Synthesis, biological evaluation, in silico modeling and crystallization of novel small monocationic molecules with potent antiproliferative activity by dual mechanism

Article abstract of DOI:10.1016/j.ejmech.2020.112797

Seeking for new anticancer drugs with strong antiproliferative activity and simple molecular structure, we designed a novel series of compounds based on our previous reported pharmacophore model composed of five moieties. Antiproliferative assays on four

Full text of DOI:10.1016/j.ejmech.2020.112797

European Journal of Medicinal Chemistry 207 (2020) 112797
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Synthesis, biological evaluation, in silico modeling and crystallization
of novel small monocationic molecules with potent antiproliferative
activity by dual mechanism
a
,
,
Lucía Serran- Aguilera 1, Elena Mariotto b 1, Gianluca Rubbini a,
ꢀ
a
c
c
ꢀ
Francisco Fermín Castro Navas , Carmen Marco , María Paz Carrasco-Jimenez ,
d
d
e, f
ꢀ
Marco Ballarotto , Antonio Macchiarulo , Ramon Hurtado-Guerrero
,
Giampietro Viola b, Luisa Carlota Lopez-Cara a
,
*
a Department of Pharmaceutical and Organic Chemistry, Faculty of Pharmacy, Campus Cartuja S/n. University of Granada, 18010, Granada, Spain
b Department of Woman’s and Child’s Health, Laboratory of Oncohematology, University of Padova, 35128, Padova, Italy
c Department of Biochemistry and Molecular Biology I, Faculty of Sciences, 18071, Granada, Spain
d Department of Pharmaceutical Sciences, University of Perugia, Via Del Liceo 1, Perugia, 06123, Italy
e Institute of Biocomputation and Physics of Complex Systems (BIFI), University of Zaragoza, Institute for Biocomputation and Physics of Complex Systems
(BIFI) and Laboratorio de Microscopías Avanzada (LMA), Mariano Esquillor S/n, Campus Rio Ebro, Edificio IþD; Fundacion ARAID, 50018, Zaragoza, Spain
f Copenhagen Center for Glycomics, Department of Cellular and Molecular Medicine, School of Dentistry, University of Copenhagen, Copenhagen, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
Seeking for new anticancer drugs with strong antiproliferative activity and simple molecular structure,
we designed a novel series of compounds based on our previous reported pharmacophore model
composed of five moieties. Antiproliferative assays on four tumoral cell lines and evaluation of Human
Received 5 May 2020
Received in revised form
28 August 2020
Accepted 30 August 2020
Available online 6 September 2020
Choline Kinase CKa1 enzymatic activity was performed for these compounds. Among tested molecules,
those ones with biphenyl spacer showed betters enzymatic and antiproliferative activities (n-v). Docking
and crystallization studies validate the hypothesis and confirm the results. The most active compound (t)
induces a significant arrest of the cell cycle in G0/G1 phase that ultimately lead to apoptosis, following
the mitochondrial pathway, as demonstrated for other choline kinase inhibitors. However additional
assays reveal that the inhibition of choline uptake could also be involved in the antiproliferative outcome
of this class of compounds.
Keywords:
Antitumoral drug
Choline kinase inhibition
Choline uptake
© 2020 Elsevier Masson SAS. All rights reserved.
1. Introduction
increased lipogenesis, in order to satisfy the demand in the syn-
thesis of biomolecules that proliferation requires [2,3]. Among the
Cancer is among the main world causes of mortality and
morbidity, further exacerbated by the occurrence of drug resistance
[1]. For this reason, despite being an intensively investigated dis-
ease, new drugs and therapeutic strategies are still urgently needed
to more efficiently eradicate cancer cells. their transcriptional and
mutational heterogeneity, cancer cells derived from different types
of tumors frequently share similar metabolic abnormalities, such as
changes in energy metabolism, from oxidative phosphorylation to
aerobic glycolysis, increased protein and nucleotide synthesis and
lipid-related alterations, it has been described that many types of
cancer show an abnormal metabolism of phospholipids containing
choline and/or ethanolamine [4,5], which can be a starting point in
the design of new antitumor strategies.
The involvement of choline (Cho) metabolism and cancer has
been demonstrated by magnetic resonance spectroscopy images,
which highlight the presence of the “choline phenotype” in mul-
tiple tumors, characterized by high levels of phosphocholine and in
general of choline-containing compounds [4].
Phosphatidylcholine (PC) is not only the most abundant phos-
pholipid in cell membranes but also constitutes a fundamental part
of lipoproteins, biliary lipids and lung surfactants [6,7], besides
being a direct substrate for the synthesis of sphingomyelins (SM)
and an essential source of second messengers. PC can be
* Corresponding author.
E-mail address: lcarlotalopez@ugr.es (L.C. Lopez-Cara).
1
Equal contributing Authors.
https://doi.org/10.1016/j.ejmech.2020.112797
0223-5234/© 2020 Elsevier Masson SAS. All rights reserved.

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
synthesized from choline, an organic cation that cannot cross
plasma membrane freely, needing therefore active transporters to
enter into the cell [8,9]. Once choline is inside the cell, it can follow
several pathways becoming incorporated to diverse cellular com-
ponents. One of these is the Kennedy pathway or CDP-choline
pathway, which leads to the synthesis of PC. This pathway con-
sists of three steps catalysed by different enzymes. The first step of
PC synthesis is catalysed by the cytosolic enzyme choline kinase
(CK) which drive the ATP-dependent phosphorylation of choline in
the presence of magnesium [10]. The human choline kinases family
N,N-methylphenyl fragment. In order to set a SAR electron with-
drawing group trifluoromethyl, (-CF3) and a phenyl ring have been
introduced, obtaining the second and third series of compounds (A,
B and C) as a substituent R3.
The cationic heads of the three series have been chosen based on
the pharmacophore and the chemical structure-activity relations
previously established by our research group [13,20,21,25]. In
particular, the cationic heads of the series A, B and C were chosen
on the basis that: a) they allow us to study whether the phenyl
group attached to the exocyclic nitrogen of the pharmacophore can
be dispensed with and whether the fusion of a phenyl ring to the
pyridine ring optimizes the pharmacophore; b) they have demon-
strated their suitability for the design of CK inhibitors over the last
few years. Consequently, functional groups derived from pyridine
(dimethylamino- and pyrrolidinopyridine) and quinolines (phenyl-
or cycloalkylaminoquinoline) and quinuclidine were selected for
this study.
comprises three isoforms, CK
residues, 50 kDa) that both originate from chok-
(395 residues, 45 kDa) encoded by chok- gene. Elevated expression
of CK has been reported for many tumors where it correlates with
a
1 (457 residues, 52 kDa), CK
a2 (439
a
gene, and CK
b
b
a
an unfavourable prognosis [11,12], thus becoming a new tumor
metabolic marker involved in the “choline phenotype”.
Recently our group has shown that in HepG2 cells the EB-3D
CK
a
1 inhibitor [13e16] interferes with PC biosynthesis through 1)
The purpose of this study is twofold. On one hand, the aim is to
optimise the pharmacophore by studying whether any of the five
functional groups that define it is dispensable, for which the frag-
ment of 1-benzyl-4-(N-methylanilino)pyridinium will be pharma-
comodulated. On the other hand, the aim is to achieve a greater
the choline uptake and 2) the CDP-choline synthesis pathway. This
effect is accompanied by a significant decrease in cholesterol
biosynthesis, as well as alterations in the expression of proteins
related to lipid homeostasis. We also found that, surprisingly, EB-
3D reduces CK
a
protein levels. All these alterations could be
degree of affinity for human CKa1 and, consequently, a better
explained by the modulation of the AMPK signalling pathway [16].
Our results conclude that EB-3D has an antitumor activity under-
lying a deregulation of lipid metabolism [14]. As mentioned above,
it is possible that some of the CK inhibitors could inhibit not only
the enzyme but also choline uptake in accordance with some au-
thors which reported the dual behaviour of certain choline kinase
inhibitors [28,29].
inhibitory and antiproliferative power than that shown by the
prototypes from virtual screening [22].
Within the A and B (Chart 1) series we distinguish thirteen
compounds benzyl and p-trifluoromethylbenzyl derivatives
depending on whether the cationic head is derived from pyridine
(4-(N,N-dimethylamino)pyridine, 4-pyrrolidinopyridine), quino-
line (7-chloro-4-(N-methylanilino)quinoline, 7-chloro-4-(perhy-
The deconvolution studies in monocationic asymmetrical
compounds determined that the adenine fragment occupies the
ATP binding site and that the pyridinium fragment, though its
positive charge is delocalized over the aromatic ring, mimics the
positive charge present in choline [17,18]. This situation was similar
for biscationic asymmetrical compounds with biphenyl group as a
linker [19]. However, the biscationic asymmetrical compounds
with a linker of 4 carbon atoms showed an unprecedented binding
droazepin-1-yl)quinoline),
quinoline) or quinuclidine.
7-chloro-4-(N-methylaniline)
The A and B series (Chart 1) aim to clarify whether these simple
structures show affinity for the enzyme as pointed out by the
pharmacophore or whether, on the contrary, they require the
introduction of more functional groups to improve their binding.
Another objective will be to determine the affinity differences be-
tween pyridine and quinoline derivatives, both on the basis of the
phenyl ring fused with pyridine and on the basis of the exocyclic
amino substituents in position 4.
mode to CKa1 [19]. On the other hand, recent docking studies have
showed that the same behaviour occurs in asymmetrical biscationic
compounds with a pyridinium and quinolinium cationic heads
[13,20,21].
In the C series (Chart 1) the structures were optimized by
introducing one more phenyl group into the spacer, with the pur-
Moreover, we recently proposed a pharmacophore model
composed of five moieties (RRRPN: the positive charge of the
pyridinium ring - feature P; a nitrogen atom - feature N; and three
aromatic groups - features R) that are included in the 1-benzyl-4-
(N-methylaniline) pyridinium fragment [22,23]. The comparison
with compounds with a phenyl ring condensed to the pyridine
(quinolinium derivatives) has showed that the extended conjuga-
tion is not essential for the enzyme inhibition [22,23]. At the same
time, Zech et al. [24] through another computational study
described a pharmacophore model consisting of small molecules
without the second positive charge, with tertiary amines (dia-
zepanes or piperazines) that present a better safety profile. In
addition, a dose-dependent correlation between the enzymatic
inhibition, the antiproliferative effect and apoptosis has been
observed for these molecules [24].
These results prompted us to synthesise and evaluate a series of
small monocationic molecules wherein our pharmacophore is
included, but also other ones wherein this pharmacophore has
been modified with other moieties, in order to establish a
structure-activity relationship (SAR) and refine the pharmaco-
phoric hypothesis. Hence, we design and synthesized firstly a series
of compounds with phenyl pyridinium and (7-chloro)-quinolinium
moieties N-substituted in 4 position with cycloalkylamines or with
pose of increasing the enzyme affinity through p-p interactions
with the aromatic residues present at the choline site. This small
modification will allow a comparison of the affinity, inhibitory ac-
tivity and antiproliferative capacity of these derivatives with those
of the A and B series.
2. Chemistry
These compounds a-v have been synthesized using microwaves
in one easy step. To a prepared solution of benzyl bromide, 1-
(bromomethyl)-4-(trifluoromethyl)benzene or 4-(bromomethyl)-
1,10-biphenyl in butanone or acetonitrile as the solvent, was added
the corresponding intermediates 1e3, (7-chloro)-4-subtituted
quinoline 4, quinuclidine 5, or 3-quinuclidinol 6 [13,20,21,25] (in
a molar ratio of 1:1). The reaction was performed in a sealed tube a
under microwave irradiation (140 ꢀC, 28e33 min). (Scheme 1).
3. Biological results and discussion
3.1. In vitro antiproliferative activities
Table 1 summarizes the growth inhibitory effects of derivatives
aev and reference compounds RSM932A and MN58b (Chart 1)
2

