Naphthalene diimide-polyamine hybrids as antiproliferative agents: Focus on the architecture of the polyamine chains

Article abstract of DOI:10.1016/j.ejmech.2017.01.025

Naphthalene diimides (NDIs) have been widely used as scaffold to design DNA-directed agents able to target peculiar DNA secondary arrangements endowed with relevant biochemical roles. Recently, we have reported disubstituted linear- and macrocyclic-NDIs that bind telomeric and non-telomeric G-quadruplex with high degree of affinity and selectivity. Herein, the synthesis, biological evaluation and molecular modelling studies of a series of asymmetrically substituted NDIs are reported. Among these, compound 9 emerges as the most interesting of the series being able to bind telomeric G-quadruplex (ΔTm = 29 °C at 2.5 μM), to inhibit the activity of DNA processing enzymes, such as topoisomerase II and TAQ-polymerase, and to exert antiproliferative effects in the NCI panel of cancer cell lines with GI₅₀values in the micro-to nanomolar concentration range (i.e. SR cell line, GI₅₀= 76 nM). Molecular mechanisms of cell death have been investigated and molecular modelling studies have been performed in order to shed light on the antiproliferative and DNA-recognition processes.

Full text of DOI:10.1016/j.ejmech.2017.01.025

European Journal of Medicinal Chemistry 128 (2017) 107e122
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Naphthalene diimide-polyamine hybrids as antiproliferative agents:
Focus on the architecture of the polyamine chains
Andrea Milelli a *, Chiara Marchetti b, Maria Laura Greco c, Federica Moraca d,
,
d
a
a
a
a
ꢀ
Giosue Costa , Eleonora Turrini , Elena Catanzaro , Nibal Betari , Cinzia Calcabrini ,
Claudia Sissi c, Stefano Alcaro d, Carmela Fimognari a, Vincenzo Tumiatti a,
Anna Minarini b
a Department for Life Quality Studies, Alma Mater Studiorum-University of Bologna, 47921 Rimini, Italy
b Department of Pharmacy and Biotechnology, Alma Mater Studiorum-University of Bologna, 40126 Bologna, Italy
c Department of Pharmaceutical and Pharmacological Sciences, University of Padova, 35131 Padova, Italy
d
ꢀ
Dipartimento di Scienze della Salute, Universita 'Magna Græcia' di Catanzaro, 88100 Catanzaro, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Naphthalene diimides (NDIs) have been widely used as scaffold to design DNA-directed agents able to
target peculiar DNA secondary arrangements endowed with relevant biochemical roles. Recently, we
have reported disubstituted linear- and macrocyclic-NDIs that bind telomeric and non-telomeric G-
quadruplex with high degree of affinity and selectivity. Herein, the synthesis, biological evaluation and
molecular modelling studies of a series of asymmetrically substituted NDIs are reported. Among these,
compound 9 emerges as the most interesting of the series being able to bind telomeric G-quadruplex
Received 15 December 2016
Received in revised form
18 January 2017
Accepted 19 January 2017
Available online 22 January 2017
(
D
Tm ¼ 29 ꢀC at 2.5
mM), to inhibit the activity of DNA processing enzymes, such as topoisomerase II and
Keywords:
Naphthalene diimides
Drug design
TAQ-polymerase, and to exert antiproliferative effects in the NCI panel of cancer cell lines with GI50
values in the micro-to nanomolar concentration range (i.e. SR cell line, GI50 ¼ 76 nM). Molecular
mechanisms of cell death have been investigated and molecular modelling studies have been performed
in order to shed light on the antiproliferative and DNA-recognition processes.
DNA
G-quadruplex
Telomeric DNA
Topoisomerase
Antiproliferative
© 2017 Elsevier Masson SAS. All rights reserved.
1. Introduction
antiproliferative activities [4,5] through different mechanisms of
action among which intercalation within double-stranded DNA
Naphthalene diimides (NDIs) represent a class of electron-
deficient aromatic compounds widely used in the field of supra-
molecular and material chemistry [1]. In the last years, this scaffold
has been exploited in the design of biologically active compounds.
In particular, published reports highlighted the use of NDI as optical
imaging probe [2] and, in the medicinal chemistry field, as potential
drug candidates. Indeed, several NDIs are able to inhibit enzymes,
such as acetylcholinesterase [3], or to exert remarkable
(dsDNA) [6] and stabilization of higher order DNA secondary
structures are worth of mention [7]. Over the years, dsDNA has
represented the prominent target in the design of new anticancer
agents; however, dsDNA-directed molecules suffer from different
drawbacks, such as low selectivity towards non-cancerous cells [8].
Recently, other secondary DNA arrangements endowed with rele-
vant biological roles have been discovered, notably the G-quad-
ruplex structures [7]. Guanine-rich DNA sequences, located in
distinct regions of the genome, such as oncogene promoters and
telomeric regions, can assemble into G-quadruplex. Molecules able
to stabilize these secondary structures, located at telomeric ends,
prevent the binding of telomerase enzyme leading to senescence
and cell death, while stabilization of G-quadruplex located in the
promoter region of oncogenes results in a down-regulation of the
expression of these genes [9,10]. Further, it is important to point out
that G-quadruplex are significantly more prevalent in tumor tissues
Abbreviations: BASASA, Boltzmann Accessible Solvent Accessible Surface Area;
BCL-2, B-cell lymphoma 2; dsDNA, double-stranded DNA; hTel, human telomeric;
HSP90, heat shock protein 90; MD, Molecular Dynamics; MC, Monte Carlo method;
NDI, Naphthalene Diimide; PARP, poly(ADP)ribose polymerase; SASA, Solvent
Accessible Surface Area.
* Corresponding author.
E-mail address: andrea.milelli3@unibo.it (A. Milelli).
http://dx.doi.org/10.1016/j.ejmech.2017.01.025
0223-5234/© 2017 Elsevier Masson SAS. All rights reserved.

108
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
compared to non-transformed cells [11].
From biophysical analysis and molecular modelling studies
performed on these disubstituted and macrocyclic NDIs emerged
that G-quadruplex binding property strongly depend on a) the
length of the polymethylene linker separating the nitrogen atom(s)
of the lateral chains mounted on the NDI nucleus, b) the number of
the nitrogen atoms of the lateral chains and c) the nature of the
substituent on aromatic rings of the lateral chains. Furthermore,
molecular modelling studies point out that the NDI-DNA recogni-
tion process is driven by the establishment of both electrostatic and
hydrophobic interactions involving both NDI core and peripheral
substituted aromatic ring.
Based on these premises, we attempted to increase the cytotoxic
activity of NDIs towards cancer cells by merging the best features of
the compounds obtained so far focusing on G-quadruplex. To this
aim, we designed asymmetric substituted NDIs where one arm was
constituted by a substituted-benzylpropylendiamine and the other
harm was constituted by a polyaminechain. Indeed, it is well known
that polyamines are able to establish strong interactions with bio-
logical counterparts due to their positively charged nitrogen atoms.
The interaction of polyamines (i. e. spermine) on G-quadruplex has
been recently confirmed, evidencing different effects on the tetra-
helical stability, according to their concentration [21,22]. In addi-
tion, it is well established that the polyamine's uptake by the
polyamine transport system, which is overactive in cancer cells,
contributes to the increased antiproliferative activity of some
anticancer agents conjugate with polyamines [23].
A large number of molecules able to stabilize both telomeric and
promoter G-quadruplex have been reported in the last years and a
large number of them shares similar structural features, that is a
large planar (hetero)aromatic core decorated with two or more
harms carrying positively charged nitrogen(s) extremely favorable
in the quadruplex recognition process [12]. NDIs have been
extensively explored as scaffold for the design of G-quadruplex
binders [13e16]. Indeed, it can be easily subjected to different types
of core functionalization, thus allowing diverse structural modifi-
cations leading to di-, tri- and tetra-substituted derivatives [1,17]. In
this context, a number of tetra-substituted NDIs have been devel-
oped by Neidle's group: in particular, compound 1 (Fig. 1) stabilizes
several distinct DNA quadruplex sequences, such as hTel, HSP90 A
and B, BCL-2 and k-RAS [18]. Moreover, 1 has significant in vivo anti-
tumor activity in a MIA PaCa-2 pancreatic cancer xenograft model.
Compound 2, developed by Richter and co-workers, represents an
example of tri-substituted NDI that displays stabilization of hTel
and antiproliferative effects in colon (HT29), lung (A549) and
melanoma (SKMel-5) tumor cells [19].
As part of our drug discovery program, we have recently re-
ported some symmetrically di-substituted NDIs endowed with
remarkable antiproliferative activities. Compound 3 (Fig.1), bearing
2-methoxybenzylpropylendiamine side chains, strongly binds DNA
and induces apoptotic cell death. Compound 3 was subject to an
optimization campaign aimed at improving its pharmacological
profile [16,20]. Among the new derivatives, compounds 4 and 5
(Fig. 1) emerged as potent G-quadruplex binders. Compound 4,
bearing 2,3,4-trimethoxybenzylpropylendiamine chains, is cyto-
toxic in a large panel of cancer cell lines thank to a simultaneous
interaction with different biological targets among which the
interaction with telomeric G-quadruplex is worth of mention [20].
Macrocyclic analogue 5 strongly stabilizes hTel and c-Kit quad-
ruplex sequences with a good level of selectivity for G-quadruplex
over dsDNA [16]. Unfortunately, in spite of these encouraging
In order to identify the most favorable substituent on the aro-
matic ring of the side chain, in terms of G-quadruplex recognition,
2-methoxybenzylpropylendiamine-NDI-spermine (6) and 2,3,4-
trimethoxybenzylpropylendiamine-NDI-spermine
(7)
were
designed as prototypes (Fig. 2). After a preliminary evaluation, the
most efficient hit for the stabilization of the G-quadruplex form of
hTel (compound 6, see later), was modified in the polyamine chain
by varying the length of the chain, the number of nitrogen atoms
and the distance between them. To this aim, both natural and non-
natural polyamines were employed.
properties, compound
5 has weaker antiproliferative effects
compared to compounds 3 and 4.
In particular, we synthesized derivatives: i) 8e10 characterized
Fig. 1. Selected NDI-based G-quadruplex binders reported in literature.

A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
109
Fig. 2. Drug design leading to compounds 6e17.
by a spermidine-like chain (8, 3-3; 9, 3e4; 10, 4e3); ii) 11 and 12
characterized by a spermine-like chain (11, 3-2-3; 12, 3-3-3); iii), 13
and 14 are the N-methyl derivatives of 8 and 6, respectively, iv) 15
deriving from the substitution with oxygens of the inner nitrogen
atoms of the spermine portions of 6; v) 16 bearing only one amine
function and 17 carrying an uncharged ethyl group in the lateral
chain.
synthesized as reported in Scheme 2. Protection of the inner ni-
trogen atom(s) of spermidine and spermine following standard
procedures has led to monoBoc-triamine 33 and diBoc-tetraamine
34. Then, 33 and 34 have been selectively benzylated at only one of
the primary amine function through reaction with benzaldehyde
and in situ reduction with NaBH4 of the formed Schiff base gave 35
and 36, respectively. The free primary amino groups of 35 and 36
have been protected as trifuoroacetamide to give 37 and 38. The
secondary amino groups have been subjected to mono-alkylation
with methyl iodide to give 39 and 40 that, after deprotection of
the primary amino group gave 41 and 42, respectively. Subsequent
debenzylation of those intermediates with H2/Pd furnished 43 and
44. Compounds 49e59 and 17 have been obtained refluxing a
stoichiometric mixture of naphtalentetracarboxylic dianhydride, 2-
2. Results and discussion
2.1. Chemistry
Synthesis of the intermediates and target compounds have been
performed following the procedure depicted in Schemes 1e3. Boc-
protected (poly) amines 18e22 have been obtained through the
standard procedure reported in literature [24]. Boc-protected
polyamines 31 and 32 have been synthesized as reported in
Scheme 1. 3-amino-1-propanol or 4-amino-1-butanol have been
protected as trifluoroacetamide (23 and 24) and then activated at
the hydroxyl group as tosylate leading to 25 and 26. Reaction of the
tosylates with monoBoc-diaminopropane 27 or monoBoc-
diaminobutane 28 has led to the protected triamine 29 and 30,
respectively. In situ Boc-protection of the secondary nitrogen atom
of 29 and 30 and then trifluoroacetamide deprotection provided
diBoc-protected triamines 31 and 32, respectively. N-methyl mono-
Boc triamine 43 and N-methyl di-Boc tetraamine 44 have been
methoxybenzylpropylendiamine
45
or
2,3,4-
trimethoxybenzylpropylendiamine 46 and the Boc-protected
amines 18e22, 31, 32, 43, 44, 47 and amine 48 in DMF (Scheme
3). Acidic hydrolysis of intermediates 49e59 has finally led to the
desired products 6e16 as hydrochloride salts.
2.2. Biochemical and biophysical studies
Target compounds 6e17 have been firstly evaluated for their
ability to stabilize hTel G-quadruplex and dsDNA sequences.
Compound 3 has been used as reference compound. Fluorescence
melting experiments were performed in the 0e20 mM compound
Scheme 1. i) CF3COOEt, rt, 1 h, quantitative yields; ii) tosylchloride, Et3N, DMAP, CH2Cl2, rt, overnight, 55e60% yields; iii) THF, rt, overnight, 35e47% yields; iv) a. Boc2O, Methanol, rt,
16 h; b. NaOH 40%, Methanol, rt, 16 h, 64e87% yields.

