Synthesis, in vitro cytotoxic, anti-Mycobacterium tuberculosis and molecular docking studies of 4-pyridylamino- and 4-(ethynylpyridine)quinazolines

Article abstract of DOI:10.1016/j.molstruc.2021.130824

A series of 4-(pyridylamino)- and 4-(ethynylpyridine)quinazolines were successfully prepared via Sonogashira cross-coupling and dechloroamination reactions on the C(4)-Cl position of the requisite 2-(p-phenyl)-4-chloroquinazolines. The prepared compounds were characterized by means of ¹H- and ¹³C-NMR, FT-IR and mass spectrometry techniques. The structure of 2-(4-chlorophenyl)-4-(2-(pyridin-4-yl) ethynyl) quinazoline from the 4-(ethynylpyridine) series was confirmed by single crystal X-ray analysis which indicates monoclinic crystal system and P21/c space group. Compounds from the 4-chloro-, 4-(pyridylamino)- and 4-(ethynylpyridine)-quinazoline series were evaluated for anti-Mycobacterium tuberculosis (Mtb) properties in vitro employing rifampicin as a reference drug. The results from the Alamar Blue assay (Mtb H37Rv strain) revealed promising MIC₉₀ ranging from <0.7 to >125 μM. The cytotoxicity of the synthesised compounds was tested against the Raw 264.7 microphage cell line at a maximum concentration of 50 μM. The possible mode of interaction against the Mtb was theoretically explained through molecular 3ZXR protein and the more prominent hydrogen bond is observed between the nitrogen of the pyridine ring moiety of the 5 and 6 series with OH group of SER280. Also, a metal coordination between the methoxy benzene moiety of compound 6e and Mg²⁺ is also observed, explaining the SAR of these compounds to MtGS.

Full text of DOI:10.1016/j.molstruc.2021.130824

Journal Pre-proof
Synthesis, In Vitro Cytotoxic, Anti-Mycobacterium tuberculosis and
Molecular Docking Studies of 4-Pyridylamino- and
4-(Ethynylpyridine)quinazolines
Kabelo B. Dilebo , Njabulo J. Gumede , Winston Nxumalo ,
Thabe M. Matsebatlela , Dikgale Mangokoana ,
Ngaoko R. Moraone , Bernard Omondi , Richard M. Mampa
PII:
S0022-2860(21)00957-1
DOI:
Reference:
https://doi.org/10.1016/j.molstruc.2021.130824
MOLSTR 130824
To appear in:
Journal of Molecular Structure
Received date:
Revised date:
Accepted date:
30 October 2020
10 May 2021
30 May 2021
Please cite this article as: Kabelo B. Dilebo , Njabulo J. Gumede , Winston Nxumalo ,
Thabe M. Matsebatlela , Dikgale Mangokoana , Ngaoko R. Moraone , Bernard Omondi ,
Richard M. Mampa , Synthesis, In Vitro Cytotoxic, Anti-Mycobacterium tuberculosis and Molecu-
lar Docking Studies of 4-Pyridylamino- and 4-(Ethynylpyridine)quinazolines, Journal of Molecular
Structure (2021), doi: https://doi.org/10.1016/j.molstruc.2021.130824
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2021 Published by Elsevier B.V.

1
Highlights:
 Successful synthesis of compounds of the series 4-chloro-, 4-(pyridylamino)- and 4-
2
3
4
5
6
7
8
9
0
(ethynylpyridine)-quinazoline,
 Positive results from the Alamar Blue assay (Mtb H37Rv strain) which revealed
promising MIC ranging from <0.7 to >125 µM and
9
0
 Candidate compounds were inductively docked into the 3ZXR enzyme and the more
prominent hydrogen bond is between Nitrogen of the pyridine ring moiety of the 5 and 6
series with an OH group of SER280.
 A metal coordination between the methoxy benzene moiety of compounds in the 6 series
2
+
1
1
and Mg is also observed, explaining the SAR of these compounds to MtGS enzyme.
1

1
1
2
3
1
1
4
5
Synthesis, In Vitro Cytotoxic, Anti-Mycobacterium tuberculosis and Molecular Docking
Studies of 4-Pyridylamino- and 4-(Ethynylpyridine)quinazolines
1
2
1
3
1
1
1
6
7
8
Kabelo B. Dilebo , Njabulo J. Gumede , Winston Nxumalo , Thabe M. Matsebatlela ,
3
1
4
Dikgale Mangokoana , Ngaoko R. Moraone , Bernard Omondi and Richard M.
1
,
Mampa *
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
Department of Chemistry, University of Limpopo, Private Bag X1106, Sovenga 0727, South
Africa; kbsebashe@gmail.com (K.B.D.); winston.nxumalo@ul.ac.za (W.N.);
ngoakohitrj@gmail.com (N.R.M)
2
3
Department of Chemistry, Mangosuthu University of Technology, Durban, P.O. Box 12363,
Jacobs 4026, South Africa; Ngumede@mut.ac.za
Department of Biochemistry, Microbiology and Biotechnology, School of molecular and Life
Sciences, University of Limpopo, Private Bag X1106, Sovenga 0727, South Africa;
thabem@ul.ac.za (T.M.M.); dikgalemangokoana@gmail.com (D.M.)
School of Chemistry and Physics, University of KwaZulu-Natal, Pietermaritzburg Campus,
Private Bag X01, Pietermaritzburg 3209, South Africa; Owaga@ukzn.ac.za
4
* Correspondence: richard.mampa@ul.ac.za; Tel.: +27 (015) 268 2334

