The effectiveness of natural diarylheptanoids against Trypanosoma cruzi: Cytotoxicity, ultrastructural alterations and molecular modeling studies

Article abstract of DOI:10.1371/journal.pone.0162926

Curcumin (CUR) is the major constituent of the rhizomes of Curcuma longa and has been widely investigated for its chemotherapeutic properties. The well-known activity of CUR against Leishmania sp., Trypanosoma brucei and Plasmodium falciparum led us to investigate its activity against Trypanosoma cruzi. In this work, we tested the cytotoxic effects of CUR and other natural curcuminoids on different forms of T. cruzi, as well as the ultrastructural changes induced in epimastigote form of the parasite. CUR was verified as the curcuminoid with more significant trypanocidal properties (IC₅₀ 10.13 μM on epimastigotes). Demethoxycurcumin (DMC) was equipotent to CUR (IC₅₀ 11.07 μM), but bisdemethoxycurcumin (BDMC) was less active (IC₅₀ 45.33 μM) and cyclocurcumin (CC) was inactive. In the experiment with infected murine peritoneal macrophages all diarylheptanoids were more active than the control in the inhibition of the trypomastigotes release. The electron microscopy images showed ultrastructural changes associated with the cytoskeleton of the parasite, indicating tubulin as possible target of CUR in T. cruzi. The results obtained by flow cytometry analysis of DNA content of the parasites treated with natural curcuminoids suggested a mechanism of action on microtubules related to the paclitaxel's mode of action. To better understand the mechanism of action highlighted by electron microscopy and flow cytometry experiments we performed the molecular docking of natural curcuminoids on tubulin of T. cruzi in a homology model and the results obtained showed that the observed interactions are in accordance with the IC₅₀ values found, since there CUR and DMC perform similar interactions at the binding site on tubulin while BDMC do not realize a hydrogen bond with Lys163 residue due to the absence of methoxyl groups. These results indicate that trypanocidal properties of CUR may be related to the cytoskeletal alterations.

Full text of DOI:10.1371/journal.pone.0162926

RESEARCH ARTICLE
The Effectiveness of Natural
Diarylheptanoids against Trypanosoma cruzi:
Cytotoxicity, Ultrastructural Alterations and
Molecular Modeling Studies
Vitor Sueth-Santiago1, Julliane de B. B. Moraes2, Eliomara Sousa Sobral Alves3, Marcos
3
Andre Vannier-Santos , Celio G. Freire-de-Lima , Rosane N. Castro , Gustavo
4
1
´
´
Peron Mendes-Silva1, Catarina de Nigris Del Cistia5, Luma Godoy Magalhães6, Adriano
Defini Andricopulo , Carlos Mauricio R. Sant´Anna , Debora Decote-Ricardo *, Marco
6
1
2
a11111
´
Edilson Freire de Lima1*
´
1 Universidade Federal Rural do Rio de Janeiro, Instituto de Ciências Exatas, Departamento de Quımica,
´
BR 465, Km 07, CEP: 23.890-000, Seropedica, RJ, Brazil, 2 Universidade Federal Rural do Rio de Janeiro,
Instituto de Veterinaria, Departamento de Microbiologia e Imunologia Veterinaria, BR 465, Km 07, CEP:
´
´
´
´
´
23.890-000, Seropedica, RJ, Brazil, 3 Laboratorio de Biologia Parasitaria, Centro de Pesquisas Gonc¸alo
Moniz (CPqGM-Fiocruz), Rua Waldemar Falcão, 121, Candeal, CEP: 40.296-710, Salvador, BA, Brazil,
´
4 Universidade Federal do Rio de Janeiro, Instituto de Biofısica Carlos Chagas Filho, Ilha do Fundão,
OPEN ACCESS
´
Cidade Universitaria, CEP: 21.941-902, Rio de Janeiro, RJ, Brazil, 5 Universidade Federal Rural do Rio de
´
Janeiro, Instituto de Ciências Exatas, Departamento de Matematica, BR 465, Km 07, CEP: 23.890-000,
Citation: Sueth-Santiago V, Moraes JdBB, Sobral
Alves ES, Vannier-Santos MA, Freire-de-Lima CG,
Castro RN, et al. (2016) The Effectiveness of
Natural Diarylheptanoids against Trypanosoma
cruzi: Cytotoxicity, Ultrastructural Alterations and
Molecular Modeling Studies. PLoS ONE 11(9):
e0162926. doi:10.1371/journal.pone.0162926
´
´
´
Seropedica, RJ, Brazil, 6 Laboratorio de Quımica Medicinal e Computacional, Centro de Pesquisa e
´
´
Inovac¸ão em Biodiversidade e Farmacos, Instituto de Fısica de São Carlos, Universidade de São Paulo, CP
396, CEP: 13.560-970, São Carlos, SP, Brazil
* marco@ufrrj.br (MEFL); decotericardo@ufrrj.br (DDR)
Editor: Ashraf B. Abdel-Naim, Ain Shams
Abstract
University, EGYPT
Received: June 6, 2016
Curcumin (CUR) is the major constituent of the rhizomes of Curcuma longa and has been
widely investigated for its chemotherapeutic properties. The well-known activity of CUR
against Leishmania sp., Trypanosoma brucei and Plasmodium falciparum led us to investi-
gate its activity against Trypanosoma cruzi. In this work, we tested the cytotoxic effects of
CUR and other natural curcuminoids on different forms of T. cruzi, as well as the ultrastruc-
tural changes induced in epimastigote form of the parasite. CUR was verified as the curcu-
minoid with more significant trypanocidal properties (IC50 10.13 μM on epimastigotes).
Demethoxycurcumin (DMC) was equipotent to CUR (IC50 11.07 μM), but bisdemethoxycur-
cumin (BDMC) was less active (IC50 45.33 μM) and cyclocurcumin (CC) was inactive. In
the experiment with infected murine peritoneal macrophages all diarylheptanoids were
more active than the control in the inhibition of the trypomastigotes release. The electron
microscopy images showed ultrastructural changes associated with the cytoskeleton of the
parasite, indicating tubulin as possible target of CUR in T. cruzi. The results obtained by
flow cytometry analysis of DNA content of the parasites treated with natural curcuminoids
suggested a mechanism of action on microtubules related to the paclitaxel‘s mode of
action. To better understand the mechanism of action highlighted by electron microscopy
and flow cytometry experiments we performed the molecular docking of natural
Accepted: August 11, 2016
Published: September 22, 2016
Copyright: © 2016 Sueth-Santiago et al. This is an
open access article distributed under the terms of
the Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.
Funding: This work was supported by Fundac¸ão
Carlos Chagas Filho de Amparo à Pesquisa do
Estado do Rio de Janeiro (FAPERJ); Coordenac¸ão
´
de Aperfeic¸oamento de Pessoal de Nıvel Superior
(CAPES); Conselho Nacional de Desenvolvimento
´
´
Cientıfico e Tecnologico (CNPq); The State of São
Paulo Research Foundation (FAPESP, grant #2013/
07600-3; and CEPID-CIBFar.
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
1 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Competing Interests: The authors have declared
curcuminoids on tubulin of T. cruzi in a homology model and the results obtained showed
that the observed interactions are in accordance with the IC50 values found, since there
CUR and DMC perform similar interactions at the binding site on tubulin while BDMC do not
realize a hydrogen bond with Lys163 residue due to the absence of methoxyl groups.
These results indicate that trypanocidal properties of CUR may be related to the cytoskele-
tal alterations.
that no competing interests exist.
Introduction
Curcuma longa is a plant belonging to the Zingiberacea family, which is endemic in South and
Southeast of Asian Continent, being cultivated in many countries worldwide [1]. The dried
and powdered roots of C. longa, known as Turmeric, have many uses as textile dyes, as herbal
medicines, or as food products, e.g. in sauces such as curry; which points the great relevance of
the knowledge about the biological properties of its chemical components. The main special
metabolites present in C. longa belong to the sesquiterpene and diarylheptanoid classes. Tur-
merone, curcuphenol and curlone (Fig 1) are the main sesquiterpenes in turmeric, with the
concentrations of 30.0, 10.6 and 10,0% in the plant rhizomes, respectively [2]. Turmerone, one
of the components responsible for the aroma of the C. longa essential oil, shows important bio-
logical properties such as inhibition of platelet aggregation [3] and antidiabetic [4]. Another
important class of special metabolites present in C. longa are the diarylheptanoids, structurally
related to curcumin (CUR), which is the most representative example of this group [5]. Besides
CUR, two other minor diarylheptanoids are found in C. longa, demethoxycurcumin (DMC)
and bisdemethoxycurcumin (BDMC, Fig 1). It’s also described in the literature the detection of
a curcuminoid non-diarylheptanoid (cyclocurcumin, CC, Fig1) in the crude mixture [6]. CUR
was originally isolated in 1818 by Vogel, and its chemical structure was elucidated in 1910 by
Lampe [7]. Certainly, the majority of the biological activities assigned to turmeric since ancient
times are due to natural diarylheptanoids [8]. The α,β-unsaturated dihydropyranone moiety
present in CC is formed through a Michael-type addition cyclization of curcumin.