110
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
Scheme 2. i) a. CF3COOEt, Methanol, ꢁ78 ꢀC, 16 h; b. Boc2O, DCM, rt, 12 h; c. NaOH 40%, Methanol, rt, 12 h, 45e56% yields; ii) a. benzaldehyde, Methanol, reflux, 5 h, b. NaBH4, EtOH,
rt, 16 h, 80e91% yield; iii) CF3COOEt, Et3N, rt, 16 h, 72e92% yield; iv) CH3I, Et3N, THF, rt, 24 h, 30e66% yield; v) NaOH 40%, Methanol, rt, 16 h, 88e92% yields; vi) H2/Pd, Methanol, rt,
5 h, 68e76% yields.
Scheme 3. i) DMF, reflux, 5 h, 15e30% yields; ii) HCl 6 N, rt, overnight, quantitative yields.
concentration range: the increment of the melting temperatures of
the nucleic acid induced by all tested ligands at 2.5 M is reported
Closely related to compound 6, compounds 11 and 12 are
characterized by shorter spermine-like side chains differing in the
number of methylene groups separating the inner nitrogen atoms.
In this series, the decrease of the methylene groups from four to
three, such as in compound 12, induces a slight decrement in the
m
in Table 1. As already introduced, compounds 6 and 7, deriving from
compounds 3 and 4, respectively, were initially synthesized and
analysed to establish the more suitable substituent on the aromatic
ring for the G-quadruplex recognition: they share the same sper-
mine chain on one side of the NDI scaffold and differ for the number
of the methoxy groups on the terminal aromatic ring. From Table 1,
compound 6 emerges as a better G-quadruplex binder than com-
stabilization of telomeric G-quadruplex (
D
Tm ¼ 16.3 ꢀC) while the
lower homologue 11, characterized by a 3-2-3 chain, has a stabili-
zation capacity comparable to compound 6 (
D
Tm ¼ 22.0 ꢀC).
Among derivatives 8e10 that contain shorter spermidine-like
chain, compound 9 (3e4) emerges as the most effective G-quad-
pound 7 (6,
D
Tm ¼ 21.4 ꢀC vs 7,
D
Tm ¼ 13.1 ꢀC at 2.5
mM) high-
lighting that, in these NDI-spermine derivatives, the 2,3,4-
mehoxybenzyl group has an unfavourable effect on the G-quad-
ruplex stabilization. Further, compound 7 is a better dsDNA binder
ruplex binder with a D mM, significantly
Tm value of 29 ꢀC at 2.5
higher than the one displayed by compound 6. This result was
unexpected since we have previously reported that the elimination
of amine functions is usually accompanied with a concomitant
decrease of the quadruplex stabilizing capacity [16]. The related
derivative 8 preserves good DNA binder activity whereas com-
pound 10 shows a significantly impaired ability to stabilize the
telomeric G-quadruplex being the weaker G-quadruplex binder
(6,
D
Tm ¼ 6.7 ꢀC vs 7,
D
Tm ¼ 12.6 ꢀC). These results are surprising in
light of the previously reported data on symmetric derivatives,
where compound 4 proved to be a better G-quadruplex binder than
compound 3, suggesting a possible different binding mode towards
target DNA for the new polyamine conjugates 6 and 7 [20].

A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
111
Table 1
Effects of compounds 3, 6e17 on the thermal stability of hTel G-quadruplex and dsDNA and on the catalytic activity of topoisomerase II
a
and TAQ polymerase.
Compound
D
Tm (ꢀC)a
DDTmb
Topoisomerase II
IC50
M)c
a
TAQ Polymerase
(
m
MIC (
M)d
m
hTel
dsDNA
G-quadruplex
3
6
7
8
20.5
21.4
13.1
13.3
29
22.2
22
16.3
18
8.4
6.7
12.6
7
12.2
9.1
10.4
12.2
9.5
17.1
1.5
12.1
14.7
0.5
6.0 0.9
2.2 0.4
n.d.e
10
2
0.25 0.06
n.d.e
6.3
1.5 0.2
8.1 1.1
7.0 0.5
n.d.e
0.25 0.06
0.5 0.1
0.5 0.1
n.d.e
9
16.8
13.1
11.6
4.1
8.5
9.0
11.4
ꢁ5.3
1.5
10
11
12
13
14
15
16
17
n.d.e
n.d.e
n.d.e
n.d.e
26.1
12.9
1.4
n.d.e
n.d.e
5.5 0.5
1.0 0.2
6.7
0.9
n.d.e
n.d.e
2.4
n.d.e
n.d.e
a
Increment of the melting temperature induced by 2.5 m
M ligand concentration. Errors were 0.4 ꢀC.
b
c
DDTm was derived as the difference of the thermal stabilization of the two DNA substrates.
IC50 values are defined as the drug concentration which reduces by 50% the activity.
MIC as the minimal drug concentration at which the enzymatic activity is suppressed.
n.d. ¼ not determined.
d
e
within the spermidine-like series with a
2.5 M.
Compounds 13 (3-3-CH3) and 14 (3-4-3-CH3) were synthesized
in order to evaluate the effect of the alkylation of the primary amine
function of polyamine chain of compounds 8 and 6, respectively.
Both compounds display improved G-quadruplex binding profiles
compared to the parent compounds (13 vs 8 and 14 vs 6); unfor-
tunately, for compound 14 a larger increase of the affinity towards
D
Tm value of 13.1 ꢀC at
topoisomerase II
interference of our ligands on topoisomerase II
resulted that the tested derivatives inhibit the enzyme with com-
parable efficiency to the lead compound 3 with the notable
exception of 8 and 6 for which significantly reduced IC50 values
a
and TAQ-polymerase. We first assessed the
activity (Fig. 3A). It
m
a
were detected (IC50 of 1.5 and 2.2 mM, respectively) (Table 1). It is
well known that compounds active against topoisomerases may
work as protein inhibitors or poisons (stabilizers of protein-DNA
cleavage complex). By a cleavage assay, we confirmed that none
of the tested compounds was able to stabilize the cleavage complex
topoisomerase-DNA (Fig. 3B), thus ruling out poisoning as mecha-
nism of action.
To further validate this model, we explored the effect of the
tested derivatives on the activity of an additional DNA processing
enzyme, TAQ-polymerase (Fig. 3C, Table 1). Consistently, de-
rivatives 8 and 6 resulted the most active also against TAQ-
polymerase, with a complete suppression of enzyme activity at
dsDNA is observed (dsDNA: 6,
D
Tm ¼ 6.7 ꢀC at 2.5
mM; 14,
D
Tm ¼ 17.1 ꢀC at 2.5
mM) thus leading to a reduced selectivity.
Replacement of the inner nitrogen atoms of 6 with oxygens (as in
compound 15) induces a marked reduction of the G-quadruplex
DNA-binding activity (
D
Tm ¼ 12.9 ꢀC at 2.5
m
M) underlying that the
establishment of interaction between the secondary amine func-
tion of the polyamine side chain and DNA is involved in stabilizing
the G-quadruplex arrangements. This was further confirmed by the
worst G-quadruplex affinities of compound 16, carrying only one
amino group, and compound 17 where the polyamine has been
replaced by an alkyl chain.
0.25 mM. Interestingly, the ranking order obtained for topoisom-
erase II inhibition was essentially conserved among asymmetric
To better highlight the potential selectivity towards G-quad-
ruplex structure, we compared the thermal stabilization on the two
DNA templates in terms of DDTm (DTmG-quadruplex- DTmdsDNA)
derivatives but with an efficiency significantly higher with refer-
ence to the symmetrically substituted control 3 (10
This suggested that within asymmetric derivatives, the inhibi-
tion of the two studied protein targets, topoisomerase II and TAQ-
2 mM).
(Table 1). From this analysis, compounds 9 and 6 emerged as the
most selective ones. This suggested that they conserved spacer's
length in the polyamine chain that is beneficial for the recognition
of the telomeric G-quadruplex. At the same time, it indicated that
the presence of an additional basic functionality, such as in com-
pound 6, did not increase the interaction with the nucleic acid
neither in terms of efficiency and selectivity (see 6 vs 9 and 12 vs 8).
Compound 15 showed a good preference towards G-quadruplex
but, as above discussed, this is counterbalanced by a reduced af-
finity on both templates as a result of the substitution of two ni-
trogen atoms with oxygen atoms in the lateral chain.
These data indicated that some of the new derivatives were
more promising G-quadruplex binders in comparison to the lead
compound 3. However, in no instance we succeed to realize a total
suppression of the dsDNA recognition thus letting open the main-
tenance of a multiple target mechanism of actions. Based on these
observations, we selected the most promising G-quadruplex
binders 6 and 9, together with their homologues 8 and 10, and 15 to
explore the functional role of the basic amine functions in the side
chain on the activity of DNA processing enzymes, such as
a
polymerase, is probably related to DNA/ligand interaction rather
than to a ligand/protein interaction. In this model, the more pro-
nounced efficiency of derivatives 6 and 8 might underline a mod-
ulation of the binding mode. This point was supported by agarose
gel electrophoresis of supercoiled plasmid treated with 8e10
(Fig. 3D). The bell-shaped reduction of the DNA electrophoretic
mobility induced by tested derivatives indicates that 9 intercalates
within the dsDNA starting from higher concentration in compari-
son to 8, which, on turn, causes a significant precipitation of the
nucleic acid at the highest tested concentrations. Conversely, 10
binds plasmid DNA as evidenced by the smear of the plasmid
electrophoretic band, but without evidence of intercalation.
Thermal analysis underlined that selected compounds bind both
the ds and the telomeric G-quadruplex DNAs. Finally, we investi-
gated if the detected stabilization of the G-quadruplex form of
telomeres exerted by the selected NDI derivatives can interfere
with telomerase activity. However, working at ligands concentra-
tions that are no affecting TAQ-polymerase activity, no specific in-
hibition of telomerase by the selected compounds was observed,

112
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
Fig. 3. a) Inhibition of the relaxation of supercoiled pBR322 promoted by 1 U of II
Cleavage assay performed on pBR322 with 5 U of topoisomerase II (lane C) in the presence of Etoposide (lane et) or compound 9 (lane 9). Lanes N and L refer to the nicked and
linearized pBR322, respectively; c) Inhibition of the amplification of a fragment of pBR322 mediated by TAQ polymerase chain reactions performed in the presence of increasing
concentrations of compounds 8e10 (0.25e0.5-1.0-2.5-5.0-10-20-40 M); d) Modification of the electrophoretic mobility of supercoiled pBR322 upon addition of compounds 8e10
(0e100 M) in 10 mM Tris, 50 mM KCl, in the absence of any enzyme.
a due to addition of increasing concentrations compounds 6, 8e10, 15 in the reaction mixture; b)
a
m
m
thus ruling out this mechanism as the preferential one. This is not
surprising since we and others have already reported that NDIs
highly effective in stabilizing telomeric G-quadruplex are not able
to inhibit telomerase in vitro or in cells [16,25].
2.3.1. Binding mode of compounds 9, 11 and 16 against G-
quadruplex/ds DNAs
To get further insight into the binding mode and clarify the role
of the spermine and spermidine-like side chains in the DNA
recognition of the NDI derivatives 9, 11 and 16, molecular docking
calculations have been performed with AutoDock4.2 [30] and a
ligand/G-quadruplex DNA stoichiometry ratio of 2:1 has been
considered to respect the experimental findings. All the 100 bind-
ing poses predicted by AutoDock4.2 were then submitted to a full
energy optimization step using the AMBER* force field as imple-
mented in MacroModel ver. 10.8 [31]. The same protocol has been
adopted also against the homology model of the experimental
sequence of the dsDNA, considering a 1:1 stoichiometry ratio.
2.3. Molecular modelling studies
Compounds 9, 11 and the worst binder 16 were submitted to a
conformational search with the Monte Carlo method (MC) coupled
to the analysis of the Solvent Accessible Surface Area (SASA)
computed onto a defined NDI planar aromatic system. Among all
the generated conformations, those within 3 kcal/mol above the
energy global minimum were selected and submitted to the
Boltzmann Accessible Solvent Accessible Surface Area (BASASA)
analysis [26] (Table S1) (see Materials and Methods), while the
conformer with the lowest potential energy (Table S2) has been
submitted to docking calculations. In addition, to take into account
the G-quadruplex DNA and dsDNA flexibility, a 10 ns long Molec-
ular Dynamics (MD) simulation has been carried out as described in
the Material and Methods section. The geometrical stability of the
G-quadruplex DNA hybrid-1 and hybrid-2 forms is demonstrated
by their RMSD average calculated for the all heavy atoms of the
nucleic acid (3.22 0.15 Å and 2.69 0.14 Å, respectively) (Fig. S2A)
2.3.1.1. Hybrid-1. A geometrical analysis of the binding poses be-
tween spermidine-like NDI 9 and spermine-like NDI 11 against the
hybrid-1 G-quadruplex DNA (Fig. 4A and B respectively) reveals a
groove recognition of compounds 9 and 11, probably also due to the
steric hindrance of the residues T18:A1:A19 at the 50 and
T23:T12:T11:A13 at the 30, which stacks at the center of the top and
bottom G-tetrad planes thus hampering the p-p stacking of the NDI
core. Specifically, in the 2:1 complex of compound 9 (Fig. 4A) the
spermidine-like side chain and the NDI core globally engage eight
hydrogen bonds (H-bonds) with the phosphate groups of
G2:G22:G4 and G20:G15:G14, while globally five electrostatic in-
teractions are established between two positively charged nitrogen
atoms and the phosphate backbone of G4:G2:G14:G15:G16 resi-
and the G-tetrads atoms only (1.09
0.08 Å and 1.07 0.09 Å,
respectively) (Fig. S2B). MD trajectories were then clustered using
the GROMOS algorithm [27] of the g-cluster tool implemented in
GROMACS ver. 4.5 [28]. Three main clusters were found both from
hybrid-1 (PDB code: 2JSM) [29] (Fig. S3A) and hybrid-2 (PDB code:
2JSL) [29] (Fig. S3B) MD trajectories, but only the one most different
has been taken into account for docking simulations (Table S3).
dues. The NDI is involved also in four T-shaped
p-p stacking with
G3:G22:G2:G21 and G14:G15:G16 residues. As concern the mo-
lecular recognition of compound 11 (Fig. 4B), the three methylene
linker of the spermine chain globally engages eight H-bonds with
the phosphate groups of G21:G22:G10:T11:G10 and six