3
4
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
Abstract: A series of 4-(pyridylamino)- and 4-(ethynylpyridine)quinazolines were successfully
prepared via Sonogashira cross-coupling and dechloroamination reactions on the C(4)-Cl
position of the requisite 2-(p-phenyl)-4-chloroquinazolines. The prepared compounds were
1
13
characterized by means of H- and C-NMR, FT-IR and mass spectrometry techniques. The
structure of 2-(4-chlorophenyl)-4-(2-(pyridin-4-yl)ethynyl)quinazoline from the 4-
(ethynylpyridine) series was confirmed by single crystal X-Ray analysis which indicates
monoclinic crystal system and P21/c space group. Compounds from the 4-chloro-, 4-
(pyridylamino)- and 4-(ethynylpyridine)-quinazoline series were evaluated for anti-
Mycobacterium tuberculosis (Mtb) properties in vitro employing rifampicin as a reference drug.
The results from the Alamar Blue assay (Mtb H37Rv strain) revealed promising MIC ranging
9
0
from <0.7 to >125 µM. The cytotoxicity of the synthesised compounds was tested against the
Raw 264.7 microphage cell line at a maximum concentration of 50 µM. The possible mode of
interaction against the Mtb was theoretically explained through molecular 3ZXR protein and the
more prominent hydrogen bond is observed between the nitrogen of the pyridine ring moiety of
the 5 and 6 series with OH group of SER280. Also, a metal coordination between the methoxy
2
+
benzene moiety of compound 6e and Mg is also observed, explaining the SAR of these
compounds to MtGS.
5
5
2
3
Keywords: Mycobacterium tuberculosis; quinazoline; glutamine synthetase; molecular
docking; NMR; Sonogashira cross-coupling
5
5
4
5

5
5
6
7
1. Introduction
5
5
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
Tuberculosis (TB) is an infectious airborne disease that is caused by Mycobacterium
tuberculosis [1]. According to the World Health Organisation (WHO), TB remains one of the
most lethal diseases, with an emergence of about 10.4 million cases and 1.3 million deaths each
year [2]. The advent of drug resistant TB strains and coexistence with Human Immunodeficiency
Virus (HIV) worsens the situation [2]. Generally, treatment takes 6 to 9 months of
chemotherapeutic regimen which includes a high dose of first-line drugs, isoniazid, ethambutol,
rifampicin and pyrazinamide [2]. Failure to complete this treatment has led to the manifestation
of multi-drug resistant (MDR) strains of the TB disease [2]. This resistance has undermined the
chemotherapeutic effectiveness of the present drug regimen. The continued spread of TB is
worrying despite 100% treatment success rate for normal TB strain followed by 50% of the
MDR and 10% rate of extensively-drug resistant (XDR) strains amid extended period (up to 2
years) of treatment [2].
To this extent, TB drug candidates containing a quinazoline framework have been identified
as potential building blocks for the development of new drugs owing to their wide range of
biological properties [3]. Numerous quinazoline derivatives are known to exhibit anticancer,
antimalarial, antiviral, anti-inflammatory and antibacterial properties [4–7]. Some of these
quinazoline derivatives have been reported to inhibit the proliferation of MDR and XDR TB
strains [8]. The 2-(3-Fluorophenyl)-4-oxoquinazolin-3(4H)-yl-3-fluorobenzoate (A) shown in
Figure 1 for example, was found to inhibit several clinical isolates of MDR and XDR TB by
exhibiting MIC values of 6.5 – 6.6 µM. In addition, enzyme kinetics and molecular docking
9
0
studies suggested that the compound is capable of binding branched-chain amino acids (BCAAs)

7
8
8
8
8
8
8
8
8
8
8
9
9
9
9
0
1
2
3
4
5
6
7
8
9
0
1
2
biosynthetic pathway promoter’s i.e. acetohydroxyacid synthase (AHAS) and serine-threonine
protein kinases (STPKs) of glutamine synthetase [9, 10]. Glutamine synthetase, which is an
enzyme responsible for the catalysis of ATP-dependent condensation of ammonium and L-
glutamate at the amino acid binding site resulting in the formation of glnA1, is essential for the
growth of Mtb in both in vitro and in vivo [10]. A disturbance at the amino acid binding site stops
cell wall biosynthesis leading to cell death [10]. Likewise, Gising and coworkers designed and
synthesised trisubstituted imidazole (B) derivatives represented in Figure 1 bearing heterocyclic
and pyridyl substituents which were found to bind and inhibit the activity of glutamine
synthetase with IC values of 2.2 ± 0.3 µM to >25 µM [11]. Similarly, Nordqvist and coworkers
5
0
prepared imidazo[1,2-a]pyridine-based derivatives (C) (Figure 1) which were found to exhibit
IC50 of 3.0 ± 0.1 µM to >25 µM against glutamine synthetase [12]. Based on this literature
precedence, it was envisaged that the 4-Pyridylamino- and 4-(Ethynylpyridine)quinazolines
would also represent additional agents which may target and inhibit the activity of glutamine
synthetase.
R
N
O
O
N
F
R1
NH
HN
N
O
N
Br
N
N
F
R2
MeO
9
9
3
4
A
B
C
9
5
Figure 1: Structures of potential inhibitors of glutamine synthetase of Mycobacterium
tuberculosis
9
9
9
6
7
8
In this study, we focus on the synthesis and functionalisation of prerequisite 4-chloro-
quinazolines on the C(4) position to generate novel pyridylquinazoline derivatives. In addition,

9
9
in vitro anti-Mtb screening and cytotoxicity studies of the newly synthesised compounds were
performed and complemented by computational molecular docking studies to evaluate the
possible binding to the target enzyme, glutamine synthetase (PDB: 1HTO, 3ZXR, and 3ZXV).
1
1
00
01
1
1
02
03
2. Results
2.1. Chemistry
1
1
04
05
2.1.1. Synthesis of 2-Aryl-quinazolin-4(3H)-ones 3a–3e and 2-Aryl-4-chloroquinazolines
4a–4e
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
The first step involves the synthesis of 2-aryl-quinazolin-4(3H)-ones 3a–3e to be used as
substrates for the requisite 2-aryl-4-chloroquinazolines 4a–4e, which is reported in the literature
[13–16], Oxidative one-pot cyclocondensation reaction of anthranilamide 1a with para-
substituted benzaldehyde derivatives 2a–2e in the presence of molecular iodine (2 equiv.) in
ethanol at 80 °C led to the desired 2-aryl-quinazolin-4(3H)-ones 3a–3e as indicated in Scheme 1.
The 2-aryl-quinazolin-4(3H)-ones 3a–3e, were obtained in high yields (greater than 86%, Table
1
13
1) following a modified literature method [16] and characterised using H-NMR, C-NMR and
1
FT-IR spectroscopic techniques. The H-NMR spectra of compounds 3a–3e revealed the
presence of a broad singlet at 12.75 ppm corresponding to the N-H proton of the α-nitrogen.
Furthermore, protons in the aromatic region ranging from 7.53–8.20 ppm are assigned to the 2-
aryl substituent and the quinazoline backbone. The amide nature of the compounds was
confirmed by FT-IR spectroscopy due to the presence of C-N and C=O in the regions vmax 1531–
−
1
−1
1583 cm and v 1656–1691 cm , respectively.
max
The next step was to generate an electrophilic centre at the C4-position of the 2-aryl-
quinazolin-4(3H)-ones scaffolds 3a–3e through oxidative aromatisation to afford 2-aryl-4-