The structural similarity between the three diarylheptanoids present in C. longa compounds
gives them similar physico-chemical properties, which hampers the discrimination of each
component during their isolation from the biological matrix. Due to this fact, the commercial
curcumin is actually a mixture of curcuminoids enriched in CUR (75–81%), but containing
also DMC (15–19%) and BDMC (2–6%). The isolation of these mixed components may be
performed in several ways, such as solvent-based extractions, percolations and extractions
assisted by supercritical fluid and microwave [9].
CUR is associated with the modulation of several biological activities such as anti-inflam-
matory effects, hepatoprotective, antioxidant and notorious chemotherapeutic properties
against bacteria, fungi, protozoa and tumor cells [1,2,5]. In the application of CUR in the treat-
ment of parasitic diseases, it should be highlighted the possibility of its use in the chemotherapy
of neglected diseases. These illnesses are so called due to low investment from the pharmaceuti-
cal industry and public health systems in research and development of drug candidates that
could lead to its cure. The biggest problem of this issue is the fact that the neglected diseases
such as malaria, leishmaniasis, trypanosomiasis and other parasitic maladies, together with
tuberculosis, had only 0.1% of global investment in health research while these diseases
together contribute about 5% of the global disease burden [10]. Despite their impact on popu-
lation health, there are no effective drugs. Chagas disease, for example, is an illness with
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
2 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Fig 1. Structures of the main chemical constituents from Curcuma longa.
doi:10.1371/journal.pone.0162926.g001
relevance in the American continent, being endemic in most countries on its Southern part.
The etiologic agent of Chagas disease is Trypanosoma cruzi (Trypanosomatidae) a flagellated
protozoan circulating the peripheral blood of their vertebrate host and multiply within cells of
vital organs (e.g. heart and liver), causing extensive cell death and tissue damage [11].
In the development of new drugs for the treatment of parasitic infections selectivity is a very
important parameter. The toxicity of the new chemical entities must be evaluated both against
the cells of the disease`s etiologic agent as well as the host cells [12]. CUR is, a priori, innocuous
to the human organism since it is present in considerable amounts in human diet, thus it is
possible to ensure a relatively safe administration. Previous studies carried out by our group
have shown that CUR is active against Leishmania amazonensis, possibly due to its structural
similarity with pentamidine, a drug employed in the treatment of leishmaniasis [13]. Although
Nagajyothi and co-workers had shown in a previous study that commercial CUR was able to
provide protection against infection of T. cruzi both in vitro and in vivo, the authors performed
their work using a traded curcumin labeled as 65% pure by HPLC (probably due to the pres-
ence of minority curcuminoids). Due to this fact, the results obtained by the authors cannot be
attributed only to curcumin, as reported in the work [14]. Furthermore, Nose and co-workers
also found that CUR showed cytotoxic activity in vitro against Trypanosoma brucei [15]. These
previous results reported in the literature encouraged us to investigate more accurately the
activity of CUR as well as the other minor natural diarylheptanoids (DMC, BDMC and CC),
each in it`s pure form (purity > 95% assessed by HPLC-DAD), against the different evolution-
ary forms of T. cruzi.
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
3 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Results and Discussion
Chemistry
A mixture of curcuminoids present in commercially available curcumin after chromatographic
separation by TLC and HPLC (as shown in S1 Fig) had each component isolated, purified and
fully characterized. This mixture present in the commercial material is not resolved by thin
layer chromatography (TLC) using hexanes and ethyl acetate as eluent, but it is possible to
identify three spots when methanol and dichloromethane (2:98) are used. Several studies in the
literature report the use of commercial curcumin in biological assessments without prior puri-
fication, so the results described in these works involve actually the evaluation of the three cur-
cuminoids as a mixture (CUR, DMC and BDMC, Fig 1).
The recrystallization of the mixture in methanol yielded pure CUR, followed by separation
of the other two minor components present in the mother liquor (DMC and BDMC) by col-
umn chromatography on silica-gel. Due to the fact that CC (Fig 1) occurs in very small
amounts in natural matrix, there are no previous studies on biological evaluation of this mole-
cule. So as CC was not detected in the mixture used in this study, it was synthesized in order to
determine the influence of this scaffold modification in the trypanocidal properties of the three
diarylheptanoids.
Isolation of the curcuminoids. Commercial curcumin was subjected to recrystallization
in order to isolate pure curcumin. CUR was obtained in suitable purity after successive recrys-
tallizations from a mixture of methanol/water (7:3). The mother-liquor enriched in DMC and
BDMC was then subjected to separation by column chromatography on silica-gel. The silica
used was previously adsorbed with NaH2PO4 and activated at 200°C overnight to minimize the
oxidation of curcuminoids [16]. Dichloromethane was used as eluent and the fractions col-
lected were grouped according to their chromatographic profiles. By this way CUR, DMC and
BDMC were obtained in suitable amounts for their characterization by spectroscopic methods
of analysis and in purities over 95% (evaluated by HPLC).
Synthesis of symmetrical curcuminoids. Since BDMC was collected in small amounts
from the commercial mixture, an experimental procedure was stabilished for the synthesis of
symmetrical curcuminoids (CUR and BDMC) in suitable yield and purity [17], allowing access
to adequate quantities to carry out all the planned biological assays. The strategy used involved
the condensation reaction between acetylacetone (2,4-pentanedione) and two equivalents of
the corresponding aldehyde. However, the methylene hydrogens are much more acidic than
the methyl ones. To overcome this effect, boric oxide was used, which will form a boron-based
enolate, whose more acidic hydrogens are those on the methyl groups of acetylacetone-boron
enolate (Fig 2). This enolate will then react with the corresponding aldehyde, and the boron
complex can be broken after reflux with methanol and DMSO.
Synthesis of cyclocurcumin. Cyclocurcumin is a curcuminoid present in very low concen-
tration in the rhizomes of C. longa occurring as a racemic mixture. CC was originally described
by Kiuchi and co-workers [6], and it is a dihydropyranone that can undergo double bond
photoisomerization. The synthesis of CC was performed from pure curcumin in benzene, cata-
lyzed by trifluoracetic acid (Fig 3). This reaction condition was protected from light in order to
prevent the isomerization of the final product, which was isolated as a yellow gummy solid in
20% yield after chromatographic purification.
Biological Assays
The four natural compounds (CUR, DMC, BDMC and CC) were evaluated on the prolifera-
tion/survival of T. cruzi epimastigotes (Dm28c strain) in four different concentrations (100, 50,
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
4 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Fig 2. Synthesis of CUR and BDMC via acetylacetone-boron enolate.
doi:10.1371/journal.pone.0162926.g002
25 and 2 μM). The parasites were cultivated for 7 days in BHI (Brain Heart Infusion) medium
supplemented with 10% (v/v) fetal bovine serum. Controls were incubated solely and with
0.02% DMSO and the cultures were readily incubated with the natural compounds. Mobile epi-
mastigote forms were considered as viable forms, and were counted on Neubauer chambers
using phase contrast microscopy. In this model, all diarylheptanoids CUR, DMC and BDMC
were active, except the α,β-unsaturated dihydropyranone CC. CUR, DMC and BDMC showed
consistent dose-response curves (as shown in S2 Fig), allowing the calculation of their values of
inhibitory concentrations for 50% (IC50), as shown in Table 1. CUR was the most active com-
pound, with an IC50 = 10.13μM. These results suggest that the diarylheptanoid scaffold (C6-
C7-C6) is required for the trypanocidal activity. Similarly, Simon and coworkers [18] showed
that CC is inactive in the inhibition of MCF-7 tumor cells proliferation.
Fig 3. Synthesis of CC by acid catalyzed cyclization of CUR.
doi:10.1371/journal.pone.0162926.g003
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
5 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Table 1. IC50 values for epimastigotes of T. cruzi (Dm28c strain) and LD50 for two cell lines.
