Identification of benzoylisoquinolines as potential anti-Chagas agents

Article abstract of DOI:10.1016/j.bmc.2012.02.046

A set of three 3-benzoyl substituted isoquinolones was synthesized in good yields and assayed for in vitro trypanocidal activity against Trypanosoma cruzi, the protozoan parasite that causes Chagas' disease. Depending on the concentration evaluated, a gre

Full text of DOI:10.1016/j.bmc.2012.02.046

Bioorganic & Medicinal Chemistry 20 (2012) 2587–2594
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Identification of benzoylisoquinolines as potential anti-Chagas agents
Kendall G. Byler a, Marco Brito-Arias b, , Adrian Marquez-Navarro c, Benjamin Nogueda-Torres c,
⇑
Luis G. Torres-Bustillos b, Karina Martínez-Mayorga a,
⇑
a Torrey Pines Institute for Molecular Studies, 11350 SW Village Parkway, Port St. Lucie, FL 34987, USA
b Unidad Profesional Interdisciplinaria de Biotecnología, Instituto Politécnico Nacional, Av. Acueducto s/n Colonia La Laguna Ticoman, Mexico, 07340 DF, Mexico
c Escuela Nacional de Ciencias Biológicas, Carpio y Plan de Ayala Colonia Santo Tomás, Mexico, DF 11340, Mexico
a r t i c l e i n f o
a b s t r a c t
Article history:
A set of three 3-benzoyl substituted isoquinolones was synthesized in good yields and assayed for in vitro
trypanocidal activity against Trypanosoma cruzi, the protozoan parasite that causes Chagas’ disease.
Depending on the concentration evaluated, a greater or equivalent reduction in the number of blood-
borne trypomastigotes compared to that observed with benznidazole, a drug currently used to attack
the parasite, was observed for two of the samples. In order to assess the potential of the 3-benzoylisoqu-
inolone nucleus as a possible scaffold in the design of novel anti-trypanosomal lead structures, a compu-
tational analysis was performed using structural and inhibition information from both functional and
target assays archived in the online database, ChEMBL. Chemical space projection of the synthesized
compounds along with 3067 structures with known activities against T. cruzi shows that the isoquinol-
ones occupy a sparsely-populated region of chemical space, indicating their potential for development
as a novel class of trypanocidals. In addition, 2D and 3D structural similarity analyses revealed micromo-
lar and submicromolar inhibitors of T. cruzi in ChEMBL with high similarity to the synthesized structures.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 16 December 2011
Revised 14 February 2012
Accepted 20 February 2012
Available online 3 March 2012
Keywords:
Benzoyl isoquinolones
Trypanosoma cruzi
Chemical space
Structural similarity
1. Introduction
Although effective in the acute, short-term phase of the disease,
nifurtimox (Nfx) and benznidazole (Bnz) are not effective in the
Parasitic infections continue to be societal and health burdens,
affecting millions of people, mainly in developing countries. Proto-
zoal diseases such as leishmaniasis, malaria, sleeping sickness, and
Chagas’ disease assault populations in tropical and subtropical
regions of Latin America, Africa, and Asia, where these diseases
are endemic. The elevated cost of treatment of these diseases
contributes to high mortality rates associated with infections in
the world’s most economically-disadvantaged areas. The high
toxicity and cost associated with current treatments, along with
a lack of efficacy during the chronic phase of the disease and
emerging resistance to available therapy, highlight the need for
the development of new antiparasitic drugs. It is important to note
that remarkable initiatives,1,2 have emerged around the world.
American trypanosomiasis, or Chagas’ disease,3 is caused by the
hemoflagellate protozoan Trypanosoma cruzi. It is estimated that
10–20 million people in Latin America are currently infected, with
25–90 million at risk of acquiring the parasite.4,5 Available treat-
ment for American trypanosomiasis is presently limited to two
drugs: the nitro imidazole benznidazole and the nitrofuran nifurti-
mox (Fig. 1).
