Structural design, synthesis and structure-activity relationships of thiazolidinones with enhanced anti-Trypanosoma cruzi activity

Article abstract of DOI:10.1002/cmdc.201300354

Pharmacological treatment of Chagas disease is based on benznidazole, which displays poor efficacy when administered during the chronic phase of infection. Therefore, the development of new therapeutic options is needed. This study reports on the structural design and synthesis of a new class of anti-Trypanosoma cruzi thiazolidinones (4 a-p). (2-[2-Phenoxy-1-(4-bromophenyl) ethylidene)hydrazono]-5-ethylthiazolidin-4-one (4 h) and (2-[2-phenoxy-1-(4- phenylphenyl)ethylidene)hydrazono]-5-ethylthiazolidin-4-one (4 l) were the most potent compounds, resulting in reduced epimastigote proliferation and were toxic for trypomastigotes at concentrations below 10 μM, while they did not display host cell toxicity up to 200 μM. Thiazolidinone 4 h was able to reduce the in vitro parasite burden and the blood parasitemia in mice with similar potency to benznidazole. More importantly, T. cruzi infection reduction was achieved without exhibiting mouse toxicity. Regarding the molecular mechanism of action, these thiazolidinones did not inhibit cruzain activity, which is the major trypanosomal protease. However, investigating the cellular mechanism of action, thiazolidinones altered Golgi complex and endoplasmic reticulum (ER) morphology, produced atypical cytosolic vacuoles, as well as induced necrotic parasite death. This structural design employed for the new anti-T. cruzi thiazolidinones (4 a-p) led to the identification of compounds with enhanced potency and selectivity compared to first-generation thiazolidinones. These compounds did not inhibit cruzain activity, but exhibited strong antiparasitic activity by acting as parasiticidal agents and inducing a necrotic parasite cell death. Stop the cycle! The attachment of an aryl ring to the iminic carbon produced thiazolidinones that are conformationally more restricted than first-generation thiazolidinones. This enhanced the potency of antiparasitic thiazolidinones, as observed under treatment with compound 4 h where parasite development and invasion in host cells were substantially reduced. Copyright

Full text of DOI:10.1002/cmdc.201300354

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201300354
Structural Design, Synthesis and Structure–Activity
Relationships of Thiazolidinones with Enhanced Anti-
Trypanosoma cruzi Activity
Diogo Rodrigo Magalh¼es Moreira,*[a, e] Ana Cristina Lima Leite,[b]
Marcos Verissimo Oliveira Cardoso,[b] Rajendra Mohan Srivastava,[a]
Marcelo Zaldini Hernandes,[b] Marcelo Montenegro Rabello,[b] Luana Faria da Cruz,[c]
Rafaela Salgado Ferreira,[c] Carlos Alberto de Simone,[d] Cꢀssio Santana Meira,[e]
Elisalva Teixeira Guimaraes,[e, f] Aline Caroline da Silva,[g] Thiago Andrꢁ Ramos dos Santos,[g]
Valꢁria RÞgo Alves Pereira,[g] and Milena Botelho Pereira Soares[e, h]
Pharmacological treatment of Chagas disease is based on
benznidazole, which displays poor efficacy when administered
during the chronic phase of infection. Therefore, the develop-
ment of new therapeutic options is needed. This study reports
on the structural design and synthesis of a new class of anti-
Trypanosoma cruzi thiazolidinones (4a–p). (2-[2-Phenoxy-1-(4-
bromophenyl)ethylidene)hydrazono]-5-ethylthiazolidin-4-one
(4h) and (2-[2-phenoxy-1-(4-phenylphenyl)ethylidene)hydrazo-
no]-5-ethylthiazolidin-4-one (4l) were the most potent com-
pounds, resulting in reduced epimastigote proliferation and
were toxic for trypomastigotes at concentrations below 10 mm,
while they did not display host cell toxicity up to 200 mm. Thia-
zolidinone 4h was able to reduce the in vitro parasite burden
and the blood parasitemia in mice with similar potency to
benznidazole. More importantly, T. cruzi infection reduction
was achieved without exhibiting mouse toxicity. Regarding the
molecular mechanism of action, these thiazolidinones did not
inhibit cruzain activity, which is the major trypanosomal pro-
tease. However, investigating the cellular mechanism of action,
thiazolidinones altered Golgi complex and endoplasmic reticu-
lum (ER) morphology, produced atypical cytosolic vacuoles, as
well as induced necrotic parasite death. This structural design
employed for the new anti-T. cruzi thiazolidinones (4a–p) led
to the identification of compounds with enhanced potency
and selectivity compared to first-generation thiazolidinones.
These compounds did not inhibit cruzain activity, but exhibited
strong antiparasitic activity by acting as parasiticidal agents
and inducing a necrotic parasite cell death.
Introduction
Chagas disease, caused by Trypanosoma cruzi parasite infec-
tion, affects approximately 5–10% of the Latin American popu-
lation.[1,2] The standard treatment is based on benznidazole,
a compound able to eliminate the parasite during the acute
phase.[3] However, during the chronic phase, in which the para-
site remains located inside various cell types, benznidazole is
not able to eliminate the parasitism even following long-term
administration.[4] Human vaccination against T. cruzi infection is
not available,[5] and thus other therapies aiming to control in-
fection or reduce clinical symptoms are being investigated.[6–8]
Aiming at the development of more effective medicines,
a large number of anti-T. cruzi small molecules have been eval-
[a] Dr. D. R. M. Moreira, Prof. R. M. Srivastava
Universidade Federal de Pernambuco (UFPE)
Departamento de Quꢀmica Fundamental
50670-901, Recife, PE (Brazil)
[e] Dr. D. R. M. Moreira, C. S. Meira, Dr. E. T. Guimaraes, Dr. M. B. Pereira Soares
Fundażo Oswaldo Cruz (Fiocruz), Centro de Pesquisas Gonc¸alo Moniz
CEP 40296-710, Salvador, BA (Brazil)
[f] Dr. E. T. Guimaraes
E-mail: diogollucio@gmail.com
Universidade do Estado da Bahia (UNEB)
Departamento de CiÞncias da Vida
41150-000, Salvador, BA (Brazil)
[b] Prof. A. C. Lima Leite, Dr. M. V. O. Cardoso, Prof. M. Z. Hernandes,
M. M. Rabello
UFPE, Departamento de CiÞncias FarmacÞuticas
50740-520, Recife, PE (Brazil)
[g] A. C. da Silva, T. A. R. dos Santos, Dr. V. R. A. Pereira
FIOCRUZ, Centro de Pesquisas Aggeu Magalhaes
50670-420, Recife, PE (Brazil)
[c] L. F. da Cruz, Prof. R. S. Ferreira
Universidade Federal de Minas Gerais (UFMG)
Departamento de Bioquꢀmica e Imunologia
31270-901, Belo Horizonte, MG (Brazil)
[h] Dr. M. B. Pereira Soares
Centro de Biotecnologia e Terapia Celular, Hospital S¼o Rafael
41253-190, Salvador, BA (Brazil)
[d] Prof. C. A. de Simone
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201300354.
Universidade de S¼o Paolo (USP)
Departamento de Fꢀsica e Informꢁtica, Instituto de Fꢀsica
CEP 13560-970, S¼o Carlos, SP (Brazil)
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
1
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
uated. To date, compounds that
in some way inhibit either the
trypanosomal protease cruzain
or the ergosterol biosynthesis
are the most outstanding T. cruzi
growth inhibitors.[9–12] In fact,
some of these compounds have
exhibited strong activity in re-
ducing parasitemia in T. cruzi-in-
fected mice as well as low toxici-
ty in these animal models.[13–16]
Given this promising outlook,
the structural design of new
molecules based on these
T. cruzi molecular targets is an
attractive line of research.
Figure 1. SARs of the previously studied anti-T. cruzi agents. Thiazolidinone 18 was identified as the most potent
compound among them.
Hydrazones are well-known antiparasitic compounds.[17,18]
Based on this, our research group has been investigating cru-
zain-inhibiting hydrazones to obtain novel and potent anti-
T. cruzi agents. At least five distinct classes were investigated:
thiosemicarbazones,[19]
N-acylhydrazones,[20,21]
thiazolidi-
nones[22–24] and their transition metal complexes.[25] Within the
thiazolidinone class of compounds, we identified compound
18 after examining the importance of modifications at every
atom of the thiazolidinic ring (Figure 1). Compound 18 dis-
played an IC50 value of 10.1Æ0.09 mm to reduce the percent-
age of infected host cells, which is similar to the observed
value for benznidazole. However, thiazolidinone 18 was less ef-
ficient in reducing blood parasitemia in T. cruzi-infected mice
than benznidazole.[24] To identify new anti-T. cruzi thiazolidi-
nones with enhanced in vivo efficacy, we reasoned that some
molecular modifications in this class of compounds are neces-
sary.
