Synthesis and structure-activity relationship of 3-phenylquinoxaline 1,4-di-N-oxide derivatives as antimalarial agents

Article abstract of DOI:10.1016/j.ejmech.2007.11.024

As a continuation of our research and with the aim of obtaining new antimalarial agents, new series of 3-phenylquinoxaline 1,4-di-N-oxide derivatives have been synthesized following the classical Beirut reaction. Antiplasmodial activity was evaluated in v

Full text of DOI:10.1016/j.ejmech.2007.11.024

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 43 (2008) 1903e1910
http://www.elsevier.com/locate/ejmech
Original article
Synthesis and structureeactivity relationship of 3-phenylquinoxaline
1,4-di-N-oxide derivatives as antimalarial agents
Esther Vicente a,b, Lidia M. Lima a, Emily Bongard b, Sarah Charnaud b,
a
Raquel Villar , Beatriz Solano , Asuncion Burguete , Silvia Perez-Silanes ,
a
a
a
´
b
a
Ignacio Aldana a, , Livia Vivas , Antonio Monge
*
a Unidad de Investigacio´n y Desarrollo de Medicamentos, Centro de Investigacio´n en Farmacobiolog´ıa Aplicada (CIFA),
University of Navarra, C/Irunlarrea s/n, 31080 Pamplona, Spain
b Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine (LSHTM),
Keppel Street, London WC1E 7HT, United Kingdom
Received 3 October 2007; received in revised form 23 November 2007; accepted 27 November 2007
Available online 22 January 2008
Abstract
As a continuation of our research and with the aim of obtaining new antimalarial agents, new series of 3-phenylquinoxaline 1,4-di-N-oxide
derivatives have been synthesized following the classical Beirut reaction. Antiplasmodial activity was evaluated in vitro against Plasmodium
falciparum by the incorporation of [3H]-hypoxanthine. Cytotoxicity was tested in KB cells by AlamarBlue assay. Twenty-one of the 60 com-
pounds that were assayed against 3D7 (CQ-sensitive) showed enough activity to be also evaluated against K1 (CQ-resistant) strain. Ten of them
were shown to be more active than chloroquine in the resistant strain. The most interesting compounds are 7-(methyl or methoxy)-3-(40-fluoro or
chloro)phenylquinoxaline-2-carbonitrile 1,4-di-N-oxides because of their low IC50 and their high SI shown for the K1 strain, making them valid
new leads.
Ó 2007 Elsevier Masson SAS. All rights reserved.
Keywords: Quinoxaline; N-Oxide; Beirut reaction; Antiplasmodial; Malaria; Plasmodium falciparum
1. Introduction
of the most deadly form of malaria [3]. Thus, renewed efforts
are required to develop novel and affordable antimalarials to
overcome the detrimental effects of drug resistance [4], partic-
ularly against resistant strains of P. falciparum.
In terms of human suffering, malaria remains one of the
most important health threats to the developing world. Further-
more, the worldwide burden of malaria is increasing, partly
due to the unfortunate spread of resistance to most of the drugs
that were once effective and safe [1,2]. For nearly half a cen-
tury, chloroquine (CQ) has been the primary therapy of choice
[1] but today CQ-resistant Plasmodium falciparum is observed
in nearly all of the malaria-endemic regions, being the cause
The quinoxaline derivatives seem to have very interesting
biological properties [5e7]. Throughout the past 20 years,
many mono- and di-N-oxide derivatives of this heterocyclic
system have been prepared and their biological activities
against Mycobacterium tuberculosis [8e18], Trypanosoma
cruzi [19e21], Candida spp. [13,16], P. falciparum [22e25]
and different human tumor cell lines [26e29] have been
reported. More recently we have identified a derivative, 3-
(40-chloro)phenylquinoxaline-2-carbonitrile 1,4-di-N-oxide 5,
with potent activity against a chloroquine-resistant strain of
P. falciparum (FcB1) [24]. In our continuing efforts to identify
new antimalarial drug candidates, we herein describe the
Abbreviations: CQ, chloroquine; ND, non data; P.f., Plasmodium falcipa-
rum; Qx, quinoxaline derivative; RI, resistance index; SD, standard derivation;
SI, selectivity index.
* Corresponding author. Tel.: þ34 948 425653; fax: þ34 948 425652.
E-mail address: ialdana@unav.es (I. Aldana).
0223-5234/$ - see front matter Ó 2007 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2007.11.024

1904
E. Vicente et al. / European Journal of Medicinal Chemistry 43 (2008) 1903e1910
synthesis, the in vitro biological evaluation and the study of
structureeactivity relationship of quinoxaline 1,4-di-N-oxide
derivatives designed as potential antimalarial agents
(Scheme 1).
optimization of therapeutic efficacy. Trifluoromethyl group
was recently described as an important subunit used to re-
place the chlorine atom in chloroquine (CQ), resulting in
a new derivative with faster cell-membrane penetration and
intracellular accumulation [32]. In addition, it appears that
the introduction of a trifluoromethyl substituent at the R6/
R7 position contributes to good physicochemical properties
and to therapeutic efficacy in an animal model [33]. Trifluor-
omethyl group (CF3) is also present in the current antimalar-
ial drugs, mefloquine and halofantrine; it is also found in the
scaffold of tafenoquine (8-aminoquinoline derivative) and
arteflene (an endoperoxide which is structurally simpler
than artemisinin), promising compounds in advanced stages
of clinical development [34].
In series II and III, the replacement of carbonitrile moiety
by a carboxylic acid subunit (49e53) or by an ester group
(54e63), respectively, was performed in order to study the
structural contribution of the carbonitrile group, present in
the lead-compound 5, to the antimalarial activity. These mod-
ifications were proposed in an attempt to study the importance
of the pKa, molecular volume and hybridization of C-atom
(from sp to sp2) in position 2 of the 1,4-di-N-oxide quinoxaline
scaffold, within the optimization process of prototype 5.