Compounds
Curcumin (CUR)
IC50 T. cruzi(μM)
10.13
LD50 Mø(μM)
LD50 Splenocytes(μM)
> 100
> 100
> 100
> 100
> 100
> 100
> 100
> 100
Demethoxycurcumin (DMC)
Bisdemethoxycurcumin (BDMC)
Cyclocurcumin (CC)
11.07
45.33
> 100
doi:10.1371/journal.pone.0162926.t001
The natural products have been subjected to a cytotoxicity assay on two distinct murine cell
types. These compounds were tested upon macrophages (Mø), which are host cells during T.
cruzi infection and splenocytes as lymphocytes are highly sensitive to external agents. The
maintenance of cell viability at presence of the curcuminoids may indicate their low toxicity to
mammalian cells. Peritoneal macrophages harvested from Balb/c mice were treated for 48
hours with concentrations of each test substance (S3 Fig). Splenocyte viability assays were per-
formed in cells obtained from BALB/c mice and grown from a 5x105 cells.mL-1 suspension (S4
Fig). For both cell types, viability was assessed by the trypan blue dye exclusion method,
because colorimetric methods such as XTT were not suitable due to influence of the curcumin
chromophoric properties. The results obtained on the evaluation of CUR, DMC, BDMC and
CC against T. cruzi epimastigotes as well as the results of two host cell types treated with a
100 μM concentration of each compound are summarized in Table 1.
The toxicity of diarylheptanoid derivatives against epimastigotes of T. cruzi and the mainte-
nance of cell viability in types of murine cells allowed the realization of an experimental infec-
tion in vitro. This experiment aims to study the effect of CUR on the amounts of
trypomastigotes of T. cruzi released from infected macrophages in vitro (Fig 4). Then, murine
peritoneal macrophages were infected with metacyclic trypomastigotes of Dm28c strain. Try-
pomastigote release in the supernatant was measured before and after 100 μM CUR, DMC,
BDMC and CC as well as benznidazole (Benz), used as reference drug for 7 days. The results
shown in Fig 4 demonstrate the inhibition of the three natural diarylheptanoid on the release
of trypomastigotes from infected mammalian cells, probably by a decrease in the proliferation
of intracellular amastigotes, which then do not reach a density that allows them to differentiate
into trypomastigotes leading to host cell disruption.
Since the natural diarylheptanoid displayed selective toxicity against the different develop-
mental forms of T. cruzi, we decided to elucidate the mechanisms underlying the trypanocidal
activity of the compounds. Studies about the changes produced in subcellular structures have
provided valuable information that enables a detailed understanding about the mechanisms of
action involved in antiparasitic agents at cellular level [19].
The action of the compounds on the survival and proliferation of protozoan may be related
to its effects on its metabolic pathways such as synthesis of ergosterol and polyamines, interfer-
ing with cellular architecture [20–23]. Different microbicides can trigger parasite surface pro-
trusions [19]. The treatment of T. cruzi with CUR (Fig 5) led to the formation of multiple
shortened flagella, which were also previously observed in T. cruzi incubated with arjunolic
acid which interfered with the microtubule function [24]. The rounded cell bodies displaying
multiple longitudinal invaginations suggest uncompleted cell division in curcumin-treated T.
cruzi epimastigotes. Similar images were obtained in Leishmania amazonensis promastigotes
cultured with vinblastine which presented remarkable microtubule alterations [25]. The
induced projections (Fig 5) may be associated with the flagellum membrane phospholipid con-
tent as reported by Santa-Rita et al [26]. The detachment of portions of the flagellar membrane
was reported in Trypanosoma and Leishmania following treatments with sterol biosynthesis
inhibitors [21], phorbol esters [27,28]. Fibronectin [29] and cationized ferritin-induced
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
6 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Fig 4. Murine peritoneal macrophages were infected with 105 T. cruzi trypomastigotes (Dm28c
strain). Parasite cultures were treated with 100 μM of benznidazole, CUR, DMC, BDMC and CC. After 7
days, the number of released trypomastigotes in the culture medium were determined. The data shown were
obtained from three independent experiments. Cultures were compared using unpaired T-Student test
(Graph Pad Prism). *P<0.05.
doi:10.1371/journal.pone.0162926.g004
shedding [30] are also associated with flagellar membrane displacement. The flagellar and cell
body alterations may be due at least in part to the action of CUR and derivatives upon cytoskel-
eton organization. In this regard CUR was reported to produce microtubules disorganization
in cancer cells [31–33] and in Plasmodium falciparum [34]. Similar flagellar membrane detach-
ment was reported after parasite incubation with semi-purified subfraction obtained of
Fig 5. Scanning electron microscopy of T. cruzi epimastigotes. Untreated control cells (A) displayed the
usual elongated morphology with smooth cell surface. Parasites incubated with 10.13 μM curcumin for 24 h
(B-D) presented reduced cell volume (B) as well as cell body rounding with multiple longitudinal invaginations
involving the anterior portion of the parasite (C,D). Some cells displayed multiple shortened flagella (C,
arrows). The protrusion of flagellar membrane was detected (D, arrows).
doi:10.1371/journal.pone.0162926.g005
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
7 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Anthemis tinctoria [35]. The blebbing and shedding of parasite membranes must also be kept
in mind since these microsomes or vesicles may bind to non-infected cells, which are recog-
nized by anti-T. cruzi antibodies [36]. Interestingly the extracellular vesicles also known as exo-
somes may play pivotal roles in protozoal parasitic infections, taking part in T. cruzi immune
evasion and chagasic cardiomyopathy [37].
Parasites treated with the curcuminoids DMC and BDMC, showed an apparent decrease in
cell volume (Figs 6 and 7), presumably due to loss of cytoplasmic contents, possibly promoted
by cytoskeleton disorganization.
CUR-treated epimastigotes displayed disorganized and/or displaced microtubules both
within flagella in the basal body vicinity (Fig 8) and the membrane protrusions were devoid of
cytoplasmic core, showing displacement from microtubules. These alterations point out to an
impaired cytoskeleton functioning including the membrane connection.
Since the detachment of the flagellar membrane was reported in T. cruzi and L. amazonensis
treated with sterol biosynthesis inhibitors, the results obtained in microscopy study led us to
investigate the sterol content of epimastigotes treated with CUR. Sterol 14α-demethylase
(CYP51) is the enzyme involved in conversion of lanosterol to ergosterol in the parasite cell by
oxidative removal of the sterol 14α-methyl group, being a key step in the biosynthesis of cell
membrane 14 α-demethylated sterols and consequently in the maintenance of the cell viability
[38]. Thus, this enzyme is a promising target for chemotherapy of Chagas disease [39]. The
method reported by Pinto and co-workers [40] for steroid determination on fungi was adapted
to T. cruzi epimastigotes. After saponification followed by n-hexane extraction of a T. cruzi epi-
mastigote pellet cultivated in the presence of two different concentrations of CUR, the organic
layer was concentrated and submitted to RP-HPLC analysis (as shown in S5 Fig). Although the
observations on the electron microscopy has indicated a possible inhibition of sterol biosynthe-
sis that could justify the cytotoxic activity of CUR against T. cruzi, the chromatographic analy-
sis of lanosterol/ergosterol ratio showed no significant difference between treated and
untreated control parasites. Furthermore, the analysis of sterol content of the positive control,
in which the parasites were treated with posaconazole, a known T. cruzi 14 α-demethylase
inhibitor, showed a significant increase in the ratio lanosterol/ergosterol (S5 Fig). These results
indicate that another mechanism of action must be associated with the trypanocidal activity of
this natural product. It seems that in both treatments the parasite membrane-cytoskeleton link-
age has been compromised.
The flow cytometry analysis of DNA content of the parasites showed that untreated control
exhibited a typical histogram for a normal cell population, with most cells in the G1 phase and
a smaller population in G2 phase (as seen in S6 Fig). Otherwise, treatment with CUR and DMC
resulted in an accumulation of cells in the G2/M phase, with reduction of the number of cells in
G1 phase, in a concentration dependent manner (Fig 9). The cultures treated with CUR, DMC
and BDMC at a concentration of 100 μM led to an accumulation of, respectively, 45%, 52%
and 40% of cells in the G2 phase, while in the non-treated cells, G2 population was 20%. The
ability to arrest cells in the G2/M phase is a characteristic of microtubule-interacting agents
[41]. For this reason, we used paclitaxel as positive control, a known microtubule drug [42]
that is active against Trypanosomatidae organisms [43,44]. Consistently with its action on
microtubules, paclitaxel promoted an accumulation of 57% of cells in the G2 phase (Fig 9).