Figure 2. Thiazolidinones previously investigated (top) and proposed here
(bottom). Crystal structures (Ortep-3) highlight the difference of molecular
planarity between the two series.
Previously, it was observed that thiazolidinone 18 inhibits
cruzain activity but not its homologous in mammalian cells
(cathepsin L).[24] However, it was several times less potent than
KB2, a high-efficient cruzain inhibitor.[13] A comparison of cru-
zain docking between 18 and KB2 has revealed that the major
difference is that KB2 assumes a T-shaped conformation. There-
fore, we hypothesized that thiazolidinones displaying a confor-
mation resembling the KB2 might enhance activity against cru-
zain, and consequently, against parasite cells. To achieve this, it
was necessary to disrupt the existent planarity between the
phenoxyl group and the thiazolidinic ring. Based on literature
findings,[26,27] the attachment of an aryl ring to the iminic
carbon should produce thiazolidinones with greater conforma-
tional restriction (i.e., high rotational energy barrier) than ob-
served in 18, and, consequently, produce thiazolidinones with
the desirable T-shaped conformation (Figure 2).
alkoxy, halogen atoms and six-membered rings. In addition,
compounds containing two or three substituents attached to
the phenyl ring were prepared. By evaluating the antiparasitic
activity of compounds 4a–p, we identified that thiazolidinones
had stronger anti-T. cruzi activity compared with the previously
reported thiazolidinone 18. More importantly, we found that
compounds 4h and 4l are as potent as benznidazole in the in-
hibition of T. cruzi infection in host cells. As expected, based
on its high in vitro antiparasitic activity, mice treated orally
with thiazolidinone 4h had substantially reduced blood parasi-
temia, similar to the observations made with benznidazole-
treated mice.
Results
Based on this structural design, new thiazolidinones denoted
as compounds 4a–p were synthesized and evaluated as anti-
parasitic agents as well as cruzain inhibitors. Initially, com-
pound 4a was prepared, which contains a phenyl ring at-
tached to the iminic carbon. To determine the importance of
this phenyl for the activity, compounds containing a pyridinyl
instead of a phenyl ring were synthesized. Thereafter, substitu-
ents attached to the phenyl ring were examined, such as, alkyl,
Synthesis
The route used for synthesizing thiazolidinones 4a–p is shown
in Scheme 1. The synthesis was initiated with the reaction be-
tween commercially available ketones and bromines, which
yielded a-bromoketones 1a–p.[28] These were reacted with
phenol in basic conditions to give the respective b-ketoethers
2a–p. To prepare biphenyl compound 2l, a Suzuki cross-cou-
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
2
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Scheme 1. Synthesis of thiazolidinones 4a–q. Reagents and conditions: a) Br2,
Et2O/dioxane (1:1), À58C, 5 h. b) Phenol, K2CO3, butanone, reflux, 5–10 h,
42–72%. c) Phenylboronic acid, Pd(OAc)2, P(Ph)3, MeOH, dioxane, 908C, 20 h,
50%. d) Thiosemicarbazide, EtOH, AcOH, reflux, 2 h, 43–83%. e) Ethyl 2-bro-
mobutyrate, KOAc, EtOH, reflux, 6–8 h, 40–81%. f) Ethyl 2-bromoacetate,
KOAc, EtOH, reflux, 6 h, 75%. Ph=phenyl, Py=pyridinyl.
pling reaction was performed between the bromophenyl 2h
and phenylboronic acid by using Pd(OAc)2 and P(Ph)3 as cata-
lysts.[29] Reactions of b-ketoethers 2a–p with thiosemicarbazide
under acidic conditions at reflux provided thiosemicarbazones
3a–p with yields varying from 43 to 83% and acceptable puri-
ties (approx. 95%) by simple precipitation following recrystalli-
zation. Cyclizations of thiosemicarbazones 3a–p to thiazolidi-
nones 4a–p were carried out with ethyl-2-bromobutyrate in
the presence of potassium acetate at reflux for six to eight
hours, and thiazolidinones 4a–p were isolated as colorless
solids with purities >98% determined by elemental analysis.
The compounds were chemically characterized by 1H and
Figure 3. A) Structures of the possible diastereoisomers and B) Ortep-3 rep-
resentation for the crystal structure of thiazolidinone 4q.
In the crystalline structure depicted in Figure 3, N2 and the
sulfur atom are on the same side, characterizing a Z configura-
1
tion for the C1=N2 bond. Thiazolidinone 4q showed H NMR
chemical shifts similar to thiazolidinones 4a–p; therefore, it is
fair to suggest that the major isomer for thiazolidinones 4a–p
has a Z configuration in regard to the C4=N1 bond. Important
to note is that all compounds were used in the pharmacologi-
cal tests as isomer mixtures.
1
13C NMR, IR and HRMS (ESI). The H NMR spectra showed that
thiazolidinones 4a–p are composed of two diastereomers, and
through analysis of the HPLC chromatograms, we determined
that the ratio of the isomer mixture 4a/4h is 90:10.
Antiparasitic activity against extracellular forms and
cytotoxicity against host cells
Next, we aimed to define the configuration of the major
isomer by crystallographic analysis. We did not succeed in crys-
tallizing thiazolidinones 4a–p suitable for X-ray analysis. Be-
cause previously we observed that thiazolidinic derivatives
without substituents attached to the heterocyclic ring are
more prone to afford crystals suitable for X-ray measurements,
we decided to prepare thiazolidinone 4q for X-ray analysis.
After recrystallization and chromatographic purification of 4q,
its major isomer was isolated, and a single crystal suitable for
X-ray analysis was collected. As shown in the Ortep-3 represen-
tation of thiazolidinone 4q, carbon C5 and nitrogen N2 are on
the same side, characteristic of a Z configuration in regard to
the C4=N1 bond (Figure 3). Interestingly, the phenyl ring at-
tached to the iminic carbon is coplanar to the thiazolidinic
ring, while the phenoxy ring is oriented to a different side. An-
other interesting observation is that the double bond of C1 is
located in an exocyclic position in relation to the heterocyclic
ring, which gives rise to an isomerism in the C1=N2 bond.[30–32]
First, compounds 4a–p were evaluated against epimastigotes
and trypomastigotes of T. cruzi. Antiparasitic activity was deter-
mined by counting the parasite number in a Neubauer cham-
ber and calculating the concentration of the test compound
resulting in 50% inhibition (IC50, epimastigotes) or 50% cyto-
toxicity (CC50, trypomastigotes). Cytotoxicity in host cells was
determined in mouse splenocytes, measured by the incorpora-
tion of [3H]-thymidine, and results were expressed as the high-
est non-cytotoxic concentration (HNC). Benznidazole was used
as reference antiparasitic drug and exhibited a CC50 value of
6.0 mm against trypomastigotes. As cut-off, compounds with
CC50 values ꢀ6.0 mm against trypomastigotes were considered
potent anti-T. cruzi compounds. The results are reported in
Table 1.
We first analyzed the antiparasitic activity against trypomas-
tigotes. The nonsubstituted phenyl derivative 4a was active
but was less potent than benznidazole. The replacement of
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
3
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
resulted in a fourfold increase in
activity and produced a deriva-
tive that was as potent as benz-
nidazole. In contrast, attaching
two (4i) or three chloro (4j) sub-
stituents decreased anti-T. cruzi
activity.
Table 1. Anti-T. cruzi activity, cytotoxicity and cruzain inhibition.
Compd
R
T. cruzi
Splenocytes
Cruzain
Inhibition
[%][d]
CC50 [mm][a]
IC50 [mm][b]
HNC [mm][c]
trypomastigotes
epimastigotes
4a
4b
4c
4d
17.7
25.0
>100
67.8
>283
28
64Æ2
5Æ5
Compound 4l, carrying a bi-
phenyl group, displayed an ac-
tivity similar to benznidazole-
treated parasites. In practice, the
4-bromophenyl derivative 4h
and biphenyl derivative 4l are
>100
>100
97.2
28
11Æ2
ND
>272
36.1Æ0.6
equipotent
antitrypanosomal
4e
4 f
4g
4h
4i
82.0
78.9
17.3
4.0
ND
>100
11.3
3.9
>261
>269
258
58Æ5
agents. The 4-phenoxy derivative
4m exhibited a CC50 value of
68.5 mm, being inactive in prac-
tice. In comparison to the non-
substituted phenyl derivative 4a,
the piperidinyl derivative 4n was
slightly more active. The same
was observed when a morpholin-
yl (4o) or a thiomorpholinyl (4p)
substituent was attached. Over-
all, 4h was identified as the
most potent cidal agent against
trypomastigote, while six other
compounds (4a, 4g, 4l, and
4n–p) also showed considerable
37Æ2
50.4Æ0.3
40Æ13
25Æ2
>231.4
237
50.8
ND
4j
ND
>100
21.9
13Æ3
4k
4l
74.9
10.8
>100
>244
>233
0
11.1
18Æ2
antiparasitic
activity
(CC50 <
10 mm), but they were less
potent than benznidazole.
