Imidazopyridinium cations: A new family of azonia aromatic heterocycles with applications as DNA intercalators

Article abstract of DOI:10.1016/j.dyepig.2016.11.041

Two novel imidazopyridinium cations formed by a hexacyclic azonia aromatic system have been synthesized. Spectrofluorimetric titrations, circular dichroism measurements, theoretical simulations and fluorescence-based thermal denaturation experiments on these materials have shown the interesting fluorescence properties and DNA-binding ability by intercalation, with a marked preference for AT-rich sequences. Compound 2 presents the highest fluorescence quantum yield (0.32 in 5% DMSO/water and 0.46 in MeOH) and affinity for DNA (binding constant of ～4.5 × 10⁵ M^?1). Moreover, the potential of these compounds for cell staining has been investigated in living HeLa cells by confocal microscopy imaging. This analysis showed the remarkable capacity of both compounds for uptake and accumulation by living cells.

Full text of DOI:10.1016/j.dyepig.2016.11.041

Accepted Manuscript
Imidazopyridinium cations: A new family of azonia aromatic heterocycles with
applications as DNA intercalators
Pedro Bosch, Verónica García, Başak S. Bilen, David Sucunza, Alberto Domingo,
Francisco Mendicuti, Juan J. Vaquero
PII:
S0143-7208(16)30903-2
10.1016/j.dyepig.2016.11.041
DYPI 5610
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 17 October 2016
Revised Date: 21 November 2016
Accepted Date: 22 November 2016
Please cite this article as: Bosch P, García V, Bilen BS, Sucunza D, Domingo A, Mendicuti F, Vaquero
JJ, Imidazopyridinium cations: A new family of azonia aromatic heterocycles with applications as DNA
intercalators, Dyes and Pigments (2016), doi: 10.1016/j.dyepig.2016.11.041.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Imidazopyridinium cations: a new family of azonia aromatic
heterocycles with applications as DNA intercalators
Pedro Bosch,1 Verónica García,2 Başak S. Bilen,3 David Sucunza,1,* Alberto Domingo,2
Francisco Mendicuti,3 Juan J. Vaquero1,*
1Departamento de Química Inorgánica y Química Orgánica, 2Departamento de Biología de
Sistemas, 3Departamento de Química Analítica, Química Física e Ingeniería Química,
Universidad de Alcalá, 28871- Alcalá de Henares, Madrid, Spain.
david.sucunza@uah.es, juanjose.vaquero@uah.es
ABSTRACT: Two novel imidazopyridinium cations formed by a hexacyclic azonia aromatic
system have been synthesized. Spectrofluorimetric titrations, circular dichroism measurements,
theoretical simulations and fluorescence-based thermal denaturation experiments on these
materials have shown the interesting fluorescence properties and DNA-binding ability by
intercalation, with a marked preference for AT-rich sequences. Compound 2 presents the highest
fluorescence quantum yield (0.32 in 5% DMSO/water and 0.46 in MeOH) and affinity for DNA
(binding constant of 4.5 × 105 M–1). Moreover, the potential of these compounds for cell
staining has been investigated in living HeLa cells by confocal microscopy imaging. This
analysis showed the remarkable capacity of both compounds for uptake and accumulation by
living cells.
KEYWORDS: Imidazopyridinium cations, azonia aromatic heterocycles, DNA intercalation,
cell staining
1

ACCEPTED MANUSCRIPT
GRAPHICAL ABSTRACT:
2

ACCEPTED MANUSCRIPT
1. Introduction
Polycyclic heteroarenes are usually regarded as representative DNA intercalators, especially if
they possess electron-deficient or charged aromatic cores.1 This class of DNA-binding ligand
can affect many biological properties of DNA, including replication and transcription, and they
therefore have potential as chemotherapeutic drugs.2 The binding of these molecules to DNA is
characterized by the insertion of planar aromatic rings between the DNA base pairs, which in
turn enables the π-stacking interaction. On the other hand, a positive charge enhances the
propensity of a molecule to bind to DNA due to this association results in cation release from
the DNA backbone and the solvation of these cations in the aqueous solution contributes to the
overall binding energy. 3 In most cationic dyes, this positive charge is established by an
exocyclic ammonium functionality or by an endocyclic pyridinium moiety. However, cationic
dyes with an endocyclic quaternary bridgehead nitrogen atom represent a promising platform
for the design of DNA-binding ligands.4
Coralyne is probably the most widely studied example of this latter kind of polyaromatic azonia
cation. 5 Interest in this synthetic alkaloid analog is due to several significant properties
associated with its structure, a penta-substituted dibenzo[a,g]quinolizinium system, such as
cytotoxicity,6 intercalative DNA binding,7 potent topoisomerase inhibition,8 and applications as
a fluorescent probe.9 This interest has also led to a significant amount of research on the
synthesis and study of coralyne analogs and other related azonia aromatic heterocycles that
contain a quinolizinium system as the core heterocycle.10 In this sense, recent significant
contributions by Ihmels and col.,11 Becq and col.,12 and us13 have shown that these cations are a
representative family of DNA intercalators with fluorescence properties, with some examples
used as fluorescent probes.14
3

ACCEPTED MANUSCRIPT
Figure 1. Structure of coralyne (A), the general structure of quinolizinium-based azonia cations (B) and
imidazopyridinium cations 1 and 2.
Less attention has been paid to other polyaromatic azonia cations that contain a quaternary
bridgehead nitrogen atom in the core heterocycle. For instance, imidazopyridinium derivatives
constitute an interesting family of cations whose DNA-binding capacity has hardly been
studied.15 As part of our research in this area,16 we report here the synthesis of two novel
imidazopyridinium derivatives 1 and 2, which have shown relatively high fluorescence quantum
yields and DNA-binding ability by intercalation.
Both compounds are formed by a hexacyclic azonia system. Although in the literature a fused
bicyclic system is considered to be the minimum size for this type of compound,17 results
obtained previously by our group show that a minimum number of four fused rings is needed to
produce an interaction with DNA in these systems with a quaternary bridgehead nitrogen
atom.16,18 This observation can be explained by an enhanced interaction between DNA base
pairs and the π system, so the extent of the π system is a key parameter that determines the
strength of the DNA association. The systems differ in the presence of an additional nitrogen in
the latter compound, a change that should provide further information about the effect of
replacing a methylene group next to the quaternary nitrogen atom of quinolizinium with another
4

ACCEPTED MANUSCRIPT
nitrogen. Similar binding constants have usually been found in compounds with this structural
difference.7
Moreover, the capacity for cell staining of compound 2 was investigated in living HeLa cells by
confocal microscopy imaging. The results of this analysis show a remarkable capacity of this
cation for live-cell uptake and staining of interphase nuclei and metaphase chromosomes.
2. Methods, materials and protocols
2.1. Synthetic procedures
2.1.1. General information. Reagents of the highest commercial quality were purchased and
used without further purification, unless stated otherwise. Reactions were monitored by thin-
layer chromatography (TLC) carried out on 0.25 mm E. Merck silica gel plates (60FS-254)
using UV light to visualize the compounds. Column chromatography was performed using silica
gel (60 F254, 70–200 mm) as the stationary phase. All melting points were determined in open
capillary tubes on a Stuart Scientific SMP3 melting point apparatus and are uncorrected. IR
1
13
spectra were obtained on a Perkin-Elmer FTIR spectrum 2000 spectrophotometer. H and C
NMR spectra were recorded with either Varian Mercury VX-300, Varian Unity 300 or 500
MHz spectrometers at room temperature. Chemical shifts are given in ppm (δ) downfield from
TMS. Coupling constants (J) are in hertz (Hz) and signals are described as follows: s, singlet; d,
doublet; t, triplet; br, broad; m, multiplet; ap, apparent etc. High-resolution analysis (TOF) was
performed on an Agilent 6210 time-of-flight LC/MS. 3-Methylimidazo[1,5-a]pyridine (3)19 and
20
O-mesytilenesulfonylhydroxylamine (MSH)
were prepared following the procedures
described in the literature.
2.1.2. 2-(2-Ethoxy-2-oxoethyl)-3-methylimidazo[1,5-a]pyridin-2-ium bromide (4)
Ethyl bromoacetate (0.081 mL, 0.75 mmol) was added to a solution of 3 (88.1 mg, 0.67 mmol)
in dry acetone (2 mL). The reaction mixture was stirred and heated under reflux for 4 h. The
reaction mixture was cooled down to room temperature and the product was filtered off to give
5