1
1
1
1
1
1
1
21
22
23
24
25
26
27
chloroquinazolines 4a–4e (Table 1). The 2-aryl-quinazolin-4(3H)-ones 3a–3e in DMF solution
was treated with excess thionyl chloride following a literature method to afford 2-aryl-4-chloro-
quinazolines 4a–4e [16]. The reaction was driven to completion by refluxing at 80 °C for 2 h.
The full conversion of the starting materials was confirmed by the absence of an N-H signal from
1
the H-NMR spectra of the 2-aryl-4-chloro-quinazolines 4a–4e, also indicated in Scheme 1. The
FT-IR spectra of compounds 4a–4e revealed the absence of the C=O stretch, thus confirming
successful formation of the C(4)-Cl bond.
O
O
O
Cl
NH2
(i)
NH
(ii)
N
+
NH2
N
N
R
R
R
1
a
4a-4e
2
a-2e
3a-3e
1
28
1
1
1
29
30
31
Scheme 1. Synthesis of 2-aryl-quinazolin-4(3H)-ones 3a–3e and 2-aryl-4-chloro-
quinazolines 4a–4e. Reagents and conditions: (i) I , 80 °C, EtOH, 6 h; (ii) SOCl , DMF,
2
2
80 °C, 2 h. R = -Cl, -Br, -F, -NO and -OMe
2
1
32
Table 1. Yields of compounds 3a–3e [13] and 4a–4d [14], 4e [15].
Comp %Yiel Comp %Yiel
ound d(R) ound d(R)
7 (- 88 (-
Cl) Cl)
93 (-
Br)
90 (-F)
8
3
a
4a
91 (-
3b
4b
4c
Br)
3
c
93 (-F)

9
0 (-
84 (-
NO2)
89 (-
OMe)
3d
4d
4e
NO2)
93 (-
3e
OMe)
1
33
2.1.2. Amination of 4a–4e with 2-Amino-3-nitropyridine
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
The synthesis of 2-aryl-N-(3-nitropyridin-2-yl)quinazolin-4-amine 5a–5e was achieved
through a slight modification of a literature procedure [17]. The Paumo et al. [17] procedure
employs a combination of THF-i-PrOH and HCl which protonates the N-1 position of the 4-
chloroquinazolines and renders the heterocyclic ring electron deficient and C-4 more
electrophilic. Based on this observation, we employed 2-amino-3-nitropyridine and the 2-aryl-4-
chloroquinazolines following the same conditions as in the literature procedure, unfortunately
only trace amounts of the desired product were achieved. The reason for the low yield was
attributed to the poor solubility of 2-amino-3-nitropyridine in the solvent mixture used, i.e. THF-
i-PrOH. Therefore, in order to improve the solubility, a mixture of THF-DMF (3:1) with 98%
sulfuric acid serving as a catalyst was applied to afford 2-aryl-N-(3-nitropyridin-2-yl)quinazolin-
4-amine derivatives 5a–5e (Scheme 2) in high yields (Table 2). The successful synthesis of
1
13
compounds 5a–5e was confirmed by H-NMR, C-NMR, FT-IR spectroscopy and HRMS
spectroscopic techniques, all included in the Supplementary Material as Figures s1 to s15. The
1
H-NMR spectra showed the appearance of an N-H signal at around 12.60 ppm, with protons
from the nitropyridyl ring showing multiplicities of doublet of doublets, doublet and a multiplet
1
3
ranging from 6.73 ppm to 8.40 ppm in the aromatic region. The C-NMR spectra showed the
quaternary carbon of the nitro-deshielded group resonating at around 149.00 ppm. The presence

−
1
1
1
51
52
of the N-H group was confirmed by a broad N-H stretch in the region 3461–3464 cm on FT-IR
spectra.
O N
2
N
Cl
N
HN
N
(i)
N
N
R
R
5a-5e
4a-4e
1
53
1
1
54
55
Scheme 2. Amination of 4a–4e with 2-amino-3-nitropyridine. Reagents and conditions:
(i) 2-NH -3-NO C H N, 98% sulfuric acid, 65 °C, 5 h.
2
2
5
3
1
56
Table 2. Yields of compounds 5a–5e.
Comp
ound
%
R
Yield
-
5
a
95
78
92
66
84
Cl
Br
F
-
-
-
5b
5
c
5d
NO2
O
5
e
Me