4m
68.5
>100
>244
31Æ2
Next, we analyzed the antipar-
asitic activity against epimasti-
gotes. Benznidazole, which was
used as the reference drug, ex-
hibited an IC50 value of 4.8 mm.
In comparison, the nonsubstitut-
ed phenyl derivative 4a was sev-
eral times less potent (IC50 =
25 mm). The 4-bromo derivative
4h was very active in inhibiting
epimastigote and displayed an
IC50 value of 3.9 mm. In fact, this
compound was as potent as
benznidazole. In contrast, the bi-
phenyl derivative 4l was active
78Æ2
4n
4o
4p
12.7
15.9
10.8
1.8
>100
31.9
57.3
22.8
22.0
(0.7Æ0.4)
15Æ2
94.3Æ0.9
(4.0Æ0.5)
Bdz
Sap
6.0
–
4.8
–
96.1
1.0 mgmLÀ1
–
–
[a] Determined 24 h after incubation of Y strain trypomastigotes with the test compounds. [b] Determined
5 days after incubation of Dm28c-strain epimastigotes with the test compounds. [a,b] Only values with a stan-
dard deviation <10% were included. [c] Highest noncytotoxic (HNC) concentration for mouse splenocytes
after 24 h of incubation in the presence of the test compounds. [d] Compounds were tested at 100 mm and
the percent inhibition of catalytic activity was determined; values in parenthesis are IC50 values [nm] and repre-
sent the meanÆSD of three measurements. ND=not determined due to lack of activity; NT=not tested;
Bdz=benznidazole; Sap=saponin.
in
inhibiting
epimastigotes
the phenyl ring (4a) by 2- and 4-pyridinyl (4b and 4c, respec-
tively) produced inactive thiazolidinones. The attachment of
a methyl (4d), methoxy (4e) and tert-butyl (4k) substituent at
the 4-position of the phenyl ring decreased the antiparasitic
activity. The attachment of 4-fluoro (4 f) also decreased the ac-
tivity; however, attaching a 4-chloro (4g) produced a com-
pound as active as the nonsubstituted phenyl derivative (4a).
A further increase in potency was achieved by replacing the
nonsubstituted phenyl ring (4a) with a 4-bromo (4h), which
(IC50 =11.1 mm) but was only half as potent as benznidazole.
After determining the antiparasitic activity, we evaluated the
cytotoxicity in host cells. Saponin was used as the reference
drug in this assay, while benznidazole was evaluated to com-
pare the cytotoxicity. As shown in Table 1, compounds 4a–p
were less cytotoxic than saponin. In comparison to benznida-
zole, only five derivatives 4b, 4c, 4n, 4o, and 4p were equally
or more cytotoxic, whereas the rest of the derivatives were less
cytotoxic. Derivatives 4h and 4l, which are the most potent
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
4
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Figure 4. Thiazolidinone 4h reduces the parasite development and invasion process in host cells. A,B) T. cruzi-infected macrophages were treated and incu-
bated for 4 days. A) Infected cells in percent. B) Mean number of intracellular amastigotes per 100 infected macrophages. C) Macrophages were simultaneous-
ly exposed to trypomastigotes and treatment. The cell culture was incubated for 2 h, washed and the percent of infected cells was determined after 2 h. Bdz
= benznidazole, AmpB = amphotericin B. Data are the mean ÆS.E.M. (error bars) of two independent experiments performed. Panels A and B: ***, p <0.001;
panel C: ***, p <0.001; **, p <0.01.
anti-T. cruzi compounds among the series, did not exhibit cyto-
Table 2. In vitro infection, cytotoxicity in macrophages and selectivity
index (SI).
toxicity for splenocytes at concentrations up to 200 mm.
Compd
IC50 [mm][a]
CC50 [mm][b]
SI[c]
In vitro infection and selectivity index
4h
4l
Bdz
5.2Æ0.54
10.1Æ0.09
13.9Æ0.39
230.5Æ5
233.3Æ9
250Æ9
44
23
18
After evaluating the antiparasitic activity against the extracellu-
lar parasite, we analyzed the activity against the intracellular
parasite. In this assay, T. cruzi-infected macrophages were treat-
ed and incubated for 96 h. Benznidazole was used as a refer-
ence drug, and the results are summarized in Figure 4A,B. In
comparison to untreated cells, treatment with 4h reduced the
percentage of infected cells (p<0.001) as well as the mean
number of intracellular amastigotes per 100 macrophages.
Next, we investigated whether 4h inhibits the T. cruzi invasion
process in macrophages. To study this, macrophages were in-
fected with trypomastigotes and simultaneously treated. After
2 h, unbound parasites were removed, and the cells were incu-
bated for 2 h. Benznidazole and amphotericin B were used as
reference drugs (Figure 4C). In comparison to untreated cells,
treatment with 4h or 4l reduced the percentage of infected
macrophages (p<0.01); however, this reduction was smaller
than observed under amphotericin B treatment (p<0.001). In
this parasite invasion assay, benznidazole did not show signifi-
cant activity.
[a] IC50 determined in T. cruzi-infected macrophages after incubation for
4 days. [b] CC50 determined in mouse splenocytes after incubation for
24 h. Data represent the meanÆSD of two independent experiments per-
formed in duplicate. [c] Determined as CC50/IC50. Bdz=benznidazole.
marin (Z-FR-AMC).[33] Initially, compounds were screened at
100 mm, and only compounds with an inhibition value >70%
were chosen for IC50 value determination (see Table 1). We ob-
served that cruzain activity was not substantially inhibited by
the presence of most thiazolidinones, except for derivatives 4n
and 4p, which showed IC50 values of 0.7Æ0.4 nm and 4.0Æ
0.5 nm, respectively.
To define the structural determinants for cruzain inhibition
observed for compounds 4n and 4p, molecular docking was
performed. The binding mode for these ligands was deter-
mined by the highest (most positive) score among the possible
solutions for each ligand. These calculations were generated
according to the Goldscore fitness function. In order to identify
the molecular reasons for the two extreme affinities towards
the cruzain target, we selected the two most potent com-
pound, (R)-4n and (R)-4p, in addition to the weak cruzain in-
hibitor (S)-4b. We performed a detailed analysis of the inter-
molecular interactions observed in the docking solutions, for
these molecules. Although the two most potent molecules are
very similar to each other, they differ by the replacement of
a carbon atom on (R)-4n for a sulfur atom on (R)-4p. The dif-
ference between the active (R)-4n and the inactive cruzain in-
hibitor (S)-4b is the presence of a 2-pyridinyl instead of a 4-(pi-
peridinyl)phenyl ring, in addition to the inversion of the chiral
center. The comparative results can be found in Figure 5.
The difference between the binding modes of (R)-4n and
(S)-4b is show in Figure 6 and Table 3. The residues of cruzain,
Because the activity of 4h was concentration-dependent, we
determined its IC50 value and selectivity index (SI) in the in
vitro infection assay (see Table 2). Compound 4h was twice as
potent as benznidazole in reducing the percentage of infected
cells. In contrast, 4l was less potent, exhibiting the same po-
tency as benznidazole. Compound 4h has an SI value of 44,
which is two times higher than the value for 4l as well as
benznidazole.
Cruzain inhibition and docking analysis
To investigate the mechanism of action, compounds were
tested against cruzain. The inhibition of cruzain enzymatic ac-
tivity by all compounds was measured using a competition-
based assay with the substrate Z-Phe-Arg-aminomethylcou-
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
5
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
docking scores, demonstrating
that the molecules with more
stable or positive docking scores
(in silico affinity) are the most
potent cruzain inhibitors.
Electron microscopy of parasite
morphology
We analyzed the T. cruzi cellular
membranes and organelles fol-
lowing thiazolidinone treatment.
In this assay, trypomastigotes
were treated with 4h (4.0 mm;
CC50 value), incubated for 24 h
and then cells were analyzed by
transmission electron microsco-
py (TEM). In comparison to un-
treated parasites, the treatment
induced the formation of an
atypical dilatation of the Golgi
complex and endoplasmatic re-
ticulum (ER), as well as some dis-
tentions of the ER perinuclear
membrane (Figure 7). In addi-
tion, treatment produced the
formation of numerous and
atypical vacuoles within the cy-
toplasm and in close proximity
to the Golgi complex, which is
commonly observed within para-
site autophagy. In comparison to
untreated parasites, no altera-
tions were observed in the kinet-
oplast and cell nucleus in the
treated parasites.