more problematic chronic stage. Moreover, these two drugs are
mutagenic and carcinogenic as well as significantly toxic, produc-
ing a variety of disturbing side effects. Both compounds are pro-
drugs that require conversion to the active component in the
parasite; the mechanism of Nfx activity involves a reductive
metabolite that produces toxic free radicals.6 In the search for
alternative approaches to combat this disease, new therapeutic
targets have been identified and validated.6 For Trypanosoma cruzi,
recently identified biological targets include, but is not limited to:
farnesyl pyrophosphate synthase (FPPS), trans-sialidase (TS), cruz-
ain, trypanothione reductase (TR), glucose-6-phosphate dehydro-
genase (GPDH), glyceraldehyde-3-phosphate dehydrogenase
(GAPDH), and a-hydroxyacid dehydrogenase (HDH) and a number
of compounds have been examined for their ability to inhibit these
enzymes. Among them, 2,3-dihydro-3-deoxy-N-acetylneuraminic
acid (DANA) has been crystallized with T. cruzi trans-sialidase7
and phenothiazine derivatives have been assayed for T. cruzi trypa-
nothione reductase activity.8
Alternatively, a convenient option currently employed in many
drug discovery efforts involves repurposing drugs marketed for the
treatment of other conditions. In the search for anti-Chagas agents;
risedronate,9 camptothecin,10 and allopurinol11 have been shown
to inhibit the progression of the disease (Fig. 1).
⇑
The efficacy of 2-substituted quinolines found in medicinal
Corresponding authors.
E-mail address: kmartinez@tpims.org (K. Martínez-Mayorga).
plants against Leishmania spp.12,13 and T. cruzi14 has motivated
0968-0896/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2012.02.046

2588
K. G. Byler et al. / Bioorg. Med. Chem. 20 (2012) 2587–2594
NH
O
O
O2N
N
O
N
N
S
N
O
NO2
PO3H2
OH
O
O
PO3H2
N
HN
O
N
N
N
H
N
N
O
OH
Figure 1. Anti-Chagas drugs benznidazole and nifurtimox and approved drugs with recently reported activity in T. cruzi (l to r): risedronate, camptothecin and allopurinol.
2. Results and discussion
O
2.1. Synthesis of benzoylisoquinolones
The synthesis of benzoylisoquinolone derivatives 4a–c was car-
N
ried out according to Scheme 1, starting from the condensation reac-
tion between potassium phthalimide 1 and 2-bromacetophenone 2
in DMF at room temperature to provide 3 in good yields. The next
step involved the reaction of compound 3 with sodium methoxide
solution under reflux to provide benzoyl isoquinolone 4a. The
formation of this heterocycle proceeded via the Gabriel-Colman
rearrangement and its characterization was achieved by X-ray
diffraction.23 Modification of the benzoyl isoquinolone nucleus at
position 3 was performed under standard conditions to provide
the nitro derivative 4b, followed by reduction under catalytic hydro-
genation to yield the amino derivative 4c in quantitative yields.
N
2-(E)-prop-1-enylquinoline
4-methoxy-2-n-propylquinoline
O
N
N
O
4-methoxy-2-(E)-prop-1-enylquinoline
2-(1,2-trans-epoxypropyl)-quinoline
Figure 2. Drugs with known activity against enzymatic targets in Trypanosoma
cruzi.
2.2. Trypanocidal activity of benzoylisoquinolones
Assays for trypanocidal activity of the isoquinolones were car-
ried out at the three concentrations described below, including
the drug benznidazole as a reference standard. The efficacy of com-
pound 4b was observed to increase at a high concentration
(161 lM), reaching a maximum trypanocidal activity of 47% after
24 h incubation. Compound 4c exhibited greater trypanocidal
the exploration of quinolines as privileged scaffolds against para-
sitic diseases and viral infections. Synthetic quinolines obtained
and assayed by Fakhfakh, et al. showed promising in vitro activity15
against leishmania/HIV co-infections. Although these compounds
were less active than the reference compound, suramin, they repre-
sent new lead molecules for further optimization (Fig. 2). Related
chemical structures, also explored as anti-leishmania agents, are
naphthylisoquinoline alkaloids. Ponte-Sucre, et al. showed that
naphthylisoquinoline alkaloids and synthetic analogs are effective
against intracellular amastigotes in the submicromolar range.16
Public domain repositories of bioactive compounds17 and asso-
ciated data provide a valuable means of collecting and sharing
information, performing data mining, and avoiding duplicated
efforts—ultimately aiding in the discovery of new medications.
Examples of such sources are the ChEMBL database,18 PubChem
Central19,20 and BindingDB.21 Retrieval of chemical structures
along with antiparasitic activity data reported in the literature
can be obtained from the ChEMBL website (https://www.ebi.ac.
uk/chembl) and is freely available for analysis.
In this work, we present the analysis of the structural and phys-
icochemical properties of compounds tested against T. cruzi in
enzymatic and functional assays reported in the ChEMBL database.
3-benzoylisoquinolones were synthesized and evaluated for their
ability to reduce the number of live parasites in infected blood.