ACCEPTED MANUSCRIPT
a pale brown solid. This solid was washed several times with acetone and Et2O to give the
desired product in 60% yield (120 mg). M.p: 212–214 ºC. IR (KBr) ν cm–1: 3106, 2907, 1737,
1
1655, 1226, 1026, 748, 439. H NMR (300 MHz, DMSO d6) δ 8.48 (dd, J = 7.1, 1,1 Hz, 1H),
8.17 (d, J = 0.9 Hz, 1H), 7.86 (dd, J = 9.0, 1.1 Hz, 1H), 7.35–7.09 (m, 2H), 5.58 (s, 2H), 4.22 (q,
J = 7.1 Hz, 2H), 3.32 (s, 3H), 1.25 (t, J = 7.1 Hz, 3H). 13C NMR (75 MHz, DMSO d6) δ 165.8,
135.3, 127.5, 123.5, 122.5, 117.8, 116.5, 113.3, 61.7, 49.0, 13.4, 8.6. HRMS (ESI-TOF) for
+
C12H15N2O2 m/z (Calc.): 219.1125, m/z (Found): 219.1127.
2.1.3. 2-Amino-3-methylimidazo[1,5-a]pyridin-2-ium 2,4,6-trimethylbenzenesulfonate (5)
A solution of MSH (242 mg, 1.12 mmol) in dry CH2Cl2 (4 mL) was added dropwise to a
solution of 3 (99 mg, 0.74 mmol) in dry CH2Cl2 (4 mL). The reaction mixture was stirred at
room temperature for 5 h. The solid was filtered off and washed with CH2Cl2 and petroleum
ether to give compound 5 in 67% yield (172 mg). M.p: 241–244 ºC. IR (KBr) ν cm–1: 3254,
1
3053, 2926, 1646, 1559, 1379, 1081, 860, 679. H NMR (300 MHz, DMSO d6) δ 8.37 (d, J =
7.1 Hz, 1H), 8.04 (s, 1H), 7.72 (d, J = 9.0 Hz, 1H), 7.30–7.15 (m, 3H), 7.09 (t, J = 6.2 Hz, 1H),
13
6.70 (s, 2H), 2.80 (s, 3H), 2.47 (s, 6H), 2.14 (s, 3H). C NMR (75 MHz, DMSO d6) δ 142.1,
135.7, 135.3, 132.9, 129.3, 125.9, 123.1, 122.4, 117.2, 115.6, 112.0, 22.2 (2C), 19.8, 7.8.
+
HRMS (ESI-TOF) for C8H10N3 m/z (Calc.): 148.0857, m/z (found): 148.0859.
2.1.4. Dibenzo[f,h]pyrido[1',2':3,4]imidazo[1,2-b]isoquinolin-16-ium bromide (1)
9,10-Phenanthrenequinone (84 mg, 0.4 mmol) and Et3N (0.055 mL, 0.4 mmol) were added to a
suspension of 4 (120 mg, 0.4 mmol) in EtOH (5 mL). The reaction mixture was stirred and
heated under reflux for 24 h. The solvent was removed in vacuo to give a residue, which was
triturated with Et2O. The residue was washed with Et2O and cold acetone. The product was
purified by flash column chromatography (98:2 CH2Cl2/MeOH) on alumina to give the target
compound as a yellow solid in 30% yield (47.5 mg). M.p: 280 ºC (decomposition). IR (KBr) ν
cm–1: 3397, 3026, 1654, 1534, 1427, 1220, 750, 717. 1H NMR (500 MHz, DMSO d6) δ 10.51 (s,
1H), 9.92 (s, 1H), 9.48 (d, J = 7.0 Hz, 1H), 8.82 (d, J = 8.4 Hz, 1H), 8.71–8.48 (m, 4H), 8.01 (d,
6

ACCEPTED MANUSCRIPT
J = 9.3 Hz, 1H), 7.75–7.51 (m, 5H), 7.25 (t, J = 6.9 Hz, 1H). 13C NMR (126 MHz, DMSO d6) δ
152.2, 146.6, 142.1, 138.8, 135.9, 135.5, 133.9, 133.3, 132.9, 132.1, 131.6, 129.4, 129.2, 128.9,
128.8, 126.5, 126.3, 125.8, 125.1, 124.5, 124.3, 123.9, 123.7, 122.3, 119.7, 22.3 (2C), 19.9.
+
HRMS (ESI-TOF) for C23H15N2 m/z (Calc.): 319.1230, m/z (Found): 319.1232.
2.1.5.
Dibenzo[f,h]pyrido[1',2':3,4]imidazo[1,2-b]cinnolin-10-ium
2,4,6-
trimethylbenzenesulfonate (2)
9,10-Phenanthrenequinone (43.2 mg, 0.21 mmol) and sodium acetate (17 mg, 0.21 mmol) were
added to a suspension of 5 (60 mg, 0.17 mmol) in EtOH (4 mL). The reaction mixture was
stirred and heated under reflux for 24 h. The reaction mixture was cooled down to room
temperature. The solvent was removed in vacuo to give a residue, which was triturated with
cold water and filtered off. The residue was washed with cold water and EtOAc several times.
The resulting solid was dissolved in the minimum amount of MeOH and cold Et2O was added
with cooling (ice bath) until a precipitate had formed. The solid was filtered off to give
compound 2 in 50% yield (44.2 mg). M.p: 177–179 ºC. IR (KBr) ν cm–1: 3044, 1648, 1602,
1
1519, 1446, 1318, 1187, 1085, 758, 677. H NMR (500 MHz, DMSO d6) δ 10.52 (s, 1H), 9.53
(dd, J = 6.9, 0.7 Hz, 1H), 9.34 (s, 1H), 8.95 (dd, J = 8.1, 1.2 Hz, 1H), 8.81 (d, J = 8.5 Hz, 1H),
8.72–8.70 (m, Hz, 2H), 8.19 (d, J = 9.2 Hz, 1H), 7.91–7.87 (m, 1H), 7.84–7.77 (m, 4H), 7.58 (t,
13
J = 6.9 Hz, 1H), 6.70 (s, 2H), 2.47 (s, 6H), 2.13 (s, 3H). C NMR (126 MHz, DMSO d6) δ
146.3, 142.3, 135.6, 135.3, 131.9, 131.8, 131.7, 130.4, 129.9, 129.2, 129.0, 128.4, 128.3, 127.7,
125.4, 124.9, 124.7, 124.5, 124.3, 123.7, 123.4, 121.5, 117.2, 115.7, 113.8, 108.6, 22.2 (2C),
+
19.7. HRMS (ESI-TOF) for C22H14N3 m/z (Calc.): 320.1182, m/z (Found): 320.1159.
2.2. Ligand-DNA binding
2.2.1. Instrumentation. Absorption spectra were recorded on a UV-Vis Uvikon 941 (Kontron
Instruments) spectrophotometer. Steady-state fluorescence measurements were carried out using
a PTI Quanta Master spectrofluorimeter equipped with a Xenon flash lamp as a light source,
single concave grating monochromators and Glan–Thompson polarizers in the excitation and
7

