Antimalarial agents against both sexual and asexual parasites stages: structure-activity relationships and biological studies of the Malaria Box compound 1-[5-(4-bromo-2-chlorophenyl)furan-2-yl]-N-[(piperidin-4-yl)methyl]methanamine (MMV019918) and analogues

Article abstract of DOI:10.1016/j.ejmech.2018.03.024

Therapies addressing multiple stages of Plasmodium falciparum life cycle are highly desirable for implementing malaria elimination strategies. MMV019918 (1, 1-[5-(4-bromo-2-chlorophenyl)furan-2-yl]-N-[(piperidin-4-yl)methyl]methanamine) was selected from the MMV Malaria Box for its dual activity against both asexual stages and gametocytes. In-depth structure-activity relationship studies and cytotoxicity evaluation led to the selection of 25 for further biological investigation. The potential transmission blocking activity of 25 versus P. falciparum was confirmed through the standard membrane-feeding assay. Both 1 and 25 significantly prolonged atrioventricular conduction time in Langendorff-isolated rat hearts, and showed inhibitory activity of Ba²⁺ current through Ca_v1.2 channels. An in silico target-fishing study suggested the enzyme phosphoethanolamine methyltransferase (PfPMT) as a potential target. However, compound activity against PfPMT did not track with the antiplasmodial activity, suggesting the latter activity relies on a different molecular target. Nevertheless, 25 showed interesting activity against PfPMT, which could be an important starting point for the identification of more potent inhibitors active against both sexual and asexual stages of the parasite.

Full text of DOI:10.1016/j.ejmech.2018.03.024

Accepted Manuscript
Antimalarial agents against both sexual and asexual parasites stages: structure-
activity relationships and biological studies of the Malaria Box compound 1-
[5-(4-bromo-2-chlorophenyl)furan-2-yl]-N-[(piperidin-4-yl)methyl]methanamine
(MMV019918) and analogues
Alessandra Vallone, Sarah D'Alessandro, Simone Brogi, Margherita Brindisi, Giulia
Chemi, Gloria Alfano, Stefania Lamponi, Soon Goo Lee, Joseph M. Jez, Karin J.M.
Koolen, Koen J. Dechering, Simona Saponara, Fabio Fusi, Beatrice Gorelli, Donatella
Taramelli, Silvia Parapini, Reto Caldelari, Giuseppe Campiani, Sandra Gemma,
Stefania Butini
PII:
S0223-5234(18)30267-8
10.1016/j.ejmech.2018.03.024
EJMECH 10289
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 24 July 2017
Revised Date: 6 March 2018
Accepted Date: 8 March 2018
Please cite this article as: A. Vallone, S. D'Alessandro, S. Brogi, M. Brindisi, G. Chemi, G. Alfano,
S. Lamponi, S.G. Lee, J.M. Jez, K.J.M. Koolen, K.J. Dechering, S. Saponara, F. Fusi, B. Gorelli, D.
Taramelli, S. Parapini, R. Caldelari, G. Campiani, S. Gemma, S. Butini, Antimalarial agents against
both sexual and asexual parasites stages: structure-activity relationships and biological studies of the
Malaria Box compound 1-[5-(4-bromo-2-chlorophenyl)furan-2-yl]-N-[(piperidin-4-yl)methyl]methanamine
(MMV019918) and analogues, European Journal of Medicinal Chemistry (2018), doi: 10.1016/
j.ejmech.2018.03.024.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

ACCEPTED MANUSCRIPT
Antimalarial agents against both sexual and asexual parasites stages: structure-
activity relationships and biological studies of the Malaria Box compound 1-[5-(4-
bromo-2-chlorophenyl)furan-2-yl]-N-[(piperidin-4-yl)methyl]methanamine
(MMV019918) and analogues
1

ACCEPTED MANUSCRIPT
Antimalarial agents against both sexual and
asexual parasites stages: structure-activity
relationships and biological studies of the
Malaria Box compound 1-[5-(4-bromo-2-
chlorophenyl)furan-2-yl]-N-[(piperidin-4-
yl)methyl]methanamine (MMV019918) and
analogues
⊥
,1,2,4
⊥ ,2,4,5
1,2,4
Alessandra Vallone
Sarah D’Alessandro
Simone Brogi,
Margherita
1,2,4
1,2,4
1,2,4
1,2
6
Brindisi,
Giulia Chemi,
Gloria Alfano,
Stefania Lamponi, Soon Goo Lee,
6
7
7
3
Joseph M. Jez, Karin J.M. Koolen, Koen J. Dechering, Simona Saponara, Fabio
3
3
2,4,8
2,4,8
9
Fusi, Beatrice Gorelli, Donatella Taramelli,
Silvia Parapini,
Reto Caldelari,
1,2,4
1,2,4
1,2,4
Giuseppe Campiani,*
Sandra Gemma,*
Stefania Butini
1
2
Department of Biotechnology, Chemistry and Pharmacy, European Research Centre
3
for Drug Discovery and Development (NatSynDrugs), Department of Life Sciences,
University of Siena, via Aldo Moro 2, 53100, Siena, Italy
1

