4-Aminoquinoline Antimalarials Containing a Benzylmethylpyridylmethylamine Group Are Active against Drug Resistant Plasmodium falciparum and Exhibit Oral Activity in Mice

Article abstract of DOI:10.1021/acs.jmedchem.7b01537

Emergence of drug resistant Plasmodium falciparum including artemisinin-tolerant parasites highlights the need for new antimalarials. We have previously shown that dibemequines, 4-amino-7-chloroquinolines with dibenzylmethylamine (dibemethin) side chains, are efficacious. In this study, analogues in which the terminal phenyl group of the dibemethin was replaced with a 2-pyridyl group and in which the 4-amino-7-chloroquinoline was either maintained or replaced with a 4-aminoquinoline-7-carbonitrile were synthesized in an effort to improve druglikeness. These compounds exhibited significantly improved solubility and decreased lipophilicity and were potent against chloroquine-sensitive (NF54) and -resistant (Dd2 and 7G8) P. falciparum strains with 5/6 having IC₅₀ < 100 nM against the NF54 strain. All inhibited both β-hematin (synthetic hemozoin) formation and hemozoin formation in the parasite. Parasitemia was reduced by over 90% in P. berghei infected mice in 3/6 derivatives following oral dosing at 4 × 30 mg/kg, with microsomal metabolic stability data suggesting that this could be attributed to highly active metabolites.

Full text of DOI:10.1021/acs.jmedchem.7b01537

Subscriber access provided by READING UNIV
Article
4-Aminoquinoline Antimalarials Containing a
Benzylmethylpyridylmethylamine Group are Active against Drug
Resistant Plasmodium falciparum and Exhibit Oral Activity in Mice
Mukesh C. Joshi, John Okombo, Samkele Nsumiwa, Jeffrey Ndove, Dale
Taylor, Lubbe Wiesner, Roger Hunter, Kelly Chibale, and Timothy J Egan
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01537 • Publication Date (Web): 29 Nov 2017
Downloaded from http://pubs.acs.org on December 1, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
4ꢀAminoquinoline Antimalarials Containing a Benzylmethylpyridylmethylamine Group
are Active against Drug Resistant Plasmodium falciparum and Exhibit Oral Activity in
Mice
9
Mukesh C. Joshi,†‡|| John Okombo,†‡ Samkele Nsumiwa,† Jeffrey Ndove,§ Dale Taylor,§
Lubbe Wiesner,§ Roger Hunter,† Kelly Chibale†⊥ and Timothy J. Egan†*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
† Department of Chemistry, University of Cape Town, Rondebosch 7701, South Africa
⊥ South African Medical Research Council Drug Discovery and Development Research Unit,
Department of Chemistry and Institute of Infectious Disease and Molecular Medicine,
University of Cape Town, Rondebosch 7701, South Africa
§ Department of Medicine, Division of Clinical Pharmacology, University of Cape Town,
Observatory, 7925, South Africa
‡ these authors contributed equally to the work.
|| current address: Motilal Nehru College, Benito Juarez Marg, South Campus, University of
Delhi, New Delhiꢀ110043, India
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 41
1
2
3
4
5
6
7
8
9
ABSTRACT: Emergence of drug resistant Plasmodium falciparum including artemisininꢀ
tolerant parasites highlights the need for new antimalarials. We have previously shown that
dibemequines, 4ꢀaminoꢀ7ꢀchloroquinolines with dibenzylmethylamine (dibemethin) side
chains, are efficacious. In this study, analogues in which the terminal phenyl group of the
dibemethin was replaced with a 2ꢀpyridyl group and the 4ꢀaminoꢀ7ꢀchloroquinoline was
either maintained or replaced with a 4ꢀaminoquinolineꢀ7ꢀcarbonitrile were synthesized in an
effort to improve drugꢀlikeness. These compounds exhibited significantly improved solubility
and decreased lipophilicity and were potent against chloroquineꢀsensitive (NF54) and
resistant (Dd2 and 7G8) P. falciparum strains with 5/6 having IC50s <100 nM against the
NF54 strain. All inhibited both βꢀhematin (synthetic hemozoin) formation and hemozoin
formation in the parasite. Parasitemia was reduced by over 90% in P. bergheiꢀinfected mice
in 4/7 derivatives following oral dosing at 4 × 30 mg/kg, with microsomal metabolic stability
data suggesting that this could be attributed to highly active metabolites.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 3 of 41
Journal of Medicinal Chemistry
1
2
3
INTRODUCTION
4
5
6
7
8
Inhibition of the heme detoxification pathway in Plasmodium falciparum is an ideal
target for antimalarial drugs owing to its critical role in parasite survival, absence of a human
equivalent and nonꢀgenomic derivation (hence immutability). Chloroquine (CQ) and a
number of related aminoquinolines are proposed to target this pathway, inhibiting the
formation of inert crystalline hemozoin.1ꢀ3 However, the efficacy of many of these quinolines
has been compromised to a greater or lesser extent by the spread of CQꢀresistant (CQR)
parasites which harbor mutations in the chloroquine resistance transporter gene (Pfcrt) whose
cognate protein (alongside other transporters) mediates drug extrusion from the parasite
digestive vacuole (DV).4, 5 At present, artemisinin combination therapy (ACT) constitutes the
recommended treatment option against malaria and though still largely effective, there is a
growing body of evidence that artemisininꢀtolerant P. falciparum parasite strains are
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
emerging, and consequently threatening the therapeutic utility of the ACT regimen.6,
Therefore, novel strategies to address the challenge of antimalarial resistance and develop
cheap new drugs with good safety profiles are urgently needed.
One approach to discovery of novel and potent antimalarial agents involves structural reꢀ
engineering of existing CQꢀlike drugs by, for instance, coupling them to resistance reversers.
Seminal work on this strategy has been reported by Peyton and coꢀworkers, in which they
proposed preservation of the fundamental pharmacophore for hemeꢀbinding (4ꢀ
aminoquinoline), inhibition of hemozoin formation (4ꢀaminoꢀ7ꢀchloroquinoline) and drug
accumulation (a tertiary amino group in the sideꢀchain) in covalent linkage with the
pharmacophore for a chloroquine chemosensitizing or resistance reversal agent (RA), namely
two suitably positioned aromatic groups with an amino group separated by a short chain.8ꢀ10
The rationale behind the potency of these molecules is that the cellular transport machinery
would fail to recognize the hybrid leading to vacuolar accumulation, or the molecule would
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 41
1
2
3
4
5
6
7
8
9
directly target the function of the efflux protein(s) thus blocking its extrusion. Indeed, the
feasibility of this approach has since been interrogated by other groups who have generated
various modifications of such compounds. For instance, conjugates based on
pentacycloundecylamine, dihydropyrimidinone, astemizole and acridines with potent activity
against CQR parasite strains have been reported.11ꢀ14 Recently, our group also synthesized
and evaluated dibemequine analogues containing a 4ꢀaminoꢀ7ꢀchloroquinoline nucleus with a
dibenzylmethylamine side chain. These compounds exhibited strong in vitro antiplasmodial
and in vivo antimalarial efficacy and notably, were further able to inhibit PfCRTꢀmediated
drug extrusion from the parasite DV.15
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In an attempt to further improve the drugꢀlike properties of these compounds we have
now made various structural alterations to the prototype molecule (Figure 1) to afford new
analogues. First, the terminal phenyl ring in the dibenzylmethylamine side group was
replaced with an o-pyridyl ring with the aim of improving aqueous solubility via introduction
of an extra hydrogen bond acceptor. This was additionally expected to potentially improve
vacuolar drug accumulation through pH trapping. Secondly, the chloroꢀ substituent at carbon
7 (C7) on the 4ꢀaminoquinoline core was replaced by a nitrile group to afford analogues with
lower molecular weight (MW) and lipophilicity while still retaining βꢀhematin inhibition
activity that critically depends on the substituent at this position.16 This, we envisaged, would
lower the potential for hERG channel inhibition (a frequent liability among aromatic weak
bases such as these) and improve druggability by improving solubility.17 Finally, one of the
original parent series compounds, 1, was prepared as a hydrochloride salt since its
corresponding free base had previously been shown to possess low intrinsic solubility, which
limited oral bioavailability. Here, we report potent in vitro activity, improved
physicochemical properties and retention of antimalarial efficacy of these modified
dibemequines in vivo as a further validation of the reversedꢀCQ strategy. In addition, we
ACS Paragon Plus Environment

