Trypanocidal agents with low cytotoxicity to mammalian cell line: A comparison of the theoretical and biological features of lapachone derivatives

Article abstract of DOI:10.1016/j.bmc.2006.04.046

Starting from α- and β-lapachones, in this work we compared the biological and theoretical profile of several oxyran derivatives of lapachone as potential trypanocidal agents. Our biological results showed that the oxyrans tested act as trypanocidal agents against Trypanosoma cruzi with minimal cytotoxicity in the VERO cell line compared to naphthoquinones. The oxyran derivative of α-lapachone (7a) showed to be one of the most potent compounds. In our molecular modeling study, we analyzed the C-ring moiety and the redox center of β-lapachone molecule as the moieties responsible for the trypanocidal and cytotoxic effects on mammalian cell line. The computational methods used to delineate the structural requirements for the trypanocidal profile pointed out that the transposition of the C-ring moiety of β-lapachone, combined with its oxyran ring, introduced important molecular requirements for trypanocidal activity in the HOMO energy, HOMO orbital coefficient, LUMO density, electrostatic potential map, dipole moment vector, and calculated log P (c log P) parameter. This study could lead to the development of new antichagasic medicines based on α-lapachone analogs.

Full text of DOI:10.1016/j.bmc.2006.04.046

Bioorganic & Medicinal Chemistry 14 (2006) 5459–5466
Trypanocidal agents with low cytotoxicity to mammalian cell line:
A comparison of the theoretical and biological features
of lapachone derivatives
Vitor F. Ferreira,a,* Alessandra Jorqueira,a Alessandra M. T. Souza,a Milton N. da Silva,a
Maria C. B. V. de Souza,a Robson M. Gouveˆa,a Carlos R. Rodrigues,b,* Antonio V. Pinto,c
d
d
d
d
´
Helena C. Castro, Dilvani O. Santos, Humberto P. Araujo and Saulo C. Bourguignon
a
ˆ
Universidade Federal Fluminense, Instituto de Quı´mica, Departamento de Quı´mica Organica,
24020-150 Nitero´i, Rio de Janeiro, Brazil
bUniversidade Federal do Rio de Janeiro, Faculdade de Farma´cia, 21944-970, Rio de Janeiro, Brazil
cUniversidade Federal do Rio de Janeiro, Nu´cleo de pesquisas de Produtos naturais, 21944-970, Rio de Janeiro, Brazil
dUniversidade Federal Fluminense, Instituto de Biologia, Departamento de Biologia Celular e Molecular,
24020-150 Nitero´i, Rio de Janeiro, Brazil
Received 19 March 2006; revised 24 April 2006; accepted 28 April 2006
Available online 24 May 2006
Abstract—Starting from a- and b-lapachones, in this work we compared the biological and theoretical profile of several oxyran
derivatives of lapachone as potential trypanocidal agents. Our biological results showed that the oxyrans tested act as trypanocidal
agents against Trypanosoma cruzi with minimal cytotoxicity in the VERO cell line compared to naphthoquinones. The oxyran deriv-
ative of a-lapachone (7a) showed to be one of the most potent compounds. In our molecular modeling study, we analyzed the C-ring
moiety and the redox center of b-lapachone molecule as the moieties responsible for the trypanocidal and cytotoxic effects on mam-
malian cell line. The computational methods used to delineate the structural requirements for the trypanocidal profile pointed out
that the transposition of the C-ring moiety of b-lapachone, combined with its oxyran ring, introduced important molecular require-
ments for trypanocidal activity in the HOMO energy, HOMO orbital coefficient, LUMO density, electrostatic potential map, dipole
moment vector, and calculated logP (clogP) parameter. This study could lead to the development of new antichagasic medicines
based on a-lapachone analogs.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
parasitic infection remains part of a group of neglected
diseases, for which chemotherapy needs to be developed.6
American trypanosomiasis (Chagas’ disease) is an impor-
tant endemic disease caused by Trypanosoma cruzi, which
affects more than 20 million people in Central and South
America.1,2 Currently, the treatment of this disease
involves the use of the nitroheterocyclic benznidazole,
which produces severe side effects.3–5 In addition, the effi-
ciency of the treatment depends on the susceptibility of
different T. cruzi strains. Despite the progress made in
the study of T. cruzi biochemistry and physiology, this
Quinones have been studied for antitumor,7 molluscicid-
al,8–10 antiparasitic,11 anti-inflammatory,12 antifun-
gic,13,14 antimicrobial,15 and trypanocidal16–18 activities.
Literature points out that the biological profiles of these
molecules are centered on their ortho- or para-quinonoid
moiety. This group generally accepts one and/or two
electrons (redox cycling) to form the corresponding
radical anion or dianion species in situ.19 Thus, the
semi-quinone radicals accelerate intracellular hypoxic
conditions by producing superoxide anion.20–22 Due to
this mechanism, quinones may present cytotoxicity in
the mammalian cells, possibly by affecting enzymes such
as topoisomerases, a group of enzymes that are critical
for DNA replication in cells.23
Keywords: b-Lapachone; Trypanocidal agent; a-Lapachone; Cytotox-
icity; Computational.
