Trypanosoma cruzi: Effect and mode of action of nitroimidazole and nitrofuran derivatives

Article abstract of DOI:10.1016/S0006-2952(02)01663-5

With the aim of determining the actual target(s) of nitro-group bearing compounds considered as possible leads for the development of drugs against Chagas' disease, we studied in parallel nitrofurans and nitroimidazoles. We investigated nine representative compounds for the following properties: efficacy on different Trypanosoma cruzi strains, redox cyclers, inhibition of respiration, production of corresponding nitroso derivatives and intracellular thiol scavengers. Our results indicate that nifurtimox and related compounds act as redox cyclers, whereas the most active in the series, the 5-nitroimidazole megazol essentially acts as thiol scavenger particularly for trypanothione, the cofactor for trypanothione reductase, an essential enzyme in the detoxification process.

Full text of DOI:10.1016/S0006-2952(02)01663-5

Biochemical Pharmacology 65 (2003) 999–1006
Trypanosoma cruzi: effect and mode of action of nitroimidazole
and nitrofuran derivatives
a
Juan Diego Maya , Soledad Bollo , Luis J. Nu n˜ ez-Vergara , Juan A. Squella ,
b
b
b
a
a
c
Yolanda Repetto , Antonio Morello , Jacques P e´ ri e´ , G e´ rard Chauvi e` re
c,*
a
Faculty of Medicine, ICBM, Program of Molecular and Clinical Pharmacology, University of Chile, P.O. Box 70086, Santiago 7, Chile
b
Laboratory of Bioelectrochemistry, Faculty of Chemical & Pharmaceutical Sciences, University of Chile, Santiago, Chile
c
Groupe de Chimie Organique Biologique, Universit e´ Paul Sabatier, UMR-CNRS 5068, Toulouse, France
Received 4 June 2002; accepted 26 November 2002
Abstract
With the aim of determining the actual target(s) of nitro-group bearing compounds considered as possible leads for the development of
drugs against Chagas’ disease, we studied in parallel nitrofurans and nitroimidazoles. We investigated nine representative compounds for
the following properties: efficacy on different Trypanosoma cruzi strains, redox cyclers, inhibition of respiration, production of
corresponding nitroso derivatives and intracellular thiol scavengers. Our results indicate that nifurtimox and related compounds act
as redox cyclers, whereas the most active in the series, the 5-nitroimidazole megazol essentially acts as thiol scavenger particularly for
trypanothione, the cofactor for trypanothione reductase, an essential enzyme in the detoxification process.
#
2003 Elsevier Science Inc. All rights reserved.
Keywords: Nitroheterocycles; Trypanosoma cruzi; Redox cycling; Intracellular trypanothione scavengers; Respiration inhibitors; Thiols depletion
1
. Introduction
released to the suspending medium [5–7]. These and other
experiments [8,9] suggest that intracellular reduction of
ꢀ
ꢁ
Trypanosoma cruzi is the etiological agent of Chagas’
nifurtimox followed by redox cycling, yielding O2 and
H O is the mode of action of this compound against
disease, a chronic illness affecting many people in Latin
America [1]. Currently, this pathology is treated with
nitroheterocyclic agents such as nifurtimox and benznida-
zole. These two drugs are effective against the circulating
form of the parasite (trypomastigotes) during the acute
phase of the disease, but not during the chronic stage.
Additionally, they produce serious adverse effects includ-
ing mutagenesis [2,3], if used in long-term therapy. There
are also differences in sensitivity between the different
strains of the parasite [4]. A characteristic ESR spectrum
corresponding to the nitro-anion radical appears when
nifurtimox is added to intact T. cruzi cells; also, H O is
2
2
T. cruzi.
Aweak signal in the ESR spectrum, corresponding to the
nitro-anion radical appears with the addition of benznida-
zole to NADPH-containing homogenates of T. cruzi [7].
ꢁ
ꢀ
Benznidazole does not stimulate O2 and H O genera-
2
2
tion under the same experimental conditions [10]. Benz-
nidazole inhibits the growth of T. cruzi at concentrations
ꢀ
ꢁ
that do not stimulate O2
and H O production. This
2
2
result indicates that the trypanocidal effect of benznidazole
does not depend on oxygen radicals as with nifurtimox
[10]. Probably reduced metabolites of benznidazole are
involved through covalent binding to macromolecules in
its trypanocidal and toxic effects [5,11–14].
Megazol [1-methyl-2-(5-amino-1,3,4-thiadiazole)-5-
nitroimidazole] is a nitroheterocyclic derivative with anti-
bacterial and antiparasitic activity particularly against
trypanosomes [15,16]. The combination of megazol with
suramin [17–19] or melarsoprol [20] in the treatment
of T. brucei infections potentiates its trypanocidal effect.
Under experimental conditions, megazol can generate
2
2
*
Corresponding author. Tel.: þ33-561-55-68-07;
fax:þ33-561-55-60-11.
E-mail address: chauvier@cict.fr (G. Chauvie
`
re).
Abbreviations: DMF, dimethylformamide; DMSO, dimethylsulfoxide;
THF, tetrahydrofuran; ITBA, iodide tetrabutylammonium; EDTA, ethyle-
0
nediamine tetraacetic acid; HEPPS, N-(2-hydroxyethyl)-1-piperazine -N -
3-propane sulfonic acid; GSH, glutathione; T(SH)
2
, trypanothione; Ln,
natural logarithm; k
c
, culture growth constant; Epc, reduction cathodic peak
potential.