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
Chart 1. General structure of series A, B and C (a-v). Reference compounds MN58b and RSM932A.
against a panel of four human cancer cell lines. The results pre-
sented in Table 1 indicate that inhibition of cell growth was highly
dependent upon the presence of a second phenyl ring as R3 sub-
stituent, being the most active compounds belonging to series C.
The first remark is that the quinuclidinium moiety (h, i, q and r)
drastically decrease the activity, which is in agreement with our
previous results [13,20,21,25].
Moreover, we can also observe that the less active compounds
are those bearing a phenyl ring without substituent. The presence
of substituent groups in position 4 (R3), such as trifluoromethyl
(-CF3) or a second phenyl group, significantly increases the anti-
proliferative activity, with this increment being higher for the C
series. Hence, the enhancement of potency seems to be related to
the increase in lipophilicity/volume of the compounds (see the
values of clogP in Table 1), that can better fit into the markedly
hydrophobic enzyme pocket.
fluorescence spectroscopy. This technique is very useful due to the
presence of tryptophan residues at the choline binding site. All
compounds were tested with CK
study their selectivity.
a1 and some also with CKb to
The selected compounds belong to the A (a, b, d, and e) and to
the C series (n, o, t and v). In addition, in order to evaluate the effect
of the positive charge and benzyl group of the different compounds
on protein affinity, control experiments were performed with non-
quaternized forms of the cationic heads. That is, 4-
dimethylaminopyridine (DMAP), 4-pyrrolidinopyridine (PPIR), 7-
chloro-4-(perhydroazepin-1-yl)quinoline (QUINAZE) and 7-
chloro-4-(N-methylanilino)quinoline (QUINPHEN) (Fig. 1).As
depicted in Table 2, the four reference compounds showed very
little affinity (35e82 mM) for CK thus demonstrating the impor-
tance of the positive charge to achieve efficient ligand binding.
However, although their effect is not very noticeable, some affinity
for the enzyme is still shown thanks to the aromatic rings. In
general, the least related compounds were those of the A series (Kd
Compounds with quinolinic rings (c-e, j-m and s-v) lead to
lower values of GI50 (mM) with respect to pyridinium derivatives (a-
b, f-g and n-p). Moreover, the worse results are obtained when the
substituent in 4 position is a N-methylaniline. (s-t vs v). Finally, the
presence of a chlorine atom in position 7 of quinolinium moiety
greatly increases the activity up to 200-fold (in RS4; 11 t vs u).
However, this effect is negligible when the chlorine atom is found
in para position of the N-methylaniline fragment (l vs m), indicating
a non-proper accommodation of the compounds over the enzy-
between 0.4 and 24
family were recognized by the protein with a very similar affinity
(0.185e1.7 M, Table 2). A remarkable improvement in affinity is
also observed in quinoline derivatives over pyridinium for all
classes, suggesting that some type of favorable interaction with the
protein must be established by the quinoline rings. Consequently, it
mM, Table 2) while the compounds from the C
m
can be said that the compounds bind to CKa1 with an increasing
matic pocket of CK
a
1.
affinity as the complexity of their chemical structure increases,
which could be related to the type and number of interactions they
are able to establish with the residues of the active site.
3.2. Affinity and selectivity of compounds a-v for CK
a
1
For Series A, the quaternization by a benzyl group of the
endocyclic nitrogen noticeably increases the affinity with the pro-
tein, in accordance with the pharmacophore model. Also, the
The evaluation of the affinity of monocationic derivatives for the
different isoforms of CK (CK 1 and CK ) was carried out by
a
b
Scheme 1. General synthetic pathway for final compounds a-v.
3

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
Table 1
In vitro inhibitory effects of compounds a-v.
Series
Comp.
GI50
HeLa
(m
M)a
clogPb
HT-29
RS4; 11
CCRF-CEM
A
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
ꢁ0.0630
0.0509
2.19
3.31
3.84
44.2 0.4
66.0 0.6
19.6 0.8
16.2 1.1
16.3 4.6
23.2 0.9
42.1 1.9
90.8 2.1
82.3 4.4
15.0 0.6
4.7 0.37
2.1 0.2
2.6 0.09
0.74 0.02
0.95 0.41
0.91 0.12
29.9 4.9
27.5 2.9
0.28 0.07
0.36 0.05
14.8 1.8
1.7 0.06
0.83 0.1
1.9 0.1
>100
30.3 2.3
8.3 3.0
1.23 0.06
0.03 0.008
3.0 0.6
7.9 0.9
7.6 0.26
>100
54.3 0.9
5.3 0.8
3.4 1.3
0.56 0.11
2.2 0.06
18.0 2.4
12.0 2.6
>100
38.0 0.6
4.1 0.16
4.6 0.14
4.5 0.16
11.2 0.37
8.1 0.22
>100
>100
12.9 0.37
5.4 1.4
3.0 0.12
1.3 0.12
0.29 0.01
0.16 0.07
0.23 0.03
24.7 2.3
34.0 1.5
0.04 0.003
0.09 0.009
17.5 0.30
0.85 0.12
0.4 0.2
R
B
0.82
0.934
0.353
0.103
3.07
4.19
4.73
5.47
1.82
1.93
4.36
1.35
1.10
4.08
4.44
5.19
5.73
R
>100
1.0 0.28
3.2 0.5
>100
3.0 1.1
F3C
4.6 0.37
2.1 0.16
2.0 0.05
0.43 0.03
0.18 0.10
0.30 0.05
11.8 3.5
16.9 3.5
0.04 0.01
0.15 0.06
8.6 1.8
2.7 0.3
1.3 0.03
0.83 0.14
0.22 0.03
0.14 0.02
2.6 0.75
1.7 0.8
0.05 0.006
0.06 0.002
12.8 4.5
1.4 0.17
0.17 0.04
1.0 0.3
C
R
s
t
u
v
1.7 0.06
0.61 0.20
0.21 0.09
RSM932Ac
MN58bc
1.9 0.4
a
GI50 ¼ compound concentration required to inhibit tumor cell proliferation by 50%. Values are the mean SE from the dose-response curves of at least three independent
experiments carried out in triplicate.
b
The values of clogP were calculated with Chemdraw 15.0.
Data taken from ref. 36 Scientific Reports. 2016, 6, 2379.
c
Table 2
Tryptophan fluorescence spectroscopy choline kinase data. Selected compounds for
fluorescence assays and Kd values versus CK
a
1 and CK
b.
DMAP
35 18
a
PPIR
49
QUINAZE
QUINPHEN
82 12
e
K
d (mM)a CK
Series A
a
1
3
45
3
b
d
K
K
d (mM) CK
a
b
1
14 0,7
e
24 3.5
e
0.40 0.04
33 5 (83)c
0.89 0,047
36 5 (41)c
d (mM)b CK
Series C
n
O
t
v
Fig. 1. Unquaternized control ligands.
K
K
d (mM) CK
d (mM) CK
a
b
1
1.7 0.045
e
0.67 0,002
e
0.37 0.80
0.185 0.024
20 5 (108)c
63 3 (171)c
a,bKds values of indicated compounds for CK
a1 or CKb are represented as mean S.D.
increment in lipophilicity of compounds d and e (clog P Table 1)
supports the influence of this molecular property on the binding
affinity. Lastly, the slightly greater affinity shown by compound
d with respect to e suggests that the phenyl ring of the N-methyl-
aniline group of the pharmacophoric hypothesis is not essential to
achieve an efficient binding to the protein.
of at least three independent experiments.
c
Fold increase respect to CKa1.
the chemical fragments present in the pharmacophore (three aro-
matic rings: benzyl, pyridine and N-methylaniline, positive charge
and exocyclic nitrogen), the most important to achieve an efficient
binding to CK are the delocalized positive charge in the hetero-
aromatic ring, the benzyl group attached to the charged atom and
the exocyclic nitrogen. The remaining functional group, i.e. benzene
from the N-methylaniline group does not appear to be indispens-
able, but if present, in some cases (compound v) it may contribute
slightly to improving the affinity of the compounds.
The affinity for CKa1 of the compounds of the series C is greater
than that of the compounds of the previous series. Noticeably, the
improvement in affinity in relation to their analogues of the series A
is greater in the Kd pyridine derivatives (n and o), than the quin-
oline derivatives (t and v). Again, it is observed that quinoline de-
rivatives bind better to the protein than pyridines and specifically,
the compound with 7-chloro-4-(N-methylanilino)quinoline (v)
shows a greater affinity than t.
Consequently, it can be inferred that the affinity of A-series
compounds for CK can be improved by introducing an additional
phenyl group into the spacer, as observed for the final C-series
compounds. In addition, the similarity of affinities between com-
pounds with and without the N-methylaniline phenyl group
demonstrates once again that it is not essential to achieve protein
binding. However, if present, it improves the binding affinity, as
does the additional phenyl of the quinolinium ring.
On the other hand, as it has recently been shown that a potent
anticarcinogenic effect can only be achieved when the expression
of CK
[26]. So, the selective binding of some of these new synthetic de-
rivatives to the isoform CK 1 was evaluated. For this, fluorescence
experiments were carried out against the isoform CK . According to
the Kd values obtained, these compounds are selective for CK 1,
a1 is specifically decreased without affecting the levels of CKb
a
b
a
with compound t showing the highest selectivity (Table 2). The
selective character of the ligands is attributed to the greater
All these data can be translated into the fact that probably of all
4