1
57
2.1.3. Sonogashira-Cross Coupling of 4a–4e
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
The first attempt following Sonogashira cross-coupling conditions of the 2-aryl-4-chloro-
quinazolines 4a–4e at the C(4)-Cl position using Pd(PPh ) , CuI, Et N and 1.5 equiv. of 4-
3
4
3
ethynyl-pyridine hydrochloride in dry THF at mild temperatures (40–60 °C) was unsuccessful,
and only the starting materials (4-ethynylpyridine and 4-chloroquinazolines) were isolated. Air
and moisture sensitivity of the Pd(0) source, tetrakis(triphenylphosphine)palladium(0) was
attributed as contributing to the failure of Sonogashira cross-coupling in this case. In the second
attempt, the Pd(0) source was replaced with PdCl (PPh ) whilst keeping other reagents constant
2
3 2
but this procedure also led to the recovery of starting materials. Further attempts in high boiling
solvents such as DMF and dioxane also proved unsuccessful. Generally, the Sonogashira
reaction requires a nucleophilic terminal alkyne which reacts with a CuX (X = Br, I) salt and
successively gives a copper acetylide [18]. The failure of the second attempt using Et N, was
3
attributed to the basic conditions not being favourable for deprotonating the hydrochlorinated 4-
ethynlpyridine substrate. Consequently, trace amounts of our desired product was obtained when
using K CO , Na CO and NaOH as our base during coupling. Since the Sonogashira cross-
2
3
2
3
coupling reaction mechanism requires a relatively strong base to deprotonate the terminal
acetylene, we then opted for Cs CO [16]. In a one-pot reaction synthesis, 4-chloro-2-(4-
2
3
chlorophenyl)quinazoline (4a, 1 equiv.) was reacted with PdCl (PPh ) (10%), CuI (10%) and
2
3 2
Cs CO (2 equiv.) in 3:1 THF-water and the reaction was left to stir for 24 h under at mild heat
2
3
(40–60 °C) under N to produce a 98% yield of 6a (Table 3) after recrystalisation from diethyl
2
ether, as indicated in Scheme 3. Synthesis of derivatives 6b–6e followed that of 6a while varying
the para-position of the 2-aryl group (Table 3). Successful synthesis of 6a–6e was confirmed
1
13
using H, C-NMR, FT-IR spectroscopy and HRMS spectroscopic techniques (see Figures s16

1
1
1
1
1
1
1
1
80
81
82
83
84
85
86
to s30 in the Supplementary Information). The H-NMR showed two additional doublet peaks
1
3
resonating at 7.60 and 8.74 ppm. The C-NMR spectra showed carbons of the π-spacer (C≡C)
resonating at 88.22 and 93.71 ppm. Further characterisation using FT-IR spectroscopy showed
−
1
the presence of Csp-Csp band at vmax 2219.44 cm . The HRMS (ESI+) revealed m/z M+ and
M2+ molecular ion peaks due to the presence of chlorine isotopes 35 and 37 for 6a and bromine
isotopes 79 and 81 for 6b (see HRMS spectra as Figures s18 and s21 for 6a and 6b, respectively
in the Supplementary Information).
Cl
(
ii)
N
N
N
N
N
R
4a-e
6a-e
R
1
87
1
1
1
88
89
90
Scheme 3. Sonogashira cross-coupling with 4-ethynylpyridine. Reagents and solutions:
(ii) 4-ethynylpyridine hydrochloride, PdCl (PPh ) , CuI, Cs CO , THF-water, 60 °C, 24
2
3 2
2
3
h.
1
91
Table 3. Yields of compounds 6a–6e.
Comp
ound
%
R
Yield
-
6
a
98
90
Cl
Br
-
6b

-
-
6
c
94
94
79
F
6d
NO2
-
6e
OMe
1
92
2.2. Crystal Structure Interpretation
1
1
1
1
1
1
93
94
95
96
97
98
The solid-state geometry of 4-(ethynylpyridine)quinazolines 6a–6e was established by
means of single-crystal X-ray diffraction (XRD). Crystals of compound 6a were grown by slow
evaporation of a chloroform/hexane solution. The crystal data and structure refinement for
compound 6a are shown in Table 4. The molecular structure of compound 6a is given in Figure 2
and the selected bond distances and angles are included in Table 5.
1
99
Table 4. Crystal data and structure refinement for compound 6a.
Identification code
Empirical formula
Formula weight
Temperature
Shelx
C H ClN
3
2
1
12
341.79
100(2) K
Wavelength
0.71073 Å
Monoclinic
P 21/c
Crystal system
Space group

a = 3.8537(5) Å
b = 21.924(3) Å
c = 18.377(2) Å
α = 90°
β = 93.536(6)°
γ = 90°
Unit cell dimensions
3
Volume
Z
1549.7(3) Å
4
3
Density (calculated)
Absorption coefficient
F(000)
1.465 Mg/m
−
1
0.254 mm
704
3
Crystal size
0.600 × 0.040 × 0.040 mm
1.447 to 26.661°
Theta range for data collection
Index ranges
−4 ≤ h ≤ 4, −27 ≤ k ≤ 27, −23 ≤ l ≤23
Reflections collected
Independent reflections
Completeness to theta = 25.242°
Refinement method
Data / restraints / parameters
24266
3227 [R(int) = 0.0367]
0.992
Full-matrix least-squares on F2
3227 / 0 / 226
2
Goodness-of-fit on F
1.16
Final R indices [I > 2sigma(I)]
R indices (all data)
R1 = 0.0490, wR2 = 0.1162
R1 = 0.0583, wR2 = 0.1204
n/a
Extinction coefficient
−
3
Largest diff. peak and hole
0.570 and −0.271 e.Å

2
2
00
01
Figure 2. Crystal structure of compound 6a.
Table 5. Selected bond distances (Å), bond angles (°) and dihedral angles (°).
Bond Distances (Å)
C(3)-C(6)
C(6)-C(7)
1.433(3)
1.194(3)
1.442(3)
1.318(3)
1.323(3)
1.362(3)
1.746(2)
1.335(3)
1.344(3)
C(7)-C(8)
C(8)-N(2)
C(15)-N(3)
C(14)-N(3)
C(19)-Cl(1)
C(1)-N(1)
C(5)-N(1)
Bond angles (°)
C(15)-C(16)-C(21)
C(15)-C(16)-C(17)
C(7)-C(8)-N(2)
C(3)-C(6)-C(7)
Dihedral angles (°)
120.4(2)
120.7(2)
118.1(2)
175.8(2)