ACCEPTED MANUSCRIPT
emission paths. Detection was performed with a photomultiplier cooled by a Peltier system. Slit
widths were selected at 6 nm for both excitation and emission paths and polarizers were fixed at
the ‘magic angle’ condition. Fluorescence decay measurements were performed on a time-
correlated single-photon-counting FL900 Edinburgh Instruments Spectrometer.
A
monochromatic NanoLed (Horiba) emitting at 335 nm was used as the excitation source.
Double concave grating monochromators were employed at both excitation and emission paths.
Photons were detected by a red-sensitive cooled photomultiplier (by a Peltier system). The data
acquisition was carried out by using 1024 channels of a multichannel analyzer with a time
window width of 200 ns. A total of 10,000 counts in the maximum peak channel were taken for
each measurement. The instrumental response function was regularly achieved by measuring the
scattering of a Ludox solution. Intensity fluorescence profiles were fitted to multi-exponential
decay functions by using the iterative deconvolution method.21 Right angle geometry and
rectangular 10 mm path cells were used for all the absorption and fluorescence measurements.
Circular dichroism spectra at 25 ºC were obtained on a JASCO-715 spectropolarimeter.
Recorded spectra were the average of 3 scans taken at a speed of 50 nm·min–1 with a 0.125 s
time response. The sensitivity and resolution were fixed at 20 mdeg and 0.5 nm, respectively.
Measurements were performed on samples in a 0.1 cm (1 cm) path quartz cell while monitoring
the CD (ICD) spectra in the region where DNA (ligand) mainly absorbs.
2.2.2. Thermodynamics. The association constant K of a fluorescent ligand (L) to n effective
binding sites of large molecule (M) under the assumption that these n binding sites of M per
ligand (L) are independent and equivalent and can be related to the fluorescence intensity (I) by
the following equation:
2
1+ n K[L]
(
+
K[M]0
−
1+ nK[L] + K[M] - 4 nK2 [L]0[M]0
)
(
)
0
0
0
(1)
I
= I + I − I ×
0 ∞ 0
(
)
2K n[L]0
where [M]0 and [L]0 are the initial concentrations of M and L.
8

ACCEPTED MANUSCRIPT
The normalized difference between the fluorescence intensity (I) for L in the presence of M and
in its absence (I0), can by written as:
2
1/ [L] + n K
+
K R
−
1/ [L] + nK + K.R
−
4 nK2 R






(
)
(
)
∆I
I∞ − I0
0
0
(2)
=
×
I0
I0
2K n
where R is the [M]0/[L]0 molar ratio. As M corresponds to DNA, [M]0 is the DNA concentration
per mol of base pairs and n is the number of base pairs per bound ligand molecule. Equations
are derived in the ESI.
2.2.3. Modeling. The minimum binding energy (MBE) structure for the L/DNA complex was
obtained by sequentially approaching each ligand (depicted in Figure 1) to a DNA fragment
containing twelve CGCGAATTCGCG sequence nucleotides (percent of the ctDNA bases) by the
major groove, with their center of mass on the y axis in a coordinate system as depicted in Figure 2.
The possibility of binding to grooves was not considered as a result of the ICD spectra intensity and
sign obtained. Optimized MBE structures for the 1/ and 2/DNA complexes were used as starting
conformations for Molecular Dynamics (MD) simulations. Histories of several parameters related
to the structure of the complexes were obtained from the analysis of MD trajectories. Calculations
were performed with Sybyl X-2.022 and the Tripos Force Field.23 Further details and protocols for
these calculations are described in the supporting information and elsewhere.13c,16,24
9

ACCEPTED MANUSCRIPT
Figure 2. Cartesian system used to define the L-to-DNA approach. The pair of central AT and TA bases were
almost parallel to the xy ligand plane. The most favorable L orientation is depicted.
2.2.4. Materials for titrations. Stock concentrated solutions of ligands 1 and 2 were dissolved in
DMSO (Aldrich, HPLC grade) and diluted with a phosphate-buffered saline solution (PBS) at pH =
7.4 to reach concentrations in the10–5 molar concentration range. The final solutions contained 5%
DMSO. A 2 mg/mL calf thymus DNA (ctDNA, high molecular weight sodium salt, Aldrich)
solution was prepared in Milli-Q water. For titration purposes a ctDNA of 2.36 × 10–3 mol/per base
pair stock solution was prepared in the buffer from the previous solution. This concentration (per
pair of bases) was determined by UV-Vis (ε260nm = 13,200 M–1cm–1).25 For titrations, aliquots of this
stock DNA solution were added to a 10 mm quartz cuvette containing 2.5 mL of the above ligand
solution. After each addition the content was stirred for approximately 10 minutes. The titration
reached 1/ and 2/DNA base pair molar ratios of approximately 1:7 and 1:3, respectively. The effect
of the ligand dilution never represented more than 5% error on the ligand concentration.
2.3. Thermal denaturation experiments
Melting curve studies were carried out following previously described procedures.13c,16, 26
Synthetic oligodeoxynucleotides with one strand 5′-end labeled with the fluorophore 6-
carboxyfluorescein (6-FAM, F) and the complementary strand 3′-end labeled with the
fluorescence quencher tetramethylrhodamine (TAMRA, Q) were synthesized by Bonsai
Technologies Group S.A. (Avda. Valdelaparra 27, 28108 Alcobendas, Madrid, Spain) on either
the 0.2 or 1 µmol scale and were purified by high-performance liquid chromatography to give a
single peak. Annealing of each F oligonucleotide with its complementary Q oligonucleotide at a
final duplex concentration of 0.1 µM in TAE buffer (0.08 M Tris-acetate, 2 mM EDTA, pH 7.5)
was accomplished in a PTC-100 thermocycler (MJ Research, Inc. Waltham, MA) by first
heating to 95 °C for 5 minutes and then gradually cooling to 20 °C at a rate of 1 °C min–1.
Correct annealing was confirmed because when the oligonucleotides form a double-stranded
structure, F and Q are placed in proximity so that the fluorescence of F is quenched. The
melting reactions were carried out in 96-well plates loaded with the double-stranded
10

ACCEPTED MANUSCRIPT
preannealed oligonucleotides in a total volume of 20 µL of TAE buffer at a final concentration
of 0.1 µM either in the absence or in the presence of the compound under study. The
compounds were dissolved in dimethyl sulfoxide and serially diluted in the same solvent to
yield final concentrations in the range from 10–5 to 10–8 M. The mixtures were incubated in all
cases for 60 minutes at 37 °C before the melting assay was started in a 7500 Fast Real-Time
PCR System (ABI Prism, Applied Biosystems) by increasing the temperature in small steps of 1
°C min–1 up to 95 °C. Upon heating, a temperature is reached at which the double strand melts
so that groups F and Q are separated and the fluorescence increases. By recording the changes
in fluorescence at 517 nm in the FAM channel obtained as a function of temperature and
calculating the midpoint of the transition (Tm), we were able to estimate the increases in melting
temperatures (∆Tm) brought about by ligand binding at increasing ligand concentrations. The
following oligonucleotides were used: 5′-F-d(CAATTAAATATAAC) and 5′-F-
d(GCGCGGCGTCCGGGCC), together with their respective 3′-Q-labeled complementary
strands. These sequences were designed to provide nearly all of the possible subsequences of
four bases formed by A and T or G and C, respectively, that could be recognized by the
compound while at the same time minimizing the possibility of incorrect annealing. The raw
data recovered from the instrument for each plate consisted of an array of FAM fluorescence
values for each of the 96 wells at each temperature.
2.4. Live-cell studies
HeLa cells were maintained at 37 °C and 5% CO2 in Dulbecco’s modified Eagle’s medium
(DMEM) supplemented with 10% fetal calf serum and 100 units/mL penicillin and
streptomycin. For live-cell staining, cells were grown on coverslips and stained by addition of
the compound under the conditions indicated in each case. Static images and time-lapse video
recordings were obtained using a BioRad confocal microscope. Cytotoxicity experiments were
carried out using the standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
(MTT) assay.27 The cells were seeded into a 24-well plate at a density of 1.5 × 104 cells per well
and incubated in medium containing the compound under study at different concentrations for
11

ACCEPTED MANUSCRIPT
24 or 48 hours. To each well, 100 µL of MTT was added and the plates were incubated at 37 °C
for 4 hours to allow MTT to form formazan crystals on reacting with metabolically active cells.
The medium with MTT was removed from the wells. Intracellular formazan crystals were
dissolved by adding 100 µL of DMSO to each well and the plates were shaken for 10 minutes.
The absorbance was recorded using a multiwell plate reader.
3. Results and discussion
3.1. Synthesis of cations 1 and 2
Compounds 1 and 2 were readily obtained in three steps from commercially available 2-
aminomethylpyridine. In this way, this starting material was treated with acetic anhydride and
p-toluensulfonic acid to give the methylimidazopyridine 3,19 which was aminated using O-
(mesitylsulfonyl)hydroxylamine (MSH)20 to give derivative 4 or alkylated using ethyl
bromoacetate to give derivative 5. Finally, 4 and 5 were reacted with 9,10-phenanthrenequinone
and base in a Westphal condensation to give the desired chromophores 1 and 2. It is worth
noting that the Westphal reaction of 4 produced product 1 with loss of the ester group as a result
of in situ hydrolysis and decarboxylation.18a
Scheme 1. Synthesis of cations 1 and 2
12