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
flexibility of Trp420 at CK
(Trp353) at CK , since in the isoform
tryptophan, instead of the phenylalanine of the isoform
reason, the selectivity of the compounds could be linked to their
binding mode to the CK, more specifically to shift of the tryptophan
residue needed to host the ligand at the site of the choline. Thus, in
view of the selectivity data it could be hypothesized that to host the
compound t, it might be necessary for the shift of Trp420 to in-
crease the size of the cavity (given the volume of the compound) or
to interact with some functional group of the ligand. In the case of
a
1 compared to that of its counterpart
1 there is a leucine behind
. For this
in affinity and inhibitory potency.
The crystalline structures of CKa1 in complex with each of the
compounds (Fig. 2) were obtained with a resolution between 1,55-
2,30 Å (Table 4) and good statistic parameters (R ¼ 0,189e0,201,
Rfree ¼ 0,214e0,257, Table 4) after an iterative process of con-
struction and refining.
Both crystalline structures are homodimeric, each monomer is
made up of a small N-terminal domain and a larger C-terminal
domain. While the ATP binding site is located in a cleft formed by
the residues of both N- and C-terminal domains, the choline site is
in a deep, hydrophobic pocket of the C-terminal domain. Surpris-
ingly, presence of two molecules of compound o at the choline
binding site was observed (Fig. 2b). Conversely, only one molecule
of compound b was observed (Fig. 2a). This is probably due to the
fact that the positioning of its ligands in this common site is more
favorable from an energetic point of view, and only occasionally
does it obtain enough energy to introduce the ligands in the second
site.
b
a
b
the isoform
a such displacement would be possible, while in the
case of the isoform CK
phenylalanine.
b, it would be prevented by the bulkier
3.3. Inhibition of human CKa1
Taking into account the second catalytic mechanism proposed
for CK [27], this implies that potential inhibitors that have been
designed to act at the level of the choline site must hinder the
The compounds belonging to the A and B series were those that
showed a weaker affinity and inhibitory activity. According to the
nucleophilic attack of Asp306 on the phosphate
g of the ATP or, if
this occurs, to be effective, they should not be displaced from the
active site after phosphorylation of that residue. By inspecting the
properties of the choline binding site, adequate ligand affinity is
achieved with interactions engaging aromatic residues of this hy-
drophobic pocket. Accordingly, functional groups amenable to such
interactions should be brought together in a single ligand according
to the geometrical features reported in the pharmacophore model,
including a positive charge, three aromatic rings and an exocyclic
nitrogen [22].
evidence provided by the crystal of the complex CK
affinity may be due to the formation of weak interactions with the
residues of the choline site. Thus, only two not parallel in-
teractions are observed with Tyr354 and Phe345 (Fig. 3a). The rest
of the amino acids that make up the hydrophobic site where b is
inserted do not contribute to the stabilization of the ligand. These
amino acids are Tyr440, Trp420, Trp423 and Tyr437.
a1/b, the lack of
p-p
The low inhibitory power may also be due to the proximity of
the compound to Asp306. This amino acid, which has been pro-
The results of the effect of tested compounds on the CK
a
1 ac-
posed as a catalytic base, performs a nucleophilic attack at the g
tivity are shown in Table 3. In general, it can be observed that the
effects on the inhibition are better for the more voluminous the
ligand, i.e. starting from series A to series B and then to series C. The
highest inhibition values were obtained for compounds t and v
phosphate of the ATP and is phosphorylated. This ligand, which
does not establish strong interactions with the residues of the
active site, could easily be displaced from the hydrophobic pocket
by the phosphorylated residue. Therefore, the choline can enter the
choline pocket and become phosphorylated.
(94% and 76% to 30 mM respectively), in good agreement with their
cytotoxic potency (Table 1). Considered that both compounds
presented a low value of Kd, these results confirm that the greater
the affinity of the ligand for the protein, the greater the inhibitory
A similar mode of interaction must exist for compound a, since it
shows affinity and inhibition values very similar to b. However, the
quinoline ring of compounds c, d and e, as well as the phenyl of the
N-methylaniline group present in e would lead to a diverse inter-
action profile in which aromatic systems would participate in the
formation of additional interactions. As a result, the affinity for CK
should be greater. However, according to the weak inhibition dis-
played, such additional interactions do not appear to be sufficient
to keep them stably attached to the active site, so Asp306 phos-
phorylation could also cause their expulsion from the choline
pocket.
action on CK
Albeit the series A compounds bind to the choline site with an
affinity of 0.4e24 M (Table 2), no significant differences are
a1.
m
observed between the A and B series, indicating that the intro-
duction of the trifluoromethyl group as an electron-withdrawing
group has a poor effect on enzyme inhibition. In this case, the in-
crease in inhibition in quinoline derivatives is not too evident, so
that compound a inhibits in higher values to compound e. While for
compounds b and d occurs the opposite, where there is an 11% at
The crystalline structure obtained for the compounds of the
50
mM, increase in inhibitory activity for compound d (38%) in
series C was formed by the complex CKa1/o, since the compound n
relation to b (28%) (Table 3).
had already been crystallized in recently published works [18] and
the insolubility of the compounds t and u at high concentrations
prevented their crystallization. Consequently, docking studies were
also undertaken to explain the binding mode of the compound u,
which in turn could be similar to that of the compound t given the
structural similarity between the two.
In the case of compounds of series C, fluorescence indicated that
the introduction of an additional phenyl in the spacer led to an
increase in enzyme affinity with respect to series A derivatives.
According to inhibition data (Table 3), this observation is confirmed
for quinoline derivatives. Specifically, compound t has a IC50 of
6.74
m
M, while the IC50 of the compound v is 12.55
m
M. It is worth to
According to the Kd values that were obtained for them, the
affinity improved considerably with respect to the derivatives of
the series A, this improvement being especially accentuated for the
case of the pyridine derivatives.
In the crystal of the compound o (Fig. 3b) not only two mole-
cules are observed at the site of the choline, but also parallel in-
note that compounds with quinuclidinium moiety as cationic head
(h-i and q-r) moderately inhibit the enzyme and have weak anti-
proliferative effects, that can be likely ascribed to the low lip-
ophilicity of these derivatives (Table 1).
3.4. Crystallographic studies
teractions
(molecule 2). In addition, the binding of molecule 1 is reinforced
with unparallel interactions with residues Phe435 and Tyr440.
Trp420, Trp423 and Tyr437 residues are also part of the hydro-
phobic pocket where both molecules are inserted. Regarding these
p-p with the residues Tyr354 (molecule 1) and Tyr333
Crystallographic studies with at least one representative from
each family (b and o compounds) were conducted to have experi-
mental evidence in this regard and to explain the gradual increase
p-p
5

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
Table 3
Inhibition of CK
a
1 activity by compounds a-v.
Series
a
Compounds
% Inhibition CK
a
1
IC50 CK
a
1(m
M)
10
m
M
30
m
M
50 mM
A
B
C
D
E
e
37.20 8.00
19.57 6.03
35.02 0,79
27.53 5.41
28.38 0.14
e
nd
nd
nd
nd
nd
11,08 1,62
31,54 8,07
35,81 2.68
23.08 4.14
27,84 1,72
e
A
B
38,7 10,83
18,58 2,00
F
25.42 1.66
e
33.53 0.26
e
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
G
H
I
J
K
L
8.07 2.01
e
24.67 2.42
e
e
24.92 3.87
52.68 0.65
e
34.74 3.71
e
M
10.41 2.54
14.50 6.05
N
O
P
Q
R
S
18.85 1.47
e
26.45 5,00
28,03 1,06
e
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
e
C
26.57 5.54
45.59 3.06
36.87 2.56
74.39 0.93
38.61 5.30
30.19 7.20
53.03 1.00
46.28 7.02
94.44 0.33
76.00 2.19
T
V
6.74 0.13
12.55 0.16
a
IC50 compound concentration that inhibits 50% of the enzyme activity.
Fig. 2. Crystals from CKa1 selected for the compounds: a) compound b, b) compound o.
interactions, the bonding of the series C compounds is stronger
than in the case of the series A compounds, probably thanks to the
additional phenyl of the spacer. The result, therefore, is an
improvement in affinity to CKa1 which is particularly notable in the
case of pyridine derivatives and should lead to an increase in po-
tency as inhibitors.
Crystallographic studies previously carried out with the com-
pound n and other derivatives with dimethylaminopyridine as
cation head (see Protein Data Bank, PDB IDs: 4BR3 and 4CGA) have
shown that, although there is only one molecule occupying to the
introduces molecule 1 (with an ordinary arrangement where it
establishes interactions parallel with Tyr354), and then
p-p
molecule 2 (in the new cavity that Trp420 and Tyr333 open).
However, inhibition assays show that both n and o are not able
to interrupt phosphocholine synthesis, so the interactions previ-
ously described do not seem to be sufficient to prevent the enzyme
from expelling them from the active site.
3.5. Docking studies
site of the choline, it has parallel interactions of the type p-p with
Based on docking studies conducted with quinoline derivative v
(Fig. 5), and depending on the observed Kd values, their binding
mode to the choline binding site could occur in two different ways.
In the first pose, the cationic head is inserted deep into the pocket
of the choline, while in the second binding mode it is exposed to
the solvent. However, in both cases, it is observed that their mo-
the residues Tyr354, Phe435 and Trp420 (Fig. 4). This would explain
the better affinity of the compound n with respect to its equivalent
a. On the other hand, the comparison of the conformation adopted
by the residues that make up the hydrophobic pocket where both
compounds are inserted shows the remarkable rotation of Trp420
and Tyr333 residues that allow the accommodation od the second
molecule of compound. So much so that it can be postulated that
this conformational change would require an extra energy contri-
lecular recognition will be reinforced by extra interactions of the
p-
cation type, parallel and non-parallel through the additional
p-p
phenyl of the spacer, quinolinium group and N-methylaniline. This
may explain their better affinity to n and o compounds, as well as
bution that would be sufficient reason to think that CKa1
6

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
Table 4
X-ray crystal structure data collection and refinement statistics. Values in parentheses refer to the highest resolution shell. Ramachandran plot
statistics were determined with PROCHECK.
b
O
Space group
P212121
P43212
Wavelength (Å)
Resolution (Å)
Cell dimensions (Å)
0.97
0.92
20e2.15 (2.27e2.15)
a ¼55.67
b ¼ 118.77
c ¼ 131.00
48047
99.9 (100)
0.092 (0.715)
16.7 (3.1)
8.2 (8.2)
0.191/0.233
0.0086
61e1.80 (1.90e1.80)
a ¼61.37
b ¼ 61.37
c ¼ 221.07
40416
99.9 (99.9)
0.1 (0.653)
12.6 (2.8)
8.5 (8.7)
0.189/0.230
0.011
Unique reflections
Completeness
Rmerge
I/s(I)
Redundancy
Rwork/Rfree
RMSD from ideal geometry, bonds (Å)
RMSD from ideal geometry, angles (º)
<B> overall (Å2)
1.527
39.65
84.96
44.74
1.385
22.06
29.95
37.07
<B> ligand (Å2)
<B> solvent (Å2)
<B> ethylene glycol (Å2)
Ramachandran plot:
Most favoured (%)
Additionally allowed (%)
Outliers (%)
66.71
38.31
96.9
2.9
0.1
97.7
2.30
0.00
7A06
PDB ID
7A04
Fig. 3. a) Crystalline structure of the complex CKa1/b. b) Crystalline structure of the complex CKa1/o. The hydrophobic residues with which parallel interactions are formed p-p are
shown in green. The rest of the residues that make up the hydrophobic pocket have been indicated in white. (For interpretation of the references to colour in this figure legend, the
reader is referred to the Web version of this article.)
their e and d analogues. Specifically, the residues involved in the
interaction -cation are Tyr440 in the first pose, and Trp420 in the
second. On the other hand, the parallel interactions of the first
same way as compound 14 (BR33) [19] (Fig. 6). The superimposi-
tion however shows that the residues involved in the recognition of
the three ligands are not the same. In this sense, the parallel
p
p-p
pose would be established with residues Phe435 and Trp420,
whereas according to the second pose, only the Tyr354 would be
interaction p-p that forms molecule 2 of compound o with Tyr333
is remarkable. This residue moves backwards to accommodate in
the most stable way possible that molecule (Fig. 6d). In the case of
engaged in a
p-p interaction. Finally, the non-parallel p-p in-
teractions would form with Phe435, Tyr354 and Trp423 residues in
the first pose, while in the second one Trp420, Phe435, Tyr440 and
Trp428 residues seem to intervene.
b, it does not establish parallel
acid, as previously indicated. On the other hand, compound BR33
forms three interactions, of which two are parallel with Tyr354
p-p interactions with any amino
p-p
Table 5 summarizes the interactions that the members of each
series establish with the residues from the choline binding site.
and Tyr333 residues. The amino acid Trp420 is responsible for the
third interaction (Fig. 6b). Interestingly, molecule 1 of compound o
superimposes its phenyl group attached directly to the cationic
head with the phenyl of compound BR33 (Fig. 6c). This arrange-
3.6. Comparison of the crystals of series A and C between
ment allows it to establish parallel interactions
p-p with Tyr354.
themselves and with others previously published
Molecule 1 of o forms non-parallel interactions with Phe435
p- p
and with Tyr440, respectively. If the poses of molecules 1 and 2 of
In this group of compounds, two types of ligands can be
the ligand o are compared, it can be seen that their biphenyl groups
differentiated, b and o, which are introduced into the hill site in the
7

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
Fig. 4. Interactions of compound n and analogues with residues from the choline site. a) Superposition of compound n (PDB ID: 4CG9) and its derivatives from PDB code crystals:
4CGA and 4BR3. b) Residues from the active site with which they form parallel interactions . c) Superposition of n and derivatives with o and the compound BR-33. d)
p-p
Conformational changes of the residues of the active site (Trp420 and Tyr 333) against the derivatives of dimethylaminopyridine (DMAP) and pyrrolidinopyridine (PPIR).
Fig. 5. First and second pose adopted by the compound v in the docking studies performed. All interactions established are indicated in the figures.
Table 5
Interaction profile of the members of each series. Table 5.
Compounds
n-p
s-v
b
P
-
p
Parallel
Unparallel
W420, F435, Y354
W423
Y333, Y437, Y440
Y354 (molecule 1), Y333 (molecule 2)
F435 (molecule 1, Phe’, Y440 (molecule 1, Phe’)
W420, W423, Y437
e
Interactions
Hydrophobic pocket
Relevant aspects
Y354, F345
W420, W423, Y437, Y440
Y354 helps to stabilize all ligands regardless of the chemical nature of the spacer and cation head
8