Page 5 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
present data on metabolite identification studies on of these compounds and a mechanistic
investigation with respect to targeting the hemoglobin degradation pathway.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1: Prototype compound showing the structure activity relationship (SAR) that was
explored to afford various analogues evaluated in this study. 1: R = Cl, X = CH, linker m; 2:
R = Cl, X = N, linker o; 3: R = Cl, X = N, linker m; 4: R = Cl, X = N, linker p; 5: R = CN, X
= N, linker o; 6: R = CN, X = N, linker m; 7: R = CN, X = N, linker p.
CHEMISTRY
The target molecules (2 – 7) were synthesized by reaction of the appropriate Nꢀ(nꢀ
(aminomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridineꢀ2ꢀyl)methanamine where n = 2 – 4 containing
an o, m or p linker group (8, 9 and 10 respectively) with an excess of commercially available
4,7ꢀdichloroquinoline or 4ꢀchloroꢀ7ꢀcyanoquinoline in anhydrous Nꢀmethylꢀ2ꢀpyrrolidone in
the presence of anhydrous K2CO3 with triethylamine as a base (Scheme 1).15 Each reaction
was carried out in sealed cycloaddition tubes under a N2 atmosphere with a gradual increase
of temperature to 90 °C. Once this temperature was reached, the reaction was allowed to stir
for an additional 48 h. The products were then extracted from an alkaline aqueous solution
into ethyl acetate and purified by column chromatography to afford modest yields (23 to
46%). All products were characterized by infrared, 1H and 13C NMR, and mass spectrometry.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 41
1
2
3
4
5
6
7
8
Solid intermediates were further characterized by elemental combustion analysis. All final
products were subjected to highꢀresolution mass spectrometry to confirm identity, and HPLC
to confirm purity (see Supporting Information for further details).
9
8, 9 and 10 were prepared in three steps (Scheme 1). First, 1ꢀ(azidomethyl)ꢀ2, 3 and 4ꢀ
(bromomethyl)benzenes (11, 12 and 13) were prepared from the appropriate dibromoxylene
(14, 15 and 16) via a substitution reaction with sodium azide as previously described.18 The
appropriate azido intermediate was then reacted with Nꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine
(17) to afford a series of Nꢀ(nꢀ(azidomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamines
(n = 2 – 4) (18, 19, 20) using anhydrous K2CO3 as base in anhydrous acetonitrile at 0 °C,
followed by overnight reflux. These were then converted to the corresponding Nꢀ(nꢀ
(aminomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridineꢀ2ꢀyl)methanamine (n = 2 – 4) via a Staudinger
reduction involving reaction with PPh3 in THF for 30 min at room temperature, followed by
the addition of water and reflux for an additional 6ꢀ10 h. All the intermediates were purified
by SiO2 gel column chromatography in a good to moderate yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NaN3
i
N3
NH
Br
N3
N
+
N
Br
Br
ii, iii
N
18 ꢀ 20
14 ꢀ 16
11 ꢀ 13
17
iv, v
Cl
H2N
HN
N
N
N
+
N
N
X
N
vi
X
2 ꢀ 7
X = Cl or CN
8 ꢀ 10
Scheme 1: Reagents and conditions: (i) DMF, RT, 16 h; (ii) CH3CN, K2CO3 (1.5 equiv.), 0 °C; (iii) ꢀ,
16 h; (iv) PPh3 (1.2 equiv.), THF, RT, 0.5 h; (v) H2O, ꢀ, 6 h; (vi) K2CO3, TEA, N2, 90 °C, 48 h (X =
Cl, CN).
ACS Paragon Plus Environment

Page 7 of 41
Journal of Medicinal Chemistry
1
2
3
PHYSICOCHEMICAL AND BIOLOGICAL PROFILING
4
5
6
7
8
Solubility, Membrane Permeability and Lipophilicity. As an initial measure of the
effectiveness of the structural modifications in improving the drugꢀlike characteristics of this
series, solubility, permeability and partitioning properties of the new derivatives were
measured. The compounds generally exhibited moderate to high solubility in aqueous
medium at pH 6.5 (Table 1) ranging from 52 to 197 ꢁM. With the exception of 3 and 6,
replacement of the terminal phenyl ring with a pyridyl ring substantially increases solubility
of the free base relative to the parent hydrochloride compound 1⋅HCl. In view of their weak
base properties, they exist predominantly in their ionized protonated state at lower pH, with
the consequence that they are likely to stay in solution under assay conditions. Replacement
of the chloro with a cyano group at C7 on the quinoline ring did not seem to further influence
solubility.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Replacement of the terminal phenyl ring with a pyridyl group caused a small, but
significant decrease in permeability relative to the parent compound 1, consistent with the
decreased lipophilicity of the ring. Nonetheless, the derivatives all still showed high
membrane permeability at pH 6.5 in the PAMPA assay (logPapp > −5) and are predicted to
permeate the gastrointestinal wall via passive diffusion efficiently (Table 1). Replacement of
the 7ꢀCl group with a CN group had little effect on permeability.
The series exhibited moderately lipophilic experimental partition coefficients, with
logD7.4 values ranging between 2.47 and 3.80. Under acidic conditions (pH 5.0), all
compounds had a substantially reduced logD, consistent with ionization of the weakly basic
groups. The pyridodibemequines had somewhat lower logD5.0 than the parent compound 1,
with the 7ꢀcyano derivatives showing no substantial difference from the 7ꢀCl derivatives.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 41
1
2
3
Table1: Physicochemical profile of the seven test compounds and standards.
4
5
6
Permeability
Partition Coefficient
7
8
9
Kinetic Solubility
pH 6.5 (ꢁM)
Compound
Average Log Papp (SD)
pH4.0
pH6.5
LogD7.4
3.56
LogD5.0
1.33
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
51.8
164.7
59.8
ND
ꢀ3.60 (0.16)
1.HCl
ꢀ6.14 (0.34) ꢀ4.56 (0.03)
ꢀ3.81 (0.03) ꢀ4.10 (0.26)
ꢀ3.85 (0.32) ꢀ4.56 (0.10)
ꢀ6.03 (0.12) ꢀ4.01 (0.06)
3.79
0.90
2
3
4
5
6
7
2.47
0.82
196.6
172.4
72.4
2.85
0.84
3.80
0.81
ND
ꢀ3.60 (0.16)
2.75
0.85
143.5
ꢀ4.51 (0.40) ꢀ4.60 (0.04)
2.84
1.00
<5
Reserpine
>150
1.50
2.50
Hydrocortisone
Verapamil
ꢀ3.70 (0.35) ꢀ3.80 (0.35)
ꢀ5.90 (0.28) ꢀ4.40 (0.28)
Warfarin
Propranolol
ND: not determined
hERG Channel Inhibition. A concern with lipophilic weak base compounds is their
potential cardiotoxicity, with QT prolongation caused by inhibition of hERG channels.
Gleeson has shown that compounds with MW below 400 and cLogP below 4 are statistically
less likely to exhibit hERG toxicity.17 In accordance with this, predicted hERG IC50 values
(Table 2) obtained using StarDropTM software showed an improvement in predicted pIC50
(6.22 – 6.43) compared to the parent compound 1 (6.55). The predicted pIC50 values compare
favorably with the clinical antimalarials CQ (6.00) and lumefantrine (6.64) and are
substantially lower than that of halofantrine (7.37), an antimalarial with known
cardiotoxicity. One of the compounds, 2, was selected for experimental measurement of
ACS Paragon Plus Environment