The reaction scheme involves the Gabriel-Colman rearrangement,22
and subsequent electrophilic aromatic substitution at position 3 of
the benzoyl ring. The structural and physicochemical properties of
3-benzoylisoquinolones were analyzed in relation to those reported
in the literature and extracted from the ChEMBL database.
activity compared to that observed for Bnz at concentrations of
18 and 36
lM, although there was a decrease in activity relative
to both Bnz and 4b at 178
lM (Fig. 3). In addition to eliciting a de-
crease in the number of live bloodstream parasites, the motility of
the remaining parasites was markedly reduced upon introduction
of Bnz. The parasites treated with 4c also exhibited the weak motil-
ity observed in the parasites treated with Bnz. Compound 4a elic-
ited no decrease in the number of bloodstream parasites at any
concentration.
2.3. Physicochemical properties of compounds evaluated
against T. cruzi
A total of 3067 compounds with reported activity data from
inhibition assays against specific enzymes in T. cruzi, and against
the whole organism in functional assays, were retrieved from the
ChEMBL database. The ranges of activity values and the number
of compounds per data set are summarized below in Table 1.
A principal component analysis (PCA) was carried out as
described in the Section 4 using physicochemical descriptors calcu-
lated in MOE 2009.1024 for the combined set of 3067 structures in
order to examine the region of chemical space occupied by the
benzoylisoquinolones in relation to structures screened against T.
cruzi. To project the structures into a chemical space representa-

K. G. Byler et al. / Bioorg. Med. Chem. 20 (2012) 2587–2594
2589
R
O
O
O
i
Br
+
N
NK
O
O
O
2
3
1
OH
O
O
R
ii
NH
4a-c
4a; R= H
4b; R= NO2
4c; R= NH2
iii
iv
i) DMF. ii) NaOMe-MeOH iii) HNO3-H2SO4. iv) H2-Pd/EtOH.
Scheme 1. Synthesis of benzoylisoquinolone derivatives.
contribution to each principal component (PC) as a percentage of
normalized transform coefficients. Of the size and shape descrip-
tors, radius of gyration contributes much more substantially to
all activity models versus molecular weight, reflecting the large
variance in the linear size of compounds tested in the inhibition as-
says. The prominence of the Kier flexibility index among the
descriptors also reveals a significant variance in the flexibility of
assayed compounds. The charge-based descriptors are featured
much less prominently, owing to smaller variations in the charge
distributions in the data sets when compared with those of the size
and shape descriptors. Nevertheless, the polarizability descriptor,
bpol, and the charge descriptors, ASA, ASA+, VSA_POS and VSA_P-
POS, make consistent contributions to the PLS regressions and
were included in a reduced set of ten descriptors that were then
used in a second PCA to produce principal components for the
activity-correlated chemical space projection.
Figure 3. Anti-trypanosomal activity of compounds 4b-c in Trypanosoma cruzi after
24 h of incubation at 4 °C. Statistical significance was determined at P <0.05.
The factor loadings of the principal components were used to
identify the descriptors with significant contributions and these
were used in the second PCA. The first three PC’s, with a total ex-
plained variance of 80%, were then plotted as a three-dimensional
activity-correlated chemical space into which all structures were
projected, as shown in Figure 5. Details of the visualization of the
chemical space and recent applications are reported elsewhere.25,26
The three benzoylisoquinolones appear as gray spheres near the
periphery of the distribution, with the nitro-substituted structure,
tion that would reflect properties relevant to inhibitory activity,
partial least squares (PLS) regressions of activity data were
employed to evaluate the relative contribution of each descriptor
to the prediction of activity in each data set and across all sets.
The descriptors identified as having the highest correlations with
activity across all data sets are shown in Figure 4, along with their
Table 1
Structures from ChEMBL database used in the structural analysis with associated inhibition value ranges.