Figure 5. Superimposition of the docking solutions on cruzain for compounds (R)-4n (orange stick), (R)-4p (red
stick), (S)-4b (blue stick) and the crystal structure of “KB2” co-crystallized ligand (gray line). A) Full and B) active
site view. Inset: chemical structure of KB2.[13]
Figure 6. Detailed view of the docking solutions on cruzain, for A) (R)-4n and B) (S)-4b. Residues involved in hy-
drophobic interactions (green), hydrogen bonds (cyan), and p–p T-shaped interactions (orange) are highlighted. A
detailed view of the p–p T-shaped interaction (orange) of the HIS162 residue with (R)-4n is shown in the box.
which participate in hydrophobic interactions, are highlighted
Table 3. Molecular interaction of cruzain with (R)-4n, (R)-4p and (S)-4b.
in green, whereas those that participate in hydrogen bonds
are highlighted in cyan. Finally, the residues involved in p–p T-
shaped interactions are highlighted in orange. As shown in
Figure 5, the cruzain binding mode for (R)-4n and (R)-4p is
very similar. Table 3 provides a list of the molecular interactions
and shows that the two compounds have many interactions in
common. This similarity is also revealed when we analyze the
docking score values for (R)-4n and (R)-4p, which are 55.39
and 56.48, respectively. In contrast, through comparison with
the values for (R)-4n and (R)-4p, the lower affinity of (S)-4b for
cruzain can also be identified. The data presented in Table 3
demonstrates that the p–p T-shaped interaction seems to be
mainly responsible for the greater stability and the positioning
of the complex formed between (R)-4n or (R)-4p and cruzain.
The docking score calculated for (S)-4b is lower (49.59) than
those calculated for the potent compounds (R)-4n or (R)-4p.
Therefore, this result indicates that the most potent cruzain in-
hibitors (R)-4n and (R)-4p are also those with the highest
Residues
Compounds[a]
(R)-4n
(R)-4p
(S)-4b
Gly23
Cys25
Trp 26
Ser61
HC
–
HC
–
HC
2.4[b]
HC
–
–
HC
–
HC
HC
HC
HC
HC
2.6[b]
HC
49.59
HC
HC
3.1[b]
HC
3.1[b]
HC
HC
HC
–
HC
PIT
HC
55.39
HC
HC
2.9[b]
HC
3.0[b]
HC
HC
HC
–
HC
PIT
HC
56.48
Ser64
Gly65
Gly66
Leu67
Met68
Ala138
Leu160
Asp161
His162
Gly163
Scores
[a] HC=hydrophobic contacts, PIT=p–p T-shaped interaction. [b] Hydro-
gen bond distances [ꢃ] between donor and acceptor.
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
6
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
with 12 mm of 4h for 48 h resulted in 9.56 and 16.2% of para-
sites positively stained for PI and PI+annexin V, respectively;
whereas 1.94% parasite cells were stained only for annexin V.
These results indicate that thiazolidinone treatment increases
the number of PI staining, which is characteristic of a parasite
cell death caused by a necrotic process.
Toxicology in mice
A single-dose toxicological study was carried out for thiazolidi-
none 4h. The compound was given orally by gavage in unin-
fected female BALB/c mice (n=3/group) at the doses of 150,
300 or 600 mgkgÀ1. Mice were monitored during 14 days. To
the tested doses, neither mortality nor gross signs associated
to toxicity were observed (data not shown). Doses higher
600 mgkgÀ1 could not be tested due to the limited solubility
of 4h in 20% DMSO/saline, so the maximum tolerated dose in
BALB/c mice was not calculated.
In another experiment, 600 mgkgÀ1 of 4h was administrated
to a group of uninfected mice (n=6) and blood samples were
collected 24 h after treatment. Fourteen biochemical compo-
nents of the sera were measured and the levels were com-
pared to the negative control group (receiving vehicle; data
are summarized in table S1 of the Supporting Information). In
comparison to the negative control, treatment with 4h altered
three biochemical components of the sera (p<0.05): alanine
aminotransferase, amylase and creatine. Alanine aminotransfer-
ase and amylase are biochemical components involved in liver
function, suggesting that in this dose, compound 4h is poten-
tially hepatotoxic. However, between untreated and treated
groups, no statistical differences were observed for other ana-
lyzed components.
Figure 7. Electron microscopy analysis of parasite morphology. A) Untreated
trypomastigotes analyzed by TEM. B–D) Trypomastigotes incubated for 24 h
with 4h (4.0 mm). The formation of atypical vacuoles in the cytoplasm is
shown(B). Arrows highlight alterations in the ER membrane (C). Alterations
in the Golgi complex is indicated by stars (D). Untreated infected macro-
phages are observed after 6 h of infection (E). F) Infected macrophages incu-
bated for 6 h with 4h (5.0 mm). Alterations in the Golgi apparatus and the
formation of atypical vacuoles are visible (indicated by arrows). N=nucleus;
K=kinetoplast; GA=Golgi apparatus; ER=endoplasmatic reticulum. Scale
bars: A) 1 mm; B,C) 0.5 mm; D) 0.2 mm; E) 1 mm; F) 0.5 mm.
Infection in mice
Given the potent in vitro antiparasitic activity as well as low
toxicity in mice, we tested thiazolidinone 4h in T. cruzi-infected
mice (acute model). In this assay, Y strain trypomastigotes were
inoculated in female BALB/c mice (n=6/group). Five days after
infection, mice were treated orally by gavage once a day for
five consecutive days. Compound 4h was administrated at
Next, infected macrophages were analyzed by TEM. In com-
parison to untreated infected macrophages, thiazolidinone
treatment led to the formation of atypical cytosolic vacuoles as
well as alterations in the Golgi complex morphology. These re-
sults indicate that 4h is a parasiticidal agent.[34]
a
dose of 125 mmolkgÀ1 (55 mgkgÀ1) and 250 mmolkgÀ1
(110 mgkgÀ1). The positive and negative control groups re-
ceived benznidazole (250 mmolkgÀ1, 65 mgkgÀ1) and 20%
DMSO/saline, respectively. Blood parasitemia was analyzed,
and the results are shown in Figure 9.
Parasite death process
After ascertaining that 4h is a parasiticidal agent, the process
of parasite death induction was studied more closely. Trypo-
mastigotes were treated with different concentrations of 4h
during 24, 48 and 72 h incubation and then labeled with an-
nexin V and propidium iodide (PI). Experiments were analyzed
by flow cytometry (Figure 8). In untreated parasites, most cells
were negative for annexin V and PI staining, demonstrating cell
viability. In comparison to untreated parasites, a significant and
concentration-dependent increase in the number of PI-positive
parasites was observed under treatment with 4h. Treatment
In the negative control group, blood parasitemia was ob-
served after day 6 of parasite inoculation and peaked on
day 10. In the positive control group, which received benznida-
zole, very low blood parasitemia was observed during the ex-
periment, indicating that infection erradication was achieved.
In comparison to the negative control group, treatment with
4h significantly reduced blood parasitemia in all tested doses
(p<0.001). All infected and treated mice survived and did not
show any behavioral alterations until the end of the experi-
ment (data not shown).
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
7
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
a dose of 250 mmolkgÀ1, 97% of
blood parasitemia was reduced
(p<0.001).
Drug combination
In another set of experiments,
the antiparasitic activity of thia-
zolidinone 4h alone and in com-
bination with benznidazole was
investigated in the in vitro infec-
tion assay. For this assay, com-
pound concentrations were se-
lected based on the IC50 values.
As shown in Table 5, neither
thiazolidinone 4h nor benznida-
zole alone were unable to cure
the infection in macrophage cul-
ture. In contrast, the combina-
tion of 4h plus benznidazole re-
duced the number of infected
macrophages to greater extent
than each compound used
alone. The combination of the
two compounds each at a con-
centration of 40 mm cured the in-
fection in macrophages. More
importantly, this was achieved
without affecting host cell viabil-
ity (data not shown).
Figure 8. Thiazolidinones cause T. cruzi death by a necrotic process. Flow cytometry examination of trypomasti-
gotes treated with 4h within 48 h incubation. A) Untreated trypomastigotes; B) 4.0 mm; C) 8.0 mm; D) 12 mm. Two
independent experiments were performed. Data are representative of one experiment.
Table 4. Summary of the in vivo antiparasitic activity.
Compd
Dose
[mmolkgÀ1
Blood parasitemia reduction in mice [%][a]
]
8 dpi
10 dpi
4h
4h
Bdz
125
250
250
79.4
97.7
98.6
75.4
97
>99
[a] Data taken from Figure 9, and values were calculated using the equa-
tion (vehicle groupÀtreated group)/vehicle group]ꢄ100%. Dpi=days
post-infection; Bdz=benznidazole.
Table 5. Summary of the in vitro antiparasitic activity of the drug combi-
nation.