ACCEPTED MANUSCRIPT
4
5
Centro Interuniversitario di Ricerche sulla Malaria (CIRM), University of Perugia,
Perugia, Italy
Dipartimento di Scienze Biomediche, Chirurgiche e Odontoiatriche, Università di
Milano, Via Pascal 36, 20133 Milan, Italy
6
Department of Biology, Washington University in St. Louis, St. Louis, MO 63130
7
TropIQ Health Sciences, Transistorweg 5-C02, 6534 AT Nijmegen, The Netherlands
8
Dipartimento di Scienze Farmacologiche e Biomolecolari, Università di Milano, Via
Pascal 36, 20133 Milan, Italy
9
Institute of Cell Biology, University of Bern, Baltzerstrasse 4, 3012 Bern
*
To whom correspondence should be addressed: Sandra Gemma, gemma@unisi.it;
Giuseppe Campiani, campiani@unisi.it.
⊥
These authors equally contributed to the present work.
Abstract. Therapies addressing multiple stages of Plasmodium falciparum life cycle are
highly desirable for implementing malaria elimination strategies. MMV019918 (1, 1-[5-
(4-bromo-2-chlorophenyl)furan-2-yl]-N-[(piperidin-4-yl)methyl]methanamine)
was
selected from the MMV Malaria Box for its dual activity against both asexual stages and
gametocytes. In-depth structure-activity relationship studies and cytotoxicity evaluation
led to the selection of 25 for further biological investigation. The potential transmission
blocking activity of 25 versus P. falciparum was confirmed through the standard
membrane-feeding assay. Both 1 and 25 significantly prolonged atrioventricular
2

ACCEPTED MANUSCRIPT
2+
conduction time in Langendorff-isolated rat hearts, and showed inhibitory activity of Ba
current through Ca 1.2 channels. An in silico target-fishing study suggested the enzyme
v
phosphoethanolamine methyltransferase (PfPMT) as a potential target. However,
compound activity against PfPMT did not track with the antiplasmodial activity,
suggesting the latter activity relies on a different molecular target. Nevertheless, 25
showed interesting activity against PfPMT, which could be an important starting point for
the identification of more potent inhibitors active against both sexual and asexual stages
of the parasite.
Keywords: Malaria, Gametocytes, Plasmodium falciparum, inhibition, structure-activity
relationships
1. Introduction
Despite recent advances in the field of antimalarial treatment, malaria is still
responsible for a high number of infections and related deaths in tropical and sub-
tropical regions.[1] The antimalarial agenda has currently shifted from malaria control to
eradication/elimination strategies. To reach this ambitious goal and to complement
appropriate vector-control measures, such as larvicidal compounds and insecticide-
treated bed nets, prioritization of new approaches, targeting the transmission stages of
Plasmodium falciparum life cycle (i.e. the gametocytes) is necessary.[2] Development of
dual acting drugs active against asexual intraerythrocytic parasites, responsible for the
clinical symptoms of malaria, and gametocytes, the sexual forms of the parasite
responsible for malaria transmission, is a challenging goal but necessary for the
implementation of the above-mentioned strategies.
3

ACCEPTED MANUSCRIPT
High-throughput screening (HTS) assays have been successfully applied to P.
falciparum asexual stages for the discovery of novel hits endowed with innovative
modes of action.[3-14] Moreover, such approaches have been adapted for the
discovery of compounds active as gametocytocidal agents.[15-24] Screening
campaigns using these assays allowed a better analysis and characterization of the
anti-gametocyte properties of known antimalarials and for the discovery of potential hit
compounds useful for the development of transmission-blocking therapies.[25] The
implementation of novel methodologies of phenotypic screening for gametocytocidal
compounds was recently boosted by the non-profit foundation Medicine for Malaria
Venture (MMV), which produced and delivered a collection of 400 commercial
compounds divided into 200 drug-like and 200 probe-like structures, i.e., the “Open
Access Malaria Box”.[26] The structures present in the Malarial Box were selected using
results from phenotypic screening of large compound collections against P.
falciparum.[3, 27, 28] These phenotypic screens identified nearly 20,000 compounds
that were active against P. falciparum 3D7 and endowed with reasonable toxicity.
Selection criteria for compounds included in the Malaria Box were commercial
availability, maximization of the structural diversity, and potency.
A cell-based assay with P. falciparum gametocytes expressing a potent luciferase and
a lactate dehydrogenase (LDH)-based gametocyte assay were used to screen known
antimalarial compounds and those of the MMV Malaria Box.[23, 25, 26] From this initial
study, compound MMV019918 (1, Chart 1)[26] demonstrated dual activity against P.
falciparum asexual and sexual stages with IC50 < 1000 nM; an in vitro IC50 < 1000 nM
against late stages Plasmodium gametocytes consistently found in different published
4

ACCEPTED MANUSCRIPT
assays (and also in the structurally related compound N-[(5-(3-chloro-4-
fluorophenyl)furan-2-yl)methyl]-N’,N’-dimethylpropane-1,3-diamine
(MMV020505,
2));[25] activity on both male and female NF54 gametes; and properties favorable for
drug development, including a low molecular weight and a scaffold suitable for
optimization of efficacy. Starting from compound 1 and combining synthetic and
biological efforts, we investigated the structure-activity relationships (SAR) of this class
of compounds with the aim of understanding the general structural features important
for the observed activity. Our goal was also to investigate and possibly improve the
toxicity profile of the prepared analogues since the selectivity index (SI) of 1 was rather
low based on the available information, and a certain degree of hERG activity (83% of
inhibition at 50 µM) was also known for this compound.[25]
Chart 1. Structure of the reference compounds 1,2 and of the analogue 25.
The SAR investigation of 1 was based on the possibility of disconnecting its scaffold
into three modules: a hydrophobic head, a heterocyclic linker, and a pendant basic
chain (Chart 1). We introduced different substituents, at the hydrophobic head to
5

