Synthesis and anti-Mycobacterium tuberculosis activity of imide-β-carboline and carbomethoxy-β-carboline derivatives

Article abstract of DOI:10.1016/j.ejmech.2019.111935

A series of methyl β-carboline carboxylates (2a-g) and of imide-β-carboline derivatives containing the phthalimide (4a-g), maleimide (5b, g) and succinimide (6b, e, g) moiety were synthesized, and evaluated for their activity against Mycobacterium tuberculosis H₃₇Rv. The most active β-carboline derivatives against the reference strain were assayed for their cytotoxicity and the activity against resistant M. tuberculosis clinical isolates. Farther, structure-activity relationship (SAR) studies were carried out using the three and four-dimensional approaches for starting to understand the way of β-carboline activity in M. tuberculosis. All 19 β-carboline derivatives were assayed, firstly, by determining the minimum inhibitory concentration (MIC) using resazurin microtiter assay plate (REMA) in M. tuberculosis H₃₇Rv. Then, five derivatives (2c, 4a, 4e, 4g, 6g), which showed MIC ≤ 125 μg/mL, were assayed in nine resistant M. tuberculosis clinical isolates (five MDR, three isoniazid monoresistant and one isoniazid plus streptomycin resistant). The MIC values against the resistant clinical isolates ranged from 31.25 to >250 μg/mL. All five derivatives were non-cytotoxic to the VERO cell line, determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, at the tested concentration (selectivity index ranged from <1.74 to 14.4). Our study demonstrated that (2c) and (6g) derivatives had better anti-M. tuberculosis activity, especially against resistant clinical isolates, what makes them scaffold candidates for further investigations about their anti-tuberculosis activity. The SAR study conducted with the 19 β-carboline derivatives showed the importance of steric effects for the synthesized β-carbolines against M tuberculosis, and these models can be used for future proposition of new derivatives, increasing the chances of obtaining potentially anti-tuberculosis compounds.

Full text of DOI:10.1016/j.ejmech.2019.111935

European Journal of Medicinal Chemistry 187 (2020) 111935
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Synthesis and anti-Mycobacterium tuberculosis activity of
imide- -carboline and carbomethoxy- -carboline derivatives
b
b
,
b
,
, 1
Mariana Aparecida Lopes-Ortiz a 1, Manuela Ribeiro Panice c
,
d
e
a
~
Eduardo Borges de Melo , Joao Paulo Ataide Martins , Vanessa Pietrowski Baldin ,
a
a, f
ꢀ
^
Claudia Terencio Agostinho Pires , Katiany Rizzieri Caleffi-Ferracioli
,
Vera Lúcia Dias Siqueira a f, Regiane Bertin de Lima Scodro a g, Maria Helena Sarragiotto c,
,
,
, f, g, *
Rosilene Fressatti Cardoso a
a
ꢀ
~
^
ꢀ
ꢀ
Programa de Pos-Graduaçao em Biociencias e Fisiopatologia, Universidade Estadual de Maringa (UEM), Avenida Colombo 5790, 87020-900, Maringa, PR,
Brazil
b
c
ꢀ
ꢀ
ꢀ
ꢀ
Uninga e Centro Universitario Inga, Rod. PR 317, 6114, 87035-510, Maringa, PR, Brazil
ꢀ
~
ꢀ
Programa de Pos-Graduaçao em Química, UEM, Maringa, PR, Brazil
Laboratorio de Química Medicinal e Ambiental Teorica (LQMAT), Universidade Estadual do Oeste do Parana, Rua Universitaria, 2069, 85819-110, Cascavel,
d
ꢀ
ꢀ
ꢀ
ꢀ
PR, Brazil
e Departamento de Química, Universidade Federal de Minas Gerais, 31270-901, Belo Horizonte, MG, Brazil
f
ꢀ
ꢀ
Departamento de Analises Clínicas e Biomedicina, UEM, Maringa, PR, Brazil
g
ꢀ
~
^
ꢀ
Programa de Pos-Graduaçao em Ciencias da Saúde, UEM, Maringa, PR, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of methyl
phthalimide (4a-g), maleimide (5b, g) and succinimide (6b, e, g) moiety were synthesized, and evaluated
for their activity against Mycobacterium tuberculosis H37Rv. The most active -carboline derivatives
against the reference strain were assayed for their cytotoxicity and the activity against resistant
M. tuberculosis clinical isolates. Farther, structure-activity relationship (SAR) studies were carried out
b-carboline carboxylates (2a-g) and of imide-b-carboline derivatives containing the
Received 24 September 2019
Received in revised form
29 November 2019
Accepted 30 November 2019
Available online xxx
b
using the three and four-dimensional approaches for starting to understand the way of
activity in M. tuberculosis. All 19 -carboline derivatives were assayed, firstly, by determining the min-
imum inhibitory concentration (MIC) using resazurin microtiter assay plate (REMA) in M. tuberculosis
g/mL, were assayed in nine
b-carboline
b
Keywords:
b
-Carboline derivatives
H37Rv. Then, five derivatives (2c, 4a, 4e, 4g, 6g), which showed MIC ꢀ 125
m
Cytotoxicity
Resazurin microtiter assay plate
Resistance
resistant M. tuberculosis clinical isolates (five MDR, three isoniazid monoresistant and one isoniazid plus
streptomycin resistant). The MIC values against the resistant clinical isolates ranged from 31.25 to
Tuberculosis
Structure-activity relationships
>250
mg/mL. All five derivatives were non-cytotoxic to the VERO cell line, determined by 3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, at the tested concentration (selectivity index
ranged from <1.74 to 14.4). Our study demonstrated that (2c) and (6g) derivatives had better anti-
M. tuberculosis activity, especially against resistant clinical isolates, what makes them scaffold candidates
for further investigations about their anti-tuberculosis activity. The SAR study conducted with the 19
b-
carboline derivatives showed the importance of steric effects for the synthesized -carbolines against M
b
tuberculosis, and these models can be used for future proposition of new derivatives, increasing the
chances of obtaining potentially anti-tuberculosis compounds.
© 2019 Elsevier Masson SAS. All rights reserved.
1. Introduction
ꢀ
~
^
* Corresponding author. Programa de Pos-Graduaçao em Biociencias e Fisiopa-
Tuberculosis still remains a serious public health problem, and is
responsible for the major number of deaths by infectious diseases
worldwide. In 2018, it was estimated that approximately 10 million
people developed tuberculosis, and 1.2 million died due to the
ꢀ
tologia, Universidade Estadual de Maringa (UEM), Avenida Colombo 5790, 87020-
ꢀ
900, Maringa, PR, Brazil.
E-mail addresses: E-mail address: rfressatticardoso@gmail.com (R.F. Cardoso).
rfcardoso@uem.br (R.F. Cardoso).
1
The authors contributed equally to the execution of the study.
https://doi.org/10.1016/j.ejmech.2019.111935
0223-5234/© 2019 Elsevier Masson SAS. All rights reserved.

2
M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
disease [1].
2. Results and discussion
The chemotherapeutic agents for treating patients with tuber-
culosis have been the same since the 70’s. The current treatment
consists in 2 months of a combination of isoniazid (INH), rifampicin
(RIF), ethambutol (EMB), and pyrazinamide (PZA), known as the
attack phase, which is intended to reduce the patient bacterial load.
In the sequence, there is the maintenance phase, composed of 4
months of INH and RIF, to avoid disease recurrence or reactivation.
It is a time-consuming treatment with adverse reactions and drug-
drug interactions, what contributes to the lack of patients adhesion
to complete the treatment, and, consequently, increase of Myco-
bacterium tuberculosis resistance [2].
2.1. Chemistry
The methodology for the synthesis of 1-substitutedphenyl
imide-
maleimide (5), and succinimide (6) moieties at C-3 is outlined in
Scheme 1. The 1-substituted -carboline-3-carbohydrazides 3a-g,
intermediates for the preparation of compounds 4e6, were pre-
pared by the Pictet-Spengler reaction of -tryptophan methyl ester
b-carboline derivatives 4e6 containing a phthalimide (4),
b
L
1a with aromatic aldehydes containing electron donating and
electron withdrawing groups, followed by oxidation of 1,2,3,4-
The increase of resistance added of multidrug-resistant (MDR-
tuberculosis), extensively drug-resistant (XDR-tuberculosis), and
the recent cases of totally-resistant strains (TDR-tuberculosis) are a
threat to disease control, which demands urgent development of
new and more effective drugs. One of the newest anti-tuberculosis
drug, bedaquiline, was approved in the market to treat resistant
disease cases, however resistant bacilli lineages to this drug have
already been reported [2].
The search for new natural or synthetic compounds that inhibit
the in vitro and in vivo M. tuberculosis growth is of paramount
importance for choosing potential candidates for new anti-
tuberculosis drugs.
tetrahydro-
the corresponding methyl
treatment of 2a-g with hydrazine hydrate, as described in our
previous study, afforded the carbohydrazides 3a-g [9].
b
-carbolines obtained with sulfur, in xylene, to give
b
-carboline-3-carboxylates 2a-g [9]. The
The
b-carboline-3-N-(1,3-dioxo-1,3-dihydro-isoindol-2-yl)-car-
boxamides (4a-g) were obtained from the reaction of carbohy-
drazides 3a-g with phthalic anhydride, under reflux in xylene,
according the methodology described by Roston et al. [21]. The
formation of 4a-g was evidenced by the presence of aromatic hy-
drogens at dH 7.75e8.07, in the 1H NMR spectra, and of the addi-
tional signals at dC 129.4e129.5 (C-2a’’/C-6a’’), 123.8e123.9 (C-3’’/
C-600), 135.3e135.5 (C-4’’/C-500) and dC 165.3e165.5 (C]O), in the
13C NMR spectra, which were assigned to the phthalimide moiety.