Thus, the effects of curcuminoids in the cell cycle can be explained, at least partially, by the dis-
turbance of microtubules, in agreement with the mechanism of interaction with tubulin pro-
posed by the electron microscopy experiments. The arising of sub-G1 peaks possibly
correspond to apoptotic cells, what is more prominent in paclitaxel than in the curcuminoids
treated cultures and the small sub-G1populations indicate that curcuminoids promote low apo-
ptosis (Fig 9).
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
8 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Fig 6. Scanning Electron Microscopy of T. cruzi epimastigotes. Parasite were cultured with 11.07 μM
DMC for 24 h (A, B). There was apparent reduction of cell volume and membrane protrusions over cell body
and flagellar surfaces (arrowheads).
doi:10.1371/journal.pone.0162926.g006
Tubulin Homology Model
The β-α-β-heterotrimer was constructed from T. cruzi α- and β-tubulin subunits, using as tem-
plates the B, C and D chains, respectively, from the tubulin-epothilone A complex deposited in
the PDB (4I50 code) [42]. Table 2 presents details about the models of tubulin:PDB code of the
template, sequence identity, GMQE (Global Model Quality Estimation), which is a quality esti-
mation that is expressed as a number between zero and one, reflecting the expected accuracy of
a model built with that alignment and template [45], QMEAN4, which is a composite scoring
function for the estimation of the global and local model quality, consisting of four structural
descriptors [46], and the quaternary structure information of the models.
The heterotrimer was constructed with the Swiss PDB-Viewer 4.01 program by superposition
of each adequate monomeric model of T. cruzi tubulin subunit with the B, C and D chains of its
tetrameric template, 4I50. This trimeric model of T. cruzi tubulin, after energy minimization with
the GROMOS96forcefield [47], available in Swiss PDB-Viewer 4.01, presented a very low RMSD
value when superimposed to 4I50, 0.36 Å, and was used for the subsequent docking study.
Fig 7. Scanning electron microscopy of T. cruzi epimastigotes. Parasites were cultured for 72 h with
11.07 μM DMC (A e B) and 45.33 μM BDMC (C, D). Parasite cell bodies displayed wavy (A) or grooved (B)
surfaces and reduced cell volumes (A-D). Surface blebbing was observed on both cell body and flagellar
membranes (A-C). Large portions of protruded flagellar membrane were observed (D, arrow).
doi:10.1371/journal.pone.0162926.g007
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
9 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Fig 8. Scanning electron microscopy showing T. cruzi epimastigotes incubated with curcumin for 48
h (A) and transmission electron microscopy of parasites treated for 72 h (B-E). Curcumin at 10.13 μM
induced the formation of flagellar protrusions (A, arrowheads). The paraflagellar rod (B White arrowhead) is
partially disorganized (B, black arrowhead). (E). Besides intraflagellar disorganization (C, arrowhead) and
curcumin-treated cells displayed flagellar membrane detachment (C, arrows). Such disorganization was
observed in flagella (F) within the flagellar pocket lumen (D, arrowhead) and may be associated to basal
body (E, black arrowhead) alterations, which included ectopic microtubules duplets (white arrowhead).
doi:10.1371/journal.pone.0162926.g008
Fig 9. Percentage of cells in different phases of cell cycle. Cells were treated with CUR and DMC at 10,
50 and 100 μM concentrations and with 100 μM BDMC. Paclitaxel (20 μM) was employed as positive control,
and DMSO 0.02% as negative control. Results are the mean of the triplicates with standard deviation
represented by error bars. Statistical significance relative to control * P<0.05, ** P<0.01 and *** P<0.001.
The data shown were obtained from three independent experiments.
doi:10.1371/journal.pone.0162926.g009
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
10 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Table 2. T. cruzi tubulin chains models.
Tubulin Chain
β-tubulin
PDB code of template (chain)
Sequence identity (%)
GMQE
0.98
QMEAN4
-1.61
4I50 (B)
4I50 (C)
4I50 (D)
85.27
84.04
84.74
α-tubulin
0.94
-1.38
β-tubulin
0.93
-1.47
doi:10.1371/journal.pone.0162926.t002
Molecular docking. Redocking experiments were successful for both colchicine and vin-
blastine with the four scoring functions available in GOLD 5.2. The default score function,
ChemPLP [48], was chosen for the subsequent docking procedure with CUR, DMC and
BDMC because this fitness score function is claimed to be generally more effective than the
other ones for both pose prediction and virtual screening. With ChemPLP, the RMSD between
the best ranked docking poses and the corresponding co-crystal structures for vinblastine and
colchicine were 0.98 and 0.21 Å, respectively, indicating the very good quality of the ligand
pose prediction with this score function.
As expected from the results previously obtained by Chakraborti and collaborators [33]
with the B. taurus tubulin, all three curcumin derivatives were able to effectively dock into the
T. cruzi tubulin at the “curcumin binding site” located in the interface between the β- and α-
tubulin subunits. The corresponding scores of the best poses correlated quite well with the
observed IC50 data, indicating a similarity of interactions for CUR and DMC, and a poorer
interaction profile for BDMC (Table 1). Analysis of the resulting poses shows the reason for
the different profiles: although the binding poses of the three derivatives are very similar, CUR
and DMC establish hydrogen bonds with exactly the same amino acid residues, whereas
BDMC is not able to make a hydrogen bond with the side chain of Lys163 because of the
absence of the methoxy groups (Fig 10C). The absence of one methoxy group in DMC makes
no difference for the interaction profile, because only one of the methoxy groups in CUR is
involved in this hydrogen bond (with Lys163) (Table 3 and Fig 10).
Conclusions
The results obtained herein revealed natural curcumin as a hit compound for the development
of new effective chemotherapeutic agents for the treatment of Trypanosoma cruzi infections.
Selective toxic effects were observed for curcumin as well as for the two other minor natural
diarylheptanoids present in turmeric, demethoxycurcumin and bisdemethoxycrucumin,
against epimastigote forms of T. cruzi. The three compounds showed inhibitory effects onthe
release of trypomastigotes from infected macrophages in vitro. These compounds had no cyto-
toxic activity against two different murine cell types, indicating a selective toxicity against T.
cruzi cells. The inactivity of dihydropyranone derivative cyclocurcumin indicates the impor-
tance of diarylheptanoid scaffold for the trypanocidal activity. Electron microscopy studies
showed significant ultrastructural changes in parasite cells pointing out the disarray of the
cytoskeleton as the probable mechanism of action, which does not appear to be associated with
the interference in the sterol synthesis as these compounds did not inhibit T. cruzi 14α-
demethylase (CYP51), and no changes were observed in the ratio between ergosterol/lanosterol
from T. cruzi epimastigotes treated with curcumin. The presence of multiple flagella and the
uncompleted cell division observed in treated T. cruzi epimastigotes may be related to impaired
cytoskeleton functioning in cytokinesis process, possibly related to tubulin polymerization pro-
cess, as well as basal body division. The flow cytometry analysis of DNA content of the para-
sites showed that their treatments with curcuminoids resulted in an accumulation of cells in
the G2/M phase, with consequent reduction of cells in G1 phase. This set of experimental
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
11 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Fig 10. Superposition of best qualified poses of CUR (Entry A), DMC (Entry B) and BDMC (Entry C)
after docking into T. cruzi tubulin. Hydrogen atoms were omitted for clarity; carbon atoms in green:
tubulin; yellow: CUR; cyan: DMC; magenta: BDMC. (image generated with PyMOL, DeLano Scientific LLC)
doi:10.1371/journal.pone.0162926.g010
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
12 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
Table 3. ChemPLP scores and hydrogen bond distances for curcumin derivatives from docking into T. cruzi tubulin.
Ligand
ChemPLP score
Hydrogen bond distances (Å)
Glu69
2.90
Lys163
2.83
2.63
-
Ser165
2.35
His396
2.88
CUR
DMC
51.56
50.06
46.49
2.99
2.43
2.92
BDMC
2.96
2.57
3.05
doi:10.1371/journal.pone.0162926.t003
observations is consistent with the action of microtubule-interacting agents as paclitaxel, used
as positive control. Accordingly, the effects of curcuminoids in epimastigotes cells cycle can be
correlated with the results highlighted in electron microscopy experiments.
Docking studies with a T. cruzi tubulin β-α-β trimer model showed that the three natural
diarylheptanoid (CUR, DMC and BDMC) are able to effectively interact with curcumin bind-
ing site previously described in the literature, presenting an interaction profile that may explain
the differences in the experimentally observed activity for each derivative. These results explain
the effects of the three diarylheptanoids against proliferative epimastigote and amastigote
forms of T. cruzi. The effect against epimastigotes was directly observed and for amastigotes it
was indirectly assessed by the strong inhibitory effect of diarylheptanoids on the release of try-
pomastigotes from host infected cells. Other mechanisms of action, however, cannot be
excluded and further studies are necessary.