ACCEPTED MANUSCRIPT
explore the role of the two halogen substituents, replaced the central linker with different
heterocyclic scaffolds, and replaced the cyclic piperidine moiety with linear lateral
chains. Compound 25, showing comparable potency with respect to 1, characterized by
the presence of and extra aromatic ring at the hydrophobic head and endowed with
reasonable TC , was then selected for further biological studies.
50
2
. Results and Discussion
.1.Chemistry
We identified a synthetic strategy for the preparation of the hit compound 1 that was
2
also suitable for the preparation of a broad series of analogues bearing different
substitution patterns at the phenyl ring.
a
Scheme 1. General synthesis of compounds 1, 8-12
a
Reagents and conditions: (a) Bis(triphenylphosphine)Pd(II) dichloride, aq. Na CO ,
2
3
dimethoxyethane (DME), EtOH, reflux, 12 h, 33-83%; (b) BCl , dichloromethane (DCM),
3
-
78 °C, 3 h, 55%; (c) i. MeOH, reflux, 12 h; ii. NaBH , 0 °C, 1 h; (d) 1 N HCl, MeOH, 25
4
6

ACCEPTED MANUSCRIPT
°
(
C, 4 h, 60-75% over two steps; (e) i. MeOCH
f) CeCl , NaI, MeCN, reflux, 12 h, 87%; (g) NaBH
MeSO Cl, Et N, DCM, 25 °C, 12 h, 99%; (i) NaN , N,N-dimethylformamide (DMF), 90
2
PPh
3
Cl, NaN(SiMe
3 2
) , 25 °C, 12 h, 40%;
3
4
, MeOH, 25 °C, 1 h, 100%; (h)
2
3
3
°
C, 12 h, 70%; (j) Pd/C, H (1 atm), MeOH, 25 °C, 3 h, 95%.
2
As displayed in Scheme 1, the first step to the synthesis of 1 and analogues 8-12 was
a Suzuki-Miyaura coupling between aryliodides 3a-f and the organoboronic acid 4,[29]
to afford aldehydes 5a-f. In the case of derivatives 5a,c,d, this palladium-catalyzed
cross coupling reaction turned out to be exquisitely regioselective for the formation of
reaction products arising from attack at the most reactive iodine with respect to other
halides. In the subsequent steps of the synthesis, aldehydes 5a-f and 6 (in turn
obtained from 5f by deprotection) were reacted with the primary amine 7 and the
resulting imine intermediates were reduced using sodium cyanoborohydride. Final
removal of the piperidine protecting group upon exposure to acidic conditions led to the
final compounds 1 and 8-12.
Amine 7 was prepared starting from N-Boc-piperidone 13, which was subject to a
Wittig reaction using (methoxymethyl)triphenylphosphonium chloride and sodium
bis(trimethylsilyl)amide[30] resulting in an intermediate vinylmethyl ether that was
hydrolyzed using cerium chloride and sodium iodide[31] to afford aldehyde 14. This
latter compound was reduced to the corresponding alcohol, which was converted into
methanesulfonate and reacted with sodium azide to furnish 15. The final catalytic
hydrogenation, using palladium on carbon, led to the formation of the desired
piperidinyl-methanamine 7 in satisfactory yield.
Because of the unavailability of the starting aryliodides, a second set of compounds,
modified at the hydrophobic head, was synthesized using a different chemical route
(Scheme 2). For the synthesis of aldehydes 18a-e, the Meerwein arylation protocol,[32]
7

ACCEPTED MANUSCRIPT
consisting in a copper-catalyzed arylation of furan derivatives with arenediazonium
salts, was conveniently used. According to this procedure, anilines 16a-e were
converted into the corresponding diazonium salts, and these latter intermediates were
coupled in situ with furan-carboxaldehyde 17. Intermediats 18a,b were also submitted to
a Suzuki coupling to afford the biphenyl-derivatives 19a,b, respectively. Aldehydes 18a-
e, and 19a,b were finally converted into final compounds 21-26 using the protocol
previously described for the introduction of the pendant basic chain.
Starting from 16a,b, the Meerwein arylation protocol was also applied to the thiophen-
carboxaldehyde 27 to obtain intermediates 28a,b. These latter compounds were then
converted to the final compounds 29,30, representing a first set of analogues of the hit
compound 1 bearing modification of the heterocyclic linker.
a
Scheme 2. Synthesis of analogues 21-26 and 29,30
8

ACCEPTED MANUSCRIPT
a
Reagents and conditions: (a) i. 16a-e, HCl 37%, NaNO
2
, H
2
O, 0 °C, 1 h; ii. 17 (or 27),
, aq. Na CO , PhB(OH)
CuCl , acetone, 25 °C, 12 h, 70-90%; (b) Pd(OAc) , PPh
2
2
3
2
3
2
,
toluene, reflux, 12 h, 63-85%; (c) i. 7, MeOH, reflux, 12 h; ii. NaBH , 0 °C, 1 h; (d) 1 N
4
HCl, MeOH, 25 °C, 4 h, 51-80% over two steps.
For the synthesis of the benzothiophene derivative 32 (Scheme 3), aldehyde 31[33,
34] was submitted to the reductive amination reaction, which upon acidic Boc
deprotection, led to the preparation of the desired final compound 32. Five-membered
heterocycles 37 and 38 were prepared starting from arylhydrazide 33. Cyclization of 33
was performed through a Pellizzari-like reaction, consisting in the condensation
between acyl hydrazine 33 with benzyl amine.[35] Only the dehalogenated product 34
was obtained from this reaction. The alcoholic function of 35, introduced by a Prince
electrophilic addition using formaldehyde, was activated to the corresponding chloride
36a which was alkylated with amine 7. Oxidative N-debenzylation[36] followed by acid-
mediated deprotection of the Boc-group furnished the triazole 37. For the synthesis of
oxadiazole 38, the chloromethyl intermediate 36b was obtained by cyclization of 33 in
3
the presence of chloroacetic acid and POCl .[37]
Another substitution of the central linker is represented by the pyridine derivative 43
(Scheme 4). A C2 selective Suzuki-Miyaura arylation was applied starting from the
commercially available 2-bromo-6-methylpyridine 39 and 4-trifluoromethylphenylboronic
acid 40. The methyl group at C6 of the pyridine ring was then subjected to a
bromination reaction using N-bromosuccinimide (NBS) and azobisisobutyronitrile
(AIBN). The brominated intermediate 42 was used to alkylate amine 7. The final Boc
deprotection gave in high yield the final derivative 43.
a
Scheme 3. Synthesis of central linker analogues 32,37,38
9

