High selective leishmanicidal activity of 3-hydroxy-2-methylene-3-(4-bromophenyl)propanenitrile and analogous compounds

Article abstract of DOI:10.1016/j.ejmech.2006.07.013

Sixteen not new aromatic compounds were prepared by one-pot reaction i.e. through Baylis-Hillman reaction and were the first time evaluated against promastigote Leishmania amazonensis and infected mammalian cells. Most of the compounds were selectively mo

Full text of DOI:10.1016/j.ejmech.2006.07.013

European Journal of Medicinal Chemistry 42 (2007) 99e102
http://www.elsevier.com/locate/ejmech
Short communication
High selective leishmanicidal activity of 3-hydroxy-2-methylene-
3-(4-bromophenyl)propanenitrile and analogous compounds
a
R.O.M.A. de Souza , V.L.P. Pereira , M.F. Muzitano , C.A.B. Falcao ,
a
b
b
~
*
B. Rossi-Bergmann b, , E.B.A. Filho , M.L.A.A. Vasconcellos
c
c,
*
a Nu´cleo de Pesquisas de Produtos Naturais, Universidade Federal do Rio de Janeiro, Bloco H, CCS, Ilha do Fund~ao,
Rio de Janeiro, RJ 21941-590, Brazil
b Instituto de Biof´ısica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil
c Departamento de Qu´ımica, Universidade Federal da Para´ıba, Campus I, Jo~ao Pessoa, PB 58059-900, Brazil
Received 18 January 2006; received in revised form 6 July 2006; accepted 21 July 2006
Available online 28 September 2006
Abstract
Sixteen not new aromatic compounds were prepared by one-pot reaction i.e. through BayliseHillman reaction and were the first time eval-
uated against promastigote Leishmania amazonensis and infected mammalian cells. Most of the compounds were selectively more active against
amastigotes than the reference drug sodium stibogluconate (Pentostam, IC50 ¼ 44.7 mM). We found that 3-hydroxy-2-methylene-3-(4-bromophe-
nyl)propanenitrile (13) was the most active (IC50 ¼ 12.5 mM) and safer compound (0.0 (0.9); % macrophage LDH release), being the lead com-
pound.
Ó 2006 Elsevier Masson SAS. All rights reserved.
Keywords: Leishmanicide; Leishmania amazonensis; BayliseHillman adducts
1. Introduction
immunopathologies and degrees of morbidity and mortality
[2]. There is so far no vaccine approved for clinical use.
Toxicity and resistance to the pentavalent antimonials,
which have been the mainstay of treatment of both visceral
leishmaniasis (VL) and cutaneous leishmaniasis (CL) during
the last 60 years are critical problems [3]. Although new drugs
have become available in recent years, including lipid formu-
lations of amphotericin B, the oral drug miltefosine for VL,
and topical paromomycin for CL, these are not entirely satis-
factory due to high cost, reported side effects or ineffective-
ness [4,5]. Thus, the currently available chemotherapy is far
from satisfactory, urging the search for new safe, affordable
and effective drugs.
The general BayliseHillman reaction [6] affords adducts
such as 1 (Fig. 1) from simple starting materials in a one-
pot reaction. All the atoms from the substrate are present in
the product (total atoms economy). This reaction was first re-
ported in 1972 [7], and involves the coupling of alkenes con-
taining electron withdrawing groups (EWG) with aldehydes,
ketones or imines. Tertiary amines are normally employed as
The leishmaniases are a complex of diseases caused by dif-
ferent species of the protozoan parasite Leishmania and a ma-
jor public health problem in many developing countries where
350 million people live at risk of infection [1]. The parasite ex-
ists in two forms: the flagellate promastigote in the female
phlebotomine sandfly vector, and the amastigote in the mam-
malian host. Amastigotes are obligate intracellular parasites of
macrophages (and are rarely of other cell types), where they
survive and multiply within the phagolysosome compartment.
The disease has traditionally been classified in three
different clinical forms, visceral (VL), cutaneous (CL) and
mucocutaneous leishmaniasis (MCL), which have different
* Corresponding authors. Tel.: þ55 83 32167413; fax: þ55 83 32167437
(M.L.A.A.V.).
E-mail addresses: bartira@biof.ufrj.br (B. Rossi-Bergmann), mlaav@
quimica.ufpb.br (M.L.A.A. Vasconcellos).