Following this line of thinking, the well-known natural or syn-
thetic compounds of the
their biological activities [3]. The
b
-carboline class, have been studied for
-carboline alkaloids were pri-
To prepare the
1-yl)-carboxamides 5a and 5g, containing the phenyl and 4-
fluorophenyl groups, respectively, at the position-1 of the -car-
boline nucleus, the carbohydrazides 3a and 3g were treated with
maleic anhydride in acetic acid and anhydrous sodium acetate,
under reflux [21]. The formation of maleimides 5a and 5g was
confirmed by the signals in the 1H and 13C NMR spectra corre-
sponding to the maleimide moiety at dH 7.23e7.27 (brs, H-3’’/H-400),
dC 133.8e133.9 (CH, C-3’’/C-400) and dC 168.2e168.3 (C]O).
b-carboline-3-N-(2,5-dioxo-2,5-dihydro-pirrol-
b
marily isolated from Peganum harmala, an African plant from
Zygophillaceae family, which is used in traditional medicine [3] and
has several biological activities, including antimicrobial [4,5],
antiviral [6], antiparasite [7,8], and antiproliferative [3,9,10]. Besides
that, some reports have shown their activities against Mycobacte-
rium intracellulare [11], Mycobacterium smegmatis [12], and also in
M. tuberculosis H37Rv [12e14]. In M. tuberculosis H37Rv, these
compounds displayed significant activity with minimum inhibitory
b
The preparation of b-carboline-3-N-(2,5-dioxo-pyrrolidin-1-yl)-
concentration (MIC) below 6.25
bacillus [14].
m
g/mL, and selectivity towards the
carboxamides (6) was firstly attempted by the reaction of 3a-g with
succinic anhydride, under anhydrous toluene reflux, as described
by Brosse et al. [22]. However, in this condition only the formation
of the non-cyclized intermediates was observed for all carbohy-
drazides 3a-g. The synthesis of 6b, 6e and 6g was possible by
adding catalytic amount of p-toluenesulphonic acid in the reaction
of the corresponding starting carbohydrazides with succinic an-
hydride. The formation of compounds 6b, 6e and 6g was confirmed
by the presence of signals of succinimide moiety at dH 2.84e2.89,
relative to hydrogens H-300 and H-400, and at 26.2e26.3 and dC
174.3e174.4 relative to C-300 and C-400, and to the imide carbonyl
group, respectively.
Previous studies of our research group demonstrated that 1-
substituted-phenyl- -carbolines, with an amino or guanidinium
b
group-terminated side chain at C-3, showed in vitro activity against
M. tuberculosis H37Rv (ATCC 27294), with MIC values ranging from
24.9 to 58.3
Besides
m
g/mL [15].
b
-carbolines, cyclic imides containing the phthalimide,
maleimide or succinimide subunits, have been described as growth
inhibitors of M. tuberculosis [16e20]. In these studies, a series of
phthalimide derivatives evaluated against M. tuberculosis H37Rv,
showed to be effective against the bacillus with minimum inhibi-
tory concentration (MIC) ranging from 3.9 to 12.5 mg/mL [17,18,20],
and were considered new lead compounds to be studied for the
treatment of susceptible and multidrug resistant tuberculosis [20].
Taking into account, the anti-M. tuberculosis activity showed by
2.2. Biological assays
Our research efforts have focused on the development of
b-
some
the present study, we synthetized and evaluated the activity of
other novel imide- -carboline derivatives against M. tuberculosis
37Rv. Intermediates methyl -carboline-3- carboxylates were also
assayed for their anti-tuberculosis activities. In addition, some of
the -carboline derivatives with higher anti-M. tuberculosis H37Rv
activity were assayed for their cytotoxicity and the activity against
resistant M. tuberculosis clinical isolates. Structure-activity rela-
tionship (SAR) studies were carried out in order to increase the
b
-carbolines and cyclic imide derivatives in the literature, in
carboline alkaloids derivatives with anti-M. tuberculosis activity.
Herein, the present study is innovative in describing the in vitro
activity of new b-carboline derivatives, which can be considered as
potential scaffolds for future development of anti-tuberculosis
agents.
b
H
b
b
Five
studied, displayed anti-M. tuberculosis H37Rv activity (MIC 125
mL) (Table 1). After the screening in the reference strain
M. tuberculosis H37Rv, we assayed the activity of all five -carboline
b-carboline derivatives (2c, 4a, 4e, 4g, 6g), from the 19
mg/
b
understanding about the mechanism of action of
rivatives as anti-M. tuberculosis compounds.
b-carboline de-
derivatives against nine resistant M. tuberculosis clinical isolates
(five MDR, three INH monoresistant and one INH and SM resistant)
(Table 2). The use of MDR isolates in our study, with resistance to
INH (MIC values ranging from 2 to 8
mg/mL) and RIF (MIC values
ranging from 2 to 32 g/mL), previously determined by REMA, can
m

M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
3
Scheme 1. Reagents and Conditions: a) CH3OH, H2SO4, reflux, 48 h; b) Aldehyde (R1CHO), TFA, CH2Cl2, rt, 48 h; c) Sulfur, xylene, 48 h. d) NH2NH2$H2O, EtOH, reflux, 48 h; e) phthalic
anhydride, xylene, reflux, 24 h; f) maleic anhydride, AcONa, acetic acid, reflux, 24 h; g) succinic anhydride, TsOH, dry toluene, reflux, 48 h.
bring further understanding on the activity of these derivatives in
strains that cause tuberculosis that are difficult to treat, and cause a
form of the disease that brings global distress.
et al. observed that the most active anti-tuberculosis tested com-
pounds that have been described in the literature demonstrated
cLogP values ranging from 2 to 6 [25]. In respect to this charac-
teristic, the compounds studied in the present study, 2c, 4a, 4e, 4g
and 6g, showed cLogP values as 4.28, 4.88, 4.84, 5.04 and 1.90,
respectively.
From the five
b-carboline derivatives tested against resistant
clinical isolates, the succinimide-
b
-carboline 6g and the methyl b-
carboline carboxylate 2c showed MIC 31.25 mg/mL against two and
four MDR clinical isolates, respectively.
In this regard it is important to mention, a weak in vitro activity
of a given compound compared to another one does not mean it
will presents weak activity in vivo against a specific microorganism.
Some of these compounds may show higher activity in vivo, by their
metabolic transformation into a more active intermediate com-
The MIC values (31.25
mg/mL) against resistant clinical isolates
are really interesting and are similar to some values found for the
susceptible reference strain H37Rv with different
rivatives [13,14]. In the literature, some synthetic
b
b
-carboline de-
-carboline de-
rivatives have shown excellent activity against M. tuberculosis
37Rv, such as manadomanzamine A and manadomanzamine B
(MIC 1.9 g/mL and 1.5 g/mL, respectively) [23], nostocarboline
and nine dimmers of nostocarboline (MIC ranging from 2.5
pound, as occurs with INH and PZA. Also, b-carbolines interactions,
H
or even an adjuvant effect with old anti-tuberculosis drugs, and the
interaction with the immune system of the patient should be
considered for further investigation with these molecules [26].
m
m
to > 10
to > 50
m
g/mL) [11], sixteen diarylpiperazine (MIC ranging from 1.5
mg/mL) [14], and natural derivatives as manzamine alkaloids
(MIC ranging from 0.4 to 30.2
cited studies above tested the
m
g/mL) [13]. However, none of the
2.3. Structure-activity relationship (SAR)
b
-carboline derivatives in resistant
M. tuberculosis clinical isolates as performed in the present study.
Some -carboline derivatives were also tested against non-
tuberculous mycobacteria (NTM) in other studies. We can name the
Wahba et al. study [11], in which the amidation of -carboline
compounds resulted in potent activity against M. intracellulare. The
nostocarboline and its nine dimmers derivatives above mentioned
also showed to have activity against M. smegmatis (MIC ranged
In an attempt to understand how these new
b-carbolines
b
interact with M. tuberculosis, SAR studies were carried out based on
the MIC values obtained, increasing the chances of obtaining
potentially anti-tuberculosis compounds.
b
The 3D and 4D SAR studies resulted in two qualitative models.
The use of PLS-DA method in the Pirouette 4 program allowed the
prediction equations to be submitted to cross-validation processes.
The 3D model (1) was formed by 8 molecular interaction field (MIF)
descriptors (equation (1)) and 3 latent variables that coded 63.45%
of the information (LV1: 14.29%, LV2: 31.11%, LV3: 18.04%). The 4D
model (2) is formed by 5 descriptors that originate 3 latent vari-
ables and accumulate 82.30% of the information (LV1: 30.08%, LV2:
20.39%, LV3: 31.83%). The obtained results for the adjustment (R2)
and internal prediction ability (Q2LOO) indicate that both explain
and predict information at the levels recommended in the litera-
ture [27e29].
from 12.5 to > 100
The five compounds that displayed better anti-M. tuberculosis
activity (2c, 4a, 4e, 4g, 6g), among the tested -carboline de-
mg/mL) [12].
b
rivatives were evaluated for their cytotoxicity in VERO cell lines
(Table 2). All of them were non-cytotoxic to the VERO cell line at the
tested concentration and the selectivity index (SI) ranged from
<1.74 to 14.4.