ACCEPTED MANUSCRIPT
3.2. Ligand-DNA binding
3.2.1 Absorption and emission spectra. The values for the absorption maxima and molar
absorptivities at the excitation wavelength (λex) used for monitoring the emission spectra of 1 and 2
in a 5% DMSO/water mixture and in methanol are provided in Table 1. The absorption spectra for
dilute solutions of ligands 1 and 2 in a 5% DMSO/water mixture at 25 ºC are shown in Figure 1S
(ESI). The spectra in Figure 3 represented by dashed lines (also in Figure 2S of ESI) correspond
to ligands 1 and 2 in very dilute (3.0 × 10–5 and 2.31 × 10–5 mol/L concentrations, respectively)
buffer (5% DMSO) solutions with λex of 440 and 470 nm, respectively. The spectrum of 1
contains a band at 515 nm and this is accompanied by a shoulder at 488 nm. In the case of 2,
however, a single band centered at 538 nm can be observed. The fluorescence quantum yields
are summarized in Table 1 and these were obtained using a Rhodamine 101 (inner salt, Sigma-
Aldrich, for fluorescence) solution in ethanol as standard for both compounds in 5% DMSO/water
and methanol at 25 ºC.28
[DNA] = 0
[DNA] = 0.142 mM
[DNA] = 0
[DNA] = 0.054 mM
500
550
600
650
500
550
600
650
700
λ
(nm)
λ
(nm)
Figure 3. Emission spectra at 25 ºC for ligands 1 (left) and 2 (right) in buffer solutions ([L] = 3.0 and 2.31
–5
× 10 mol/L, respectively) and upon the addition of aliquots of DNA.
Hardly any shift of the emission maxima was observed upon DNA addition during the titration
experiments for any of the ligands. However, as shown in Figure 3, significant changes in the
fluorescence intensity were observed. A fluorescence enhancement was observed during the
13

ACCEPTED MANUSCRIPT
titration of 1 whereas intensity quenching was observed for 2. Furthermore, the spectrum of 1
exhibited a new band at 482 nm, the intensity of which increased upon addition of DNA. A
small shoulder was also evident at 560 nm, but this also appeared in the spectrum of 2.
Excitation spectra (Figure 3S, ESI) recorded at the maximum emission intensities for 1 and 2
exhibited features similar to those in the corresponding absorption spectra; an increase and a
decrease in the intensity of the spectra were observed upon DNA addition for 1 and 2,
respectively.
The observations outlined above indicate a binding interaction for both ligands with DNA. The
normalized variation of the fluorescence intensity of 1 and 2, as (I – I0)/I0 (I is the integral of the
emission band and the subscript zero signifies in the absence of DNA), versus DNA/L molar
ratios is depicted in Figure 4. Similar representations obtained from the intensities measured at the
maximum of the low energy bands from the excitation spectra of 1 (440 nm) and 2 (460 nm) are
depicted in the supporting information (Figure 3S). Figure 4S in the supporting information also
shows the variation of the fluorescence anisotropy r (eq 10S, ESI), monitored at the emission
maximum upon excitation at the same wavelength where the emission spectra were collected,
versus molar ratios at 25 ºC.
1.6
1.4
1.2
∆I/I
1.0
0.8
0.6
0.4
0.2
0.0
0
1
2
3
4
5
6
7
molar ratio
14

ACCEPTED MANUSCRIPT
Figure 4. Normalized variation of the fluorescence intensity for ligands 1 (ꢀ) and 2 () in 5% DMSO
and 95% buffer solutions ([1] = 3.0 × 10–5 and [2] = 2.31 × 10–5 mol/L) at 25 ºC versus DNA/ligand molar
ratios (R) during titrations.
The r values for solutions of 1 and 2 at 25 ºC in the absence of DNA denote different
orientations of the absorption and emission transition moments, molecule sizes and/or rotational
diffusion in a low viscosity aqueous solvent at 25 ºC. The r values for 1 and 2 increased
significantly with [DNA], which is a typical trend for the formation of 1:DNA and 2:DNA
complexes whose rotational relaxation times are larger when compared to those of the free
ligands 1 and 2.29
The curves depicted in Figure 4 for titrations of 1 and 2 are the result of the fitting of the
experimental data to equation 2. This provides a value for the binding constant of ca. (6.7±3.6) ×
104 M–1 for 1. This value is of the same order of magnitude as those for some other derivatives of
quinolizinium DNA intercalators previously reported by us and others.7,16 The average number of
nucleic base pairs per bound ligand was 0.9±0.4. However, compound 2 exhibited an association
constant of (4.5±1.1) × 105 M–1, which is one order of magnitude larger, and a surprising n value of
0.6±0.1, which denotes strong DNA-2 interactions. The curves shown in Figure 3S in the
supporting information were obtained from excitation spectra and these reproduce the adjustment to
equation 2 by using K and n values previously obtained from analysis of the emission spectra.
Table 1. Selected photophysical data in 5% DMSO/water and in methanol (parentheses) at 25 ºC.
Fluorescence quantum yields (φ ) were obtained by using Rhodamine 101 in ethanol as a standard.28
f
ε (M–1cm–1)
φf
λem, max
Absorption
spectra
(peaks and
shoulders), λex
(nm)
Code
Structure
in 5% DMSO/water at λex
(in methanol at 458 nm)
at λex
244(p),280(p),
299(s),340(s),
365(p),385(p),
416(s),
1
2580 (3,440)
0.17 (0.16)
528
440(p),463(s)
15

ACCEPTED MANUSCRIPT
246(p),279(s),
314(p),340(s),
358(s),441(s),
470(p),485(s)
470
8940 (5,570)
0.32 (0.46)
536
2
Fluorescence intensity decay profiles for ligands 1 and 2 were fitted to double exponential
functions with lifetime components of 6.9 ns (9.4 ns) and 17.1 ns (26.1 ns) for ligand 1 (2) and a
weighted average lifetime (<τ>) of 8.9 ns (10.0 ns) (see ESI for <τ> definition).30 The presence of
DNA makes the profiles more complex (the sum of three exponential decays). In addition, a
decrease in <τ> with the DNA/L molar ratios was observed for both complexes, as shown by the
data in Table 2. This decrease is more significant for ligand 2.
Table 2. Lifetime for the ligands in 5% DMSO and 95% buffer solutions in the absence and in the presence
of DNA at different DNA/ligand molar ratios at 25 ºC. A NanoLed emitting at 335 nm was used as the
excitation source.
DNA/L ratios
DNA/L ratios
<τ>, ns
<τ>, ns
L = 1
0
L = 2
0
8.9
7.9
7.6
7.4
7.4
7.4
7.3
10.0
8.4
8.3
7.5
5.5
5.8
0.4
0.8
1.2
1.6
2.4
3.1
0.5
1.0
1.5
2.5
4.0
3.2.2 Circular dichroism measurements. The circular dichroism (CD) spectrum in the region of the
DNA absorption provides information about alterations in its secondary structure upon L to DNA
interaction.31 As can be seen in Figure 5, ctDNA shows the typical B-form CD spectrum with
positive and negative Cotton effects centered at 278 and 245 nm. Bands are attributed to base
stacking and right-handed polynucleotide helicity, respectively. Changes in both bands are
16