C(3)-C(6)-C(7)-C(8)
-13(5)
2
02
2
2
2
2
2
2
2
2
2
2
03
04
05
06
07
08
09
10
11
12
The molecule crystallises with one molecule in the asymmetric unit. The molecule contains
a quinazoline moiety, a 4-ethynylpyridine moiety and a chlorobenzene ring, all which define
three sets of planes. The 4-chlorobenzene ring, the 4-ethynylpyridine moiety and the quinazoline
moiety planes are not co-planar. This phenomenon has been observed in related compounds from
literature [19,20]. The dihedral angles between the chlorobenzene ring and the 4-ethynylpyridine
moiety is 12.67(6)°, that between the 4-ethynylpyridine moiety and the quinazoline moiety is
35.33(6)° while that between chlorobenzene ring and 4-ethynylpyridine 37.58(6)°. In the crystal,
the molecules are stabilized via a C—H…N intermolecular interactions (C…N = 2.7881(4) Å
and <C—H…N = 100°) and a strong π…π intermolecular interactions between adjacent
quinazoline moieties (Cg…Cg = 3.6034, symmetry code = 1- x, -y, -z).
2
2
13
14
2.3. Biological Evaluation
2.3.1. In Vitro Anti-Mycobacterium Tuberculosis Activity
2
2
2
2
2
2
2
2
15
16
17
18
19
20
21
22
Pyridine containing compounds serve a great purpose in inhibiting the progression of Mtb
cells [21]. Such derivatives are represented by a small commercially available first-line drug
isoniazid and the second-line drugs ethionamide and prothionamide [22,23]. The conjugated π-
system of the pyridine ring is reported to bring about receptor π-π intermolecular interactions and
2
hydrogen bond interaction through the sp -hybridised nitrogen [24]. Accordingly, in vitro anti-
Mycobacterium tuberculosis activity of compounds 4a–4e, 5a–5e and 6a–6e were assayed
against the H Rv strain in the concentration range of 0.723–125 µM according to the procedure
3
7
explained in Section 3.6. Activity was assessed at MIC , which is the minimum concentration
9
0

2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
required to inhibit at least 90% of the bacterial population, employing rifampicin as a reference
drug. As an aid for structural activity relationship (SAR) analysis, compounds which exhibited
MIC , ≤10 µM were rendered active (relative to rifampicin, Table 6); compounds with MIC
90
9
0
ranging from 10–25 µM were considered to have moderate activity; MIC ranging from 25 µM–
9
0
(>125 µM) was considered as poor activity. Consequently, a structure activity relationship
(SAR) study was explored on the 4-position of the quinazoline scaffold and the 4-position of the
2-aryl substituents. Firstly, the 4-chloroquinazolines 4a–4e were tested against Mtb H Rv strain
37
and no significant activity was observed (MIC , >125 µM). We then hypothesised that the poor
9
0
activity was attributed to the poor solubility of these compounds 4a–4e in DMSO. To improve
the solubility, a 3-nitropyridylamino group was introduced on the C(4) position of the 4-
chloroquinazolines. Introduction of the amine group resulted in moderate improvement in
activity for compounds 5b–5e, with MIC ranging from 20.65–109.09 µM. Another contributing
9
0
factor was hypothesised to be the presence of the -NO group ortho to the amine group. This is
2
true because the -NO group is reported to enhance anti-Mtb activity, as it binds –NH groups of
2
amino acids found in active sites of Mtb proteins [21]. A second pyridine moiety, 4-
ethynylpyridine, introduced at the C(4) position led to significant improvement in potency, with
MIC ranging from <0.72 µM–23 µM for compounds 6a–b and 6d–6e.
9
0
From both series of 5a–5e and 6a–6e, compounds 5e and 6e, both containing a methoxy
group on the para position of the 2-aryl substituent appeared to show good potency with MIC90
of 20.65 µM and <0.72 µM, respectively. This moderate to significantly higher activity was
hypothesised to be due to the presence of the polar effect of the methoxy oxygen which is
responsible for hydrogen bonding during inhibition (see Figure 2). Other compounds with good
activity emanated from the 6a–6e series i.e. compounds 6a and 6b, MIC , 9.82 µM and 9.39
9
0

2
2
2
2
2
2
2
46
47
48
49
50
51
52
µM, respectively. Apart from the biologically influential π-conjugated system of the pyridine
ring, these two compounds contain halogen groups i.e. 4-Cl and 4-Br, which are responsible for
halogen bond interactions probably with amino acids in the active site of Mtb proteins (see
Figure 2 and Figure 3). Compounds bearing the 4-ethynylpyridine demonstrated better activity as
compared to compounds with 3-nitropyridylamino group, as evident when comparing 6a vs 5a,
6b vs 5b, 6d vs 5d, and 6e vs 5e. An exception was noted for compounds 6c vs 5c where the
introduction of an ethynylpyridine resulted in loss of activity.
2
2
2
53
54
55
2
2
2
2
2
2
2
2
2
2
2
2
56
57
58
59
60
61
62
63
64
65
66
67
Compounds 6a–6e were further evaluated for their in vitro cytotoxic effects against Raw
264.7 microphage cells using a Glomax-Multi microplate reader MTT assay. The lethal dosage
of the compounds was recorded as IC (which is the inhibitory concentration required to inhibit
5
0
50% of viable cells) at a maximum concentration of 50 µM. All assayed compounds showed low
cytotoxicity against the microphages i.e. IC > 50 µM (Table 6). The low cytotoxicity of these
5
0
compounds allowed for further investigations of crucial efficacy determining parameters such as
the selectivity index (SI) ratio. The selectivity index ratio is an acceptable parameter responsible
for measuring the gap between cytotoxicity and anti-mycobacterial activity [25]. Theoretically,
the higher the SI ratio, the more effective and safer a drug would be during in vivo treatment for
bacterial infection. According to literature [25], the SI ratio should be ≥10 for compounds to be
considered as new drug leads. A compound’s SI ratio means that, the compounds are cytotoxic
only at high concentrations and show anti-Mtb activity at lower concentrations. Accordingly,