Assay
Reported inhibition
Range
Number of unique structures
Functional
IC50
EC50
0.25nM – 17 mM
4.0 nM–10 lM
1895
318
% inhibition
ꢀ
212
Target
Cruzain
IC50
IC50
IC50
IC50
0.1 nM–10 mM
403
16
200
23
Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)
Trypanothione reductase (TR)
trans-Sialidase (TS)
100 nM–692
300 nM–133
l
l
M
M
580 nM–1 mM

2590
K. G. Byler et al. / Bioorg. Med. Chem. 20 (2012) 2587–2594
PC1
0.3%
0.3%
PC2
0.0%
0.1%
PC3
0.1%
2.0%
Descriptor
Water accessible surface area (1.4Å)
Water accessible surface area of all atoms with positive partial
charge
ASA
ASA+
FF_VSA_PPOS
KierFlex
1.8%
1.0%
0.4%
Total positive polar van der Waals surface area
Kier molecular flexibility index
13.4%
1.8%
29.8%
Sum of the absolute value of the difference between atomic
polarizabilities of all bonded atoms
Number of O and N atoms
bpol
2.8%
0.5%
4.6%
lip_acc
lip_don
logP(o/w)
logS
12.2%
13.1%
8.8%
12.9%
22.8%
31.6%
23.3%
5.9%
9.0%
34.5%
0.8%
Number of OH and NH atoms
Log of the octanol/water partition coefficient
Log of the aqueous solubility (mol/L)
Radius of gyration
16.4%
30.8%
14.7%
4.0%
rgyr
Figure 4. Descriptor contributions to principal components used in the chemical space projection.
Figure 5. Principal component space projection of combined set of T. cruzi assay compounds from ChEMBL database with isoquinolone structures as large gray spheres.
4b, closest to the main group. The amino structure, 4c, is the far-
thest outlier of the synthesized compounds, occupying a spar-
sely-populated region, suggestive of unexplored chemical space.
ROCS.28 For large datasets a reduction in the number of conformers
analyzed can be valuable. To that end, rmsd-based clustering fol-
lowed by the selection of conformers based on a penalty function
have shown useful.29 ROCS uses molecular volumes represented
by Gaussian functions to overlay structures, and returning the
sum of the Tanimoto indices of 3D structure similarity and chem-
ical feature similarity (TanimotoCombo score).30,31 Further details
of this method and applications can be found in the literature.31–33
For the 2D portion of the analysis, five molecular fingerprints
(Atom Pairs, Radial, MACCS keys, TGD, and piDAPH3) were gener-
ated for each structure, and the Tanimoto index used to rank
compounds most similar to the isoquinolones. The selection of
these fingerprints as suitable representations for similarity search-
ing has been described previously.34,35 The structures with the
highest Tanimoto scores for each fingerprint or ROCS alignment
from the functional assays are presented in Table 2, along with
2.4. Structural comparison of benzoylisoquinolones with
compounds evaluated against T. Cruzi
A structural similarity analysis of the benzoylisoquinolones was
performed against each of the ChEMBL data sets using both 2D and
3D representations in order to identify structurally similar
compounds with known activities in T. cruzi. We have previously
reported the benefits of combining 2D and 3D structural
representations.27
In the 3D analysis, twenty conformers of each isoquinolone
structure were generated using OMEGA28 and aligned to all (500
max.) conformations of each structure in the reference data using

K. G. Byler et al. / Bioorg. Med. Chem. 20 (2012) 2587–2594
2591
Table 2
Structures from ChEMBL functional sets identified as structurally similar to the benzoylisoquinolones
4a
4b
4c
O
O
O
NO2
O
O
N
O
O
ROCS alignments
Ts = 0.694
O
EC50 = 42nM
Ts = 0.730
Ts = 0.673
CHEMBL1609301
IC50 = 1.26μM
IC50 = 1.26μM
CHEMBL256155
CHEMBL256155
O2N
H
N
N
N
N
Ts = 0.140
NH
NH
Radial fingerprints
IC50 = 117μM
CHEMBL209918
O
O
Ts = 0.168
IC50 > 420μΜ
CHEMBL66761
Ts = 0.149
IC50 > 420μΜ
CHEMBL66761
OH
NH2
H
N
H2N
O
N
N
N
Ts = 0.477
O
IC50 = 10.4μM
Atom Pairs fingerprints
O2N
Ts = 0.629
IC50 > 367μΜ
CHEMBL368103
CHEMBL1164085
Ts = 0.500
IC50 = 80μM
CHEMBL508548
OMe
NH
OH
OH
N
N
O
O
O
HN
OH
HN
OH
O
O
O
NO2
O
N
O
N
Ts = 0.700
MACCS keys
EC50 = 262nM
CO2H
CO2H
CHEMBL1335703
Ts = 0.620
96% inhibition @
12μΜ
Ts = 0.660
96% inhibition @
12μΜ
CHEMBL463985
CHEMBL463985
O
O
NH
OMe
N
NH
N
OH
O
N
H
OH
TGD fingerprints
Ts = 0.810
Ts = 0.780
IC50 = 73μΜ
NO2
IC50 = 31.3μM
CHEMBL15193
Ts = 0.810
CHEMBL186474
IC50 = 22.2μM
CHEMBL488544
N
S
S
N
N
CO2H O
CO2H O
N
H
O
O
N
H
S
N
H
piDAPH3 fingerprints
N
N
O
O
Ts = 0.650
IC50 = 1.8μM
Ts = 0.650
Ts = 0.590
IC50 > 90μM
IC50 > 90μM
CHEMBL366482
CHEMBL566122
CHEMBL566122
their ChEMBL identifiers and inhibition values, while those for the
set of target assays are shown in Table 3.