0006-2952/03/$ – see front matter # 2003 Elsevier Science Inc. All rights reserved.
doi:10.1016/S0006-2952(02)01663-5

1000
J.D. Maya et al. / Biochemical Pharmacology 65 (2003) 999–1006
0
oxygen redox cycling [21] at high concentration but pro-
tein synthesis inhibition is suggested as its anti-T. cruzi
mode of action [22]. Because of its efficacy against several
strains of T. cruzi of diverse sensitivity, it has a potential
role in treatment of Chagas’ disease. However, megazol
has shown mutagenicity [23] at least according to the Ames
assay; for this reason, it is necessary to re-evaluate this
mutagenicity using more recent assays and to develop
nitroimidazoles with similar trypanocidal activity but
lower mutagenic profiles. RO150216, other nitroimidazole
derivative, has been shown to be effective in inhibiting the
culture growth of different strains of the African trypano-
some [24], and in vivo using various animal models [25].
To get more information on how these different com-
pounds may act and with the aim to define in a second step
other possible structures, the present work was undertaken
by including in the same study and according to the same
protocol several nitroimidazoles and nitrofurans deriva-
tives shown in Fig. 1. These compounds were considered
in a parallel study for their action on different strains on
T. cruzi and their effect on respiration inhibition, redox
potential, redox cycling properties and effect as thiol
scavengers particularly for trypanothione and glutathione.
Laboratories,RO150216(2-(dimethylamino)-4 -[(1-methyl-
2-nitroimidazole-5-yl)methoxy]acetanilide) and benzni-
dazole (N-benzyl-2-nitroimidazole-1-acetamide) from
Hoffman LaRoche; megazol (2-amino-5-(1-methyl-5-nitro-
2-imidazolyl)-1,3,4-thiadiazole) and other 5-nitro imida-
zoles and nitrofurans (Fig. 1) were synthesized from the
corresponding 2-cyano-5-nitroheterocycle by condensation,
in trifluoroacetic acid, with thiosemicarbazide to generate
thethiadiazole ring or with semicarbazide for theoxadiazole
ring following a method previously described [26].
Melting points were determined on an electrothermal
capillary melting point apparatus and were uncorrected.
1H and 13C NMR spectra were recorded, respectively
on Bruker AC80 (80 MHz) and AC200 (50 MHz). All
chemical shifts are reported in ppm relative to TMS as
internal standard. Mass spectra were run on a Nermag
R1010 spectrometer at P. Sabatier University. All chemicals
were purchased from Aldrich or Fluka and were used
without extra purification.
˚
DMF and DMSO were dried by stirring over 4 A
molecular sieves. THF was distilled over sodium in the
presence of benzophenone ketyl as indicator prior to use.
Reactions and column chromatographic separations were
followed by thin layer chromatography using silicagel
(
with 254 nm fluorescent indicator) on aluminum plates.
2
. Materials and methods
2.2. Megazol: 1-methyl-2-(5-amino-1,3,4-thiadiazole)-
5-nitroimidazole
2
.1. Chemistry
Nifurtimox(3-methyl-N-[(5-nitro-2-furanyl)-methylene]-
-thiomorpholinamine-1,1-dioxide) was provided by Bayer
Thiosemicarbazide (2.1 g, 23 mmol) mixed with 1-
methyl-2-cyano-5-nitroimidazole (3.28 g, 22 mmol) were
4
Fig. 1. Chemical structures of the nitroheterocyclic compounds.

J.D. Maya et al. / Biochemical Pharmacology 65 (2003) 999–1006
1001
1
heated at 608 in trifluoroacetic acid (10 mL) for 15 hr. The
resulting solution was poured into ice water (20 mL) and
neutralized with sodium hydrogenocarbonate. The yellow
precipitate formed was filtered on a sintered glass filter,
washed with water (2 Â 5 mL), and dried under vacuum,
in acetone/water 1/1 (300 mg, 90% yield) H NMR
(DMSO-d , 80 MHz): 4.54 (s, NCH ); 8.17 (s, H4). 13C
NMR (DMSO-d , 50 MHz): 35 (NCH ); 117 (q, CF3);
6
3
6
3
0
132.9 (C4); 140 (C2); 141 (C2 ); 167.7 (CO); 170 (CS).
þ
Mass spectrometry: (DCI/NH ) m/z 323 (M þ 1, 60%)
3
þ
to give after crystallization in acetone, megazol (2.5 g,
340 (M þ NH , 50%).
3
1
1
8
1
1 mmol, 50% yield); m.p. 2708; H NMR (DMSO-d ,
6
0 MHz): 4.32 (s, 3H, NCH ); 7.8 (b, 2H, NH ); 8.2 (s,
1
2.6. 2-(5-Amino-1,3,4-oxadiazolyl)-5-nitrofuran
(see Fig. 1, compound 4)
3
2
3
H). CNMR(DMSO-d ,50 MHz):35.00(s,NCH );133.1
3
6
Ã
Ã
(
C4);140.1 (C5); 141.4 (C2); 148.2(C2 );169.9(C5 ). Mass
þ
spectrometry: (EI) m/z 227 (M þ 1, 80%). Anal. calc. for
2-Carbonitrile-5-nitrofuran (700 mg, 5 mmol) and semi-
carbazide hydrochloride (560 mg, 5 mmol) diluted in tri-
fluoroacetic acid (20 mL) were stirred at 608 for 15 hr;
then the cold solution (08) was carefully neutralized with
sodium bicarbonate, and the orange solid obtained was
filtered, washed with water and dried under vacuum, to
C H N O S:C,31.53;H,2.67;N,37.15.Found:C,31.64;H,
6
6 6 2
2
.62; N, 36.41.