Page 9 of 41
Journal of Medicinal Chemistry
1
2
3
hERG inhibition in CHO cells with stable hERG expression. This compound exhibited an
4
5
6
experimental hERG IC50 of 0.87 ± 0.08 ꢁM (pIC50 = 6.06), comparable to verapamil used as
7
8
9
a control (IC50 = 0.56 ± 0.08 ꢁM). The observed value was in good agreement with the
predicted pIC50 of 6.28, supporting the prediction that these compounds have lower hERG
liability than halofantrine, but that it would nonetheless remain a concern for this series.
Potential Vacuolar Accumulation. The vacuolar accumulation ratio (VAR) and lipid
accumulation ratio (LAR) have been implicated in predicting the activity of hemozoin
inhibiting antimalarials and in determining crossꢀresistance with CQ. For this reason,
experimental LogD values were used to calculate the VAR and LAR as described previously
by Warhurst and coꢀworkers.19 These values are presented in Table 2, together with MW and
cLogP data. A notable feature is that the VAR values are more than two orders of magnitude
lower than that reported by Warhurst for CQ (143,482), while the LAR values are more than
an order of magnitude larger than CQ.19, 20 The three compounds with the largest VAR values
(1, 2 and 5) also have the highest LAR values. This does not correspond to a high cLogP in
the case of 2 or 5 as these values are comparable to the other compounds in the respective
series (3 and 4; 6 and 7 respectively). The increased LAR must therefore arise from a lower
pKa in these compounds. Notably, both have an ortho substituent arrangement in the side
chain allowing for intramolecular hydrogen bonding between the 4ꢀamino group on the
quinoline and the deprotonated tertiary amino group in the side chain.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 41
1
2
3
4
5
6
Table 2: Predicted VAR, LAR, MW, cLogP and predicted hERG inhibition pIC50 (–logIC50)
7
8
9
of test compounds and five reference antimalarial compounds
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound
VAR
LAR
Molecular cLogP
Weight
hERG
IC50
p
170
776
45
3630
6165
295
708
6309
562
692
ND
439.4
402.9
402.9
402.9
395.5
395.5
395.5
319.8
528.9
378.3
500.4
444.3
6.36
4.81
4.86
4.86
3.66
3.82
3.71
3.81
9.12
3.95
7.65
6.08
6.55
6.28
6.23
6.43
6.23
6.22
6.39
6.00
6.64
5.72
7.37
6.52
1.HCl
2
3
102
977
79
4
5
6
69
7
ND
ND
ND
ND
ND
CQ
ND
Lumefantrine
Mefloquine
Halofantrine
ND
ND
ND
N
ꢀdesbutylhalofantrine
VAR = vacuolar accumulation ratio = antilog(logD7.4 – logD5.0). LAR = lipid accumulation
ratio = antilog(logD7.4); hERG pIC50s were obtained from StarDrop™ while cLogP values
were computed with ChemDrawTM Ultra 14.0; ND = Not determined
Activity Against P. falciparum in Vitro. All compounds showed potent antiplasmodial
activity in the nanomolar range against CQS and CQR parasite strains (Table 3). Chloroꢀ
substituted analogues had higher potency across all strains than cyano, an unsurprising
observation as it has been previously shown that halogens exert optimum influence at 7ꢀ
ACS Paragon Plus Environment

Page 11 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
position on the quinoline ring due to their strong lipophilicity and a moderately strong
electronꢀwithdrawing capacity.16 Against Dd2, all compounds were more active than CQ
while only 1, 3, 4 and 7 were more potent against the 7G8 strain. The slightly weaker activity
seen with 7G8 could possibly be attributed to differences in PfCRT arising from one or more
of the seven amino acids that differ between the Dd2 and 7G8 forms of PfCRT expressed in
the two strains or differential regulation in the pathways that compensate for these mutations
in these strains.21
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
There was no indication of crossꢀresistance between the test compounds and CQ in the
two strains of CQR parasites as indicated by their resistance indices which were all below 5
in the test compounds, but above 20 for CQ. This low crossꢀresistance with CQ could be a
result of the incorporation of the pyridodibemequine moiety as a side chain. This side chain
could bind to and subsequently inhibit transport of the molecule by PfCRTCQR or bind to
PfCRTCQR but simply fail to be translocated because of the altered side chain, or the altered
side chain may result in the molecule not being recognized at all.
These compounds also showed low cytotoxicity in a mammalian cell line (CHO cells)
with a selectivity index range between 146 and 740 (Table 3), highlighting the specificity of
their activity. By contrast, emetine, a potent inhibitor of protein synthesis in eukaryotic cells,
exhibited cellular toxicity at extremely low concentrations. There was no discernible trend in
cytotoxicity attributable to any structural features of the molecules, perhaps due to the limited
number of derivatives investigated.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 41
1
2
3
4
5
6
7
8
Table 3: Antiplasmodial activity, cytotoxicity profiles and βꢀhematin inhibition activity of
9
the series
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound
In Vitro IC50 (nM)
NF54 Dd2 7G8 R.IDd2
Cytotoxicity
βHIAIC50 (SD)
R.I 7G8 CHOIC50 (ꢁM) S.Ib
(ꢁM)
a
14.4
54.4
30.1
38.5
44.5
69.6
3.1
2.5
4.1
3.9
1.3
1.1
1.9
27.3
2.6
ND
4.8
3.4
2.8
3.0
2.4
1.9
1.8
21.9
1.4
ND
9.5
40.2
8.3
660
740
275
408
218
639
146
ND
ND
ND
13.4 (1.8)
17.9 (3.2)
20.8 (0.6)
17.2 (0.3)
16.6 (0.4)
16.7 (0.1)
24.8 (2.1)
26.1 (3.4)
ND
1.HCl
136.6 184.3
124.0 83.8
150.7 114.2
2
3
15.7
27.6
50.6
11.2
ND
4
126.6 162.0 307.1
5
79.1
76.4
8.3
89.4 151.0
145.5 134.0
226.4 150.3
6
7
CQ c
5.6
14.3
ND
7.7
ND
Artesunate
ND
ND
0.095
ND
Emetine
a
b
R.I = Resistance index = [IC50 CQR strain] / [IC50 CQS strain]; SI= Selectivity index =
[IC50 CHO] / [IC50 PfNF54]; ND = Not Determined. βHIA = βꢀhematin inhibition activity; c
Chloroquine diphosphate salt.
In view of the fact that they encompass the CQ pharmacophore, these compounds were
evaluated for their ability to inhibit formation of βꢀhematin using a detergentꢀbased assay that
mimics the conditions of the parasite acidic DV. All compounds were found to be more
potent inhibitors of βꢀhematin formation than CQ (Table 3). Analysis of the relationship
between antiplasmodial activity and inhibition of βꢀhematin formation reveals no correlation
ACS Paragon Plus Environment