ACCEPTED MANUSCRIPT
a
Reagents and conditions: (a) i. 7, MeOH, reflux, 12 h; ii. NaBH , 0 °C, 1 h; (b) 1 N
4
HCl, MeOH, 25 °C, 4 h; (c) N,N-dimethylformamide dimethyl acetal (DMF-DMA), MeCN,
0 °C, 1 h, then BnNH , AcOH, 120 °C, 12 h, 60%; (d) aq. CH O, reflux, 12 h; (e)
5
2
2
pyridine, SOCl , DCM, 0 °C, 3 h, 50%; (f) POCl , ClCH CO H, reflux, 12 h, 20%; (g) 7,
2
3
2
2
K
2 3 2
CO , KI (cat.), MeCN, reflux, 1 h, 72%; (h) tBuOK, DMSO, THF, O , 25 °C, 2 h, 87%.
a
Scheme 4. Synthesis of pyridine derivative 43
10

ACCEPTED MANUSCRIPT
a
Reagents and conditions: (a) Pd(OAc) , PPh , K CO , 7:3 DMF/H O, reflux, 12 h,
2
3
2
3
2
8
1%; (b) NBS, AIBN (cat.), CCl , reflux, 12 h, 30%; (c) 7, K CO , MeCN, reflux, 12 h,
4
2
3
20%; (d) 1 N HCl, MeOH, 25 °C, 4 h, 98%.
The open-chain analogue 45 was synthesized as described in Scheme 5 starting from
intermediate 5a and the commercially available N,N’-dimethyl-1,3-propanediamine 44.
Starting from 18b, reductive amination with 44 followed by protection of the secondary
amine as Boc-derivative led to intermediate 46. The tertiary amine was subsequently
converted to a quaternary ammonium salt using benzyl bromide and iodomethane,
respectively. The final Boc-deprotection using acetyl chloride and methanol gave the
two final derivatives 47 and 48 in satisfactory yields.
2.2.Antiplasmodial and gametocytocidal activities and structure-activity
relationships
The antiplasmodial activity of the newly synthesized compounds was tested against
two laboratory P. falciparum strains, the chloroquine-sensitive (CQ-S) D10 and the
chloroquine-resistant (CQ-R) W2, according to the described procedures.[38, 39]
Activity against gametocytes (GCT) was assessed using a luciferase-based assay, as
previously described.[23]
a
Scheme 5. Synthesis of analogues modified at the protonatable side chain
11

ACCEPTED MANUSCRIPT
a
Reagents and conditions: (a) i. NH
C, 24 h, 85%; (b) (Boc) O, Et N, MeOH, 25 °C, 4 h, 43% over two steps; (c) MeI (for
7) or BnBr (for 48), acetone, reflux, 72 h, 58-60%; (d) 1 N HCl, MeOH, 25 °C, 4 h, 94-
2 2 3 2 3
(CH ) NMe , DCM, 25 °C, 1 h; ii. NaBH(OAc) , 25
°
4
2
3
95%.
Selected compounds were also tested against P. falciparum strain NF54 asexual
blood stage parasites using a Sybrgreen DNA replication assay. NF54 gametocyte
assays were performed using reporter strain NF54-HGL.[40] The results of the
biological assays are reported in Tables 1-3.
2.2.1. Hydrophobic head
In order to understand if the antiplasmodial and antigametocytocidal activities
depended upon the substitution pattern of the aromatic head, we removed one or
both alogens of 1 (compounds 9, 22 and 8, respectively). Moreover, we inverted the
position of the Cl or Br substituent obtaining the regioisomeric derivative 21, and
introduced different electron-withdrawing groups at the p-position such as –F (10), -
CN (24), and –CF
3
(23). We also placed electrondonating and H-bond
donor/acceptor groups such as -OMe and –OH (11 and 12, respectively). A second
12

ACCEPTED MANUSCRIPT
aromatic ring, giving rise to regioisomers 25 and 26 was placed along the halogen
present in the original structure in order to explore the tolerance of the system to
bulky substituents. From this study we derived interesting SARs: as shown in Table
1, the substituents at the hydrophobic head seems to play a key role for in vitro
activity against both asexual and sexual parasite stages. In particular, the presence
of both halogens is important for activity since the unsubstituted analogue 8, as well
as the mono-halogenated derivatives 9, 10 and 22 are less potent antiplasmodial
and gametocytocidal agents than the hit compound 1. Inversion of chlorine and
bromine seems not to have a significant effect on potency as demonstrated by the
regioisomeric derivative 21. Among the other electron-withdrawing substituents
investigated, the trifluoromethyl derivative 23 maintained a potency similar to 1,
while with a cyano group at para position (24) a drop of potency was observed,
especially against the gametocytes. Also the introduction of electron-donating
substituents such as methoxy- (11) or hydroxyl- (12) groups resulted in a dramatic
drop of potency against gametocytes. Interestingly, introduction of an aromatic ring
at either C2 or C4, as in compounds 25 and 26, respectively, was tolerated and
antiplasmodial and gametocytocidal activity comparable to the hit compound 1 were
observed.
Table 1. Asexual and sexual stages activity (IC , µM) of compounds modified at
50
the hydrophobic head.
a
b
a
D10
W2
GCT:
D7elo1-
NF54
GCT:
NF54-HGL
3
Structure
pfs16-
CBG99
13