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
are arranged parallel to each other (Fig. 7), in such a way that the
formation of non-parallel
favoured.
p-p interactions between them is
In conclusion, each of these interaction profiles is responsible for
a higher Kd value than that of the quinoline derivatives. In view of
the above, this observation may respond to the formation of fewer
parallel interactions p-p.
The typical disposition of dimethylaminopyridine derivatives
(DMAP) at the choline binding site involves the formation of three
parallel p-p interactions with Trp420, Tyr354 and Phe435 residues
(Fig. 4). All of these are the cause of the stronger affinity that DMAP
derivatives generally show with pyrrolidinopyridine (PPIR). An
interaction between the amino acid Trp423 and the pyridine ring of
compound n can also be seen at
p-p The DMAP fragment is
superimposed in all cases, however, the rest of the molecule is ar-
ranged differently depending on the spacer (Fig. 4). This confirms
once again that the part of the molecule that is really important for
interacting with the CK protein is the pharmacophore that has been
previously proposed.
The comparison of the residues orientation involved in the
interaction of DMAP and PPIR derivatives, showed the turnaround
of the Trp420 and Try333 residues to facilitate the binding of PPIR
derivatives (Fig. 4d).
Fig. 7. Cross arrangement of the two molecules of o.
Conversely, within the C series, we observed that the anti-
proliferative activity of compounds o and p directly correlates with
the inhibition in choline uptake, while for the compounds t and v,
the antiproliferative action is due both to the inhibition of CKa1 and
choline uptake. Further investigations are needed to better deter-
mine the relationship between the degree of inhibition of choline
uptake and the antiproliferative effect, as well as the correlation
with the compound structures, without ruling out that other
mechanisms may contribute to antiproliferative SAR.
3.7. Inhibition uptake of choline assays
Since it has been previously demonstrated that some CKa1 in-
hibitors, in addition to their effects, are able to reduce choline up-
take [28e30], we decided to investigate whether the test
compounds are also capable of inhibiting choline uptake, as
described in previous research conducted by our group [14e16].
The A series inhibitors have not been tested as they have lower
antiproliferative activity and lower capacity to inhibit the CKa1
enzyme. Compounds f, h and i were chosen from the B series. The
first two inhibit the enzyme by 33% and 8%, respectively, while the
3.8. Compound t induce G0/G1 arrest of the cell cycle and apoptosis
through the mitochondrial pathway
Previous reports indicated that molecules that are able to inhibit
choline kinase significantly inhibits cell growth in different tumoral
cell lines, in a concentration dependent manner, due to a G0/G1
arrest of the cell cycle [13,16,20].
third is unable to inhibit CK
a
1 activity (Table 3). Among those three
In this context we evaluated the cell cycle after treatment with
compound t, one of the most active derivatives in two cancer cell
lines (HeLa and HT-29). As depicted in Fig. 8, the drug induces a
significant accumulation in G0/G1 phase in both cell lines, although
highly pronounced in HT-29 cells, accompanied by the consequent
decrease of the S-phase, in a concentration-dependent manner
(Fig. 8, Panels A and B).
inhibitors, only compound f shows antiproliferative activity in HT-
29 cells. As shown in Table 6, none of these three compounds
potently inhibits choline uptake in either HT-29 cells or HepG2 cells
(the latter was chosen to corroborate results in a different cell line).
The results highlighted the hypothesis that the antiproliferative
effect obtained with B series compounds is mainly due to the in-
hibition of CK
a1 activity.
To evaluate the mode of cell death induced by compound t, we
Fig. 6. a) Superposition of the crystalline structures of compounds b (orange), o (blue) and BR33 (magenta). b) Some of the residues with which
p-p interactions are formed. c)
Superposition of the pyridinium groups of the two molecules of compound o with the pyridinium group of compound BR33 and proximity with the pyridinium of b. d) Backward
rotation of Tyr333 to accommodate molecule 2 of compound o and the N-methylaniline group of BR33. (For interpretation of the references to colour in this figure legend, the
reader is referred to the Web version of this article.)
9

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
Table 6
Inhibition of [3H]choline uptake by compounds a-v in HT-29 and HepG2 cells. Values of percentage of inhibition of choline uptake in HT-29 and HepG2 cells.
Compounds
GI50
(m
M) HT-29
Choline uptake HT-29% inhibition
Choline uptake
HepG2
% inhibition
10
mM
30
mM
10
mM
30 mM
A
B
d
e
f
h
I
N
O
P
S
T
V
>100
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
11.7 0.8
5.1 2.4
14.3 4.2
nd
70.3 0.77
83.3 0.71
nd
88.2 0.84
87.2 0.03
nd
nd
nd
nd
16.9 0.8
17.7 2.2
15.1 1.7
nd
84.3 0.79
94.8 0.24
nd
95.5 0.21
94.8 0.34
38.0 0.6
4.6 0.14
4.5 0.16
11.2 0.37
>100
14.3 1.15
9.1 0.99
17.6 0.76
nd
56.7 1.6
78.4 0.20
nd
22.7 1.06
2.4 1,29
21.6 3.8
nd
79.5 0.33
95.0 0.34
nd
>100
0.29 0.01
0.16 0.07
0.23 0.03
0.04 0.003
0.09 0.009
0.85 0.12
86.9 1.08
85.5 1.00
96.7 0.14
93.3 0.55
performed a bi-parametric cytofluorimetric analysis using propi-
dium iodide (PI) and annexin-V-FITC, which stain DNA and phos-
phatidylserine (PS) residues, respectively. As showed in Fig. 8
(panels C and E), t induces a dose- and time-dependent increase
in apoptotic cells in both cell lines cell lines. In order to evaluate the
type of apoptosis induced by the compound, we analyzed the
variation of mitochondrial potential following treatment. It is well
known that many drugs cause apoptosis through the so-called
intrinsic pathway in which mitochondria play an essential role in
the propagation of apoptosis [31]. In particular at an early stage,
apoptotic stimuli alter the mitochondrial transmembrane potential
Scientific SMP3 electrothermal melting-point apparatus and were
uncorrected. Analytical thin-layer chromatography (TLC) was per-
formed using Merck Kieselgel 60 F254 aluminium plates and
visualized by UV light or iodine. All evaporation occurred in vacuo
with a Büchi rotary evaporator and the pressure controlled by a
Vacuubrand CVCII apparatus. For flash chromatography, Merck
silica gel 60 with a particle size of 0.040e0.063 mm (230e400
mesh ASTM) was used: 500 MHz 1H and 126 MHz 13C NMR Varian
Direct Drive; a 400 MHz 1H and 101 MHz 13C NMR Varian Direct
Drive spectrometers at room temperature. Chemical shifts (d) are
quoted in parts per million (ppm) and are referenced to the residual
solvent peak. Spin multiplicities are given as s (singlet), bs (broad
singlet), d (doublet), t (triplet), ddd (doublet doublet doublet), dd
(doublet doublet), dt (doublet triplet), q (quadruplet) and m
(multiplet). High-resolution NanoAssisted Laser Desorption/Ioni-
zation (NALDI-TOF) or Electrospray Ionization (ESI-TOF) mass
spectra were carried out on a Bruker Autoflex or a Waters LCT
Premier Mass. Organic solutions were dried over anhydrous
Na2SO4. Solvents and reagents that are commercially available were
purchased from Aldrich (Sigma-Aldrich) or Alfa Aesar (Johnson
Matthey Company) and were used without further purification
unless otherwise noted.
(
Djmt) that ultimately leads to apoptosis. With this aim we treated
both HeLa and HT-29 cell with t and after 24 h the dissipation of
Djmt was monitored by flow cytometry in the presence of the
fluorescent dye JC-1. As depicted in Fig. 8 (panels D and F), com-
pound t is able to significantly reduce mitochondrial potential early
after treatment and in a concentration dependent manner in both
the two cell lines investigated.
4. Conclusions
In this paper we have presented a series of compounds (a-v)
that were initially designed and synthesized to validate our previ-
ously described pharmacophore model (RRRPN). These compounds
were tested against four tumor cell lines with good or moderate
antiproliferative properties, where the most complex structures
(series C) turned out to be the best choline kinase inhibitors. The
5.1.2. General procedure for the synthesis of the target compounds
To a prepared solution of 4-Bromobenzyl, 1-(bromomethyl)-4-
(trifluoromethyl)benzene or 4-(bromomethyl)-1,10-biphenyl in
butanone or acetonitrile as solvent, was added the corresponding
intermediates 1e3 and (7-chloro)-4-subtitutedquinoline 4 [3,6,7]
(in a molar ratio of 1:1). The reaction was performed in a sealed
tube a under microwave irradiation (140 ꢀC, 28e33 min). Target
compounds a-v were isolated after chromatography flash purifi-
cation or simple filtration and wash with diethyl ether.
affinity (Kd), selectivity (versus
b isoform) and degree of inhibition
were evaluated on the isolated human choline kinase. Our results
showed a poor enzyme inhibition, which do not correlate with
antiproliferative activity. Indeed, we observed that the compounds
are also capable of inhibiting the uptake of choline and this effect
could in part contribute to their antiproliferative activity. However,
it is still not ruled out that these compounds could also be involved
in other mechanisms related to the hitherto underestimated role of
choline kinase alpha in protein-protein interactions. So, further
studies are needed to clarify if our compounds have the additional
benefit of disrupting the surface of the protein, regardless of their
action on enzymatic activity.
5.1.2.1. 1-benzyl-4-(dimethylamino)pyridin-1-ium
bromide
(a).
Following general procedure, after workup as described previously,
compound a was isolated as a white solid. Yield: 88%, mp
210e212 ꢀC. 1H NMR (500 MHz, CD3OD)
d 3.26 (s, 6H), 5.38 (s, 2H),
7.01 (d, J ¼ 7.8 Hz, 2H), 7.44e7.38 (m, 5H), 8.23 (d, J ¼ 7.8 Hz, 2H). 13
C
NMR (125 MHz, CD3OD) d 38.98 (2C), 60.38 (1C), 107.74 (2C), 127.93
(2C), 128.88 (1C), 129.06 (2C), 134.79 (1C), 141.73 (2C), 156.75 (1C).
HRMS-m/z[M ꢁ Br]þ calcd for C14H17N2þ [M ꢁ Br]þ: 213.1392;
found, 213.1387.
5. Experimental protocols
5.1. Chemistry
5.1.2.2. 1-benzyl-4-(pyrrolidin-1-yl)pyridin-1-ium
bromide
(b).
5.1.1
Following general procedure, after workup as described previously,
Materials and Methods.
Melting points were taken in open capillaries on a Stuart
compound b was isolated as a white solid. Yield: 82%, mp
226e228 ꢀC. 1H NMR (500 MHz, CD3OD)
d
2.12 (q, J ¼ 7.0, 4H), 3.56
10