Page 13 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
between parasite IC50 and βꢀhematin inhibition IC50. On the other hand, we observed a
significant increase in in vitro activity (pIC50) with increasing cLogP in the CQS strain (R2 =
0.91, P = 0.008, see Figure 2a). A similar correlation was also observed in both CQR strains,
but neither reached significance at the 95% confidence level (Dd2: R2 = 0.51, P = 0.071;
7G8: R2 = 0.55, P = 0.057). These findings suggest that the potency of these quinolines likely
hinges more on access to the vacuolar compartment than on strength of inhibition of
hemozoin. This may reflect the larger variations in cLogP values than in βHIA IC50s in this
series.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2: Plot showing correlation between: (a) biological activity (pIC50) and lipophilicity
(cLogP) in the CQS NF54 strain and (b) RI and cLogP in the two CQR strains, Dd2 (solid
circles and line) and 7G8 (open squares and dashed line).
While no correlations were seen between parasite growth IC50 or RI and either the VAR
or LAR, a trend of increasing RI with cLogP was seen in both the Dd2 and 7G8 strains
relative to NF54, albeit only significant in the case of the 7G8 strain (R2 = 0.91, P = 0.0007)
and not the Dd2 strain (R2 = 0.44, P = 0.10). Interestingly, in this series the effect of high
lipophilicity seems to be to reduce activity in the CQR strains, the opposite of what has been
reported for other quinolines.19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 41
1
2
3
4
5
6
7
8
Intracellular Hemozoin Inhibition. To establish whether this series of compounds
actually inhibits intracellular hemozoin formation in P. falciparum, we used a cellular heme
fractionation assay to determine whether or not there was a doseꢀdependent effect of the
compounds on the various heme components (hemoglobin, free heme and hemozoin) in the
parasite.22 Exposure of synchronized ring stage parasites to the test compounds for 32 h
caused statistically significant doseꢀdependent decreases in the fraction of total heme present
as hemozoin and a corresponding increase in the fractions of “free” or exchangeable heme for
all the compounds in a manner similar to CQ (Figure 3). This doseꢀdependent profile
persisted for all compounds when the amount of heme per cell, rather than proportion, was
determined by dividing the heme measurement (expressed as mass of heme Fe) by cell count
obtained through flow cytometry (Supplementary Figure 2). Parasite survival curves overlaid
on the free heme Fe per cell showed that an increase in free heme corresponded to a decrease
in parasite survival (Supplementary Figure 2). There was no significant perturbation in the
quantities free heme or hemozoin in parasites exposed to pyrimethamine, which is a nonꢀ
hemozoin inhibiting control.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In Vivo Antimalarial Efficacy. In view of the favorable in vitro activity and safety
profiles, we went on to screen a selection of compounds for in vivo oral activity against
Plasmodium berghei using the Peters 4 day suppressive test performed at 30 mg/kg doses
over 1ꢀ4 days postꢀinfection.23 All the compounds tested, except for 2, 5 and 6 achieved
>90% suppression of parasitemia according to this protocol, which is comparable with CQ
(Table 4). For comparison, the activity of the free base of 1 reported in our previous study is
also shown.15 The new series appears to exhibit similar efficacy.
ACS Paragon Plus Environment

Page 15 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3: Effects of dose of CQ (a), the parent dibemequine 1.HCl (b) and modified
dibemequines 2 (c), 3 (d), 4 (e), 5 (f), 6 (g) and 7 (h) on hemoglobin (left panels), free heme
(middle panels) and hemozoin (right) panels in the CQS NF54 strain of P. falciparum.
Increase relative to control in Hb and heme and decrease in hemozoin: * P < 0.05; ** P <
0.01; *** P < 0.001.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 41
1
2
3
Table 4: In vivo antimalarial activity of test compounds in P. bergheiꢀinfected mice.
4
5
6
7
8
9
Compound Dose (mg/kg) % Parasitized RBCs
% of Control
97.5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.HCl
4 × 30
4 × 30
4 × 30
4 × 30
4 × 30
4 × 30
4 × 30
4 × 30
3 × 30
0.26 ± 0.05
1.8 ± 0.3
82.4
2
97.8
3
0.22 ± 0.11
0.88 ± 0.28
5.4 ± 1.5
91.4
4
47.1
5
87.3
6
1.3 ± 1.2
98.0
7
0.20 ± 0.18
0.9 ± 0.1
91.2
CQ
1a
99.9
0.10 ± 0.02
10.2 ± 1.1
Untreated
a Reference 15.
Metabolic Stability and Putative Metabolite Identification. Finally, the metabolic
stability of the entire series was tested. All the compounds exhibited very short degradation
halfꢀlives after single point incubation with mouse liver microsomes and high hepatic
clearance ratios (>0.90) suggesting notable susceptibility to hepatic metabolism (Table 5).
These compounds were thus all metabolically unstable with fractions metabolized
comparable to the metabolically labile standard midazolam. Varying the arrangement of the
side chain linker from the orthoꢀ to the metaꢀ or paraꢀ position had no significant effect on
compound stability as is sometimes expected for such changes arising from steric effects on
enzyme access to metabolically labile sites on the molecule. This high metabolic
susceptibility has implications on in vivo activity as plasma exposure is likely to be poor and
ACS Paragon Plus Environment

Page 17 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
bioavailability low following dosing. To further understand this rapid compound clearance,
mass spectrometry scan data from the metabolic stability assay for each compound was
screened for the presence of fragments corresponding to putative metabolites and routes/sites
of metabolism. The metabolite profiles of all compounds appeared to be similar, with Nꢀ
dealkylation on the tertiary amine seemingly the predominant pathway (metabolites 1 and 2).
Other metabolites detected included products arising from oxidation of the quinoline core or
pyridyl ring (metabolite 5), Nꢀdealkylation together with oxidative deamination (metabolite
6) as well as aldehyde reduction (metabolite 3) and oxidation of the aldehyde to carboxylic
acid (metabolite 4) as depicted in Figure 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5: In vitro metabolism in mouse liver microsomes
a
Compound
% remaining
(after 30 min)
Projected t1/2
(min)
Intrinsic
Clearance CL
(mL/min/kg)
1106.62
Predicted EH
3.57
2.32
2.96
2.08
2.08
1.91
3.63
1.33
93.55
6.15
5.49
5.89
5.35
5.35
5.10
6.22
4.74
172.25
0.92
0.93
0.93
0.93
0.93
0.94
0.90
0.94
0.32
1.HCl
1223.91
2
1138.53
3
1250.06
4
1249.33
5
1348.68
6
1101.28
7
Midazolamb
MMV390048b
1424.93
42.03
a Fraction metabolized; b metabolically labile and inert standards respectively.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 41
1
2
Metabolite 6
3
O
m/z = 296, 287
4
HN
5
6
R
N
7
8
N-dealkylation + oxidative
deamination + aldehyde
reduction
N-dealkylation + oxidative
deamination
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
X
X
HN
N
N
OH
HN
N
Oxidation
O
HN
N
R
R
N
R
Metabolite 5
Parent Compound
N-dealkylation
Metabolite 3
m/z = 409, 417, 418
R: Cl, CN
X: N, CH
m/z = 298, 289
HN
N
NH
N-dealkylation + oxidative
R
deamination + oxidation of aldehyde
to carboxylic acid
Metabolite 1
N-dealkylation
m/z = 312, 302
N-dealkylation
HN
NH2
O
HN
OH
R
N
Metabolite 2
R
N
m/z = 297, 288
Metabolite 4
m/z = 312, 303
Figure 4: Proposed metabolic fragmentation pattern of the representative compounds with
putative metabolite structures and masses based on mass spectrometric measurements.
DISCUSSION
The hypothesis that replacement of the terminal phenyl ring in the dibemequine series
with an oꢀpyridyl ring and of the 7ꢀCl group on the quinoline with a 7ꢀCN group would not
substantially reduce activity against Plasmodium either in vitro or in vivo was borne out in
this study. Compounds bearing a 7ꢀCl group had IC50 values against the CQS NF54 strain of
P. falciparum about twice that of the parent dibemequines, while the 7ꢀCN bearing
derivatives exhibited IC50s about four times higher than the parent compounds. In a four day
Peters test at 30 mg/kg, all the compounds measured were indistinguishable from each other,
from the parent dibemequine 1 and from CQ. Resistance indices (RIs) for the CQR Dd2 and
ACS Paragon Plus Environment