2
2
2
68
69
70
selectivity index ratios for compounds 6a–6e were calculated as SI = IC / MIC . In this study,
90 90
only compounds 6b and 6e suited the literature SI window with SI ratios of 12.16 and >150.24
respectively (Table 6).
2
71
Table 6. Anti-mycobacterial activity (MIC in µM) of compounds 4a–4e, 5a–5e and 6a–6e.
9
0
Entry
MIC90
(µM)
>125
>125
>125
>125
>125
>125
40.84
85.71
109.09
20.65
9.82
Compound Structure
R
Compound ID
IC50
SI
Number
1
2
3
4
5
6
7
8
9
-Cl
-Br
-F
4a
4b
4c
4d
4e
5a
5b
5c
5d
5e
6a
6b
6c
6d
6e
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
5.22
12.16
ND
2.52
Cl
N
N
R
ND
-NO2
-OMe
-Cl
ND
ND
ND
O N
2
-Br
ND
N
HN
-F
ND
N
N
-NO2
-OMe
-Cl
ND
R
1
1
1
1
1
1
0
1
2
3
4
5
ND
51.30
114.17
52.36
59.04
-Br
9.39
N
N
N
-F
>125
23.46
<0.72
0.12
-NO2
-OMe
R
108.17 >150.24
Rifampicin

2
2
72
73
ND = Not determined
2.3.2. Molecular Docking Studies
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
The accuracy of docking experiments were measured by performing cross-docking of
compound 2 and 3 (compound numbers taken as represented from Gising et al [11]) co-
crystallized to x-ray crystal structures 3ZXV and 3ZXR, respectively. These two co-crystallized
structures were designed by Gising et al [11], and were tested against Mycobacterium
tuberculosis glutamine synthetase (MtGS) enzyme both in vitro and in silico. There is a huge
body of evidence in the literature postulating that the accuracy of docking results is system
dependent [26,27]. Also, docking results are only able to correctly discriminate between
compounds which bind to the receptor and those that do not show any affinity to the target
receptor correctly, as previously reported by both Graves et al. and Mobley et al. [28,29]. Hence,
methods such as MMGBSA (Kollman et al [30]), MMPBSA (Kuhn et al. [31]) and recently
water map score (Robinson et al. [32]), and Free Energy Perturbation (FEP+) (Ciordia et al. [33])
have been developed and used successfully by various academic groups and some big pharma
companies to overcome this shortcoming. The relative binding affinities from FEP+ can be
correctly correlated with in vitro (experimental) values, a recent focus in molecular modeling.
However, in this work only docking by virtue of cross-docking and classical Glide XP docking
were employed.
2
2
2
2
2
90
91
92
93
94
This approach is aimed at using docking to establish the binding mechanism of the
candidate compounds designed, and their ability to block the biosynthesis of poly-L-
glutamate−glutamine by MtGS. In comparison to compound 2 and 3 designed by Gising et al.
Figure 3 shows the cross-docking poses of compound 2 and 3 from Gising et al [11], overlaid on
the native conformers of the x-ray crystallographic structures of MtGS. Figure 3a shows an

2
2
2
2
2
3
3
95
96
97
98
99
00
01
overlay of 2 (RMSD of 1.34Å) on the native conformer; and the binding mode is not correctly
reproduced. A conformational change is evident on 2 when cross-docked with 3ZXV. However,
in Figure s27a classical Glide XP docking of 2 to 3ZXR reproduces the native conformation of 2
on the x-ray crystal structure (3ZXV). The more prominent hydrogen bond is between Nitrogen
of the pyridine ring moiety of 2 and the OH group of SER280. Also, a hydrogen bond between
the NH group of the pyridine moiety and the carbonyl group of LYS361 is observed. Gising et
2
al [11], also observed the same binding mode and similar hydrogen bonds.
3
02

(a)
(b)
3
03
3
3
3
3
3
04
05
06
07
08
Figure 3. (a) The cross-docking pose of Compound 2 (grey carbons) overlay to the native
conformer (green carbons) from the co-crystalized structure of MtGS PDB ID (3ZXV), with an
2
+
RMSD of 1.34 Å. Mg ions are shown in purple spheres. (b) The cross-docking pose of
Compound 3 (pink carbons) overlay to the native conformer (grey carbons) from the co-
crystalized structure of MtGS PDB ID (3ZXR), with an RMSD of 0.28 Å.
3
3
3
3
3
3
09
10
11
12
13
14
The cross-docking of compound 3 on 3ZXR is shown in Figure 3 (b), the bound
conformation of 3 and the native co-crystalized structure of 3 overlays perfectly with an RMSD
of 0.28 Å. SER280 hydrogen bonds with the Nitrogen atom of the pyridine ring of 3.
Furthermore, figure s27b has revealed that classical Glide XP is able to reproduce the native
bound conformation of 3 to MtGS (3ZXR). The binding mode and the hydrogen bond observed
in figure 3b and s27b are consistent with those observed by Gising et al [11].

3
3
3
15
16
17
We further cross-docked AMP co-crystalized with MtGS from PDB ID 1HTO. AMP was
not well superposed over the native conformer resulting in an RSMD of 2.24Å (see Figure 4a).
Then, AMP was further docked on 3ZXR using classical Glide XP.
(a)
(b)
3
3
3
3
18
19
20
21
Figure 4. (a) AMP in the native conformation (grey carbons ) and is not well overlaid with
cross-docked conformation (green carbons) into the active site cavity of 1HTO. Manganese 2+
ion is shown as dark purple spheres. (b) AMP (green carbons) bound to amino acids of 3ZXR,
Magnesium 2+ ions are shown as light purple spheres.
3
3
3
3
22
23
24
25
It is evident that the binding mode of AMP to 1HTO and 3ZXR is not the same and the
active site cavities are quite different including cofactors such as manganese (1HTO) and
magnesium (3ZXR). The evidence observed above, thus validates 3ZXR as an x-ray crystal
structure of choice for further studies.
3
3
26
27
Molecular docking has been previously used successfully as a computer-aided drug
design (CADD) method for the design of Mycobacterium tuberculosis drug candidates [11].