According to the literature, compounds that present high lip-
ophilicity tend to have better activity against M. tuberculosis, which
can be explained by the high amount of mycolic acids in the
mycobacteria cell wall structure. Applying the cLogP value for
correlating to lipophilicity, in which cLogP value increases linearly
when the permeability of a compound increases [24], Fernandes
Class ¼ ꢁ0.220ꢂ(150_TIP-TIP) ꢁ0.425ꢂ(169_TIP-
TIP) þ0.345ꢂ(176_TIP-TIP) þ0.243ꢂ(325_DRY-
TIP) ꢁ0.205ꢂ(403_O-TIP) þ0.474ꢂ(439_N1-TIP) ꢁ0.408ꢂ(459_N1-

4
M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
Table 1
Molar Mass and Minimal Inhibitory Concentration (MIC) of b-carboline derivatives against Mycobacterium tuberculosis H37Rv.
Compounds
R1
R2
Molar Mass
MIC
mg/mL
MIC mM
2a
2b
2c
2d
2e
2f
Ph
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
302.11
432.12
345,15
348.11
347.09
336.07
320.10
432.12
>250
250
125
250
250
>250
>250
125
>827.51
578.54
362.16
718.16
720.27
>743.89
>781.01
289.27
4-OMe-Ph
4-Me2N-Ph
3-OMe, 4-OH-Ph
3-NO2-Ph
2-Cl-Ph
2g
4a
4-F-Ph
Ph
4b
4c
4d
4e
4f
4-OMe-Ph
4-Me2N-Ph
3-OMe, 4-OH-Ph
3-NO2-Ph
2-Cl-Ph
462.31
475.16
478.13
477.11
466.08
450.11
382.11
>250
>250
250
>540.76
>526.14
522.87
261.99
536.39
277.71
654.26
125
250
4g
5a
4-F-Ph
125
Ph
250
5g
6b
6e
6g
4-F-Ph
400.10
414.13
429.11
402.11
250
624.84
4-OMe-Ph
3-NO2-Ph
4-F-Ph
>250
>250
125
>603.68
>582.60
310.86
TIP) þ0.290ꢂ(463_N1-TIP) R2 ¼ 0.936; RMSEC ¼ 0.133;
forms, which allow a proper interpretation of the models in relation
to the most and least important descriptors as to the presence or
absence of activity. In model 1 (more information of each descriptor
is available in Table 3), descriptors with information encoded by the
steric probe (TIP) were observed, being present in all descriptors,
with 3 totally steric descriptors (150_TIP-TIP, 169_TIP-TIP and
176_TIP-TIP).
Q2
¼ 0.878; RMSECV ¼ 0.169
(1)
LOO
Class ¼ 0.770ꢂ(14_21_12_NH3_LJ)þ0.509ꢂ(16_11_13_NH3_C)þ
0.337ꢂ(14_7_15_NH3_LJ)þ0.005ꢂ(22_18_10_NH3_LJ)þ
0.046ꢂ(11_7_12_NH3_LJ) R2 ¼ 0.816; RMSEC ¼ 0.213;
Q2
¼ 0.674; RMSECV ¼ 0.256
(2)
LOO
Despite the two most important descriptors (459_N1-TIP and
439_N1-TIP) also present important contributions for hydrogen
In both models the equations are presented in their auto scaled

M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
5
Table 3
Characteristics of the selected descriptors for Model 1.
Sign
Descriptor
Field
Distance range (angstrons)
þ
150
169
TIP-TIP
TIP-TIP
3.60e4.00
11.20e11.60
e
þ
þ
176
325
403
TIP-TIP
DRY-TIP
O-TIP
14.00e14.40
17.20e17.60
10.80e11.20
e
þ
e
439
459
N1-TIP
N1-TIP
6.40e6.80
14.40e14.80
þ
463
N1-TIP
16.00e16.40
bonds, the steric characteristic may be considered more important.
However, the importance of hydrophobicity appears in the
descriptor 325_DRY-TIP, which favors the activity, as already indi-
cated in the literature.
The graphs obtained by PLS-DA analysis (Fig. 1) showed that
these descriptors actually present an excellent discriminatory po-
wer, with active and inactive compounds grouping separately into
two quadrants defined by LV1.
Fig. 2 shows the descriptors formed around an active compound
(2b). It can be clearly observed that the steric components forming
the model (green spheres) are located at the ends of the structures.
For the descriptor DRY-TIP pharmacophore distances are not well
defined as for the others, but it is observed that the hydrophobicity
(yellow dots) is defined by the central part of the structure.
It is interesting to note that these results are in agreement with
several other 3D-QSAR studies already performed with several
compounds that act on the H37Rv strain. The model obtained by Pan
et al. [30] for a set of quinoxaline 1,4-di-N-oxide derivatives showed
a steric contribution of 52.3% of the information.
Shah et al. [31] obtained a model for 2,3-dideoxy hex-2-
enopyranosid-4-uloses with 66.5% steric information. Patel et al.
[32] obtained a model for 4- (adamantan-1-yl) quinolone de-
rivatives where the steric contribution corresponded to 60.40% of
the model. Also, Wang et al. [33] observed that bulky electroposi-
tive groups tended to render nitrofuranyl methyl N-heterocycles
derivatives more active against the strain studied, in addition to the
anti-tuberculosis also having a positive relationship with cLogP.
Another interesting fact observed is that in all the reported
studies the 3D-QSAR CoMFA program was used. Thus, it can be
proposed that the results observed for the anti-tuberculosis activity
are a specific characteristic of the agents against the studied strain,
and it is not a dependent effect of the algorithms for calculating the
descriptors of a specific program.
The discriminatory power of the selected descriptors for model
2 showed to be lower (Fig. 3), with the quadrants defined jointly by
LV1 and LV2. The adjustment and prediction ability were also
inferior to model 1, but the model is still suitable for classification
purposes. The two most important descriptors are those closest to
the conformational ensemble profiles (Fig. 4). The descriptor
named 14_21_12_NH3_LJ is a Lennard-Jones descriptor, while the
descriptor 16_11_13_NH3_C is a Coulomb descriptor.
However, although all the coefficients of the model were posi-
tive, Coulomb’s descriptor values were negative for all analyzed
derivatives, indicating that the increased polarization of the mol-
ecules can cause loss of activity. This strengthens the interpretation
performed for model 1, thus gaining prominence the importance of
steric effects and liposolubility in the discrimination.
The results observed also strengthens the hypothesis that the
steric effect is an important characteristic for this type of molecules,
independent of the program or methodology used to generate the
descriptors, since the MIF descriptors are obtained based on a set of
different conformations for each molecule, not just a single

6
M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
Fig. 1. Plot of the loading and score vectors of 3D model obtained by PLS-DA approach.
Fig. 2. GRIND selected descriptors of Model 1 associated with an active compound 2b. Red lines: positive descriptors; blue lines: negative descriptors. (For interpretation of the
references to color in this figure legend, the reader is referred to the Web version of this article.)
optimized geometry, as commonly performed in 3D-QSAR.
against resistant clinical isolates, what makes them scaffold can-
didates for further investigations about their anti-tuberculosis ac-
tivity. The SAR studies performed using 3D and 4D methodologies
indicated that the activity was influenced mainly by the steric
characteristics (size and shape) and lipophilicity of the molecules,
meeting the previous information available in the literature. These
3. Conclusion
The present study demonstrated that five of the
b-carboline
derivatives tested showed anti-M. tuberculosis activity, especially

M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
7
Fig. 3. Plot of the loading and score vectors of 4D model obtained by PLS-DA approach. In this model, all descriptors have positive signs.
Fig. 4. Full 4D model in three-dimensional space surrounding the CEP of one active compound 2b. Blue: Coulomb descriptor; pink: Lennard-Jones descriptor. (For interpretation of
the references to color in this figure legend, the reader is referred to the Web version of this article.)
models can be used for future proposition of new derivatives,
increasing the chances of obtaining potentially anti-tuberculosis
compounds.
4.2.1.1. 1-(Phenyl)-b-carboline-3- N-(1,3-dioxo-1,3-dihydro-isoindol-
2-yl)-carboxamide (4a). Yield: 85%; mp: 262.4e265.3; IR (KBr)
nmax: 3377 (NH), 1622 (C]N), 1703 and 1739 (C]O), 1460e1595
(C]C) cmꢁ1; 1H NMR (300 MHz, DMSO‑d6):
d 8.91 (s, 1H, H-4), 8.46
(d, J ¼ 7.2 Hz, 1H, H-5), 7.33 (t, J ¼ 7.2 Hz, 1H, H-6), 7.51e7.75 (m, 5H,
H-7, H-8, H-30, H-40 and H-50), 8.25 (d, J ¼ 7.2 Hz, 2H, H-20and H-60),
8.02e8.06 (m, 2H, H-300 and H-600), 7.97e8.01 (m, 2H, H-400 and H-
500), 11.10 (s, 1H, NH), 12.00 (s, 1H, 9-NH); 13C NMR (75.5 MHz):
4. Experimental
4.1. General methods
d
141.5 (C, C-1), 134.7 (C, C-3), 114.4 (CH, C-4), 129.7 (C, C-4a), 121.3
All reagents were purchased from commercial suppliers. The
reactions were monitored by thin layer chromatography conducted
on Whatman TLC plates (Silica Gel 60 F254). NMR spectra were
recorded in an in Varian spectrometer model Mercury plus BB at
300.0 MHz (for 1H) and 75.5 MHz (for 13C) with TMS and deuterated
solvents, dimethyl sulfoxide (DMSO‑d6) as internal standard. EI-MS
spectra were recorded in a Thermoelectron Corporation Focus-DSQ
II spectrometer. Melting points were determined in Microquímica
apparatus model MQAPF-301, and were not corrected.