Page 19 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7G8 strains were comparable to, or better than the parent compound 1.HCl. All were below
5, suggesting no significant crossꢀresistance with CQ which had RIs above 20. All of the
compounds exhibited relatively low cytotoxicity against mammalian (CHO) cells, with IC50s
close to, or above 10 ꢁM. Selectivity indices for P. falciparum were all above 100 for the
NF54 strain. Thus, these derivatives retained their excellent activity and selectivity against
malaria parasites.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The prediction that these compounds would exhibit improved drugꢀlike properties over
the parent compounds was to some extent supported. All showed an improvement in kinetic
solubility relative to 1.HCl, albeit a small improvement. Similarly, small improvements in
predicted hERG liability were also seen. It should be noted that none of these improvements
were significant and the predicted hERG pIC50 of the parent compound 1 was itself not that
much different from CQ and was considerably lower than halofantrine. Overall, the success
of this strategy with regard to solubility and hERG liability was marginal and came at the cost
of somewhat reduced permeability.
All of the compounds were found to inhibit βꢀhematin formation in the NPꢀ40 detergentꢀ
based assay. The range of IC50 values was, however, very narrow (16.6 – 24.8 ꢁM) and
comparable to both the parent compound 1.HCl and CQ (13.4 and 26.1 ꢁM). No correlation
was found between IC50s for βꢀhematin and parasite growth inhibition. Rather, activity
against the parasite is strongly correlated with cLogP. This indicates that the activity of
compounds is predominantly determined by access to the interior of the cell as has been
observed before. Nonetheless, measurement of “free” heme and hemozoin levels in treated
parasites showed that “free” heme levels clearly rise and hemozoin levels drop with
increasing dose and that parasite survival is correlated with the increase in “free” heme,
strongly suggesting that hemozoin inhibition is the basis of activity of these compounds. The
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 41
1
2
3
4
5
6
7
8
9
effects of the very small differences in βꢀhematin inhibition activity in these compounds may
in fact be masked by the much larger influence of lipophilicity on access to the target.
In view of the good in vivo activity of these compounds, their high level of metabolic
instability in liver microsomes was noteworthy and suggested the activity was due to the
metabolites rather than the parent. The principal metabolites (1 and 2) retain the key
pharmacophore of the 4ꢀaminoquinoline antimalarials, namely a 4ꢀaminoꢀ7ꢀchloroquinoline
core with side chain bearing a basic amino group. We predicted that they are still likely to be
βꢀhematin inhibitors and will inhibit hemozoin in the parasite. Interestingly, their lower MW
and cLogP may in fact be beneficial for improved solubility. In view of this, we have gone on
to synthesize a series of the predicted metabolites 1 and 2 from compounds 1 – 4. These
predicted metabolites retain the βꢀhematin inhibitory properties of the parent compounds with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IC50 values <24 ꢁM and are active against P. falciparum in culture, with IC50 values ranging
between 9.2 and 25 nM in the metabolite 1 series and 9.8 – 231 nM in the metabolite 2 series
in the NF54 strain. Cytotoxicity in CHO cells covers a range of IC50 values from 5.0 – 7.4
ꢁM in metabolite series 1 and 10.7 – 168 ꢁM in metabolite series 2. These data confirm that
the metabolites are selectively strongly active against the parasite. Detailed studies are
underway and will be reported in a further publication.
CONCLUSIONS
Replacement of the terminal phenyl ring in the dibemequines with a 2ꢀpyridyl ring and
replacement of the 7ꢀCl group with a 7ꢀCN group resulted in a series of compounds that
retained both in vitro and in vivo antimalarial activity, had acceptable parasite selectivity and
minimal crossꢀresistance in CQR parasite strains. The resulting decrease in lipophilicity was
associated with only a slight decrease in predicted and experimental hERG liability, but the
ACS Paragon Plus Environment