According to our previous results, compound 3-phenylqui-
noxaline-2-carbonitrile 1,4-di-N-oxide 5, was identified as
a lead-compound with promising in vitro anti-falciparum
activity. In attempts to establish the structural requirements
necessary for inhibition of P. falciparum, three new series of
quinoxaline derivatives were proposed based on structural
modifications of the lead-compound 5. These modifications
can be summarized as follows: (a) variations in the electronic
pattern of the substituent present in 1,4-di-N-oxide quinoxa-
line ring and in the phenyl group linked to C-4 (series I);
and (b) substitution of nitrile subunit, present in the lead-
compound 5, by an ester group in series II and by a carboxylic
acid moiety in series III.
In series I, the modification made in 4-chloro-phenyl sub-
unit of lead-compound 5 was performed in order to understand
the relevance of the electronic parameters in the optimization
of the antimalarial activity. It was based on the replacement of
the chlorine atom (sp ¼ 0.23), present in the lead-compound 5,
by new substituents such as: a fluorine atom (sp ¼ 0.06; 17e
22); a methoxy group (sp ¼ ꢀ0.27; 23e28); or a trifluorome-
thoxy group (sp ¼ 0.35; 29e34) [30].
2. Chemistry
The introduction of a fluorine atom (35e41) or a trifluoro-
methyl (42e48) group in C-7 (R7) of the 1,4-di-N-oxide qui-
noxaline system was proposed so as to be able to modify the
electronic profile of the quinoxaline ring (Scheme 1), improv-
ing the interactions of the fluorinated molecule with the re-
ceptor site [31]. In addition, the electronegative centre
which originates due to the proposed modifications (H versus
F and CF3) can introduce new hydrogen bond interactions (as
hydrogen acceptor) and alter physicochemical properties such
as solubility and log P. This could result in an improvement
of activity, an increase in stability, or in a reduction of toxic-
ity to mammalian cells, thereby contributing to the
The aforementioned compounds were prepared according
to the synthetic process illustrated in Scheme 2. The starting
compounds, 5-substituted or 5,6-disubstituted benzofuroxanes,
BF(aei), were obtained by previously described methods [27].
The functionalized carbonitrile derivatives 1e48 were ob-
tained by the classical Beirut reaction [24,35]. The appropriate
benzofuroxane and the corresponding arylacetonitrile were
dissolved in dry dichloromethane in the presence of triethyl-
amine, which acted as the catalyst (Scheme 2). When the
reaction was finished, the solvent was evaporated to dryness
under reduced pressure and a yellow crude solid or brown
oil was obtained. The purification methods used for each com-
pound are shown in Section 7.
The synthesis of compounds 54e63 was carried out explor-
ing the condensation of benzofurazan oxide with ethyl
functionalized benzoylacetate in the presence of potassium
carbonate, using acetone as the solvent [36]. The attempts to
hydrolyze these esters using different basic conditions in order
to obtain carboxylic acids derivatives 49e53 were not success-
ful, resulting in decarboxylation. Due to the aforementioned,
the syntheses of the acids were then performed; this was
carried out by means of hydrolysis of the carbonitrile group,
using 2 M KOH in ethanol, and good yields were obtained.
Under these conditions, the methyl acetate group of 13 is
also hydrolyzed, and the compound obtained has a carboxylic
acid group instead of the methyl ester substituent in para-
position of the benzene ring.
First approximation looking for antimalarial activity, 1-16, and
R6/R7 = H/H, H/Cl,
H/CH3,H/OCH3,
F/F, Cl/Cl,CH3/CH3
First
R2 = CN
First
R7
R6
New
R6/R7 = H/F, H/CF3
O
N
New
+
R2 = COOH, COOCH2CH3
R2
+
N
O
First
W
W=H,Cl,CH3,COOCH3
New
W=F,OCH3,OCF3
The formation of isomeric quinoxaline 1,4-di-N-oxide was
observed in the case of monosubstituted benzofuroxanes.
According to previous reports [6], we have observed that
7-substituted quinoxaline 1,4-di-N-oxide derivatives were
New 3-phenylquinoxaline 1,4-di-N-oxide derivatives, 17-67
Scheme 1. Design of new antimalarial agents from our previous results [24].

E. Vicente et al. / European Journal of Medicinal Chemistry 43 (2008) 1903e1910
1905
O
O
+
CN
N
R7
R6
+
R7
R6
N
i
O
+
N
N
O
W
Series I
1-48
BF(a-i)
iii
ii
a
b
c
R6=R7=H
R6=H, R7=Cl
R6=H, R7=CH3
O
O
O
O
+
+
N
d
R6=H, R7=OCH3
R6=R7=Cl
R6=R7=CH3
R6=R7=F
R6=H, R7=F
R6=H, R7=CF3
N
R7
R6
OH
O
e
f
+
+
N
N
O
g
h
i
O
W
W
Series II
49-54
Series III
55-67
Scheme 2. Synthesis of quinoxaline-2-carbonitrile, ethyl carboxylate and carboxylic acid di-N-oxides. Conditions: (i) arylacetonitrile, CHCl2, Et3N, 24 h; (ii) 2 M
KOH in ethanol, 30 min; and (iii) ethyl 4-W-benzoylacetate, K2CO3, acetone, 2 h.
prevailing over the 6-isomer, or in the case of the methoxy
substituent, only the 7-isomer was formed. In practice, the
workup and purification permitted the isolation of the 7-
isomer [29].
All of the compounds were chemically characterized by
thin layer chromatography (TLC), infrared (IR) and nuclear
magnetic resonance (1H NMR) spectra as well as by elemental
microanalysis.
