72-h incubation. Quinols are dou- least in part, by reaction with T(SH)2, TryP, and TDPX, thereby
ble Michael acceptors, suggesting that biological molecules inhibiting the pivotal, unique, and essential antioxidant system
with vicinal dithiols would be preferential targets. Indeed, in these parasites. Quinols are clearly not “magic bullets” spe-
monothiols such as glutathione and cysteine form 14-fold cifically inhibiting one target, but take advantage of “polyphar-
weaker 2:1 complexes than the 1:1 complex with T(SH)2. This macology” (56, 57). As a result, selectivity is an issue that will be
can be attributed to the fact that the reaction in dilute solution addressed in future studies.
is essentially intramolecular with dithiols due to the higher
Acknowledgments—We thank Pharminox Ltd., Nottingham, UK, for
provision of quinol inhibitors, Irene Hallyburton and Bhavya Rao for
excellent assistance in performing the growth inhibition assays, and
Dr. Stephen Patterson for advice.
effective concentration of the second thiol group once the first
Michael addition is formed.
The two trypanothione-dependent peroxidase systems com-
prise a series of dithiol-containing proteins (TryR, TryX, TryP,
and TDPX) that switch between reduced (dithiol) and oxidized
(disulfide) states during catalysis (19, 36, 47–49). Both classes
of tryparedoxin peroxidase (TryP and TDPX) are inhibited in
concentration- and time-dependent manners by quinols,
REFERENCES
1. Stuart, K., Brun, R., Croft, S., Fairlamb, A., Gu¨rtler, R. E., McKerrow, J.,
Reed, S., and Tarleton, R. (2008) J. Clin. Invest. 118, 1301–1310
8532 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 286•NUMBER 10•MARCH 11, 2011

Quinols Inhibit the Trypanothione Peroxidase System
2. Pecoul, B. (2004) Plos Med. 1, e6
B., Westwell, A. D., and Stevens, M. F. G. (2003) J. Med. Chem. 46,
532–541
3. Balasegaram, M., Harris, S., Checchi, F., Ghorashian, S., Hamel, C., and
Karunakara, U. (2006) Bull. World Health Org. 84, 783–791
4. Balasegaram, M., Young, H., Chappuis, F., Priotto, G., Raguenaud, M. E.,
and Checchi, F. (2009) Trans. R. Soc. Trop. Med. Hyg. 103, 280–290
5. Fairlamb, A. H., Blackburn, P., Ulrich, P., Chait, B. T., and Cerami, A.
(1985) Science 227, 1485–1487
31. Mukherjee, A., Westwell, A. D., Bradshaw, T. D., Stevens, M. F., Carmi-
chael, J., and Martin, S. G. (2005) Br. J. Cancer 92, 350–358
32. Pallis, M., Bradshaw, T. D., Westwell, A. D., Grundy, M., Stevens, M. F.,
and Russell, N. (2003) Biochem. Pharmacol. 66, 1695–1705
33. Greig, N., Wyllie, S., Patterson, S., and Fairlamb, A. H. (2009) FEBS J. 276,
376–386
6. Ariyanayagam, M. R., and Fairlamb, A. H. (2001) Mol. Biochem. Parasitol.
115, 189–198
34. Wyllie, S., Oza, S. L., Patterson, S., Spinks, D., Thompson, S., and Fairlamb,
A. H. (2009) Mol. Microbiol. 74, 529–540
7. Fairlamb, A. H., and Cerami, A. (1992) Annu. Rev. Microbiol. 46, 695–729
8. Dormeyer, M., Reckenfelderba¨umer, N., Ludemann, H., and Krauth-Sie- 35. Torrie, L. S., Wyllie, S., Spinks, D., Oza, S. L., Thompson, S., Harrison, J. R.,
gel, R. L. (2001) J. Biol. Chem. 276, 10602–10606
9. Augustyns, K., Amssoms, K., Yamani, A., Rajan, P. K., and Haemers, A.
(2001) Curr. Pharm. Des. 7, 1117–1141
Gilbert, I. H., Wyatt, P. G., Fairlamb, A. H., and Frearson, J. A. (2009)
J. Biol. Chem. 284, 36137–36145
36. Ko¨nig, J., and Fairlamb, A. H. (2007) FEBS J. 274, 5643–5658
10. Flohe´, L., Hecht, H. J., and Steinert, P. (1999) Free Radic. Biol. Med. 27, 37. Fairlamb, A. H., Henderson, G. B., Bacchi, C. J., and Cerami, A. (1987) Mol.