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
Fig. 8. Percentage of cells in each phase of the cell cycle in HeLa (Panel A) and Jurkat cells (Panel B) treated with compound t at the indicated concentrations for 24 h. Cells were
fixed and labelled with PI and analyzed by flow cytometry as described in the Experimental Section. Data are presented as mean SEM of three independent experiments. Flow
cytometric analysis of apoptotic cells after treatment of HeLa (Panels A and B) and Jurkat (Panels C and D) cells with t at the indicated concentrations after incubation for 24 (Panels
A and C) or 48 h (Panels B and D). The cells were harvested and labelled with annexin-V-FITC and PI and analyzed by flow cytometry.
Assessment of mitochondrial membrane potential (Djmt) (Panel A) and production of ROS (Panel B) after treatment of HeLa cells with compound t. Cells were treated with the
indicated concentration of compound for 3, 6, 12 and 24
h and then stained with the fluorescent probe JC-1 for analysis of mitochondrial potential or with 2,7-
dichlorodihydrofluorescein diacetate (H2-DCFDA).
(t, J ¼ 6.9 Hz, 4H), 5.37 (s, 2H), 6.87 (d, J ¼ 7.7 Hz, 2H), 7.38e7.45 (m,
135.79 (1C),140.94 (1C),141.61 (1C),146.34 (1C),157.10 (1C). HRMS-
5H), 8.21 (d, J ¼ 7.7 Hz, 2H). 13C NMR (125 MHz, CD3OD)
d
24.76
m/z[M ꢁ Br]þ calcd for C20H20ClNþ2 : 323.1310; found, 323.1315.
(2C), 48.30 (2C), 60.38 (1C), 108.28 (2C), 127.84 (2C), 128.81 (1C),
129.00 (2C), 134.96 (1C), 141.51 (2C), 153.84 (1C). HRMS-m/z
[M ꢁ Br]þ calcd for C16H19N2þ: 239.1548; found, 239.1542.
5.1.2.4. 4-(azepan-1-yl)-1-benzyl-7-chloroquinolin-1-ium bromide
(d). Following general procedure, after workup as described pre-
viously, compound d was isolated as a white solid. Yield: 53%, mp
5.1.2.3. 1-benzyl-7-chloro-4-(pyrrolidin-1-yl)quinolin-1-ium
bro-
225e227 ꢀC. 1H NMR (400 MHz, CD3OD)
d 1.76e1.72 (m, 4H),
mide. (c). Following general procedure, after workup as described
previously, compound c was isolated as a white solid. Yield: 79%,
2.08e2.01 (m, 4H), 4.09 (t, J ¼ 5.5 Hz, 4H), 5.80 (s, 2H), 7.10 (d,
J ¼ 3.8 Hz, 1H), 7.28e7.30 (m, 2H), 7.43e7.35 (m, 3H), 7.64 (dd,
J ¼ 1.6, 9.2 Hz, 1H), 7.98 (d, J ¼ 2.04 Hz, 1H), 8.37 (d, J ¼ 9.2 Hz, 1H),
mp 256e258 ꢀC. 1H NMR (401 MHz, CD3OD)
d 2.16 (s, 4H), 4.01 (s,
4H), 5.76 (s, 2H), 6.86 (d, J ¼ 7.6 Hz, 2H), 7.23 (d, J ¼ 7.0 Hz, 1H),
7.41e7.26 (m, 2H), 7.63 (pt, J ¼ 7.6 Hz, 1H), 7.88e7.80 (pt, J ¼ 7.5 Hz,
1H), 7.92 (d, J ¼ 8.8 Hz, 1H), 8.51 (d, J ¼ 7.6 Hz, 1H), 8.61 (d,
8.51 (d, J ¼ 7.8 Hz, 1H). 13C NMR (150 MHz, CD3OD)
d 27.63 (2C),
28.29 (2C), 55.23 (2C), 58.44 (1C), 104.25 (1C), 118.85 (1C), 118.94
(1C), 126.53 (1C), 127.86 (2C), 129.76 (1C), 130.38 (2C), 131.52 (1C),
135.68 (1C), 140.94 (1C), 142.06 (1C), 146.16 (1C), 160.68 (1C).
HRMS-m/z[M ꢁ Br]þ calcd for C22H24N2Clþ: 351.1628; found,
351.1631.
J ¼ 8.6 Hz, 1H). 13C NMR (101 MHz, CD3OD)
d (126 MHz, CD3OD):
26.50 (2C), 54.44 (2C), 58.58 (1C), 103.85 (1C), 118.84 (1C), 119.28
(1C), 126.87 (1C), 127.82 (2C), 129.73 (1C), 130.37 (2C), 131.16 (1C),
11

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
5.1.2.5. 1-benzyl-7-chloro-4-(methyl(phenyl)amino)quinolin-1-ium
bromide (e). Following general procedure, after workup as
described previously, compound e was isolated as a white solid.
(2C), 128.97 (1C), 129.85 (2C), 139.00 (1C), 139.81 (1C), 140.12 (1C),
144.96 (1C), 155.81 (1C). HRMS-m/z[M
ꢁ
Br]þ calcd for
C
21H19ClN2Fþ3 : 391.1185; found, 391.1185.
Yield: 87%, mp 259e261 ꢀC. 1H NMR (600 MHz, CD3OD)
d 2.46 (s,
3H), 5.82 (s, 2H), 6.89 (d, J ¼ 7.4 Hz, 1H), 7.30e7.28 (m, 2H),
7.44e7.37 (m, 8H), 7.80 (dd, J ¼ 1.9, 9.0 Hz, 1H), 8.08 (d, J ¼ 1.9 Hz,
1H), 8.55 (d, J ¼ 7.5 Hz, 1H), 8.62 (d, J ¼ 9.0 Hz, 1H). 13C NMR
5.1.2.11. 4-(azepan-1-yl)-7-chloro-1-(4-(trifluoromethyl)benzyl)qui-
nolin-1-ium bromide (k). Following general procedure, after
workup as described previously, compound k was isolated as a
white solid. Yield: 78%, mp 146e148 ꢀC. 1H NMR (300 MHz, CD3OD)
(125 MHz, CD3OD) d 19.82 (1C), 57.64 (1C), 100.48 (1C), 117.12 (1C),
118.36 (1C), 125.17 (2C), 125.63 (1C), 126.38 (2C), 127.71 (1C), 128.48
(1C), 129.03 (2C), 129.86 (1C), 130.47 (2C), 134.20 (1C), 138.52 (1C),
139.38 (1C), 140.67 (1C), 147.65 (1C), 155.98 (1C). HRMS-m/z
[M ꢁ Br]þ calcd for C23H20N2Clþ: 359.1315; found, 359.1306.
d
1.74e1.70 (m, 4H), 2.00e2.10 (m, 4H), 4.07 (t, J ¼ 5.6 Hz, 4H), 5.87
(s, 2H), 7.12 (d, J ¼ 7.8 Hz, 1H), 7.45 (d, J ¼ 8.1 Hz, 2H), 7.62 (dd,
J ¼ 9.2, 2.0 Hz, 1H), 7.70 (d, J ¼ 8.0 Hz, 2H), 7.91 (d, J ¼ 2.0 Hz, 1H),
8.38 (d, J ¼ 9.2 Hz, 1H), 8.52 (d, J ¼ 7.8 Hz, 1H). 13C NMR (101 MHz,
CD3OD) d 28.70 (4C), 54.69 (2C), 57.13 (2C), 103.70 (1C), 118.00 (1C),
5.1.2.6. 4-(dimethylamino)-1-(4-(trifluoromethyl)benzyl)pyridin-1-
ium bromide (f). Following general procedure, after workup as
described previously, compound f was isolated as a white solid.
118.35 (1C), 126.07 (1C), 126.59 (1C), 126.62 (1C), 127.83 (2C), 131.05
(2C), 139.75 (1C), 140.61 (1C) 141.45 (1C), 145.68 (1C), 160.22 (1C).
HRMS-m/z[M ꢁ Br]þ calcd for C23H23ClN2Fþ3 : 419.1502; found,
419.1512.
Yield: 75%, mp 216e218 ꢀC. 1H NMR (300 MHz, CD3OD)
d 3.37 (m,
6H), 5.50 (s, 2H), 7.03 (d, J ¼ 7.6 Hz, 2H), 7.56 (d, J ¼ 8.0 Hz, 2H), 7.74
(d, J ¼ 8.2 Hz, 2H), 8.26 (d, J ¼ 7.6 Hz, 2H). 13C NMR (75 MHz, CD3OD)
5.1.2.12. 7-chloro-4-(methyl(phenyl)amino)-1-(4-(trifluoromethyl)
benzyl)quinolin-1-ium bromide (l). Following general procedure,
after workup as described previously, compound l was isolated as a
yellow solid. Yield: 81%, mp 227e229 ꢀC. 1H NMR (500 MHz,
d
40.36 (2C), 60.49 (1C), 109.09 (2C), 127.20 (2C), 129.59 (2C), 132.50
(1C), 132.52 (1C), 140.13 (1C), 142.74 (2C), 157.66 (1C). HRMS-m/z
[M ꢁ Br]þ calcd for C15H16N2Fþ3 : 281.1266; found, 281.1272.
CD3OD)
d
2.44 (s, 3H), 5.95 (s, 2H), 6.91 (d, J ¼ 7.5 Hz,1H), 7.24e7.53
5.1.2.7. 4-(pyrrolidin-1-yl)-1-(4-(trifluoromethyl)benzyl)pyridin-1-
ium bromide (g). Following general procedure, after workup as
described previously, compound g was isolated as a white solid.
(m, 7H), 7.71 (d, J ¼ 8.1 Hz, 2H), 7.81 (dd, J ¼ 9.0, 1.9 Hz, 1H), 8.05 (d,
J ¼ 2.0 Hz, 1H), 8.62 (dd, J ¼ 12.6, 8.3 Hz, 2H). 13C NMR (101 MHz,
CD3OD)
d (126 MHz, CD3OD): 19.75 (1C), 56.85 (1C), 100.73 (1C),
Yield: 84%, mp 74.8e76.3 ꢀC. 1H NMR (300 MHz, CD3OD)
d
2.14 (m,
117.20 (1C), 118.04 (2C), 125.18 (2C), 125.79 (3C), 126.95 (2C), 127.84
(1C), 130.48 (3C), 133.82 (1C), 138.64 (1C), 138.82 (1C), 139.33 (1C),
140.95 (1C), 147.87 (1C), 156.15 (1C). HRMS-m/z[M ꢁ Br]þ calcd for
4H), 3.56 (m, 4H), 5.50 (s, 2H), 6.87 (d, J ¼ 7.6 Hz, 2H), 7.58 (d,
J ¼ 8.0 Hz, 2H), 7.76 (d, J ¼ 8.2 Hz, 2H), 8.25 (d, J ¼ 7.6 Hz, 2H). 13
C
NMR (75 MHz, CD3OD)
d
26.17 (2C), 49.15 (2C), 60.98 (1C), 109.69
C
24H19ClN2Fþ3 : 427.1189; found, 427.1195.
(2C), 125.39 (1C), 127.31 (2C), 129.67 (2C), 130.60 (1C), 140.76 (1C),
143.22 (2C), 154.99 (1C). HRMS-m/z[M ꢁ Br]þ calcd for C17H18N2F3þ:
307.1422; found, 307.1417.
5.1.2.13. 7-chloro-4-((4-chlorophenyl)(methyl)amino)-1-(4-(tri-
fluoromethyl)benzyl)quinolin-1-ium bromide (m). Following gen-
eral procedure, after workup as described previously, compound m
was isolated as a yellow solid. Yield: 88%, mp 298e300 ꢀC. 1H NMR
5.1.2.8. 1-(4-(trifluoromethyl)benzyl)quinuclidin-1-ium bromide (h).
Following general procedure, after workup as described previously,
compound h was isolated as a white solid. Yield: 71%, mp
(300 MHz, CD3OD)
d 3.83 (s, 3H), 6.04 (s, 2H), 7.47e7.33 (m, 4H),
7.52 (dd, J ¼ 12.5, 8.5 Hz, 4H), 7.63 (d, J ¼ 9.3 Hz, 1H), 7.73 (d,
238.3e239 ꢀC. 1H NMR (300 MHz, CD3OD)
d
2.03 (m, 6H), 2.18 (m,
J ¼ 8.1 Hz, 2H), 8.04 (d, J ¼ 1.9 Hz, 1H), 8.89 (d, J ¼ 7.5 Hz, 1H). 13
C
1H) 3.53 (m, 6H), 4.53 (s, 2H), 7.78 (d, J ¼ 8.1 Hz, 2H), 7.86 (d,
NMR (101 MHz, CD3OD) d 44.54 (1C), 57.33 (1C), 106.29 (1C), 118.03
J ¼ 8.1 Hz, 2H). 13C NMR (75 MHz, CD3OD)
d
21.48 (1C), 25.05 (3C),
(1C), 118.40 (1C), 125.86 (1C), 125.90 (1C), 126.51 (1C), 127.13 (4C),
129.56 (1C), 130.66 (4C), 133.85 (2C), 138.50 (1C), 138.53 (1C),
145.99 (1C), 147.41 (1C), 158.57 (1C). HRMS-m/z[M ꢁ Br]þ calcd for
56.22 (3C), 67.65 (1C), 127.25 (2C), 133.18 (1C), 133.45 (1C), 134.42
(1C), 135.20 (2C). HRMS-m/z[M ꢁ Br]þ calcd for C15H19NF3þ:
270.1470; found, 270.1475.
C
24H18ClN2Fþ3 : 461.0799; found, 461.0803.
5.1.2.9. 1-(4-(trifluoromethyl)benzyl)-3-hydroxy-quinuclidin-1-ium
bromide (i). Following general procedure, after workup as
described previously, compound i was isolated as a white oil. Yield:
5.1.2.14. 1-([1,10-biphenyl]-4-ylmethyl)-4-(dimethylamino)pyridin-1-
ium bromide (n). Following general procedure, after workup as
described previously, compound n was isolated as a white solid.
68%. 1H NMR (400 MHz, CD3OD)
d
2.03 (m, 2H), 2.19 (m, 1H) 3.21
Yield: 95%, mp 222e224 ꢀC. 1H NMR (500 MHz, CD3OD)
d 3.26 (s,
(dd, J ¼ 12.0 Hz, 1H), 3.45e3.37 (m, 5H), 3.74 (m. 1H), 4.21 (m, 1H),
6H), 5.42 (s, 2H), 7.02 (d, J ¼ 7.8 Hz, 2H), 7.36 (t, J ¼ 7.4 Hz, 1H),
4.55 (d, J ¼ 8.1 Hz, 1H), 4.59 (s, 2H), 7.76 (d, J ¼ 8.1 Hz, 2H), 7.85 (d,
7.51e7.40 (m, 4H), 7.61 (d, J ¼ 7.4 Hz, 2H), 7.69 (d, J ¼ 8.2 Hz, 2H),
J ¼ 8.1 Hz, 2H). 13C NMR (100 MHz, CD3OD)
d
18.85 (1C), 22.44 (1C),
8.26 (d, J ¼ 7.8 Hz, 2H). 13C NMR (125 MHz, CD3OD)
d 40.36 (2C),
28.39 (1C), 54.80 (1C), 56.21 (1C), 64.68 (1C), 65.15 (1C), 67.58 (1C),
127.16 (2C), 132.60 (1C), 133.43 (1C), 133.69 (1C), 135.02 (2C).
61.47 (1C), 109.14 (2C), 127.99 (2C), 128.86 (1C), 128.92 (2C), 129.90
(2C), 130.00 (2C), 135.07 (1C), 141.36 (1C), 143.11 (2C), 143.48 (1C),
158.04 (1C). HRMS-m/z[M ꢁ Br]þ calcd for C20H21Nþ2 : 289.1705;
found, 289.1698.
HRMS-m/z[M
286.1413.
ꢁ
Br]þ calcd for C15H19NFþ3 : 286.1419; found,
5.1.2.10. 7-chloro-4-(pyrrolidin-1-yl)-1-(4-(trifluoromethyl)benzyl)
quinolin-1-ium bromide (j). Following general procedure, after
workup as described previously, compound j was isolated as a
yellow solid. Yield: 91%, mp 221e223 ꢀC. 1H NMR (300 MHz,
5.1.2.15. 1-([1,10-biphenyl]-4-ylmethyl)-4-(pyrrolidin-1-yl)pyridin-1-
ium bromide (o). Following general procedure, after workup as
described previously, compound o was isolated as a white solid.
Yield: 94%, mp 196e198 ꢀC. 1H NMR (300 MHz, CD3OD)
d 2.12 (q,
CD3OD)
d
2.22 (m, 4H), 4.16 (m, 4H), 5.91 (s, 2H), 6.95 (d, J ¼ 7.7 Hz,
J ¼ 6.8 Hz, 4H), 3.56 (t, J ¼ 6.9 Hz, 4H), 5.41 (s, 2H), 6.87 (d, J ¼ 7.8 Hz,
2H), 7.36 (t, J ¼ 7.4 Hz, 1H), 7.48e7.43 (m, 4H), 7.61 (d, J ¼ 8.2 Hz,
2H), 7.68 (d, J ¼ 8.4 Hz, 2H), 8.24 (d, J ¼ 7.8 Hz, 2H). 13C NMR
1H), 7.47 (d, J ¼ 8.1 Hz, 2H), 7.79e7.64 (m, 3H), 7.93 (d, J ¼ 1.9 Hz,
1H), 8.55 (d, J ¼ 7.7 Hz, 1H), 8.65 (d, J ¼ 9.3 Hz, 1H). 13C NMR
(101 MHz, CD3OD)
d
25.27 (2C), 56.40 (2C), 65.45 (1C), 102.16 (1C),
(125 MHz, CD3OD) d 24.71 (2C), 48.70 (2C), 60.08 (1C), 108.28 (2C),
102.62 (1C), 117.14 (1C), 117.86 (1C), 125.57 (1C), 125.79 (1C), 126.95
126.57 (2C),127.43 (1C),127.47 (2C),128.48 (2C),128.58 (2C),133.78
12