Due to the fact that the replacement of the carbonitrile
group by either an ester or a carboxylic acid unit failed to im-
prove the antiplasmodial activity, confirming the pharmaco-
phoric character of the carbonitrile moiety, novel carbonitrile
derivatives were designed and obtained for the purpose of in-
vestigating the contribution of an aromatic substitution pattern
in the optimization of lead-compound 5. Thus, compounds
having structural modifications at positions 6 (R6) and 7
(R7) of quinoxaline scaffold or at para-position (W) of phenyl
group, linked to C3, were synthesized and evaluated. In this
case, 20 of the 48 quinoxaline-2-carbonitrile derivatives as-
sayed against 3D7 strain improved the previously established
cut-off value (IC50 < 2 mM).
3. Pharmacology
3.1. In vitro anti-P. falciparum activity and cytotoxicity
in mammalian cells
Among the 60 assayed compounds against 3D7, 21 of
them showed enough activity to qualify for further studies.
These derivatives were assayed against a chloroquine-resis-
tant strain, K1, almost all of them showing IC50 values be-
low 2 mM. With these values, their resistance indexes (RI)
were calculated (RI ¼ IC50[K1]/IC50[3D7]) and are shown
in Table 2.
The previous lead-compound, 5, showed an IC50 of
1.01 mM for the 3D7 strain and 0.41 mM for the K1 strain.
In the sensitive strain, 10 of the 60 tested compounds and,
in the resistant strain, six of the 21 assayed compounds in
this work improved the activity of the lead-compound, with
derivative 42 being the one that showed the smallest value
of IC50 for both strains (0.33 and 0.18 mM, respectively).
In addition, the cytotoxicity of the 21 most active com-
pounds was tested in KB cells. Only seven of them demon-
strated certain cytotoxicity, thus two thirds showed
interesting results. With these values, their selectivity indexes
(SI) or cytotoxicity:antiplasmodial ratios were calculated with
respect to both strains (SI ¼ IC50[KB]/IC50[P.f.]). These data
are shown in Table 2. Compounds showing SI values above
10 were selected for further in vivo evaluation. According to
the established cut-off, this work identified seven derivatives
which were selective enough for both strains, and six more
The modified quinoxaline 1,4-di-N-oxide derivatives, 17e
63, and the previous analogues, 1e16, were evaluated against
3D7 strain (P. falciparum chloroquine-sensitive) and the
results are listed in Table 1. Compounds were selected for fur-
ther cytotoxicity evaluation and tested against P. falciparum
K1 (chloroquine-resistant strain) if their IC50 values were be-
low 2 mM against 3D7 strain. The second cut-off to make these
candidates valid new leads against malaria is an SI value
above 10.
4. Results
In an attempt to study the pharmacophoric profile of the
carbonitrile group in the antimalarial activity of lead-com-
pound 5, novel series of compounds were designed and syn-
thesized, and the corresponding antimalarial activity was
investigated against a CQ-sensitive P. falciparum strain
(3D7). These series were proposed with substituting the car-
bonitrile moiety with carboxylic acid subunit in series II
(49e53) and with an ester group in series III (54e63). The re-
sults are shown in Table 1. Only one of these compounds, 59,
showed an IC50 value below 2 mM, the cut-off established for
this work.

1906
E. Vicente et al. / European Journal of Medicinal Chemistry 43 (2008) 1903e1910
Table 1 (continued)
Table 1
Structure and antiplasmodial activity of 3-phenylquinoxaline 1,4-di-N-oxide
derivatives against P. falciparum 3D7
ID
R2
W
R6
H
R7
H
IC50 ꢁ SD (mM)
ND
53
COOH
COOH
O
54
55
56
57
58
59
COOC2H5
COOC2H5
COOC2H5
COOC2H5
COOC2H5
COOC2H5
H
H
H
H
H
H
H
Cl
3.22 ꢁ 0.34
>10
>10
R2
R7
R6
N+
H
CH3
OCH3
F
H
N+
O
F
>10
>7
Cl
CH3
Cl
CH3
W
1.18 ꢁ 0.05
ID
R2
W
R6
R7
IC50 ꢁ SD (mM)
60
61
62
63
COOC2H5
COOC2H5
COOC2H5
COOC2H5
H
Cl
H
H
H
H
H
H
H
H
>10
>10
>10
>7
1
2
3
4
CN
CN
CN
CN
H
H
H
H
H
H
H
H
H
Cl
2.35 ꢁ 0.04
CH3
CF3
>5
1.55 ꢁ 0.76
1.47 ꢁ 0.24
CH3
OCH3
CQ
0.0135
5
6
7
8
CN
CN
CN
CN
Cl
Cl
Cl
Cl
H
H
H
H
H
Cl
1.01 ꢁ 0.37
Values shown are the mean IC50 ꢁ SD (mM) from 2 to 3 independent experi-
ments, each one performed in triplicate. CQ ¼ chloroquine.
>5
0.48 ꢁ 0.13
0.73 ꢁ 0.03
CH3
OCH3
9
CN
CN
CN
CN
CH3
CH3
CH3
CH3
H
H
H
H
H
Cl
3.46 ꢁ 1.08
which were selective for only the resistant strain, with SI
values ranging from 10.31 to 50.87.