Many of the structures identified as most similar to the isoqui-
nolones by Tanimoto score in the target assays had scores slightly

2592
K. G. Byler et al. / Bioorg. Med. Chem. 20 (2012) 2587–2594
Table 3
Structures from ChEMBL target sets identified as structurally similar to the benzoylisoquinolones.
4a
4b
4c
OH
O
O
N
N
OMe
O
O
O
ROCS alignments
Ts = O0.626
Ts = 0O.638
Ts = 0.638
cruzain IC50 = 2μM
cruzain IC50 >> 1μM
TR IC50 > 50μM
CHEMBL324722
CHEMBL115581
CHEMBL134615
NO2
O
OH
O
N
N
H2N
HO
O
N
H
NH
OH
Radial fingerprints
Ts = 0.175
OH
Ts =O0.168
cruzain IC50 = 860nM
Ts = 0.773
CHEMBL373586
cruzain IC50 > 60μM
cruzain IC50
>
CHEMBL31574
100μg/mL
CHEMBL66761
O
O
O
H
N
O2N
HO2C
O
O
N
H
NO2
O
Ts = 0.396
Atom pairs fingerprints
OH
OH
OH
O
OH
Ts =O0.542
Ts = 0.433
TR IC50 > 50μM
CHEMBL118432
TS IC50 = 710μM
TS IC50 = 14μM
CHEMBL53418
CHEMBL418068
O
O
O
O
HO2C
O2N
OH
N
O
N
O
N
OH
Ts = 0.72H0
Ts = 0H.580
MACCS Keys
O
Ts = 0.620
cruzain IC50 >>
10μM
cruzain IC50 >> 10μM
CHEMBL326294
cruzain IC50 > 10mM
CHEMBL258538
CHEMBL118305
OH
H
OH
H
Cl
H
N
N
N
B
B
O
O
S
OH
S
OH
O
O
O
O
OH
HO
Cl
Ts = 0.830
Ts = 0.790
TGD fingerprints
cruzain IC50 > 500μM
cruzain IC50 > 500μM
CHEMBL1231476
CHEMBL1231476
Ts = 0.820
cruzain IC50 = 73μM
CHEMBL259139
O
O
O
piDAPH3 fingerprints
OH
OH
OH
OH
OH
OH
Ts = O0.620
Ts = O0.590
Ts = O0.560
TS IC50 = 710μM
TS IC50 = 710μM
TS IC50 = 710μM
CHEMBL53418
CHEMBL53418
CHEMBL53418
lower than, but comparable to, those of the functional assays for all
the fingerprints used, save for the radial fingerprints, which consis-
tently yielded poor Tanimoto scores. The majority of similar struc-
tures in the target sets were obtained from the cruzain set,
although the activity values for these compounds were rather poor
in most cases, with inhibitory concentrations ranging from 1–10
half maximal effective concentration of 42 nM, as determined by
a cell-based luminescence high-throughput screen of amastigotes
in NIH3T3 mouse fibroblasts. The most similar compound to the
amino product, 4c, using MACCS keys, was CHEMBL463985, with
96% inhibition (at 12
skeletal myoblasts.
lM) of Tulahuen C2C4 amastigotes in L6 rat
lM. By comparison, the most similar structures in the functional
The similarity matches from the target assays yielded only one
inhibition assays had slightly higher similarity indices to the ben-
zoylisoquinolones, with the best ROCS alignment for 4b, a nitro-
substituted isoquinoline compound (CHEMBL1609301), having a
micromolar inhibitor to 4c, CHEMBL324722, an indoline dione
with a reported IC50 value of 2
lM to cruzain. We observed no
other low-concentration inhibitors in the matches from the target

K. G. Byler et al. / Bioorg. Med. Chem. 20 (2012) 2587–2594
2593
sets; rather, the most similar structures had poor inhibition values.
4.2.1. Statistical analysis
Therefore, based on this analysis, the specific target of interaction
of the compound studied here cannot be suggested.