Figure 9. Course of acute infection and response to treatment in T. cruzi-
infected mice. Female BALB/c mice (n=6/group) were infected with trypo-
mastigotes and treated for five consecutive days with thiazolidinone 4h by
oral gavage once a day. Blood parasitemia was monitored by counting the
number of trypomastigotes. One single experiment. Error bars for S.E.M.; sig-
nificance: ***, p<0.001 compared to untreated (vehicle) group.
4h [mm]
Bdz [mm]
Cell infection inhibition [%][a]
none
5.5
5.5
11
14
none
14
28
40
49.7
45.2
81.5
94.9
100
40
The peak parasitemia was employed to calculate the per-
centage of parasitemia reduction (Table 4). In comparison to
the negative control, treatment with 4h at a dose of 125
mmol/kg reduced blood parasitemia by 75.4% (p<0.001). At
[a] Determined 4 days after macrophage infection with Y strain trypomas-
tigotes. Inhibition [%] was determined in comparison to untreated infect-
ed cells. Data are from one single experiment. Bdz=benznidazole.
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
8
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
rest of the chemical series. Therefore, the observed inhibitory
properties for cruzain are likely due to the substituents present
in 4n and 4p. For compounds 4n and 4p, the attachment of
a piperidinyl or thiomorpholinyl produced compounds with in-
hibitory property against cruzain; however, this was abolished
when a morpholinyl substituent was attached. In fact, cruzain
docking of 4n showed that the piperidinyl group is oriented in
a hydrophobic pocket with the participation of p–p T-shaped
interactions. The same is not observed for compounds 4b and
4o, which explains their lack of inhibitory activity against cru-
zain.
Discussion
There is a need for novel anti-T. cruzi drugs. Thiazolidinones
are heterocyclic compounds well known for their antiprotozoal
activities.[35–38] After screening the activity of 60 thiazolidinic
derivatives, we previously identified a potent anti-T. cruzi thia-
zolidinone (18).[24] This compound selectively inhibited the try-
panosomal protease cruzain but not its mammalian homolo-
gous cathepsin L. Of note, compound 18 achieved inhibitory
property against cruzain without exhibiting nonspecific and
promiscuous binding properties, a characteristic observed for
thiazolidinones of very low molecular weight.[39] Moreover, this
compound presented low cytotoxicity towards host cells, no
apparent toxicity in mice and reduced blood parasitemia in
mice when administrated orally. As a limitation, thiazolidinone
18 was less efficient in reducing acute infection than benznida-
zole. Therefore, compound 18 was used here as a structural
prototype for the synthesis of a series of novel thiazolidinones.
The employed structural planning for the design of com-
pounds 4a–p aimed at causing a disruption in the molecular
planarity between the phenoxy group and the thiazolidinic
ring. This was achieved by the attachment of an aryl group in
the iminic carbon of this class of compounds. As highlighted in
Figure 2, this produced derivatives with higher conformational
restriction compared with the prototype (18). The pharmaco-
logical evaluation of compounds 4a–p confirmed that the dis-
ruption of the planarity changed the bioactive profile of this
compound class. Importantly, this led to the identification of
derivatives with an enhanced antiparasitic activity compared
with known thiazolidinones, such as compound 18.
Regarding the antiparasitic activity for trypomastigotes, we
found that the attachment of a phenyl ring produced an
active thiazolidinone, which was, however, less potent than
benznidazole. Replacing the phenyl with a pyridinyl ring was
deleterious for antiparasitic activity, suggesting certain struc-
tural requirements for placing an aryl ring at the iminic carbon.
In fact, the investigation of substituents attached to the
phenyl ring revealed interesting SARs. We found substituents
that retained (4-Cl, 4-morpholinyl), enhanced (4-Br, 4-phenyl, 4-
thiomorpholinyl) or removed (4-CH3, 4-CH3O, 4-F) activity
against trypomastigotes in comparison to the nonsubstituted
thiazolidinone. Among the substituents that led to an en-
hanced anti-T. cruzi activity, 4-bromo (4h) and biphenyl (4l)
were the most promising in terms of potency and selectivity.
These compounds were able to inhibit the proliferation of epi-
mastigotes and were toxic for T. cruzi but displayed low cyto-
toxicity towards host cells. Though compounds 4h and 4l
both have hydrophobic and bulky substituents, other substitu-
ents containing similar properties, such as tert-butyl or phen-
oxy, did not produce active antiparasitic agents. This implies
that there are unknown structural requirements involved in 4h
and 4l that provide the observed antiparasitic activity.
Compound 4h was selected as an anti-T. cruzi lead com-
pound because it exhibited the highest selectivity among the
thiazolidinones studied here. Under 4h treatment, in vitro par-
asite development and invasion in host cells was substantially
reduced. This activity was more pronounced even than that
observed with benznidazole-treated parasites. Moreover, the
treatment with this compound caused alterations in the Golgi
apparatus and the ER morphology of T. cruzi, whereas little or
no effects were observed in the kinetoplast and cell nucleus
morphology. Therefore, this compound exerts its antiparasitic
activity by altering organelle morphology, which ultimately de-
stroys parasite cells, similar to the mode of action of a parasiti-
cidal agent. In agreement to this, we observed that 4h caused
parasite cell death through a necrotic process.
With regard to the activity in infected mice during the acute
phase, 4h reduced the blood parasitemia in a dose-dependent
manner and exhibited a potency similarly to that observed in
benznidazole-receiving mice. When compared to untreated in-
fected mice, thiazolidinone 4h reduced 97% of blood parasite-
mia, while at the same dose, the reduction observed for the
previously reported thiazolidinone 18 was 89%.[24] Therefore,
a potency enhancement of the in vivo antiparasitic activity was
achieved from the first-generation thiazolidinone 18 to the
new generation described here (4h). Regarding toxicity in
mice, 4h was not lethal in doses up to 600 mgkgÀ1, which is
several times higher than the dose used to reduce blood para-
sitemia. Altogether, these results reinforced the notion that
thiazolidinone 4h is a selective anti-T. cruzi agent. A pharmaco-
kinetic analysis as well as identification of its mechanism of
action should be determined in the course of further investiga-
tion. However, our preliminary results already show that the
combination of thiazolidinone 4h plus benznidazole is additive
in reducing in vitro T. cruzi infection. This finding indicates that
thiazolidinone 4h could be a suitable partner for anti-Chagas
drug combinations. This is an important parameter for new
anti-Chagas drug candidates, since an effective treatment will
likely contain a drug combination to improve the efficacy of
the treatment and to reduce the chance of developing parasite
resistance.[41]
Conclusions
Moreover, we observed that only 4n (piperidinyl) and 4p
(thiomorpholinyl) inhibited cruzain activity whereas other de-
rivatives did not. The literature describes some diastereoiso-
mers with different cruzain inhibitory properties.[40] However,
the isomeric ratios for 4n and 4p are not different from the
Thiazolidinones are a family of well-known antiparasitic com-
pounds. Here, we prepared a series of new thiazolidinones 4a–
p, which were designed to be potential anti-Trypanosoma cruzi
compounds targeting the trypanosomal protease cruzain. In
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
9
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Cell culture: Epimastigotes of a Dm28c strain (discrete typing unit I)
were maintained at 268C in liver infusion tryptose (LIT) medium
(Life Technologies, Carlsbad, CA, USA) supplemented with 10%
fetal bovine serum (FBS, Life Technologies), 1% hemin (Sigma–Al-
drich, St. Louis, MO, USA), 1% R9 medium (Sigma–Aldrich), and
50 mgmLÀ1 gentamycin (Novafarma, Anꢀpolis, GO, Brazil). Metacy-
clic trypomastigotes of Y strain (discrete typing unit II) were ob-
tained from the supernatant of infected LLC-MK2 cells and main-
tained in RPMI-1640 medium (Sigma–Aldrich) supplemented with
10% FBS and 50 mgmLÀ1 gentamycin at 378C and 5% CO2. Spleno-
cytes were collected from BALB/c mice and cultivated in RPMI-
1640 medium supplemented with 10% FBS and 50 mgmLÀ1 genta-
mycin. Peritoneal exudate macrophages were elicited by intraperi-
toneal injection of sodium thioglycollate in BALB/c mouse.
fact, we found these compounds are antiparasitic agents and
have a high degree of selectivity. Our structure–activity rela-
tionship (SAR) studies revealed structural determinants for anti-
parasitic activity and led to the identification of thiazolidinone
derivatives, which displayed similar potencies to benznidazole.
Specifically, we demonstrated that thiazolidinone 4h has
strong antiparasitic activity, low cytotoxicity toward host cells
and achieves its anti-T. cruzi activity as a parasiticidal agent.