A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
113
Fig. 4. 2:1 binding mode against the hybrid-1 form of G-quadruplex DNA receptor (PDB code: 2JSM) of the NDI polyamine derivatives (A) 9, (B) 11 and the worst binder (C) 16.
Nucleic acid is displayed as grey ribbon while interacting residues as stick coloured by atom type. Polyamine NDIs 9, 11 and 16 are shown as yellow, cyan and iceblue sticks
respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
electrostatic interactions with T11:T12:G10:G15:G21:G22. Only one
T-shaped stacking is established, instead, between the 2-
with residues G22:G4:G14:G15:G16 and the phosphate backbone
of G4:G8:G15:G16, respectively. A lower number of electrostatic
interactions are observed also for compound 11. In this case, in fact,
the spermine-like side chain of compound 11 is globally involved in
p-p
methoxybenzyl group and T23. The role of the spermidine-like
and spermine side chains in G-quadruplex DNA recognition is
better highlighted comparing the binding mode of compounds 9
and 11 with that of the worst binder 16, which bears only one
ionized nitrogen atom without any polyamine chain. The absence of
spermidine or spermine side chains reduces the number of H-
bonds rather than the electrostatic interactions. Compound 16
engages, in fact, five H-bonds and five electrostatic interactions
with the terminal positively charged nitrogen atom of the three
methylene side chain and the phosphate groups of
G4:G20:G22:G14:G21.
ten
H-bonds
with
the
phosphate
backbone
of
G2:G4:G22:G4:T18:G16:T17 residues and only four electrostatic
interactions with G2:G4:T18:T17 (Fig. 5A). In addition, the 2-
methoxybenzyl group of compound 11 is involved in three T-sha-
ped
p-p stacking with G14:G14. The reduced number of electro-
static interactions that occurs in the hybrid-2 topology for both
compounds 9 and 11 could be attributed to an increased water
solvent exposure of both spermidine and spermine-like side chains
with respect to the hybrid-1 topology.
2.3.1.2. Hybrid-2. Similarly to hybrid-1, also the binding mode
against hybrid-2 conformation (PDB code: 2JSL) (Fig. 5) is charac-
terized by a groove recognition, since the steric hindrance of
T11:T12:A13:A1 and T23:A7:T6 at the 5’-end and the 3’-end
2.3.1.3. ds-DNA. Concerning the dsDNA recognition of compounds
9, 11 and 16, despite the presence of an intercalative binding site,
the most populated clusters found the major and the minor grooves
as the most involved binding sites in the recognition of the NDIs
polyamine (Fig. S4) with both spermidine and spermine-like side
chains of compounds 9 and 11 involved in a much lower number of
H-bonds and electrostatic interactions with the phosphate back-
bone with respect to G-quadruplex DNA (Fig. S4A).
respectively, hinders the
p-p stacking with the NDI core. With
respect to the hybrid-1 topology, the molecular recognition of
compound 9 towards hybrid-2 is characterized by a slightly lower
number of both H-bonds and electrostatic interactions of the
spermidine side chain that is more solvent exposed (Fig. 5A). Seven
H-bonds and four electrostatic interactions are engaged, in fact,

114
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
Fig. 5. 2:1 binding mode against the hybrid-2 form of G-quadruplex DNA receptor (PDB code: 2JSL) of the NDI polyamine derivatives (A) 9, (B) 11 and the worst binder (C) 16.
Nucleic acid is displayed as light-cyan ribbon while interacting residues as stick coloured by atom type. Polyamine NDIs 9, 11 and 16 are shown as yellow, cyan and iceblue sticks
respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
2.4. Antiproliferative and proapoptotic activities
468 (breast cancer) with a GI50 value of 85 nM. The less sensitive
cell line is UO-31 (renal cancer) characterized by a GI50 value of
All the synthesized compounds were evaluated for anti-
proliferative activity in the National Cancer Institute's 60 cell lines
screen. Compounds 6, 8e11, 13e17 were selected for screening at
five concentrations using 10-fold dilutions (10ꢁ4 ꢁ 10ꢁ8 M) on the
full panel of approximately 60 human cancer cell lines derived from
9 human cancer cell types that have been grouped into disease
subpanels including leukemia, non-small cell lung, colon, central
nervous system, melanoma, ovarian, renal, prostate, and breast
tumor cell lines (Table 2 and Table S4). The antiproliferative activity
of the compounds was given as GI50 (the molar concentration of the
compound that inhibits 50% net cell growth). An evaluation of the
data reported pointed out that all the tested compounds exhibit
antiproliferative activity against all the human cell lines in the
micro to sub-micromolar range of concentration. In general, com-
pounds characterized by a spermidine-like side chain, such as
compounds 8 and 9, show the best GI50 values. The most active
compound of the series (with the exception of some cell lines) is
1.86 mM. Compound 8 is the most active compound in some lines:
in particular, in OVCAR-8 it is an order of magnitude more active
than compound 9. Compound 10, characterized by a 4-3 chain, is, in
general, less active than 8 and 9 displaying also the worst GI50 value
of all the series towards SNB-75 (CNS cancer) cell line
(GI50 ¼ 16.74
mM). Compound 6, carrying a spermine chain,
although being less active than compound 9, is characterized by a
good level of antiproliferative activity. At the opposite, its homo-
logue 11 shows the less promising activity in some cell lines (i.e.
GI50 ¼ 10.8
trogen atom has a detrimental effect on the growth inhibition ac-
tivity (6 vs 14). In particular, despite its good Tm value, compound
mM in UO31 cell line). Methylation of the terminal ni-
D
14 is the less active compound of the series in most of the cell lines.
Surprisingly, reduction of the number of nitrogen atoms or their
replacement with oxygens does not have a detrimental effect as
high as observed with compound 14.
Next, the ability of compounds 8 and 9 to induce apoptotic cell
death was evaluated. Compound 8 induced a significant increase in
apoptotic events only at the highest concentration used (14.5% vs
6.6% of untreated cells) while compound 9 showed 17.9% of
compound 9, which is characterized by the higher value of
DTm
towards G-quadruplex. Compound 9 does not show any particular
cell line selectivity being active in the sub-micromolar range of
concentration in all the cell lines assayed. The most sensitive cell
lines are SR (leukemic) with a GI50 value of 78 nM and MDA-MB-
apoptotic cells at 3.2
cells (Fig. 6A).
mM and 22.2% at 6.4 mM vs 5.8% of untreated

A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
115
Table 2
Overview of the in vitro antiproliferative screening results for compounds 6, 8e11, 13e17.
b
b
cpd
6
MG_MIDa
Cell lines (GI50
Most sensitive
,
m
M)
cpd
13
14
15
16
17
MG_MIDa
Cell lines (GI50
Most sensitive
, mM)
b
b
(GI50
,
mM)
(GI50
, mM)
Less sensitive
Less sensitive
0.680
breast
renal
1.61
leukemia
SR
(0.270)
breast
renal
UO-31
(5.88)
NSCL
NCI-eH522
(5.62)
NSCL
HOP-62
(2.66)
ovarian
ADR-RES
(2.90)
MDA-MB 468
(0.133)
leukemia
SR
(0.028)
leukemia
SR
(0.076)
leukemia
SR
(0.038)
leukemia
MOLT-4
(0.173)
UO-31
(2.675)
renal
TK-10
(1.61)
renal
UO-31
(1.86)
renal
CAKI-I
(13.15)
renal
8
0.621
0.360
2.18
3.02
1.10
1.19
0.95
T47-D
(0.342)
ovarian
OVCAR-4
(0.197)
leukemia
SR
(0.167)
ovarian
OVCAR-4
(0.152)
9
10
11
3.65
NSCL
HOP-92
(2.654)
CAKI-I
(21.7)
a
MG_MID ¼ mean graph midpoint: arithmetic mean value for all tested cancer cell lines.
b
GI50 ¼ concentration (
mM) that inhibits 50% net cell growth.
Fig. 6. A) Apoptotic cells after 24 h treatment with compounds 8 or 9. **P < 0.01; ***P < 0.001. B) Caspase-3 activity of Jurkat cells after 24 h treatment with compound 9 or 1 h with
valinomycin (Val). *P < 0.05; **P < 0.01; ***P < 0.001. C) Fraction of Jurkat cells with cleaved PARP after 24 h treatment with compounds 8 or 9. *P < 0.05; ***P < 0.001.
Moreover, the activation of caspase-3 was evaluated. This
enzyme showed an enhanced activity: a 2.6-fold increase was
3. Conclusion
observed in cells treated with compound 9 at 1.6
increase at 6.4 M (Fig. 6B).
m
M and a 5.6-fold
In summary, with the aim of expanding the structure-activity
relationship studies of compound 3 and to obtain compounds
endowed with increased antiproliferative activity, a series of hybrid
NDI-polyamines 6e17 has been synthesized. The replacement of a
2-methoxybenzylpropylendiamine group of compound 3 with a
spermidine chain has a positive effect in the G-quadruplex recog-
nition process, as compound 9 is more active than the parent
compound. Furthermore, this substitution has a positive effect also
concerning the selectivity, since compound 9 is characterized by a
lower affinity than compound 3 to non-quadruplex DNA, and the
inhibitory activity toward other anticancer targets, such as topo-
m
The activation of caspase-3 was further confirmed by poly(ADP)
ribose polymerase (PARP) cleavage in Jurkat cells. PARP, an enzyme
important in genome surveillance, is one of the natural substrates
cleaved by caspases during apoptosis. PARP is an important re-
porter of caspase-3 activation and both compounds 8 and 9 induced
PARP cleavage starting from the concentration of 3.2 mM. However,
their proapoptotic potential was different: compound 8 showed a
slight increase (7.2%) of cells with cleaved PARP compared to
compound 9 that induced 16.6% of cells with cleaved PARP versus
4.8% of untreated cells (Fig. 6C).
isomerase IIa and TAQ -polymerase. Compound 9 exhibits high
Both compounds 8 and 9 significantly altered cell-cycle resi-
dence. At 3.2 mM, compound 8 caused an accumulation of cells in
the G2/M phase (27.8% vs 20.8% of untreated cells) accompanied by
a slight compensatory decrease of cells in the G0/G1 phase (56.1% vs
62.7% of untreated cultures) (Fig. 7A). Similarly, compound 9, at the
antiproliferative activity against about 60 human tumor cell lines
with GI50 values reaching nanomolar concentrations in leukemia
and breast cancer cell lines (i.e. SR and MDA-MB-468). The high
level of in vitro and cellular activities suggest that compound 9 may
represent a promising hit for the development of new anti-
proliferative agents. Furthermore, the binding mode of the most
interesting compounds 9 and 11 against G-quadruplex and dsDNAs
has been elucidated, confirming the importance of the spermidine
and spermine side chains in engaging H-bond and electrostatic
interactions with the phosphate backbone of the G-quadruplex
same treatment concentration (3.2 mM), induced an accumulation
of cells in the G2/M phase (20.8% vs 17.1% of untreated cells) and a
decrease of cells in the G0/G1 phase (59.7% vs 68.6% of control)
(Fig. 7B).