0223-5234/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2006.07.013

100
R.O.M.A. de Souza et al. / European Journal of Medicinal Chemistry 42 (2007) 99e102
OH OH OH OH
CN
CO CH
CN
CO CH
2
2
3
3
O N
2
O N
2
1
2
3
4
OAc
OH
OH
CN
F C
OH
CN
CO Me
2
CN
O
O
O N
2
F C
3
3
5
6
7
8
OCH OH
3
OH
OH
OCH OH
3
CN
H CO
CN
CN
3
CN
HO
OCH3
9
10
11
12
OH
OH
OH
OH
CN
CO Me
CO CH
2
CO Me
2
2
3
Br
N
F
13
16
14
15
Fig. 1. The synthesized MBH adducts 1e16.
catalyst and 1,4-diazabicyclo[2.2.2]octane (DABCO) is the
usual choice (Scheme 1). These adducts have been extensively
used as intermediates in organic synthesis for a variety of ap-
plications [6].
(4-bromophenyl)propanenitrile [15] (13) is a very active and
promising new lead compound against Leishmania.
2. Chemistry
In our continuing search for bioactive substances [8,9] and
in connection with our efforts towards the study of reactivity
of BayliseHillman reaction (BHR) [10,11], we initiated an in-
vestigation on the bioactivities of the aromatic BayliseHill-
man adducts against parasites and vectors transmitting
tropical diseases. First, we described an efficient synthetic pro-
tocol for the preparation of aromatic adducts that were shown
to be active against malarial parasites [12,13]. Then we pre-
sented a new class of molluscicidals against the snail Bio-
mphalaria glabrata that transmits schistosomiasis [14].
The purpose of this work is to evaluate the selective activity
of the aromatic BayliseHillman adducts presented in Fig. 1
against Leishmania amazonensis, the main causal agent of
diffuse CL, a rare clinical form very refractory to current chemo-
therapy. We also present here that 3-hydroxy-2-methylene-3-
It is very important to point out that although all the com-
pounds presented in Fig. 1 are not new, they can be efficiently
synthesized in one step through the BayliseHillman reaction
between the commercial precursor aldehydes and acrylonitrile
or methylacrylate, in presence of 1,4-diazabicicle [2.2.2]oc-
tane (DABCO) as catalyst and DMSO as solvents. This proto-
col is easy to scale up. Some compounds were prepared as
described previously by us or others [12,16e18]. The acetate
derivative 5 was prepared through a classical acetylation of the
precursor alcohol 3 [12] with acetic anhydride.
3. Experimental protocols
3.1. Preparation of the adducts
The reactions were carried out using dimethyl sulfoxide
(DMSO, 1 ml), aldehyde (0.6 mmol) and acrylonitrile
(0.6 mmol) at room temperature until all starting aldehyde was
consumed, as indicated by TLC analysis using ethyl acetate/hex-
ane (2:8 by volume). The mixtures were evaporated and filtered
through silica gel. Purification of the products was done by silica
gelcolumnchromatography using a 30 cm ꢀ 1.5 cm columnand
15 g of silica gel (230e400 mesh) with ethyl acetate/hexane (1:9
by volume) as eluent (50e95% purified yield). The adducts were
characterized from 1H NMR comparison of the compounds de-
scribed in literature.
N
XH
X
N
EWG
R
EWG
1
+
DABCO
R
R
2
R
1
2
Scheme 1. The BayliseHillman reaction: R1 ¼ alkyl, aryl, heteroaryl; R2 ¼ H,
CO2R, alkyl, X ¼ O, NTs, NSO2Ph; EWG ¼ electron withdrawing group:
COR, CHO, CN, CO2R, others.

R.O.M.A. de Souza et al. / European Journal of Medicinal Chemistry 42 (2007) 99e102
101
Table 1
Antileishmanial activity and cytotoxicity against macrophages of compounds
1e16 and reference drug Pentostam (50 mM)
3.2. Instrumentation and measurements
The antileishmanial activity of the synthesized adducts was
fluorimetrically determined in vitro against both promastigote
and amastigote forms of L. amazonensis (Josefa strain) trans-
fected with the green fluorescence protein (GFP), as previously
described [19]. Briefly, promastigotes were plated in triplicates
at 105 parasites/well with 50 mM of drugs in medium DMEM
containing 5% serum and 1% hybri-max DMSO. After 72 h
at 27 ꢁC, the fluorescence intensity was measured using
a plate-reader fluorometer [20]. For antiamastigote activity,
106 mouse peritoneal macrophages were infected with
5 ꢀ 106 fluorescent promastigotes for 4 h at 37 ꢁC, washed
and cultured for a further 72 h with the test compounds in 1%
DMSO. A relatively long incubation period was used to allow
for various cycles of parasite replication. Controls were 1%
DMSO alone or Pentostam (Welcome). The fluorescence inten-
sity of the cell monolayers was measured as indicative of para-
site growth [21]. Maximum and minimum inhibitory activities
were fluorescence units of uninfected macrophages and in-
fected cells without drugs, respectively. The IC50 values were
calculated by linear regression analysis.