Most thiazolidinones, including active compound 4h, did not
inhibit cruzain activity, but these compounds affected the
Golgi apparatus as well as ER morphology and produced atypi-
cal cytosolic vacuoles, which ultimately is followed by a necrot-
ic parasite cell death. Consistent with in vitro antiparasitic ac-
tivity, thiazolidinone 4h reduced parasitemia in a mice model
of acute infection. Importantly, this compound displayed low
toxicity in mice and exhibited additive antiparasitic activity
when combined with benznidazole.
Cytotoxicity in splenocytes: Splenocytes of BALB/c mice (200 mL)
were placed into 96-well plates at 5ꢄ106 cells/well. Compounds
were added in a serial dilution (1.1, 3.3, 11, 33 and 100 mgmLÀ1) in
triplicate. To each well, an aliquot of compound suspended in di-
methyl sulfoxide (DMSO) was added. Negative (untreated) and pos-
itive (saponin, Sigma–Aldrich) controls were measured in each
plate, which was incubated for 24 h at 378C and 5% CO2. After in-
cubation, [3H]-thymidine (1.0 mCimLÀ1, PerkinElmer, Waltham, MA,
USA) was added to each well, and the plate was returned to the in-
cubator. Cells were then transferred to a filter paper using a cell
Experimental Section
Chemistry
harvester and measured using
a liquid scintillation counter
Synthetic protocols and spectral data for compounds are described
in the Supporting Information.
(WALLAC 1209, Rackbeta Pharmacia, Stockholm, Sweden). [3H]-Thy-
midine incorporation [%] was measured, and the highest noncyto-
toxic concentration was determined using triplicates.
Antiproliferative activity for epimastigotes: Epimastigotes were
counted in a hemocytometer, and 200 mL were dispensed into 96-
well plates at 106 cells/well. Compounds were added in a serial di-
lution (1.1, 3.3, 11, 33 and 100 mgmLÀ1) in triplicate. The plate was
incubated for 5 days at 268C, and aliquots of each well were col-
lected for counting the number of viable parasites using a Neuba-
uer chamber. The percentage of inhibition was calculated in rela-
tion to untreated cultures. IC50 values were calculated using nonlin-
ear regression on Prism 4.0 (GraphPad). Results are from one single
experiment.
Docking
The structures of all compounds were obtained by application of
the RM1 method,[42] available as part of the SPARTAN 08’ pro-
gram,[43] using internal default settings for convergence criteria.
Some of these new molecules were synthesized as racemic mix-
tures; therefore, the molecular modeling treated the two isomers
(R and S) independently, when appropriate, and the docking proce-
dure used both isomers for each compound. Docking calculations
and analysis were carried using the structure of T. cruzi cruzain
(PDBID: 3IUT) as the target, which is composed of a co-crystallized
complex with inhibitor (referred as “KB2”).[13] The active site was
defined as all atoms within a radius of 6.0 ꢃ from the co-crystal-
lized ligand. Residues Gln19, Cys25, Ser61, Leu67, Met68, Asn70,
Asp161, His162, Trp184 and Glu208 were treated as flexible during
the calculations, using a conformation library for each one. The
GOLD 5.1 program[44] was used for docking calculations, followed
by Binana program,[45] which was used to analyze the molecular in-
teractions present in the best docking solutions, using default set-
ting, except for hydrogen bond distance, which was changed to
a maximum of 3.5 ꢃ. Figures were generated with Pymol (version
1.3r1 edu).[46]
Toxicity for trypomastigotes: Trypomastigotes were collected from
the supernatant of LLC-MK2 cells, and 200 mL were aliquoted into
96-well plate in an axenic media at 4ꢄ105 cells/well. Compounds
were added in a serial dilution (1.1, 3.3, 11, 33 and 100 mgmLÀ1) in
triplicate. The plate was incubated for 24 h at 378C and 5% of CO2.
Aliquots of each well were collected, and the number of viable par-
asites, based on parasite motility, was counted in a Neubauer
chamber. The percentage of inhibition was calculated in relation to
untreated cultures. CC50 calculation was also carried out using non-
linear regression with Prism 4.0 (GraphPad). Two independent ex-
periments were performed.
Intracellular parasite development: Macrophages were seeded at 2ꢄ
105 cells/well for 24 h in a 24-well plate with rounded coverslips on
the bottom in RPMI-1640 medium supplemented with 10% FBS.
Macrophages were infected with trypomastigotes at a ratio of 10
parasites per macrophage for 2 h. Unbound trypomastigotes were
removed by successive washes with saline. Each compound was
dissolved in 5% DMSO and saline in a serial dilution, in triplicate.
Compounds remained on the cell culture for 6 h prior to removal
and addition of fresh media. The plate was incubated for 4 days at
378C and 5% CO2. Cells were fixed in MeOH, stained with Giemsa
and manual counting of at least 100 cells per slide was done using
an optical microscope (Model CX41, Olympus, Tokyo, Japan). The
Biology
Animals: Female BALB/c mice, aged 6–8 weeks, were supplied by
the animal house of Centro de Pesquisas GonÅalo Moniz (Fundażo
Oswaldo Cruz, Bahia, Brazil) and Centro de Pesquisas Aggeu Magal-
haes (Fundażo Oswaldo Cruz, Pernambuco, Brazil). Mice were
maintained in sterilized cages under a controlled environment, re-
ceiving a balanced diet for rodents and water ad libitum. All ex-
periments were carried out in accordance with the recommenda-
tions of ethical guidelines and were approved by the local Animal
Ethics Committee.
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&10
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
percentages of infected macrophages and mean numbers of amas-
tigotes per 100 infected macrophages were determined. To calcu-
late IC50 values, the percentage of infected macrophages in com-
parison to untreated infected macrophages was used.
tion and analyses were performed using FACS Calibur flow cytome-
ter (Becton Dickinson, San Jose, CA, USA) and FlowJo software
(Tree Star, Ashland, OR, USA), respectively. A total of 20000 events
were acquired. Two independent experiments were performed.
Drug combination evaluation: The combination of thiazolidinone
4h plus benznidazole was evaluated in the in vitro infection assay
(described above). To this end, compounds were tested in the
same concentration of their IC50 values as well as in higher concen-
trations.
Acute toxicity in mice: Uninfected BALB/c mice (female, 7–9 weeks
old) were randomly divided in groups (n=3). Oral administration
of 4h was given by gavage in one single dose of 150, 300 or
600 mgkgÀ1 in 20% DMSO/saline as vehicle. After treatment, sur-
vival was monitored for 14 days. The same experiment was repeat-
ed using n=6 per group. Heparinized blood samples were collect-
ed 24 h after treatment for biochemical analysis of serum compo-
nents. Readings were performed in an Analyst platform (Hemagen
Diagnostics, Colombia, MD, USA) for 16 biochemical components.
Biochemical readings were compared to negative control (received
vehicle only).
Parasite invasion: Macrophages (105 cellsmLÀ1) were plated onto
13 mm glass coverslips in a 24-well plate and maintained for 24 h.
Y strain trypomastigotes were added at 107 cellsmLÀ1, followed by
addition of compound 4h. Amphotericin B (Fungizone, Life Tech-
nologies) and benznidazole (Lafepe, Recife, PE, Brazil) were used as
reference compounds. The plate was incubated for 2 h at 378C
and 5% CO2. Unbound parasites were then removed by successive
washes with saline, and the RPMI medium was replaced. After 2 h
of incubation, the number of infected cells was counted by optical
microscopy using standard Giemsa staining.
Infection in mice: BALB/c mice (female, 6–8 weeks old) were infect-
ed with bloodstream trypomastigotes (Y strain) by intraperitoneal
injection of 104 parasites in 100 mL of saline solution. Mice were
randomly divided in groups (n=6 per group). After day 5 of post-
infection, treatment with 4h was given orally by gavage once
a day for five consecutive days. For the positive control group,
benznidazole was given orally. As recommended by standard pro-
tocols, infection was monitored daily by counting the number of
motile parasites in 5 mL of fresh blood sample drawn from the lat-
eral tail vein.[47] Mortality was recorded daily until day 30 after the
end of treatment.
Inhibition of cruzain activity: Recombinant cruzain was kindly pro-
vided by Dr. Anna Tochowicz and Dr. James H. McKerrow from the
University of California, San Francisco (CA, USA). Cruzain activity
was measured by monitoring the cleavage of the fluorogenic sub-
strate Z-Phe-Arg-aminomethylcoumarin (Z-FR-AMC, Sigma–Aldrich)
using a Synergy 2 microplate reader (Biotek) from the Center of
Flow Cytometry and Fluorimetry in the Biochemistry and Immunol-
ogy Department (UFMG, Brazil) with filters for l=340 nm (excita-
tion) and l=440 nm (emission). All assays were performed in 0.1m
NaOAc (pH 5.5) and in the presence of 5 mm dithiothreitol. The
final concentration of cruzain was 0.5 nm, and the substrate con-
centration was 2.5 mm (Km =1.0 mm). Assays were conducted in
presence of 0.01% Triton X-100. Compounds were screened at
100 mm, after pre-incubation for 10 min with the enzyme prior to
the addition of Z-FR-AMC. In all assays, fluorescence was moni-
tored for 5 min after addition of the substrate and activity was cal-
culated in relation to DMSO control. If the cruzain inhibition was
higher than 70% at 100 mm, IC50 values were determined by using
at least seven inhibitor concentrations. Each compound concentra-
tion was tested in triplicate and data were analyzed with Prism 5.0
(GraphPad).