116
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
Fig. 7. Cell-cycle distribution of Jurkat cells following 24 h of treatment with compounds 8 (A) or 9 (B). **P < 0.01; ***P < 0.001.
DNA, particularly with the hybrid-1 topology. Overall, in the
hybrid-2 form, in fact, a reduced number of electrostatic in-
teractions has been observed both for 9 and 11, probably due to a
greater water solvent exposure of the spermidine and spermine
side chains. Surprisingly, the major and the minor groove of the
dsDNA are the binding site mainly involved in the molecular
recognition.
(100 MHz, CDCl3) d 26.8, 29.9, 39.4, 62.3, 115.6, 157.8; HMRS (esi) m/
z calcd for C6H11F3NO2 [MþH] 186.0742, found 186.0569.
4.1.3. General procedure for the synthesis of 25 and 26
Et3N (2 eq), tosyl chloride (1.2 eq) and a catalytic amount of
DMAP were added to a cooled solution of 23 or 24, respectively (1
eq) in DCM and the resulting mixture was stirred at room tem-
perature for 16 h. Removal of the dried solvent gave a residue that
was purified by flash chromatography using as eluent a mixture of
ethyl acetate/petroleum ether (6:3), providing the desired products
25 and 26.
4. Experimental section
4.1. Chemistry
4.1.1. General information
4.1.3.1. 3-(2,2,2-trifluoroacetamido)propyl 4-methylbenzenesulfonate
(25). Yellow solid, 55% yield; 1H NMR (400 MHz, CDCl3)
Uncorrected melting point was taken in glass capillary tubes on
a Buchi SMP-20 apparatus. The elemental analysis was performed
with Perkin Elmer elemental analyzer 2400 CHNESI. MS spectra
were recorded on Perkin Elmer 297 and Waters ZQ 4000 and
Micromass Q-TOF Ultima Global. 1H NMR and 13C NMR were
recorded on Varian VRX 200 and 400 instruments. Chemical shifts
are reported in parts per million (ppm) relative to peak of tetra-
methylsilane (TMS) and spin multiplicities are given as s (singlet),
br s (broad singlet), d (doublet), t (triplet), q (quartet) or m
(multiplet). Chromatographic separations were performed on silica
gel columns by flash (Kieselgel 40, 0.040 e 0.063 mm, Merck) col-
umn chromatography. Reactions were followed by thin layer
chromatography (TLC) on Merck (0.25 mm) glass-packed pre-
coated silica gel plates (60 F254) and then visualized in an iodine
chamber or with a UV lamp.
d
1.94e2.00 (m, 2H), 2.46 (s, 3H), 3.45e3.50 (m, 2H), 4.11 (t, 2H,
J ¼ 5.6 Hz), 7.36e7.38 (m, 2H), 7.78e7.80 (m, 2H); 13C NMR
(100 MHz, CDCl3)
d 20.4, 30.4, 37.4, 68.2, 115.9, 120.9, 122.7, 139.7,
143.5, 157.8; HMRS (esi) m/z calcd for C12H15F3NO4S [MþH]
326.0674 found 326.0681.
4.1.3.2. 4-(2,2,2-trifluoroacetamido)butyl 4-methylbenzenesulfonate
(26). Yellow solid, 60% yield; 1H NMR (400 MHz, CDCl3)
d
1.60e1.71 (m, 4H), 2.38 (s, 3H), 3.24e3.27 (m, 2H), 4.02 (t, 2H,
J ¼ 6.8 Hz), 7.29e7.33 (m, 2H), 7.68e7.63 (m, 2H); 13C NMR
(100 MHz, CDCl3)
d 26.5, 29.6, 39.7, 69.3, 116.1, 120.6, 123.0, 139.9,
144.5, 157.9; HMRS (esi) m/z calcd for C13H17F3NO4S [MþH]
340.0830, found 340.0839.
4.1.4. General procedure for the synthesis of 27 and 28
The appropriate diamine (10 eq) was dissolved in DCM and a
solution of Boc2O (1 eq) was added dropwise and the stirring was
continued for 16 h at room temperature. Removal of the solvent
gave residue that was dissolved in DCM and washed with brine
(3 ꢂ 10 ml). The organic fraction was dried providing the desired
products 27 and 28.
4.1.2. General procedure for the synthesis of 23 and 24
To a cooled solution (0 ꢀC) of the appropriate aminoalcohol (1
eq) was added dropwise ethyltrifluoroacetate (1.2 eq) and the
resulting mixture was stirred at room temperature of 2 h. The
solvent was evaporated in vacuum and the product obtained was
used without further purification.
4.1.2.1. 2,2,2-Trifluoro-N-(3-hydroxypropyl)acetamide
(23).
1.70e1.78
4.1.4.1. tert-Butyl (3-aminopropyl)carbamate (27). Yellow oil; 72%
Yellow oil, quantitative yield; 1H NMR (400 MHz, CDCl3)
d
yield; 1H NMR (400 MHz, CDCl3)
d
1.40 (s, 9H), 1.80e1.85 (m, 2H)
2.00 (brs, 2H, exch D2O), 2.61e2.68 (m, 2H), 2.90e2.98 (m, 2H), 5.90
(brs, 1H, exch D2O); 13C NMR (100 MHz, CDCl3)
28.5, 33.3, 38.5,
(m, 2H), 3.18 (t, 2H, J ¼ 5.1 Hz), 3.69 (brs, 1H, exch D2O), 3.52 (t, 2H,
J ¼ 4.8 Hz), 8.74 (brs, 1H, exch D2O); 13C NMR (100 MHz, CDCl3)
d
d
30.5, 38.4, 61.0, 115.9, 157.8; HMRS (esi) m/z calcd for C5H9F3NO2
39.4, 79.1, 156.4. HMRS (esi) m/z calcd for C8H19N2O2 [MþH]
[MþH] 172.0585, found 172.0576.
175.1447, found 175.1441.
4.1.2.2. 2,2,2-Trifluoro-N-(4-hydroxybutyl)acetamide
(24).
4.1.4.2. tert-Butyl (4-aminobutyl)carbamate (28). Yellow oil; 60%
Yellow oil, quantitative yield; 1H NMR (300 MHz, CDCl3)
yield; 1H NMR (400 MHz, CDCl3)
d
1.38 (s, 9H), 1.52e1.59 (m, 4H)
2.34 (brs, 2H, exch D2O), 2.72e2.79 (m, 2H), 3.19e3.24 (m, 2H), 5.95
(brs, 1H, exch D2O); 13C NMR (100 MHz, CDCl3)
27.6, 28.3, 30.7,
d
1.53e1.66 (m, 4H), 3.31 (t, 2H, J ¼ 10.0 Hz), 3.65 (t, 2H, J ¼ 7.2 Hz),
3.79 (brs, 1H, exch D2O), 8.02 (brs, 1H, exch D2O); 13C NMR
d

A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
117
40.5, 41.8, 79.8, 156.5; HMRS (esi) m/z calcd for C9H21N2O2 [MþH]
(m, 5H); ESI-MS m/z: 322 (M þ H)þ.
189.1603, found 189.1610.
4.1.7.2. tert-Butyl (4-((3-aminopropyl)(tert-butoxycarbonyl)amino)
4.1.5. General procedure for the synthesis of 29 and 30
To a cooled solution (0 ꢀC) of 25 or 26 (1 eq) respectively in THF
was added a solution of 27 or 28 (1.5 eq) in THF. The reaction
mixture was stirred for 6 h at 0 ꢀC and then for 16 h at rt. Removal of
the solvent gave a crude product that was purified by flash chro-
matography using as eluent a mixture of DCM/methanol/33%
aqueous ammonia (9:1:0.05), providing the desired products 29
and 30.
butyl)(3-(benzylamino)propyl) carbamate (36). Yellow oil, 81%
yield; 1H NMR (400 MHz, CDCl3)
d 1.44 (s, 18H), 1.62e1.64 (m, 4H),
1.69e1.78 (m, 4H), 2.59 (t, 2H, J ¼ 6.8 Hz), 2.66 (t, 2H, J ¼ 6.6 Hz),
3.13e3.22 (m, 8H), 3.75 (s, 2H), 7.27e7.29 (m, 5H); ESI-MS m/z: 494
(M þ H)þ.
4.1.8. General procedure for the synthesis of 37 and 38
To a stirred solution of 35 or 36 (1 eq) in methanol, Et3N (1 eq)
and ethyltrifluoroacetate (1 eq) were added and the stirring was
continued for 16 h at room temperature. The solvent was removed
and the crude product was purified by flash chromatography using
as eluent a mixture of DCM/methanol/33% aqueous ammonia
(9:1:0.05), providing the desired products 37 and 38.
4.1.5.1. tert-Butyl
butyl)carbamate (29). Yellow oil, 35%yield; 1H NMR (400 MHz,
CDCl3) 1.44 (s, 9H), 1.52e1.54 (m, 4H), 1.72e1.75 (m, 2H),
2.62e2.65 (m, 2H), 2.82e2.85 (m, 2H), 3.11 (m, 2H), 3.46e3.49 (m,
2H); 13C NMR (100 MHz, CDCl3)
27.1, 27.3, 28.0, 39.3, 40.2, 47.8,
(4-((3-(2,2,2-trifluoroacetamido)propyl)amino)
d
d
48.8, 49.5, 78.7, 111.5, 114.4, 117.2, 120.1, 156.0, 156.8, 157.5; HMRS
(esi) m/z calcd for C14H27F3N3O3 [MþH] 342.2005, found 342.2015.
4.1. 8.1. tert-Butyl (3-(benzylamino)propyl)(3-(2, 2, 2-
trifluoroacetamido)propyl)carbamate (37). Yellow oil, 72% yield;
1H NMR (400 MHz, CDCl3)
d 1.42 (s, 9H), 1.69e1.73 (m, 4H), 2.61 (t,
4.1.5.2. tert-Butyl (3-((4-(2,2,2-trifluoroacetamido)butyl)amino)pro-
pyl)carbamate (30). Yellow oil, 47% yield; 1H NMR (400 MHz,
2H, J ¼ 6.8 Hz), 3.18e3.20 (m, 2H), 3.27e3.29 (m, 4H), 3.76 (s, 2H),
7.29e7.31 (m, 5H); ESI-MS m/z: 418 (M þ H)þ.
CDCl3)
d
1.41 (s, 9H), 1.54e1.68 (m, 6H), 2.64 (t, 4H, J ¼ 6.8 Hz), 3.07
(t, 2H, J ¼ 6.4 Hz), 3.27e3.30 (m, 2H); 13C NMR (100 MHz, CDCl3)
4.1.8.2. tert-butyl(3-(benzylamino)propyl)(4-((tert-butox-
d
25.7, 26.1, 26.4, 27.4, 28.8, 37.5, 39.0, 39.2, 78.5, 111.9, 114.7, 117.6,
ycarbonyl)(3-(2,2,2-trifluoroacetamido) propyl)amino)butyl)carba-
120.4, 156.9, 157.2, 157.7, 158.0; HMRS (esi) m/z calcd for
14H27F3N3O3 [MþH] 342.2005, found 342.2011.
mate (38). Yellow oil, 92% yield; 1H NMR (400 MHz, CDCl3)
d 1.44 (s,
C
18H), 1.46e1.48 (m, 6H), 1.67e1.70 (m, 4H), 2.62 (t, 2H, J ¼ 7.0 Hz),
3.11e3.14 (m, 4H), 3.27e3.31 (m, 4H), 3.78 (s, 2H), 7.30e7.32 (m,
5H); ESI-MS m/z: 588 (M þ H)þ.
4.1.6. General procedure for the synthesis of 31 and 32
To a stirred solution of 29 or 30 (1 eq) in Methanol was added
slowly a solution of NaOH 40% p/p. The reaction mixture was
allowed to stir at rt for 16 h. After removal of the solvent, the re-
sidual was taken up with DCM and washed with brine (3 ꢂ 10 ml).
The organic layer was dried to obtain the desire products 31 and 32.
4.1.9. General procedure for the synthesis of 39 and 40
37 or 38 (1 eq) was dissolved in THF, Et3N (1 eq) and methyl
iodide (1.5 eq) were added and the resulting mixture was stirred for
24 h h at room temperature. The solvent was removed under vac-
uum and the crude product was purified by flash chromatography
using as eluent a mixture of DCM/methanol/33% aqueous ammonia
(9:1:0.04), providing the desired products 39 and 40.
4.1.6.1. tert-Butyl (3-aminopropyl)(4-((tert-butoxycarbonyl)amino)
butyl)carbamate (31). Yellow oil, 87% yield, 1H NMR (400 MHz,
D2O)
(m, 6H), 2.93e2.94 (m, 2H); 13C NMR (100 MHz, D2O)
28.2, 38.4, 39.9, 43.6, 46.3, 79.2, 155.7; HMRS (esi) m/z calcd for
17H36N3O4 [MþH] 346.2706, found 342.2709.
d
1.28 (s, 18H),1.32e1.35 (m, 4H), 1.44e1.52 (m, 2H), 2.41e2.52
d
27.1.27.9,
4.1.9.1. tert-butyl(3-(benzyl(methyl)amino)propyl)(3-(2,2,2-
trifluoroacetamido)propyl)carbamate (39). Yellow oil, 66% yield; 1H
C
NMR (400 MHz, CDCl3)
d 1.44 (s, 9H), 1.58e1.64 (m, 6H), 2.20 (s,
3H), 2.22e2.23 (m, 2H), 3.26e3.27 (m, 4H), 3.42 (s, 2H), 7.31e7.33
4.1.6.2. tert-Butyl (4-aminobutyl)(3-((tert-butoxycarbonyl)amino)
(m, 5H); ESI-MS m/z: 432 (M þ H)þ.
propyl)carbamate (32). Yellow oil, 64% yield; 1H NMR (400 MHz,
D2O) d 1.30 (s,18H),1.39e1.41 (m, 4H),1.53e1.58 (m, 2H), 2.49e2.61
(m, 6H), 2.97e2.99 (t, 2H, J ¼ 6.8 Hz); 13C NMR (100 MHz, D2O)
4.1.9.2. tert-butyl(3-(benzyl(methyl)amino)propyl)(4-((tert-butox-
ycarbonyl)(3-(2,2,2-trifluoroacetamido)propyl)amino)butyl)carba-
d
22.9, 23.5, 24.5, 25.6, 27.4, 38.5, 44.8, 46.5, 80.8, 157.9; HMRS (esi)
mate (40). Yellow oil, 30% yield; 1H NMR (400 MHz, CDCl3)
d 1.41 (s,
m/z calcd for C17H36N3O4 [MþH] 346.2706, found 346.2714.
18H), 1.43e1.46 (m, 4H), 1.65e1.67 (m, 4H), 2.17 (s, 3H), 3.13e3.15
(m, 6H), 3.26e3.29 (m, 6H), 3.46 (s, 2H), 7.27e7.29 (m, 5H); ESI-MS
m/z: 604 (M þ H)þ.
4.1.7. General procedure for the synthesis of 35 and 36
To a solution of 33 or 34 (5 eq) in toluene was added benzal-
dehyde (1 eq) and the resulting mixture was refluxed for 5 h. The
solvent was removed in vacuo and the residue was taken up with
ethanol. NaBH4 (3 eq) was added and the solution was stirred for
12 h at room temperature. After evaporation of the solvent, the
crude product was dissolved in DCM and washed with brine
(2 ꢂ 10 ml). The organic phase was dried, evaporated and purified
by flash chromatography using as eluent a mixture of DCM/meth-
anol/33% aqueous ammonia (9:1:0.1), providing the desired prod-
ucts 35 and 36.
4.1.10. General procedure for the synthesis of 41 and 42
To a stirred solution of 39 or 40 in methanol was added a so-
lution of NaOH 40% p/p. The resulting mixture was allowed to stir
for 16 h at room temperature. After removal of the solvent, the
residue was dissolved in DCM and was washed with water
(3 ꢂ 10 ml). The organic layer was dried providing the desire
products 41 and 42.
4.1.10.1. tert-Butyl (3-aminopropyl)(3-(benzyl(methyl)amino)propyl)
4.1.7.1. tert-Butyl
(3-aminopropyl)(3-(benzylamino)propyl)carba-
1.41
carbamate (41). Yellow oil, 88% yield; 1H NMR (400 MHz, CDCl3)
d 1.43 (s, 9H), 1.61e1.64 (m, 2H), 1.71e1.74 (m, 2H), 2.16 (s, 3H), 2.34
mate (35). Yellow oil; 80% yield; 1H NMR (400 MHz, CDCl3)
d
(s, 9H), 1.74e1.77 (m, 4H), 2.61e2.65 (m, 2H), 2.73e2.78 (m, 2H),
3.13 (brs, 3H, exch D2O), 3.21e3.27 (m, 4H), 3.81 (s, 2H), 7.29e7.32
(t, 2H, J ¼ 7.2 Hz), 2.67 (t, 2H, J ¼ 6.6 Hz), 3.16e3.19 (m, 4H), 3.46 (s,
2H), 7.28e7.31 (m, 5H); ESI-MS m/z: 336 (M þ H)þ.