966–984
11. Krauth-Siegel, R. L., Meiering, S. K., and Schmidt, H. (2003) Biol. Chem. 38. Studier, F. W. (2005) Protein Expr. Purif. 41, 207–234
384, 539–549 39. Gething, M. J., and Davidson, B. E. (1972) Eur. J. Biochem. 30, 352
Biochem. Parasitol. 24, 185–191
12. Aboagye-Kwarteng, T., Smith, K., and Fairlamb, A. H. (1992) Mol. Micro- 40. Berry, J. M., Bradshaw, T. D., Fichtner, I., Ren, R., Schwalbe, C. H., Wells,
biol. 6, 3089–3099
G., Chew, E. H., Stevens, M. F., and Westwell, A. D. (2005) J. Med. Chem.
48, 639–644
13. Krieger, S., Schwarz, W., Ariyanayagam, M. R., Fairlamb, A. H., Krauth-
Siegel, R. L., and Clayton, C. (2000) Mol. Microbiol. 35, 542–552
14. Lu¨demann, H., Dormeyer, M., Sticherling, C., Stallmann, D., Follmann, H.,
and Krauth-Siegel, R. L. (1998) FEBS Lett. 431, 381–385
15. Tetaud, E., Giroud, C., Prescott, A. R., Parkin, D. W., Baltz, D., Biteau, N.,
Baltz, T., and Fairlamb, A. H. (2001) Mol. Biochem. Parasitol. 116,
171–183
41. Chew, E. H., Matthews, C. S., Zhang, J., McCarroll, A. J., Hagen, T., Ste-
vens, M. F., Westwell, A. D., and Bradshaw, T. D. (2006) Biochem. Biophys.
Res. Commun. 346, 242–251
42. Chew, E. H., Lu, J., Bradshaw, T. D., and Holmgren, A. (2008) FASEB J. 22,
2072–2083
43. Mortimer, C. G., Wells, G., Crochard, J. P., Stone, E. L., Bradshaw, T. D.,
Stevens, M. F., and Westwell, A. D. (2006) J. Med. Chem. 49, 179–185
44. Callister, M. E., Pinhu, L., Catley, M. C., Westwell, A. D., Newton, R.,
Leaver, S. K., Quinlan, G. J., Evans, T. W., Griffiths, M. J., and Burke-
Gaffney, A. (2008) Br. J. Pharmacol. 155, 661–672
16. Budde, H., Flohe´, L., Hecht, H. J., Hofmann, B., Stehr, M., Wissing, J., and
Lu¨nsdorf, H. (2003) Biol. Chem. 384, 619–633
17. Henderson, G. B., Fairlamb, A. H., and Cerami, A. (1987) Mol. Biochem.
Parasitol. 24, 39–45
18. Trujillo, M., Budde, H., Pin˜eyro, M. D., Stehr, M., Robello, C., Flohe´, L., 45. Alphey, M. S., Bond, C. S., Tetaud, E., Fairlamb, A. H., and Hunter, W. N.
and Radi, R. (2004) J. Biol. Chem. 279, 34175–34182
19. Alphey, M. S., Ko¨nig, J., and Fairlamb, A. H. (2008) Biochem. J. 414, 46. Lion, C. J., Matthews, C. S., Wells, G., Bradshaw, T. D., Stevens, M. F., and
375–381 Westwell, A. D. (2006) Bioorg. Med. Chem. Lett. 16, 5005–5008
20. Hillebrand, H., Schmidt, A., and Krauth-Siegel, R. L. (2003) J. Biol. Chem. 47. Ko¨nig, J., Lotte, K., Plessow, R., Brockhinke, A., Baier, M., and Dietz, K. J.