Materials and Methods
Ethics Statement
This study was carried out in strict accordance with the recommendations in the Guide for the
Care and Use of Laboratory Animals of the National Institutes of Health (USA). The protocol
was approved by the Committee on the Ethics of Animal Experiments of the Health Science
Center of the Federal University of Rio de Janeiro (CEUA-CCS, Permit Number: IMPPG 038/
16) and all efforts were made to minimize suffering.
Chemistry
Melting points were determined on a Büchi B-510 apparatus and are uncorrected. The
1H-NMR (500 MHz) and 13C-NMR (125 MHz) spectra were recorded on a Bruker Ultrashield-
Plus Spectrometer (BrukerBioSpin GmbH, Rheinstetten, Germany) operating at 500 MHz for
1H and 125 MHz for 13C. 1H and 13C-NMR shifts (δ) are reported in parts per million (ppm)
with respect to DMSO-d6 (2.50 ppm for 1H and 39.7 ppm for 13C). Chemical shifts (δ) were
reported in ppm and coupling constants (J) in Hertz [Hz]. Signal multiplicity was assigned as
singlet (s), doublet (d), doublet of doublets (dd), triplet (t), quartet (q), multiplet (m) and
broad signal (bs). All pectroscopic data are available in S1 File. All the reactions involving
microwave instrumentation used the Discover SP system (CEM Inc., Matthews, NC, USA) and
were performed in open vessel mode. Analytical thin-layer chromatography (TLC) was per-
formed on precoated silica gel plates (0.25 mm layer thickness) in an appropriate solvent and
the spots were visualized under UV light (254 nm and 365 nm).
Isolation of curcuminoids from commercial source. 2g of commercial curcumin pur-
chased from Aldrich1 were dissolved in 200 mL of methanol at heating. After complete solubi-
lization, 50 mL of distilled water were added. The solution was filtered and after cooling, the
precipitate formed was collected by vacuum filtration. After successive recrystallizations, 1.2 g
of pure curcumin were collected as an orange amorphous solid (60% yield). m.p.: 182–185°C
(Lit.: 183°C)[6]. 1H-NMR δ (ppm): 9.70 (s, 2H, -OH), 7.57 (d, 2H, J = 15 Hz), 7.35 (s, 2H), 7.18
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
13 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
(d, 2H, J = 10 Hz), 6.86–6.77 (m, 4H), 6.08 (s, 1H), 3,85 (s, 6H, -OCH3) (Fig A in S1 File); 13
C
NMR δ (ppm): 183.72, 149.84, 148.46, 141.22, 126.82, 123.64, 121.57, 116.19, 111.80, 101.37,
56.16 (Fig B in S1 File). EM/HR (369.1336) (Fig C in S1 File).
The mother liquor from curcumin recrystallization was submitted to vacuum until complete
removal of methanol, resulting in 0.72 g of an orange solid, which was used for the isolation of
the curcuminoids through a silica column. 100 g of silica gem 70–200 mesh were suspended in
100 mL distilled water. 5 g of NaH2PO4 were solubilized on this suspension. The system was
stirred for 30 minutes at room temperature. Then were added 100 mL of acetone. After 10 min-
utes of stirring, the silica was filtered and kept in an oven at 200°C overnight. The silica was
then employed in a chromatographic column used to isolate the natural curcuminoids, using
dichloromethane as eluent. The fractions were grouped according to their TLC profile and
after the removal of the solvent, the curcuminoids could be obtained in suitable purities.
Demethoxycurcumin (DMC) was isolated as an amorphous orange solid in 14.25% yield. m.
p. 179–183°C (Lit.: 181–182°C)[6]. 1H-NMR δ (ppm): 10.08 (bs, 1H, -OH), 9.72 (bs, 1H, -OH),
7.61–7.56 (m, 4H), 7.36 (s, 1H), 7.17 (s, 1H) 6.86–6.71 (m, 5H), 6.07 (s, 1H), 3,87 (s, 3H,
-OCH3) (Fig D in S1 File); 13C NMR δ (ppm): 183.78, 183.65, 160.31, 149.85, 148.49, 141.22,
140.88, 130.86, 123.72, 121.52, 121.31, 116.41, 116.17, 111.70, 101.46, 56.16 (Fig E in S1 File);
EM/HR (339.1230) (Fig F in S1 File). Bisdemethoxycurcumin (BDMC) was isolated as a red-
dish solid in 3.0% yield. m.p. 230–233°C. (Lit.: 232–234)[6]. 1H-NMR δ (ppm): 10.09 (s, 2H,
-OH), 7.58–7.54 (m, 4H), 6.83 (d, 4H, 8Hz), 6.71 (d, 2H, J = 18 Hz), 6.05 (s, 1H) (Fig G in S1
File); 13C NMR δ (ppm): 183.68, 160.28, 140.84, 130.82, 126.27, 121.23, 116.37, 101.45 (Fig H
in S1 File); EM/HR (309.1124) (Fig I in S1 File).
General procedure for the synthesis of symmetrical curcuminoids. In a round bottom
flask with 10 mL capacity containing a magnet, 2,4-pentonedione (1 g, 0.1 mol) and boric
anhydride (696 mg, 0.1 mol) were added, followed by catalytic amounts of acetic acid and mor-
pholine. Then were added two equivalents of appropriate aldehyde and the flask was submitted
to microwave irradiation (300 W) for 5 minutes in the open vessel mode. After the aldehyde
were consumed (monitored by TLC), the oil formed was solubilized with aid of ultrasound in 5
mL of methanol. After the formation of an intense red solution, 5 mL of DMSO were added
and the flask was connected to a reflux apparatus and heated at 100°C for 1h. Addition of cold
water gives an orange solid which was separated by vacuum filtration on a Buchner funnel and
purified by column chromathography on silica using suitable eluent.
Procedure for the cyclocurcumin synthesis. In a round bottom flask with 50 mL capacity,
100 mg of purified curcumin was suspended in 20 mL of dry benzene. 0.6 mL of trifluoracetic
acid was added and the system was stirred at room temperature and protected from light for
72h. The mixture was concentrated under vacuum and subjected to chromatograph. 21mg
(21% yield) of cyclocurcumin (CC) was isolated as a yellow oil. 1H-NMR δ (ppm): 7.32 (d, 1H,
J = 10 Hz), 7.06 (d, 1H, J = 5 Hz), 7.02 (m, 4H), 6.93 (d, 1H, J = 10 Hz), 6.47 (d, 1H, J = 15Hz)
5.86 (s, 1H), 5.78 (s, 1H), 5.62 (s, 1H), 5.41 (dd, 1H, J1 = 15Hz, J2 = 5Hz), 3.98 (s, 3H), 3.96 (s,
3H), 2.96 (dd, 1H, J1 = 15Hz, J2 = 10Hz), 2.68 (dd, 1H, J1 = 15Hz, J2 = 5Hz) (Fig J in S1 File);
13C NMR δ (ppm): 193.07, 168.90, 147.54, 146.61, 146.70, 146.23, 137.67, 130.38, 127.76,
122.52, 119.91, 118.93, 114.79, 114.55, 109.11, 109.00, 105.56, 80.88, 56.07, 56.00, 43.25 (Fig K
in S1 File); EM/HR (369.1337) (Fig L in S1 File).
Biological Assays
Parasites. Epimastigotes of T. cruzi (Dm28c strain) were cultured at 27°C in BactoTM Brain
Heart infusion (BHI, Becton Dickinson Company, USA) supplemented with 10 μg.mL-1
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
14 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
hemin, 0.02 g.L-1 folic acid (from Sigma-Aldrich, USA) and 10% (v/v) of heat inactivated fetal
bovine serum (FCS, Gibco/Life technologies).
Anti-epimastigote effect. Epimastigotes forms of T. cruzi, Dm28c strain (2x105 cells.mL-1),
were grown for 7 days in BHI culture medium at 26°C. Some cultures were treated with curcu-
min (CUR), demethoxycurcumin (DMC), bisdemethoxycurcumin (BDMC) or cyclocurcumin
(CC) at concentrations of 100, 50, 25 and 2 μM. Viable forms were counted in Neubauer cham-
bers under phase contrast microscopy using 40X objectives. Data representing three independent
experiments.
Cytotoxic effect on murine macrophages. Macrophages harvested from the peritoneal
cavity of BALB/c mice were plated at a concentration of 5x105 cells.mL-1 in triplicates. Cells
were treated for 48 h with the indicated doses of CUR, DMC, BDMC and CC ranging from 2
to 100 μM. Cell viability was assessed by the dye exclusion method using trypan blue and the
results show data representative of three independent experiments.