2.3. 1-Methyl-2-(5-amino-1,3,4-oxadiazolyl)-5-
nitroimidazole (see Fig. 1, compound 1)
1
give compound 4 (450 mg, 45% yield); m.p. 2408, H
2-Cyano-1-methyl-5-nitroimidazole (600 mg, 4 mmol)
and semicarbazide hydrochloride (500 mg, 4.5 mmol) were
mixedin 10 mL oftrifluoroaceticacid andrefluxedat708 for
NMR (DMSO-d , 80 MHz): 6.33 (s, NCH ); 7.30 (d, H4);
6
2
7.80 (d, H3); 13C NMR (DMSO-d , 50 MHz): 111.3 (C3);
6
0
140 (C2); 114.9 (C4); 133.5 (C2); 150.6 (C5); 150.8 (C 2);
0
1
5 hr. The solution was then poured into ice water (10 mL)
157.0 (C 5). Mass spectrometry: (DCI/NH ) m/z 231
3
þ
and carefully neutralized with sodium bicarbonate. The
yellow solid formed was filtered and washed with diethyl
ether and water. After crystallization in an ethanol/water
(M þ 2NH , 100%).
3
2.7. 2-Formyl-5-nitrofuran-semicarbazine
(see Fig. 1, compound 5)
mixture (2/1 v/v) compound 1 was obtained (420 mg, 50%
1
yield); m.p. 2888, H NMR (DMSO-d , 80 MHz): 4.25 (s,
6
1
3
NCH ); 7.94 (s, NH ); 8.11 (s, H4); C NMR (DMSO-d6,
50 MHz): 34.53 (NCH ); 130.65 (C4); 140.05 (C5); 141.34
Semicarbazide hydrochloride (444 mg, 4 mmol) dis-
solved in DMSO (5 mL) was added to sodium hydride
(110 mg, 4 mmol) suspended in DMSO (3 mL). After
5 min, when all hydrogen had evolved, 5-nitrofuran-2-
carboxaldehyde (570 mg, 4 mmol) diluted in DMSO
(2 mL) was added. After 18 hr at room temperature, the
brown solution was poured into water (50 mL). The dark
precipitate was filtered, and then purified after dissolving in
acetone by flash chromatography (CH Cl /methanol: 90/
3
2
3
0
0
(
C2); 157 (C2 ); 159.77 (C5 ). Anal. calc. for C H N O : C,
6
6 6 3
34.29; H, 2.88; N, 39.99. Found: C, 33.79; H, 3.05%.
2.4. 1-Methyl-2-(5-acetamido-1,3,5-thiadiazolyl)-5-
nitroimidazole (see Fig. 1, compound 2)
Acetyl chloride (2 mL) was added by syringe through a
septum cap on megazol (226 mg, 1 mmol) to 30 mL of
anhydrous THF under argon. The mixture was stirred at
room temperature for 3 hr. Then the THF was evaporated
under vacuum. The residual solid was washed with water,
then recrystallized in ethanol/acetone 2/1, to obtain com-
pound 2 (160 mg, 60% yield). 1H NMR (DMSO-d6,
2
2
10) to give compound 5 (400 mg, 55% yield); m.p. 2408,
1
H NMR (DMSO-d , 80 MHz): 6.57 (s, NH); 7.22 (d, H4);
6
7.77 (d, H3); 10.76 (s, CH ¼¼ N). 3C NMR (DMSO-d6,
1
50 MHz): 112.0 (C4); 115.1 (C3); 127.3 (CH ¼¼ N); 151.1
(C2); 150.0 (C5); 155.8 (CONH ). Mass spectrometry:
2
þ
(DCI/NH ) m/z 216 (M þ 18, 100%).
3
8
H4); C NMR (DMSO-d , 50 MHz): 22.33 (CH CO);
0 MHz): 2.24 (s, COCH ); 4.39 (s, NCH ); 8.25 (s,
3 3
1
3
2.8. Assays on cultures
6
3
3
1
5.23 (CH N); 133.09 (C4); 140.6 (C2); 141.04 (C5);
3
0
0
53.93 (CONH); 159.74 (C2 ); 169.17 (C ).
2.8.1. Parasites and culture growth inhibition
6
A suspension of 3 Â 10 T. cruzi epimastigotes per mL,
2.5. 1-Methyl-2-(5-trifluoromethylacetamido-1,3,5-
thiadiazolyl)-5-nitroimidazole (see Fig. 1, compound 3)
from Tulahu e´ n and LQ strains and CL-Brener clone, were
cultured at 288 in monophasic Diamond’s culture medium
supplemented with 4 mM of hemin [27]. Inactivated bovine
calf serum at a final concentration of 4% was added. Drugs
dissolved in DMSO (final concentration of 1%) were added
to the culture media to give 0.5–100 mM final concentra-
tions. Parasite growth was followed by nephelometry for
7–10 days [28,29]. Nephelometry readings were directly
proportional to the concentration of parasites. No toxic
effect of DMSO alone was observed.
To megazol (226 mg, 1 mmol) suspended in THF (5 mL)
was added trifluoracetic anhydride (2 mL); the suspension
dissolved instantly. After 5 min at room temperature,
THF and trifluoroacetic anhydride were evaporated under
vacuum, and a saturated solution of sodium hydrogeno-
carbonate was added (3 mL) and then acetone (5 mL).
The yellow solid formed was filtered and recrystallized

1002
J.D. Maya et al. / Biochemical Pharmacology 65 (2003) 999–1006
The growth culture constant (k ) for each drug concen-
c
excitation wavelengths were 480 and 385 nm, respectively.
The method detects reduced free thiols, but not disulfides
[31].
tration employed was calculated using an exponential
epimastigote growth curve (regression coefficient > 0:97,
P < 0:05). The slope resulting from plotting the natural
logarithm (Ln) of the nephelometry reading vs. time, in
2.8.4. Statistical analysis
ꢁ1
hours, corresponds to the k (hr ). The ic
is defined as
the drug concentration needed to diminish the k in 50%;
Lineal regression analysis, correlation analysis, and
ANOVA and Dunnett’s multiple comparison tests were
performed using Prism GraphPad software (GraphPad
Software Inc.).
c
kc50
c
it is calculated by lineal regression analysis from the k
6
c
values and the concentrations used; 80 Â 10 parasite cells
correspond to 1 mg protein or 12 mg of fresh weight.