(C, C-4b), 122.1 (CH, C-5), 120.4 (CH, C-6), 128.7 (CH, C-7, C-20 and C-
60), 112.8 (CH, C-8), 141.2 (C, C-8a), 137.3 (C, C-9a), 129.1 (C, C-10),
128.5 (CH, C-30, C-40and C-50), 129.5 (C, C-2a’’ and C-6a’’), 123.8 (CH,
C-300 and C-600), 135.3 (C-400 and C-500), 164.1 (C, C]O), 165.3 (C, C]
O). EI-MS: calcd for C26H16N4O3 432.122, found: 432.956 [Mþ$].
4.2.1.2. 1-(4-Methoxyphenyl)-b-carboline-3-N-(1,3-dioxo-1,3-
dihydro-isoindol-2-yl)-carboxamide (4b). Yield: 89%; mp: > 280.0
(decomp.); IR (KBr) nmax: 3290 (NH), 1625 (C]N), 1670 and 1735
4.2. Synthesis
(C]O), 1440e1566 (C]C) cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
;
d
8.85 (s, 1H, H-4), 8.45 (d, J ¼ 7.8 Hz, 1H, H-5), 7.31e7.36 (m, 1H, H-
4.2.1. General procedure for the synthesis of 1-(substituted-phenyl)-
6), 7.60e7.65 (m, 1H, H-7), 7.74 (d, J ¼ 7.8 Hz, 1H, H-8), 8.25 (d,
J ¼ 7.8 Hz, 2H, H-20and H-60), 7.22 (d, J ¼ 7.8, 2H, H-30 and H-50), 3.90
(s, 3H, OCH3), 8.02e8.06 (m, 2H, H-300 and H-600), 7.97e8.01 (m, 2H,
H-400 and H-500), 11.08 (s, 1H, NH), 12.96 (s, 1H, 9-NH); 13C NMR
b
-carboline-3- N-(1,3-dioxo-1,3-dihydro-isoindol-2-yl)-
carboxamides (4a-g)
A solution of carbohydrazides 3a-g (2.0 mmol) and phthalic
anhydride (3.0 mmol) in xylene was refluxed for 24 h under stir-
ring. Then, the reaction mixture was cooled, and the resulting solid
was filtered off and washed with cold ethanol to furnish the
phthalimides 4a-g.
(75.5 MHz):
d 141.5 (C, C-1), 134.5 (C, C-3), 113.9 (CH, C-4), 129.5 (C,
C-4a), 121.3 (C, C-4b), 122.2 (CH, C-5), 120.5 (CH, C-6), 128.7 (CH, C-
7), 112.8 (CH, C-8), 141.2 (C, C-8a), 137.2 (C, C-9a), 129.5 (C, C-10),
130.4 (CH, C-20 and C-60), 114.2 (CH, C-30 and C-50), 160.1 (C, C-40),

8
M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
55.4 (OeCH3), 129.5 (C, C-2a’’ and C-6a’’), 123.9 (CH, C-300 and C-600),
135.5 (C-400 and C-500), 164.2 (C, C]O), 165.5 (C, C]O).). EI-MS:
calcd for C27H18N4O4 462.133, found: 462.951 [Mþ$].
127.5 (CH, C-7), 112.5 (CH, C-8), 140.5 (C, C-8a),136.9 (C, C-9a), 132.6
(C, C-10), 129.5 (C, C-20), 129.6 (CH, C-30), 132.2 (CH, C-40), 128.9 (CH,
C-50), 130.8 (CH, C-60), 129.5 (C, C-2a’’ and C-6a’’), 123.8 (C-300 and C-
600), 135.3 (CH, C-400 and C-500), 164.0 (C, C]O), 165.3 (C, C]O). EI-
MS: calcd for C26H15ClN4O3 466.083, found: 466.002 [Mþ$].
4.2.1.3. 1-(4-Dimethylaminophenyl)-
dihydro-isoindol-2-yl)-carboxamide
b
-carboline-3-N-(1,3-dioxo-1,3-
(4c). Yield: 84%; mp:
297.6e299.8; IR (KBr) nmax: 3377 (NH), 1608 (C]N), 1668 and 1745
4.2.1.7. 1-(4-Fluorophenyl)-
isoindol-2-yl)-carboxamide (4g). Yield: 90%; mp:
(decomp.); IR (KBr) nmax: 3330 (NH), 1623 (C]N), 1706 and 1733
(C]O), 1444e1562 (C]C) cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
b
-carboline-3-N-(1,3-dioxo-1,3-dihydro-
(C]O), 1458e1560 (C]C) cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
;
>
280.0
d
8.89 (s, 1H, H-4), 8.31 (d, J ¼ 8.1 Hz, 1H, H-5), 7.34 (t, J ¼ 8.1 Hz, 1H,
H-6), 8.31 (d, J ¼ 8.1 Hz, 1H, H-7), 7.72 (d, J ¼ 8.1 Hz, 1H, H-8), 8.45
(d, J ¼ 8.1 Hz, 2H, H-20and H-60), 7.63 (d, J ¼ 8.1, 2H, H-30 and H-50),
3.44 (s, 1H, N(CH3)2), 7.99e8.03 (m, 2H, H-300 and H-600), 7.99e8.03
(m, 2H, H-400 and H-500), 11.12 (s, 1H, NH), 12.05 (s, 1H, 9-NH); 13C
;
d
8.89 (s, 1H, H-4), 8.45 (d, J ¼ 7.8 Hz, 1H, H-5), 7.33 (t, J ¼ 7.8 Hz, 1H,
H-6), 7.62 (t, J ¼ 7.8 Hz, 1H, H-7), 7.72 (d, J ¼ 7.8 Hz, 1H, H-8), 8.30
(dd, J ¼ 9.0 and 8.7 Hz, 2H, H-20and H-60), 7.49 (dd, J ¼ 9.0 and
8.7 Hz, 2H, H-30 and H-50), 9.88 (s, 1H, OH), 8.02e8.06 (m, 2H, H-300
and H-600), 7.97e8.01 (m, 2H, H-400 and H-500),11.13 (s,1H, NH),12.03
NMR (75.5 MHz):
d 141.5 (C, C-1), 137.3 (C, C-3), 114.4 (CH, C-4),
129.8 (C, C-4a),121.1 (C, C-4b),130.0 (CH, C-5 and C-7), 120.5 (CH, C-
6), 112.7 (CH, C-8), 142.7 (C, C-8a), 134.7 (C, C-9a), 133.3 (C, C-10),
122.2 (CH, C-20 and C-60), 121.2 (CH, C-30 and C-50), 162.0 (C, C-40),
129.5 (C, C-2a’’ and C-6a’’), 123.8 (CH, C-300 and C-600), 135.4 (C-400
and C-500), 164.1 (C, C]O), 165.3 (C, C]O). EI-MS: calcd for
(s, 1H, 9-NH); 13C NMR (75.5 MHz):
d 141.5 (C, C-1), 137.3 (C, C-3),
114.5 (CH, C-4), 129.8 (C, C-4a), 121.1 (C, C-4b),122.2 (CH, C-5), 120.5
(CH, C-6), 128.9 (CH, C-7), 112.7 (CH, C-8), 140.2 (C, C-8a), 133.6 (C,
C-10), 131.2 (CH, C-20 and C-60), 115.6 (CH, C-30 and C-50), 162.7 (C, C-
40),129.5 (C, C-2a’’ and C-6a’’),123.8 (CH, C-300 and C-600),135.4 (C-400
and C-500), 164.1 (C, C]O), 165.4 (C, C]O). HRMS-ESI: calcd for
450.421, found: 449.924. EI-MS: calcd for C26H15FN4O3 450.113,
found: 450.948 [Mþ$].
C
28H21N5O3 475.164, found: 475.0 [Mþ$].
4.2.1.4. 1-(3-Methoxy-4-hydroxyphenyl)-
b
-carboline-3-N-(1,3-
dioxo-1,3-dihydro-isoindol-2-yl)-carboxamide (4d). Yield: 83%;
mp: 210.8e212.4; IR (KBr) nmax: 3319 (NH), 1623 (C]N), 1666 and
1743 (C]O), 1463e1595 (C]C) cmꢁ1
;
1H NMR (300 MHz,
4.2.2. General procedure for synthesis 1-(substituted-phenyl)-b-
carboline-3-N-(2,5-dioxo-2,5-dihydro-pirrol-1-yl)-carboxamides
(5a and g)
To a solution of the carbohydrazides 3a and 3g (2.0 mmol) and
maleic anhydride (2.0 mmol) in acetic acid (20 mL), sodium acetate
(2.5 mmol) was added. The resulting solution was refluxed for 24 h,
then treated with cold-water. The solid formed was filtered off and
washed with cold water to furnish the products 5a and 5g,
respectively.
DMSO‑d6):
d
8.82 (s, 1H, H-4), 8.42 (d, J ¼ 7.8 Hz, 1H, H-5), 7.32 (t,
J ¼ 7.8 Hz, 1H, H-6), 7.62 (t, J ¼ 7.8 Hz, 1H, H-7), 7.70e773 (m, 2H, H-
8 and H-50), 7.04 (d, J ¼ 7.8 Hz, 1H, H-20), 7.63 (d, J ¼ 7.8, 1H, H-60),
8.01e8.05 (m, 2H, H-300 and H-600), 3.85 (s, 1H, OCH3), 9.43 (s, 1H,
OH), 10.90 (s, 1H, NH), 12.00 (s, 1H, 9-NH); 13C NMR (75.5 MHz):
d
141.4 (C, C-1), 134.5 (C, C-3), 113.7 (CH, C-4), 129.3 (C, C-4a), 121.3
(C, C-4b), 122.0 (CH, C-5), 120.3 (CH, C-6), 128.7 (CH, C-7), 112.7 (CH,
C-8), 141.4 (C, C-8a), 141.9 (C, C-9a), 128.5 (C, C-10), 115.5 (CH, C-20),
147.9 (CH, C-30), 147.8 (CH, C-40), 113.1 (CH, C-50), 121.7 (C, C-60),
129.5 (C, C-2a’’ and C-6a’’), 123.8 (CH, C-300 and C-600), 135.4 (C-400
and C-500), 164.2 (C, C]O), 165.4 (C, C]O). EI-MS: calcd for
4.2.2.1. 1-(Phenyl)-
1-yl)-carboxamide (5a). Yield: 70%; mp: 187.5e189.2; IR (KBr) nmax
3337 (NH), 1626 (C]N), 1697 and 1730 (C]O), 1460e1595 (C]C)
cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
8.87 (s, 1H, H-4), 8.45 (d,
b-carboline-3- N-(2,5-dioxo-2,5-dihydro-pyrrol-
26H16N4O3 448.117, found: 448.963 [Mþ$].