ACCEPTED MANUSCRIPT
c
IC (µM)
5
0
d
1
8
9
1
0.77
3.5
2.3
9.1
8.0
7.9
1.2
8.6
4.4
6.0
n.t.
n.t.
n.t.
n.t.
n.t.
O
HN
Br
Cl
NH
n.t.
n.t.
n.t.
n.t.
O
HN
NH
1.1
O
HN
Br
NH
0
2.1
O
HN
F
NH
11
12
21
10.0
4.4
>15
9.6
11.8
8.3
n.t.
O
HN
HN
OMe
NH
0.26
n.t.
0.29
n.t.
O
OH
NH
0.51
1.87
1.1
O
HN
Cl
Br
NH
2
2
3
2.1
4.3
1.5
5.9
1.9
n.t.
n.t.
n.t.
n.t.
O
HN
Cl
NH
NH
2
0.41
HN
O
3
F C
NH
24
1.1
6.3
18.4
n.t.
n.t.
HN
O
NC
14

ACCEPTED MANUSCRIPT
2
5
6
0.88
1.0
2.4
3.0
1.4
1.4
0.67
n.t.
0.043
n.t.
O
HN
Cl
NH
2
O
HN
Cl
NH
Chloroquine
Epoxomycin
0.026
-
0.31
-
n.t.
-
-
-
0.007
0.030
-
-
0.0007
n.t.
Methylene blue
-
a
b
c
CQ-S strain; CQ-R strain ; IC values are the mean of at least three experiments in
50
duplicate; differences in the experimental conditions, parasite strain/stage, and methods
of detection may account for the different IC observed in the two gametocyte assays.
50
d
Standard deviations were within 20% of the mean; n.t. = not tested.
2.2.2. Heterocyclic linker:
The role of the heterocyclic tether on activity was evaluated through the synthesis of
several 5-membered heterocyclic systems as isosteric derivatives of the furan ring such
as thiophene- (29, 30), triazole- (37), and oxadiazole- (38) derivatives (Table 2). The
pyridine-derivative 43 was conceived in order to explore the effect on activity of a 6-
membered ring bearing an heteroatom able to function as H-bond acceptor analogously
to the furan oxygen, while the fusion of the pendant phenyl ring with the thiophene of 29
resulted in the benzothiophene-derivative 32. The sulfur-based heterocycles were
tolerated and in some cases led to a slight improvement of activity. Also the six-
membered pyridine-derivative (43) showed reasonable potency, especially against the
CQ-S strain, while the other heterocycles investigated led to a dramatic drop of potency.
Table 2. Asexual and sexual stages activity (IC , µM) of compounds modified at
50
the heterocyclic linker.
15

ACCEPTED MANUSCRIPT
a
b
a
D10 W2 GTC
NF54 GTC
3D7elo1-pfs16-CBG99 NF54-HGL
Structure
c
IC50 (µM)
2
9
0
0.73 2.2 0.85
0.53 1.4 0.83
0.28
0.42
0.020
S
HN
Br
Cl
NH
NH
3
0.028
HN
S
Cl
Br
Br
NH
3
2
7
1.0
1.5 1.4
0.73
0.053
n.t.
HN
S
N
NH
d
3
>13 >13 >13
>13 7.9 >26
n.t.
N
HN
N
H
Cl
NH
38
n.t.
n.t.
n.t.
n.t.
N
HN
N
O
Cl
Br
NH
43
H
0.46 3.4 4.0
N
N
3
F C
a
b
c
CQ-S strain; CQ-R strain; IC values are the mean of at least three experiments in
50
duplicate; differences in the experimental conditions, parasite strain/stage, and methods
of detection may account for the different IC observed in the two gametocyte assays.
50
d
Standard deviations were within 20% of the mean; n.t. = not tested.
2.2.3. Pendant basic chain:
Derivative 45, bearing the same basic chain as 2, showed a potency comparable to 1
Table 3). The ammonium quaternary salts 47 and 48 were synthesized in an attempt to
(
increase potency since quaternary ammonium salt derivatives were previously
displayed as interesting antiplasmodial agents.[41] Both compounds displayed a drop of
16

ACCEPTED MANUSCRIPT
potency against gametocytes. Interestingly, benzyl-derivative 48 maintained its potency
against the asexual forms of the parasite, but not against the gametocytes.
Table 3. Asexual and sexual stages activity (IC , µM) of compounds modified at
50
basic side chain.
a
b
D10 W2 GTC:
D7elo1-pfs16-CBG99
3
Structure
c
IC (µM)
5
0
4
5
7
1.2
1.4 1.1
5.5 >24
O
HN
Br
Cl
N
Cl-
4
2.5
Me
Me
Me
N
HN
O
Cl
Br
48
0.55 0.66 4.8
Cl-
Me
Me
N
HN
O
Cl
Br
a
b
c
CQ-S strain; CQ-R strain; IC values are the mean of at least three experiments in
50
duplicate;. Standard deviations were within 20% of the mean.
For all compounds synthesized, molecular properties calculated in silico by QikProp
(version 4.3; Schrödinger, LLC: New York, 2015) such as lipophilicity, polar and solvent
accessible surface areas, solubility and membrane permeability were all in the
recommended range values for known drugs (Table S1 in the Supplementary
Information file).
2.3.In vitro cytotoxicity.
17