Compounds
% Amastigote
inhibition
% Promastigote
inhibition
% Macrophage
LDH release
1
48.1 (ꢂ2.9)
14.4 (ꢂ5.5)
93.9 (ꢂ7.1)
80.6 (ꢂ1.4)
21.5 (ꢂ8.2)
86.0 (ꢂ0.9)
89.0 (ꢂ0.0)
9.1 (ꢂ3.2)
42.2 (ꢂ0.9)
37.6 (ꢂ3.4)
47.0 (ꢂ7.6)
67.3 (ꢂ4.8)
59.3 (ꢂ0.6)
62.8 (ꢂ1.2)
67.6 (ꢂ5.8)
25.5 (ꢂ0.5)
6.6 (ꢂ1.8)
0.0 (ꢂ3.7)
0.0 (ꢂ0.9)
41.7 (ꢂ0.0)
41.6 (ꢂ3.4)
0.0 (ꢂ5.6)
36.3 (ꢂ3.5)
36.2 (ꢂ4.3)
0.0 (ꢂ7.8)
0.0 (ꢂ2.1)
6.1 (ꢂ1.8)
0.0 (ꢂ0.4)
26.6 (ꢂ0.0)
0.0 (0.9)
2
3
4
5
6
7
8
9
10
19.7 (ꢂ0.9)
14.7 (ꢂ3.6)
35.3 (ꢂ2.9)
95.6 (ꢂ8.3)
84.3 (ꢂ0.1)
8.9 (ꢂ0.69)
33.6 (ꢂ2.2)
28.7 (ꢂ6.6)
35.3 (ꢂ2.7)
30.0 (ꢂ7.1)
22.3 (ꢂ9.9)
24.0 (ꢂ8.5)
17.5 (ꢂ3.9)
39.6 (ꢂ2.3)
32.4 (ꢂ0.9)
32.4 (ꢂ4.0)
0.0 (ꢂ0.0)
11
12
13
14
13.2 (ꢂ4.7)
0.0 (ꢂ2.5)
0.0 (ꢂ1.8)
13.2 (ꢂ0.5)
15
16
Pentostam
Values are means of triplicate samples; standard deviations are given in
parentheses.
Nitric oxide (NO) production was determined in the super-
natants of macrophages cultures that were incubated for 48 h
in the presence of 50 mM of the drugs. Interferon-gamma
(IFN-g, 10 U/ml, BD Biosciences) was used as a positive con-
trol. The NO by-product nitrite was measured using a modifica-
tion of the Griess method [22].
that displayed IC50 of 7.9 mM, 11.2 mM and 12.5 mM, respec-
tively (Table 3), whereas the most selectively active compound
was 13. The steep antiamastigote doseeresponse curve of
compound 7 (see activities in Tables 1 and 3) may be due to
its concomitant toxicity to macrophages, shared by the other
trifluoromethyl compound 6.
For cytotoxicity, drugs were tested at 50 mM on uninfected
adherent mouse peritoneal macrophages for 48 h at 37 ꢁC. The
release of the cytoplasmic enzyme lactate dehydrogenase
(LDH) into the culture medium was measured using an assay
kit (Doles Reagentes, Brazil). Maximum and minimum release
values were in cells cultured with 2% Triton X-100 or 1%
DMSO, respectively.
The preliminary analysis of the structure activity relation-
ship (SAR) of the aromatic 1e16 compounds (Fig. 1) revealed
that the presence of a high electron-attractor group such as NO2
in the aromatic moiety significantly increased the antiamasti-
gote activity but also rendered them more toxic to macrophages
(1 and 2 versus 3 and 4, Table 1). The OH group acetylation in 3
4. Results and discussion
Table 2
Production of nitric oxide by macrophages
The antileishmanial activity (promastigote and amastigote)
and the cytotoxicity results obtained in the compounds 1e16
are presented in Table 1.