:
C
;
d
4.2.1.5. 1-(3-Nitrophenyl)-
isoindol-2-yl)-carboxamide (4e). Yield: 80%; mp:
(decomp.); IR (KBr) nmax: 3369 (NH), 1623 (C]N), 1706 and 1733
(C]O), 1444e1560 (C]C) cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
b
-carboline-3-N-(1,3-dioxo-1,3-dihydro-
J ¼ 7.5 Hz, 1H, H-5), 7.33 (t, J ¼ 7.5 Hz, 1H, H-6), 7.55e7.73 (m, 5H, H-
7, H-8 and H-40), 8.22 (d, J ¼ 7.5 Hz, 2H, H-20and H-60), 8.22 (d,
J ¼ 7.22 Hz, 2H, H-30 and H-40), 7.23 (brs, 2H, H-300 and H-400), 10.90
>
280.0
;
(s, 1H, NH), 12.00 (s, 1H, 9-NH); 13C NMR (75.5 MHz):
d 141.6 (C, C-
d
8.99 (s, 1H, H-4), 8.50 (d, J ¼ 7.8 Hz, 1H, H-5), 7.38 (t, J ¼ 7.8 Hz, 1H,
1), 134.7 (C, C-3), 114.4 (CH, C-4), 129.7 (C, C-4a), 121.2 (C, C-4b),
122.2 (CH, C-5), 120.5 (CH, C-6), 128.8 (CH, C-7 and C-40), 112.8 (CH,
C-8), 141.2 (C, C-8a), 137.1 (C, C-9a), 129.1 (C, C-10), 128.9 (CH, C-20
and C-60), 129.5 (CH, C-30, C-40and C-50), 133.9 (C, C-300 and C-400),
164.2 (C, C]O), 168.3 (C, C]O). EI-MS: calcd for C22H14N4O3:
382.107, found: 382.927 [Mþ$].
H-6), 7.69 (t, J ¼ 7.8 Hz, 1H, H-7), 7.73 (d, J ¼ 7.8 Hz, 1H, H-8), 8.30
(brs, 1H, H-20), 8.43 (dd, J ¼ 7.8 and 1.8 Hz, 1H, H-40), 7.96 (t,
J ¼ 7.8 Hz, 1H, H-50), 8.67 (d, J ¼ 7.8 Hz, 1H, H-60), 7.97e8.07 (m, 4H,
H-30, H-400, H-500 and H-600), 11.24 (s, 1H, NH), 12.19 (s, 1H, 9-NH); 13C
NMR (75.5 MHz):
d 141.6 (C, C-1), 138.5 (C, C-3), 115.3 (CH, C-4),
129.5 (C, C-4a),121.4 (C, C-4b),122.4 (CH, C-5),120.6 (CH, C-6),129.1
(CH, C-7), 112.6 (CH, C-8), 138.9 (C, C-8a), 129.4 (C, C-10), 123.7 (CH,
C-20), 148.3 (CH, C-30), 123.8 (C, C-40), 130.2 (CH, C-50), 135.4 (CH, C-
60),129.4 (C, C-2a’’ and C-6a’’),123.8 (CH, C-300 and C-600),135.3 (C-400
and C-500), 163.9 (C, C]O), 165.3 (C, C]O). EI-MS: calcd for
4.2.2.2. 1-(4-Fluorophenyl)-
dihydro-pyrrol-1-yl)-carboxamide
182.5e184.5; IR (KBr) nmax: 3350 (NH), 1623 (C]N), 1687 and 1730
(C]O), 1461e1510 (C]C) cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
b
-carboline-3- N-(2,5-dioxo-2,5-
(5g). Yield: 80%; mp:
C
26H15N5O5 477.107, found: 477.921 [Mþ$].
;
d
8.87 (s, 1H, H-4), 8.45 (d, J ¼ 8.0 Hz, 1H, H-5), 7.34 (t, J ¼ 8.0 Hz, 1H,
4.2.1.6. 1-(2-Chlorophenyl)-
b
-carboline-3-N-(1,3-dioxo-1,3-dihydro-
H-6), 7.63 (t, J ¼ 8.0 Hz, 1H, H-7), 7.72 (d, J ¼ 8.0 Hz, 1H, H-8), 8.30
(dd, J ¼ 8,7 and 8.4 Hz, 2H, H-20 and H-60), 7.49 (dd, J ¼ 8.7 and 8.4,
2H, H-30 and H-50) 7.27 (brs, H-300 and H-400, 10.91 (s, 1H, NH), 12.01
isoindol-2-yl)-carboxamide (4f). Yield: 90%; mp: 194.9e196.3; IR
(KBr) nmax: 3269 (NH), 1623 (C]N), 1664 and 1737 (C]O),
1460e1598 (C]C) cmꢁ1
;
1H NMR (300 MHz, DMSO‑d6):
d
8.98 (s,
(s, 1H, 9-NH); 13C NMR (75.5 MHz):
d 141.5 (C, C-1), 134.5 (C, C-3),
1H, H-4), 8.48 (d, J ¼ 7.8 Hz, 1H, H-5), 7.31e7.37 (d, J ¼ 7.8 Hz, 1H, H-
6), 7.58e7.67 (m, 1H, H-7 e H-8), 7.72e7.78 (m, 2H, H-30and H-40),
7.58e7.67 (m, 2H, H-50 and H-60), 7.95e8.04 (m, 2H, H-300 and H-600)
7.95e8.04 (m, 2H, H-400 and H-500), 10.95 (s, 1H, NH), 11.85 (s, 1H, 9-
114.3 (CH, C-4), 129.8 (C, C-4a), 121.1 (C, C-4b), 122.1 (CH, C-5), 120.5
(CH, C-6),128.8 (CH, C-7),112.7 (CH, C-8), 140.1 (C, C-8a),137.2 (C, C-
9a), 133.5 (J ¼ 2.9 Hz, C, C-10), 131.2 (J ¼ 8.3, CH, C-20 and C-60), 115.5
(J ¼ 21.9, CH, C-30 and C-50), 162.7 (J ¼ 246.9, C, C-40), 133.8 (C, C-300
and C-400), 164.1 (C, C]O), 168.2 (C, C]O). EI-MS: calcd for
NH); 13C NMR (75.5 MHz):
d
141.4 (C, C-1), 136.2 (C, C-3), 115.2 (CH,
C-4), 129.5 (C, C-4a), 121.0 (C, C-4b), 122.4 (CH, C-5), 120.4 (CH, C-6),
C
22H13FN4O3: 400.097, found: 400.945 [Mþ$].

M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
9
4.2.3. General procedure for synthesis of 1-(substituted-phenyl)-
carboline-3-N-(2,5-dioxo-pirrolidin-1-yl)-carboxamides (6b, e, g)
b
-
compounds were then assayed against nine resistant
M. tuberculosis clinical isolates (five MDR, three INH monoresistant
and one INH and streptomycin (SM) resistant). All resistant clinical
isolates used in the study were selected from the mycobacterial
collection from the laboratory of Medical Bacteriology, Department
To a solution of the carbohydrazides 3b, 3e and 3g (2.0 mmol)
and succinic anhydride (2.0 mmol) in anhydrous toluene, a catalytic
amount of p-toluene sulphonic acid was added. The resulting so-
lution was refluxed for 48 h with a Dean-Stark apparatus in order to
remove the water formed during the reaction. Then, the reaction
mixture was cooled, and the resulting solid was filtered off, and
washed with 15% aqueous Na2CO3 and water to furnish the prod-
ucts 6b, 6e and 6g, respectively.
ꢀ
of Clinical Analysis and Biomedicine, State University of Maringa,
ꢀ
Parana, Brazil. The isolates were previously biochemically identi-
fied as M. tuberculosis complex and the drug susceptibility testing
€
carried out by Lowenstein-Jensen proportion method [33] and
genotyped by spoligotyping [34] and Mycobacterial Interspersed
Repetitive Unit - Variable Number Tandem Repeat (MIRU-VNTR)
(Table 2) [35].
4.2.3.1. 1-(4-Methoxyphenyl)-b-carboline-3-N-(2,5-dioxo-pirrolidin-
1-yl)-carboxamide (6b). Yield: 80%; mp: 241.9e243.4; IR (KBr)
nmax: 3330 (NH), 1623 (C]N), 1695 and 1728 (C]O), 1461e1593
The reference strain and resistant clinical isolates were grown in
Middlebrook 7H9 broth medium (Difco Laboratories, Detroit, USA),
added of 0.2% glycerol and supplemented with oleic acid, albumin,
dextrose and catalase (OADC) enrichment (BBL/Becton-Dickinson,
Sparks, MD, USA) (7H9-OADC) for 15 days at 35 ꢃC. Standardized
bacilli inoculums, with turbidity equivalent to McFarland scale 1
(3.0 ꢂ 108 CFU/mL), were prepared and diluted 1:20 in 7H9-OADC.
Ethical approval was not required for the study.