Statistical analysis: To determine the statistical significance of each
group in the in vitro and in vivo experiments, the one-way ANOVA
test and the Bonferroni for multiple comparisons were used. A
p value <0.05 was considered significant.
Supporting Information
Please see the Supporting Information for synthetic protocols,
spectral data for compounds and a table with toxicological results.
CCDC 936127 contains the supplementary crystallographic data
(4q) of this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre (Cambridge, UK)
via www.ccdc.cam.ac.uk).
Electron microscopy analysis: Trypomastigotes (3ꢄ107 cellsmLÀ1
)
were treated with 4h (4.0 mm ) for 24 h. Parasites were fixed in 2%
formaldehyde and 2.5% glutaraldehyde (Electron Microscopy Sci-
ences, Philadelphia, PA, USA) in sodium cacodylate buffer (0.1m,
pH 7.2) for 1 h at RT, washed 3ꢄ with sodium cacodylate buffer
(0.1m, pH 7.2), and post-fixed with a 1% solution of osmium tetr-
oxide (Sigma–Aldrich) for 1 h. After dehydration with acetone, try-
pomastigotes were embedded in Poly/Bed (PolyScience, Dallas, TX,
USA), sectioned, stained with uranyl acetate and lead citrate and
analyzed using a JEOL TEM-1230 transmission electron microscope
(Acworth, GA, USA). Mouse macrophages were infected with
Y strain trypomastigotes. Cell cultures were washed with saline to
remove unbound parasites, followed by the addition of 4h
(5.0 mm) and incubation for 6 h at RT. Cell cultures were processed
as described above for collecting TEM images.
Acknowledgements
This study was supported by the Brazilian agencies Fundażo de
Amparo ꢂ Pesquisa do Estado da Bahia (FAPESB, Brazil), Consel-
ho Nacional de Desenvolvimento Cientꢀfico e Tecnolꢃgico (CNPq,
Brazil), Fundażo de Amparo ꢂ Pesquisa do estado de Minas
Gerais (FAPEMIG, Brazil) and the European Union consortium
ChemBioFight. A.C.L.L., M.B.P.S., C.A.S. and M.Z.H. are CNPq fel-
lows. D.R.M.M. and C.S.M. hold a FAPESB scholarship. We thank
the Physics Institute, University of Sao Paulo (Brazil) for allowing
the use of the diffractometer and the Mass Spectrometry Unit of
the Centro de Pesquisas GonÅalo Moniz (CPqGM). The authors ac-
knowledge Dr. Phillipe J. Eugster (University of Lausanne, Switzer-
land) for recording mass spectra and Dr. Kyan J. Allahdadi (Hos-
pital S¼o Rafael, Brazil) for proofreading the manuscript. R.S.F. is
thankful to Dr. Anna Tochowicz (University of California, San
Francisco (UCSF), USA) for providing recombinant cruzain.
Flow cytometry analysis: Trypomastigotes (107 cellsmLÀ1
) were
treated with 4h (4.0, 8.0 and 12 mm) and incubated for 72 h at
378C. Aliquots were collected in the intervals of 24–72 h and incu-
bated for 45 min with propidium iodide (PI) and annexin V using
the annexin V–FITC apoptosis detection kit (BioLegend, San Diego,
CA, USA) according to the manufacturer instructions. Data acquisi-
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
11
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
[23] M. Z. Hernandes, M. M. Rabello, A. C. L. Leite, M. V. O. Cardoso, D. R. M.
Moreira, D. J. Brondani, C. A. Simone, L. C. Reis, M. A. Souza, V. R. A. Per-
eira, R. S. Ferreira, J. H. McKerrow, Bioorg. Med. Chem. 2010, 18, 7826–
7835.
Keywords: antiprotozoal agents
·
biological activity
thiazolidinones
·
·
hydrazones
·
medicinal chemistry
·
Trypanosoma cruzi
[24] D. R. M. Moreira, S. P. Costa, M. Z. Hernandes, M. M. Rabello, G. B. O.
Filho, C. M. de Melo, L. F. da Rocha, C. A. de Simone, R. S. Ferreira, J. R.
Fradico, C. S. Meira, E. T. Guimar¼es, R. M. Srivastava, V. R. A. Pereira,
M. B. P. Soares, A. C. L. Leite, J. Med. Chem. 2012, 55, 10918–10936.
[25] C. L. Donnici, M. H. Araffljo, H. S. Oliveira, D. R. M. Moreira, V. R. A. Pereira,
M. A. Souza, M. C. Castro, A. C. L. Leite, Bioorg. Med. Chem. 2009, 17,
5038–5043.
[26] M. Ishikawa, Y. Hashimoto, J. Med. Chem. 2011, 54, 1539–1554.
[27] E. J. Barreiro, A. E. Kümmerle, C. A. M. Fraga, Chem. Rev. 2011, 111,
5215–5246.
[1] J. A. Urbina, Clin. Infect. Dis. 2009, 49, 1685–1687.
[2] M. P. Barrett, R. J. Burchmore, A. Stich, J. O. Lazzari, A. C. Frasch, J. J. Caz-
zulo, S. Krishna, Lancet 2003, 362, 1469–1480.
[3] A. Rassi, A. Rassi, Jr., J. A. Marin-Neto, Lancet 2010, 375, 1388–1402.
[4] F. S. Machado, L. A. Jelicks, L. V. Kirchhoff, J. Shirani, F. Nagajyothi, S. Mu-
kherjee, R. Nelson, C. M. Coyle, D. C. Spray, A. C. de Carvalho, F. Guan,
C. M. Prado, M. P. Lisanti, L. M. Weiss, S. P. Montgomery, H. B. Tanowitz,
Cardiol. Rev. 2012, 20, 53–65.
[5] S. Gupta, X. Wan, M. P. Zago, V. C. Sellers, T. S. Silva, D. Assiah, M.
Dhiman, S. NuÇez, J. R. Petersen, J. C. Vꢀzquez-Chagoyꢀn, J. G. Estrada-
Franco, N. J. Garg, PLoS Neglected Trop. Dis. 2013, 7, e2018.
[6] M. B. P. Soares, R. S. Lima, B. S. F. Souza, J. F. Vasconcelos, L. L. Rocha,
R. R. Dos Santos, S. Iacobas, R. C. Goldenberg, M. P. Lisanti, D. A. Iacobas,
H. B. Tanowitz, D. C. Spray, A. C. Campos de Carvalho, Cell Cycle 2011,
10, 1448–1455.
[28] Ö. Ö. Güven, ARKIVOC 2007, 142–147.
[29] G. M. Allan, N. Vicker, H. R. Lawrence, H. J. Tutill, J. M. Day, M. Huchet, E.
Ferrandis, M. J. Reed, A. Purohit, B. V. Potter, Bioorg. Med. Chem. 2008,
16, 4438–4456.
[30] P. Vicini, A. Geronikaki, K. Anastasia, M. Incerti, F. Zani, Bioorg. Med.
Chem. 2006, 14, 3859–3864.
[31] M. H. Bolli, S. Abele, C. Binkert, R. Bravo, S. Buchmann, D. Bur, J. Gatfield,
P. Hess, C. Kohl, C. Mangold, B. Mathys, K. Menyhart, C. Müller, O.
Nayler, M. Scherz, G. Schmidt, V. Sippel, B. Steiner, D. Strasser, A. Treiber,
T. Weller, J. Med. Chem. 2010, 53, 4198–4211.
[7] S. G. Macambira, J. F. Vasconcelos, C. R. Costa, W. Klein, R. S. Lima, P. Gui-
mar¼es, D. T. Vidal, L. C. Mendez, R. Ribeiro-Dos-Santos, M. B. P. Soares,
FASEB J. 2009, 23, 3843–3850.
[32] A. P. Liesen, T. M. de Aquino, C. S. Carvalho, V. T. Lima, J. M. de Araffljo,
J. G. de Lima, A. R. de Faria, E. J. de Melo, A. J. Alves, E. W. Alves, A. Q.
Alves, A. J. S. Góes, Eur. J. Med. Chem. 2010, 45, 3685–3691.
[33] R. S. Ferreira, C. Bryant, K. K. Ang, J. H. McKerrow, B. K. Shoichet, A. R.
Renslo, J. Med. Chem. 2009, 52, 5005–5058.