*
Corresponding authors. Tel.: +55 21 26292345; fax: +55 21 26292362
(V.F.F.); e-mail addresses: cegvito@vm.uff.br; rangel@pharma.
ufrj.br
0968-0896/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.04.046

5460
V. F. Ferreira et al. / Bioorg. Med. Chem. 14 (2006) 5459–5466
b-Lapachone (1) is a 1,2-naphthoquinone (Fig. 1)
isolated from the bark of the Lapacho tree (Tabebuia
avellanedae).23,24 As other quinones, 1 possesses a
the o-quinone moiety into oxazoles (2a), imidazoles
(2b),31 and phenazine (3).32 The electron density on the
redox center of quinones structures is, indeed, affected
by the addition of different substituents to the system.33
variety
of
pharmacological
effects,
including
trypanocidal activity.25–28 However, this molecule is also
cytotoxic against several cell lines.29,30
In this work, we focused on the trypanocidal and cyto-
toxic effect on mammalian cell line of several oxyran
derivatives of lapachones and naphthoquinones. Addi-
tionally, a theoretical comparison between these com-
pounds regarding the redox center and C-ring moiety
(pyran ring) was performed to determine the structural
and stereoelectronic features that could lead to com-
pounds presenting trypanocidal activity with low cyto-
toxicity against mammalian cells (Scheme 1). In order
to analyze the role of the C-ring moiety of 1, we used
the following modifications: (a) the elimination of the
C-ring with the maintenance of the oxygen atom as is
the case of derivative 4 that possesses a methoxy group;
(b) the contraction of the C-ring from 6 to 5 members
(nor-b-lapachone) (5); (c) the replacement of the hydro-
gen closest to the dimethyl group in the C-ring by a
methanesulfonate group, a more bulky and polar substi-
tuent (6); and (d) the transposition of the C-ring moiety
(a-lapachone) (7) (Scheme 1). Considering the theoreti-
cal studies performed and the importance of the redox
center for b-lapachone activity (1), we synthesized the
oxyrans forms (1a, 4a, 5a, 6a, and 7a) from the
corresponding naphthoquinones (1, 4, 5, 6, and 7)
(Scheme 1). Theoretical parameters (HOMO’s energy,
HOMO orbital coefficients distribution, LUMO density,
dipole moment, dipole moment vector and
lipophilicity—clogP) for all derivatives were calculated
by semi-empirical method, and correlated with the
trypanocidal and cytotoxicity effects.
Although the development of new effective compounds
for the treatment of Chagas’ disease is of interest, stud-
ies for designing new trypanocidal agents based on 1
were decelerated due to its high cytotoxicity against
mammalian cells. On the other hand, the identification
of 1 as a DNA topoisomerase I inhibitor increased the
interest in developing itself as an anti-cancer agent.23
More recently, searching for new compounds with
reduced cytotoxicity while maintaining the trypanocidal
profile of 1, some derivatives were prepared modified at
the redox center. Interestingly, Neves-Pinto et al.
increased the trypanocidal activities of 1 by turning
NHPh
N
Ph
Ph
O
N
N
O
X
N
Ph
O
O
O
1
3
2a, X = O
2b, X = NH
Figure 1. Some important derivatives of 1.
O
O
O
O
CH2Cl2/Et2O
CH2N2
4
OCH3
4a
CH2Cl2/Et2O
OCH3
CH2N2
oxyran
insertion
O
O
O
O
1a
1
O
O
O
O
5a
B
5
O
A
O
C
O
O
O
O
O
O
Redox center
modification
Ring
modification
6a
O
SO3Me
6
O
SO3Me
O
O
O
O
CH2Cl2/Et2O
CH2N2
7
7a
O
O
Scheme 1. General synthetic routes and structural features of compounds.

V. F. Ferreira et al. / Bioorg. Med. Chem. 14 (2006) 5459–5466
5461
2. Chemistry
2.1. General procedures
initially with the compounds at six different concentra-
tions (50–0.3 lM), each one in quadruplicate, in order
to determine the antiproliferative 50% inhibitory con-
centration against T. cruzi (IC50). The final concentra-
tion of DMSO in the experiments never exceeded 1%.
The cultures were evaluated after 72 h of incubation.