278, 6809–6815 (2003) J. Biol. Chem. 278, 24409–24420
21. Schlecker, T., Comini, M. A., Melchers, J., Ruppert, T., and Krauth-Siegel, 48. Wood, Z. A., Poole, L. B., Hantgan, R. R., and Karplus, P. A. (2002) Bio-
R. L. (2007) Biochem. J. 405, 445–454 chemistry 41, 5493–5504
22. Schlecker, T., Schmidt, A., Dirdjaja, N., Voncken, F., Clayton, C., and 49. Wood, Z. A., Schro¨der, E., Harris, J., and Poole, L. B. (2003) Trends
Krauth-Siegel, R. L. (2005) J. Biol. Chem. 280, 14385–14394 Biochem. Sci. 28, 32–40
23. Comini, M. A., Krauth-Siegel, R. L., and Flohe´, L. (2007) Biochem. J. 402, 50. Pin˜eyro, M. D., Pizarro, J. C., Lema, F., Pritsch, O., Cayota, A., Bentley,
43–49 G. A., and Robello, C. (2005) J. Struct. Biol. 150, 11–22
24. Wilkinson, S. R., Horn, D., Prathalingam, S. R., and Kelly, J. M. (2003) 51. Alphey, M. S., Leonard, G. A., Gourley, D. G., Tetaud, E., Fairlamb, A. H.,
J. Biol. Chem. 278, 31640–31646 and Hunter, W. N. (1999) J. Biol. Chem. 274, 25613–25622
(2000) J. Mol. Biol. 300, 903–916
25. Cunningham, M. L., Zvelebil, M. J., and Fairlamb, A. H. (1994) Eur. 52. Alphey, M. S., Gabrielsen, M., Micossi, E., Leonard, G. A., McSweeney,
J. Biochem. 221, 285–295
S. M., Ravelli, R. B., Tetaud, E., Fairlamb, A. H., Bond, C. S., and Hunter,
W. N. (2003) J. Biol. Chem. 278, 25919–25925
26. Fairlamb, A. H., Henderson, G. B., and Cerami, A. (1989) Proc. Natl. Acad.
Sci. U.S.A. 86, 2607–2611
53. Bond, C. S., Zhang, Y., Berriman, M., Cunningham, M. L., Fairlamb, A. H.,
and Hunter, W. N. (1999) Structure 7, 81–89
27. Mukhopadhyay, R., Dey, S., Xu, N., Gage, D., Lightbody, J., Ouellette, M.,
and Rosen, B. P. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 10383–10387
28. Wyllie, S., Cunningham, M. L., and Fairlamb, A. H. (2004) J. Biol. Chem.
279, 39925–39932
54. Jones, D. C., Ariza, A., Chow, W. H., Oza, S. L., and Fairlamb, A. H. (2010)
Mol. Biochem. Parasitol. 169, 12–19
55. Kang, S. W., Rhee, S. G., Chang, T. S., Jeong, W., and Choi, M. H. (2005)
Trends Mol. Med. 11, 571–578
29. Bradshaw, T. D., Matthews, C. S., Cookson, J., Chew, E. H., Shah, M.,
Bailey, K., Monks, A., Harris, E., Westwell, A. D., Wells, G., Laughton, 56. Hopkins, A. L., Mason, J. S., and Overington, J. P. (2006) Curr. Opin. Struct.
C. A., and Stevens, M. F. G. (2005) Cancer Res. 65, 3911–3919
Biol. 16, 127–136
30. Wells, G., Berry, J. M., Bradshaw, T. D., Burger, A. M., Seaton, A., Wang,
57. Hopkins, A. L. (2008) Nat. Chem. Biol. 4, 682–690
MARCH 11, 2011•VOLUME 286•NUMBER 10
JOURNAL OF BIOLOGICAL CHEMISTRY 8533

Metabolism:
Antitumor Quinol PMX464 Is a Cytocidal
Anti-trypanosomal Inhibitor Targeting
Trypanothione Metabolism
Janine König, Susan Wyllie, Geoffrey Wells,
Malcolm F. Stevens, Paul G. Wyatt and Alan
H. Fairlamb
J. Biol. Chem. 2011, 286:8523-8533.
doi: 10.1074/jbc.M110.214833 originally published online January 6, 2011
Access the most updated version of this article at doi: 10.1074/jbc.M110.214833
Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites.
Alerts:
• When this article is cited
• When a correction for this article is posted
Click here to choose from all of JBC's e-mail alerts
This article cites 57 references, 19 of which can be accessed free at
http://www.jbc.org/content/286/10/8523.full.html#ref-list-1