118
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
4.1.10.2. tert-butyl(4-((3-aminopropyl)(tert-butoxycarbonyl)amino)
4.1.12.3. tert-Butyl (3-((tert-butoxycarbonyl)amino)propyl)(3-(7-(3-
((2-methoxybenzyl)amino)propyl)-1,3,6,8-tetraoxo-3,6,7,8-
tetrahydrobenzo[lmn][3,8]phenanthrolin-2(1H)-yl)propyl)carbamate
butyl)(3-(benzyl(methyl)amino) propyl)carbamate (42). Yellow
solid, 92% yield; 1H NMR (400 MHz, CDCl3)
d 1.41 (s, 18H), 1.45e1.46
(m, 4H), 1.68e1.71 (m, 4H), 2.15 (s, 3H), 2.33 (t, 2H, J ¼ 6.0 Hz),
2.73e2.75 (m, 2H), 3.00e3.03 (m, 4H), 3.12e3.15 (m, 4H), 3.44 (s,
2H), 7.21e7.27 (m, 5H); ESI-MS m/z: 508 (M þ H)þ.
(51). Brown oil; 15% yield; 1H NMR (400 MHz, CDCl3)
d 1.43 (s,18H),
1.62e1.73 (m, 2Hþ1H exch D2O), 1.95e2.01 (m, 4H), 2.71 (t, 2H,
J ¼ 6.4 Hz), 3.10e3.12 (m, 2H), 3.30e3.34 (m, 4H), 3.77 (s, 2H), 3.83
(s, 3H), 3.88e4.30 (m, 4H), 4.76 (brs, 1H, exch D2O), 6.83e6.86 (t,
2H, J ¼ 7.2 Hz), 7.17e7.21 (m, 2H), 8.72e8.26 (m, 4H); ESI-MS m/z:
759 (M þ H)þ.
4.1.11. General procedure for the synthesis of 43 and 44
A solution of 41 or 42 dissolved in methanol, was hydrogenated
over 10% Pd on charcoal under stirring for 2 h. After the con-
sumption of the starting material, the solution was filtered in order
to remove the catalyst and the solution was dried under vacuum.
The crude product was purified by flash chromatography using as
4.1.12.4. tert-butyl(4-((tert-butoxycarbonyl)amino)butyl)(3-(7-(3-
((2-methoxybenzyl)amino)propyl)-1,3,6,8-tetraoxo-3,6,7,8-
tetrahydrobenzo[lmn][3,8]phenanthrolin-2(1H)-yl)propyl)carbamate
(52). Yellow oil,19% yield; 1H NMR (400 MHz, CDCl3)
d 1.43 (s, 18H),
eluent
a mixture of DCM/methanol/33% aqueous ammonia
(8:2:0.2), providing the desired products 43 and 44.
1.52e1.56 (m, 4H), 1.95e2.02 (m, 4H), 2.06 (brs, 1H, exch D2O),
2.63e2.67 (m, 6H), 2.71e2.75 (m, 2H), 3.12 (brs, 1H, exch D2O), 3.79
(s, 2H), 3.84 (s, 3H), 4.29 (t, 4H, J ¼ 7.0 Hz), 6.84 (t, 2H, J ¼ 7.4 Hz),
7.19 (t, 2H, J ¼ 6.8 Hz), 8.73 (d, 4H, J ¼ 7.2 Hz); 13C NMR (100 MHz,
4.1.11.1. tert-Butyll (3-aminopropyl)(3-(methylamino)propyl)carba-
mate (43). Yellow oil, 68% yield; 1H NMR (400 MHz, CDCl3)
d 1.40 (s,
CDCl3)
d 27.2, 27.8, 28.1, 28.4, 29.6, 38.8, 39.0, 40.4, 46.4, 46.9, 48.9,
9H), 1.61e1.69 (m, 4H), 1.72 (brs, 1H, exch D2O), 2.38 (s, 3H), 2.56 (t,
2H, J ¼ 6.6 Hz), 2.67 (t, 2H, J ¼ 6.6 Hz), 3.16e3.23 (m, 4H); ESI-MS m/
z: 246 (M þ H)þ.
49.3, 55.2, 78.9, 110.1, 120.3, 126.4, 126.6, 128.0, 128.1, 129.7, 130.9,
130.9, 156.0, 157.5, 162.8, 162.9; ESI-MS m/z: 773 (M þ H)þ.
4.1.12.5. tert-butyl(3-((tert-butoxycarbonyl)amino)propyl)(4-(7-(3-
((2-methoxybenzyl)amino)propyl)-1,3,6,8-tetraoxo-3,6,7,8-
tetrahydrobenzo[lmn][3,8]phenanthrolin-2(1H)-yl)butyl)carbamate
4.1.11.2. tert-butyl(3-aminopropyl)(4-((tert-butoxycarbonyl)(3-
(methylamino)propyl)amino)butyl) carbamate (44). Yellow oil, 76%
yield; 1H NMR (400 MHz, CDCl3)
d 1.42 (s, 18H), 1.45e1.47 (m, 4H),
(53). Yellow oil; 15% yield; 1H NMR (400 MHz, CDCl3)
d 1.42 (s, 9H),
1.60e1.63 (m, 2H), 1.64e1.66 (m, 2H), 1.94 (brs, 3H, exch D2O), 2.41
(s, 3H), 2.54e2.56 (m, 2H), 2.66 (t, 2H, J ¼ 6.6 Hz), 3.14e3.22 (m,
8H); ESI-MS m/z: 418 (M þ H)þ.
1.64e1.73 (m, 4H), 1.78e1.85 (m, 2H), 1.95e2.02 (m, 2H), 2.3 (brs,
1H, exch D2O), 2.70e2.75 (m, 6H), 3.20e3.22 (m, 2H), 3.79 (s, 2H),
3.84 (s, 3H), 4.22 (t, 2H, J ¼ 7.6 Hz), 4.29 (t, 2H, J ¼ 6.4 Hz), 5.15 (brs,
1H exch D2O), 6.84 (t, 2H, J ¼ 7.2 Hz), 7.19 (t, 2H, J ¼ 6.8 Hz), 8.73 (s,
4.1.12. General procedure for the synthesis of 49e59 and 17
An equimolar mixture of the appropriate diamine 45e46 (1 eq),
1,4,5,8-naphthalene-tetracarbocylic dianhydride (1 eq) and (poly)
amines 18e22, 31, 32, 43, 44, 47 and 48 (1 eq) DMF was refluxed for
2 h. After cooling down, removal of the solvent gave residue that
was purified by flash chromatography using as eluent a mixture of
toluene/DCM/methanol/33% aqueous ammonia (4:5:1:0.03)
providing the desired products 49e59 and 17.
4H). 13C NMR (100 MHz, CDCl3)
d 25.7, 27.0, 28.2, 28.4, 29.5, 38.9,
39.0, 40.5, 46.4, 47.4, 48.9, 49.1, 55.2, 79.0, 120.3, 126.5, 126.6, 128.0,
128.1, 129.7, 130.4, 130.9, 156.2, 157.5, 162.8; ESI-MS m/z: 773
(M þ H)þ.
4.1.12.6. tert-butyl(2-((tert-butoxycarbonyl)(3-(7-(3-((2-
methoxybenzyl)amino)propyl)-1,3,6,8-tetraoxo-3,6,7,8-
tetrahydrobenzo[lmn][3,8]phenanthrolin-2(1H)-yl)propyl)amino)
ethyl)(3-((tert-butoxycarbonyl)amino)propyl)carbamate
(54).
4.1.12.1. tert-butyl(4-((tert-butoxycarbonyl)(3-((tert-butox-
ycarbonyl)amino)propyl)amino)butyl)(3-(7-(3-((2-methoxybenzyl)
amino)propyl)-1,3,6,8-tetraoxo-3,6,7,8-tetrahydrobenzo[lmn][3,8]
phenanthrolin ꢁ2(1H)-yl)propyl)carbamate (49). Brown oil; 15%
Brown oil; 16% yield; NMR (200 MHz, CDCl3) 1.46 (s, 27H), 1.66
d
(brs, 1H, exch D2O), 1.89e1.98 (m, 6H), 2.69e2.76 (t, 2H, J ¼ 6.8 Hz),
2.88e3.26 (m, 10H), 3.78 (s, 2H), 3.84 (s, 3H), 4.20e4.32 (m, 4H),
5.12 (brs, 1H, exch D2O), 6.81e6.88 (t, 2H, J ¼ 7.0 Hz), 7.16e7.19 (t,
2H, J ¼ 6.3 Hz), 8.74 (s, 4H); ESI-MS m/z: 902 (M þ H)þ.
yield; 1H NMR (400 MHz, CDCl3)
d 1.43 (s, 27H), 1.62e1.73 (m,
2Hþ1H exch D2O), 1.95e2.01 (m, 4H), 2.71 (t, 2H, J ¼ 6.4 Hz),
3.10e3.12 (m, 2H), 3.30e3.34 (m, 12H), 3.77 (s, 2H), 3.83 (s, 3H),
3.88e4.30 (m, 4H), 4.76 (brs, 1H, exch D2O), 6.83e6.86 (t, 2H),
7.17e7.21 (m, 2H), 8.72e8.26 (m, 4H); 13CNMR (100 MHz, CDCl3)
4.1.12.7. tert-butyl(3-((tert-butoxycarbonyl)(3-((tert-butox-
ycarbonyl)amino)propyl)amino)propyl)(3-(7-(3-((2-methoxybenzyl)
amino)propyl)-1,3,6,8-tetraoxo-3,6,7,8-tetrahydrobenzo [lmn][3,8]
phenanthrolin-2(1H)-yl)propyl)carbamate (55). Brown oil; 20%
d
25.4, 28.5, 28.1, 38.8, 39.2, 46.7, 48.9, 79.6, 106.5, 123.9, 126.3,
126.3, 126.9, 129.2, 130.9, 131.4, 142.4, 152.5, 153.9, 163.5; HMRS
yield; NMR (400 MHz, CDCl3)
d
1.43 (s, 27H), 1.62e1.73 (m, 2Hþ1H
(esi) m/z calcd for C50H69N6O11 [MþH] 929.5024, found 929.5013.
exch D2O), 1.95e2.01 (m, 2H), 2.71 (t, 2H, J ¼ 6.4 Hz), 3.10e3.12 (m,
2H), 3.30e3.34 (m, 12H), 3.77 (s, 2H), 3.83 (s, 3H), 3.88e4.30 (m,
4H), 4.76 (brs, 1H, exch D2O), 6.83e6.86 (t, 2H), 7.17e7.21 (m, 2H),
8.72e8.26 (m, 4H); HMRS (esi) m/z calcd for C49H66N6O11 [MþH]
914.4790, found 914.4797.
4.1.12.2. tert-butyl(4-((tert-butoxycarbonyl)(3-((tert-butox-
ycarbonyl)amino)propyl)amino)butyl)(3-(1,3,6,8-tetraoxo-7-(3-
((2,3,4-trimethoxybenzyl)amino)propyl)-3,6,7,8-tetrahydrobenzo
[lmn][3,8]phenanthrolin-2(1H)-yl)propyl)carbamate
(50).
1.43 (s, 27H),
Brown oil; 16% yield; 1H.NMR (400 MHz, CDCl3)
d
1.45e1.52 (m, 4H), 1.94e1.99 (m, 4Hþ1H exch D2O), 2.74 (t, 2H,
J ¼ 6.8 Hz), 3.08e3.31 (m, 10H), 3.73 (s, 2H), 3.84 (s, 9H), 4.21 (t, 2H,
J ¼ 7.0 Hz), 4.29 (t, 2H, J ¼ 7.2 Hz), 5.16 (brs, 1H, exch D2O), 6.57 (d,
2H, J ¼ 8.4 Hz), 6.92 (d, 2H, J ¼ 8.4 Hz), 8.74 (s, 4H); 13CNMR
4.1.12.8. tert-butyl(3-((tert-butoxycarbonyl)(methyl)amino)pro-
pyl)(3-(7-(3-((2-methoxybenzyl)amino)
propyl)-1,3,6,8-tetraoxo-
3,6,7,8-tetrahydrobenzo[lmn][3,8]phenanthrolin-2(1H)-yl)propyl)
carbamate (56). Yellow oil; 16% yield; 1H NMR (400 MHz, CDCl3)
(100 MHz, CDCl3)
d
25.7, 28.2, 28.4, 38.7, 39.0, 46.5, 48.6, 55.9, 60.7,
d
1.42 (s, 9H), 1.96e2.01 (m, 6H), 2.59 (s, 3H), 2.73 (t, 2H, J ¼ 6.8 Hz),
61.0, 79.4, 106.9, 123.9, 126.0, 126.4, 126.6, 128.9, 130.8, 130.9, 142.0,
152.9, 153.7, 162.8; HMRS (esi) m/z calcd for C52H73N6O13 [MþH]
989.5236, found 898.5230.
2.85e2.88 (m, 2H), 3.23e3.35 (m, 4H), 3.80 (s, 2H), 3.82 (s, 3H), 4.17
(t, 2H, J ¼ 7.6 Hz), 4.26 (t, 2H, J ¼ 7.2 Hz), 6.81e6.83 (m, 2H),
7.18e7.24 (m, 3H), 8.72 (s, 4H); ESI-MS m/z: 773 (M þ H)þ.