[34] M. A. Vannier-Santos, S. L. De Castro, Curr. Drug Targets 2009, 10, 246–
260.
[35] C. Pizzo, C. Saiz, A. Talevi, L. Gavernet, P. Palestro, C. Bellera, L. B. Blanch,
D. Benꢅtez, J. J. Cazzulo, A. Chidichimo, P. Wipf, S. G. Mahler, Chem. Biol.
Drug Des. 2011, 77, 166–177.
[36] H. E. Cummings, J. Barbi, P. Reville, S. Oghumu, N. Zorko, A. Sarkar, T. L.
Keiser, B. Lu, T. Rückle, S. Varikuti, C. Lezama-Davila, M. D. Wewers, C.
Whitacre, D. Radzioch, C. Rommel, S. Seveau, A. R. Satoskar, Proc. Natl.
Acad. Sci. USA 2012, 109, 1251–1256.
[37] N. Zelisko, D. Atamanyuk, O. Vasylenko, P. Grellier, R. Lesyk, Bioorg. Med.
Chem. Lett. 2012, 22, 7071–7074.
[38] T. L. de B. Moreira, A. F. S. Barbosa, P. Veiga-Santos, C. Henriques, A.
Henriques-Pons, S. L. Galdino, M. C. de Lima, I. R. Pitta, W. de Souza,
T. M. de Carvalho, Int. J. Antimicrob. Agents 2013, 41, 183–187.
[39] T. Mendgen, C. Steuer, C. D. Klein, J. Med. Chem. 2012, 55, 743–753.
[40] N. C. Romeiro, G. Aguirre, P. Hernꢀndez, M. Gonzꢀlez, H. Cerecetto, I.
Aldana, S. Pꢁrez-Silanes, A. Monge, E. J. Barreiro, L. M. Lima, Bioorg. Med.
Chem. 2009, 17, 641–652.
[8] R. A. Cutrullis, T. J. Poklꢁpovich, M. Postan, H. L. Freilij, P. B. Petray, Int.
Immunopharmacol. 2011, 11, 1024–1031.
[9] J. M. Kraus, H. B. Tatipaka, A. S. McGuffin, N. K. Chennamaneni, M.
Karimi, J. Arif, C. L. Verlinde, F. S. Buckner, M. H. Gelb, J. Med. Chem.
2010, 53, 3887–3898.
[10] A. Gerpe, G. Alvarez, D. Benꢅtez, L. Boiani, M. Quiroga, P. Hernꢀndez, M.
Sortino, S. Zacchino, M. Gonzꢀlez, H. Cerecetto, Bioorg. Med. Chem.
2009, 17, 7500–7509.
[11] S. P. Fricker, R. M. Mosi, B. R. Cameron, I. Baird, Y. Zhu, V. Anastassov, J.
Cox, P. S. Doyle, E. Hansell, G. Lau, J. Langille, M. Olsen, L. Qin, R. Skerlj,
R. S. Wong, Z. Santucci, J. H. McKerrow, J. Inorg. Biochem. 2008, 102,
1839–1845.
[12] M. E. Caputto, A. Ciccarelli, F. Frank, A. G. Moglioni, G. Y. Moltrasio, D.
Vega, E. Lombardo, L. M. Finkielsztein, Eur. J. Med. Chem. 2012, 55, 155–
163.
[13] K. Brak, I. D. Kerr, K. T. Barrett, N. Fuchi, M. Debnath, K. Ang, J. C. Engel,
J. H. McKerrow, P. S. Doyle, L. S. Brinen, J. A. Ellman, J. Med. Chem. 2010,
53, 1763–1773.
[14] P. S. Doyle, C. K. Chen, J. B. Johnston, S. D. Hopkins, S. S. Leung, M. P. Ja-
cobson, J. C. Engel, J. H. McKerrow, L. M. Podust, Antimicrob. Agents Che-
mother. 2010, 54, 2480–2488.
[15] F. S. Buckner, M. T. Bahia, P. K. Suryadevara, K. L. White, D. M. Shackle-
ford, N. K. Chennamaneni, M. A. Hulverson, J. U. Laydbak, E. Chatelain, I.
Scandale, C. L. Verlinde, S. A. Charman, G. I. Lepesheva, M. H. Gelb, Anti-
microb. Agents Chemother. 2012, 56, 4914–4921.
[16] F. Villalta, M. C. Dobish, P. N. Nde, Y. Y. Kleshchenko, T. Y. Hargrove, C. A.
Johnson, M. R. Waterman, J. N. Johnston, G. I. Lepesheva, J. Infect. Dis.
2013, 208, 504–511.
[17] X. Du, C. Guo, E. Hansell, P. S. Doyle, C. R. Caffrey, T. P. Holler, J. H. McKer-
row, F. E. Cohen, J. Med. Chem. 2002, 45, 2695–2707.
[18] D. C. Greenbaum, Z. Mackey, E. Hansell, P. Doyle, J. Gut, C. R. Caffrey, J.
Lehrman, P. J. Rosenthal, J. H. McKerrow, K. Chibale, J. Med. Chem. 2004,
47, 3212–3219.
[19] A. C. L. Leite, R. S. de Lima, D. R. M. Moreira, M. V. O. Cardoso, A. C. G.
Brito, L. M. F. Santos, M. Z. Hernandes, A. C. Kiperstok, R. S. de Lima,
M. B. P. Soares, Bioorg. Med. Chem. 2006, 14, 3749–3457.
[20] J. M. dos Santos Filho, A. C. L. Leite, B. G. de Oliveira, D. R. M. Moreira,
M. S. Lima, M. B. P. Soares, L. F. C. C. Leite, Bioorg. Med. Chem. 2009, 17,
6682–6891.
[41] S. Cencig, N. Coltel, C. Truyens, Y. Carlier, Int. J. Antimicrob. Agents 2012,
40, 527–532.
[42] G. B. Rocha, R. O. Freire, A. M. Simas, J. J. Stewart, J. Comput. Chem.
2006, 27, 1101 – 1111.
[43] Spartan ‘08 Tutorial and User’s Guide; Wavefunction: Irvine, CA, 2008:
http://www.wavefun.com/products/spartan.html (accessed: August
2013).
[44] Gold software, version 5.1, Cambridge Crystallographic Data Centre:
http://www.ccdc.cam.ac.uk (accessed: August 2013).
[45] J. D. Durrant, J. A. Mccammon, J. Mol. Graphics Modell. 2011, 29, 888–
893.
[46] W. L. DeLano. The PyMOL Molecular Graphics System, DeLano Scientific,
San Carlos, CA, 2002: http://www.pymol.org (accessed: August 2013).
[47] A. J. Romanha, S. L. Castro, M. N. Soeiro, J. Lannes-Vieira, I. Ribeiro, A.
Talvani, B. Bourdin, B. Blum, B. Olivieri, C. Zani, C. Spadafora, E. Chiari, E.
Chatelain, G. Chaves, J. E. Calzada, J. M. Bustamante, L. H. Freitas-Junior,
L. I. Romero, M. T. Bahia, M. Lotrowska, M. B. Soares, S. G. Andrade, T.
Armstrong, W. Degrave, Z. A. Andrade, Mem. Inst. Oswaldo Cruz 2010,
105, 233–238.
[21] J. M. dos Santos Filho, D. R. M. Moreira, C. A. de Simone, R. S. Ferreira,
J. H. McKerrow, C. S. Meira, E. T. Guimar¼es, M. B. P. Soares, Bioorg. Med.
Chem. 2012, 20, 6423–6433.
[22] A. C. L. Leite, D. R. M. Moreira, M. V. O. Cardoso, M. Z. Hernandes, V. R. A.
Pereira, R. O. Silva, A. C. Kiperstok, M. S. Lima, M. B. P. Soares, ChemMed-
Chem 2007, 2, 1339–1345.
Received: September 4, 2013
Published online on && &&, 0000
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&12
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
Stop the cycle! The attachment of an
aryl ring to the iminic carbon produced
thiazolidinones that are conformational-
ly more restricted than first-generation
thiazolidinones. This enhanced the po-
tency of antiparasitic thiazolidinones, as
observed under treatment with com-
pound 4h where parasite development
and invasion in host cells were substan-
tially reduced.
D. R. M. Moreira,* A. C. Lima Leite,
M. V. O. Cardoso, R. M. Srivastava,
M. Z. Hernandes, M. M. Rabello,
L. F. da Cruz, R. S. Ferreira,
C. A. de Simone, C. S. Meira,
E. T. Guimaraes, A. C. da Silva,
T. A. R. dos Santos, V. R. A. Pereira,
M. B. Pereira Soares
&& – &&
Structural Design, Synthesis and
Structure–Activity Relationships of
Thiazolidinones with Enhanced Anti-
Trypanosoma cruzi Activity
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&13
&
ÞÞ
These are not the final page numbers!