Non-treated epimastigote forms control corresponded
to 100% of living cells with no drug treatment. Results,
calculated using the Probit statistical program, are
presented in Table 1. Cultures containing non-treated
epimastigote forms and 1.0% DMSO (v/v) were also
included as negative controls, while the lead compound
(1) was used as positive control. Crystal violet was also
used as an additional molecule for comparison such as
described elsewhere.37,38
Compound 1, 4-methoxy-1,2-naphthoquinone (4), nor-
b-lapachone (5), and a-lapachone (7) were prepared by
standard procedures.34 Reactions of the naphthoqui-
nones with diazomethane leading to oxyrans had been
previously described and their structures unequivocally
established.24,35,36 Briefly, the oxyrans (1a, 4a, 5a, 6a
and 7a) were prepared from the corresponding naphtho-
quinones (1, 4, 5, 6 or 7), by adding an ethereal solution
of freshly prepared diazomethane (excess). The mixtures
were kept at 5 °C for 48 h, and then the solvent was
evaporated under reduced pressure. The resulting com-
pounds were purified using a silica-gel chromatography
column eluted with a gradient mixture of hexane and
ethyl acetate. Spectroscopic data for compounds 1a,
4a, 5a, and 7a have been already reported elsewhere.36,37
3.2. Toxicity assay in mammalian cell
The cytotoxic effects of all compounds were evaluated in
VERO cells (ATCC, CRL-1586TM), a mammalian cell
line from the kidney of African green monkeys (Cercopi-
thecus aethiops), based on modifications over Margis
and Borojevic methods.39 Briefly, cells were maintained
at 37 °C in Dulbecco’s modified minimum essential
medium (Sigma Chemical Co., St. Louis, MO), supple-
mented with 10% fetal bovine serum (Cultilab, Campi-
nas, SP, Brazil) and Hepes buffer (Sigma) in 24-well
plates (Nuclon) (105 cells/well) during 12 h, for adhe-
sion. Then the compounds diluted in 100% DMSO solu-
tion were added to the cells at six different
concentrations, each one in quadruplicate, during 72 h.
Controls containing non-treated cells and 1.0% DMSO
(v/v) were also included. After 72 h-incubation, treated
and untreated cells were washed twice with PBS and
fixed with 0.2% formaldehyde in PBS for 15 min. Cells
were stained for 1 h at room temperature with 0.2%
CBBR-250 solution and then the stain was eluted from
the cells with 1.0 ml of 1% sodium dodecyl sulfate dur-
ing another hour.30 Quantification was performed using
the correlation (r) between cell number and the CBBR-
250 absorbance. Compounds’ toxicity on VERO cells is
shown in Table 1, expressed as the concentration that
induces 50% of VERO Cell death (CC50), calculated
using the Probit statistical program. Non-treated
cultures and 1.0% DMSO (v/v) were also included as
2.1.1. 2-Methyl sulfonate-3-dihydro-2,2-dimethyl-spiro-
[4H-1-oxafenanthrene-6,20-oxyran]-5(6H)-one
(6a).
,
Obtained in 65% as slightly yellow oil. IR mmax (cmꢀ1
1
film): 1652 (C@O), 1124 (SO2), H NMR (300 MHz,
CDCl3) d 3.51 (dd, 1H, J = 11.7, 5.5), 3.15 (dd, 1H,
J = 17.7, 5.5), 2.82 (dd, 1H, J = 17.7, 11.7), 7.12 (1H,
ddd, J = 7.4, 2.1, 1.6), 7.43 (1H, ddd, J = 12.9, 7.4,
1.6), 7.38 (1H, ddd, J = 12.9, 7.2, 2.1), 7.80 (1H, ddd,
J = 7.2, 2.1, 1.6), 3.43 (d, 1H, J = 8.0), 3.10 (d, 1H,
J = 8.0), 3.96 (3H, s), 1.85 (3H, s), 1.56 (3H, s). 13C
NMR (75 MHz, CDCl3) d 78.3, 61.0, 19.5, 107.6,
190.1, 55.1, 128.3, 122.9, 130.6, 128.1, 123.7, 135.9,
160.5, 62.1, 55.8, 21.4, 28.1.
3. Biology
3.1. Trypanocidal assay
Stock solutions of all quinones (1, 4–7) and oxyrans (1a,
4a–7a) (5 mM) were prepared in dimethylsulfoxide
(DMSO). Active T. cruzi Dm28c epimastigote forms
from 1st and 3rd days cultures in BHI medium were
counted in a Neubauer chamber using an optical micros-
copy (Olympus BX41). The parasites were incubated
Table 1. Comparison of trypanocidal activity (IC50), cytotoxicity against VERO cells (CC50), and molecular electronic properties (EHOMO, dipole,
clogP) of naphthoquinones and oxyrans derivatives
Compound
Experimental
Theoretical
a
b
IC50 (lM)
CC50 (lM)
EHOMO (eV)
Dipole (Debye)
clogP
1
0.9
11
<3.1
22
ꢀ9.22
ꢀ9.60
ꢀ9.23
ꢀ9.51
ꢀ9.61
ꢀ9.02
ꢀ9.37
ꢀ9.03
ꢀ9.31
ꢀ9.35
5.56
5.59
5.24
6.31
0.55
4.15
4.30
3.93
5.54
2.33
1.66
1.04
1.24
0.61
0.90
1.58
0.96
1.17
0.54
0.83
4
5
>50
>50
>50
12
<3.1
>50
>50
>50
32
6
7
1a
4a
5a
6a
7a
4.0
3.5
43
25
>50
>50
1.3
a IC50 represents the antiproliferative 50% inhibitory concentration against T. cruzi Dm28c epimastigote forms.
b Compounds’ cytotoxicity against VERO cells was expressed as the cytotoxic concentration that leads to 50% of VERO cells death (CC50).