ACCEPTED MANUSCRIPT
usually sensitive to the mode of interaction with small molecules.32 Some authors state that while
groove binding and the electrostatic interaction of small molecules with DNA show little or no
perturbations in both bands, intercalation due to strong base stacking interactions and stable DNA
conformations usually enhances their intensities. Some other authors, however, assert that
intercalated complexes that disrupt interactions between bases should cause a decrease in the
observed band intensities.33 As depicted in Figure 5, the intensities of both bands decrease upon
increasing the concentration of 1. Compound 2, however, exhibits a decrease in the intensity of the
band located at 245 nm and an increase in the band at 278 nm.
15
10
5
15
10
5
DNA
0.4
2
1
0.4
DNA
0
0
θ
θ
Only DNA
1:20
1:10
1:5
1:2.5
only DNA
-5
-5
1-20
1-10
1-5
0.4
0.4
-10
-10
-15
1-2.5
DNA
DNA
-15
230 240 250 260 270 280 290 300
240
260
280
300
λ
(nm)
λ (nm)
Figure 5. CD spectra (ellipticity, θ) in the region of DNA absorption at 25 ºC for ligands 1 (left) and 2 (right)
in 5% DMSO and 95% buffer solutions in the presence of DNA ([DNA] = 7.20 × 10–4 M and 6.70 × 10–4 M,
respectively) at several ligand/DNA molar ratios.
However, as ligands 1 and 2 both also absorb in this region, an ICD signal from the DNA/L
complex would also appear in this zone. This is undoubtedly observed in the CD spectra for
ligand/DNA solutions at very low DNA concentrations (Figure 5S of ESI). As a consequence, we
believe that the CD spectra in this region consist of the overlapping of both CD and ICD spectra
from DNA and L complexed, respectively, which casts doubt on previous conclusions reported by
other authors.32,33 The observation of an ICD spectrum at λ ≥ 300 nm, i.e., in the region where 1
and 2 predominantly absorb and DNA does not, is a consequence of ligand complexation with
DNA. The magnitude and sign of the ICD spectrum depends on the orientation of the transition
moment inside the intercalation site and on the lateral displacements of the intercalator relative to
17

ACCEPTED MANUSCRIPT
the helix axis. An intercalator exhibits a weak negative ICD when the transition moment for the
corresponding absorption band is oriented along the long axis of the binding DNA pocket (parallel
to the bisector of the base pairs or perpendicular to the pseudo-dyad axis). However, a positive
strong ICD is expected when the transition moment is perpendicular to the long axis of the binding
pocket. The ICD signal for a groove-bound ligand is positive and one or two orders of magnitude
larger than for an intercalated ligand.7,31a,34
40
30
20
10
0
20
15
10
5
1.0
0.5
1.0
0.5
1
2
0.0
θ
0.0
0.04
0.1
0.2
0.4
-0.5
-1.0
-1.5
0.05
0.1
0.2
0.4
θ
-0.5
-1.0
300
350
400
300 350 400 450 500
only ligand
λ
, nm
λ
, nm
0
only ligand
0.05
0.1
θ
θ
0.05
0.1
-5
0.2
-10
-20
-10
-15
0.2
0.4
0.4
absorption
absorption
450 500
300
350
400
300 320 340 360 380 400 420
λ
, nm
λ
, nm
Figure 6. Ellipticity
θ for ligands 1 and 2 in 5% DMSO and 95% buffer solutions in the presence of DNA
([DNA] = 7.20 × 10–4 M and 6.70 × 10–4 M, respectively) at several ligand/DNA molar ratios at 25 ºC.
The CD spectra, represented as molar ellipticity, are superimposed. Also superimposed are the absorption
spectra for ligands 1 and 2 (---, black).
The ICD spectra for isolated 1 and 2 in solution and in the presence of DNA at different L:DNA
molar ratios, at constant [DNA], are shown in Figure 6. No signals were observed for isolated 1
and 2. However, CD spectra for 1/DNA or 2/DNA solutions at any of the ratios used exhibit Cotton
effects, albeit with relatively low intensity, with the intensities depending on the ligand
concentration. The spectra for 1/DNA and 2/DNA solutions at any of the ratios are also consistent
with the absence of exciton coupling, which excludes a ligand aggregation during association to
DNA.7,34 Molar ellipticity [
θ
] allows visualization of the ICD spectra in two different regions and
18

ACCEPTED MANUSCRIPT
these correspond to distinguishable bands that appear in the absorption spectra. For ligand 1, one of
these bands, which is negative, extends below 350 nm; the other one, which is almost zero, covers
wavelengths greater than 350 nm. However, this wavelength limit between the two bands occurs
at around 380 nm for 2. The ICD spectrum for the high energy band of 2 is also negative, whereas
the signal for the low energy bands is slightly positive. Assuming that the transition moments of the
bands that appear at lower wavelengths are almost perpendicular to the long ligand axis, the
negative sign of the ICD bands reveals an intercalation of ligands 1 and 2 with their short-
molecular axis preferentially parallel to the DNA binding pocket.7,31a,34a However, the virtual
absence or slight positive sign of the ICD signal for the low energy band (attributed to interaction
with the ligand transition moment oriented nearly along the long ligand axis) must agree with a
transition moment oriented at nearly 45º (or slightly <45º) to the base pair binding pocket. The
absence of large positive
interactions.
θ values also excludes the presence of groove binding ligand/DNA
Figure 7. Total binding energies (ꢀ) and van der Waals (ꢀ) and electrostatic () contributions as a function
of the oo' distance measured along the y coordinate for ligands 1 and 2 approaching DNA as intercalators.
The MBE structures for 1 and 2 are represented and, once minimized again, these were used as starting
conformations for the MD simulations.
19

ACCEPTED MANUSCRIPT
3.2.3 Molecular mechanics and molecular dynamics. The total energy binding energy is depicted in
Figure 7 along with the van der Waals and electrostatic contributions for 1 and 2, with the most
favorable orientation intercalating the DNA by the major groove parallel to the plane bases as a
function of the oo' distance along the y coordinate. Intercalation takes place with no significant
potential barriers. It can be seen from Figure 7 that electrostatic interactions are mostly responsible
for the stabilization. Minima binding energy (MBE) structures for 1/DNA and 2/DNA complexes
obtained by MM calculations are also depicted. Compound 1 intercalates in such a manner that its
main axis is oriented around 45º to the TA and AT central basis binding pocket and is slightly
outside the major groove face. However, the main long axis of 2 is almost parallel to the bases
pocket axis and slightly outside the minor groove face.
Analysis of the 1.0 ns MD trajectories at 300 K on the water-solvated MBE optimized structures
depicted in Figure 7 showed that ligand-DNA interactions are favorable for both intercalations
throughout the whole trajectory (Figure 6S, ESI). As with MM, the most important contributions
to stabilization are electrostatic. The van der Waals contribution represents barely 10% of the total
binding energy.
10
8
90
80
70
60
50
40
30
20
10
0
90
80
70
60
50
40
30
20
10
0
o
o
o
2
1
1
6
2
4
1
2
0
y
-2
-4
-6
-8
2
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
time (ns)
time (ns)
time (ns)
Figure 8. (Left panel) Histories for the y coordinate (bottom lines) for the center of 1 (black) and 2 (gray)
ligands (o') and the end-to-end DNA helix distance per nucleotide unit (upper red lines) for 1/DNA (red) and
2/DNA (light red) complexes obtained from the analysis of the 1 ns MD trajectory at 300 K starting from the
minimized MBE structures. Dashed lines are the values for the initial structures. (Middle and right panels)
Histories for the averaged angles between the plane of both AT and TA pairs of bases and the plane of the 1
(middle) or 2 (right) ligand rings (red) and the angle between the bisector of pocket bases and the long ligand
axis (blue) from the analysis of the MD trajectories.
The histories for the y coordinates of the center of ligands are depicted in the left-hand panel of
Figure 8. The MBE structure at time zero and at the end of the trajectory for the 1/DNA complex is
represented in Figure 9. The 1 (2) intercalator tends to move early inside the base pair pocket close
20

ACCEPTED MANUSCRIPT
to the y coordinate zero and is slightly outside the major (minor) groove side. The averages of the y
coordinate throughout the MD trajectories were 0.9±2.0 and –2.3±0.9 Å, respectively. This
movement also involved some rotation in the case of 1 and a subsequent increase in the angle
between the binding pocket axis and the long axis of 1. In fact, after ∼0.45 ns 1 was oriented
almost perpendicular to the pocket with the phenanthrene group lying slightly outside the pocket by
the major groove face (see Figure 9) to decrease the angle slightly at the end of the trajectory. The
average of the binding pocket axis and the long axis angle throughout the trajectory was 64±20º,
which is reasonably consistent with the weak negative high energy band whose transition moment
was assumed to be nearly perpendicular to the long axis of 1. The average of the angle between the
binding bases pocket and the long axis of the ligand 2, however, was 8±5°, which means that it
does not reproduce the signs and intensity of the bands monitored in the ICD spectrum for 2/DNA
solutions.
Figure 9. Snapshots of the 1 ns MD trajectory at 300 K showing the location of ligand 1 at time = 0 ns and at
1 ns from the beginning of the trajectory. Detail of the location of ligand 1 at the end of the 1 ns trajectory
(right).
In addition, the angles between the AT and TA base pair planes and the plane of both ligands 1 and
2 were close to zero (averages of 12±6º and 11±7º, respectively) – as one would expect for ligands
21