10
11
12
>10
1.30 ꢁ 0.55
0.81 ꢁ 0.13
CH3
OCH3
13
14
15
16
CN
CN
CN
CN
COOCH3
COOCH3
COOCH3
COOCH3
H
H
H
H
H
Cl
6.54 ꢁ 1.03
5. Discussion
>20
2.15 ꢁ 0.80
3.16 ꢁ 0.45
CH3
OCH3
As shown in Table 1, the replacement of carbonitrile moi-
ety by either a carboxylic acid group or an ethyl-ester group
resulted in an important loss of activity, with the exception of
compound 59. These results appear to indicate that the
17
18
19
20
21
22
CN
CN
CN
CN
CN
CN
F
F
F
F
F
F
H
H
Cl
3.91 ꢁ 0.92
5.83 ꢁ 0.60
0.88 ꢁ 0.13
0.87 ꢁ 0.19
>10
H
H
CH3
OCH3
Cl
H
Cl
CH3
CH3
1.26 ꢁ 0.16
Table 2
Antiplasmodial activity, cytotoxicity and selectivity of selected compounds
23
24
25
26
27
28
CN
CN
CN
CN
CN
CN
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
H
H
Cl
3.14 ꢁ 0.44
H
>10
ID
IC50 ꢁ SD (mM)
P.f. K1
SI
RI
H
CH3
OCH3
Cl
2.77 ꢁ 0.65
2.29 ꢁ 0.12
>20
H
KB cells
11.65 ꢁ 0.97
KB/3D7
KB/K1
3D7/K1
Cl
CH3
3
0.74 ꢁ 0.04
0.72 ꢁ 0.39
0.41 ꢁ 0.08
0.19 ꢁ 0.03
0.47 ꢁ 0.30
0.95 ꢁ 0.06
2.29 ꢁ 0.53
0.37 ꢁ 0.05
0.36 ꢁ 0.06
0.93 ꢁ 0.02
0.27 ꢁ 0.09
0.93 ꢁ 0.09
0.38 ꢁ 0.08
1.25 ꢁ 0.26
>8
7.52
7.12
15.76
14.47
23.57
49.46
37.53
26.19
7.39
0.48
0.49
0.41
0.39
0.64
0.73
2.83
0.42
0.42
0.73
0.21
1.31
0.51
1.35
CH3
3.02 ꢁ 0.31
4
10.46 ꢁ 1.16
9.73 ꢁ 1.63
9.25 ꢁ 0.68
17.51 ꢁ 0.04
24.98 ꢁ 1.36
16.95 ꢁ 1.31
11.19 ꢁ 0.74
18.54 ꢁ 1.14
18.84 ꢁ 1.56
2.76 ꢁ 0.16
2.35 ꢁ 0.26
8.12 ꢁ 1.64
8.25 ꢁ 0.62
2.61 ꢁ 0.40
2.82 ꢁ 0.18
2.85 ꢁ 0.50
2.76 ꢁ 0.19
3.17 ꢁ 0.56
13.45 ꢁ 0.49
5
9.63
29
30
31
32
33
34
CN
CN
CN
CN
CN
CN
OCF3
OCF3
OCF3
OCF3
OCF3
OCF3
H
H
Cl
1.27 ꢁ 0.17
0.71 ꢁ 0.34
0.75 ꢁ 0.28
0.93 ꢁ 0.48
>10
7
19.27
23.99
19.22
20.93
12.72
21.31
14.95
2.17
H
8
H
CH3
OCH3
Cl
11
12
19
20
22
29
30
31
32
40
42
43
44
45
47
59
CQ
POD
H
Cl
CH3
30.44
50.87
20.36
10.31
2.53
CH3
>5
35
36
37
38
39
40
41
CN
CN
CN
CN
CN
CN
CN
H
H
H
H
H
H
H
H
F
F
F
F
F
F
F
>10
F
Cl
5.78 ꢁ 0.20
4.21 ꢁ 0.57
>10
3.31
CH3
OCH3
OCF3
COOCH3
10.83
8.87
1.54
21.41
6.58
>20
1.70 ꢁ 0.19
>20
0.18 ꢁ 0.04
0.52 ꢁ 0.24
1.77 ꢁ 0.08
0.87 ꢁ 0.09
0.55 ꢁ 0.01
>10
8.55
4.32
15.67
5.48
1.56
0.55
0.79
1.13
1.21
0.28
42
43
44
45
46
47
48
CN
CN
CN
CN
CN
CN
CN
H
H
H
H
H
H
H
H
CF3
CF3
CF3
CF3
CF3
CF3
CF3
0.33 ꢁ 0.12
0.66 ꢁ 0.03
1.56 ꢁ 0.19
0.72 ꢁ 0.09
2.44 ꢁ 0.30
1.93 ꢁ 0.31
2.03 ꢁ 0.62
1.77
4.40
F
Cl
3.64
24.45
6.97
CH3
OCH3
OCF3
COOCH3
>60
110
0.003
>50
8150
0.682
161
50.52
Values shown are the mean IC50 ꢁ SD (mM) from 2 to 3 independent experi-
ments, each one performed in triplicate.
49
50
51
52
COOH
COOH
COOH
COOH
H
Cl
H
H
H
H
H
H
H
H
ND
CQ ¼ chloroquine;
>10
>10
ND
POD ¼ podophylotoxin; P.f. ¼ Plasmodium falciparum; SI ¼ selectivity index;
RI ¼ resistance index.
CH3
OCH3

E. Vicente et al. / European Journal of Medicinal Chemistry 43 (2008) 1903e1910
1907
hybridization of the carbon atom linked to position 2 of the
quinoxaline nucleus is an important structural requirement
for the antiplasmodial activity. These data also demonstrated
that the antiplasmodial activity cannot be improved by the
use of an acidic substituent at C-2 because when evaluated
under the same experimental conditions as those used for
5, none of the compounds of series II (49e53) was found
to be active (Table 1). However, the introduction of a di-
methyl substituent in the quinoxaline ring, resulting in com-
pound 59, appears to compensate the poor activity produced
by the ester moiety, thereby maintaining the activity of com-
pound 59 (IC50 ¼ 1.18 mM) similar to that of the lead-com-
pound (IC50 ¼ 1.01 mM) for sensitive strain. In spite of the
fact that this compound turned out to be the least cytotoxic
of all those which were assayed (IC50 > 60 mM for KB cells),
its antiplasmodial activity in resistant strain (IC50 > 10 mM
for K1 strain) was not sufficient enough to merit a more ad-
vanced study.