A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
119
4.1.12.9. tert-butyl(4-((tert-butoxycarbonyl)(3-(methylamino)pro-
pyl)amino)butyl)(3-(7-(3-((2-methoxybenzyl)amino)propyl)-1,3,6,8-
tetraoxo-3,6,7,8-tetrahydrobenzo[lmn][3,8]phenanthrolin-2(1H)-yl)
propyl)carbamate (57). Yellow oil; 23% yield; 1H NMR (400 MHz,
(s, 3H), 3.19e3,24 (m, 6H), 6.74 (d, 1H, J ¼ 9.2), 7.13 (d, 1H, J ¼ 8.0),
8.42e8.47 (m, 4H); 13C NMR (100 MHz, D2O)
22.7, 23.7, 24.0, 24.2,
d
36.5, 37.6, 43.7,44.5, 45.3, 45.9, 47.0, 55.9, 60.7, 61.2, 107.9, 116.0,
125.4, 125.5, 125.6, 126.7, 130.9, 140.7, 151.8, 154.4, 163.5, 163.6; ESI-
CDCl3)
d
1.43 (s, 18H), 1.48e1.52 (m, 6H), 1.97e2.01 (m, 4H), 2.58 (s,
MS m/z: 345 (Mþ2H)2þ
.
3H), 2.73 (t, 2H, J ¼ 6.8), 2.77e2.79 (t, 2H), 3.09e3.27 (m, 8H), 3.80
(s, 2H), 3.83 (m, 3H), 4.15e4.19 (m, 2H), 4.28 (t, 2H, J ¼ 6.8 Hz),
6.82e6.86 (m, 2H), 7.17e7.21 (m, 2H), 8.74 (s, 4H); ESI-MS m/z: 844
(M þ H)þ.
4.1.13.3. 2-(3-((3-aminopropyl)amino)propyl)-7-(3-((2-
methoxybenzyl)amino)propyl)benzo[lmn] [3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (8). Orange solid; quantitative yield; m.p.
>250 ꢀC; 1H NMR (400 MHz, D2O)
d 2.15-2-19 (m, 6H), 3.13e3.24
4.1.12.10. 2-(3-(4-(3-aminopropoxy)butoxy)propyl)-7-(3-((2-
methoxybenzyl)amino)propyl)benzo[lmn] [3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (58). Brown oil; 26% yield; 1H NMR
(m, 8H), 3.89 (s, 3H), 4.18e4.27 (m, 6H), 6.83 (t, 1H, J ¼ 7.7), 7.01 (d,
1H, J ¼ 8.2), 7.32e7.35 (m, 2H), 8.44e8.47 (m, 4H); 13C NMR
(100 MHz, D2O)
d 23.7, 23.8, 24.3, 36.6, 36.7, 44.1, 44.7, 45.5, 55.6,
(400 MHz, CDCl3)
d
1.41 (s, 9H), 1.53e1.54 (m, 4H), 1.70e1.72 (m,
111.0, 118.0, 120.7, 125.0, 125.2, 130.9, 131.6, 157.6, 162.9, 163.0; ESI-
2H), 1.89 (brs, 1H, exch D2O), 1.93e2.03 (m, 4H), 2.70 (t, 2H,
J ¼ 6.8 Hz), 3.17e3.19 (m, 2H), 3.19e3.33 (m, 2H), 3.34e3.44 (m,
4H), 3.54 (t, 2H, J ¼ 6.0 Hz), 3.76 (s, 2H), 3.81 (s, 3H), 4.25e4.31 (m,
4H), 4.99 (brs, 1H, exch D2O), 6.81 (t, 2H, J ¼ 8.0 Hz), 7.16 (t, 2H,
MS m/z: 280 (Mþ2H)2þ
.
4.1.13.4. 2-(3-((4-aminobutyl)amino)propyl)-7-(3-((2-
methoxybenzyl)amino)propyl)benzo[lmn] [3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (9). Brown solid; quantitative yield; m.p.
J ¼ 6.2 Hz), 8.69 (s, 4H); 13C NMR (100 MHz, CDCl3)
d 26.3, 26.4,
28.2, 28.3, 28.4, 28.7, 38.7, 39.1, 46.5, 49.0, 55.20, 68.7, 69.2, 70.6,
70.7, 77.4, 110.1, 120.3, 126.5, 126.6, 126.7, 128.1, 128.2, 129.7, 130.8,
130.8155.9, 157.6, 162.6, 162.8; HMRS (esi) m/z calcd for C35H43N4O7
[MþH] 631.3132, found 631.3139.
>250 ꢀC; 1H NMR (400 MHz, D2O)
d 1.62e1.68 (m, 4H), 2.00e2.04
(m, 4H), 2.91 (t, 2H, J ¼ 7.0), 2.96e3.01 (m, 4H), 3.07 (t, 2H, J ¼ 7.6),
3.74 (s, 3H), 4.03e4.10 (m, 4H), 4.13 (s, 2H), 6.71 (t, 1H, J ¼ 8.2), 6.87
(d, 1H, J ¼ 7.6), 7.13e7.19 (m, 2H), 8.34e8.36 (m, 4H); 13C NMR
(100 MHz, D2O)
d 22.7, 23.8, 23.9, 24.2, 37.6, 37.8, 38.7, 44.0, 45.2,
4.1.12.11. tert-butyl(3-(7-(3-((2-methoxybenzyl)amino)propyl)-
1,3,6,8-tetraoxo-3,6,7,8-tetrahydrobenzo[lmn][3,8]phenanthrolin-
2(1H)-yl)propyl)carbamate (59). Yellow oil, 16% yield; 1H NMR
46.6, 47.0, 55.4, 111.0, 118.0, 120.7, 125.1, 125.2, 125.3, 130.9, 131.5,
157.6, 163.2; ESI-MS m/z: 287 (Mþ2H)2þ
.
(400 MHz, CDCl3)
d
1.44 (s, 9H), 1.94e1.97 (m, 4H), 2.72 (t, 2H,
4.1.13.5. 2-(4-((4-aminobutyl)amino)butyl)-7-(3-((2-
methoxybenzyl)amino)propyl)benzo[lmn][3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (10). Brown solid, quantitative yield, m.p. .
J ¼ 7.0 Hz), 3.18e3.19 (m, 2H), 3.77 (s, 2H), 3.82 (s, 3H), 4.29 (t, 4H,
J ¼ 6.2 Hz); HMRS (esi) m/z calcd for C33H37N4O7 [MþH] 601.2662,
found 601.2671.
>250 ꢀC; 1H NMR (400 MHz, D2O)
d 1.84e2.11 (m, 4H), 2.12e2.21
(m, 4H), 3.11e3.22 (m, 8H), 3.90 (s, 3H), 4.24 (t, 2H, J ¼ 7.0), 4.21 (t,
4.1.12.12. 2-Ethyl-7-(3-((2-methoxybenzyl)amino)propyl)benzo[lmn]
2H, J ¼ 6.8), 4.29 (s, 2H), 6.88 (t, 1H, J ¼ 7.6), 7.03 (d, 1H, J ¼ 8),
[3,8]phenanthroline-1,3,6,8(2H,7H)-tetraone
(17). Brown solid,
1.12 (t,
8.49e8.50 (m, 4H); 13C NMR (100 MHz, D2O)
d 23.3, 23.7, 24.0, 24.2,
36.6, 37.6, 40.1, 44.0, 44.5, 46.7, 47.3, 55.4, 111.1, 118.1, 120.8, 125.1,
quantitative yield, m.p. >250 ꢀC; 1H NMR (400 MHz, D2O)
d
3H, J ¼ 7.2 Hz), 2.01e2.05 (m, 4H), 3.00 (t, 2H, J ¼ 7.4 Hz), 3.75 (s,
3H), 3.84e3.89 (m, 2H), 4.02 (t, 2H, J ¼ 6.2 Hz), 4.13 (s, 2H), 6.73 (t,
1H, J ¼ 7.4 Hz), 6.87 (d, 1H, J ¼ 8.4 Hz), 7.15e7.19 (m, 2H), 8.15 (s,
125.5, 130.8, 130.9, 131.6, 157.7, 163.2, 163.4; ESI-MS m/z: 287
(Mþ2H)2þ
.
4H). 13C NMR (100 MHz, D2O)
d
15.65, 26.74, 36.45, 45.51, 46.43,
4.1.13.6. 2-(3-((2-((3-aminopropyl)amino)ethyl)amino)propyl)-7-(3-
((2-methoxybenzyl)amino) propyl)benzo[lmn][3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (11). Orange solid, quantitative yield; m.p.
54.77, 109.87, 117.37, 120.16, 123.79, 125.63, 130.25, 130.98, 155.83,
160.16; ESI-MS m/z: 473 (M þ H).
>250 ꢀC; 1H NMR (400 MHz, D2O)
d 2.05e2.22 (m, 6H), 3.11e3.22
4.1.13. General procedure for the synthesis of 6e16
(m, 4H), 3.24e3.35 (m, 4H), 3.57 (s, 4H), 3.89 (s, 3H), 4.18 (t, 2H,
J ¼ 7.2), 4.23 (t, 4H, J ¼ 6.8), 6.85 (t, 1H, J ¼ 7.5), 7.02 (d, 1H, J ¼ 8.1),
7.27e7.34 (m, 2H), 8.38e8.44 (m, 4H); 13C NMR (100 MHz, D2O)
To solution of 49e59 in methanol (20 ml), HCl 3 M (20 mL) was
added and the resulting solution was stirred overnight at room
temperature. Following solvent removal, the residue was washed
with Et2O (5 ꢂ 20 ml). The resulting solid was filtered and dried to
afford 6e16 as hydrochloride salt.
d
23.6, 24.0, 24.3, 36.4, 37.6, 37.7, 43.2, 43.4, 44.0, 45.1, 46.0, 46.6,
55.4, 111.1, 118.1, 120.8, 125.5, 125.6, 131.0, 131.6, 157.7, 163.3; ESI-MS
m/z: 302 (Mþ2H)2þ
.
4.1.13.1. 2-(3-((4-((3-aminopropyl)amino)butyl)amino)propyl)-7-(3-
((2-methoxybenzyl)amino)propyl)benzo[lmn][3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (6). Brown solid, quantitative yield; m.p.
4.1.13.7. 2-(3-((3-((3-aminopropyl)amino)propyl)amino)propyl)-7-
(3-((2-methoxybenzyl)amino) propyl)benzo[lmn][3, 8]
phenanthroline-1,3,6,8(2H,7H)-tetraone (12). Brown solid, quanti-
>250 ꢀC; 1H NMR (400 MHz, D2O)
d
1.82e1.84 (m, 4H), 2.12e2.18
tative yield; m.p. >250 ꢀC; 1H NMR (400 MHz, D2O)
d 2.07e2.18 (m,
(m, 6H), 3.02e3.23 (m, 12H), 3.90 (s, 3H), 4.21e4.29 (m, 6H), 7.86 (t,
1H, J ¼ 7.4), 7.04 (d, 1H, J ¼ 8.2), 7.27e7.36 (m, 2H), 8.56e8.58 (m,
6H), 3.09e3.25 (m,12H), 3.88 (s, 3H), 4.18e4.78 (m, 6H), 6.86 (d,1H,
J ¼ 7.4), 7.01 (d, 1H, J ¼ 8.4), 7.24e7.41 (m, 2H), 8.37e8.52 (m, 4H);
4H); 13C NMR (100 MHz, D2O)
d
22.8, 23.7, 24.0, 24.3, 24.8, 34.6,
ESI-MS m/z: 308 (Mþ2H)2þ
.
36.7, 37.0, 37.7, 37.9, 44.1, 45.3, 46.6, 47.0, 55.5, 111.0, 118.1, 120.7,
125.0, 125.1, 125.2, 125.3, 130.9, 131.6, 157.6, 163.1, 163.2, 163.3; ESI-
4.1.13.8. 2-(3-((2-methoxybenzyl)amino)propyl)-7-(3-((3-(methyl-
amino)propyl)amino)propyl)benzo [lmn][3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (13). Yellow solid, quantitative yield; m.p.
MS m/z: 315 (Mþ2H)2þ
.
4.1.13.2. 2-(3-((4-((3-aminopropyl)amino)butyl)amino)propyl)-7-(3-
((2,3,4-trimethoxybenzyl)amino)propyl)benzo[lmn][3,8]phenanthro-
line-1,3,6,8(2H,7H)-tetraone (7). Brown solid, quantitative yield,
>250 ꢀC; 1H NMR (400 MHz, D2O)
d 2.15e2.21 (m, 6H), 3.12e3.35
(m, 11H), 3.89 (s, 3H), 4.18e4.48 (m, 6H), 6.86 (t, 1H, J ¼ 7.5), 7.02 (d,
1H, J ¼ 8.0), 7.30e7.34 (m, 2H), 8.50e8.59 (m, 4H); 13C NMR
m.p.>250 ꢀC; 1H NMR (400 MHz, D2O)
d
1.81e1.85 (m, 4H),
(100 MHz, D2O) d 22.5, 24.0, 24.2, 32.8, 43.1, 44.0, 44.5, 45.4, 45.7,
2.07e2.16 (m, 6H), 3.10e3.21 (m,12H), 3.69 (s, 3H), 3.74 (s, 3H), 3.92
46.7, 55.4, 63.5, 111.1, 118.1, 120.8, 125.5, 125.6, 131.0, 131.6, 157.7,