(C]C) cmꢁ1; 1H NMR (300 MHz, DMSO‑d6):
d 8.81 (s, 1H, H-4), 8.43
(d, J ¼ 7.3 Hz, 1H, H-5), 7.33 (t, J ¼ 7.3 Hz, 1H, H-6), 7.61 (t, J ¼ 7.3 Hz,
1H, H-7), 7.72 (d, J ¼ 7.3 Hz, 1H, H-8), 8.20 (d, J ¼ 7.3 Hz, 2H, H-20and
H-60), 7.20 (d, J ¼ 7.3 Hz, 2H, H-30and H50), 2.88 (brs, 2H, H-300 and
H-400), 10.83 (s, 1H, NH), 11.94 (s, 1H, 9-NH); 13C NMR (75.5 MHz):
d
141.5 (C, C-1), 134.8 (C, C-3), 113.7 (CH, C-4), 129.6 (C, C-4a), 121.1
(C, C-4b), 122.1 (CH, C-5), 120.2 (CH, C-6), 128.7 (CH, C-7), 112.8 (CH,
C-8), 141.1 (C, C-8a), 137.4 (C, C-9a), 129.5 (C, C-10), 130.3 (CH, C-20
and C-60), 114.1 (CH, C-30 and C-50), 160.1 (CH, C-40), 55.4 (OCH3),
26.3 (C, C-300 and C-400), 163.4 (C, C]O), 174.4 (C, C]O). EI-MS: calcd
for C23H18N4O4: 414.133, found: 414.047 [Mþ$].
4.3.2. Minimum inhibitory concentration (MIC)
Anti-M. tuberculosis activities of 19 new
were determined by resazurin microtiter assay plate (REMA) [36] in
three independent assays at different days. Firstly, -carboline de-
b-carboline derivatives
b
rivatives stock solutions were prepared in dimethyl sulfoxide
(Sigma-Aldrich, St Louis, MO, USA). The derivatives stock solutions
4.2.3.2. 1-(3-Nitrophenyl)-
yl)-carboxamide (6e). Yield: 40%; mp: 199.6e199.5; IR (KBr) nmax
3373 (NH), 1623 (C]N), 1701 and 1712 (C]O), 1460e1593 (C]C)
cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
8.93 (s, 1H, H-4), 8.46 (d,
J ¼ 8.1 Hz,1H, H-5), 7.34 (t, J ¼ 8.1 Hz,1H, H-6), 7.63 (t, J ¼ 8.1 Hz,1H,
H-7), 7.70 (d, J ¼ 8.1 Hz, 1H, H-8), 8.94 (s, 1H, H-20), 8.40 (dd, J ¼ 8.1
and 2.1 Hz, 1H, H), 7.92 (t, J ¼ 8.1 Hz, 1H, H50), 8.64 (d, J ¼ 8.1 Hz, 1H,
H-60), 2.84 (brs, 2H, H-300 and H-400), 11.00 (s, 1H, NH); 13C NMR
b-carboline-3- N-(2,5-dioxo-pirrolidin-1-
:
were diluted in 7H9-OADC to 1000
derivatives were carried out in 96-well sterile microplates and
100 L of previously standardized mycobacterial inoculum were
added to obtain final derivatives concentrations from 0.98 to
250
g/mL. The microplates were incubated for 7 days at 35 ꢃC in
normal atmosphere. After the incubation time, 30 L of freshly
mg/mL. New dilutions of all
;
d
m
m
m
prepared 0.01% resazurin solution (Acros, Morris Plains, NJ, USA)
were added to each well. The microplates were incubated for
additional 24 h at 35 ꢃC for subsequent visual reading. A color
change from blue to pink, indicated mycobacterial growth, and the
MIC was defined as the lowest concentration of the derivatives that
(75.5 MHz):
d 141.6 (C, C-1), 137.5 (C, C-3), 115.1 (CH, C-4), 130.3 (C,
C-4a), 121.1 (C, C-4b), 122.4 (CH, C-5), 120.6 (CH, C-6), 129.1 (CH, C-
7), 112.8 (CH, C-8), 138.7 (C, C-8a), 138.6 (C, C-9a), 135.5 (J ¼ 3.3 HZ,
C, C-10), 123.7 (CH, C-20), 148.3 (C, C-30), 123.7 (CH, C-40), 130.3 (CH,
C-50), 135.5 (CH, C-60), 26.3 (C, C-300 and C-400), 163.3 (C, C]O), 174.4
(C, C]O).). EI-MS: calcd for C22H15N5O5: 429.107, found: 429.010
[Mþ$].
prevented color change. Medium, b-carboline derivatives sterility
and mycobacterial growth with and without 2.5% (v/v) DMSO
controls were included in all assays. Isoniazid (Sigma-Aldrich, St
Louis, MO, USA) was used as the reference drug in all assays.
4.2.3.3. 1-(4-Fluorophenyl)-
yl)-carboxamide (6g). Yield: 79%; mp: 281.3e283.6; IR (KBr) nmax
3348 and 3236 (NH), 1623 (C]N), 1699 and 1731 (C]O),
1461e1593 (C]C) cmꢁ1 1H NMR (300 MHz, DMSO‑d6):
8.72 (s,
b-carboline-3-N-(2,5-dioxo-pirrolidin-1-
:
4.3.3. Cytotoxicity
The cytotoxicity of b-carboline derivatives was determined in
;
d
three independent assays at different days, using VERO epithelial
cells (ATCC CCL81). The cells were cultured in 10% Fetal Bovine
Serum (FBS) Dulbecco’s Modified Eagle’s Medium (DMEM) and
incubated at 37 ꢃC under 5% CO2 until confluent growth. In a
transparent 96-well microplate, 5 ꢂ 104 cell/well were added, and
incubated at 37 ꢃC under 5% CO2 for 24 h. After, the medium was
1H, H-4), 8.46 (d, J ¼ 8.0 Hz, 1H, H-5), 7.35 (t, J ¼ 8.0 Hz, 1H, H-6),
7.63 (t, J ¼ 8.0 Hz, 1H, H-7), 7.72 (d, J ¼ 8.0 Hz, 1H, H-8), 8.29 (dd,
J ¼ 8.7 and 8.7 Hz, 2H, H-20and H-60), 7.490 (dd, J ¼ 8.7 and 8.7 Hz,
2H, H-30and H50), 2.89 (brs, 2H, H-300 and H-400), 10.88 (s, 1H, NH),
12.01 (s, 1H, 9-NH); 13C NMR (75.5 MHz):
d 141.5 (C, C-1), 134.5 (C,
C-3), 114.2 (CH, C-4), 129.8 (C, C-4a), 121.1 (C, C-4b), 122.2 (CH, C-5),
120.5 (CH, C-6),128.8 (CH, C-7),112.7 (CH, C-8),140.1 (C, C-8a),137.4
(C, C-9a), 135.5 (J ¼ 3.3 HZ, C, C-10), 1311 (J ¼ 8.8 Hz, CH, C-20 and C-
60), 115.5 (J ¼ 21.6 Hz, CH, C-30 and C-50), 162.6 (J ¼ 246.6 Hz, C, C-40),
26.3 (C, C-300 and C-400), 163.3 (C, C]O), 174.3 (C, C]O). EI-MS: calcd
for C22H15FN4O3: 402.113, found: 402.035 [Mþ$].
removed and different concentrations of the
b-carboline de-
rivatives (1.95e1000 g/mL), diluted in DMEM, were added to the
m
microplate wells. All medium was removed after 24-h incubation at
37 ꢃC under 5% CO2, and the microplate wells washed twice with
phosphate-buffered saline (PBS). Then, 3-(4,5-dimethylthiazol-2-
yl)-2,5-diphenyltetrazolium bromide (MTT) was diluted in PBS
(2 mg/mL), and 50
microplate was incubated at 37 ꢃC under 5% CO2 for 4 h. After, MTT
was removed and 150 L of DMSO was added to the microplate
mL were added to each microplate well. The
4.3. Biological assays
m
4.3.1. Mycobacterium tuberculosis reference strain and clinical
isolates
wells, and absorbance was determined at 550 nm. The CC50 value
was defined as the highest drug concentration that 50% of the cells
are viable in comparison to the control. The selectivity index (SI)
was calculated by the CC50/MIC ratio against the reference strain
The pan-susceptible M. tuberculosis H37Rv (ATCC 27294) refer-
ence strain was used for screening the
b-carboline derivatives
compounds activities (Table 1). The most anti-tuberculosis active
H37Rv and clinical isolates.

10
M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
4.3.4. Three and four-dimensional structure-activity relationships
studies
Declaration of competing interest
In order to improve understanding on the way of
b
-carboline
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
derivatives action in M. tuberculosis, structure-activity relationship
(SAR) studies were carried out using the three and four-
dimensional approaches. Thus, three-dimensional structures were
built in HyperChem 7 (Hyper Co.), and calculations of B3LYP/
6e311þþg (d,p) approach were carried in Gaussian 09 (Gaussian
Inc.). Using these structures, field descriptors of each derivative
were obtained in Pentacle (Molecular Discovery Ltd), which uses
the GRIND approach for calculating 3D field descriptors in an
alignment-independent way. The used descriptors were computed
using the AMANDA/MACC2 algorithms [26e28,37].