ACCEPTED MANUSCRIPT
After investigation of SAR, we selected a representative set of analogues
characterized by various degrees of antiplasmodial and gametocytocidal potency for
cytotoxicity assays. Cell viability was evaluated in vitro against NIH3T3 after 24 h of
contact, by Neutral Red Uptake (NRU) test. The cell line has been chosen because is a
standard fibroblast cell line.
Results of cytotoxicity tests are reported in Table 4. Compounds 11, 23 and 25
resulted the least toxic of the series. The reasonable toxicity of compound 25, combined
to the structure of the inhibitor presenting an extra aromatic ring that could be further
decorated, prompted us to select this compound for further biological investigation as
detailed in the next paragraphs.
Table 4. Cytotoxic activity against 3T3 cell lines of selected compounds evaluated after
24 h of incubation by NRU test.
Cpd
11 12 21 23 25 26 29 30 45
a
TC (µM) 53 7 <5 50 20 5 <5 <5 5
50
a
Data are expressed as mean of three experiment repeated in six replicate. s.d. are
within 5% of the mean
2.4.Standard membrane-feeding assay.
The evaluation of the ability of compounds to inhibit the development of oocysts in the
mosquito and thus reduce the mosquito’s ability to transmit malaria parasites can be
confidently assessed using the standard membrane-feeding assay (SMFA) which
remains the assay of choice for this kind of investigation. The assay consists in feeding
cultured P. falciparum gametocytes to Anopheles mosquitoes in the presence of the test
compound and measuring subsequent mosquito infection. The hit compound 1 was
recently shown to have an EC50 in the SMFA assay of 0.07 µM.[21] Evaluation of
18

ACCEPTED MANUSCRIPT
compound 25 in the SMFA assay resulted in an IC50 of 0.21 µM on infection intensity,
and 0.83 µM on infection prevalence at a baseline infection intensity of 30
oocysts/midgut (Figure 1). These data confirm that the gametocytocidal activity of 25 as
observed in the NF54 gametocyte luciferase assay translates into an inability to infect
the mosquito vector. The IC50 on infection intensity is slightly higher than expected on
basis of the NF54 gametocyte luciferase assay. This may be explained by the difference
in incubation time (24 h in the SMFA vs 72 h in the gametocyte assay) and a relatively
slow mode of action of the compound.
Figure 1. Compound 25 blocks transmission in the SMFA. A) Compound effect on the
infection intensity. The figure shows luminescence activities 8 d post-infection of
mosquitoes fed on NF54-HGL gametocytes that were pre-incubated with compound 24
h prior to feeding. A total of 24 mosquitoes was analyzed for each experimental feed,
and all feeds were performed in duplicate. The figure shows the mean of the average
signal intensities of each feed. Error bars indicate the standard error of the mean B)
compound effect of infection prevalence, or the proportion of mosquitoes that carry at
least one oocyst. The figure shows average data from two independent feeds. Error
bars indicate standard deviations.
19

ACCEPTED MANUSCRIPT
2.5.Evaluation of activity of compound 25 against P. berghei liver stages.
We also investigated if compound 25 could have activity against the liver stages of P.
berghei parasites. Primary mouse hepatocytes have been infected with mCherry
expressing P. berghei sporozoites. 2 h post infection (hpi) the cells were exposed to
different concentrations of 25 for 48 hpi with medium change at 24 hpi. Neither parasite
numbers (Figure 2A) nor parasite sizes (Figure 2B) were significantly influenced at the
concentration tested. Compound 25 was also used to analyze detached cell formation,
which is an in vitro assay for final development (equivalent to merosome formation in
vivo). Living parasites were counted at 48 h and from the same wells detached cells at
65 hpi evaluated. Neither concentration tested significantly inhibited detached cell
formation, as indicated by the similar rates of detached cells compared to DMSO-
treated control cells at 48 h (Figure 2C).
20

ACCEPTED MANUSCRIPT
Figure 2. Compound 25 does not inhibit P. berghei liver stage schizont development in
vitro. Different concentrations of 25 were employed to treat mouse primary hepatocytes
infected with mCherry expressing P. berghei sporozoites were treated with different
concentrations of 25. A. Parasite numbers per well at 48 h are shown (n = 4, SD)
indicating no liver stage activity of compound 25. B. Parasite sizes of one replicate of
treated cells from A) are shown. C. Detached cells compared to DMSO-treated control
cells at 48 h One-wayANOVA with Dunnet’s Multiple Comparisons (* p
.01 *** p 0.001, ns/not significant > 0.05) showed no significant difference compared
to the control treated cultures.
≤ 0.05, ** p ≤
0
≤
2.6.Evaluation of cardiac toxicity and calcium antagonist activity of 1 and 25.
Considering the potential hERG liability previously highlighted for compound 1, we
decided to evaluate the cardiotoxic potential of both compounds 1 and 25. Accordingly,
their effect on cardiac mechanical function and electrocardiogram (ECG) in
Langendorff-isolated rat hearts was assessed, as previously described.[39, 42]
Compounds 1 and 25 reduced left ventricular pressure (LVP) in a concentration-
dependent manner (Figure 3), leaving unaltered coronary perfusion pressure (CPP,
data not shown). They significantly prolonged atrioventricular conduction time (PQ) and
intraventricular conduction time (QRS) in ECG at the maximum concentration tested (10
µM; Table 5). However, QTc values did not vary over the drug range tested (Table 5).
Compound 1 significantly decreased heart rate (HR), increased cycle length (RR)
interval (Table 5) and caused a sinus arrhythmia in 3 out of 4 hearts.
v
Since channel blockers, by inhibiting the Ca 1.2 channel currents, may prolong
conduction and refractoriness in the atrioventricular node,[43] the increased PQ interval
21