2.8.5. Electrochemical measurements
Mixed media cyclic voltammetry (DMF:citrate buffer/
2
.8.2. Oxygen uptake
Tulahu e´ n strain T. cruzi epimastigotes were harvested by
00 g centrifugation, followed by washing and re-suspen-
sion in 0.05 M sodium phosphate buffer, pH 7.4, contain-
ing 0.107 M sodium chloride. Respiration measurements
were carried out polarographically with a Clark no. 5331
electrode (Yellow Springs Instruments) in a 53 YSI model
60:40) was performed employing KCl 0.3 M and ITBA
0.1 M, pH ¼ 9, as support electrolytes. A three-electrode
BAS 50 W potentiostat system was employed using a
mercury drop, Ag/AgCl, and platinum wire as working,
reference and auxiliary electrodes, respectively [32,33].
The solutions were degassed with nitrogen for 10 min
before each measurement at 258.
5
(
2
Simpson Electric Co) [30]. The chamber volume was
mL and the temperature was 288. The amount of parasite
used was equivalent to 2 mg of protein. In order to evaluate
redox cycling, mitochondrial respiration was inhibited
with 20 mM sodium cyanide. The ickc50 equivalent con-
centration corresponds to the final concentration used in
the oxygen uptake experiments. This concentration was
calculated considering that the ickc50 (culture growth
2.8.6. Determination of k the second-order decay rate
2
constant of the nitro radical anion
The return to forward peak current ratio Ipa/Ipc for the
ꢀ
ꢁ
reversible first electron transfer (RNO /RNO2 ) is mea-
2
sured for each cyclic voltammogram, varying the scan rate
from 0.1 to 10 V/s according to the procedure previously
described [34]. Then a theoretical approach is applied to
6
experiments) was determined using 3 Â 10 parasites/
mL, equivalent to 0.0375 mg protein/mL as initial parasite
6
determine k value which involve the drug concentration in
2
mass; 80 Â 10 parasites/mL, equivalent to 1 mg protein/
the radical stability [35].
mL, was used in the oxygen uptake experiments. In order to
maintain the parasite mass-drug ratio constant in these two
types of experiments, the original ickc50 was corrected by
this 26-fold parasite mass increase in the oxygen uptake
experiments. Values are expressed as mean Æ SD for three
independent experiments. No effect of DMSO alone was
observed.
3. Results and discussion
The nitroheterocyclic agents employed in this study are
grouped into three different chemical families (Fig. 1):
(i) 2-nitroimidazoles (benznidazole as prototype), (ii) 5-
nitroimidazoles (megazol as prototype), and (iii) nitrofur-
2
.8.3. HPLC glutathione and trypanothione measurement
To form glutathione and trypanothione fluorescent deri-
ans (nifurtimox as prototype). The ick 50 values for each
drugs and strain are given in Table 1. All drugs tested
c
vatized adducts, Tulahu e´ n strain epimastigotes, equivalent
to 1 mg protein, were suspended in 40 mM HEPPS, 2 mM
EDTA buffer, pH 8, with 2 mM monobromobimane and
incubated at 708 for 5 min. To precipitate proteins, 4 M
methanesulfonic acid was added. After 10 min incubation
at 48, the sample was centrifuged at 10,000 g for 3 min. A
Table 1
Culture growth inhibition by nitro derivative compounds
Drug
ick 50
c
Tulahu e´ n (mM)
LQ (mM)
Brener (mM)
2
(
1
0 mL sample was applied to a Lichrospher 100 RP-1 8
5 mm) reverse phase column and eluted during 60 min at a
mL/min flow. The eluting gradient was: 0–20 min, 90%
of solvent A (0.25% w/v lithium a-camphorsulfonate, pH
.35) and 10% solvent B (25% w/v 1-propanol in solution
A); 20–40 min, linear gradient of 10–50% solvent B; from
0 to 60 min, 50% solvent B isocratic in solvent A,
Benznidazole
11.44 Æ 0.1
10.53 Æ 1.3
3.75 Æ 0.1
38.11 Æ 0.1
7.46 Æ 0.3
34.16 Æ 2.9
9.91 Æ 0.2
74.99 Æ 1.3
12.26 Æ 6.0
16.95 Æ 0.7
10.30 Æ 1.1
4.24 Æ 0.6
67.30 Æ 2.9
10.29 Æ 0.1
–
12.77 Æ 1.1
10.01 Æ 0.4
2.91 Æ 0.2
32.27 Æ 2.7
7.44 Æ 0.7
31.16 Æ 0.5
10.44 Æ 0.3
80.17 Æ 6.0
13.13 Æ 0.7
RO150216
Megazol
1
2
2
3
Nifurtimox
12.28 Æ 0.2
>100
4
4
5
returning to initial conditions in 30 min. All measures
were performed using a Merck-Hitachi high-pressure chro-
matograph L-6200 Intelligent Pump, with a F-1050 fluor-
escence detector and a D-2500 integrator. Emission and
20.16 Æ 2.1
The ick 50 values for each drug compound studied in the Tulahu e´ n, and
LQ strains and clone Brener of Trypanosoma cruzi. Values are expressed as
c
mean Æ SD for three independent experiments. For details, see Section 2.