All results were expressed as the mean percentage (%) of lysis of
a triplicate. Data were analyzed using Dunnett’s method to com-
pare the treatment groups with the control. Statistical significance
was determined as P < 0.05.
3. Conclusions
4.3. Computational methods
The present work illustrates the potential of the easily-synthe-
sized 3-benzoyl-4-hydroxyisoquinolin-1(2H)-one scaffold as anti-
Chagas agents. Two of the samples evaluated exhibited greater
trypanocidal activity than the reference drug benznidazole at the
assayed concentrations. Considering that nitro derivatives usually
show different degrees of toxicity, further studies regarding poten-
tial side effects should be evaluated in the case of 4b.
It is unclear from the structural similarity analysis the specific
target with which the benzoylisoquinolones may be interacting
to effect their trypanocidal action, as all but one structure with
significant similarity have poor activities in their respective target
inhibition assays.
The region occupied by the synthesized compounds in the prop-
erty space projection lies at the outer edge of the distribution of
compounds with known activity in T. cruzi. And in the case of 4c,
this region is also sparsely populated, suggesting that the isoquino-
lone scaffold represents an under-explored domain for trypanoci-
dals, with possibilities for the discovery of novel active compounds.
This work leads to further explore the potential of ben-
zoylisoquinolones as anti-Chagas agents; preparation of analogues
and structure–activity relationships studies are warranted.
4.3.1. Data selection
An assay search of the ChEMBL database using the keywords
Trypanosoma cruzi (CHEMBL368) yielded three major sets of assay
data, which were downloaded 10/2011 as individual sets, accord-
ing to the reported measure of inhibition. The compound data for
the functional assays:% inhibition (212 unique structures from 62
assays), EC50 (318 unique structures from one assay, CHEM-
BL1614129), and IC50 (1895 unique structures from 171 assays)
were converted to three-dimensional structures in MOE 2009.10
and minimized with the Merck ‘94 force field. In addition to the
functional assays, compounds in ChEMBL used in specific target as-
says in T. cruzi were also downloaded and their structures prepared
as previously described: [glyceraldehyde-3-phosphate dehydroge-
nase (GAPDH), CHEMBL5926, IC50 values (100 nM-629 lM), 16 un-
ique structures from one assay, CHEMBL1023015]; [cruzain,
CHEMBL3563, IC50 values (0.1 nM–10 mM), 403 unique structures
from 29 assays]; [trypanothione reductase (TR), CHEMBL5131, IC50
(300 nM–133 lM), 200 unique structures from 16 assays]; [trans-
sialidase (TS), CHEMBL5126, IC50 (580 nM–1 mM), 23 unique
structures from 1 assay, CHEMBL1154902].
4. Methods
4.3.2. Principal Component Analysis
Twenty-four physicochemical descriptors (Supplementary
Table S1) were calculated for all minimized structures in MOE
2009.10 and used in PC analysis. Partial least squares regressions
of the PC’s (as normalized composite descriptors) were performed
using each set of target inhibition data, as well as for the functional
inhibition data. The (descriptor) factor loadings of the three PC’s
with the largest PLS coefficients in each regression were averaged,
weighting by the normalized PLS coefficient for each PC, and scaled
to inspect the relative importance of each descriptor to the predic-
tion of inhibition for each target. Since the functional assays re-
ported different measures of activity, three separate regressions
were performed as described and combined, weighting by the
number of compounds in each set. Ten descriptors that contributed
significantly to the regressions in each data set were selected for a
second PCA, which was used in the chemical space projection.
4.1. Synthesis of benzoylisoquinolones
The lack of purity of the compounds prevents us from unambig-
uously characterize the samples. However, our previously reported
X-ray structure of compound 4a,23 synthesized in a similar man-
ner, validates the synthetic method employed. Additional studies
are under way to fully characterize the chemical structures and
to prepare of analogues for SAR studies. The NMR spectra for 4b
and 4c are provided as Supplementary data, as well as the crystal-
lographic information for compound 4a.23
4.2. In vitro trypanocidal assay
The NINOA strain of T. cruzi employed in this study was isolated
from an acute case of Chagas’ disease in the city of Oaxaca, Mexico.
The live parasite was maintained in the laboratory in the triato-
mine vector Meccus pallidipennis (Insecta:Hemiptera) and in female
white mice, Mus musculus.
All compounds were initially dissolved in dimethyl sulfoxide
(DMSO) at a concentration of 10 mg/mL, followed by serial dilu-
tions with distilled water to achieve the desired concentrations.