5462
V. F. Ferreira et al. / Bioorg. Med. Chem. 14 (2006) 5459–5466
negative controls, while the lead compound (1) was used
as positive control. Crystal violet was also used as an
additional molecule for comparison.
diazomethane to produce the oxyran derivatives (4a,
5a, 6a, and 7a).
5.2. Drug screening
4. Molecular modeling and theoretical comparison
All compounds were prepared and tested in vitro against
epimastigote forms of T. cruzi. Our naphthoquinones’
results showed only 4 (IC50 = 11 lM) with a significant
trypanocidal activity (IC50 < 50 lM) compared to the
lead compound 1 (IC50 = 0.9 lM) (Table 1) and crystal
violet (IC50 < 3.1 lM). These data pointed out the
importance of C-ring moiety since modifications in this
region interfered directly with the trypanocidal activity
in all naphthoquinones tested, including 4 that presented
a ꢁ10-fold lower activity compared to 1.
All computations were performed using SPARTAN’04
(wavefunction Inc., Irvine, CA, 2000). Structures were
minimized and the equilibrium geometry was obtained
in vacuum using a semi-empirical AM1 module. In or-
der to evaluate the electronic properties of the AM1
minimal energy conformations, they were submitted to
a single-point ab initio calculation with a 6-31-G* basis
set of the SPARTAN’04 package. The three-dimension-
al isosurfaces of the molecular electrostatic potential
maps (MEPs) at the van der Waals contact surface rep-
resented electrostatic potentials superimposed onto a
surface of constant electron density (0.002 e/au3). They
were generated in a range from ꢀ65 to +23 kcal/mol.
These color-coded isosurface values provide an indica-
tion of the overall molecular size and location of nega-
tive (red) or positive (blue) electrostatic potentials. The
electronic properties (HOMO’s energy, HOMO orbital
coefficients distribution, LUMO density, dipole mo-
ment, dipole moment vector, and lipophilicity—clogP)
were calculated for all compounds. Theoretical logP
(clogP) was calculated at AM1 semi-empirical level
using the Villar method, included in Spartan.
The evaluation of the influence of the redox center in
1a–7a derivatives’ biological profile showed that all
oxyrans were still able to inhibit the parasite growth
with IC50 ranging from 1.3 to 43 lM. These data are
in agreement with the literature that describes oxyrans
as pharmacologically active compounds.45 Although in
a significant lower level, the derivative 1a still presented
a trypanocidal effect (IC50 = 12 lM) when compared to
1 (IC50 = 0.9 lM). This result may reinforce the impor-
tance of C-ring moiety’s structural feature that may be
sustaining the biological activity of 1a. Interestingly,
a-lapachone (7) is described as a non-producing free
radical molecule,46 and generated 7a (IC50 = 1.3 lM),
the most potent compound among the oxyrans tested
in this study.
5. Results and discussion
5.1. Rationale
The cytotoxicity (CC50) against mammalian cells (Vero
cells) was investigated and revealed all naphthoqui-
nones, except for the derivative 5, presenting a lower
cytotoxicity than 1 (CC50 < 3.1 lM) and crystal violet
(CC50 = 8.9 mM). Despite showing no trypanocidal ef-
fect, derivative 5 presented a significant degree of cyto-
toxicity against mammalian cells (CC50 < 3.1 lM)
(Table 1). This result suggests different structural
features to rule each 5 derivative’s biological profile
(trypanocidal activity and cytotoxicity).
The search for new compounds with activities against
T. cruzi, with low toxicity to mammalian cells, and with
increased efficacies during the chronic phase of Chagas’
disease continues. The C-ring moiety of a- and b-lapac-
hones is an important structural feature for its biological
activities. For instance, the natural product Rhicanth-
one, isolated from Rhinacanthus nasusts and used in folk
medicine for the treatment of cancer, hepatitis, and skin
diseases, is an isomer of 1 with different mode of
action.40,41
All oxyrans tested presented lower cytotoxicity values
(CC50 = 25 to >50 lM) when compared to the corre-
sponding naphthoquinones (Table 1). In agreement with
the literature, these data implied the direct correlation of
the redox center with the cytotoxic level in the naphtho-
quinones, as its replacement by an oxyran ring led to
compounds less toxic to the mammalian cells. Interest-
ingly the most potent compound (7a) presented a lower
cytotoxicity profile, which leads to a high selectivity
(CC50/IC50) against parasites compared to 1. Thus, our
biological results pointed 7a as a lead compound for de-
sign of new trypanocidal agents with low cytotoxicity
against mammalian cells.
In order to identify other structural feature involved in
the trypanocidal activity of 1 than the redox center, we
tested a set of related compounds presenting a modified
C-ring moiety (1, 4–7, 1a, and 4a–7a) against T. cruzi.