J.D. Maya et al. / Biochemical Pharmacology 65 (2003) 999–1006
1003
produced an inhibitory effect against T. cruzi epimasti-
gotes, but differences in potency were observed. Megazol
was the most potent, but only three times more potent than
nifurtimox depending on the parasite strain used. The ickc50
values were slightly higher for the LQ than for the Tula-
hu e´ n strains or for the CL-Brener clone. The 2-nitroimi-
dazole RO150216 shows growth inhibition potency similar
to benznidazole. Albeit, it has a good inhibitory activity in
the three strains tested. These results are consistent with
previously reported T. cruzi strains susceptibility patterns
and 5 produced an increase of oxygen uptake, more visible
upon cyanide inhibition, indicating a clear redox cycling
(P < 0:01). Megazol, benznidazole and its derivatives, did
not have any effects on redox cycling as previously
reported [2,37].
In contrast, megazol and drugs 2 and 3 inhibited oxygen
uptake (P < 0:01) in a time-dependent fashion (Fig. 2).
2-Nitroimidazoles did not modify the epimastigote oxygen
uptake. Similar results were found with other strains (data
not shown). It is important to note that the use of ickc50
equivalent drug concentrations validates a possible rela-
tionship between oxygen redox cycling or respiration
inhibition with the parasite growth inhibition.
[18,13].
The antichagasic drugs acts upon the nonreplicative
trypomastigote and the intracellular amastigote forms. In
this study we used the noninfective epimastigote form, as
has been done in several other studies with other nitroi-
midazole and nitrofuran derivatives as well as other natural
and synthetic potential trypanocidal agents, which have
also proved to be efficient against the trypomastigote and
amastigote forms [36]. Moreover, the epimastigote form is
a suitable model for studying the mode of action of the
compounds employed in the present study. Since the low
molecular weight thiol content in the trypomastigote and
amastigote forms of T. cruzi is lower than in the epimas-
tigote form [14], should render these compounds more
effective in the infective and intracellular replicative forms.
It has been proposed that nifurtimox and, possibly,
benznidazole, can induce redox cycling, which could
explain their toxic effects, through the production of super-
oxide radical anion and then hydroxyl radicals. To evaluate
this possibility, we performed oxygen uptake experiments
with or without cyanide, which is a mitochondrial respira-
tion inhibitor. Table 2 shows Tulahu e´ n epimastigote oxy-
gen uptake, with or without 20 mM sodium cyanide, in the
presence of ickc50 equivalent drug concentration. When
compared with controls, only nifurtimox and derivatives 4
Reduction cathodic peak potential (E , Table 2) indicate
pc
one-electron transfer to nitro groups to form nitro-anion
radicals, while k values (Table 2) indicate radical decay,
that is, its stability, according to:
2
Epc
ꢀ
ꢁ
2
RNO þ e!RNO
2
k2
ꢀ
ꢁ
þ
2
RNO2 þ 2H !RNO þ RNO þ H O
2
2
Under anaerobic conditions, the electrochemically
formed anion radical dismutates more rapidly in the 5-
nitroimidazole group of megazol than in the nitrofuran of
nifurtimox, whereas 2-nitroimidazoles show intermediate
dismutation kinetics (Table 2). Thus, there may be a
relationship between the biological activity of megazol
and nitroso derivative formation, which is reflected here
through dismutation (Table 2).
There was no correlation between electrochemical
potential and growth inhibition. However, the electroche-
mical potentials of nitrofurans (Table 2) were directly
correlated with oxygen uptake under mitochondrial
respiratory inhibition by cyanide (Pearson’s correlation
coefficient ꢁ0.99). This is in agreement with their redox
Table 2
Nitroderivatives: effect upon oxygen uptake, redox cycling, bioavailability, electrochemical potential and dismutation ability
a,b
a,b,c
Drug
Respiration
Oxygen redox cycling
Bioavailability
Cyclic voltammetry
d
e
f
Log P
E
pc (mV)
k
2
(L/mol/s)
Control
31.3 Æ 1.7
26.6 Æ 7.9
29.0 Æ 5.5
18.3 Æ 2.1
28.1 Æ 0.4
2.1 Æ 0.5
2.3 Æ 0.1
1.5 Æ 0.1
2
Benznidazole
1.2
0.93
ꢁ730
ꢁ740
ꢁ733
ꢁ742
ꢁ758
ꢁ753
ꢁ630
ꢁ617
ꢁ594
8.9 Â 10
7.0 Â 10
1.9 Â 10
7.8 Â 10
3.3 Â 10
2.7 Â 10
5.3 Â 10
2.3 Â 10
8.9 Â 10
2
3
2
3
3
1
2
2
RO150261
Megazol
*
*
*
0.5 Æ 0.5
0.22
1
1.2 Æ 0.5
1.9 Æ 0.3
2.4 Æ 0.3
ꢁ0.37
0.17
1.69
*
2
26.2 Æ 5.7
8.2 Æ 0.5
*
*
*
3
*
**
Nifurtimox
40.0 Æ 0.8
26.2 Æ 3.7
34.1 Æ 6.0
4.0 Æ 1.5
ꢁ0.34
ꢁ0.42
0.23
**
4
5
6.0 Æ 0.4
*
*
6.6 Æ 0.2
The values correspond to the mean Æ SD of three independent experiments. For details, see Section 2.
a
Final drug concentrations used correspond to the ick 50 equivalent concentrations as defined in Section 2.
Respiration and oxygen uptake values are expressed as nanoatoms-gram of oxygen/min/mg protein.
c
b
c
Sodium cyanide (20 mM) was added as well as the respective drug at the ick 50 equivalent concentration.
Bioavailability is expressed as log P, calculated using the demo SRC’s LogKow (KowWin) Program software [40].
c
d
e
f
Epc corresponds to the reductioncathodic peak potential expressed as milivolts and measured as described in Section 2.
Dismutation constant (k ) is expressed as L/mol/s. See Section 2.
2
**
Dunnett’s multiple comparison test: P < 0:05; Dunnett’s multiple comparison test: P < 0:01.