ACCEPTED MANUSCRIPT
that intercalate. Another characteristic upon intercalation is the distortion of the DNA double helix
and a slight unwinding and lengthening of the DNA chains. In fact, from the initial DNA
fragment, whose end-to-end DNA distance per nucleotide unit was 4.1 Å, the intercalation led to an
increase in this value to 5.8±0.5 Å and 4.9±0.3 Å for 1/DNA and 2/DNA complexes, respectively.
The latter values correspond to the averages over the entire trajectory.
A longer trajectory of up to 3 ns complex was performed for 2/DNA. The analysis showed how
the average of the angle between the binding bases pocket and the long axis of 2 increase with
time (Figure 7S, ESI). Indeed, after 2 ns of the trajectory ligand 2 rotated slightly in such a
manner that the angle between the long axis and the binding pocket increased and even reached
values of 70º. The average of this angle during the last nanosecond of the trajectory was
40±9.0º. The structures for the 2/DNA complex at the end of the 3 ns trajectory are shown in
Figure 8S of the ESI.
3.3. Thermal denaturation experiments
Figure 10. Melting curves (left panel) and plot of the experimental ∆Tm versus ligand concentration on a
logarithmic scale (right panel) obtained for reaction mixtures of compound 2 and double-stranded DNA
oligonucleotides of different base compositions and sequences. Oligonucleotide AT (blue line).- 5'-
CAATTAAATATAAC-3' and its complementary. Oligonucleotide GC (orange line).- 5'-
GCGCGGCGTCCGGGCC-3' and its complementary. Each data point is the average of two separate
experiments. Ligand concetrations (M): 1.95 E-08, 3.91 E-08, 7.81 E-08, 1.56 E-07, 3.13 E-07, 6.25 E-
07, 1.25 E-06, 2.5 E-06, 5.0 E-06, 1.0 E-05 and 2.0 E-05.
22

ACCEPTED MANUSCRIPT
DNA binding studies on compound 2 were carried out by FRET-based thermal denaturation
experiments using custom synthetic oligonucleotides with defined AT- or GC-rich sequences
(Figure 10). The duplex containing the GC-rich sequence (GC) produced a simple monophasic
profile with a melting temperature (Tm) of 62 °C, which changed slightly and gradually upon
addition of increasing concentrations of 2. In fact, at a ligand concentration of 20 ꢀM, the ∆Tm
reached a maximum of ~10 °C. In contrast, the Tm of the AT-rich oligo, AT, rapidly increased
upon addition of 2 and the melting curve was shifted by almost 24 °C at the same concentration
of 20 ꢀM (Figure 10). This difference suggests a preference of 2 for DNA sequences that are
rich in AT base-pairs.
3.4. Live-cell studies
The toxicity of compounds 1 and 2 on living cells in culture was determined by the standard
MTT assay using the HeLa cell line after 24 and 48 hours of exposure. The apparent lethal dose
50% (LD50) is close to 1.0 × 10–5 in all cases (see ESI).
The capacity of compounds 1 and 2 for live-cell staining, uptake and intracellular distribution
was studied by confocal microscopy imaging and time-lapse video recording, using HeLa cells
grown on coverslips in Dulbecco’s Modified Eagle’s Medium supplemented with 10% fetal
bovine serum (DMEM, 10% FBS). The analysis showed the remarkable capacity of both
compounds for uptake and accumulation by living cells. Cells treated with compound 1 showed
a disruption of internal structures and a fluorescent staining compatible with apoptotic
fragmentation of the nucleus (Figure 11, upper panels). On the other hand, compound 2 showed
a clear staining pattern of interphase nuclei and metaphase chromosomes (Figure 11, lower
panels).
23

ACCEPTED MANUSCRIPT
Figure 11. Living HeLa cells in culture after 1 hour of exposure to compounds 1 and 2. Upper panels:
compound 1, 12.80 mM in DMSO, diluted 1:400 in culture medium (32 µM final). Lower panels:
compound 2, 56.70 mM in DMSO, diluted 1:400 in culture medium (142 µM final). Confocal microscopy
images (excitation argon laser 488 nm, emission filter 500–575 nm). Left panels: wide field with corner
marks indicating the area magnified on the right. Lower right panel: detail of a region with two cells in
mitosis (arrows) showing bright fluorescent chromosomes in metaphase plates stained with compound 2.
The scale bar represents 50 µm in all panels.
24

ACCEPTED MANUSCRIPT
The uptake and intracellular distribution of compound 2 appears to be a complex process and is
apparently due to active biological accumulation rather than passive diffusion and binding. The
final staining is strongly concentrated inside the nuclei. Mitotic metaphase chromosomes are
also brightly stained (Figure 11, lower panel).
The intracellular and nuclear staining patterns were complex and clearly change at different
incubation times. After experiments at fixed incubation times, it was decided to investigate the
complete dynamic process by time-lapse video recording using a confocal microscope equipped
with a thermostatic chamber. A composition made with selected frames in order to follow the
changes in a set of cells is shown in Figure 12.
Figure 12. Frames selected from a time-lapse, confocal video microscopy recording of living HeLa cells
in culture exposed to compound 2, 56.70 mM in DMSO, diluted 1:1000 in culture medium (56.7 µM
final). The complete sequence was recorded at 2 frames per minute. Excitation argon laser 488 nm,
emission filter 500–575 nm, gain settings were kept constant during the recording.
In the initial steps, the fluorescence is accumulated in disperse, vesicle-like cytoplasmic
structures, which is compatible with an active endocytic mechanism rather than passive
diffusion of the compound through the plasma membrane. A progressive increase is clearly
visible in the time-lapse video recording. The nucleus is initially not differentiated but the
fluorescence subsequently begins to increase inside. The nuclei become very brightly stained,
with a reproducible pattern of fluorescence accumulation in the periphery and defined internal
areas, apparently concentrating most of the compound within the cell. The stained nuclei
undergo dramatic structural changes. Initially they appear to collapse and acquire a round shape,
with an apparent reduction in diameter following a progressively homogeneous, compact and
bright staining. After a short time the diameter of the nuclei begins to increase and the internal
25

ACCEPTED MANUSCRIPT
fluorescence gradually fades away. At this stage the content of the nuclei appears homogeneous,
with an apparent loss of all internal structures but without signs of apoptotic fragmentation.
The same process occurs in all of the cells and follows the same sequence and timing, but it is
not synchronous in different cells. A clear example can be seen in the two cells near to the lower
edge of the frames in Figure 12. This feature indicates the participation of some active cellular
process in the uptake and accumulation rather than a simple diffusion and binding of the
compound.
Once the compound enters the nucleus, very likely through the nuclear pores, the fluorescence
pattern suggests strong binding to DNA and a progressive accumulation radiating from the entry
points until saturation of the complete content of the nucleus is achieved. The dramatic
morphological changes in the nuclei suggest a progressive disorganization of the chromatin as
compound 2 accumulates, possibly due to a massive displacement of proteins bound to DNA.
Conclusions
Photophysical and DNA binding experiments have shown that novel imidazopyridinium cations
1 and 2 exhibit interesting fluorescence properties and DNA-binding ability by intercalation.
Compound 2 presents the highest fluorescence quantum yield (0.32 in 5% DMSO/water and
0.46 in MeOH) and affinity for DNA (binding constant of 4.5 × 105 M–1), with a significant
preference for AT-rich sequences. Moreover, the uptake and intracellular distribution of these
compounds by living cells has been investigated by confocal microscopy static and time-lapse
imaging. This analysis showed disruption of internal structures and a fluorescent staining
compatible with apoptotic fragmentation of the nucleus for compound 1 and a clear staining
pattern of interphase nuclei and metaphase chromosomes for compound 2. The findings
described here with both parent heterocycles support further studies on some representative
derivatives in order to modulate the promising properties found for 1 and 2.
26