120
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
163.6; ESI-MS m/z: 287 (Mþ2H)2þ
.
0.1% acetic acid pH 8.0, added of 0.1% SDS). The electrophoretic run
was performed at 4.9 V/cm for 90 min min and finally the bands
were stained with ethidium bromide (0.5 g/ml in H2O). When the
topoisomerase cleavage activity was tested, the reaction was per-
formed with 5 U of enzyme. Before loading, the reaction mixture
4.1.13.9. 2-(3-((2-methoxybenzyl)amino)propyl)-7-(3-((4-((3-(meth-
ylamino)propyl)amino)butyl) amino)propyl)benzo[lmn][3,8]
phenanthroline-1,3,6,8(2H,7H)-tetraone (14). Brown solid, quanti-
m
tative yield, m.p. >250 ꢀC; 1H NMR (400 MHz, D2O)
d
1.81e1.83 (m,
was treated with proteinase K (1.6 mg), Na2EDTA (12.5 mM) and SDS
4H), 2.11e2.21 (m, 6H), 2.76 (s, 3H), 3.14e3.25 (m, 12H), 3.90 (s, 3H),
4.23e4.29 (m, 6H), 6.85 (t, 1H, J ¼ 7.6) 7.02 (d, 1H, J ¼ 8.0),
(0.5%) (Sigma), for 30 min min at 45 ꢀC.
8.66e8.72 (m, 4H); 13C NMR (100 MHz, D2O)
d
22.4, 22.7, 24.2, 32.8,
4.2.3. TAQ polymerase inhibition assay
44.0, 44.4, 45.3, 45.7, 46.7, 46.9, 55.4, 110.1, 118.1, 120.8, 125.4, 125.6,
The 1065e906 fragment of plasmid pBR322 (Fermentas) was
amplified by PCR using the primers 30GYRA (50 TGA GGA TCC GCC
TGG ACA GCA TGG 30) and 50GYRA (50 GTC GAA TTC TCG GCG AGA
131.0, 131.6, 157.7, 163.6; ESI-MS m/z: 643 (M þ H)þ.
4.1.13.10. 2-(3-(4-(3-aminopropoxy)butoxy)propyl)-7-(3-((2-
methoxybenzyl)amino)propyl)benzo [lmn][3,8]phenanthroline-
1,3,6,8(2H,7H)-tetraone (15). Orange solid, quantitative yield; m.p.
AGC AGG 30) at 0.5
m
M concentration in the presence/absence of
increasing concentrations of tested derivatives (0e20 M) using
m
2.5 ng of plasmid and 2 U of TAQ polymerase enzyme (Fermentas).
The reaction was carried out in an Eppendorf thermocycler by
repeating the following PCR scheme for 25 cycles: 30 s at 94 ꢀC;
30 s at 58 ꢀC and 30 s at 72 ꢀC. PCR products were resolved on a 2%
agarose gel in 0.5 X TBE (4.2 V/cm for 1.5 h) stained by ethidium
bromide and visualized on a Geliance apparatus.
189e191 ꢀC; 1H NMR (400 MHz, D2O)
d 1.47e1.49 (m, 4H),
1.89e1.95 (m, 4H), 2.17e2.22 (m, 2H), 3.89 (t, 2H, J ¼ 7.4), 3.12 (t, 2H,
J ¼ 7.2), 3.40e3.48 (m, 4H), 3.54e3.62 (m, 4H), 3.89 (s, 3H), 4.10 (t,
2H, J ¼ 6.9), 4.18 (t, 2H, J ¼ 7.2) 4.28 (2, 2H), 6.86 (t, 1H, J ¼ 7.6)7.01
(d, 1H, J ¼ 8), 8.35e8.42 (m, 4H); 13C NMR (100 MHz, D2O)
d 23.9,
25.2, 26.5, 27.0, 37.4, 43.9, 46.4, 55.4, 67.5, 68.0, 70.2, 70.3, 110.9,
117.9,120.6,124.6,124.8,124.8,124.9,130.6,130.8,131.4,131.5,157.6,
162.3, 162.7.; ESI-MS m/z: 632 (M þ H)þ.
4.2.4. Telomerase repeat amplification protocol assay (TRAP-lig)
Telomerase activity was monitored using a modified Telomere
Repeat Amplification Protocol (TRAP) assay. 100 ng of substrate TS
(50 AAT CCG TCG AGC AGA GTT 30) was elongated by the telomerase
4.1.13.11. 2-(3-aminopropyl)-7-(3-((2-methoxybenzyl)amino)propyl)
benzo[lmn][3,8]phenanthroline-1,3,6,8(2H,7H)-tetraone
Yellow solid, quantitative yield, m.p. >250 ꢀC; 1H NMR (400 MHz,
D2O)
(16).
present in 5 m
g of HeLa cells protein extract at 37 ꢀC for 30 min in
the presence/absence of increasing drug concentrations. Ligands
were removed by QIA quick nucleotide purification extraction kit
(Qiagen) and, after addition of 100 ng of reverse primer ACX (50 GCG
CGG CTT ACC CTT ACC CTT ACC CTA ACC 30) and 2U TAQ polymerase,
the amplification of the products was performed by PCR (33 cycles
of 30 s at 92 ꢀC, 30 s at 58 ꢀC and 45 s at 72 ꢀC). The reaction
products were loaded onto a 10% polyacrylamide gel in 0.5x TBE
and visualized by Sybr Green I staining (Euroclone).
d
2.10e2.16 (m, 4H), 3.09e3.15 (m, 4H), 3.87 (t, 2H, J ¼ 7.4),
3.87 (s, 3H), 4.15e4.22 (m, 4H), 4.25 (s, 2H), 6.84 (t, 1H, J ¼ 7.6), 6.99
(d, 1H, J ¼ 8.4), 7.25e7.30 (m, 2H), 8.44e8.50 (m, 4H); 13C NMR
(100 MHz, D2O)
d 23.9, 25.3, 37.2, 37.5, 37.6, 37.8, 44.0, 46.5, 55.4,
110.9, 118.0, 120.7, 125.0, 125.1, 130.9, 131.5, 157.6, 163.1.; ESI-MS m/
z: 502 (M þ H)þ.
4.2. Biology
4.2.5. Cell cultures
4.2.1. Fluorescence melting assay
Fluorescence melting curves were determined in a Roche
LightCycler480 (lecc 488 nm, lem 520 nm) in a total reaction volume
Jurkat T leukemia cells were purchased from ATCC (LGC Stan-
dard, Teddington, UK). Cells were grown in suspension and prop-
agated in RPMI 1640 supplemented with 10% heat-inactivated
of 20
ml containing 0.25
mM of the selected sequences (HTS:
bovine serum, 1% L-glutamine solution, and 1% penicillin/strepto-
50Dabcyl-AGG GTT AGG GTT AGG GTT AGG GT-FAM 30 and dsDNA:
50-GGA TGT GAG TGT GAG TGT GAG GG-FAM 30 previously annealed
to its complementary strand 50 Dabcyl CC CTC ACA CTC ACA CTC
ACA TCC-30) and increasing ligand concentrations in LiP buffer
(10 mM LiOH, 50 mM KCl, pH 7.0 with H3PO4). In a typical experi-
ment, samples were maintained at 30ꢀ for 5 min before being
slowly heated to 95 ꢀC (1 ꢀC/min), kept at this temperature for
5 min and then annealed at the same rate. Recordings were taken
during both the annealing and melting steps. Tm values were
determined from the first derivatives of the melting profiles using
the Roche LightCycler software. Each curve was repeated at least
mycin (all purchased from Biochrom, Berlin, Germany). Cells were
incubated at 37 ꢀC, 5% CO2 and maintained in exponential growth
by splitting the cultures every third day to
a density of
0.5 ꢂ 106 cells.
4.2.6. Induction of apoptosis
After treatment with increasing concentrations of 8 or 9
(0.0e6.4 M) for 24 h, cells were incubated with Guava Nexin Re-
m
agent (Merck Millipore, Hayward, CA, SA), containing 7-amino-
actinomycin D and annexin V-phycoerythrin, for 20 min at room
temperature in the dark and then analyzed via flow cytometry.
three times and errors were 0.4 ꢀC.
DTm was calculated by sub-
tracting the Tm value recorded in the presence of the ligand from
4.2.7. Caspase-3 activity
the corresponding value in the absence of ligand. DDTm represents
Activity of caspase-3 was evaluated with Caspase-3 Colorimetric
Protease Assay Kit (Thermo Fisher Scientific, Carlsbad, CA, USA),
according to manual instructions. Briefly, after 24 h treatment with
the difference of
DTm between the G-4 and dsDNA recorded at
2.5 M ligand concentration.
m
increasing concentrations of 9 (0.0e6.4
m
M), 3 ꢂ 106 cells/sample
4.2.2. Topoisomerase II
0.125 g of pBR322 (Inspiralis) were incubated with increasing
concentrations (0.5e100
M) of tested compounds for 1 h h at 37 ꢀC
in the presence/absence of 1 U of human topoisomerase II
were washed in phosphate buffer 1x (PBS, Sigma Aldrich, St. Louis,
MO, USA), re-suspended in Cell Lysis Buffer and kept on ice for
10 min. Cellular lysates were centrifuged at 10000 ꢂ g for 1 min.
The supernatant was incubated for 2 h at 37 ꢀC in the dark with 2x
m
m
a
(Inspiralis) in the required buffer. Reactions were stopped by the
addition of loading buffer (10 mM Tris, 1 mM EDTA; 200 mM NaCl,
bromophenol blue, xylene cyanol, pH 8.0) and the products were
resolved on a 1% agarose gel in 1X TAE (10 mM Tris, 1 mM EDTA,
Reaction Buffer containing DTT 10 mM and 200 mM DEVD-pNA. The
substrates are composed of the chromophore p-nitroanilide (pNA),
and the synthetic tetrapeptide, DEVD (Asp-Glue-Val-Asp). Upon
cleavage of the specific substrate by Caspase-3, free pNA light

A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
121
absorbance was quantified at 405 nm, using the microplate reader
Victor X3 (Perkin Elmer, Waltham, MA, USA), against a negative
control processed such as the samples but without the DEVD
addition to the reaction mixture. Caspase-3 activity was expressed
as fold increase of treated cells respect to untreated cells. Cells
using the NAMD 2.9 code [36]. The parm99 Amber force field [37]
including the recent parmbsc0 nucleic acids parameters were
used to treat the G4-DNA and dsDNA. The systems were placed in a
10.0 Å layer cubic water box using the TIP3P explicit water model
and Kþ ions were added to neutralize the net charge. The SHAKE
algorithm was applied to all covalent bonds involving hydrogen
atoms and 2 fs integration time step was considered. The long-
range electrostatic interactions were computed using the Particle
Mesh Ewald (PME) method with a 1.0 Å grid spacing in periodic
boundary conditions. All the receptors were thus subjected to a
double minimization step using the conjugate gradient algorithm
in the following conditions: (i) solvent and ions minimization,
keeping all the solute fixed (2000 steps); (ii) minimization of the
entire system, without any restriction (20000 steps). The mini-
mized systems were equilibrated at 298.15 K through 5 ns in NVT
ensemble followed by 5 ns of equilibration in NPT ensemble at
1 atm. Finally, a further equilibration step of 5 ns was performed in
the NVE ensemble. Production run of 10 ns was performed in NPT
conditions at 1 atm and 298.15 K while trajectory energies and
coordinates were saved every 40 ps. Trajectories clusterization have
been performed using the g-cluster tool of GROMACS ver. 4.5 [28]
and the GROMOS algorithm using a RMSD cut-off of 0.2 nm.
treated with 100.0 mM valinomycin (Sigma Aldrich) for 1 h were
used as positive control.
4.2.8. Analysis of PARP expression
The expression of the cleaved form of PARP was analyzed. After
treatment with 8 or 9 (0.0e3.2
m
M) for 24 h, 1 ꢂ 106 cells were fixed
by 4% paraformaldehyde and permeabilized by using 90% ice-cold
methanol. Cells were then incubated with the anti-cleaved PARP
antibody labeled with fluorescein isothiocyanate (1:100, Thermo
Fisher, Carlsbad, CA, USA). After washing, samples were analyzed
via flow cytometry and the mean fluorescence was recorded.
4.2.9. Cell-cycle analysis
Cells were treated for 24 h with increasing concentrations of 8 or
9 (0.0e3.2 mM) and fixed in ice-cold ethanol. After washing, sam-
ples were incubated with Cell Cycle Reagent (Merck Millipore,
Hayward, CA, SA) containing propidium iodide and incubated at
room temperature in the dark for 30 min. Cells were then analyzed
via flow cytometry.
4.3.3. Polyamine NDIs preparation
The 3D molecular structures of the most interesting 9, 11 and 16
were built using the Maestro Build Panel ver. 10.2 [33] and all the
hydrogen atoms were added. Their protonation state has been set
according with that experimentally calculated at pH 7.0 revealing
the ionization of all the primary and secondary aliphatic amine
groups of both the spermine and spermidine-like side chains.
Specifically, three amine groups were protonated for compounds 9,
four amine groups were ionized for compounds 11 and two amine
groups for 16. The so obtained NDI derivatives 9, 11 and 16 were
submitted to 3000 steps Polak-Ribiere conjugate gradient mini-
mization (PRCG) (0.05 kJ/Å mol convergence threshold) with
MacroModel ver. 10.8 [38] using the force field AMBER* and their
conformational degrees of freedom were explored using the MC
method. In particular, 10000 conformations were generated for
each ligand and the GB/SA water implicit solvent model was
considered. The total number of conformers identified for each
ligand has been reported in Table S1 of the Supporting Information.
In addition, conformers found within 3 kcal/mol above the global
minimum were subjected to BASASA (Table S1), while only the
energetically lowest conformer of each ligand (the global mini-
mum) has been taken into account for docking calculations.
4.2.10. Flow cytometry
All flow cytometric analyses were performed using Guava
EasyCyte 6 2L cytometer (Guava Technologies, Merck Millipore).
Approximately 5000 cells were evaluated for each sample.
4.2.11. Statistical analysis
Results are expressed as mean
SEM of three experiments.
Differences between treatments were assessed using one way
ANOVA and Dunnett as post-test. All statistical analyses were per-
formed using GraphPad InStat (GraphPad Prism, San Diego, CA,
USA). P < 0.05 was considered significant.
4.3. Molecular modelling
4.3.1. DNA receptors preparation and optimization
The NMR structures of the 23-nt hybrid-1 (PDB code: 2JSM) [29]
and 25-nt hybrid-2 (PDB code: 2JSL) [29] forms of the DNA G-
quadruplex (G4-DNA) with the human telomeric sequence 50-
TAG3T2AG3T2AG3T2AG3-30
and 50-TAG3T2AG3T2AG3T2AG3T2-30
respectively, have been downloaded from the Protein Data Bank
website [32]. Subsequently, the two telomeric sequences were
modified in order to achieve the same 23-nt sequence 50-
AG3T2AG3T2AG3T2AG3T-30 of G4-DNA adopted in the biophysical
studies using the graphical user interface of Maestro ver. 10.2 [33].
The so obtained hybrid-1 and hybrid-2 models were then treated
with the Protein Preparation Wizard tool, all the hydrogen atoms
added and the correct bond orders assigned. As concern the duplex
DNA (dsDNA), we built the tridimensional model having the same
experimental sequence using the graphical user interface of
Maestro ver. 10.2 [33] and then submitted to Protein Preparation
Wizard treatment [34]. Specifically, with the aim to maintain the
intercalative binding site between the GC base pair, we choose as
template structure the 6 pb crystallographic dsDNA (PDB code:
1Z3F) [35] complexed with the anticancer agent ellipticine and
superimposed it with the GC base pair of the homology dsDNA
(Fig. S1).
4.3.4. Docking setup
Molecular docking calculations against the hybrid-1 and hybrid-
2 forms of G4-DNA human telomeric sequences and the experi-
mental sequence of dsDNA were carried out using the AutoDock4.2
software package [30]. Ligands and both G4-DNA and dsDNA re-
ceptors were converted to apposite AD4 PDBQT file formats and the
GaisteigereMarsili partial charges were then assigned. In order to
explore the entire conformational space of both nucleic acids, the
3D grid box dimensions have been defined including the entire
DNA macromolecules. Specifically, grid points of 126 ꢂ 126 ꢂ 126
for the G4 DNA and of 70 ꢂ 70 ꢂ 70 for the dsDNA with a 0.375 Å
spacing were computed by mean of AutoGrid4.2. For each target,
100 docking poses were generated leaving other default docking
parameters, only the rmsd tolerance (rmstol) was increased from
the default value of 2.0 Å to 4.5 Å. Docking conformations were
clustered on the basis of their rmsd. All the obtained complexes
were subjected to a 3000 multi minimization phase using the
AMBER* force field with MacroModel ver. 10.8 [30].
4.3.2. Molecular dynamics of G4-DNA and dsDNA
The previously prepared 23-nt G4-DNA receptors and the
dsDNA have been submitted to Molecular Dynamics simulation