Page 21 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
compounds were all highly metabolically labile. Indeed, the observed activity of these
compounds in mice suggests that the metabolites are responsible for this in vivo activity.
6
7
8
9
EXPERIMENTAL SECTION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General Methods. Solvents, acids, and other commercially obtained chemicals were
purchased from Kimix Chemicals or SigmaꢀAldrich. Thin layer chromatography was used to
monitor reactions using preꢀcoated silica gel plates (Merck F54 aluminum backed). Silica for
column chromatography was obtained from Merck, South Africa. Melting points were
1
determined using a ReichertꢀJung Thermovar hot stage microscope and are uncorrected. H
and 13C NMR spectra were recorded on a Varian Mercury spectrometer at 300 MHz or a
Varian Unity spectrometer at 400 MHz in d3ꢀchloroform with references 7.26 ppm for 1H and
13
77.00 ppm for C. Infrared spectra were recorded on a PerkinꢀElmer Paragon 1000 FTꢀIR
spectrophotometer in the range 3600ꢀ800 cmꢀ1. Mass spectra were recorded on a VG
Micromass 16F spectrometer operating at 70 eV with an accelerating voltage of 4 kV and a
variable temperature source. Accurate mass determinations were performed on a Kratos
Limited MS9/50 spectrometer. All mass spectra were obtained using electronꢀimpact
techniques. All final products were subjected to high resolution mass spectrometry to confirm
identity and analytical HPLC to confirm purity using a Spectra Series HPLC with
PhenomenexꢀLuna, 3 ꢁm C18 column, with a run time of 13 min (see Supporting
Information). Purity exceeded 95% in all cases. Synthesis of 1 was described in a previous
publication.
Procedure for the Syntheses of the Pyridodibemequine Analogues 2 ꢀ 7. To a solution
of the appropriate Nꢀ(nꢀ(aminomethyl)benzyl)ꢀN-methylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine 8, 9 or
10 (2.63 g, 13.28 mmol) in NMP (8 ml), was added dry K2CO3 (1.37 g, 9.96 mmol) and
triethylamine (2.31 ml, 16.70 mmol) in a sealed tube at room temperature and stirred for 5ꢀ10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 41
1
2
3
4
5
6
7
8
9
min under N2 atmosphere. Then, 4,7ꢀdichloroquinoline or 4ꢀchloroꢀ7ꢀcyanoquinoline (0.8 g,
3.3 mmol) was added to the reaction mixture under N2 atmosphere. After addition of the
quinoline, the sealed tube was capped and put into an oil bath with gradually heating to 90
°C. Once the temperature reached 90 °C, the reaction was allowed to stir for 48 h. Progress
was monitored by TLC using EtOAc: hexane (8: 2) as a mobile phase while UV, iodine and
ninhydrin were used to monitor the spots. After completion of reaction, the mixture was
allowed to cool to room temperature and then poured into saturated brine solution (60 ml)
and extracted with EtOAc (5ꢀ6 times with 50 ml each). The organic layer was collected and
dried over anhydrous MgSO4 and excess organic solvent was removed under vacuum. The
crude product was then purified by using SiO2 gel column chromatography. EtOAc:
petroleum ether (25 : 75 to 35 : 65) were used to remove unreacted 4,7ꢀdichloroquinoline or
4ꢀchloroꢀ7ꢀcyanoquinoline. Then using 40ꢀ70 % EtOAc in petroleum ether other unwanted
impurities were separated. Finally pure EtOAc or EtOAc: trimethylamine (95: 5) were used
to elute the product. Preparative TLC was used to further purify the products using EtOAc:
MeOH: EtOH (90: 6: 4) as mobile phase as required. Finally purity was confirmed by HPLC.
7ꢀChloroꢀNꢀ(2ꢀ((methyl(pyridinꢀ2ꢀylmethyl)amino)methyl)benzyl)quinolinꢀ4ꢀamine
(2). White solid powder (33%); m.p.: 125ꢀ127 °C; IR: 3241 (NH), 2922 (CH), 1578 (C=C),
1611 (C=N), 753 (C−N); ¹H NMR (CDCl3, 300 MHz) δ 8.52 (1H, ddd, J = 6.0, 3.0 1.8 Hz),
8.46 (1H, d, J = 5.7 Hz), 8.02 (1H, d, J = 1.8 Hz), 7.76 (1H, d, J = 9.0 Hz), 7.57 (1H, td, J =
7.5, 1.8 Hz), 7.43ꢀ7.40 (1H, m), 7.34ꢀ7.26 (3H, m), 7.20ꢀ7.15 (2H, m), 7.06 (1H, dd, J = 9.0,
2.1 Hz), 6.56 (1H, d, J = 5.7 Hz), 4.49 (2H, d, J = 5.4 Hz), 3.77 (2H, s), 3.70 (2H, s), 2.19
(3H, s); ¹³C NMR (CDCl3/CD3OD (4 drops), 101 MHz) δ 158.0, 151.1, 150.8, 149.4, 148.5,
137.1, 137.0, 136.4, 135.0, 132.0, 130.5, 128.3, 127.9, 127.8, 124.9, 123.8, 122.6, 122.4,
117.9, 99.0, 63.9, 61.3, 46.5, 42.3; HRMS m/z [M + H]+ calcd for C24H24ClN4 403.1684,
found 403.1682; HPLC purity: 97.7%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 23 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7ꢀChloroꢀNꢀ(3ꢀ((methyl(pyridinꢀ2ꢀylmethyl)amino)methyl)benzyl)quinolinꢀ4ꢀamine
(3). Light yellow solid (23%); m.p.: 50ꢀ52 °C; IR: 3272 (NH), 2924 (CH), 1611 (C=C), 1579
(C=N), 758 (C−N); ¹H NMR (CDCl3, 300 MHz) δ 8.51 (1H, dd, J = 3.9, 0.6 Hz), 8.32 (1H, d,
J = 6.0 Hz), 8.14 (1H, d, J = 9.0 Hz), 8.03 (1H, d, J = 2.1 Hz), 7.62 (1H, td, J = 7.5, 1.8 Hz),
7.44ꢀ7.15 (7H, m), 6.43 (1H, d, J = 6.0 Hz), 4.61 (2H, d, J = 4.5 Hz), 3.67 (2H, s), 3.60 (2H,
s), 2.24 (3H, s); ¹³C NMR (CDCl3, 101 MHz) δ 159.5, 150.6, 150.5, 149.1, 147.9, 140.0,
137.1, 136.3, 135.6, 128.9, 128.6, 128.0, 127.8, 126.2, 125.7, 122.9, 121.9, 121.7, 117.1,
99.6, 63.3, 61.9, 47.6, 42.5; HRMS m/z [M + H]+ calcd for C24H24ClN4 403.1684, found
403.1688; HPLC purity: 98.29%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7ꢀChloroꢀNꢀ(4ꢀ((methyl(pyridinꢀ2ꢀylmethyl)amino)methyl)benzyl)quinolinꢀ4ꢀamine
(4). Light yellow powder (36%); m.p.: 115ꢀ118 °C; IR: 3272 (NH), 2923 (CH), 1609 (C=C),
1578 (C=N), 758 (C−N); ¹H NMR (CDCl3, 300 MHz) δ 8.54 (1H, ddd, J = 5.1, 1.8, 0.9 Hz),
8.37 (1H, d, J = 6.0 Hz), 8.07 (1H, d, J = 9.0 Hz), 8.01 (1H, d, J = 2.1 Hz), 7.69 (1H, td, J =
9.0, 1.8 Hz), 7.50 (1H, d, J = 7.8 Hz), 7.37ꢀ7.31 (5H, m), 7.18 (1H, ddd, J = 7.2, 4.8, 0.9 Hz),
6.44 (1H, d, J = 6.0 Hz), 4.58 (2H, d, J = 4.8 Hz), 3.69 (2H, s), 3.59 (2H, s), 2.24 (3H, s); ¹³C
NMR (CDCl3, 101 MHz) δ 159.7, 150.7, 150.3, 149.1, 147.6, 139.0, 136.3, 135.8, 135.8,
129.6, 127.5, 127.5, 125.8, 123.0, 121.9, 121.8, 117.1, 99.5, 63.4, 61.7, 47.4, 42.5; HRMS
m/z [M + H]+ calcd for C24H24ClN4 403.1684, found 403.1682; HPLC purity: 98.6%.
4ꢀ((2ꢀ((Methyl(pyridinꢀ2ꢀylmethyl)amino)methyl)benzyl)amino)quinolineꢀ7ꢀ
carbonitrile (5). Off white solid powder (46%); m.p.: 145ꢀ148 °C; IR: 3259 (NH), 2962
(CH), 2230 (C≡N), 1588 (C=N), 1572 (C=C), 754 (C−N); ¹H NMR (CDCl3, 300 MHz) δ
8.55 (1H, d, J = 5.7 Hz), 8.51 (1H, dd, J = 5.7, 1.5 Hz), 8.36 (1H, d, J = 1.2 Hz), 7.99 (1H, d,
J = 8.7 Hz), 7.61 (1H, td, J = 7.5, 1.8 Hz), 7.43ꢀ7.41 (1H, m), 7.31ꢀ7.18 (6H, m), 6.65 (1H, d,
J = 5.7 Hz), 4.50 (2H, d, J = 5.4 Hz), 3.78 (2H, s), 3.71 (2H, s), 2.17 (3H, s); ¹³C NMR
(CDCl3, 101 MHz) δ 158.1, 153.2, 151.2, 149.0, 146.2, 137.0, 136.7, 136.5, 133.2, 131.8,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 41
1
2
3
4
5
6
7
8
129.5, 128.3, 127.9, 125.2, 124.2, 123.2, 122.7, 121.7, 118.2, 112.9, 100.5, 63.5, 61.1, 45.8,
42.1; HRMS m/z [M + H]+ calcd for C25H24N5 394.2026, found 394.2032; HPLC purity:
96.0%.
9
4ꢀ((3ꢀ((Methyl(pyridinꢀ2ꢀylmethyl)amino)methyl)benzyl)amino)quinolineꢀ7ꢀ
carbonitrile (6). Off white solid powder (34%); m.p.: 153ꢀ158 °C; IR: 3271 (NH), 2927
(CH), 2228 (C≡N), 1587 (C=N), 1570 (C=C), 756 (C−N); ¹H NMR (CDCl3, 300 MHz) δ
8.58 (1H, d, J = 5.4 Hz), 8.53 (1H, d, J = 4.8 Hz), 8.34 (1H, s), 8.03 (1H, d, J = 8.4 Hz), 7.64
ꢀ7.18 (8H, m), 6.55 (1H, d, J = 5.4 Hz), 5.88 (1H, bs, NH), 4.56 (2H, d, J = 4.8 Hz), 3.74
(2H, s), 3.69 (2H, s), 2.31 (3H, s); ¹³C NMR (CDCl3/CD3OD (3 drops) 101 MHz) δ 158.7,
151.9, 150.8, 148.7, 146.9, 139.1, 137.4, 137.1, 134.2, 128.9, 128.7, 128.0, 126.4, 125.5,
123.6, 122.7, 122.5, 121.6, 118.5, 113.0, 101.1, 62.7, 61.9, 47.2, 42.3; HRMS m/z [M + H]+
calcd for C25H24N5 394.2026, found 394.2034; HPLC purity: 99.4%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4ꢀ((4ꢀ((Methyl(pyridinꢀ2ꢀylmethyl)amino)methyl)benzyl)amino)quinolineꢀ7ꢀ
carbonitrile (7). Off white solid powder (29%); m.p.: 152ꢀ155 °C; IR: 3289 (NH), 2661
(CH), 2228 (C≡N), 1588 (C=N), 1570 (C=C), 758 (C−N); ¹H NMR (CDCl3, 300 MHz) δ
8.57 (2H, m), 8.35 (1H, s), 8.05 (1H, d, J = 8.7 Hz), 7.70 (1H, td, J = 7.8, 1.5 Hz), 7.57 – 7.50
(2H, m), 7.43 – 7.33 (4H, m), 7.20 (1H, dd, J = 6.7, 5.6 Hz), 6.55 (1H, d, J = 5.4 Hz), 6.08
(1H, br, s, NH), 4.56 (2H, d, J = 5.1 Hz), 3.75 (2H, s), 3.67 (2H, s), 2.30 (3H, s); ¹³C NMR
(CDCl3/CD3OD (3 drops) 101 MHz) δ 158.9, 151.6, 150.9, 148.6, 146.6, 138.1, 137.1, 136.2,
134.0, 129.7, 127.3, 125.5, 123.6, 122.7, 122.5, 121.5, 118.4, 113.0, 101.1, 62.9, 61.6, 47.0,
42.2, ; HRMS m/z [M + H]+ calcd for C25H24N5 394.2026, found 394.2029;HPLC purity:
99.0%.
Procedure for the Syntheses of Amine Intermediates 8 ꢀ 10. To a stirred solution of
the appropriate Nꢀ(nꢀ(azidomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine 18, 19 or
20 (2.0 g, 7.49 mmol) in THF (30 ml) was added PPh3 (1.57 g, 5.99 mmol) at room
ACS Paragon Plus Environment