Cytotoxic effect on murine splenocytes. Splenic lymphocytes were obtained from male
BALB/c mice and plated at 5x105 cells.mL-1 in triplicate. Cells were treated for 48h with the
indicated concentrations of CUR, DMC, BDMC and CC ranging from 2 to 100 μM. Viability
was evaluated by the exclusion method using trypan blue and the data shown is representative
of three independent experiments.
Inhibition of trypomastigotes release of T. cruzi in vitro. 2.104 murine peritoneal mac-
rophages were infected with 105 (5 parasites per host cell) chemically-induced metacyclic
forms of Trypanosoma cruzi (Dm28c strain) obtained as described [49]. Some cultures were
treated with Benznidazole, CUR, DMC, BDMC and CC at the concentration of 100 μM. After
7 days the number of trypomastigotes released into the supernatant of cultured were deter-
mined. The cultures were made in triplicate and the data are representative of three indepen-
dent experiments. The culture were compared using the T-Student test unpaired (Prism Graph
Pad) pà > 0.05.
Transmission electron microscopy. Epimastigotes of Trypanosoma cruzi, Dm28c strain,
were treated with curcuminoids at its IC50 concentrations (CUR 10.13 μM, DMC 11.07 μM
and BDMC 45.33 μM) and maintained in brain heart infusion (BHI) supplemented with 10%
fetal bovine serum at 25°C ( 1°C). The protozoa were washed in PBS and fixed in Karnovsky
for 24 h. Next, the samples were post-fixed in 1% osmium tetroxide, 0.8% potassium ferricya-
nide and 5 mM CaCl2 sheltered from light for 40 minutes at room temperature and then dehy-
drated in increasing concentrations of acetone (15–100%). The samples were embedded in
epoxy Polybed (Polysciences, Inc) resin. After 72 hours at 60°C, the samples were sectioned
with ultramicrotome (Reichert, Leica), contrasted in aqueous solutions of 2% uranyl acetate for
20 minutes and 1% lead citrate for 5 minutes and observed under a transmission electron
microscope (Zeiss EM 109 a 80 kV) [19].
Scanning electron microscopy. Parasites were fixed as described above and adhered to
coverslips previously coated with a solution of poly-L-lysine. Next, dehydration was performed
in series of increasing acetone concentrations (15–100%). The samples were subjected to criti-
cal point drying and metallized with an approximately 20 nm-thick gold layer to be observed
in a scanning electron microscope JEOL model JEOL [19].
Sterol extraction. 6x106 epimastigotes of T. cruzi (Dm28c strain) were grown for 7 days in
30mL of BHI culture medium at 26°C. Some cultures were treated with curcumin (CUR, 5 and
10 μM), posaconazole (0.05 and 0.1 μM). The cultures were submitted to centrifugation at
2100 rpm for 10 minutes. The supernatants were discarded and the pellets with T. cruzi cells
were submitted to sterol extraction. The wet pellets of T. cruzi cells (wild and treated with cur-
cumin) were adjusted to 100 mg, then 3 mL of alcoholic potassium hydroxide solution (25%
m/v) was added, followed by a vigorous agitation for 1 minute. The cell suspensions were
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
15 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
incubated in a hot plate at 80°C during 1 hour. The tubes were cooled to room temperature.
Sterols were extracted by addition of 1 mL of water and 3 mL of n-hexane in each tube, fol-
lowed by a vigorous agitation for exactly 3 minutes. The organic phase was transferred to a
glass tube and the n-hexane was removed to dryness under vacuum. The extracted sterols were
dissolved in spectroscopic grade n-hexane prior to HPLC analysis [40].
Sterol analysis. Analysis of the isolated lipid fractions of T. cruzi epimastigotes were car-
ried out in an HPLC Shimadzu (LC-20AT), oven CTO 20A; Detector PDA (SPD-M20A), auto
injector Sil-10AF and controller CBM-20A. The method used was the reverse phase (C18-col-
umn Betasil-THERMO, 25 cm x 4.6 mm x 5 μm) at isocratic mode, using methanol (98%) and
acetonitrile (2%) as eluent, at a flow of 1.1 mL.min-1. The detection was carried out at 280 nm
(to ergosterol) and 243 nm (to lanosterol). The injection volume was 20 μL and the oven tem-
perature was 27°C. The steroids identification (ergosterol e lanosterol) were made through
comparison with the retention times and UV-curves with commercially-available steroids
(Sigma-Aldrich).
Flow cytometry analysis of the cell cycle. Epismatigotes forms of T. cruzi (Dm28c strain)
were grown at 28°C in BHI medium at a concentration of 2x106 cells.mL-1 in 12-well culture
plates (TPPTM, Switzerland). Test compounds (CUR, DMC and BMDC) were added in tripli-
cate at concentrations of 10, 50 and 100 μM. The controls used were the vehicle (DMSO
0.02%) and paclitaxel at 20 μM concentration. After 48 h of incubation, parasites were fixed in
ethanol 70% for 30 min at -20°C. Next, cells were centrifuged at 200 rcf for 8 min and washed
with PBS. Then, the samples were incubated for 10 min with a permeabilization buffer (0.2 M
Na2HPO4 and 0.005% Triton-X, pH 7.8). Finally, cells were centrifuged at 200 rcf for 8 min
and stained in a solution of 50 μg.mL-1 propidium iodide (PI) and 0.2 μg.mL-1 RNAse for 30
min at room temperature. DNA content was analyzed measuring PI fluorescence in a BD
Accuri C5 cytometer (Becton Dickinson Company, USA) at 488 nm excitation wavelength and
585/40 nm emission filter.
Statistical analysis. Statistical analyses were performed with GraphPad Prism 4 software,
using one-way ANOVA test. Results were expressed as mean standard error (S.E.). Differ-
ences between control and treated group were considered statistically significant when Pꢀ0.05.
Molecular Modeling
Based on the experimental results that are indicative of tubulin as a possible target for the active
compounds, we implemented a molecular docking study with T. cruzi tubulin to identify possi-
ble molecular reasons for the difference in activities observed for CUR, DMC and BDMC.
Ligands that are well-known to disturb tubulin-microtubule dynamics are colchicine, vinblas-
tine and taxol, for which specific binding sites were identified in tubulin. Jackson and collabo-
rators have shown that CUR also binds to mammalian tubulin, inducing mitotic catastrophe,
and impeding normal endothelial cell proliferation [50]. A large number of drugs are known to
inhibit tubulin by binding at one of the three characterised tubulin ligand sites—taxol, colchi-
cine and vinca-binding sites. Due to its action, it would be possible for CUR to bind at the col-
chicine or vinca-binding sites, but Chakraborti et al. [33], using fluorescence spectroscopy with
tubulin isolated from goat brains, showed that CUR binds to a fourth binding site located 32 Å
away from the colchicine-binding site.
There is no available crystallographic structure for T. cruzi tubulin, so it was necessary for
the docking study the construction of a 3D model. Microtubules are composed of two globular
protein subunits, α- and β-tubulin, that form an α,β-heteropolymer. There are sequences for
both T. cruzi α- and β-tubulin obtainable in the UniProtKB/Swiss-Prot protein sequence data-
base (Q27352 for the α-tubulin and P08562 for β-tubulin), which we used separately for
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
16 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
construction of homology models with the automated mode of the protein structure homol-
ogy-modeling server, Swiss-Model [51]. In order to obtain complete binding sites located at
the interfaces between α- and β-tubulin subunits and also between β- and α-tubulin subunits, a
β-α-β-heterotrimer was constructed with the modelled subunits.
For the three possible binding sites for CUR and its derivatives, there are crystallographic
structures of tubulin with ligands co-crystallized in the colchicine and vinca-binding sites. As a
previous test for the ability of the docking program GOLD 5.2 (CCDC Software Ltd., Cam-
bridge, UK) to find reliable solutions for the docking into tubulin, we implemented redocking
experiments at both sites. All of the scoring functions available in the program were tested
through redocking procedures with the PDB crystallographic structures 4EB6 (Ovisaries tubu-
lin co-crystallized with vinblastine) and 4O2B (Bostaurus tubulin co-crystallized with colchi-
cine). Hydrogen atoms were added to proteins structures based on ionization and tautomeric
states defined by the program. To allow the best orientation of hydrogen bonds involving the
serine, threonine, tyrosine and lysine side chains, they were set free to rotate during the docking
procedure. In the course of the searching procedure, 100,000 genetic operations (crossover,
migration, mutation) were used for each docking run. Radius of binding sites for the enzymes
was set up to 10 Å around atoms from adequate amino acids selected based on literature infor-
mation for each binding site.