*

1004
J.D. Maya et al. / Biochemical Pharmacology 65 (2003) 999–1006
Fig. 2. Time effect of prototype nitroderivatives upon T. cruzi epimastigotes, Tulahu e´ n strain respiration. Control (continuous line, open circles), nifurtimox
dotted line, closed triangles), megazol (dotted line, closed circles), drug 3 (continuous line, and open squares). Drug 2 (not shown) followed the same pattern
(
than drug 3. Values correspond to mean Æ SD of three independent experiments and expressed as nanoatom gram of oxygen/min/mg protein. Percent
variation with respect to the control is shown in the right axis.
cycling behavior. Considering that megazol and its deri-
vatives are poor oxygen redox cyclers and their respiration
inhibitory effect is time-dependent (Fig. 2), it may be
assumed that their action is mediated by electrophilic
When homogenates of drug-treated parasites were incu-
bated with an active T. cruzi’s trypanothione reductase
preparation and further derivatized with monobromobi-
mane followed by separation of the reaction products by
HPLC, no significant increase in the reduced thiol con-
centrations shown in Fig. 4 was observed. Thus, indicating
that the loss of reduced thiols was not due to oxidation,
but rather to conjugation with electrophilic metabolites
[14]. Appropriate controls showed no trypanothione or
glutathione degradation by T. cruzi homogenates.
Nifurtimox can produce anion radicals and interfere
with oxygen metabolism, as shown in Fig. 2. Megazol
has a more complex behavior. In the first place, it inhibits
mitochondrial respiration in a time-dependent fashion, and
does not induce an increase in oxygen uptake, even when
tested with cyanide. Second, it produces considerable
depletion of T(SH)2 and to a lesser extend, of GSH
(Fig. 3). The same trend has been seen in other strains
(data not shown). Finally, comparison of results obtained
with the similar compounds, megazol and 1 reveals another
important feature: the bioavailability of the drug.
metabolites; based on their faster dismutation (k values,
2
Table 2) and possibly also through a second electron
transfer to the primary radical anion. This greater ability
of megazol to form the nitroso species more rapidly than
nifurtimox has also been shown using another technique,
pulse radiolysis [38]. Such electrophilic metabolites may
bind to epimastigote intracellular macromolecules. To test
this, we determined the intracellular glutathione (GSH)
and trypanothione (T(SH )) thiol contents of the Tulahu e´ n
2
strain.
In the presence of ickc50 equivalent concentrations,
megazol elicited a progressive decrease of the thiol content
in 2 hr (Fig. 2), with a half-life of 88 min for T(SH)2
(regression coefficient 0.85). GSH did not decrease as
much as T(SH) , and its half-life was >200 min. All the
2
tested nitroheterocyclic compounds induced a decrease of
thiol content; T(SH) was more affected than GSH (Fig. 4).
2
The ability of nitrofurans to decrease thiols was compar-
able to that of nifurtimox, as previously reported [14,39],
and to that of 2-nitroimidazoles and benznidazole as shown
in the present paper. Moreover, the RO150216 compound
presented a very high thiol scavenging capacity compared
with all the compounds tested. However, it is similar to
benznidazole’s thiol scavenging capacity [14]. In general,
RO150261 behaves similarly as benznidazole in growth
inhibition potency, respiration inhibition and dismutation
capacity as well as bioavailability. No difference in GSH
content was observed when 4 and 5 were compared to
controls (P > 0:05).
Indeed both of these compounds have similar properties
with respect to their ability to produce the nitroso form by
dismutation; compound 1 appears to be an even more
efficient thiol scavenger than megazol (Fig. 4). By contrast,
megazol is at least 10 times more efficient with respect to
inhibition of culture growth than compound 1 (Table 1).
Actually, the only structural difference between the two
compounds is the replacement of an oxygen atom in 1 by a
sulfur atom in megazol; this change is enough to raise the
log P value from ꢁ0.37 for compound 1 to 0.22 in megazol
[40] (Table 2). It is well known that compounds with nega-
tive log P values cross cell membranes very poorly [41].

J.D. Maya et al. / Biochemical Pharmacology 65 (2003) 999–1006
1005
Fig. 3. Effect of megazol upon thiol content in T. cruzi epimastigotes. Reduced trypanothione T(SH)
2
regression coefficient 0.85, P < 0:05; controls and
GSH-megazol regression coefficients <0.5, P > 0:05. Trypanothione half-life is 88 min. T(SH)
GSH-megazol (*).
2
control (&), T(SH) megazol (*), GSH control (~),
2
A specific feature of megazol might therefore be related to
its greater bioavailability. As for compounds 2 and 3,
which also have positive log P values (0.17 and 1.69,
respectively), compound 2 is a prodrug of megazol after
hydrolysis of the acetyl group, whereas compound 3 is not
since the trifluoroacetyl group is only poorly hydrolyzed by
cytosolic peptidases (not shown).
In conclusion, this work sheds some light on the differ-
ences between nitrofurans, exemplified by nifurtimox and
nitroimidazoles, with megazol as a representative struc-
ture. Whereas the former compounds are essentially active
as redox cycling agents, leading to oxidative stress (upper
part on Fig. 2), the 5-nitroimidazoles are converted to the
nitroso form, resulting from dismutation, as shown here,
possibly also via a second one-electron reduction (lower
part in Fig. 2). The resulting electrophilic species are
Therefore, the action of both classes of compounds
involves oxygen metabolism, in the case of nitrofurans
by inducing oxidative stress, and by trypanothione deple-
tion for nitroimidazoles.
Furthermore, the comparison of the effects produced by
similar compounds suggests an additional feature respon-
sible for the high efficacy of megazol, namely its suitable
amphiphilic balance. On these grounds, other nitroimida-
zoles are currently under investigation, as well as the
ability to act as thiol scavenger of host glutathione, toxicity
and in vivo assays for the lead compound megazol.