On the other hand, the results obtained clearly demon-
strated that it was possible to optimize the antiplasmodial
activity of the prototype 5 keeping the carbonitrile group
linked to C-2 of quinoxaline scaffold, particularly when
a monosubstitution was carried out at quinoxaline scaffold
by a methyl group, followed by the introduction of a chlorine
atom in the para-position of the phenyl ring, linked at C-3
of quinoxaline nucleus, as observed for compound 7
(Table 1).
In short, the results illustrated in Table 1 indicate that the
monosubstitution at 1,4-di-N-oxide-quinoxaline scaffold by
an electron donating group, such as methyl (3, 7, 11, 15, 19,
25, 31) or methoxy (4, 8, 12, 16, 20, 26, 32), appeared to
enhance the antimalarial activity when compared with unsub-
stituted compound (1, 5, 9, 13, 17, 23, 29). In contrast, the
monosubstitution at 1,4-di-N-oxide-quinoxaline scaffold by
a halogen atom, such as chlorine (2, 6, 10, 14, 18, 24, 30)
or fluorine (35e41), resulted in decreased activity. Curiously,
the disubstitution at R6 and R7 of the quinoxaline ring, gener-
ally resulted in a significant loss of activity (Table 1). Despite
the electron withdrawing nature of the trifluoromethyl group,
the derivatives containing this substituent (42e48) showed the
most interesting values of IC50 against P. falciparum 3D7;
this fact might be due to its larger volume which could
contribute even more to hydrophobicity than in the case of
fluorine [31].
With regard to the substitution carried out at position para
of the phenyl ring (W), linked to C-3 of the quinoxaline
scaffold, the most active compounds were the ones which pre-
sented a chlorine atom at para (7 and 8). The substitution by
an OCF3 group (30, 31, 32) also showed promising activity
(parameters sp and p being very similar). The worst results
came from the derivatives with W ¼ OCH3 (23e28, 39, 46)
and COOCH3 (13e16, 41, 48); these groups differ from
each other considerably with regard to sp but their p is very
close to zero in both cases.
chloroquine shows an IC50 50 times higher in the K1 strain
than in the 3D7 strain, we obtained equal or better activity
in the resistant strain that in the sensitive strain for almost
all of our most active compounds. As observed in Table 2, al-
most all of the compounds tested against K1 showed a ten-
dency towards lower IC50 values against K1 strain than
against 3D7 strain. A comparison of the IC50 values of the
resistant and sensitive strains of P. falciparum suggests rela-
tively low levels of cross-resistance to CQ [37]. The resistance
index (RI) values, calculated as IC50(K1)/IC50(3D7), were
lower than that found for CQ (Table 2). Ten of the derivatives
that were studied, bearing an electron withdrawing group on
the phenyl substituent (W) such as 5, 7, 8, 19, 20, 29 and
31, or a CF3 group on the quinoxaline ring (R6/R7) such as
42, 43 and 47, showed superior activity for K1 strain in
comparison with chloroquine (IC50(Qx)/IC50(CQ) ꢂ 1). In
this way, the 2-carbonitrile 1,4-di-N-oxide quinoxaline
derivatives seem to be a novel promise antimalarial lead-
candidates.
With regard to the cytotoxicity of the other compounds, the
most cytotoxic derivatives are those which include an OCF3
group in position W of the phenyl substituent (29e32, 40)
or a CF3 group in position R7 of the quinoxaline ring (42e
45), with IC50 values ranging from 2.35 to 8.25 mM. Within
these derivatives is the compound that shows the lowest IC50
values for both 3D7 and K1 strains of P. falciparum (42).
Curiously, the compound that links both aforementioned
groups (47) presents an SI value of 13.45, very superior to
that of its analogues.
By means of the SI value, which relates the cytotoxicity
with the antiplasmodial activity of the derivatives studied,
we discovered the most interesting structures. Among the
seven compounds whose SI is greater than 10 for both strains
(7, 8, 11, 19e22), it should be pointed out that four of them
presented an IC50 in the K1 strain inferior to that of CQ and
in addition, their SI for the same strain was greater than 30.
These four compounds present an electron releasing group
(methyl, 7 and 19, or methoxy, 8 and 20) in position R7 of
the quinoxaline ring and a halogen atom (chlorine, 7 and 8,
or fluorine, 19 and 20) in position W of the phenyl substituent.
These compounds are now being considered for further evalu-
ation in vivo against CQ-sensitive and CQ-resistant Plasmo-
dium berghei lines.
Although the mechanism of action of our compounds has
not yet been investigated, previous results with some quinoxa-
line derivatives with antimalarial activity correlated with
a high inhibitory of b-hematin formation, indicating that this
class of compounds may interfere with the heme polymeriza-
tion process within the food vacuole [24,37].
6. Conclusions
Screening of the in vitro antiplasmodial activity and
cytotoxicity of the novel series, quinoxaline-2-carbonitrile
1,4-di-N-oxide derivatives, allowed us to identify interesting
antimalarial candidates based on their potency, selectivity
and low cytotoxicity, making them valid new leads for
The most interesting results were obtained from evaluating
the selected compounds against K1 strain (CQ-resistant) and
mammalian cells (KB), as illustrated in Table 2. While

1908
E. Vicente et al. / European Journal of Medicinal Chemistry 43 (2008) 1903e1910
synthesizing new compounds that might improve the previous
prototypes. Compounds 7-methyl or 7-methoxy 3-phenylqui-
noxaline-2-carbonitrile 1,4-di-N-oxide with a chlorine (7 and
8) or a fluorine atom (19 and 20) in para-position (W) of
the phenyl substituent were shown to be more active than
CQ against K1 strain; in addition, their selectivity indexes
for the same strain were greater than 30. These SI values
that were obtained suggest parasite-specific effects, perhaps
by inhibition of the heme polymerization process, as discussed
above. The aforementioned compounds are now being consid-
ered for further evaluation in vivo against CQ-sensitive and
CQ-resistant P. berghei lines.