ACCEPTED MANUSCRIPT
Acknowledgments
Financial support from the Spanish Ministerio de Economía y Competitividad (projects
CTQ2014-52488-R and CTQ2015-64425-C2-1-R), Instituto de Salud Carlos III (FEDER funds,
ISCIII RETIC REDINREN RD012/20021/0014) and University of Alcalá (project
CCG2015/EXP-011) is gratefully acknowledged. P. B. also thanks the University of Alcalá for
his grant.
1 (a) Xie Y, Tam VK, Tor Y. The interactions of small molecules with DNA and RNA. In: Mayer G,
editor. The chemical biology of nucleic acids, John Wiley & Sons, Ltd, 2010, p. 115–140. (b) Hannon
MJ. Supramolecular DNA recognition. Chem Soc Rev 2007; 36: 280–295.
2 Ihmels H, Thomas L. Intercalation of organic ligands as a tool to modify the properties of DNA. In: Jin
J-I, Grote J, Editors. Materials science of DNA, CRC Press; 2011, p. 49–75.
3 Paul A, Bhattacharya S. Chemistry and biology of DNA-binding small molecules. Curr Sci 2012; 102:
212–231.
4 Granzhan A, Ihmels H. Playing around with the size and shape of quinolizinium derivatives: versatile
ligands for duplex, triplex, quadruplex and abasic site-containing DNA. Synlett 2016; 27: 1775–1793.
5 Bhadra K, Kumar GS. Therapeutic potential of nucleic acid-binding isoquinoline alkaloids: Binding
aspects and implications for drug design. Med Res Rev 2011; 31: 821–862.
6 Cho MJ, Repta AJ, Cheng CC, Zee-Cheng KY, Higuchi T, Pitman IH. Solubilization and stabilization
of the cytotoxic agent coralyne. J Pharm Sci 1975; 64: 1825–1830.
7 Ihmels H, Faulhaber K, Vedaldi D, Dall’Acqua F, Viola G. Intercalation of organic dye molecules into
double-stranded DNA. Part 2: the annelated quinolizinium ion as a structural motif in DNA intercalators.
Photochem Photobiol 2005; 81: 1107–1115.
8 Makhey D, Gatto B, Yu C, Liu A, Liu LF, LaVoie EJ. Coralyne and related compounds as mammalian
topoisomerase I and topoisomerase II poisons. Bioorg Med Chem 1996; 4: 781–791.
9
Mateos E, Cebolla VL, Membrado L, Vela J, Gálvez EM, Matt M, Cossio FP. Coralyne cation, a
fluorescent probe for general detection in planar chromatography. J Chromatogr A 2007; 1146: 251–257.
10 Sucunza D, Cuadro AM, Alvarez-Builla J, Vaquero JJ. Recent advances in the synthesis of azonia
aromatic heterocycles. J Org Chem 2016; 81: 10126–10135.
11
(a) Berdnikova DV, Aliyeu TM, Paululat T, Fedorov YV, Fedorova OA, Ihmels H.
DNA-Ligand interactions gained and lost: light-induced ligand redistribution in a supramolecular
cascade. Chem Commun 2015; 51: 4906–4909. (b) Benner K, Ihmels H, Kölsch S, Pithan PM. Targeting
abasic site-containing DNA with annelated quinolizinium derivatives: The influence of size, shape and
substituents. Org Biomol Chem 2014; 12: 1725–1734. (c) Bortolozzi R, Ihmels H, Thomas L, Tian M,
Viola G. 9-(4-Dimethylaminophenyl)benzo[b]quinolizinium: A near-infrared fluorophore for the
multicolor analysis of proteins and nucleic acids in living cells. Chem–Eur J 2013; 19: 8736–8741. (d)
Faulhaber K, Granzhan A, Ihmels H, Otto D, Thomas L, Wells S. Studies of the fluorescence light-up
effect of amino-substituted benzo[b]quinolizinium derivatives in the presence of biomacromolecules.
Photochem Photobiol Sci 2011; 10: 1535–1545. (e) Granzhan A, Ihmels H, Viola G. 9-Donor-substituted
acridizinium salts: versatile environment-sensitive fluorophores for the detection of biomacromolecules. J
Am Chem Soc 2007; 129: 1254–1267.
12 (a) Norez C, Jayle C, Becq F, Vandebrouck C. Bronchorelaxation of the human bronchi by CFTR
activators. Pulm Pharmacol Ther 2014; 27: 38–43. (b) Norez C, Bilan F, Kitzis A, Mettey Y, Becq F.
Proteasome-dependent pharmacological rescue of cystic fibrosis transmembrane conductance regulator
revealed by mutation of glycine 622. J Pharmacol Exp Ther 2008; 325: 89–99. (c) Marivingt-Mounir C,
Norez C, Dérand R, Bulteau-Pignoux L, Nguyen-Huy D, Viossat B, Morgant G, Becq F, Vierfond J-M,
Mettey Y. Synthesis, SAR, crystal structure, and biological evaluation of benzoquinoliziniums as
activators of wild-type and mutant cystic fibrosis transmembrane conductance regulator channels. J Med
Chem 2004; 47: 962–972.
13 (a) Abengozar A, Abarca B, Cuadro AM, Sucunza D, Álvarez-Builla J, Vaquero JJ. Azonia aromatic
cations by ring-closing metathesis: synthesis of azaquinolizinium cations. Eur J Org Chem 2015; 19:
4214–4223. (b) Marcelo G, Pinto S, Cañeque T, Mariz IFA, Cuadro AM, Vaquero JJ, Martinho JMG,
Maçôas EMS. Nonlinear Emission of Quinolizinium-Based Dyes with Application in Fluorescence
Lifetime Imaging. J Phys Chem A 2015; 119: 2351–2362. (c) Abarca B, Custodio R, Cuadro AM,
27