ACCEPTED MANUSCRIPT
obtained in presence of both compounds prompted us to investigate the effects of 1 and
2
+
2
5 on Ca 1.2 channel currents. Compound 1 inhibited I
(Ba current through Ca 1.2
v
Ba1.2 v
channels), measured at 0 mV from a holding potential (V ) of −50 mV, in a
h
concentration-dependent manner (Figure 4). At 10 µM concentration, inhibition
amounted to 57%. Compound 25 blocked the current at 30 µM concentration and
showed an IC value of 3.3 ± 0.5 µM (n = 5).
50
1
00
8
6
4
2
0
0
0
0
0
1
2
5
-
7.0
-6.5
-6.0
-5.5
-5.0
Log [drug] (M)
Figure 3. Effects of 1 and 25 on left ventricular pressure in Langendorff perfused rat
hearts. Concentration-effect relationship of 1 and 25 on LVP. On the ordinate scale,
response is reported as mmHg. Each value represents mean ± SEM (n = 4 hearts).
Table 5. Effects of 1 and 25 on HR, RR, PQ, QRS and QTc in Langendorff Perfused
a
Rat Hearts
Cpd (µM)
HR (BPM)
RR (ms)
PQ (ms)
QRS (ms)
QTc (ms)
1
2
1
None
247.05±6.87
275.90±5.43
256.15±7.19
242.23±5.93
217.15±3.38
235.08±7.41
42.63±2.08
41.13±1.64
42.29±2.30
14.35±0.47
13.31±0.62
14.25±0.48
80.07±1.94
76.45±1.15
81.51±1.95
5
0.1
22

ACCEPTED MANUSCRIPT
2
1
2
1
2
1
5
5
5
5
276.19±5.33
258.35±6.79
275.47±4.14
257.71±7.88
276.40±3.59
218.32±4.37
232.48±6.38
217.02±3.57
233.90±7.19
216.73±2.48
41.24±1.65
42.50±2.19
41.59±1.60
47.90±2.95
45.47±2.58
14.20±0.61
15.80±0.39
13.70±0.43
17.46±1.15
14.65±0.91
74.92±1.01
80.93±1.56
77.21±2.25
82.10±2.16
78.56±3.86
1
3
10
222.74±14.1* 272.57±17.8* 67.50±1.85**
23.75±2.84** 80.99±3.78
2
259.41±6.79 232.20±5.83 65.08±10.15** 17.91±1.33** 83.77±4.58
a
HR (frequency), RR (cycle length), PQ (atrio-ventricular conduction time), QRS
intraventricular conduction time), QTc (corrected overall action potential duration).
(
Bazett's formula normalized to average rat RR (QTc=QT/(RR/ f)1/2) [J. Kmecova, J.
Klimas, Heart rate correction of the QT duration in rats, Eur. J. Pharmacol. 641 (2010)
187–192] was used to correct QT.Each value represents mean ± SEM (n = 4 hearts). *
P < 0.05, ** P < 0.01, repeated measures ANOVA and Dunnett’s post test.
100
75
50
25
0
1
2
5
-7.0
-6.5
-6.0
-5.5
-5.0
-4.5
Log [drug] (M)
Figure 4. Effect of compounds 1 and 25 on IBa1.2 in single rat tail artery myocytes.
Concentration-dependent effect of 1 and 25 at the peak of IBa1.2 trace. On the ordinate
scale, response is reported as percentage of control. Data points are mean ± SEM (n =
3–6 cells, isolated from at least 3 animals.).
2.7.In silico approach for target identification of compounds 1 and 25.
23

ACCEPTED MANUSCRIPT
Despite recent advances in the field, target identification remains a challenging task
that could be approached through several methodologies such as direct biochemical
methods, genetic interactions, computational approaches, or a combination of methods.
We performed a preliminary computational investigation aimed at identifying a potential
target of compounds 1 and 25 from structures in the Protein Data Bank (PDB). As a part
of a broad investigation aimed at repurposing FDA-approved drugs against malaria, we
prepared all the P. falciparum X-ray crystal structures present in PDB at May 2016, to
be used in a high-throughput docking (HTD) procedure for identifying potential
antimalarial drug-targets targeted by known drugs. Employing a reverse-docking
procedure (also known as reverse virtual screening or target fishing),[44] we retrieved
from the PDB database (and prepared by Protein Preparation Wizard as previously
reported[42, 45]) all enzymes and proteins relevant for the plasmodium biology, some of
which are established drug targets. Compounds 1 and 25 were then used to “fish” for
potential targets by using Glide software.[46-48] The results of our in silico study for
these compounds indicated a very limited number of potential targets with low docking
scores among the examined docking complexes. Surprisingly, for one protein (P.
falciparum phosphoethanolamine methyltransferase; PfPMT) we observed for both
compounds significant docking scores and calculated ligand-binding energies, as found
by Glide and Prime (see Experimental part for further details).[46, 49-51] In fact, for 1
and 25 we found computational scores (1, Glide XP score = -10.52 kcal/mol,
7.38 kcal/mol; 25, Glide XP score = -11.28 kcal/mol, Gbind = -67.54 kcal/mol)
comparable to those found for Sinefungin (Glide XP score = -13.16 kcal/mol,
4.77 kcal/mol) and S-adenosylmethionine (SAM) (Glide XP score = -10.57 kcal/mol,
∆Gbind = -
5
∆
∆Gbind = -
8
24