122
A. Milelli et al. / European Journal of Medicinal Chemistry 128 (2017) 107e122
Structural basis for telomeric G-quadruplex targeting by naphthalene diimide
5. Associated content
ligands, J. Am. Chem. Soc. 134 (2012) 2723e2731.
[16] C. Marchetti, A. Minarini, V. Tumiatti, F. Moraca, L. Parrotta, S. Alcaro, R. Rigo,
C. Sissi, M. Gunaratnam, S.A. Ohnmacht, S. Neidle, A. Milelli, Macrocyclic
naphthalene diimides as G-quadruplex binders, Bioorg Med. Chem. 23 (2015)
3819e3830.
Further Molecular Modelling analysis, complete compounds
evaluation at the NCI 60 human cancer cell lines, elemental analysis
and 1H NMR and 13C NMR spectra for selected compounds.
[17] S.L. Suraru, F. Würthner, Strategies for the synthesis of functional naphthalene
diimides, Angew. Chem. Int. Ed. Engl. 53 (2014) 7428e7448.
[18] S.A. Ohnmacht, C. Marchetti, M. Gunaratnam, R.J. Besser, S.M. Haider, G. Di
Acknowledgement
ꢁ
Vita, H.L. Lowe, M. Mellinas-Gomez, S. Diocou, M. Robson, J. Sponer, B. Islam,
R.B. Pedley, J.A. Hartley, S. Neidle,
showing anti-tumour activity in an in vivo model for pancreatic cancer, Sci.
Rep. 5 (2015) 11385.
A G-quadruplex-binding compound
This work was supported by RFO (University of Bologna) and the
Italian Ministry of Education (Funding for Investments of Base
Research) (code FIRB-IDEAS RBID082ATK_002). We thank the Na-
tional Cancer Institute (Bethesda, MD) for the anticancer assays.
G.C. was supported by the UMG grant dedicated to the memory of
Anna and Ninì Barbieri.
[19] F. Doria, M. Nadai, M. Folini, M. Di Antonio, L. Germani, C. Percivalle, C. Sissi,
N. Zaffaroni, S. Alcaro, A. Artese, S.N. Richter, M. Freccero, Hybrid ligand-
alkylating agents targeting telomeric G-quadruplex structures, Org. Biomol.
Chem. 10 (2012) 2798e2806.
[20] A. Milelli, V. Tumiatti, M. Micco, M. Rosini, G. Zuccari, L. Raffaghello, G. Bianchi,
V. Pistoia, J. Fernando Díaz, B. Pera, C. Trigili, I. Barasoain, C. Musetti,
M. Toniolo, C. Sissi, S. Alcaro, F. Moraca, M. Zini, C. Stefanelli, A. Minarini,
Structure-activity relationships of novel substituted naphthalene diimides as
anticancer agents, Eur. J. Med. Chem. 57 (2012) 417e428.
Appendix A. Supplementary data
[21] M.A. Keniry, E.A. Owen, Insight into the molecular recognition of spermine by
DNA quadruplexes from an NMR study of the association of spermine with the
thrombin-binding aptamer, J. Mol. Recognit. 26 (2013) 308e317.
[22] H. Sun, J. Xiang, Y. Liu, L. Li, Q. Li, G. Xu, Y. Tang, A stabilizing and denaturing
dual-effect for natural polyamines interacting with G-quadruplexes depend-
ing on concentration, Biochimie 93 (2011) 1351e1356.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2017.01.025.
References
ꢂ
[23] J.M. Barret, A. Kruczynski, S. Vispe, J.P. Annereau, V. Brel, Y. Guminski,
[1] S.V. Bhosale, C.H. Jani, S.J. Langford, Chemistry of naphthalene diimides, Chem.
Soc. Rev. 37 (2008) 331e342.
[2] Z. Hu, R.L. Arrowsmith, J.A. Tyson, V. Mirabello, H. Ge, I.M. Eggleston,
J.G. Delcros, A. Lansiaux, N. Guilbaud, T. Imbert, C. Bailly, F14512, a potent
antitumor agent targeting topoisomerase II vectored into cancer cells via the
polyamine transport system, Cancer Res. 68 (2008) 9845e9853.
[24] E. Bonaiuto, A. Milelli, G. Cozza, V. Tumiatti, C. Marchetti, E. Agostinelli,
C. Fimognari, P. Hrelia, A. Minarini, M.L. Di Paolo, Novel polyamine analogues:
from substrates towards potential inhibitors of monoamine oxidases, Eur. J.
Med. Chem. 70 (2013) 88e101.
[25] M. Micco, G.W. Collie, A.G. Dale, S.A. Ohnmacht, I. Pazitna, M. Gunaratnam,
A.P. Reszka, S. Neidle, Structure-based design and evaluation of naphthalene
diimide G-quadruplex ligands as telomere targeting agents in pancreatic
cancer cells, J. Med. Chem. 56 (2013) 2959e2974.
S.W. Botchway, G.D. Pantos, S.I. Pascu,
A fluorescent Arg-Gly-Asp (RGD)
peptide-naphthalenediimide (NDI) conjugate for imaging integrin
a(v)b(3)
in vitro, Chem. Commun. (Camb) 51 (2015) 6901e6904.
[3] V. Tumiatti, A. Milelli, A. Minarini, M. Rosini, M.L. Bolognesi, M. Micco,
V. Andrisano, M. Bartolini, F. Mancini, M. Recanatini, A. Cavalli, C. Melchiorre,
Structure-activity relationships of acetylcholinesterase noncovalent inhibitors
based on a polyamine backbone. 4. Further investigation on the inner spacer,
J. Med. Chem. 51 (2008) 7308e7312.
[4] E. De Gianni, E. Turrini, A. Milelli, F. Maffei, M. Carini, A. Minarini, V. Tumiatti,
T.D. Ros, M. Prato, C. Fimognari, Study of the cytotoxic effects of the new
synthetic Isothiocyanate CM9 and its fullerene derivative on human T-leu-
kemia cells, Toxins (Basel) 7 (2015) 535e552.
[26] S. Alcaro, A. Artese, G. Costa, S. Distinto, F. Ortuso, L. Parrotta, Conformational
studies and solvent-accessible surface area analysis of known selective DNA
G-Quadruplex binders, Biochimie 93 (2011) 1267e1274.
[27] K.G. Xavier Daura, Bernhard Jaun, Dieter Seebach, Wilfred F. van Gunsteren,
A.E. Mark, Peptide folding: when simulation meets experiment, Angew. Chem.
- Int. Ed. 38 (1999) 236e240.
[5] Y. Wang, X. Zhang, J. Zhao, S. Xie, C. Wang, Nonhematotoxic naphthalene
diimide modified by polyamine: synthesis and biological evaluation, J. Med.
Chem. 55 (2012) 3502e3512.
ꢂ
[28] S. Pronk, S. Pall, R. Schulz, P. Larsson, P. Bjelkmar, R. Apostolov, M.R. Shirts,
[6] V. Tumiatti, A. Milelli, A. Minarini, M. Micco, A. Gasperi Campani, L. Roncuzzi,
D. Baiocchi, J. Marinello, G. Capranico, M. Zini, C. Stefanelli, C. Melchiorre,
Design, synthesis, and biological evaluation of substituted naphthalene imides
and diimides as anticancer agent, J. Med. Chem. 52 (2009) 7873e7877.
[7] M.L. Bochman, K. Paeschke, V.A. Zakian, DNA secondary structures: stability
and function of G-quadruplex structures, Nat. Rev. Genet. 13 (2012) 770e780.
[8] L.H. Hurley, DNA and its associated processes as targets for cancer therapy,
Nat. Rev. Cancer 2 (2002) 188e200.
[9] S. Balasubramanian, S. Neidle, G-quadruplex nucleic acids as therapeutic
targets, Curr. Opin. Chem. Biol. 13 (2009) 345e353.
[10] S. Balasubramanian, L.H. Hurley, S. Neidle, Targeting G-quadruplexes in gene
promoters: a novel anticancer strategy? Nat. Rev. Drug Discov. 10 (2011)
261e275.
J.C. Smith, P.M. Kasson, D. van der Spoel, B. Hess, E. Lindahl, GROMACS 4.5: a
high-throughput and highly parallel open source molecular simulation tool-
kit, Bioinformatics 29 (2013) 845e854.
[29] A.T. Phan, V. Kuryavyi, K.N. Luu, D.J. Patel, Structure of two intramolecular G-
quadruplexes formed by natural human telomere sequences in Kþ solution,
Nucleic Acids Res. 35 (2007) 6517e6525.
[30] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,
A.J. Olson, AutoDock4 and AutoDockTools4: automated docking with selective
receptor flexibility, J. Comput. Chem. 30 (2009) 2785e2791.
€
[31] V. MacroModel, Schrodinger, LLC, New, N. York, 2015.
[32] Protein Data Bank website. http://www.rcsb.org/.
€
[33] Maestro, Version 10.2, Schrodinger, LLC, New York, NY, 2015.
€
[34] Schrodinger Suite 2015-2 Protein Preparation Wizard; Epik version3.2,
[11] G. Biffi, D. Tannahill, J. Miller, W.J. Howat, S. Balasubramanian, Elevated levels
of G-quadruplex formation in human stomach and liver cancer tissues, PLoS
One 9 (2014) e102711.
[12] S.A. Ohnmacht, S. Neidle, Small-molecule quadruplex-targeted drug discov-
ery, Bioorg Med. Chem. Lett. 24 (2014) 2602e2612.
€
€
Schrodinger, LLC, New York, NY, 2015; Impact version 6.7, Schrodinger,LLC,
€
New York, NY, 2015; Prime version 4.0, Schrodinger, LLC, New York, NY, 2015.
[35] A. Canals, M. Purciolas, J. Aymamí, M. Coll, The anticancer agent ellipticine
unwinds DNA by intercalative binding in an orientation parallel to base pairs,
Acta Crystallogr. D. Biol. Crystallogr. 61 (2005) 1009e1012.
[13] F. Cuenca, O. Greciano, M. Gunaratnam, S. Haider, D. Munnur, R. Nanjunda,
W.D. Wilson, S. Neidle, Tri- and tetra-substituted naphthalene diimides as
potent G-quadruplex ligands, Bioorg Med. Chem. Lett. 18 (2008) 1668e1673.
[14] M. Gunaratnam, S. Swank, S.M. Haider, K. Galesa, A.P. Reszka, M. Beltran,
F. Cuenca, J.A. Fletcher, S. Neidle, Targeting human gastrointestinal stromal
tumor cells with a quadruplex-binding small molecule, J. Med. Chem. 52
(2009) 3774e3783.
[36] J.C. Phillips, R. Braun, W. Wang, J. Gumbart, E. Tajkhorshid, E. Villa, C. Chipot,
ꢂ
R.D. Skeel, L. Kale, K. Schulten, Scalable molecular dynamics with NAMD,
J. Comput. Chem. 26 (2005) 1781e1802.
ꢂ
ꢂ
[37] A. Perez, I. Marchan, D. Svozil, J. Sponer, T.E. Cheatham, C.A. Laughton,
M. Orozco, Refinement of the AMBER force field for nucleic acids: improving
the description of alpha/gamma conformers, Biophys. J. 92 (2007) 3817e3829.
€
[38] MacroModel, Version 10.8, Schrodinger, LLC, New York, NY, 2015.
[15] G.W. Collie, R. Promontorio, S.M. Hampel, M. Micco, S. Neidle, G.N. Parkinson,