Since the literature had showed the strong influence of
the redox center on biological activity, we decided also
to prepare a set of closely related compounds to deter-
mine a new lead structure. It was interesting to assess
the influence of the redox center by replacing one
carbonyl by an oxyran ring. Many authors have pointed
out that oxyran derivatives are trypanocidal agents,
which target some specific proteinases.42 Natural
E-64,43 for example, and several others reported synthet-
ic derivatives44 show potent trypanocidal activity. This
transformation was carried out from the readily avail-
able naphthoquinones (1, 4, 5, 6, and 7), which were
straightforwardly treated with an ethereal solution of
5.3. Molecular modeling and theoretical studies
5.3.1. Naphthoquinones analysis. Considering the
computational studies we evaluated structural and
electronic properties of the current compounds to gain
insight on their role in modulating the trypanocidal

V. F. Ferreira et al. / Bioorg. Med. Chem. 14 (2006) 5459–5466
5463
and cytotoxicity profiles. The overall analysis showed
that the HOMO energy and lipophilicity (clogP) data
decreased for most of the naphthoquinones (except for
5) as much as the trypanocidal effect did, different from
that observed for some naphthoquinones described by
Molfetta et al. (Table 1).47 In contrast, both dipole mo-
ments (0.55–6.31 D), and dipole moment vector that was
oriented at the same direction for all compounds (except
for 7), presented no clear or direct correlation with try-
panocidal activity.
Due to the presence of the electron-withdrawing group,
derivative 6 showed a decrease of the electronic density
represented by the intense blue color in the ring systems
as noted in the molecular modeling studies (Fig. 2). The
replacement of the hydrogen closest to the dimethyl
groups in the C-ring by the substituent methanesulfo-
nate, a more polar and voluminous substituent, raised
the possibility of steric hindrance as this group clearly
occupied the largest area in this compound (Fig. 2 and
Table 1). Thus, both 6 electronic effect and steric hin-
drance may be deleterious to trypanocidal activity. Sim-
ilar to compounds 4 and 5, the dipole moment vector of
6 is directed to the oxygen atom in C-ring (not shown).
The molecular electrostatic potential maps (MEP),
HOMO orbital coefficient, and LUMO density showed
an overall similarity for all naphthoquinone derivatives,
except for 7 (Fig. 2). In a more specific analysis of the
naphthoquinones group, the elimination of the C-ring
of 1 with the maintenance of an oxygen atom as in 4 re-
vealed no significant changes in the electronic density in
the MEP in comparison to 1 (Fig. 2). This result rein-
forced the suggestion that the structural requirement
represented by C-ring is important to the trypanocidal
activity.
The last naphthoquinone studied, derivative 7, showed
the electrophilic center distributed in both carbonyls
(Fig. 2). The C-ring (pyran) moiety transposition in 7
also induced a significant decrease of the trypanocidal
activity (Table 1). This reinforces the importance of
the right orientation of the pyran ring to an ideal
interaction with the target present in the parasite. The
absence of activity in 7 is also in agreement with the
experimental pharmacological data described in the
literature.43 We also analyzed the three-dimensional
coefficient isosurface of both orbitals of 7 to determine
the atomic contribution on the frontier HOMO orbitals.
Our analysis revealed the highest HOMO contribution
for this compound in the unsaturated bond in the
B- and C-ring junction closest to the oxygen atom of
C-ring (Fig. 2). This result is different from that
observed for other derivatives, where these contribu-
tions were distributed among the ring systems (Fig. 2).
Compound 5, which presents the contraction of C-ring
from six to five members, demonstrated the influence
of the restricted flexibility and its impact in the trypan-
ocidal profile. Interestingly, derivative 5 showed no sig-
nificant variation in its electronic properties, which is
comparable to 1 (Fig. 2). However, 5 was not active
against T. cruzi and curiously the cytotoxicity had in-
creased. Thus, this result led us to infer that the six-
membered C-ring arrangement of 1 may define an oxy-
gen-specific orientation, which is probably significant
to the interaction with the target in T. cruzi. This region
is more restricted in 5 and differently oriented in 4, due
to the conformational freedom of the methoxy group,
which apparently compromised the antiparasite activity
(Fig. 2 and Table 1).
The analysis of
7 LUMO contributions showed
no significant difference from those of other naphtho-
quinones of this series (data not shown).
5.3.2. Oxyrans analysis. In order to assess structural
features and electronic properties that could affect the
Figure 2. Comparison of the naphthoquinones (1 and 4–7) and the respective compounds with oxyran ring (1a and 4a–7a). (A) Front (up) and back
view (down) of molecular electrostatic potential energy isosurfaces (MEP) superimposed onto total electron density of 0.002 e/a.u.3 The color code is
in the range of ꢀ65.0 (deepest red) to ꢀ23.0 (deepest blue) kcal/mol. (B) HOMO orbital coefficient and (C) LUMO encoded onto a surface of
constant electron density (0.002 e/a.u.3) (LUMO density).