7.3. General procedure for preparation of ethyl 3-
phenylquinoxaline-2-carboxylate 1,4-di-N-oxide
derivatives (55e67) [36]
The corresponding functionalized ethyl benzoylacetate
(7.34 mmol) and de potassium carbonate (9.55 mmol) were
added to a solution of 7.34 mmol of benzofurazan oxide
(BF) in 50 mL of acetone. The suspension was stirred at
room temperature for 2 h. The quinoxaline 1,4-di-N-oxide de-
rivatives were isolated by addition of 50 mL of water followed
by extraction with dichloromethane (5 ꢄ 40 mL). The organic
layer was dried (Na2SO4), filtered and evaporated to dryness.
The residue was purified by recrystallization from a mixture
of methanol/ether/n-hexane (2:4:4). Yields: 15e64%.
7. Experimental protocols
7.1. General
7.4. General procedure for preparation of 3-
phenylquinoxaline-2-carboxylic acid 1,4-di-N-oxide
derivatives (49e54)
The 1H NMR spectra were recorded on a Bruker 400 Ultra-
shield instrument (400 MHz), using TMS as the internal stan-
dard and with DMSO-d6 and CDCl3 as the solvents; the
chemical shifts are reported in ppm (d) and coupling constants
(J ) values are given in Hertz (Hz). Signal multiplicities are
represented by: s (singlet), d (doublet), t (triplet), q (quadru-
plet), dd (double doublet) and m (multiplet). The IR spectra
were performed on a Thermo Nicolet Nexus FTIR (Madison,
A mixture of carbonitrile derivative (2.6 mmol), ethanol
(15 mL) and a solution of 2 M KOH in ethanol (5 mL) was
stirred at room temperature for 30 min. The mixture was
diluted with 30 mL of water and washed with CH2Cl2
(3 ꢄ 20 mL). The aqueous layer was neutralized with 3 N
HCl and the solid which precipitated was collected by filtra-
tion in order to give the desired carboxylic acid in moderate
to good yields (45e70%).
USA) in KBr pellets; the frequencies are expressed in cmꢀ1
.
Elemental microanalyses were obtained on an Elemental
Analyzer (Leco CHN-900, Tres Cantos, Madrid, Spain) from
vacuum-dried samples. The analytical results for C, H, and
N, were within ꢁ 0.4 of the theoretical values.
7.5. P. falciparum in vitro culture and parasite growth
inhibition assays
Alugram SIL G/UV254 (layer: 0.2 mm) (Macherey-Nagel
GmbH & Co. KG. Postfach 101352. D-52313 Duren, Ger-
¨
All parasite clones, isolates and strains were acquired from
MR4 (Malaria Research and Reference Reagent Resource
Center, Manassas, Virginia, USA). Strains/isolates used in
this study were: the drug sensitive 3D7 clone of the NF54 iso-
late (unknown origin); and the chloroquine, pyrimethamine
and cycloguanyl resistant K1 strain (Thailand). In vitro culture
of P. falciparum was carried out following standard methods
[38] with modifications as described [39]. In vitro parasite
growth inhibition was assessed by the incorporation of [3H]-
hypoxanthine based on the method used by Desjardins et al.
[40] and modified as described [41]. Briefly, stock drug solu-
tions were dissolved in 100% dimethylsulfoxide (Sigma, Dor-
set, UK) and 50 mL of a 3-fold dilution series (10.0, 3.33, 1.11,
0.370, 0.124, and 0.0412 mg/mL) of the drugs prepared in
assay medium (RPMI 1640 supplemented with 0.5% Albumax
II (Invitrogen), 0.2% w/v glucose, 0.03% L-glutamine, and
5 mM hypoxanthine) added to each well of 96-well plates in
triplicate. Fifty microlitres of asynchronous (65e75% ring
stage) P. falciparum culture (0.5% parasitemia) or uninfected
erythrocytes (blank) were added to each well reaching a final
volume of 100 mL per well, a final hematocrit of 2.5% and
final dimethylsulfoxide concentrations ꢂ0.01%. Plates were
incubated at 37 ꢃC in 5% CO2, and 95% air mixture for
24 h, at which point 10 mL (0.2 mCi/well) of [3H]-hypoxan-
thine (PerkineElmer, Hounslow, UK), was added to each
well. After an additional 24 h incubation period, the
many) was used for thin layer chromatography and silica gel
60 (0.040e0.063 mm) for Column flash Chromatography
(Merck).
Chemicals were purchased from E. Merck (Darmstadt, Ger-
´
many), Panreac Quımica S.A. (Montcada i Reixac, Barcelona,
Spain), SigmaeAldrich Quımica, S.A. (Alcobendas, Madrid),
´
Acros Organics (Janssen Pharmaceuticalaan 3a, 2440 Geel,
Belgie) and Lancaster (Bischheim-Strasbourg, France).
¨
Benzofurazan oxide (BF) was prepared as reported [27].