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
(1C), 139.94 (1C), 141.62 (2C), 141.98 (1C), 153.77 (1C). HRMS-m/z
(2C), 127.09 (2C), 127.32 (1C), 127.44 (2C), 128.51 (2C), 130.08 (1C),
133.18 (1C), 139.64 (1C), 139.93 (1C), 140.69 (1C), 141.57 (1C), 144.72
(1C), 159.30 (1C). HRMS-m/z[M ꢁ Br]þ calcd for C28H28N2Clþ:
427.1941; found, 427.1933.
[M ꢁ Br]þ calcd for C22H23N2þ: 315.1861; found, 315.1863.
5.1.2.16. 1-([1,10-biphenyl]-4-ylmethyl)-4-((4-chlorophenyl)(methyl)
amino)pyridin-1-ium bromide (p). Following general procedure,
after workup as described previously, compound p was isolated as a
white solid. Yield: 84%, mp 199e201 ꢀC. 1H NMR (401 MHz, CD3OD)
5.1.2.21. 4-(azepan-1-yl)-1-([1,10-biphenyl]-4-ylmethyl)-quinolin-1-
ium bromide (u). Following general procedure, after workup as
described previously, compound u was isolated as a yellow solid.
d
3.52 (s, 3H), 5.46 (s, 2H), 6.93 (d, J ¼ 7.9 Hz, 2H), 7.41e7.29 (m, 3H),
7.46e7.39 (m, 3H), 7.49 (d, J ¼ 8.2 Hz, 2H), 7.63e7.53 (m, 3H), 7.67
Yield: 72%, mp 223e225 ꢀC$1H NMR (500 MHz, CD3OD):
d 1.73 (m,
(d, J ¼ 8.3 Hz, 2H), 8.33 (d, J ¼ 7.9 Hz, 2H). 13C NMR (125 MHz,
4H), 2.07 (m, 4H), 4.05 (m, 4H), 5.00 (s, 2H), 7.25 (d, J ¼ 7.5 Hz, 1H),
7.35 (t, J ¼ 7.4 Hz, 1H), 7.39 (d, J ¼ 8.3 Hz, 2H), 7.43 (t, J ¼ 7.6 Hz, 2H),
7.60 (d, J ¼ 7.2 Hz, 2H), 7.67 (m, 3H), 8.12 (d, J ¼ 1.9 Hz, 1H), 8.67 (d,
J ¼ 7.5 Hz, 1H), 8.21 (d, J ¼ 9.1 Hz, 1H). 13C NMR (126 MHz, CD3OD):
CD3OD)
d (126 MHz, CD3OD): 39.94 (1C), 60.50 (1C), 108.98 (2C),
126.57 (2C), 127.46 (2C), 127.53 (2C), 127.97 (2C), 128.58 (2C), 128.61
(2C), 130.68 (2C), 133.38 (1C), 134.31 (1C), 139.99 (1C), 142.00 (1C),
142.16 (1C), 142.42 (1C), 157.10 (1C). HRMS-m/z[M ꢁ Br]þ calcd for
d
23.35 (1C), 25.65 (2C), 53.36 (2C), 47.81 (1C), 57.11 (1C), 105.27
C
25H22N2Clþ: 385.1472; found, 385.1475.
(1C), 117.99 (1C), 118.69 (1C), 126.30 (1C), 126.55 (2C), 127.18 (2C),
127.38 (1C),127.50 (2C),128.55 (2C),129.35 (1C),133.06 (1C),139.93
(1C), 140.17 (1C), 140.57 (1C), 141.70 (1C), 145.95 (1C), 160.90 (1C).
HRMS-m/z[M ꢁ Br]þ calcd for C28H29N2Clþ: 413.1785; found,
413.1792.
5.1.2.17. 1-([1,10-biphenyl]-4-ylmethyl)quinuclidin-1-ium
bromide
(q). Following general procedure, after workup as described pre-
viously, compound q was isolated as a white solid. Yield: 71%, mp
281e283 ꢀC. 1H NMR (300 MHz, CD3OD)
d
7.78 (d, J ¼ 8.3 Hz, 2H),
7.66 (dd, J ¼ 8.4, 1.1 Hz, 2H), 7.59 (d, J ¼ 8.3 Hz, 2H), 7.53e7.43 (m,
2H), 7.44e7.33 (m, 1H), 4.43 (s, 2H), 3.62e3.41 (m, 6H), 2.17 (m, 1H),
5.1.2.22. 1-([1,10-biphenyl]-4-ylmethyl)-7-chloro-4-(methyl(phenyl)
amino)quinolin-1-ium bromide (v). Following general procedure,
after workup as described previously, compound v was isolated as a
yellow solid. Yield: 81%, mp 283e285 ꢀC. 1H NMR (500 MHz,
2.04e1.98 (m, 6H). . 13C NMR (151 MHz, CD3OD):
d 21.48 (1C), 24.89
(3C), 55.86 (1C), 68.79 (3C), 127.41 (1C), 128.17 (2C), 128.84 (2C),
129.25 (1C), 130.15 (2C), 134.75 (2C), 141.14 (1C), 144.88 (1C).
CD3OD)
d
2.47 (s, 3H), 5.87 (s, 2H), 6.94 (d, J ¼ 7.5 Hz, 1H), 7.48e7.37
HRMS-m/z[M ꢁ Br]þ calcd for C20H24
N
þ: 278.1909; found,
(m, 10H), 7.60 (d, J ¼ 7.2 Hz, 2H), 7.67 (d, J ¼ 8.3 Hz, 2H), 7.84 (dd,
278.1910.
J ¼ 1.9, 9.1 Hz,1H), 8.17 (d, J ¼ 1.8 Hz, 1H), 8.63 (d, J ¼ 8.5 Hz, 2H). 13
C
NMR (125 MHz, CD3OD) d 19.75 (1C), 57.36 (1C), 100.59 (1C), 117.21
5.1.2.18. 1-([1,10-biphenyl]-4-ylmethyl)-3-hydroxyquinuclidin-1-ium
bromide (r). Following general procedure, after workup as
described previously, compound r was isolated as a white solid.
(1C), 118.31 (1C), 125.18 (2C), 125.63 (1C), 126.55 (2C), 127.00 (2C),
127.37 (1C), 127.51 (2C), 127.79 (1C), 128.54 (2C), 130.48 (2C), 133.11
(1C), 133.91 (1C), 138.58 (1C), 139.41 (1C), 139.93 (1C), 140.78 (1C),
141.69 (1C), 147.65 (1C), 156.00 (1C). HRMS-m/z[M ꢁ Br]þ calcd for
Yield: 63%, mp 260e262 ꢀC. 1H NMR (300 MHz, CD3OD)
d 7.81 (d,
J ¼ 8.0 Hz, 2H), 7.69 (d, J ¼ 7.7 Hz, 2H), 7.62 (d, J ¼ 8.0 Hz, 2H), 7.50 (t,
J ¼ 7.5 Hz, 2H), 7.42 (t, J ¼ 7.2 Hz, 1H), 4.33e4.72 (m, 2H), 4.04e4.34
(m, 1H), 3.76 (ddd, J ¼ 11.6, 8.2, 2.7 Hz, 1H), 3.08e3.66 (m, 5H), 2.35
C
29H24N2Clþ: 435.1628; found, 435.1626.
5.2. Biological assays and computational studies
(m, 1H), 2.01e2.32 (m, 2H), 1.91 (ddd, J ¼ 20.1, 9.7, 4.6 Hz, 2H). 13
C
NMR (75 MHz, CD3OD)
d
18.99 (1C), 22.59 (1C), 28.68 (1C), 54.74
5.2.1. Antiproliferative assays in cancer cells
(1C), 56.06 (1C), 64.64 (1C), 65.53 (1C) 68.69 (1C), 127.31 (1C),
Human cervix carcinoma (HeLa), and human colon adenocar-
cinoma (HT-29) cells were grown in DMEM medium (Gibco, Milan,
Italy). B-acute lymphoblastic leukemia (RS4; 11) and T-acute
lymphoblastic leukemia (CCRF-CEM) cells were grown in RPMI
medium (Gibco, Milan, Italy). Both media were supplemented with
128.25 (2C),128.94 (2C), 129.34 (1C),130.24 (2C),134.81 (2C),141.19
þ
(1C), 145.06 (1C). HRMS-m/z[M ꢁ Br]þ calcd for C20H24NO
:
294.1858; found, 294.1860.
5.1.2.19. 1-([1,10-biphenyl]-4-ylmethyl)-7-chloro-4-(pyrrolidin-1-yl)
quinolin-1-ium bromide (s). Following general procedure, after
workup as described previously, compound s was isolated as a
white solid. Yield: 76%, mp 178e180 ꢀC. 1H NMR (300 MHz, CD3OD)
115 units/mL of penicillin G (Gibco, Milan, Italy), 115 mg/mL of
streptomycin (Invitrogen, Milan, Italy) and 10% FCS (Invitrogen,
Milan, Italy). Hela, Jurkat and HL-60 were purchased by DSMZ
(Braunschweig, Germany), whereas HT-29 by ATCC. Cell lines were
tested for mycoplasma contamination every 6 month by RT-PCR
analysis. Stock solutions (10 mM) of the different compounds
were obtained by dissolving them in DMSO. Individual wells of 96-
d
1.66e2.42 (m, 4H), 2.99e3.60 (m, 4H), 5.83 (s, 2H), 6.93 (d,
J ¼ 7.7 Hz, 1H), 7.26e7.51 (m, 5H), 7.53e7.75 (m, 5H), 8.04 (d,
J ¼ 2.1 Hz, 1H), 8.57 (d, J ¼ 7.7 Hz, 1H), 8.63 (d, J ¼ 9.3 Hz. 1H). 13
C
NMR (126 MHz, CD3OD):
d
26.43 (2C), 54.97 (2C), 58.29 (1C), 103.87
well tissue-culture microtiter plates were inoculated with 100 mL of
(1C), 118.84 (1C), 119.29 (1C), 126.92 (1C), 127.95 (2C), 128.42 (2C),
128.75 (1C), 128.85 (2C), 129.95 (2C), 131.16 (1C), 134.73 (1C), 141.04
(1C), 141.37 (1C), 141.65 (1C), 142.95 (1C), 146.29 (1C), 157.11 (1C).
HRMS-m/z[M ꢁ Br]þ calcd for C26H24ClNþ2 : 399.1628; found,
399.1631.
complete medium containing 8 ꢂ 103 cells. The plates were incu-
bated at 37 ꢀC in a humidified 5% CO2 incubator for 18 h prior to the
experiments. After medium removal, 100
mL of fresh medium
containing the test compound at different concentrations was
added to each well and incubated at 37 ꢀC for 72 h. Cell viability was
assayed by the (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetra-
zolium bromide (MTT) test as previously described [31]. The GI50
was defined as the compound concentration required to inhibit cell
proliferation by 50%, in comparison with cells treated with the
maximum amount of DMSO (0.25%) and considered as 100%
viability.
5.1.2.20. 1-([1,10-biphenyl]-4-ylmethyl)-4-(azepan-1-yl)-7-
chloroquinolin-1-ium bromide (t). Following general procedure,
after workup as described previously, compound t was isolated as a
white solid. Yield: 83%, mp 223e225 ꢀC. 1H NMR (300 MHz;
CD3OD)
d
1.77e1.63 (m, 4H), 2.10e2.06 (m, 4H), 4.07 (t, J ¼ 5.6, 4H),
5.81 (s, 2H), 7.12 (d, J ¼ 7.8, 1H), 7.44e7.31 (m, 5H), 7.66e7.58 (m,
5H), 8.06 (d, J ¼ 2.0 Hz, 1H), 8.37 (d, J ¼ 9.2, 1H), 8.55 (d, J ¼ 7.8, 1H).
5.2.2. Cell cycle analysis
HeLa cells were treated with the test compounds for 24 h. Cells
were harvested by centrifugation and fixed with 70% (v/v) cold
13C NMR (75 MHz; CD3OD)
d
26.86 (2C), 26.99 (2C), 53.84 (2C),
56.79 (1C), 102.37 (1C), 117.45 (1C), 117.56 (1C), 125.21 (1C), 126.52
13