Spartan’14 program [Wavefunction, Inc.] was utilised to construct and optimize CUR,
DMC and BDMC structures with the PM3 method [52]. Several poses were obtained for each
compound in all proteins, and the best-ranked pose for each one was chosen for analysis of the
interactions with the amino acid residues. In the GOLD docking program, the docking func-
tions yield fitness scores, which are dimensionless values. In each case, the score values are a
guide of how good the docking pose is, with a higher score indicating a better docking result.
Supporting Information
S1 Fig. TLC and HPLC analysis of commercial curcumin. A) TLC using hexanes/ethyl ace-
tate 50% as eluent. B) TLC using methanol/dichloromethane 2% as eluent. C) HPLC analysis
at reversed-phase column C18 using acetonitrile/water as eluent.
(TIF)
S2 Fig. T. cruzi epimastigotes (Dm28c strain) were cultivated for 7 days in BHI (Brain
Heart Infusion) culture medium supplemented with hemin, folic acid and 10% fetal bovine
serum at 26°C, and readily treated with CUR, DMC, BDMC and CC at the indicated con-
centrations ranging from 2 to 100μM. Viable forms were counted on Neubauer chambers
under phase microscopy at the seventh day. The data presented were obtained from three rep-
resentative independent experiments.
(TIF)
S3 Fig. Peritoneal macrophages obtained from BALB/c mice were cultured in triplicates at
concentration of 5x105cells.mL-1. The cells were treated for 48 hours with different concentra-
tions of CUR, DMC, BDMC and CC, ranging from 2 to 100μM. Cell viability was assessed by
exclusion method using trypan blue. The results shown were obtained from representative data
of three independent experiments.
(TIF)
S4 Fig. Lymphocytes were obtained from the spleen of BALB/c mice and cultivated at con-
centration of 5x105 cells.mL-1. The cells were treated for 48 hours with CUR, DMC, BDMC
and CC at indicated concentrations (2 to 100 μM). Cell viability was assessed by exclusion
method using trypan blue. The results shown were obtained from representative data of three
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
17 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
independent experiments.
(TIF)
S5 Fig. Representative HPLC chromatograms of unsaponified lipds from T. cruzi epimasti-
gotes. A) Negative control: epimastigotes cultivated, without treatment; B) Solvent control: epi-
mastigotes treated with DMSO 0.2%; C) Positive control: epimastigotes treated with
posaconazole 0.05 μM; D) Treatment: epimastigotes treated with curcumin 10.0 μM.
(TIF)
S6 Fig. Representative fluorescence histograms showing DNA content of cell-cycle phases
(Sub-G1, G1, S and G2) of parasites (Dm28c strain). A) Negative control (DMSO 0.02%); B)
Positive control (paclitaxel 20μM); C) Treatment with 100 μM BDMC; D) Treatment with
10 μM CUR; E) Treatment with 50 μM CUR; F) Treatment with 100 μM CUR; G) Treatment
with 10 μM DMC; H) Treatment with 50 μM DMC; I) Treatment with 100 μM DMC. The data
shown were obtained from three independent experiments.
(TIF)
S1 File. Spectral data (1H, 13C, HRMS) of CUR, DMC, BDMC and CC.
(PDF)
Author Contributions
Conceived and designed the experiments: MEFL MAVS DDR CFDL ADA CMRS.
Performed the experiments: VSS JBBM ESSA CNDC LGM GPMS.
Analyzed the data: MEFL MAVS DDR CFDL ADA CMRS VSS JBBM ESSA CNDC LGM.
Contributed reagents/materials/analysis tools: MEFL MAVS DDR CFDL ADA CMRS
MEFL MAVS DDR CFDL ADA CMRS RNC.
Wrote the paper: MEFL MAVS DDR CFDL ADA CMRS VSS.
References
1. Ammon HP, Wahl MA. Pharmacology of Curcuma longa. Planta Med. 1990; 57: 1–7.
2. Parthasarathy VA, Chempakam B, Zachariah TJ (2008) Chemistry of Spices. United Kingdom, CABI.
445 p.
3. Prakash P, Misra A, Surin WR, Jain M, Bhatta RS, Pal R, et al. Anti-platelet effects of Curcuma oil in
experimental models of myocardial ischemia-reperfusion and thrombosis. Thromb Res. 2011; 127:
111–118. doi: 10.1016/j.thromres.2010.11.007 PMID: 21144557
4. Kuroda M, Mimaki Y, Nishiyama T, Mae T, Kishida H, Tsukagawa K, et al. Hypoglycemic effects of Tur-
meric (Curcuma longa L. Rhizomes) on genetically diabetic KK-Ay mice. Biol Pharm Bull. 2005; 5:
937–939.
5. Haddad M, Sauvain M, Deharo E. Curcuma as a parasiticidal agent: a review. Planta Med. 2010; 77:
672–678. doi: 10.1055/s-0030-1250549 PMID: 21104602
6. Kiuchi F, Goto Y, Sugimoto N, Akao N, Kondo K, Tsuda Y. Nematocidal activity of turmeric: synergistic
action of curcuminoids. Chem Pharm Bull. 1993; 41: 1640–1643. PMID: 8221978
7. Miłobe¸dzka J, Kostanecki St V; Lampe V. Zur kenntnis des curcumins. Ber Dtsch Chem Ges. 1910;
43: 2163–2170.
´
8. Sueth-Santiago V, Mendes-Silva GP, Decote-Ricardo D, Lima MEF. Curcumin, the golden powder
´
from turmeric: insights into chemistry and biological activities. Quım Nova. 2015; 38: 538–552.
9. Paulucci VP, Couto RO, Teixeira TCC, Freitas LAP. Optimization of the extraction of curcumin from
Curcuma longa rhizomes. Rev Bras Farmacogn. 2012; 23: 94–100.
10. Pink R, Hudson A, Mouriès MA, Bendig M. Oportunities and challenges in antiparasitic drug discovery.
Nat Rev Drug Discov. 2005; 4: 727–740. PMID: 16138106
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
18 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
11. Clayton J. Chagas disease 101. Nature. 2010; 465: S4–S5. doi: 10.1038/nature09220 PMID:
20571553
12. Lawton P, Sarciron ME, Petavy AF. Chemotherapeutic targets for antiparasitic therapy. Drugs Fut.
2006; 31:793–809.
13. Gomes DC, Alegrio LV, Leon LL, Lima MEF. Total synthesis and anti-leishmanial activity of some cur-
cumin analogues. Arzneim Forsch Drug Rec. 2002; 52: 695–698.
14. Nagajyothi F, Zhao D, Weiss LM, Tanowitz HB. Curcumin treatment provides protection against Trypa-
nosoma cruzi infection. Parasitol Res. 2012; 110: 2491–2499. doi: 10.1007/s00436-011-2790-9
PMID: 22215192
15. Nose M, Koide T, Yabu Y, Ohta N. Trypanocidal effects of curcumin in vitro. Biol Pharm Bull. 1998; 21:
643–645. PMID: 9657056
16. Rasmussen HB, Christensen SB, Kvist LP, Karazmi A. A simple and efficient separation of the curcu-
mins, the antiprotozoal constituents of Curcuma longa. Planta Med. 2000; 66: 396–398. PMID:
10865470
17. Nichols CE, Youssef D, Harris RG, Jha A. Microwave-assisted synthesis of curcumin analogs. Arkivoc.
2006; xiii: 64–72.
18. Simon A, Allais DP, Duroux JL, Basly JP, Durand-Fontanier S, Detage C. Inhibitory effect of curcumi-
noids on MCF-7 cell proliferation and structure-activity relationships. Cancer Lett. 1998; 129: 111–
116. PMID: 9714342
19. Vannier-Santos MA, De Castro SL. Electron microscopy in antiparasitic chemotherapy: a (close) view
to a kill. Curr Drug Targets. 2009; 10: 246–260. PMID: 19275561
20. Soares-Bezerra RJ, Leon L, Genestra M. Recentes avanc¸os da quimioterapia das leishmanioses:
´
´
moleculas intracelulares como alvo de farmacos. Rev Bras Ciênc Farm. 2004; 40: 139–149.
21. Granthon AC, Braga MV, Rodrigues JC, Cammerer S, Lorente SO, Gilbert IH, et al. Alterations on the
growth and ultrastructure of Leishmania chagasi induced by squalene synthase inhibitors. Vet Parasi-
tol. 2009; 146: 25–34.