5464
V. F. Ferreira et al. / Bioorg. Med. Chem. 14 (2006) 5459–5466
trypanocidal activity, we performed the computational
studies by analyzing the addition of the oxyran ring in
the quinone moiety. The overall analysis of theoretical
parameters showed that the dipole moment, except for
6a, and the lipophilicity (clogP) decreased for most of
the oxyrans but with no clear correlation with trypano-
cidal activity.
of the variation in some structural parameters such as
dipole moment and clogP, no direct correlation could
be strictly determined to 1a when compared to other
derivatives. Interestingly, 1a MEP showed the orienta-
tion of the methylene bridge, which could impose an
important steric hindrance and limit the perfect match
with the target (Fig. 2).
The three-dimensional molecular electrostatic potential
maps (MEP) and LUMO density provide a measure of
charge distribution from the point of view of an
approaching molecule.48 In general oxyrans, MEP
and LUMO density significantly differed from the cor-
responding naphthoquinones (Fig. 2). The MEPs re-
The trypanocidal activity of oxyrans 4a, 5a, and 6a also
reinforced the importance of other structural features
than the quinone moiety for presenting this biological
profile. Unlike the naphthoquinone derivatives 4 and
5, the addition of the oxyran group simultaneously to
the ring modulation did not abolish the trypanocidal
activity in 4a and 5a. The derivative 6a still presented
a slight level of antiparasite activity, although the volu-
minous substituent may interfere with most of the favor-
able interactions with the target (Fig. 2).
vealed
a
more negative electrostatic potential
extending at the ring systems’ surface in all oxyrans
when compared with naphthoquinone derivatives
(Fig. 2). The LUMO delineated electron deficient areas,
which may be susceptible to interaction with the elec-
tron density regions of the target of the parasite. In
Figure 2, regions where the absolute value of the
LUMO is greatest are shown in blue, while regions
where it is least are shown in red. The LUMO electron
deficient (blue) area of naphthoquinones concentrated
in both carbon carbonyls and in the carbon atom bond
to the oxygen pyran ring. In oxyrans, this area was
restricted to the carbon atom closest to the pyran oxy-
gen atom (Fig. 2). In contrast, HOMO’s orbital coeffi-
cients and the dipole moment vector showed no
apparent modifications when compared to the original
quinone derivatives (Fig. 2B).
Among the planned drugs, derivative 7a was the most
potent compound with at least 40-fold activity
improvement over 7 and at a similar level to 1 (Fig. 3
and Table 1). In addition, this oxyran was also shown
to be less toxic, even at a high concentration (50 lM),
pointing it as the best candidate for a more specific study
about its mechanism of action. The molecular modeling
analysis pointed out the variation of all electronic and
molecular properties, compared to 1, which led to the
highest trypanocidal activity of this series of compounds
without a significant cytotoxicity against mammalian
cells (Fig. 3). Apparently, the presence of the oxyran
group decreased the lipophilicity (clogP = 0.83) and
the dipole moment (2.33 D) of the compound and,
interestingly enough, also the cytotoxicity of these
compounds (CC50 > 50 lM) (Table 1).
The individual analysis of oxyran 1a showed its lower
but consistent trypanocidal effect, suggesting that the
quinone system is not essential to this activity. Despite
Figure 3. Comparison of the structural and biological features of 1 and 7a. (A) Dipole moment vector orientation, (B) stereo diagram of the
superimposition of both structures from two different views, and (C) antiparasitic profile (up) and the predicted selectivity (CC50/IC50) against
parasites (down).

V. F. Ferreira et al. / Bioorg. Med. Chem. 14 (2006) 5459–5466
5465
The dipole moment and the molecule vector also com-
References and notes
pose a relevant parameter for evaluating ligands
approaching the binding site. Although compounds 1a,
4a, 5a, and 6a showed lower dipole moments and also
lower cytotoxicity against mammalian cells values than
its precursors, its vectors pointed to the same region,
showing that the dipole region is similar for all deriva-
tives (not shown). The only exception was 7a, which
showed a higher dipole moment than its quinone precur-
sor, while its vector was in an opposite direction (Fig. 3).
This vector orientation showed that the polarity of the
molecule is different when compared to that of the lead
compound (1) (Fig. 3). This feature might be extremely
relevant for maximizing the interaction with the target
in the T. cruzi, revealing 7a as the most potent com-
pound of the oxyran series, and the closest to the biolog-
ical profile of 1. This may also be important for the
selectivity degree (CC50/IC50) of 7a to parasites against
mammalian cells, which is at least 10-fold higher than
1 (Fig. 3).
1. Williams-Blangero, S.; VandeBerg, J. L.; Blangero, J.;
Correa-Oliveira, R. Front. Biosci. 2003, 8, e337.
2. Dias, J. C.; Silveira, A. C.; Schofield, C. J. Mem. Inst.
Oswaldo Cruz 2002, 97, 603.
3. Brener, Z. Pharmacol. Ther. 1979, 7, 71.
4. Schmun˜is, G. A.; Szarfman, A.; Coarasa, L.; Guilleron,
C.; Peralta, J. M. Am. Soc. Trop. Med. Hyg. 1980, 29,
170.
5. Canc¸ado, J. R.; Brener, Z. In Trypanosoma cruzi e Doenc¸a
de Chagas; Brener, Z., Andrade, Z., Eds.; Guanabara
Koogan: Rio de Janeiro, 1979; p 362.