All compounds were evaluated against trypomastigotes obtained
by cardiac puncture of infected mice at the peak of parasitemia
using benznidazole (Bnz) as a reference. Heparin was used as an
anticoagulant. The infected blood was diluted with sterile 0.85%
saline solution and the parasites were adjusted to 1 Â 106 blood-
stream trypomastigotes/mL. Bioassays were run in triplicate in
4.3.3. Structural similarity
The 3D structural similarity analysis was performed using the
OpenEye28 programs. A maximum of 500 conformations for each
structure in the 3D alignment analysis were generated using OME-
GA 2.4.1 and each of the benzoylisoquinolone conformations were
aligned to each of the conformations in each data set using ROCS
3.0.0 (MMFF94s force field), ranking alignments by Tanimoto-
Combo score. The TanimotoCombo score was multiplied by ½ in
order to compare similarities with other Tanimoto scores in the
2D analyses.
In the 2D structural similarity analysis, the Atom Pairs and
Radial fingerprints were generated using Canvas 1.4.11237 and
the TGD and piDAPH3 fingerprints, as well as the 166-bit MACCS
keys were generated with MOE 2009.10.
96-well plates containing 195
lL of the trypomastigote suspension
and 5 L of each compound at a concentration of 5, 10 and 50
l
lg/
mL. A DMSO control was assayed in parallel, with the concentra-
tion of DMSO never exceeding 1%. The plates were incubated at
4 °C for 24 h, after which the trypomastigotes were counted using
the method described by Brener.36 Anti trypanosomal activity was
expressed as a percentage of the reduction in the number of live
parasites compared to the control (non-treated parasites).
Acknowledgments
MBA is grateful to SIP and COFAA IPN for financial support.
KMM thanks the State of Florida, Executive Officer of the Gover-
nor’s Office of Tourism, Trade and Economic Development for

2594
K. G. Byler et al. / Bioorg. Med. Chem. 20 (2012) 2587–2594
16. Ponte-Sucre, A.; Faber, J. H.; Gulder, T.; Kajahn, I.; Pedersen, S. E. H.; Schultheis,
M.; Bringmann, G.; Moll, H. Antimicrob. Agents Chemother. 2006, 51, 188.
17. Bender, A. Databases: Compound bioactivities go public Nat. Chem. Biol. 2010,
6, 309.
funding. Authors thank OpenEye Scientific Software for providing
the OMEGA, ROCS and VIDA programs.
18. Gaulton, A.; Bellis, L.; Chambers, J.; Davies, M.; Hershey, A.; Light, Y.;
McGlinchey, S.; Akhtar, R.; Atkinson, F.; Bento, A. P.; Al-Lazikani, B. Nucleic
Acids Res. 2011, 40, D1100.
19. Bolton, E.; Wang, Y.; Thiessen, P. A.; Bryant, S. H. PubChem: Integrated Platform
of Small Molecules and Biological Activities In Annual Reports in Computational
Chemistry; American Chemical Society: Washington, DC, 2008; Vol. 4,.
20. Wang, Y.; Bolton, E.; Dracheva, S.; Karapetyan, K.; Shoemaker, B. A.; Suzek, T.
O.; Wang, J.; Xiao, J.; Zhang, J.; Bryant, S. H. Nucleic Acids Res. 2010, 38, D255.
21. Liu, T.; Lin, Y.; Wen, X.; Jorissen, R. N.; Gilson, M. K. Nucleic Acids Res. 2007, 35,
D198.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2012.02.046.
References and notes
1. Chatelain, E.; Ioset, J. R. Drug Des. Dev. Ther. 2011, 5, 175.
2. Scott, W. L.; O’Donnell, M. J. J. Comb. Chem. 2009, 11, 3.
3. da Costa Santiago, H.; Tanowitz, H. B.; Weiss, L. M.; Montgomery, S. P. PLoS
Neglected Trop. Dis. 2011, 5, e1136.
22. Hill, J. H. M. J. Org. Chem. 1965, 30, 620.
23. Brito-Arias, M.; Castillo-Juárez, K.; Molins, E. Acta Crystallogr., Sect. E: Struct.
Rep. Online 2006, 62, o5261.
24. MOE. MOE. Montreal: Chemical Computing Group; 2009.
25. Medina-Franco, J. L.; Martinez-Mayorga, K.; Giulianotti, M. A.; Houghten, R. A.;
Pinilla, C. Curr. Comput.-Aided Drug Des. 2008, 4, 322.
26. Lopez-Vallejo, F.; Peppard, T. L.; Medina-Franco, J. L.; Martinez-Mayorga, K.
Computational methods for the discovery of mood disorder therapies Expert
Opin. Drug Discovery 2011, 6, 1227.