7.2. General procedure for preparation of
3-arylquinoxaline-2-carbonitrile 1,4-di-N-oxide
derivatives (1e48) [24]
The corresponding arylacetonitrile (10.6 mmol) was added
to a solution of the appropriate benzofuroxane (2.4 mmol) in
dry dichloromethane (35 mL). The mixture was allowed to
stand at 0 ꢃC. Triethylamine was added dropwise (1 mL),
and the reaction mixture was stirred at room temperature in
darkness for 24 h. After evaporating to dryness under reduced
pressure, a yellow crude solid or a brown oil was obtained. It
was then precipitated and washed by adding diethyl ether (or
n-hexane), affording the target compound. The obtained
yellow precipitate was purified by recrystallization from meth-
anol or a mixture of methanol/ethyl acetate. Yields: 8e74%.

E. Vicente et al. / European Journal of Medicinal Chemistry 43 (2008) 1903e1910
[4] R.G. Ridley, Nature 424 (2003) 887e889.
1909
experiment was terminated by placing the plates in a ꢀ80 ꢃC
freezer. Plates were thawed and harvested onto glass fibre filter
mats using a 96-well cell harvester (Harvester 96, Tomtec,
Oxon, UK) and left to dry. After the addition of MeltiLex solid
scintillant (PerkineElmer, Hounslow, UK) the incorporated
radioactivity was counted using a Wallac 1450 BetaLux scin-
tillation counter (Wallac).
[5] D.J. Brown (Ed.), Quinoxalines, Supplement II, Wiley & Sons, New
York, 2004.
[6] G.W.H. Cheeseman, R.F. Cookson (Eds.), Condensed Pyrazines, Wiley
& Sons, New York, 1979.
[7] A. Carta, P. Corona, M. Loriga, Curr. Med. Chem. 12 (2005) 2259e2272.
[8] M.E. Montoya, Y. Sainz, M.A. Ortega, A.L. De Cerain, A. Monge,
Farmaco 53 (1998) 570e573.
[9] Y. Sainz, M.E. Montoya, F.J. Martinez-Crespo, M.A. Ortega, A.L. de Ce-
rain, A. Monge, Arzneim.-Forsch. 49 (1999) 55e59.
Data acquired by the Wallac BetaLux scintillation counter
were exported into a MICROSOFT EXCEL spreadsheet
(Microsoft), and the IC50 values of each drug were calculated
by using XLFit line fitting software (ID Business Solutions,
UK). Chloroquine diphosphate, as a standard drug, and control
wells with untreated infected and uninfected erythrocytes were
included in all assays.
[10] M.A. Ortega, Y. Sainz, M.E. Montoya, A.L. De Cerain, A. Monge,
Pharmazie 54 (1999) 24e25.
[11] M.A. Ortega, M.E. Montoya, A. Jaso, B. Zarranz, I. Tirapu, I. Aldana,
A. Monge, Pharmazie 56 (2001) 205e207.
[12] M.A. Ortega, Y. Sainz, M.E. Montoya, A. Jaso, B. Zarranz, I. Aldana,
A. Monge, Arzneim.-Forsch. 52 (2002) 113e119.
[13] A. Carta, G. Paglietti, M.E. Rahbar Nikookar, P. Sanna, L. Sechi,
S. Zanetti, Eur. J. Med. Chem. 37 (2002) 355e366.
[14] A. Jaso, B. Zarranz, I. Aldana, A. Monge, Eur. J. Med. Chem. 38 (2003)
791e800.
7.6. In vitro cytotoxicity assay
[15] B. Zarranz, A. Jaso, I. Aldana, A. Monge, Bioorg. Med. Chem. 11 (2003)
2149e2156.
The AlamarBlue (AccuMed International, USA) method
was used to assess cytotoxicity to KB cells as previously de-
scribed [39]. BrieXy, microtiter plates were seeded at a density
of 4 ꢄ 104 KB cells/mL in RPMI 1640 culture medium supple-
mented with 10% heat-inactivated foetal calf serum (complete
medium) (Seralab). Plates were incubated at 37 ꢃC, 5% CO2,
95% air mixture for 24 h followed by compound addition to
triplicate wells in a dilution series in complete medium. The
positive control drug was podophyllotoxin (Sigma;
IC50 ¼ 0.003 mM). Plates were incubated for a further 72 h fol-
lowed by the addition of 10 mL of AlamarBlue (AccuMed In-
ternational) to each well and incubation for 2e4 h at 37 ꢃC,
5% CO2, 95% air mixture. Fluorescence emission at 585 nm
was measured in a SPECTRAMAX GEMINI plate reader
(Molecular Devices) after excitation at 530 nm. IC50 values
were calculated using XLFit (ID Business Solutions, UK)
line fitting software.
[16] A. Carta, M. Loriga, G. Paglietti, A. Mattana, P.L. Fiori, P. Mollicotti,
L. Sechi, S. Zanetti, Eur. J. Med. Chem. 39 (2004) 195e203.
[17] A. Jaso, B. Zarranz, I. Aldana, A. Monge, J. Med. Chem. 48 (2005)
2019e2025.
[18] S. Zanetti, L.A. Sechi, P. Molicotti, S. Cannas, A. Bua, A. Deriu,
A. Carta, G. Paglietti, Int. J. Antimicrob. Agents 25 (2005) 179e181.
[19] G. Aguirre, H. Cerecetto, R. Di Maio, M. Gonzalez, W. Porcal,
G. Seoane, M.A. Ortega, I. Aldana, A. Monge, A. Denicola, Arch.
Pharm. (Weinheim, Ger.) 335 (2002) 15e21.
[20] G. Aguirre, H. Cerecetto, R. Di Maio, M. Gonzalez, M.E.M. Alfaro,
A. Jaso, B. Zarranz, M.A. Ortega, I. Aldana, A. Monge-Vega, Bioorg.
Med. Chem. Lett. 14 (2004) 3835e3839.