6. Hotez, P. J.; Remme, J. H.; Buss, P.; Alleyne, G.; Morel,
C.; Breman, J. G. Clin. Infect. Dis. 2004, 38, 871.
7. Subramanian, S.; Ferreira, M. M. C.; Trsic, M. Struct.
Chem. 1998, 9, 47.
8. dos Santos, A. F.; Ferraz, P. A. L.; de Abreu, F. C.;
Chiari, E.; Goulart, M. D. F.; Sant’Ana, A. E. G. Planta
Med. 2001, 67, 92.
9. Barbosa, T. P.; Camara, C. A.; Silva, T. M. S.; Martins,
R. M.; Pinto, A. C.; Vargas, M. D. Bioorg. Med. Chem.
2005, 13, 6464.
In brief, all the molecular modeling calculations
suggested that these compounds might act by a different
mechanism against the T. cruzi. In addition, the
cytotoxicity data also suggest that the new oxyran
derivatives are active in the parasite without critically
affecting the mammalian cells, in contrast to the
naphthoquinones. The superimposition of 1 and 7a
showed that the six-membered ring occupies different
regions, which might be extremely important to bind
adequately with the target in the parasites (Fig. 3).
10. Silva, T. M. S.; Camara, C. A.; Barbosa, T. P.; Soares, A.
Z.; da Cunha, L. C.; Pinto, A. C.; Vargas, M. D. Bioorg.
Med. Chem. 2005, 13, 193.
11. Teixeira, M. J.; Almeida, Y. M.; Viana, J. R.; Holanda
Filha, J. G.; Rodrigues, T. P.; Prata, J. R. C., Jr.; Coelho,
I. C. B.; Rao, V. S.; Pompeu, M. M. L. Phytother. Res.
2001, 15, 44.
12. Almeida, E. R. J. Ethnopharmacol. 1990, 29, 239.
13. Marr, J. J.; Docampo, R. Rev. Infect. Dis. 1986, 8,
884.
14. Garnier, S.; Wolfender, J. L.; Nianga, M.; Stoeckli-
Evans, H.; Hostettmann, K. Phytochemistry 1996, 42,
1315.
15. Abreu, F. C.; Ferreira, D. C. M.; Wadhawan, J.; Amatore,
C.; Ferreira, V. F.; da Silva, M. N.; de Souza, M. C. B. V.;
Gomes, T. S.; Ximenes, E. A.; Goulart, M. O. F.
Electrochem. Commun. 2005, 7, 767.
16. Ferreira, V. F.; Pinto, A. V.; Silva, J. S.; Gilbert, B. Mem.
Inst. Oswaldo Cruz. 1992, 87, 345.
17. Pinto, C. N.; Dantas, A. P.; De Moura, K. C. G.; Emery,
F. S.; Polequevitch, P. F.; Pinto, M. C. F. R.; De Castro,
S. L.; Pinto, A. V. Arzneim. Forsch./Drug Res. 2000, 50,
1120.
18. Menna-Barreto, R. F. S.; Henrique-Pons, A.; Pinto, A. V.;
Morgado-Diaz, J. A.; Soares, M. J.; De Castro, S. L. J.
Antimicrob. Chemother. 2005, 56, 1034.
19. Babior, B. M. Braz. J. Med. Biol. Res. 1997, 30, 141.
20. Santos, E. V. M.; Carneiro, J. W. M.; Ferreira, V. F.
Bioorg. Med. Chem. 2004, 12, 87.
21. Silva, M. N.; Ferreira, V. F.; de Souza, M. C. B. V. Quim.
Nova. 2003, 26, 407.
22. Monks, T. J.; Hanzlik, R. P.; Cohen, G. M.; Ross, D.;
Graham, D. G. Toxicol. App. Pharmacol. 1992, 112, 2.
23. Pardee, A. B.; Li, Y. Z.; Li, C. J. Curr. Cancer Drug
Targets 2002, 2, 227.
24. Pinto, A. V.; Pinto, M. C. R.; Gilbert, B.; Pellegrino, J.;
Mello, R. T. Trans. Roy. Soc. Trop. Med. Hyg. 1977, 71,
133–135.
6. Conclusion
The presence of the redox center of the quinones is de-
scribed in the literature as important to maintain the
antiproliferative activity against T. cruzi. In this work,
we designed and synthesized a set of related compounds
with modifications in the C-ring moiety and on the qui-
none moiety to correlate them with antiparasite, cyto-
toxicity, and molecular properties. This study pointed
out the naphthoquinones with an oxyran ring as a new
class of trypanocidal compounds with low cytotoxicity
against mammalian cells. Despite presenting totally dif-
ferent electronic properties, 1 and 7a showed a similar
activity profile against the parasites. These features
may account for its different levels of reactivity, and this,
on the other hand, could lead to different pharmacolog-
ical mechanisms. In agreement with this hypothesis,
compound 7a showed a lower cytotoxicity in the VERO
cell line, which shows it as a potential lead compound
for the future design of drugs to be used in the treatment
of Chagas’ disease.