Compounds
Nitrite (mM)
1
ꢃ1.6 (ꢂ0.5)
ꢃ2.1 (ꢂ0.9)
5.0 (ꢂ0.7)
2
In a general evaluation of the biological activities, the Baylise
Hillman compounds 1, 3, 4, 6, 7, 9, 11, 12 and 13 were more
active against intracellular amastigotes than against free promas-
tigotes, suggesting drug metabolization by macrophages, as
occurs with the pentavalent antimonial Pentostam [20]. The
higher activity is unlikely due to activation of nitric oxide
production by macrophages, a main leishmanicidal mecha-
nism of these cells, as none of the adducts induced significant
nitrite production in the culture medium, and in fact some even
inhibited the spontaneous NO production (negative values,
Table 2).
Compounds 1, 3, 7, 12 and 13 were either very active (3, 7,
12 and 13) and/or very safe to macrophages (1 and 13), so
their antiamastigote IC50 values were determined. The antia-
mastigote activity of those compounds were superior to Pen-
tostam (IC50 ¼ 44.7 mM), especially compounds 3, 12 and 13
3
4
ꢃ3.7 (ꢂ0.4)
0.0 (ꢂ0.0)
5
6
ꢃ4.0 (ꢂ0.4)
ꢃ3.5 (ꢂ0.1)
ꢃ2.1 (ꢂ0.7)
ꢃ2.3 (ꢂ1.3)
0.4 (ꢂ0.1)
7
8
9
10
11
12
ꢃ4.9 (ꢂ0.2)
ꢃ7.1 (ꢂ0.5)
2.5 (ꢂ0.1)
13
14
0.1 (ꢂ0.9)
15
16
ꢃ4.8 (ꢂ0.4)
ꢃ3.7 (ꢂ0.8)
29.3 (ꢂ1.3)
0.0 (ꢂ0.3)
IFN-g
Pentostam
Values are means of triplicate samples; standard deviations are given in
parentheses.

102
R.O.M.A. de Souza et al. / European Journal of Medicinal Chemistry 42 (2007) 99e102
Table 3
Antiamastigote IC50 of selected compounds
on the other different species of the protozoan parasite Leish-
mania (e.g. Leishmania chagasi, Leishmania donovani), an
endemic disease on Brazil.
Compounds
IC50 (mM)
1
49.3 (ꢂ1.2)
7.9 (ꢂ0.5)
42.8 (ꢂ1.2)
11.2 (ꢂ0.9)
12.5 (ꢂ1.0)
47.7 (ꢂ0.5)
3
Acknowledgments
7
12
This work was supported by grants from FAPESQ
13
Pentostam
~
`
(Fundac¸ao de Amparo a Pesquisa do Estado da Paraıba) and
CNPq (Conselho Nacional de Desenvolvimento Cientıfico e
´
´
´
Tecnologico) in Brazil.
lowers its cytotoxic effects (3 versus 5, Table 1). Oxygenation
of aromatic ring also seems to decrease the biological activities
(e.g. 8e11). Substitution of the nitrile group with the carboxy-
methyl group, in some cases does not alter the biological activ-
ities (e.g. 3 versus 4 and 6 versus 7). On the other hand, note that
the compound 1 is more active than 2. The effect of the aromatic
ring of carboxymethylated compounds seems to be important as
2, 14, 15 and 16 are practically devoid of activity as compared
with the active 4 and 7 compounds.
References
[1] World Health Organization, Division of Control of Tropical Diseases,
<http://www.who.int/tdr/diseases/leish/diseaseinfo.Htm/> (accessed on
January 2006).
[2] B.L. Herwaldt, Lancet 354 (1999) 1191e1199.
[3] S.L. Croft, S. Sundar, A.H. Fairlamb, Clin. Microbiol. Rev. 19 (2006)
111e126.
[4] S.L. Croft, G.H. Coombs, Trends Parasitol. 19 (2003) 502e508.
´
A molecular modeling using DTF theory and the B3LYP 6-
31þG* level, suggests that there is a great conformational
topology difference between 1 and 2. We observed in 2 an
intra-molecular hydrogen bond between the oxygen of car-
boxy group and the OH moiety on the most stable conforma-
tion. On the other hand, we did not observe hydrogen bond
between OH and the nitrogen of nitrile group of 1 in its
most stable conformation. We suppose that this 3D different
conformation can also be related to the very different leishma-
nicidal and toxicological activities of these compounds.
Compound 13 was found to be most selectively active
against amastigotes, inhibiting 84.3% of parasite growth con-
comitant with 0% release of LDH by macrophages at 50 mM.
This was likely due to the presence of the bromine atom, not
found in the less active 1. That was less toxic to macrophages
than the naphthyl group in compound 12 (26.6% LDH re-
lease). Bromine seems to be a key antileishmanial component,
as this halogen is also associated with enhanced activity of
similar chalcones [19] The IC50 of the most promising drugs
in Table 1, confirms the higher potency in relation to Pentos-
tam (Table 3).