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
ethanol. Cells were lysed with 0.1% (v/v) Triton X-100 containing
RNase A and stained with PI. A Beckman Coulter Cytomics FC500
instrument and MultiCycle for Windows software from Phoenix
Flow Systems were used to analyze the cells.
magensium nitrate. Both orthorhombic (space group P212121 for
crystals of the human ChoK
onal crystals (space group P43212 for crystals of the human
ChoK 1-compound o complex) grew within 3e14 days. The crys-
a1-compound b complex) and tetrag-
a
tals used in this study were cryoprotected in mother liquor solu-
tions containing 20% ethylenglycol and frozen in a nitrogen gas
stream cooled to 100 K. Diffraction data of the binary complexes
were collected at beamline I04e1 (Diamond, Oxford; experiment
MX8035-11) and XALOC (ALBA, Barcelona). The data was processed
and scaled using the XDS package [34] and CCP4 software [35]:
relevant statistics are given in Table 4.
5.2.3. Measurement of apoptosis by flow cytometry
The cells were treated with the test compounds and after
different times stained with both PI, to stain DNA, and annexin V-
fluorescein isothiocyanate, to stain membrane PS exposed on the
cell surface, following the instructions of the manufacturer (Roche
Diagnostics) of the Annexin-V-Fluos reagent.
5.3.2. Structure determination and refinement
5.2.4. Measurement of mitochondrial membrane potential
Mitochondrial potential was measured in HeLa cells by flow
cytometry as previously described [20], using the fluorescent dye
JC-1 (Molecular Probes).
The structure of the binary complex was solved by molecular
replacement using PDB entry 3G15 as a template. Initial phases
were further improved by cycles of manual model building in Coot
[36] and refinement with REFMAC5 [37]. The final model was
validated with PROCHECK [38], model statistics are given in Table 4.
Coordinates and structure factors have been deposited in the
Worldwide Protein Data Bank.
5.2.5. Cloning and purification of CK
Details about cloning and purification of human CKa1 and CKb
have been previously reported [18].
5.3.3. Determination of human choline kinase
The effect of the different inhibitors on human CK was assayed
in CK 1 purified as previously described [21]. In each experiment
DMSO-assays were consistently run in parallel as a control. DMSO
in no case exceeded a concentration of 0.1% in order to avoid un-
specific CK inhibition. CK activity was assayed by measuring the
rate of incorporation of 14C from [methyl-14C]choline (PerkinElmer,
Massachusetts, USA) into PC both in the presence or absence of
different inhibitor concentrations. Briefly, the final reaction mixture
contained 100 mM Tris (pH 8.5), 10 mM MgCl2, and 10 mM ATP, and
a1 activity
5.3. Tryptophan fluorescence
a
All compounds were prepared in 100% DMSO. Before running
the binding experiments, fluorescence property of the compounds
was evaluated measuring the fluorescence signal at varying con-
centrations in a specific buffer (25 mM Tris, 150 mM NaCl, pH 7.5)
containing an equivalent % of DMSO as the one used for the binding
assay. The excitation wavelength was 280 nm while the emission
was between 300 and 400 nm (typical emission range for protein
tryptophan).
No fluorescence signal was detected for the compounds that
could interfere with the binding assay.
The binding affinity constant (Kds) of the compounds against
human CKa1 and CKb, were measured by monitoring the quench-
20 ng of purified CKa1. After the samples were preincubated at
37 ꢀC for 5 min, the reaction was initiated with 1 mM [methyl-14C]
choline (4500 dpm/nmol) and incubated at 37 ꢀC for 10 min, the
final volume being 55 mL. The assay was stopped by immersing the
reaction tubes in boiling water for 3 min. Aliquots of the reaction
mixture were applied to the origin of Silica Gel plates (Analtech,
Newark, USA) in the presence of PC (0.1 mg) and choline (0.1 mg) as
carriers. The chromatography was developed in methanol/0.6%
NaCl/28% NH4OH in water (50:50:5, v/v/v) as solvent. PC was
visualized under exposure to iodine vapor, and the corresponding
spot was scraped and transferred to scintillation vials for mea-
surement of radioactivity by a Beckman 6000-TA (Madrid, Spain)
liquid-scintillation counter. At least three experiments were per-
formed in all assays. The 50% inhibitory concentrations (IC50 values)
were determined from the % activity of the enzymes at different
concentrations of synthetic inhibitors by using a sigmoidal dose-
response curve (the ED50plus v1.0 software).
ing of tryptophan fluorescence. All experiments were carried out in
a Cary Eclipse spectrofluorometer (Varian) at 25 ꢀC with the en-
zymes at 1 mM, and concentrations of compounds varying from 0.1
to 10 mM for HsCKa1 and 0.5e200 mM for HsCKb in 25 mM Tris,
150 mM NaCl, pH 7.5. Fluorescence emission spectra were recorded
in the 300e400 nm range with an excitation wavelength of 280 nm,
with slit width of 5 nm. Controls were determined by incubating
the enzymes with equivalent amounts of DMSO. As indicated pre-
viously, data analysis was performed in Prism 6 (GraphPad soft-
ware) [32] considering a model with a single binding site (Equation
(1)), where F0 is the intrinsic fluorescence of the enzyme in the
absence of quencher (Q), F1 is the observed fluorescence at a given
quencher concentration, fa is the fractional degree of fluorescence,
and Kd is the dissociation constant.
5.3.4. Docking studies
F1
F0
½Qꢃ
Details about computational studies have been previously re-
ported [22].
1 ꢁ
¼ fa ꢂ
(1)
Kd þ ½Qꢃ
5.3.5. Choline uptake assay
Choline uptake was determined as previously reported [39].
Briefly, HepG2and HT-29 cells (200,000 cells/well) were incubated
for 10 min at 37 ꢀC either in a medium containing different con-
5.3.1. Protein crystallography
Human ChoK 1 at 20 mg/mL was preincubated at room tem-
a
perature with 25 mM hit b and 30 mM hit o in buffer 25 mM Tris/
HCl, 150 mM NaCl pH 7.5 (DMSO is at 5% final concentration in the
mix). The sitting-drop vapor diffusion [33] method was used to
produce crystals by mixing 0.5 mL of the protein solution and an
centrations of CK
a1 inhibitors or with no supplement as controls.
The medium was then removed and the cells immediately exposed
to a medium containing [methyl-14C]choline (16
mM, 31 Ci/mol) for
5 min at 37 ꢀC. The incorporation of choline was stopped by me-
dium aspiration followed by two washes with ice-cold PBS con-
equal volume of mother liquor. Crystals of the human ChoKa1-
compound b complex appeared in 18% polyethylene glycol (PEG)
3350 and 0.1 M potassium chloride, whereas crystals of the human
ChoKa1-compound o complex appeared in 20% PEG 3350 and 0.2 M
taining 580 mM choline. The cells were solubilized in NaOH 0.1 N
and an aliquot used to determine the total amount of radiolabel
taken up by the cells.
14