ACCEPTED MANUSCRIPT
Sucunza D, Domingo A, Mendicuti F, Álvarez-Builla J, Vaquero JJ. Efficient synthesis of an
indoloquinolizinium alkaloid selective DNA-binder by ring-closing metathesis. Org Lett 2014; 16: 3464–
3467. (d) Cañeque T, Cuadro AM, Alvarez-Builla J, Pérez-Moreno J, Clays K, Castaño O, Andrés JL,
Vaquero JJ. Novel charged NLO chromophores based on quinolizinium acceptor units. Dyes Pigments
2014; 101: 116–121. (e) Maçôas E, Marcelo G, Pinto S, Cañeque T, Cuadro AM, Vaquero JJ, Martinho
JMG. A V-shaped cationic dye for nonlinear optical bioimaging. Chem Commun 2011; 47: 7374–7376.
(f) Núñez A, Abarca B, Cuadro AM, Alvarez-Builla J, Vaquero JJ. Ring-closing metathesis approach to
heteroaromatic cations: synthesis of benzo[a]quinolizinium salts. Eur J Org Chem 2011; 1280–1290. (g)
Núñez A, Abarca B, Cuadro AM, Alvarez-Builla J, Vaquero JJ. Ring-closing metathesis reactions on
azinium salts: straightforward access to quinolizinium cations and their dihydro derivatives. J Org
Chem 2009; 74: 4166–4176.
14 (a) Granzhan A, Ihmels H, Tian M. The benzo[b]quinolizinium ion as a water-soluble platform for the
fluorimetric detection of biologically relevant analytes. Arkivoc 2015; vi: 494–523. (b) Chang L, Liu C,
He S, Lu Y, Zhang S, Zhao L, Zeng X. Novel styryldehydropyridocolinium derivative as turn-on
fluorescent probe for DNA detection. Sens Actuators B 2014; 202: 483–488.
15 (a) Juskowiak B, Galezowska E, Koczorowska N, Hermann TW. Aggregation and G-quadruplex DNA-
binding study of 6a,12a-diazadibenzo-[a,g]fluorenylium derivative. Bioorg Med Chem Lett 2004; 14:
3627–3630. (b) Feigon J, Denny WA, Leupin W, Kearns DR. Interactions of antitumor drugs with natural
DNA: lH NMR study of binding mode and kinetics. J Med Chem 1984; 27: 450–465.
16 Suárez RM, Bosch P, Sucunza D, Cuadro AM, Domingo A, Mendicuti F, Vaquero JJ. Targeting DNA
with small molecules: a comparative study of a library of azonia aromatic chromophores. Org Biomol
Chem 2015; 13: 527–538.
17 Denny WA. Acridine-4-carboxamides and the concept of minimal DNA intercalators. In: Demeunynck
M, Bailly C, Wilson WD, Editors. Small molecule DNA and RNA binders: from synthesis to nucleic acid
complexes, Wiley-VCH; 2004, p. 482–502.
18 (a) Pastor J, Siro JG, García-Navío JL, Vaquero JJ, Alvarez-Builla J, Gago F, de Pascual-Teresa B,
Pastor M, Rodrigo MM. Azino-fused benzimidazolium salts as DNA intercalating agents. 2. J Org Chem
1997; 62: 5476–5483. (b) Matia MP, García-Navío JL, Vaquero JJ, Alvarez-Builla J. Westphal
condensation of N-aminocycloiminium salts synthesis of fused pyridazinium derivatives. Liebigs Ann
Chem 1992; 777–779.
19 Prostota Y, Kachkovsky OD, Reis LV, Santos PF. New unsymmetrical squaraine dyes derived from
imidazo[1,5-a]pyridine. Dyes Pigments 2013; 96: 554–562.
20 Mendiola J, Rincón JA, Mateos C, Soriano JF, de Frutos O, Niemeier JK, Davis EM. Preparation, use,
and safety of O-mesitylenesulfonylhydroxylamine. Org Process Res Dev 2009; 13: 263–267.
21 O'Connor DV, Ware WR, Andre JC. Applications of fast Fourier-transform to deconvolution in single
photon-counting. J Phys Chem 1979; 83:1333–43.
22 Sybyl-X 2.0, Tripos International, 1699 South Hanley Rd., St. Louis, Missouri, 63144, USA.
23 Clark M, Cramer RD III, Van Opdenbosch N. Validation of the general purpose tripos 5.2 force field. J
Comput Chem 1989; 10, 982–1012.
24 (a) Gallego-Yerga L, Blanco-Fernández L, Urbiola K, Carmona T, Marcelo G, Benito JM, Mendicuti F,
Tros de Ilarduya C, Ortiz Mellet C, García Fernández JM. Host–guest-mediated DNA templation of
polycationic supramolecules for hierarchical nanocondensation and the delivery of gene material. Chem
Eur J 2015; 21:12093–12104. (b) Carmona T, Cañeque T, Custodio R, Cuadro AM, Vaquero JJ, Mendicuti F.
Cucurbit[n]urils as a potential fine-tuned instrument for modifying photophysical properties of D–π–A+–
π–D “push–pull” chromophores. Dyes Pigments 2014; 103: 106–117.
25 McConnaughie AW, Jenkins TC. Novel acridine-triazenes as prototype combilexins: synthesis, DNA
binding, and biological activity. J Med Chem 1995; 38: 3488–501.
26
(a) Leal JF, Martínez-Díez M, García-Hernández V, Moneo V, Domingo A, Bueren-Calabuig JA,
Negri A, Gago F, Guillén-Navarro MJ, Avilés P, Cuevas C, García-Fernández LF, Galmarini CM.
PM01183, a new DNA minor groove covalent binder with potent in vitro and in vivo anti-tumour activity.
Br J Pharmacol 2010; 161: 1099–1110. (b) Leal JF, García-Hernández V, Moneo V, Domingo A, Bueren-
Calabuig JA, Negri A, Gago F, Guillén-Navarro MJ, Avilés P, Cuevas C, García-Fernández LF,
Galmarini CM. Molecular pharmacology and antitumor activity of Zalypsis in several human cancer cell
lines. Biochem Pharmacol 2009; 78: 162–170. (c) Negri A, Marco E, García-Hernández V, Domingo A,
Llamas-Saiz AL, Porto-Sandá S, Riguera R, Laine W, David-Cordonnier MH, Bailly C, García-
Fernández LF, Vaquero JJ, Gago F. Antitumor activity, X-ray crystal structure, and DNA binding
properties of thiocoraline A, a natural bisintercalating thiodepsipeptide. J Med Chem 2007; 50: 3322–
3333. (d) Reddy PM, Dexter R, Bruice TC. DNA sequence recognition in the minor groove by hairpin
pyrrole polyamide-Hoechst 33258 analogue conjugate. Bioorg Med Chem Lett 2004; 14: 3803–3807. (e)
28

ACCEPTED MANUSCRIPT
James PL, Le Strat L, Ellervik U, Bratwall C, Nordén B, Brown T, Fox KR. Effects of a hairpin
polyamide on DNA melting: comparison with distamycin and Hoechst 33258. Biophys Chem 2004; 111:
205–212. (f) Darby RAJ, Sollogoub M, McKeen C, Brown L, Risitano A, Brown N, Barton C, Brown T,
Fox KR. High throughput measurement of duplex, triplex and quadruplex melting curves using molecular
beacons and a LightCycler. Nucleic Acids Research. 2002; 30: e39.
27 Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and
cytotoxicity assays. J Immunol Methods 1983; 65: 55–63.
28 (a) Brouwer AM. Standards for photoluminescence quantum yield measurements in solution (IUPAC
Technical Report). Pure Appl Chem 2011; 83: 2213–2228. (b) F. Mendicuti F, Mattice WL. Suppression of
the influence of stray light on quantitative measurement of ID/IM for weakly emitting dilute solutions of
polymers. Polym Bull 1989; 22: 557–63.
29 Lakowicz JR. Principles of fluorescence spectroscopy. 3rd ed. New York: Springer; 2006, Ch 10, p.
361.
30 Lakowicz, JR. Principles of fluorescence spectroscopy. 3rd ed. New York: Springer; 2006; Ch 4, p. 97.
31
(a) Ardhammar M, Nordén B, Kurucsev T. DNA-drug interactions. In: Berova N, Nakanishi K, Woody
RW, editors. Circular dichroism: principles and applications, Weinheim: Wiley-VCH; 2000, p. 741–768. (b)
Johnson WC, CD of Nucleic Acids. In: Berovan N, Nakanishi K, Woody RW, editors. Circular dichroism:
principles and applications, Weinheim: Wiley-VCH; 2000, p 703–718.
32 (a) Dehkordi MN, Bordbar AK, Lincoln P, Mirkhani V. Spectroscopic study on the interaction of ct-
DNA with manganese Salen complex containing triphenyl phosphonium groups. Spectrochim Acta part A
2012; 90: 50–54. (b) Triantafillidi K, Karidi K, Novakova O, Malina J, Garoufis A. DNA binding selectivity
of oligopyridine-ruthenium(II)-lysine conjugate. Dalton Trans 2011; 40: 472–483. (c) Kypr J, Kejnovska I,
Renciuk D, Vorlickova M. Circular dichroism and conformational polymorphism of DNA. Nucleic Acids
Res 2009; 37: 1713–1725. (d) Karidi K, Garoufis A, Hadjiliadis N, Reedijk J. Solid-phase synthesis,
characterization and DNA binding properties of the first chloro(polypyridyl)ruthenium conjugated
peptide complex. Dalton Trans 2005; 728 –734.
33 Zhang S, Zhu Y, Tu C, Wei H, Yang Z, Lin L, Ding J, Zhang J, Guo Z. A novel cytotoxic ternary
copper(II) complex of 1,10-phenanthroline and L-threonine with DNA nuclease activity. J Inorg Biochem
2004; 98: 2099–2106.
34
(a) Tian M, Ihmels H, Ye S. Fluorimetric detection of Mg2+ and DNA with 9-
(alkoxyphenyl)benzo[b]quinolizinium derivatives. Org Biomol Chem 2012; 10: 3010–3018. (b) Basili S,
Bergen A, Dall’Acqua F, Faccio A, Granzhan A, Ihmels H, Moro S, Viola G. Relationship between the
structure and the DNA binding properties of diazoniapolycyclic duplex- and triplex-DNA binders:
efficiency, selectivity, and binding mode. Biochemistry 2007; 46: 12721–12736. (c) Nordén B, Kurucsev T.
Analysing DNA complexes by circular and linear dichroism. J Mol Recognit 1994; 7: 141–156.
29