[5] J.A. Valderrama, C. Zamorano, M.F. Gonzalez, E. Prinab, A. Fournet, Bi-
oorg. Med. Chem. 13 (2005) 4153e4159.
[6] D. Basavaiah, J. Rao, T. Satyanarayana, Chem. Rev. 103 (2003)
811e891.
[7] A.B. Baylis, M.E.D. Hillman, German Patent 2155113, 1972. Chem.
Abstr. 77 (1972) 34174q.
~
[8] L.S.M. Miranda, B.G. Marinho, S.G. Leitao, E.M. Matheus,
P.D. Fernandes, M.L.A.A. Vasconcellos, Bioorg. Med. Chem. Lett. 14
(2004) 1573e1575.
[9] L.S.M. Miranda, B.A. Meireles, J.S. Costa, V.L.P. Pereira,
M.L.A.A. Vasconcellos, Synlett (2005) 869e871.
[10] R.O.M.A. de Souza, M.L.A.A. Vasconcellos, Synth. Commun. 33 (2003)
1383e1399.
[11] R.O.M.A. de Souza, M.L.A.A. Vasconcellos, Catal. Commun. 5 (2004)
21e24.
[12] R.O.M.A. de Souza, B.A. Meireles, L.S. Sequeira, M.L.A.A. Vasconcellos,
Synthesis (2004)1595e1600(in this articlewe also describedthe synthesis
and the spectroscopic data of the compounds 1e4, 9e12 and 14).
[13] M.K. Kundu, N. Sundar, S.K. Kumar, S.V. Bhat, S.V.N. Biswas, Bioorg.
Med. Chem. Lett. 9 (1999) 731e736.
[14] M.L.A.A. Vasconcellos, T.M.S. Silva, C.A. Camara, R.M. Martins,
K.M. Lacerda, H.M. Lopes, V.L.P. Pereira, de R.O.M.A. de Souza,
L.T.C. Crespo, Pest. Manage. Sci. 62 (2006) 288e296.
[15] For the spectroscopic data of 13 see: L.R. Reddy, K.R. Rao Org. Prep.
Proced. Int. 32 (2000) 185e188.
[16] For the syntheses and spectroscopic data of 6 see: I. Beltaief, S. Hbaieb,
R. Besbes, H. Amri, M. Villieras, J. Villieras Synthesis (1998)
1765e1768.
5. Conclusion
[17] For the syntheses and spectroscopic data of 7, 15 and 16 compounds, see:
J. Cai, Z. Zhou, G. Zhao, C. Tang Org. Lett. 4 (2002) 4723e4725.
We present in this short communication a very important se-
lective leishmanicidal activity of the aromatic BayliseHillman
adducts. The compounds discussed in this paper are not new
but they could be prepared by one-pot reaction by a very simple
and efficient synthesis, essential for subsequent studies for
scale up. The higher potency and lesser toxicity than the refer-
ence drug Pentostam (the most used drug in the developing
countries) are enough to classify 1, 3, 7, 12 and 13 in the
new potential antileishmanial drugs. We also detach here the
brominated adduct 13 as a lead compound. Our present study
thus indicates 13 as a new alternative to the conventional che-
motherapeutic agent on L. amazonenesis. This new class of
leishmanicide drugs is very promising for our further studies
[18] For the synthesis and spectroscopic data of
8 see: F. Coelho,
W.P. Almeida, D. Veronese, C.R. Mateus, E.C. Silva Lopes,
R.C. Rossi, G.P.C. Silveira, C.H. Pavam Tetrahedron 58 (2002)
7437e7447.
~
[19] P. Boeck, C.B. Falcao, P.V. Leal, V. Cechinel-Filho, E.C. Torres-Santos,
R.A. Yunes, B. Rossi-Bergmann, Bioorg. Med. Chem. 16 (2006)
1538e1545.
[20] C. Demicheli, R. Ochoa, J.B.B. Silva, C.A.B. Falcao, B. Rossi-Bergmann,
~
A.L. Melo, R.D. Sinisterra, F. Frezard, Antimicrob. Agents Chemother. 48
(2004) 100e103.
[21] W.L. Roberts, J.D. Berman, P.M. Rainey, Antimicrob. Agents Chemo-
ther. 39 (1995) 1234e1239.
[22] S.A. Da-Silva, S.S. Costa, B. Rossi-Bergmann, Parasitology 118 (1999)
575e582.