3
3
3
3
28
29
30
31
Docking of compounds 5a–5e, 6a–6e was performed by using classical Glide XP docking on
3ZXR enzyme. Induced Fit Docking (IFD) did not reproduce the native poses, thus indicating
that the binding of these compounds understudy including compounds 2 and 3 from Gising et al
follows a lock and key approach. Therefore, IFD was not used for further studies.
(
b)
(
a)
(
d)
(
c)
3
3
3
3
32
33
34
35
Figure 5. (a) IFD binding pose of 6e in the active site cavity of 3ZXR depicting the
hydrogen bonds, and π-π interactions. (b) Ligand interaction diagram of 6e with
important amino acid residues in the active site cavity of 3ZXR and a metal
2
+
coordination of methoxy benzene moiety and Mg . (c) An overlay of compounds in the

3
3
3
36
37
38
6 series, including 6e (green carbons), 6c (pink carbons), 6d (light blue carbons), and 6b
(light brown carbons). (d) An overlay of compounds in the 5 series, including 5a (green
carbons), 5b (pink carbons), 5c (light blue carbons), and 5d (light brown carbons).
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
The Glide XP docking results have revealed the binding mode of compound 6e to MtGS
enzyme, that is facilitated by a hydrogen bond between SER280 and the Nitrogen atom of the
pyridine ring moiety of 6e. A π-π interaction between the pyridine ring of 6e and the benzene
ring of PHE232 is shown (see Fig. 5a). Furthermore, a metal coordination between the Oxygen
2
+
atom of the methoxybenzene moiety of 6e and Mg ion of MtGS is shown (see Fig. 5b). The
binding modes of the compounds in the 6 series is similar to the binding mode observed by
Gising et al [11]. Also, the validation of the docking by cross-docking follows the same binding
mechanism. This therefore, proves that the compounds in the 6 series are MtGS inhibitors.
Compounds in the 5 series, on the other hand, have revealed a different binding mode
2
+
hypothesis. They prefer to locate themselves far below the Mg ions moiety, the Oxygen of the
nitropyridine ring hydrogen bonds with SER280. Also, some π-π interactions between the
compounds in the 5 series and PHE232 and ARG364 are shown see Fig. 5d. The classical Glide
XP method employed in this work has been able to assist in determining compounds that would
be bio-active prior to synthesis and biological evaluations. It is worth noting that compounds that
were synthesized without being tested in silico for their bioactivity exhibited no activity towards
the enzyme such as compounds in the 3a–3e and 4a–4e series.
3
55
2.3.3. Calculated Physicochemical and ADME Properties
3
3
56
57
The predicted physicochemical and ADME properties using QikProp v5.7 [34] for
compounds in the 5a–5e and 6a–6e series are presented in Table 8. In particular, the predicted

3
3
3
3
3
3
3
58
59
60
61
62
63
64
central nervous system activity (CNS), octanol/gas partition coefficient (logPo/w), aqueous
solubility (logS), IC50 value for blockage of HERG K+ channels (logHERG), brain/blood
partition coefficient (logBB), and human oral absorption (%Human Oral Absorption). It is worth
noting that only compound 6b violated the Lipinski’s rule of 5 by having an aqueous solubility
<6.5. The predicted physicochemical properties have revealed that the most potent compound 6e
is drug-like and does not violate the Lipinski’s rule of 5. In fact, all the compounds in the series
are drug-like, with an exception of compound 6b.
3
3
65
66
Table 7. Predicted physicochemical and ADME properties for compound 5a–5e and
6a–6e series.
%
Hu
M
man
Oral
Molecule CNS
logPo logP logPo logS CIlo logHE PCaco Log #meta LogKh
W
a
c
d
e
f
g
h
i
j
k
l
ID
ct
w
/w
gS RG
BB
b
sa
b
Absor
m
ption
386
1
1
6.00 7.99
−6.5 −6.9
3,712. 0.09
100.00
0
6
6
b
d
1
.25
0
5.385
−7.398
2
0.754
5
7
62 71
551
6
352
7.62 9.35
−5.7 −5.7
439.44 −1.2
100.00
0
-2 .35
1
4.052
5.027
−7.396
−7.336
3
2
0.505
0.647
4
8
39 87
0
00
3
25 15.44 8.00
−6.0 −5.6
3,712. 0.03
716
100.00
0
6
c
1
.
34
8
8
31 77
0

4
341
1
1
1
1
1
1
1
5.89 7.98
−6.4 −6.0
3,712. 0.08
100.00
0
6
a
1
0
0
.79
9
5.267
4.860
4.354
3.922
4.432
3.165
4.095
−7.370
−7.342
−6.655
−6.634
−6.702
−6.700
−6.617
2
3
3
4
3
4
3
0.729
0.576
0.559
0.450
0.585
0
8
20 27
415
4
3
37
.38
0
5.84 8.45
−5.8 −5.5
3,706. −0.1
100.00
0
6
e
8
4
17 85
216 53
3
77
.78
9
7.52 10.0
79
−5.7 −6.4
965.13 −0.5
100.00
0
5
a
4
83 14
7
68
373
7.66 10.5
−5.2 −6.0
967.64 −0.8
05
100.00
0
5
e
-1 .37
0
4
46
57 25
5
422
7.67 10.0
−5.9 −7.3
950.90 −0.5
68
100.00
0
5
5
b
d
0
.24
0
6
98
17 27
5
388
8.88 11.4
−5.2 −6.2
114.15 −1.8
10
-2 .34
2
0.397 82.302
3
53
27 29
4
361
7.08 10.1
−5.4 −6.0
953.02 −0.6
21
100.00
0.487
0
5
c
0
.33
4
3
05
18 77
1