4. Lammie, P. J.; Hotez, P. J.; Bottazzi, M. E.; Franco-Paredes, C.; Ault, S. K.; Periago,
M. R. PLoS Neglected Trop. Dis. 2008, 2, e300.
5. Reithinger, R.; Tarleton, R. L.; Urbina, J. A.; Kitron, U.; Gürtler, R. E. Br. Med. J.
2009, 338.
6. Rivera, G.; Bocanegra-García, V.; Ordaz-Pichardo, C.; Nogueda-Torres, B.;
Monge, A. Curr. Med. Chem. 2009, 16, 3286.
27. Medina-Franco, J. L.; Martínez-Mayorga, K.; Bender, A.; Marín, R. M.;
Giulianotti, M. A.; Pinilla, C.; Houghten, R. A. J. Chem. Inf. Model. 2009, 49, 477.
28. OpenEye. OEChem. Santa Fe, NM USA: OpenEye Scientific Software, Inc., http://
www.eyesopen.com; 2010.
29. Yongye, A. B.; Bender, A.; Martinez-Mayorga, K. J. Comput. Aided. Mol. Des. 2010,
24, 675.
7. Buschiazzo, A.; Amaya, M. A. F.; Cremona, M. A. L.; Frasch, A. C.; Alzari, P. M.
Molecular Cell 2002, 10, 757.
8. Chan, C.; Yin, H.; Garforth, J.; McKie, J. H.; Jaouhari, R.; Speers, P.; Douglas, K. T.;
Rock, P. J.; Yardley, V.; Croft, S. L.; Fairlamb, A. H. J. Med. Chem. 1998, 41, 4910.
9. Bouzahzah, B.; Jelicks, L. A.; Morris, S. A.; Weiss, L. M.; Tanowitz, H. B. Parasitol.
Res. 2005, 96, 184.
30. Hawkins, P. C. D.; Skillman, A. G.; Nicholls, A. J. Med. Chem. 2006, 50, 74.
31. Sykes, M. J.; McKinnon, R. A.; Miners, J. O. J. Med. Chem. 2008, 51, 780.
32. Yongye, A. B.; Appel, J. R.; Giulianotti, M. A.; Dooley, C. T.; Medina-Franco, J. L.;
Nefzi, A.; Houghten, R. A.; Martinez-Mayorga, K. Biorg. Med. Chem. 2009, 17,
5583.
10. Bodley, A. L.; Shapiro, T. A. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 3726.
11. Gobbi, P.; LoPresti, M. S.; Fernández, A. R.; Enders, J. E.; Fretes, R.; Gea, S.;
Paglini-Oliva, P. A.; Rivarola, H. W. Parasitol. Res. 2007, 101, 1459.
12. Fournet, A.; Hocquemiller, R.; Roblot, F.; Cavé, A.; Richomme, P.; Bruneton, J. J.
Nat. Prod. 1993, 56, 1547.
33. Martinez-Mayorga, K.; Medina-Franco, J. L.; Giulianotti, M. A.; Pinilla, C.;
Dooley, C. T.; Appel, J. R.; Houghten, R. A. Biorg. Med. Chem. 2008, 16, 5932.
34. Waddell, J.; Medina-Franco, J. L. Biorg. Med. Chem., in press. doi:10.1016/
j.bmc.2011.11.051.
13. Fournet, A.; Ferreira, M. E.; Rojas De Arias, A.; Torres De Ortiz, S.; Fuentes, S.;
Nakayama, H.; Schinini, A.; Hocquemiller, R. Antimicrob. Agents Chemother.
1996, 40, 2447.
14. Nakayama, H.; Ferreira, M. E.; Arias, A. R. D.; Bilbao, N. V. D.; Torres, S.; Schinini,
A.; Fournet, A. Phytother. Res. 2001, 15, 630.
´
35. Yongye, A. B.; Byler, K.; Santos, R.; Martinez-Mayorga, K.; Maggiora, G. M.;
Medina-Franco, J. L. Chem. Inf. Model. 2011, 51, 1259.
36. Brener, Z. Revista do Instituto de Medicina Tropical de Sao Paolo 1962, 4, 389.
37. Glide. Glide. New York, NY: Schrödinger, LLC; 2010.
15. Fakhfakh, M. A.; Fournet, A.; Prina, E.; Mouscadet, J.-F.; Franck, X.;
Hocquemiller, R.; Figadère, B. Biorg. Med. Chem. 2003, 11, 5013.