Acknowledgments
This research was supported in part by SIDA/SAREC,
Fondecyt-Chile Grant Number 1000882 and 1020095,
ECOS-Conicyt C98S01 and CNRS-Conicyt.
We thank professor Dr. Bruce Cassels for suggestions
and critical review of the manuscript.
efficient thiol scavengers, particularly for T(SH) , the
2
cofactor of the detoxification enzyme trypanothione reduc-
tase, which has been shown to be essential for the survival
of the trypanosome [42].
References
[
[
1] Moncayo A. Chagas’ disease. Tropical disease research: progress
991–1992. Geneva: WHO; 1993. p. 67–76.
1
2] Castro JA, Diaz de Toranzo EG. Toxic effects of nifurtimox and
benznidazole, two drugs used against trypanosomiasis (Chagas’ dis-
ease). Biomed Env Sci 1988;1:19–33.
[
3] Gorla NB, Ledesma OS, Barbieri GP, Larripa IB. Thirteen-fold increase
of chromosomal aberrations nonrandomly distributed in chagasic chil-
dren treated with nifurtimox. Mutat Res 1989;224:263–7.
4] Filardi LS, Brener ZA. Nitroimidazole-thiadiazole derivative with
curative action in experimental Trypanosoma cruzi infections. Ann
Trop Med Parasitol 1982;76:293–7.
[
Fig. 4. Effect of nitroderivatives upon GSH (black bars), T(SH)
2
(white
[5] Docampo R, Stoppani AOM. Generation of superoxide anion and
hydrogen peroxide induced by nifurtimox in Trypanosoma cruzi. Arch
Biochem Biophys 1979;1911:317–21.
bars), and total thiol (scattered bars) content in T. cruzi epimastigotes,
Tulahu e´ n strain. Drug concentrations were equivalent to their respective
ick 50, and the incubation time was 2 hr before thiol determination. Values
are expressed as GSH nmol/g wet weight of parasites. Asterisks indicate no
[6] Docampo R, Moreno SNJ, Stoppani AOM, Leon W, Cruz FS, Villalta
F, Muniz RFA. Mechanism of nifurtimox toxicity in different forms of
Trypanosoma cruzi. Biochem Pharmacol 1981;30:1947–51.
c
statistical difference with control in Dunnett’s test.

1006
J.D. Maya et al. / Biochemical Pharmacology 65 (2003) 999–1006
[
7] Docampo R, Moreno SNJ. Free-radical intermediates in the antipar-
[25] Bouteille B, Chauvi e` re G. Implication du m e´ gazol dans la chimioth e´ r-
apie des trypanosomoses. Med Trop 1999;59:321–30.
asitic action of drugs and phagocytic cells. In: Pryor WA, editor.
Free radicals in biology, vol. 6. Academic Press: New York; 1984.
p. 243–88.
[26] Chauvi e` re G, Viod e´ C, P e´ ri e´ J. Nucleophilic substitution studies on
nitroimidazoles and applications to the synthesis of biologically active
compounds. J Heterocyclic Chem 2000;37:119–26.
[
8] Docampo R, Stoppani AOM. Generation of oxygen-reduction
derivatives induced by nifurtimox and other nitrocompounds in Trypa-
nosoma cruzi. Medicina (Buenos Aires) 1980;40(Suppl 1):10–6.
9] Dubin M, Moreno SNJ, Martino EE, Docampo R, Stoppani AOM.
Increased biliary secretion and loss of hepatic glutathione in rat liver
after nifurtimox treatment. Biochem Pharmacol 1983;32:483–7.
[27] Aldunate J, Ferreira J, Letelier MJ, Repetto Y, Morello A.
t-Butyl-4-hydroxyanisole,
a novel respiratory chain inhibitor.
[
Effects on Trypanosoma cruzi epimastigotes. FEBS Lett 1986;195:
295–7.
[28] Ferreira J, Coloma L, Fones E, Letelier ME, Repetto Y, Morello A,
Aldunate J. Effects of t-butyl-4-hydroxyanisole and other phenolic
antioxidants on tumoral cells and Trypanosoma parasites. FEBS Lett
1988;234:485–8.
[
10] Moreno SNJ, Docampo R, Mason RP, Leon W, Stoppani AOM.
Different behaviors of benznidazole as free radical generator with
mammalian and Trypanosoma cruzi microsomal preparations. Arch
Biochem Biophys 1982;218:585–91.
[29] Aldunate J, Traipe L, Spencer P, Morello A, Repetto Y. Effects of
hydroquinones on intact Trypanosoma cruzi epimastigotes. Comp
Biochem Physiol 1992;103C:97–100.
[
11] Masana M, Toranzo EGD, Castro JA. Reductive metabolism and
activation of benznidazole. Biochem Pharmacol 1984;33:1041–5.
12] Diaz de Toranzo EG, Castro JA, Franke de Cazzulo BM, Cazzulo JJ.
Interactions of benznidazole reactive metabolites with nuclear and
kinetoplasmic DNA, proteins and lipids from Trypanosoma cruzi.
Experientia 1988;44:880–1.
[
[30] Letelier ME, Rodriguez E, Wallace A, Lorca M, Repetto Y, Morello A,
Aldunate J. Trypanosoma cruzi: a possible control of transfusion
induced Chagas disease by phenolic antioxidants. J Exp Parasitol
1990;71:357–63.
[
13] Aldunate J, Morello A. Free radicals in the mode of action of
antiparasitic drugs. In: Aruoma OI, editor. Free radicals in tropical
diseases. Harwood Academic: Switzerland; 1993. p. 137–65.