a
b
3
3
3
3
3
3
3
3
3
3
3
67
68
69
70
71
72
73
74
75
76
77
Predicted central nervous system activity, range −2 (inactive) to +2 (active).
c
Molecular weight of the molecule, range 130.0–725.0. Predicted octanol/gas partition
d
e
coefficient, range 8.0–35.0. Predicted water/gas partition coefficient, range 4.0–45.0.
f
Predicted octanol/water partition coefficient, range −2.0–6.5. Predicted aqueous
g
solubility, range −6.5–0.5. Conformation-independent predicted aqueous solubility,
h
range −6.5–0.5. Predicted IC50 value for blockage of HERG K+ channels, range
I
concern below −5. Predicted apparent Caco-2 cell permeability in nm/sec. Caco-2 cells
j
are a model for the gut-blood barrier, range <25 poor & >500 great. Predicted
k
brain/blood partition coefficient, range −3.0–1.2.
Number of likely metabolic
l
m
reactions, 1–8. Prediction of binding to human serum albumin, range −1.5–1.5.
Predicted human oral absorption on 0 to 100% scale, >80% is high & <25% is poor.
3
3
3
3
3
3
3
3
3
3
3
3
78
79
80
81
82
83
84
85
86
87
88
89
The number of sites of metabolism for compounds 5a–5e have revealed that compounds in
this series are metabolised through aryl nitro to NH , para-hydroxylation of aryl, and pyridine
2
C2 hydroxylation. The metabolism of compound 6a–6e series are metabolised through pyridine
C2 hydroxylation, aryl nitro reduction to NH , ether dealkylation, and para-hydroxylation of
2
aryl. We also checked whether our compounds in the series 5a–5e and 6a–6e are indeed new
derivatives by performing a similarity search using QikProp v5.7 [34]. Compound 5a shows a
similarity index that spans from quinestradol 88.94, atovaquone 87.69, talniflumate 84.91,
diphenpyramide 84.40, to quinestrol 84.00, respectively. Compound 5b on the other hand shows
a similarity index that spans from quinestradol 87.23, talniflumate 86.10, atovaquone 85.97,
brovincamine 84.97, to ziprasidone 83.23. Compound 5c is a new derivative because of its
similarity index with quinestradol 89.24, atovaquone 88.15, mofezolac 85.26, diphenpyramide
84.84, and clometacin 83.96. Compound 5d shows a similarity index with cilostazol 87.65,

3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
vesnarinone 87.47, ketanserin 87.26, buspirone 84.39, and benperidol 84.15. Compound 5e
exhibits similarity index with cilostazol 88.50, quinestradol 86.17, benperidol 86.10, oxypertine
85.49, and timiperone 84.38.
Compound 6a is a new derivative because it exhibits a similarity index with known
compounds such as isofezolac 85.42, metaclazepam 83.70, oxetorone 82.73, quazepam 82.33,
and thioridazine 82.32%. Compound 6b is also a new derivative because it exhibited a similarity
index with known compounds such as isofezolac 85.02, metaclazepam 84.94, quazepam 84.05,
thioridazine 82.99, and pentagestrone 81.58%. Compound 6c shows a similarity index that spans
from isofezolac 84.08, oxetorone 84.07, midazolam 83.38, metaclazepam 83.11, and clemizole
82.67. Compound 6d exhibits a similarity index that spans from isofezolac 85.42, metaclazepam
83.70, oxetorone 82.73, quazepam 82.33, and thioridazine 82.32% when compared with known
compounds. Compound 6e has a similarity index of quinestradol 86.42, clemizole 84.59,
etoricoxib 83.89, talniflumate 83.73, and atovaquone 83.64%. Therefore, the compounds in this
series 5a–5e and 6a–6e are newly synthesised derivatives which exhibit drug like properties
because of their similarity indexes to known compounds ranging from 80–86%.
4
4
06
07
3. Materials and Methods
3.1. General
4
4
4
4
08
09
10
11
Chemical reagents were purchased from Sigma-Aldrich or Merck (Johannesburg, South
Africa) and used without further purification. Melting points were obtained using Lasec/SA-
melting point apparatus from Lasec Company, SA (Johannesburg, South Africa). IR spectra were
recorded using a Agilent Technologies Cary 600 Series FT-IR spectrometer (Agilent

4
4
4
4
4
4
12
13
14
15
16
17
Technologies Australia [M] Pty Ltd., Mulgrave, Australia). NMR spectra were obtained using a
Bruker Ascend 400 MHz NMR spectrometer (Bruker Biospin GmhH, Karlsrushe, Germany)
1
13
operating at 400 MHz ( H) and 100 MHz ( C), where chemical shifts were quoted relative to the
TMS peak. High-resolution mass spectra were recorded at an ionisation potential of 70 eV using
a Waters Synapt G2 Quadrupole Time-of-flight mass spectrometer (Waters Corp., Milford, MA,
USA) at the University of Stellenbosch Central Analytical Facility).
4
18
3.2. Synthesis of 2-Aryl-quinazoline-4-one Derivatives 3a–3e
4
4
4
4
4
4
4
19
20
21
22
23
24
25
Quinazolinone derivatives 3a–3e were prepared following a literature method [18]. To a 100
mL round bottom flask, anthranilamide 1a (1 mmol), benzaldehyde derivatives 2a–2e (1 mmol)
and iodine (2 equiv.) in ethanol (30 mL per mmol of 1a) was refluxed at 80 °C for 6 h. The
mixture was left to cool to room temperature and quenched with cold saturated sodium
metabisulfate solution. The resulting precipitate was filtered and washed thoroughly with water.
The solid product was recrystalised from acetonitrile and oven dried to yield the corresponding
quinazoline-4(3H)-ones 3a–3e [13]. The following products were prepared accordingly:
4
4
4
4
4
4
4
26
27
28
29
30
31
32
2-(4-Chlorophenyl)quinazolin-4(3H)-one (3a): A mixture of 1a (1.50 g, 11.02 mmol),
benzaldehyde (2a) (1.55 g, 11.02 mmol) and iodine (2.80 g, 22.04 mmol) in ethanol (100 mL)
yielded 3a as a white solid (2.45 g, 87%); m.p. 295.2–297.1 °C, lit (298–299 °C) [13]; FT-
−
1 1
IR(vmax) 764, 840, 1168, 1232, 1259, 1531, 1606, 1666, 3052, 3174 cm ; H-NMR (400 MHz,
DMSO-d , ppm) 7.60–7.53 (m, 1H), 7.73–7.77 (m, 3H), 7.89–7.84 (m, 1H), 8.12–8.19 (m, 3H),
6
1
3
12.63 (s, 1H); C-NMR (100 MHz, DMSO-d , ppm) 121.0, 125.2, 125.9, 126.8, 127.4, 129.8,
6
131.6, 131.9, 134.6, 148.5, 151.5, 162.2.