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
specific and selective cytotoxic antitumoral strategy as a potential mechanism
of action, Oncogene 23 (2004) 8247e8259.
[12] J. Xiong, J. Bian, L. Wang, J.Y. Zhou, Y. Wang, Y. Zhao, L. L Wu, J.J. Hu, B. Li, S. J
Chen, C. Yan, W. L Zhao, Dysregulated choline metabolism in T-cell lym-
5.3.6. Statistical analysis
Graphs and statistical analyses were performed using GraphPad
Prism software (GraphPad, La Jolla, CA, USA). All data in graphs
represented the mean of at least three independent
experiments SEM. Statistical significance was determined using
Student’s t-test or ANOVA (one- or two-way) depending on the
type of data. Asterisks indicate a significant difference between the
treated and the control group, unless otherwise specified. *p < 0.05,
**p < 0.01, ***p < 0.001, ****p < 0.0001.
phoma: role of choline kinase-
(2015) 287.
[13] S. Schiaffino-Ortega, E. Baglioni, E. Mariotto, R. Bortolozzi, L. Serran-Aguilera,
a
and therapeutic targeting, Blood Canc. J. 5
ꢀ
P. Ríos-Marco, M.P. Carrasco-Jimenez, M.A. Gallo, R. Hurtado-Guerrero,
ꢀ
C. Marco, G. Basso G. Viola A. Entrena, L.C. Lopez-Cara, Design, synthesis,
crystallization and biological evaluation of new symmetrical biscationic
compounds as selective inhibitors of human Choline Kinase a1 (ChoKa1), Sci.
Rep. 6 (2016) 23793.
ꢀ
[14] A. Sola-Leyva, L.C. Lopez-Cara, P. Ríos-Marco, A. Ríos, C. Marco, M.P. Carrasco-
ꢀ
Jimenez, Choline kinase inhibitors EB-3D and EB-3P interferes with lipid ho-
Declaration of competing interest
meostasis in HepG2 cells, Sci. Rep. 9 (2019) 5109.
[15] E. Mariotto, R. Bortolozzi, I. Volpin, D. Carta, V. Serafin, B. Accordi, Navarro
P. Luque, L.C. Lopez Cara, G. Basso, G. Viola, EB-3D a novel choline kinase
inhibitor induces deregulation of the AMPK-mTOR pathway and apoptosis in
leukemia T-cells, Biochem. Pharmacol. 155 (2018) 213e223.
[16] E. Mariotto, G. Viola, R. Ronca, Z.M. Bhujwalla, B. Accordi, V. Serafin, L. Persano,
L.C. Lopez Cara, R. Bortolozzi, The novel choline kinase alpha inhibitor EB-3D
induces cellular senescence reduce tumor growth and metatastic dissemina-
tion in breast cancer Cancers 10 (2018) 391.
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Acknowledgments
[17] M. Sahún-Roncero, B. Rubio-Ruiz, G. Saladino, A. Conejo-García, A. Espinosa,
ꢀ
ꢀ
A. Velazquez-Campoy, F.L. Gervasio, A. Entrena, R. Hurtado-Guerrero, The
The authors are grateful to the Centro de Servicios de Informa-
mechanism of allosteric coupling in choline kinase a1 revealed by the action
of a rationally designed inhibitor, Angew. Chem., Int. Ed. Engl. 52 (2013)
4582e4586.
ꢀ
tica and the Centro de Instrumentacion Científica of the University
of Granada (Spain) for the use of their computers and scientific
software. We acknowledge to “PID2019-109294RB-I00. Con-
vocatoria 2019 Proyectos de I þ D þ i - RTI Tipo B. Ministerio de
[18] M. Sahún -Roncero, B. Rubio-Ruíz, A. Conejo-García, A. Velazquez-Campoy,
A. Entrena, R. Hurtado-Guerrero, Determination of potential scaffolds for
human choline kinase a1 by chemical deconvolution studies, Chembiochem
14 (2013) 1291e1295.
ꢀ
Innovacion y Ciencia”for its financial support. R.H.-G. acknowledges
ALBA Light Source (Barcelona, Spain) synchrotron beamline XALOC
and the Diamond Light Source (Oxford, UK) synchrotron beamline
I04-1 (experiment number MX8035-11). He also acknowledges
ARAID, MEC (CTQ2013-44367-C2-2-P, BFU2016-75633-P and
[19] B. Rubio-Ruiz, A. Figuerola-Conchas, J. Ramos-Torrecillas, F. Capitan-Canadas,
P. Ríos-Marco, M.P. Carrasco, M.A. Gallo, A. Espinosa, C. Marco, C. Ruiz,
A. Entrena, R. Hurtado-Guerrero, A. Conejo-García, Discovery of a new binding
site on human choline kinase a1: design, synthesis, crystallographic studies,
and biological evaluation of asymmetrical bispyridinium derivatives, J. Med.
Chem. 57 (2014) 507e515.
ꢀ
PID2019-105451 GB-I00) and Gobierno de Aragon (E34_R17 and
[20] F. Fermin Castro-Navas, S. Schiaffino-Ortega, M.P. Carrasco-Jimenez, P. Ríos-
Marco, C. Marco, A. Espinosa, M.A. Gallo, G. Basso, R. Bortolozzi, G. Viola,
LMP58_18) with FEDER (2014e2020) funds for ‘Building Europe
ꢀ
from Aragon’ for financial support. The research leading to these
ꢀ
A. Entrena, L.C. Lopez-Cara, New more polar symmetrical bipyridinic com-
pounds: new strategy for the inhibition of choline kinase a1, Future Med. Chem
. 4 (2015) 417e436.
results has also received funding from FP7 442 (2007e2013) under
BioStruct-X (grant agreement no. 283570 and BIOSTRUCTX_5186).
[21] G. Rubbini, Ana B. Buades-Martín, María Kimatrai-Salvador, Antonio Entrena,
ꢀ
M.A. Gallo-Mezo, P. Ríos-Marco, C. Marco, E. Mattiuzzo, R. Bortolozzi,
ꢀ
E. Mariotto, F.A. Greco, A. Macchiarulo, M.P. Carrasco-Jimenez, G. Viola,
Appendix A. Supplementary data
ꢀ
L.C. Lopez-Cara, Lead optimization-hit expansion of new asymmetrical pyr-
idinium/quinolinium compounds as ChoK
a
1 inhibitors, Future Med. Chem. 10
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2020.112797.
(2018) 1769e1786.
ꢀ
ꢀ
ꢀ
[22] L. Serran-Aguilera, R. Nuti, L.C. Lopez-Cara, M.A. Gallo Mezo, A. Macchiarulo,
A. Entrena, R. Hurtado-Guerrero, Pharmacophore-based virtual screening to
discover new active compounds for human choline kinase
a1, Molecular
References
Informatics 34 (2015) 458e466.
ꢀ
[23] L. Serran-Aguilera, R. Nuti, L.C. Lopez-Cara, P. Ríos-Marco, M. P Carrasco,
C. Marco, A. Entrena, A. Macchiarulo, R. Hurtado-Guerrero, Choline kinase
active site provides features for designing versatile inhibitors, Curr. Top. Med.
Chem. 14 (2014) 2684e2693.
[1] R. Straussman, T. Morikawa, K. Shee, M. Barzily-Rokni, Z.R. Qian, J. Du,
A. Davis, M.M. Mongare, J. Gould, D.T. Frederick, Z.A. Cooper, P.B. Chapman,
D.B. Solit, A. Ribas, R.S. Lo, K.T. Flaherty, S. Ogino, J.A. Wargo, T.R. Golub,
Tumour micro-environment elicits innate resistance to RAF inhibitors through
HGF secretion, Nature 487 (2012) 500e504.
[2] B. Kalyanaraman, G. Cheng, M. Hardy, O. Ouari, M. Lopez, J. Joseph, J. Zielonka,
M.B. Dwinell, A review of the basics of mitochondrial bioenergetics, meta-
bolism, and related signaling pathways in cancer cells: therapeutic targeting
of tumor mitochondria with lipophilic cationic compounds, Redox Biol 14
(2018) 318e327.
[24] S.G. Zech, A. Kohlmann, T. Zhou, F. Li, R.M. Squillace, L.E. Parillon,
M.T. Greenfield, D.P. Miller, J. QiR, M. Thomas, Y. Wang, Y. Xu, J.J. Miret,
W.C. Shakespeare, X. Zhu, D.C. Dalgarno, Novel small molecule inhibitors of
choline kinase identified by fragment-based drug discovery, J. Med. Chem. 59
(2016) 671e686.
ꢀ
[25] S. Schiaffino-Ortega, L.C. Lopez-Cara, P. Ríos-Marco, M.P. Carrasco-Jimenez,
M. A Gallo, A. Espinosa, Carmen Marco, Antonio Entrena, Bioorg. Med. Chem.
21 (2013) 7146e7154.
[3] G. Luo, N. Liu, Int. J. Mol. Med. 28 (2018) 1e10.
[26] J. Gruber, W.C. See Too, M.T. Wong, A. Lavie, T. McSorley, M. Konrad, Balance
of human choline kinase isoforms is critical for cell cycle regulation: impli-
cations for the development of choline kinase-targeted cancer therapy, FEBS J.
279 (2012) 1915e1928.
[27] C.S. Hudson, R.M. Knegtel, K. Brown, P.A. Charlton, J.R. Pollard, Kinetic and
mechanistic characterisation of choline kinase-alpha, Biochim. Biophys. Acta
1834 (2013) 1107e1116.
[4] K. Glunde, M.F. Penet, L. Jiang, M.A. Jacobs, Z.M. Bhujwalla, Choline
metabolism-based molecular diagnosis of cancer: an update, Expert Rev. Mol.
Diagn 15 (2015) 735e747.
[5] T. Shah, B. Krishnamachary, F. Wildes, J.P. Wijnen, K. Glunde, Z.M. Bhujwalla,
Molecular causes of elevated phosphoethanolamine in breast and pancreatic
cancer cells, NMR Biomed. 31 (2018), e3936.
[6] F. Gibellini, T.K. Smith, The Kennedy Pathway–De Novo Synthesis of Phos-
phatidylethanolamine and Phosphatidylcholine IUBMB Life, vol. 62, 2010,
pp. 414e428.
[28] V. Michel, Z. Yuan, S. Ramsubir, M. Bakovic, Choline transport for phospholipid
synthesis, Exp. Biol. Med. 231 (2006) 490e504.
[29] T. Hara, A. Bansal, T.R. DeGrado, Choline transporter as a novel target for
molecular imaging of cancer, Mol. Imag. 5 (2006) 498e509.
[30] M. Inazua, T. Yamada, N. Kubota, T. Yamanaka, Functional expression of
choline transporter-like protein 1 (CTL1) in small cell lung carcinoma cells: a
target molecule for lung cancer therapy, Pharmacol. Res. 76 (2013) 119e131.
[31] F.J. Bock Tait SWG, Mitochondria as multifaceted regulators of cell death, Nat.
Rev. Mol. Cell Biol. 21 (2020) 85e100.
[7] M. Agassandian, R.K. Mallampalli, Surfactant phospholipid metabolism, Bio-
chim. Biophys. Acta 1831 (2013) 612e625.
[8] M. Cheng, Z.M. Bhujwalla, K. Glunde, Targeting phospholipid metabolism in
cancer, Front Oncol 6 (2016) 266e282.
[9] S.P. Arlauckas, A.V. Popov, E.J. Delikatny, Choline kinase alpha-Putting the
ChoK-hold on tumor metabolism, Prog. Lipid Res. 63 (2016) 28e40.
[10] R.B. Cornell, N.D. Ridgway, CTP:phosphocholine cytidylyltransferase: function,
regulation, and structure of an amphitropic enzyme required for membrane
biogenesis, Prog. Lipid Res. 59 (2015) 147e171.
[32] www.graphpad.com.
[33] H. Adachi, K. Takano, M. Morikawa, S. Kanaya, M. Yoshimura, Y. Mori,
T. Sasaki, Acta Crystallogr. D Biol. Crystallogr. 59 (2003) 194e196.
[34] W. Kabsch, Acta Crystallogr. D Biol. Crystallogr. 66 (2010) 125e132.
ꢀ
ꢀ
[11] A. Rodríguez-Gonzalez, A. Ramírez de Molina, F. Fernandez, J.C. Lacal, Choline
kinase inhibition induces the increase in ceramides resulting in a highly
15

ꢀ
L. Serran- Aguilera, E. Mariotto, G. Rubbini et al.
European Journal of Medicinal Chemistry 207 (2020) 112797
[35] A.J. McCoy, R.W. Grosse-Kunstleve, P.D. Adams, M.D. Winn, L.C. Storoni,
[38] a) R.A. Laskowski, J.A. Rullmannn, M.W. MacArthur, R. Kaptein, J.M. Thornton,
J. Biomol. NMR 8 (1996) 477e486;
R.J. Read, J. Appl. Crystallogr. 40 (2007) 658e674.
[36] P. Emsley, K. Cowtan, Acta Crystallogr.
D
Biol. Crystallogr. 60 (2004)
b) , in: http://www.ebi.ac.uk/thornton-srv/software/PROCHECK/index. html.
ꢀ
ꢀ
2126e2132.
[39] J.M. Jimenez-Lopez, M.P. Carrasco, J.L. Segovia, C. Marco, Hex-
adecylphosphocholine inhibits phosphatidylcholine biosynthesis and the
proliferation of HepG2 cells, Eur. J. Biochem. 269 (2002) 4649e4655.
[37] A.A. Vagin, R.A. Steiner, A.A. Lebedev, L. Potterton, S. McNicholas, F. Long,
G.N. Murshudov, Acta Crystallogr. D Biol. Crystallogr. 60 (2004) 2184e2195.
16