Acknowledgements
ꢀ
ꢀ
We are grateful to Laboratorio de Bacteriologia Medica,
ꢀ
ꢀ
belonging to Laboratorio de Ensino e Pesquisa em Analises Clínicas
~
(LEPAC), Coordenaçao de Aperfeiçoamento de Pessoal de Nível
Superior (CAPES), Conselho Nacional de Desenvolvimento Cientí-
ꢀ
~
ꢀ
ꢀ
fico e Tecnologico (CNPq), and Fundaçao Araucaria (State of Parana,
A qualitative SAR study based in the approach used by Ermondi
Brazil).
et al. [38] was carried out. Compounds with MIC >250
defined as inactives, and were set as ꢁ1; on the other hand, de-
rivatives showing MIC ꢀ250 g/mL were set as actives, 1. The field
mg/mL were
Appendix A. Supplementary data
m
descriptors used were based in distances at angstroms between
different regions of a molecule: hydrophobic-hydrophobic groups
(DRY-DRY), hydrogen bond acceptor-hydrogen bond acceptor
groups (OeO), hydrogen bond donor-hydrogen bond donor groups
(N1eN1), shape-shape groups (TIP-TIP), and the combinations be-
tween these (DRY-O, DRY-N1, DRY-TIP, OeN1, O-TIP, and N1-TIP).
The set of obtained descriptors was reduced using the Fractional
Factorial Design (FDD) variable selection, available in the same
program [26e28,37]. This selection was carried out by the
consecutive construction of regression models by Partial Least
Squares (PLS) with a smaller number of descriptors, but with
improved statistical quality. Next, the selected descriptors were
exported and a new selection of variables was performed using the
Ordered Predictors Selection (OPS) [39] method, implemented in
the QSAR Modeling [40] program, in order to obtain a model with a
lower number of descriptors, but more significant and easier to
interpret.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2019.111935.
References
[1] World Health Organization, Global tuberculosis report. https://www.who.int/
tb/publications/global_report/gtbr2019_main_text.pdf, 2019.
[2] G.F. dos, S. Fernandes, P.C. de Souza, L.B. Marino, K. Chegaev, S. Guglielmo,
L. Lazzarato, R. Fruttero, M.C. Chung, F.R. Pavan, J.L. dos Santos, Synthesis and
biological activity of furoxan derivatives against Mycobacterium tuberculosis,
Eur. J. Med. Chem. 123 (2016) 523e531, https://doi.org/10.1016/
j.ejmech.2016.07.039.
[3] R. Cao, W. Peng, Z. Wang, A. Xu, beta-Carboline alkaloids: biochemical and
pharmacological functions, Curr. Med. Chem. 14 (2007) 479e500, https://
doi.org/10.2174/092986707779940998.
[4] S. Wu, Y. Fu, R. Yan, Y. Wu, X. Lei, X.-S. Ye, Synthesis of neamineecarboline
conjugates for RNA binding and their antibacterial activities, Tetrahedron 66
(2010) 3433e3440, https://doi.org/10.1016/j.tet.2010.03.034.
[5] G. Nenaah, Antibacterial and antifungal activities of (beta)-carboline alkaloids
of Peganum harmala (L) seeds and their combination effects, Fitoterapia 81
(2010) 779e782, https://doi.org/10.1016/j.fitote.2010.04.004.
Using the activities as defined for the 3D-QSAR study, a 4D-
QSAR study was carried out by LQTA-QSAR methodology [41],
employing the free software web-4D-QSAR [42]. This approach was
based on the generation of a conformational ensemble profile for
each compound, followed by the calculation of 3D descriptors,
which were an average of the interaction experienced by a probe,
NHþ3 in this study, in each grid point for all conformations. Coulomb
and Lennard-Jones interactions were calculated for each derivative
tested. The matrix with the descriptors generated by LQTAgrid
module in web-4D-QSAR was treated using digital filters for mo-
lecular field interactions [43] followed by a variable reduction with
the elimination of descriptors with the absolute value of the cor-
relation coefficient between the descriptor and biological activity
lower than 0.3. Then, the resulting matrix was also submitted to the
OPS method [39].
The quality of the obtained models was assessed on its coeffi-
cient of determination (R2), the root mean square error of calibra-
tion (RMSEC), the F-test, the coefficient of determination of cross-
validation (Q2LOO), and the root mean square error of cross-
validation (RMSECV). In all steps of variable selection and con-
struction of models, descriptors were autoscaled, the most rec-
ommended data pre-processing approach for QSAR studies [44].
[6] A.S. Nazari Formagio, P.R. Santos, K. Zanoli, T. Ueda-Nakamura, L.T. Dusman
Tonin, C. V Nakamura, M.H. Sarragiotto, Synthesis and antiviral activity of
beta-carboline derivatives bearing a substituted carbohydrazide at C-3 against
poliovirus and herpes simplex virus (HSV-1), Eur. J. Med. Chem. 44 (2009)
4695e4701, https://doi.org/10.1016/j.ejmech.2009.07.005.
[7] R. Kumar, S. Khan, A. Verma, S. Srivastava, P. Viswakarma, S. Gupta, S. Meena,
N. Singh, J. Sarkar, P.M.S. Chauhan, Synthesis of 2-(pyrimidin-2-yl)-1-phenyl-
2,3,4,9-tetrahydro-1H-beta-carbolines as antileishmanial agents, Eur. J. Med.
Chem. 45 (2010) 3274e3280, https://doi.org/10.1016/j.ejmech.2010.04.004.
[8] R.H. Valdez, L.T.D. Tonin, T. Ueda-Nakamura, B.P. Dias Filho, J.A. Morgado-Diaz,
M.H. Sarragiotto, C.V. Nakamura, Biological activity of 1,2,3,4-tetrahydro-beta-
carboline-3-carboxamides against Trypanosoma cruzi, Acta Trop. 110 (2009)
7e14, https://doi.org/10.1016/j.actatropica.2008.11.008.
[9] A.S.N. Formagio, L.T.D. Tonin, M.A. Foglio, C. Madjarof, J.E. de Carvalho, W.F. da
Costa, F.P. Cardoso, M.H. Sarragiotto, Synthesis and antitumoral activity of
novel 3-(2-substituted-1,3,4-oxadiazol-5-yl) and 3-(5-substituted-1,2,4-
triazol-3-yl) beta-carboline derivatives, Bioorg. Med. Chem. 16 (2008)
9660e9667, https://doi.org/10.1016/j.bmc.2008.10.008.
[10] B. Baruah, K. Dasu, B. Vaitilingam, P. Mamnoor, P.P. Venkata, S. Rajagopal,
K.R. Yeleswarapu, Synthesis and cytotoxic activity of novel quinazolino-b-
carboline-5-one derivatives, Bioorg. Med. Chem. 12 (2004) 1991e1994,
https://doi.org/10.1016/j.bmc.2004.03.005.
[11] A.E. Wahba, J. Peng, S. Kudrimoti, B.L. Tekwani, M.T. Hamann, Structure-ac-
tivity relationship studies of manzamine A: amidation of positions 6 and 8 of
the beta-carboline moiety, Bioorg. Med. Chem. 17 (2009) 7775e7782, https://
doi.org/10.1016/j.bmc.2009.09.026.
[12] S. Bonazzi, D. Barbaras, L. Patiny, R. Scopelliti, P. Schneider, S.T. Cole, M. Kaiser,
R. Brun, K. Gademann, Antimalarial and antitubercular nostocarboline and
eudistomin derivatives: synthesis, in vitro and in vivo biological evaluation,
Bioorg. Med. Chem. 18 (2010) 1464e1476, https://doi.org/10.1016/
j.bmc.2010.01.013.
[13] K. V Rao, N. Kasanah, S. Wahyuono, B.L. Tekwani, R.F. Schinazi, M.T. Hamann,
Three new manzamine alkaloids from a common Indonesian sponge and their
activity against infectious and tropical parasitic diseases, J. Nat. Prod. 67
(2004) 1314e1318, https://doi.org/10.1021/np0400095.
Funding
No funding.
[14] A. Penta, S. Franzblau, B. Wan, S. Murugesan, Design, synthesis and evaluation
of diarylpiperazine derivatives as potent anti-tubercular agents, Eur. J. Med.
Chem. 105 (2015) 238e244, https://doi.org/10.1016/j.ejmech.2015.10.024.
[15] F.M.F. Moreira, J. Croda, M.H. Sarragiotto, M.A. Foglio, A.L.T.G. Ruiz,
J.E. Carvalho, A.S.N. Formagio, Synthesis, in vitro antiproliferative and Anti-
Ethical approval
Not required.

M.A. Lopes-Ortiz et al. / European Journal of Medicinal Chemistry 187 (2020) 111935
11
Mycobacterium tuberculosis activities of novel
Chem. Soc. 27 (2016) 1398e1405, https://doi.org/10.5935/0103-
5053.20160062.
b
-Carboline Derivatives, J. Braz.
Z. Liu, Z. Yuan, Synthesis, 3D-QSAR analysis and biological evaluation of qui-
noxaline 1,4-di-N-oxide derivatives as antituberculosis agents, Bioorg. Med.
Chem. Lett 26 (2016) 4146e4153, https://doi.org/10.1016/j.bmcl.2016.01.066.
[31] P. Shah, M. Saquib, S. Sharma, I. Husain, S.K. Sharma, V. Singh, R. Srivastava,
A.K. Shaw, M.I. Siddiqi, 3D-QSAR and molecular modeling studies on 2,3-
dideoxy hexenopyranosid-4-uloses as anti-tubercular agents targeting
alpha-mannosidase, Bioorg. Chem. 59 (2015) 91e96, https://doi.org/10.1016/
j.bioorg.2015.02.001.
[16] H. Akgun, I. Karamelekoglu, B. Berk, I. Kurnaz, G. Saribiyik, S. Oktem,
T. Kocagoz, Synthesis and antimycobacterial activity of some phthalimide
derivatives, Bioorg. Med. Chem. 20 (2012) 4149e4154, https://doi.org/
10.1016/j.bmc.2012.04.060.