Page 25 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
temperature under nitrogen atmosphere and the reaction mixture was allowed to stir for
additional 30 min. Water was added (6.79 ml, 376 mmol) and the reaction mixture was
refluxed for 6ꢀ10 h. Reaction progress was monitored by TLC using EtOAc: MeOH : Et3N (5
: 4 : 1) as a mobile phase and ninhydrin and iodine as stain. Excess THF was removed under
high vacuum while the remaining water was removed as an azeotrope with toluene. The
crude product was purified by SiO2 gel column chromatography using EtOAc and petroleum
ether as an eluent to remove impurities and finally the desired product was obtained using
EtOAc : MeOH : Et3N (85 : 10 : 5) as eluent.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
ꢀ(2ꢀ(Aminomethyl)benzyl)ꢀN-methylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine (8). Dark brown
liquid (1.40 g, 70%), ¹H NMR (CDCl3, 300 MHz) δ 8.77 (1H, ddd, J = 4.8, 1.8, 0.9 Hz), 7.74
(1H, dd, J = 7.2, 1.5 Hz), 7.64 (1H, td, J = 7.8, 1.8 Hz), 7.38 – 7.16 (5H, m), 4.21 (2H, s),
3.68 (2H, s), 3.62 (2H, s), 2.25 (3H, s); ¹³C NMR (CDCl3, 101 MHz) δ 156.9, 150.2, 137.7,
136.7, 133.5, 133.1, 131.1, 129.0, 129.0, 123.4, 122.5, 62.0, 61.4, 42.2, 42.2; EIꢀMS m/z M+
calcd for C15H19N3 241.16, found 241.05 (83).
N
ꢀ(3ꢀ(Aminomethyl)benzyl)ꢀN-methylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine (9). Light brown
solid (1.63 g, 76%); ¹H NMR (CDCl3, 300 MHz) δ 8.53 (1H, dd, J = 4.8, 0.9 Hz), 7.66 (1H,
td, J = 7.8, 1.8 Hz), 7.47ꢀ7.23 (5H, m), 7.16 (1H, ddd, J = 7.5, 4.8, 0.9 Hz), 4.07 (2H, s), 3.60
(2H, s), 3.50 (2H, s), 2.13 (3H, s); ¹³C NMR (CDCl3, 101 MHz) δ 159.0, 149.0, 139.1, 136.6,
131.4, 129.1, 128.9, 128.7, 127.4, 123.1, 122.1, 63.3, 61.7, 44.3, 42.4; EIꢀMS m/z M+ calcd
for C15H19N3 241.16, found 241.06 (73); Anal. calc. for C15H19N3: C, 74.65%; H, 7.94%; N,
17.41%. Found: C, 74.41%; H, 7.41%; N, 17.64%.
N
ꢀ(4ꢀ(Aminomethyl)benzyl)ꢀN-methylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine
(10).
Dark
brown solid (1.85 g, 92%); ¹H NMR (CDCl3, 300 MHz) δ 8.54 (1H, m), 7.68 (1H, td, J = 7.8,
2.1 Hz), 7.52ꢀ7.27 (5H, m), 7.16 (1H, m), 3.89 (2H, s), 3.69 (2H, s), 3.59 (2H, s), 2.25 (3H,
s); ¹³C NMR (CDCl3, 101 MHz) δ 161.6, 149.0, 137.5, 136.4, 135.1, 129.1, 128.2, 123.0,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 41
1
2
3
4
5
6
7
8
122.0, 63.3, 61.8, 42.5, 42.4; EIꢀMS m/z M+ calcd for C15H19N3 241.16, found 241.05 (65);
Anal. calc. for C15H19N3: C, 74.65%; H, 7.94%; N, 17.41%. Found: C, 74.57%; H, 6.98%; N,
16.91%.
9
Procedure for the Synthesis of
Nꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine (17). To a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
stirred solution of pyridineꢀ2ꢀcaboxaldehyde (8.89 mL, 93.4 mmol) in MeOH (100 ml) at
room temperature, 40% aq. methylamine (12.03 mL, 140 mmol) was added slowly and the
reaction allowed to progress at 65 °C for 2 h. The reaction mixture was allowed to cool to 0
°C, then NaBH4 (2.48 g, 65.4 mmol) was added slowly and reaction continued for 6 h at
room temperature. Reaction progress was monitored by TLC. After completion of the
reaction, the remaining solvent was removed under high vacuum. The crude reaction mixture
was worked up by the addition of a saturated solution of aq. Na2CO3 and the product
extracted into CHCl3 (3 × 100 ml) and the organic extracts dried over anhydrous MgSO4. The
solvent was then removed under reduced pressure to give a crude product (11.8 g), which was
chromatographed on silicaꢀgel using 5% MeOH + 5% TEA + DCM as eluent. The product
was dried under vacuum to give Nꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine as a yellow oil (9.2
g, 74%):24 1H NMR (300 MHz, CDCl3) δ 8.49 (dd, 1H, J = 4.8, 0.9 Hz), 7.55 (td, 1H, J = 7.5,
1.5 Hz), 7.25 (m, 1H), 7.10 (m, 1H), 3.70 (s, 2H), 2.35 (s, 3H); 13C NMR (CDCl3,75 MHz) δ
159.5, 149.2, 136.4, 122.3, 121.2, 60.00, 35.9. EIꢀMS m/z M+ calcd for C7H10N2 122.08,
found 122.08.
Procedure for the Syntheses of Azido Intermediates 18 – 20. The synthesis of 1ꢀ
(azidomethyl)ꢀ2, 3 or 4ꢀ(bromomethyl)benzene intermediates (11, 12 and 13) from
commercially available dibromoxylenes 14, 15 and 16 has been previously described.18 To a
stirred solution of crude 11, 12 or 13 (2.5 g, 11.01 mmol) and K2CO3 (3.78 g, 27.53 mmol) in
anhydrous acetonitrile (50 ml) at 0 °C, was added Nꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine
(17, 1.07 g, 8.81 mmol) and the reaction heated and allowed to progress under reflux
ACS Paragon Plus Environment