[21] C. Urquiola, M. Vieites, G. Aguirre, A. Marin, B. Solano, G. Arrambide,
P. Noblia, M.L. Lavaggi, M.H. Torre, M. Gonzalez, A. Monge,
D. Gambino, H. Cerecetto, Bioorg. Med. Chem. 14 (2006) 5503e5509.
[22] I. Aldana, M.A. Ortega, A. Jaso, B. Zarranz, P. Oporto, A. Gimenez,
A. Monge, E. Deharo, Pharmazie 58 (2003) 68e69.
[23] A. Martin, L.M. Lima, B. Solano, E. Vicente, S.P. Silanes, S. Maurel,
M. Sauvain, I. Aldana, A. Monge, E. Deharo, Exp. Parasitol. 118
(2008) 25e31.
[24] B. Zarranz, A. Jaso, I. Aldana, A. Monge, S. Maurel, E. Deharo,
V. Jullian, M. Sauvain, Arzneim.-Forsch. 55 (2005) 754e761.
[25] B. Zarranz, A. Jaso, L.M. Lima, I. Aldana, A. Monge, S. Maurel,
M. Sauvain, Braz. J. Pharm. Sci. 42 (2006) 357e361.
Acknowledgements
´
We thank the Ministerio de Educacion y Ciencia (Grant
[26] B. Zarranz, A. Jaso, I. Aldana, A. Monge, Bioorg. Med. Chem. 12 (2004)
3711e3721.
AP2003-2175 to Esther Vicente) and the University of Navarra
(PiUNA project).
[27] A. Monge, J.A. Palop, A.L. De Cerain, V. Senador, F.J. Martinez-Crespo,
Y. Sainz, S. Narro, E. Garcia, C. De Miguel, M. Gonzalez, E. Hamilton,
A.J. Barker, E.D. Clarke, D.T. Greenhow, J. Med. Chem. 38 (1995)
1786e1792.
Appendix. Supplementary data
[28] A. Monge, J.A. Palop, M. Gonzalez, F.J. Martinez-Crespo,
A.L. Decerain, Y. Sainz, S. Narro, A.J. Barker, E. Hamilton, J. Hetero-
cycl. Chem. 32 (1995) 1213e1217.
Characterization data (IR and 1H NMR) and elemental
analyses of all new compounds are available via the Internet
at http://www.sciencedirect.com. Supplementary data associ-
ated with this article can be found in the online version at
doi:10.1016/j.ejmech.2007.11.024.
[29] M.A. Ortega, M.J. Morancho, F.J. Martinez-Crespo, Y. Sainz, M.E. Montoya,
A.L. de Cerain, A. Monge, Eur. J. Med. Chem. 35 (2000) 21e30.
[30] C. Hansch, A. Leo (Eds.), Substituent Constants for Correlation Analysis
in Chemistry and Biology, Wiley & Sons, New York, 1979.
[31] D. Gimenez, C. Andreu, M.L. del Olmo, T. Varea, D. Diaz, G. Asensio,
Bioorg. Med. Chem. 14 (2006) 6971e6978.
[32] J.J. Cheng, R. Zeidan, S. Mishra, A. Liu, S.H. Pun, R.P. Kulkarni,
G.S. Jensen, N.C. Bellocq, M.E. Davis, J. Med. Chem. 49 (2006)
6522e6531.
[33] Y. Takano, F. Shiga, J. Asano, W. Hori, T. Anraku, T. Uno, Bioorg. Med.
Chem. Lett. 14 (2004) 5107e5111.
References
[1] D.J. Krogstad, Epidemiol. Rev. 18 (1996) 77e89.
[2] B.M. Greenwood, K. Bojang, C.J.M. Whitty, G.A.T. Targett, Lancet 365
(2005) 1487e1498.
[34] A. Mital, Curr. Med. Chem. 14 (2007) 759e773.
[35] M.J. Haddadin, C.H. Issidorides, Heterocycles 35 (1993) 1503e1525.
[3] S.J. Burgess, A. Selzer, J.X. Kelly, M.J. Smilkstein, M.K. Riscoe,
D.H. Peyton, J. Med. Chem. 49 (2006) 5623e5625.

1910
E. Vicente et al. / European Journal of Medicinal Chemistry 43 (2008) 1903e1910
[36] L.M. Lima, B. Zarranz, A. Marin, B. Solano, E. Vicente, S.P. Silanes,
I. Aldana, A. Monge, J. Heterocycl. Chem. 42 (2005) 1381e1385.
[37] J. Guillon, P. Grellier, M. Labaied, P. Sonnet, J.M. Leger, R. Deprez-Pou-
lain, I. Forfar-Bares, P. Dallemagne, N. Lemaitre, F. Pehourcq, J. Rochette,
C. Sergheraert, C. Jarry, J. Med. Chem. 47 (2004) 1997e2009.
J.J. Martin, F. Risco, S. Garcia-Ochoa, F.J. Gamo, L. Sanz, L. Leon,
J.R. Ruiz, R. Gabarro, A. Mallo, F.G. de las Heras, J. Biol. Chem. 279
(2004) 31429e31439.
[40] R.E. Desjardins, C.J. Canfield, J.D. Haynes, J.D. Chulay, Antimicrob.
Agents Chemother. 16 (1979) 710e718.
[38] W. Trager, J.B. Jensen, Science 193 (1976) 673e675.
[39] A. Cameron, J. Read, R. Tranter, V.J. Winter, R.B. Sessions, R.L. Brady,
L. Vivas, A. Easton, H. Kendrick, S.L. Croft, D. Barros, J.L. Lavandera,
[41] L. Vivas, A. Easton, H. Kendrick, A. Cameron, J.L. Lavandera,
D. Barros, F.G. de las Heras, R.L. Brady, S.L. Croft, Exp. Parasitol.
111 (2005) 105e114.