´
´
25. Ravelo, A. G.; Estevez-Braun, A.; Chavez-Orellana, H.;
Perez-Sacau, E.; Mesa-Siverio, D. Curr. Top. Chem. 2004,
Acknowledgments
´
4, 241–244.
26. Chiang, J.; Averboukh, L.; Pardee, A. B. J. Biol. Chem.
1993, 268, 22463–22468.
27. Frydman, B.; Marton, L. J.; Sun, S. J.; Neder, K.; Witiak,
T. D.; Liu, A. A.; Wang, H. M.; Mao, Y.; Wu, H. Y.;
Sanders, M. M.; Liu, L. F. Cancer Res. 1997, 57, 620–627.
Fellowships granted to A.J. and M.N.S. by CAPES and
to M.C.B.V.S., C.R.R., A.V.P., and V.F.F. by CNPq
(Brazil) are gratefully acknowledged. This work was
partially supported by CNPq, FAPERJ, FUJB, and
UFF.

5466
V. F. Ferreira et al. / Bioorg. Med. Chem. 14 (2006) 5459–5466
28. Dubin, M.; Fernadez Villamil, S. H.; Stoppani, A. O.
Medicina 2001, 61, 343–350.
29. Frydman, B.; Marton, L. J.; Sun, J. S. Cancer Res. 1995,
57, 620–627.
30. Pinto, A. V.; Pinto, M. C. R.; Gilbert, B.; Pellegrino, J.;
Mello, R. T. Trans. Roy. Soc. Trop. Med. Hyg. 1977, 71,
133.
31. De Moura, K. C. G.; Saloma˜o, K.; Menna-Barreto, R. F.
S.; Emery, F. S.; Pinto, M. C. F. R.; Pinto, A. V.; Castro,
S. L. Eur. J. Med. Chem. 2004, 39, 639.
32. Neves-Pinto, C.; Malta, V. R. S.; Pinto, M. C. F. R.;
Santos, R. H. A.; de Castro, S. L.; Pinto, A. V. J. Med.
Chem. 2002, 45, 2112.
33. Abreu, F. C.; Ferraz, P. A. L.; Goulart, M. O. F. J. Braz.
Chem. Soc. 2002, 13, 19.
38. de Souza, A. O.; Hemerly, F. P.; Gomes-Cardoso, L.;
Santa-Rita, R. M.; Leon, L. L.; de Castro, S. L.; Duran,
N. J. Chemother. 2004, 16, 530.
39. Margis, R.; Borojevic, R. Anal. Biochem. 1989, 181, 209.
40. Kongkathip, N.; Luangkamin, S.; Kongkathip, B.; Sang-
ma, C.; Grigg, R.; Kongsaeree, P.; Prabpai, S.; Pradid-
phol, N.; Piyaviriyagul, S.; Siripong, P. J. Med. Chem.
2004, 47, 4427.
41. Kongkathip, N.; Kongkathip, B.; Siripong, P.; Sangma,
C.; Luangkamin, S.; Niyomdecha, M.; Pattanapa, S.;
Piyaviriyagul, S.; Kongsaeree, P. Bioorg. Med. Chem.
2003, 11, 3179.
42. Cazzulo, J. J. Curr. Top. Med. Chem. 2002, 2, 1257.
43. Lacobuzio-Donahue, C. A.; Shuja, S.; Cai, J.; Peng, P.;
Murname, M. J. J. Biol. Chem. 1997, 272, 29190.
´
34. Hooker, S. C. J. Chem. Soc. 1892, 611.
35. Ferreira, V. F.; Pinto, A. V.; Pinto, M. C. R. F.; da Silva,
M. N.; Skakle, J. M. S.; de Souza, M. C. B. V.; Wardell, S.
M. S. V. Acta. Crystallogr. 2002, C58, 560.
36. Silva, M. N.; de Souza, M. C. B. V.; Ferreira, V. F.; Pinto,
A. V.; Pinto, M. C. R. F.; Wardell, S. M. S. V.; Wardell, J.
L. Arkivoc 2003, 10, 156.
44. Roush, W. R.; Gonzalez, F. V.; McKerrow, J. H.; Hansell,
E. Bioorg. Med. Chem. Lett. 1998, 8, 2809.
45. Marco-Contelles, J.; Molina, M. T.; Anjum, S. Chem. Rev.
2004, 104, 2857.
46. Krishnan, P.; Bastow, K. F. Biochem. Pharmacol. 2000,
60, 1367.
´
47. Molfetta, F. A.; Bruni, A. T.; Honorio, K. M.; da Silva, A.
37. Gadelha, F. R.; Moreno, S. N.; De Souza, W.; Cruz,
F. S.; Docampo, R. Mol. Biochem. Parasitol. 1989,
34, 117.
B. F. Eur. J. Med. Chem. 2005, 40, 329.
48. Bhattacharjee, A. K.; Majundar, D.; Guha, S. J. J. Chem.
Soc., Perkin Trans. 2 1992, 805.