[17] P.S. Phatak, R.D. Bakale, S.T. Dhumal, L.K. Dahiwade, P.B. Choudhari, V. Siva
Krishna, D. Sriram, K.P. Haval, Synthesis, antitubercular evaluation and mo-
lecular docking studies of phthalimide bearing 1,2,3-triazoles, Synth. Com-
[32] S.R. Patel, R. Gangwal, A.T. Sangamwar, R. Jain, Synthesis, biological evaluation
and 3D-QSAR study of hydrazide, semicarbazide and thiosemicarbazide de-
rivatives of 4-(adamantan-1-yl)quinoline as anti-tuberculosis agents, Eur. J.
mun.
49
(2019)
2017e2028,
https://doi.org/10.1080/
00397911.2019.1614630.
Med.
Chem.
85
(2014)
255e267,
https://doi.org/10.1016/
[18] A. Rani, A. Viljoen, M.D. Johansen, L. Kremer, V. Kumar, Synthesis, anti-
mycobacterial and cytotoxic evaluation of substituted isoindoline-1,3-dione-
j.ejmech.2014.07.100.
[33] A. Wang, Y. Yang, Y. Jun, B. Wang, K. Lv, M. Liu, H. Guo, Y. Lu, Synthesis,
evaluation and CoMFA/CoMSIA study of nitrofuranyl methyl N-heterocycles
as novel antitubercular agents, Bioorg. Med. Chem. 26 (2018) 2073e2084,
https://doi.org/10.1016/j.bmc.2018.03.004.
[34] H.O. Molhuizen, A.E. Bunschoten, L.M. Schouls, J.D. van Embden, Rapid
detection and simultaneous strain differentiation of Mycobacterium tubercu-
losis complex bacteria by spoligotyping, Methods Mol. Biol. 101 (1998)
381e394, https://doi.org/10.1385/0-89603-471-2:381.
[35] P. Supply, C. Allix, S. Lesjean, M. Cardoso-Oelemann, S. Rüsch-Gerdes,
E. Willery, E. Savine, P. de Haas, H. van Deutekom, S. Roring, P. Bifani,
N. Kurepina, B. Kreiswirth, C. Sola, N. Rastogi, V. Vatin, M.C. Gutierrez,
M. Fauville, S. Niemann, R. Skuce, K. Kremer, C. Locht, D. van Soolingen,
Proposal for standardization of optimized mycobacterial interspersed repet-
itive unit-variable-number Tandem Repeat typing of Mycobacterium tuber-
culosis, J. Clin. Microbiol. 44 (2006) 4498, https://doi.org/10.1128/JCM.01392-
06. LP e 4510.
4-aminoquinolines coupled via alkyl/amide linkers, RSC Adv.
9 (2019)
8515e8528, https://doi.org/10.1039/C8RA10532D.
[19] T. Matviiuk, F. Rodriguez, N. Saffon, S. Mallet-Ladeira, M. Gorichko, A.L. de
Jesus Lopes Ribeiro, M.R. Pasca, C. Lherbet, Z. Voitenko, M. Baltas, Design,
chemical synthesis of 3-(9H-fluoren-9-yl)pyrrolidine-2,5-dione derivatives
and biological activity against enoyl-ACP reductase (InhA) and Mycobacterium
tuberculosis, Eur. J. Med. Chem. 70 (2013) 37e48, https://doi.org/10.1016/
j.ejmech.2013.09.041.
[20] J.L. Santos, P.R. Yamasaki, C.M. Chin, C.H. Takashi, F.R. Pavan, C.Q.F. Leite,
Synthesis and in vitro anti Mycobacterium tuberculosis activity of a series of
phthalimide derivatives, Bioorg. Med. Chem. 17 (2009) 3795e3799, https://
doi.org/10.1016/j.bmc.2009.04.042.
[21] S.A.F. Rostom, Polysubstituted pyrazoles, part 6. Synthesis of some 1-(4-
chlorophenyl)-4-hydroxy-1H-pyrazol-3-carbonyl derivatives linked to
nitrogenous heterocyclic ring systems as potential antitumor agents, Bioorg.
Med.
j.bmc.2010.02.006.
Chem.
18
(2010)
2767e2776,
https://doi.org/10.1016/
[36] J. Palomino, A. Martin, M. Camacho, H. Guerra, J. Swings, F. Portaels, Resazurin
microtiter assay plate: simple and inexpensive method for detection of drug
resistance in Mycobacterium tuberculosis, Antimicrob. Agents Chemother. 46
(2002) 2720e2722, https://doi.org/10.1128/AAC.46.8.2720.
[37] A. Duran, G.C. Martinez, M. Pastor, Development and validation of AMANDA, a
new algorithm for selecting highly relevant regions in Molecular Interaction
Fields, J. Chem. Inf. Model. 48 (2008) 1813e1823, https://doi.org/10.1021/
ci800037t.
ꢀ
[22] N. Brosse, M.-F. Pinto, B. Jamart-Gregoire, Preparation of multiply protected
alkylhydrazine derivatives by Mitsunobu and PTC approaches, Eur. J. Org.
Chem. 2003 (2003) 4757e4764, https://doi.org/10.1002/ejoc.200300445.
[23] J. Peng, J.-F. Hu, A.B. Kazi, Z. Li, M. Avery, O. Peraud, R.T. Hill, S.G. Franzblau,
F. Zhang, R.F. Schinazi, S.S. Wirtz, P. Tharnish, M. Kelly, S. Wahyuono,
M.T. Hamann, Manadomanzamines A and B: A novel alkaloid ring system
with potent activity against mycobacteria and HIV-1, J. Am. Chem. Soc. 125
(2003) 13382e13386, https://doi.org/10.1021/ja030087z.
[24] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Experimental and
computational approaches to estimate solubility and permeability in drug
discovery and development settings, Adv. Drug Deliv. Rev. 46 (2001) 3e26,
https://doi.org/10.1016/S0169-409X(00)00129-0.
[25] G.F. Dos Santos Fernandes, D.H. Jornada, P.C. de Souza, C.M. Chin, F.R. Pavan,
J.L. Dos Santos, Current advances in antitubercular drug discovery: potent
prototypes and new targets, Curr. Med. Chem. 22 (2015) 3133e3161, https://
doi.org/10.2174/0929867322666150818103836.
[26] J.B. Reigada, C.M. Tcacenco, L.H. Andrade, M.J. Kato, A.L.M. Porto, J.H.G. Lago,
Chemical constituents from Piper marginatum Jacq. (Piperaceae) - antifungal
activities and kinetic resolution of (RS)-marginatumol by Candida Antarctica
lipase, Tetrahedron Asymmetry 18 (2007) 1054e1058.
[38] G. Ermondi, G. Caron, I.G. Pintos, M. Gerbaldo, M. Perez, D.I. Perez, Z. Gandara,
A. Martinez, G. Gomez, Y. Fall, An application of two MIFs-based tools
(Volsurfþ and Pentacle) to binary QSAR: the case of a palinurin-related data
set of non-ATP competitive glycogen synthase kinase 3beta (GSK-3beta) in-
hibitors, Eur. J. Med. Chem. 46 (2011) 860e869, https://doi.org/10.1016/
j.ejmech.2010.12.024.
ꢀ
[39] R.F. Teofilo, J.P.A. Martins, M.M.C. Ferreira, Sorting variables by using infor-
mative vectors as a strategy for feature selection in multivariate regression,
J. Chemom. 23 (2009) 32e48, https://doi.org/10.1002/cem.1192.
[40] J.P.A. Martins, M.M.C. Ferreira, QSAR modeling: um novo pacote computa-
cional open source para gerar e validar modelos QSAR, Quim. Nova 36 (2013)
554e560, https://doi.org/10.1590/S0100-40422013000400013.
[41] J.P.A. Martins, E.G. Barbosa, K.F.M. Pasqualoto, M.M.C. Ferreira, LQTA-QSAR: a
new 4D-QSAR methodology, J. Chem. Inf. Model. 49 (2009) 1428e1436,
https://doi.org/10.1021/ci900014f.
[42] J.P.A. Martins, M.A. Rougeth de Oliveira, M.S. Oliveira de Queiroz, Web-4D-
QSAR, A web-based application to generate 4D-QSAR descriptors, J. Comput.
Chem. 39 (2018) 917e924, https://doi.org/10.1002/jcc.25166.
[43] E.G. Barbosa, M.M.C. Ferreira, Digital filters for molecular interaction field
descriptors, Mol. Inf. 31 (2012) 75e84, https://doi.org/10.1002/
minf.201000181.
[27] P.P. Roy, K. Roy, On some aspects of variable selection for partial least squares
regression models, QSAR Comb. Sci. 27 (2008) 302e313, https://doi.org/
10.1002/qsar.200710043.
[28] L.T.F.M. Camargo, M.M. Sena, A.J. Camargo, A quantum chemical and che-
mometrical study of indolo[2,1-b]quinazoline and their analogues with
cytotoxic activity against breast cancer cells, SAR QSAR Environ. Res. 20
(2009) 537e549, https://doi.org/10.1080/10629360903278800.
[29] S. Askjaer, M. Langgård, Combining pharmacophore fingerprints and PLS-
discriminant analysis for virtual screening and SAR elucidation, J. Chem. Inf.
Model. 48 (2008) 476e488, https://doi.org/10.1021/ci700356w.
[44] R. Kiralj, M.M.C. Ferreira, Basic validation procedures for regression models in
QSAR and QSPR studies: theory and application, J. Braz. Chem. Soc. 20 (2009)
770e787, https://doi.org/10.1590/S0103-50532009000400021.
[30] Y. Pan, P. Li, S. Xie, Y. Tao, D. Chen, M. Dai, H. Hao, L. Huang, Y. Wang, L. Wang,