Page 27 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
overnight. Acetonitrile was then removed under reduced pressure and the reaction mixture
worked up with ethyl acetate (100 ml) and saturated aqueous Na2CO3 (3 × 50 ml), dried
(MgSO4) and concentrated under vacuum. The crude product was purified by SiO2 gel
column chromatography using mixtures of ethyl acetate: hexane (50: 50) to (80: 20) as eluent
to give the Nꢀ(nꢀ(azidomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamines (n = 2 – 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
ꢀ(2ꢀ(Azidomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine
(18).
Dark
yellow oil (1.45 g, 73%); ¹H NMR (CDCl3, 300 MHz) δ 8.58 (dd, 1H, J = 4.8, 0.9 Hz), 7.70
(td, 1H, J = 7.5, 1.5 Hz), 7.51 – 7.48 (m, 2H), 7.33 ꢀ 7.20 (m, 2H), 4.51 (s, 2H), 3.75 (s, 2H),
3.65 (s, 2H), 2.22 (s, 3H); ¹³C NMR (CDCl3, 101 MHz) δ 158.5, 149.0, 136.9, 136.5, 135.0
130.9, 129.5, 128.2, 127.8, 123.1, 122.1, 63.7. 59.9, 52.1, 42.4; EIꢀMS m/z M+ calcd for
C15H17N5 267.15, found 267.23 (75).
N
ꢀ(3ꢀ(Azidomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine
(19).
Light
yellow liquid (1.90 g, 96%); ¹H NMR (CDCl3, 300 MHz) δ 8.58 (dd, 1H, J = 4.8, 0.9 Hz),
7.70 (td, 1H, J = 7.5, 1.5 Hz), 7.56 (d, 1H, J = 7.8 Hz), 7.40 – 7.18 (m, 5H), 4.35 (s, 2H),
3.81 (s, 2H), 3.73 (s, 2H), 2.33 (s, 3H); ¹³C NMR (CDCl3, 101 MHz) δ 158.8, 149.0, 139.0,
136.4, 135.5, 128.9, 128.7, 128.7, 126.9, 123.0, 122.0, 63.3, 61.7, 54.7, 42.4; EIꢀMS m/z M+
calcd for C15H17N5 267.15, found 267.08 (66).
N
ꢀ(4ꢀ(Azidomethyl)benzyl)ꢀNꢀmethylꢀ1ꢀ(pyridinꢀ2ꢀyl)methanamine
(20).
Light
yellow liquid (1.62 g, 60%); ¹H NMR (CDCl3, 300 MHz) δ 8.58 (1H, dt, J = 4.8 0.9 Hz), 7.70
(1H, td, J = 7.8, 1.8 Hz), 7.58 (1H, d, J = 7.8 Hz), 7.47 (2H, d, J = 8.1 Hz), 7.31ꢀ7.26 (2H,
m), 7.21 (1H, m), 4.29 (2H, s), 3.84 (2H, s), 3.78 (2H, s), 2.35 (3H, s); ¹³C NMR (CDCl3, 101
MHz) δ 158.3, 149.0, 138.0, 136.4, 134.5, 129.4, 128.2, 123.0, 122.0, 63.3, 61.6, 54.6, 42.4;
EIꢀMS m/z M+ calcd for C15H17N5 267.15, found 267.05 (72).
In Vitro Antiplasmodial Activity Tests using pLDH. The pLDH method for measuring
parasite survival was based on that of Makler et al.25 All parasite strains were acquired from
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 41
1
2
3
4
5
6
7
8
9
MR4 (Malaria Research and Reference reagent Resource Centre, Manassas, VA). Briefly, the
assay procedure involved preparation of respective stock solutions of CQ diphosphate and
test compounds to 2 mg/mL in distilled water (for CQ) and 100% DMSO (for test
compounds) then stored at ꢀ20 °C and further dilutions prepared on the day of the
experiment. Synchronized trophozoiteꢀstage cultures of NF54 (CQS), Dd2 (CQR) and 7G8
(CQR) were prepared to 2% parasitemia and 2% hematocrit. Compounds were tested at
starting concentrations of 10,000 ng/mL (1,000 ng/mL for CQ), which were then serially
diluted 2ꢀfold in complete medium to give 10 concentrations with final volumes of 200 ꢁL in
each well. Plates were covered with a sterile lid, placed in a gas chamber and incubated at 37
°C for 48 hours after which 100 ꢁL of MalStat reagent was added to a new plate and 15 ꢁL of
reꢀsuspended culture added to each well of this plate followed by addition of 25 ꢁL of NBT
(nitro blue tetrazolium chloride). The plates were kept in the dark for about 10 min to fully
develop and absorbance measured at 620 nm on a microplate reader. Raw data were exported
to Microsoft Excel and doseꢀresponse analysis performed in GraphPad Prism v.6 (GraphPad
Software Inc., La Jolla, USA).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In
vitro
Cytotoxicity
Assay.
The
3ꢀ(4,5ꢀdimethylthiazolꢀ2ꢀyl)ꢀ2,5ꢀ
diphenyltetrazoliumbromide (MTT)ꢀassay was used to screen compounds for in vitro
cytotoxicity against Chinese Hamster Ovarian (CHO) mammalian cellꢀlines as previously
described.26 Briefly, test compounds were first prepared to a 20 mg/mL stock solution in
100% DMSO followed by further dilutions in complete medium. The initial concentration of
the compounds was 100 ꢁg/mL, which was serially diluted in complete medium with 10ꢀfold
dilutions to give 6 concentrations, the lowest being 0.001 ꢁg/mL. Emetine was used as the
reference drug in this experiment and the same dilution technique was applied to the control,
and testing was performed in triplicate on each occasion. The 50% inhibitory concentration
ACS Paragon Plus Environment

Page 29 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
(IC50) values were obtained from full doseꢀresponse curves, using nonꢀlinear doseꢀresponse
curve fitting analysis via GraphPad Prism v.6 (GraphPad Software Inc., La Jolla, USA).
6
7
8
9
In vivo Peters’ Suppressive 4 Day Test. The CQS P. berghei (ANKA strain) was used
to assess in vivo antiꢀmalarial efficacy of the test compounds. The parasites were maintained
in a C57BL/6 mouse by inoculation with 250 µl of a 1:1 (v/v) suspension of erythrocytes
infected with P. berghei in phosphate buffered saline (PBS). On the day of the experiment the
host mouse was anaesthetized intraperitoneally with a mixture of ketamine (120 mg/kg) and
xylazine (16 mg/kg). Whole blood from the host mouse was drawn by cardiac puncture into a
Vacuette® heparin tube and a suspension of P. berghei parasitized erythrocytes (1 × 107) in
PBS was prepared and the test mice were infected with 200 µL of this suspension
intraperitoneally.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The curative potential of the test compounds was evaluated using Peters’ 4ꢀday test as
described elsewhere.23 Briefly, the mice were dosed orally at 30 mg/kg 2 h postꢀinfection and
for three consecutive days (D0 to D3). CQ was used as a reference drug and was also dosed
orally at 30 mg/kg. The control group of mice was also dosed orally with PBS. Parasitemia
was determined from the third day (D4) by preparing thin blood films from the tail of each
mouse and the smears were fixed with methanol and stained with Giemsa. Equations 1 and 2
were used to calculate % parasitemia and reduction.
% ꢀꢁꢂꢁꢃꢄꢅꢆꢇꢄꢁ = ꢈꢉ.ꢉꢊ ꢋꢌꢍꢌꢎꢏꢐꢏꢑꢒꢓ ꢔꢕꢖꢎ × 100
1
2
ꢗꢉꢐꢌꢘ ꢈꢉ.ꢉꢊ ꢔꢕꢖꢎ
% ꢂꢆꢙꢚꢛꢅꢄꢜꢝ = ꢋꢌꢍꢌꢎꢏꢐꢒꢞꢏꢌ ꢉꢊ ꢋꢘꢌꢟꢒꢠꢉꢡꢋꢌꢍꢌꢎꢏꢐꢒꢞꢏꢌ ꢉꢊ ꢐꢒꢎꢐ ꢟꢉꢞꢋꢉꢢꢈꢓ × 100
ꢋꢌꢍꢌꢎꢏꢐꢒꢞꢏꢌ ꢉꢊ ꢋꢘꢌꢟꢒꢠꢉ
Parasite heme fractionation assay. The parasite heme fractionation assay examines the
doseꢀdependent effect (if any) of an inhibitor on the fate of total heme in the parasite as
previously described.3, 22 Briefly, cultures were synchronized at 48ꢀh intervals with 5%
ACS Paragon Plus Environment