22. Menna-Barreto RF, Gonc¸alves RL, Costa EM, Silva RS, Pinto AV, Oliveira MF et al. The effects on
Trypanosoma cruzi of novel synthetic naphthoquinones are mediated by mitochondrial dysfunction.
Free Radical Biol Med. 2009; 47: 644–653.
23. Rodrigues JCF, Attias M, Rodriguez C, Urbina JA, De Souza W. Ultrastructural and biochemical alter-
ations induced by 22,26-azasterol, a Δ24(25)-sterol methyltransferase inhibitor, on promastigote and
amastigote forms of Leishmania amazonensis. Antimicrob Agents Chemother. 2002; 46: 487–499.
PMID: 11796362
24. Souza-Neta LC, Menezes D, Lima MS, Cerqueira MD, Cruz FG, Martins D, et al. Modes of action of
arjunolic acid and derivatives on Trypanosoma cruzi cells. Curr Top Med Chem. 2014; 14: 1022–
1032. PMID: 24660682
25. Borges VM, Lopes UG, De Souza W, Vannier-Santos MA. Cell structure and cytokinesis alterations in
multidrug-resistant Leishmania amazonensis. Parasitol Res. 2005; 95: 90–96. PMID: 15592939
26. Santa-Rita RM, Barbosa SH, Meirelles MN, de Castro SL. Effect of the alkyl-lysophospholipids on the
proliferation and differentiation of Trypanosoma cruzi. Acta Trop. 2000; 75: 219–228. PMID:
10708662
27. De Carvalho TU, De Souza W. Effect of Phorbol-12-Myristate-13-Acetate (PMA) on the fine structure
of Trypanosoma cruzi and its interaction with activated and resident macrophages. Parasitol Res.
1987; 74: 11–17. PMID: 3325979
28. Vannier-Santos MA, Pimenta PF, De Souza W. Effects of phorbol ester on Leishmania mexicana ama-
zonensis: an ultrastructural and cytochemical Study. J Submicrosc Cytol Pathol. 1988; 20: 583–593.
PMID: 3179996
29. Vannier-Santos MA, Saraiva EM, Martiny A, Neves A, De Souza W. Fibronectin shedding by Leish-
mania may influence the parasite-macrophage interaction. Eur J Cell Biol. 1992; 59: 389–397. PMID:
1493804
30. Saraiva EM, Vannier-Santos MA, Silva-Filho FC, De Souza W. Anionic site behavior in Leishmania
and its role in the parasite-macrophage interaction. J Cell Sci. 1989; 93: 481–489. PMID: 2606939
31. Jiang J, Jin H, Liu L, Pi J, Yang F, Cai J. Curcumin disturbed cell-cycle distribution of HepG2 cells via
cytoskeletal arrangement. Scanning. 2013; 35: 253–260. doi: 10.1002/sca.21058 PMID: 23070725
32. Blakemore LM, Boes C, Cordell R, Manson MM. Curcumin-induced mitotic arrest is characterized by
spindle abnormalities, defects in chromosomal congression and DNA damage. Carcinogenesis. 2013;
34: 351–360. doi: 10.1093/carcin/bgs345 PMID: 23125222
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
19 / 20

Natural Diarylheptanoids against Trypanosoma cruzi
33. Chakraborti S, Das L, Kapoor N, Das A, Dwivedi V, Poddar A, et al. Curcumin recognizes a unique
binding site of tubulin. J Med Chem. 2011; 54: 6183–6196. doi: 10.1021/jm2004046 PMID: 21830815
34. Chakrabarti R, Rawat PS, Cooke BM, Coppel RL, Patankar S. Cellular effects of curcumin on Plasmo-
dium falciparum include disruption of microtubules. PLoS One. 2013; 8: e57302. doi: 10.1371/journal.
pone.0057302 PMID: 23505424
35. Bittencourt NLR, Ueda-Nakamura T, Dias BP, Nakamura CV. Antitrypanosomal activity of a semi-puri-
fied subfraction rich in labdane sesquiterpenes, obtained from flowers of Anthemis tinctoria, against
Trypanosoma cruzi. Pharmacol and Pharm. 2011; 2: 47–55.
36. Pinho RT, Vannier-Santos MA, Alves CR, Marino AP, Castello Branco LR. Effect of Trypanosoma
cruzi released antigens binding to non-infected cells on anti-parasite antibody recognition and expres-
sion of extracellular matrix components. Acta Trop. 2002; 83: 103–115. PMID: 12088851
37. Mantel PY, Marti M. The role of extracellular vesicles in Plasmodium and other protozoan parasites.
Cell Microbiol. 2014; 16: 344–354. doi: 10.1111/cmi.12259 PMID: 24406102
38. Doyle PS, Chen CK, Johnston JB, Hopkins SD, Leung SS. A nonazole CYP51 inhibitor cures Chagas
disease in a mouse model of acute infection. Antimicrob Agents Chemother. 2010; 54: 2480–2488.
doi: 10.1128/AAC.00281-10 PMID: 20385875
39. Choi JY, Podust LM, Roush WR. Drug strategies targeting CYP51 in neglected tropical diseases.
Chem Rev. 2014; 114:11242–11271. doi: 10.1021/cr5003134 PMID: 25337991
40. Pinto E, Afonso C, Duarte S, Vale-Silva L, Costa E, Sousa E, et al. Antifungal activity of xanthones:
evaluation of their effect on ergosterol biosynthesis by high-performance liquid chromatography.
Chem Biol Drug Des. 2011; 77: 212–222. doi: 10.1111/j.1747-0285.2010.01072.x PMID: 21244637
41. Castedo M, Perfettini J, Roumier T, Andreau K, Medema R, Kroemer G. Cell death by mitotic catastro-
phe: a molecular definition. Oncogene. 2004; 23: 2825–2837. PMID: 15077146
´
42. Prota AE, Bargsten K, Zurwerra D, Field JJ, Dıaz JF, Altmann KH, et al. Molecular mechanism of
action of microtubule-stabilizing anticancer agents. Science. 2013; 339: 587–590. doi: 10.1126/
science.1230582 PMID: 23287720
´
´
43. Dostal V, Libusova L. Microtubule drugs: action, selectivity, and resistance across the kingdoms of life.
Protoplasma. 2014; 251: 991–1005 doi: 10.1007/s00709-014-0633-0 PMID: 24652407
44. Baum SG, Wittner M, Nadler JP, Horwitz SB, Dennis JE, Schiff PB, et al. Taxol, a microtubule stabiliz-
ing agent, blocks the replication of Trypanosoma cruzi. Proc Natl Acad Sci. 1981; 78: 4571–4575.
PMID: 6117077
45. Biasini M, Bienert S, Waterhouse A, Arnold K, Studer G, Schimidt T, et al. SWISS-MODEL: modelling
protein tertiary and quaternary structure using evolutionary information. Nucleic Acids Res. 2014; 42:
W252–258. doi: 10.1093/nar/gku340 PMID: 24782522
46. Benkert P, Biasini M, Scwede T. Toward the estimation of the absolute quality of individual protein
structure models. Bioinformatics. 2011; 27: 343–350. doi: 10.1093/bioinformatics/btq662 PMID:
21134891
47. Van Gunsteren WF, Billeter SR, Eising AA, Hu¨nenberger PH, Kru¨ger P, Mark AE, et al. Biomolecular
Simulation: The GROMOS96 Manual and User Guide; vdf Hochschulverlag AG an der ETH Zu¨rich
and BIOMOS b.v.: Zu¨rich, Groningen, 1996.
48. Korb O, Stu¨tzle T, Exner TE. Empirical scoring functions for advanced protein-ligand docking with
PLANTS. J Inf Model. 2009; 49: 84–96.
49. Contreras VT, Salles JM, Thomas N, Morel CM, Goldenberg S. In vitro differentiation of Trypanosoma
cruzi under chemically defined conditions. Mol Biochem Parasitol. 1985; 16: 315–327. PMID:
3903496
50. Jackson SJ, Murphy LL, Venema RC, Singletary KW, Young AJ. Curcumin binds tubulin, induces
mitotic catastrophe, and impedes normal endothelial cell proliferation. Food Chem Toxicol. 2013; 60:
431–438. doi: 10.1016/j.fct.2013.08.008 PMID: 23939039
51. Schwede T, Koop J, Guex N, Peitsch MC. SWISS-MODEL: an automated protein homology-modeling
server. Nucleic Acids Res. 2003; 31: 3381. PMID: 12824332
52. Stewart JJP. Optimization of parameters for semiempirical methods I. Method. J Com Chem. 1989;
10: 209–220.
PLOS ONE | DOI:10.1371/journal.pone.0162926 September 22, 2016
20 / 20