14] Maya JD, Repetto Y, Agos ´ı n M, Ojeda JM, Tellez R, Gaule C, Morello
A. Effects of nifurtimox and benznidazole upon glutathione and
trypanothione in epimastigote, trypomastigote, and amastigote forms
of Trypanosoma cruzi. Mol Biochem Parasitol 1997;86:101–6.
15] Berkelhammer G, Asato G. 2-Amino-5-(1-methyl-5-nitro-2-imida-
zolyl)-1,3,4-thiadiazole: a new antimicrobial agent. Science 1968;
[31] Fairlamb AH, Henderson GB, Bacchi CJ, Cerami A. In vivo effects of
difluoromethylornithine on trypanothione and polyamine levels in
bloodstream forms of Trypanosoma brucei. Mol Biochem Parasitol
1987;24:185–91.
[
[32] N u´ n˜ ez-Vergara L, Bollo S, Alvarez AF, Bl a´ zquez M, Squella JA. Nitro
radical anion formation from nimodipine. Electroanal Chem
1993;345:121–34.
[
[
[
[33] Squella JA, Bollo S, de la Fuente J, N u´ n˜ ez-Vergara LJ. Electroche-
mical study of the nitro radical anion from nicardipine: kinetic
parameters and its interaction with glutathione. Bioelectrochem Bioe-
nergetics 1994;34:13–8.
1162:1146.
16] Filardi LS, Brener Z. Susceptibility and natural resistance of Trypa-
nosoma cruzi strains to drugs used clinically in Chagas disease. Trans
R Soc Trop Med Hyg 1987;81:755–9.
[34] Bollo S, Nu n˜ ez-Vergara LJ, Bonta M, Chauvi e` re G, P e´ ri e´ J, Squella
JA. Cyclic voltammetric studies on nitro radical anion formation from
megazol and some related nitroimidazole derivatives. J Electroanal
Chem 2001;511:46–54.
17] Bouteille B, Marie-Daragon A, Chauvi e` re G, De Albuquerque C,
Enanga B, Darde ML, Vallat M, P e´ ri e´ J, Dumas M. Effect of megazol
in Trypanosoma brucei brucei acute and subacute infection in Swiss
mice. Acta Trop 1995;60:73–80.
[35] Olmstead ML, Nicholson RS. Cyclic voltammetry theory for the
disproportionation reaction and spherical diffusion. Anal Chem
1969;41:862–4.
[
18] Enanga B, Keita M, Chauvi e` re G, Dumas M, Bouteille B. Megazol
combined with suramin: a chemotherapy regimen which reversed the
CNS pathology in a model of human African trypanosomiasis in mice.
Trop Med Int Health 1998;3:736–41.
[36] Rodrigues Coura J, De Castro SL. A critical review on Chagas disease
chemotherapy. Mem Inst Oswaldo Cruz 2002;97:3–24.
[37] Tsuhako MH, Alves MJ, Colli W, Brener Z, Augusto O. Restricted
bioreductive metabolism of a nitroimidazole-thiadiazole derivative
with curative action in experimental Trypanosoma cruzi infections.
Biochem Pharmacol 1989;38:4491–6.
[
19] Denise H, Mattews K, Lindergard G, Croft S, Barrett MP. Trypano-
somiasis and leishmaniasis: between the idea and the reality of control.
Parasitol Today 1999;15:43–5.
[
20] Jennings FW, Chauvi e` re G, Viod e´ C, Murray M. Topical chemother-
apy for experimental African trypanosomiasis with cerebral involve-
ment: the use of melarsoprol combined with the 5-nitroimidazole,
megazol. Trop Med Int Health 1996;1:363–6.
[38] Viod e´ C, De Albuquerque C, Chauvi e` re G, Hou e´ e-Levin C, P e´ ri e´ J.
Comparative study by pulse radiolysis of the radical anion derived
from compounds used in Chagas’ disease therapy. New J Chem
1997;21:1331–8.
[
[
[
[
21] Viod e´ C, Bettache N, Narimantas C, Krauth-Siegel RL, Chauvi e` re G,
Bakalara N, P e´ ri e´ J. Enzymatic reduction studies of nitroheterocycles.
Biochem Pharmacol 1999;57:549–57.
[39] Repetto Y, Opazo E, Maya JD, Agos ´ı n M, Morello A. Glutathione and
trypanothione in several strains of Trypanosoma cruzi. Effects of
drugs. Comp Biochem Physiol 1996;115B:281–5.
22] De Castro SL, Meireles M. Mechanism of action of a nitroimidazole-
thiadiazole derivative upon Trypanosoma cruzi tissue culture amas-
tigotes. Mem Inst Oswaldo Cruz 1990;85:95–9.
[40] On-line interactive demo SRC’s LogKow (KowWin) Program estab-
lished by the Environmental Science Center, Syracuse Research Cor-
poration, available on the Internet at: http://esc.syrres.com/interkow/
logkow.htm [accessed 16 September 2002].
23] Ferreira RC, Ferreira LC. CL 64,855, a potent anti-Trypanosoma cruzi
drug, is also mutagenic in the Salmonella/microsome assay. Mem Inst
Oswaldo Cruz 1986;81:49–52.
[41] Gago F, Alvarez-Builla J, Elguero J. Hydrophobicity measurements
by HPLC: a new approach to P constants. J Liq Chromatogr 1987;10:
1031–47.
24] Borowy NK, Nelson RT, Hirumi H, Brun R, Waithaka HK, Schwartz
D, Polak A. RO150216 a nitroimidazole compound active against
human and animal pathogenic African trypanosomes. Ann Trop Med
Parasitol 1988;82:13–9.
[42] Fairlamb AH, Cerami A. Identification of a novel thiol-containing
cofactor essential for glutathione reductase activity in trypanosoma-
tids. Mol Biochem Parasitol 1985;14:187–98.