ACCEPTED MANUSCRIPT
∆Gbind = -82.42 kcal/mol), especially for the docking scores. The most relevant
differences were related to the ∆Gbind, for which we observed less negative values of 1
and 25 with respect to the reference ligands. So, given the satisfactory in silico scores,
we decided to test them against PfPMT along with two related derivatives 24 and 48.
2.7.1. Evaluation of 1, 25 and analogues as inhibitors of PfPMT.
Compounds 1, 24, 25 and 48 were tested as potential inhibitors of PfPMT using a
radiochemical assay performed with purified recombinant protein, which was expressed
and purified as previously reported (Figure 5A).[52, 53] Compounds 1, 24, and 48
showed weak activity against PfPMT, PfPMT, while compound 25 displayed an
inhibitory effect against PfPMT with an IC50 = 44.3 ± 5.4
µM (Figure 5B). The
experimental data is in good agreement with the computational output. This latter
indicates compound 25 as the best performing compound among the tested molecules,
although with slight reduced activity with respect to the reference compounds [53].
Considering that at the moment only few PfPMT inhibitors have been described, mainly
related to the natural substrate SAM, compound 25 could be used as a starting point to
develop PfPMT inhibitors as potential antimalarials. Figure 6 depicts the docking output
found by Glide software of 25 into the PfPMT active site. In particular, the amino-groups
of the compound may form hydrogen bonds with D85 and R127 with the biaryl system
fitting into in a hydrophobic sub-pocket and the furan group stacking with H132. Notably,
these interactions are part of the main contacts found for the above-mentioned
reference compounds [53].
25

ACCEPTED MANUSCRIPT
Figure 5. A. Inhibitory activity against PfPMT by compounds 1,24,25,48 at 100 µM; B.
PfPMT inhibitory activity (IC50, µM) of 25. All assays were performed as described in
the methods section with values shown as mean ± standard deviation (n=3).
Figure 6. Putative mode of interaction of compound 25 (yellow sticks) within the active
site of PfPMT (deep teal cartoon; PDB ID 3UJ8). H-bonds are represented by black
26

ACCEPTED MANUSCRIPT
dotted lines. The picture was generated by PyMOL (The PyMOL Molecular Graphics
System, v1.8.0; Schrodinger, LLC, New York, 2015).
Despite the interesting findings, the inhibition of PfPMT experimentally observed was
not correlated to the antiplasmodial and gametocytocidal activity of the compound and
from these data we can rule out PfPMT as the target (or at least the main target) of this
series of compounds. Our results are in line with a study recently published[25] in which
PfPMT has been excluded as the target of 1 in an indirect yeast-mediated assay.
3. Conclusions
Here we described a detailed SAR analysis of hit compound 1 and its analogue 25 for
their activity against asexual and sexual life cycle stages of P. falciparum. From our
SAR studies, the electron withdrawing substituents at the hydrophobic head were
identified as important for activity. Moreover, the presence of nitrogen-based
heterocyclic linkers resulted detrimental for activity being the thiophene and the furan
rings the best performing linkers. A small set of side-chain analogues was explored
highlighting that the open-chain analogue retained antiplasmodial potency, while it was
interesting to note that among the two quaternary ammonium salts prepared, the
benzyl-derivate was active against the asexual life cycle stage. Among the set of
compounds synthesized, analogue 25 was selected for a more in depth biological
investigation, due to its antiplasmodial potency in the low micromolar range, reasonable
cytotoxicity and the presence of an extra aromatic ring that could offer a handle for
further scaffold decoration. In particular, compound 25 was active in the SMFA assay
with potency comparable to 1. On the other hand, 25 did not show activity against P.
berghei hepatocytic life-cycle stage at the concentration tested, even though this could
27

ACCEPTED MANUSCRIPT
be a species specific effect. Unfortunately, cardiac toxicity was observed for both 1 and
25. Even though the biological activity of 1 and 25 against gametocytes and asexuals P.
falciparum stages remains interesting and worth of further investigation, exploitation of
this scaffold will only be possible by identifying its target at the molecular level. This is a
necessary pre-requisite for investigating if dissociation of in vitro toxicity from the
specific gametocytocidal and schizontocidal activities is a feasible task. Accordingly, we
reported herein our preliminary in silico study devoted at identifying the target for this
series of compounds. Driven from this study we subsequently found that analogue 25
has an interesting in vitro inhibitory activity against PfPMT, a key enzyme for
phospholipid biosynthesis in P. falciparum.[54] Compound 25 could serve as a starting
molecule, for the structure-based design of novel inhibitors as potential antiplasmodial
and transmission-blocking agents.
4
. Experimental Section
.1.Chemistry
Starting materials and solvents were purchased from commercial suppliers and used
4
without further purification. Reaction progress was monitored by TLC using silica gel 60
F254 (0.040 - 0.063 mm) with detection by UV. Silica gel 60 (0.040 - 0.063 mm) was
1
13
used for column chromatography H NMR and C NMR spectra were recorded on a
Varian 300 MHz, or a Bruker 400 MHz spectrometer using the residual signal of the
deuterated solvent as internal standard. Splitting patterns are described as singlet (s),
doublet (d), triplet (t), quartet (q), and broad (br); the value of chemical shifts (δ) are
given in ppm and coupling constants (J) in Hertz (Hz). Mass spectra were recorded
utilizing electron spray ionization (ESI) Agilent 1100 Series LC/MSD spectrometer.
28