Synthesis and photophysical evaluation of a pyridinium 4-amino-1,8- naphthalimide derivative that upon intercalation displays preference for AT-rich double-stranded DNA

Article abstract of DOI:10.1039/c2ob06898b

The synthesis, characterisation and solid state crystal structure of a cationic 4-amino-1,8-naphthalimide derivative (1) are described. The photophysical properties of 1 are shown to vary with the solvent polarity and H-bonding ability. The fluorescence of 1 is enhanced and blue-shifted in its 1:1 complex with 5′-adenosine-monophosphate while it is partially quenched and red-shifted in its complex with 5′-guanosine-monophosphate. Linear and circular dichroism measurements show that 1 binds to double-stranded DNA by intercalation. Comparative UV-visible and fluorescence studies with double stranded synthetic polynucleotides poly(dA-dT)₂ and poly(dG-dC) ₂ show that 1 binds much more strongly to the AT polymer; 1 also has a strong preference for A-T rich sequences in natural DNA. Thermal denaturation measurements also reveal a much greater stabilisation of the double-stranded poly(dA-dT)₂ than of natural DNA. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2ob06898b

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 3033
www.rsc.org/obc
PAPER
Synthesis and photophysical evaluation of a pyridinium
4-amino-1,8-naphthalimide derivative that upon intercalation displays
preference for AT-rich double-stranded DNA†
Swagata Banerjee, Jonathan A. Kitchen, Thorfinnur Gunnlaugsson* and John M. Kelly*
Received 10th November 2011, Accepted 26th January 2012
DOI: 10.1039/c2ob06898b
The synthesis, characterisation and solid state crystal structure of a cationic 4-amino-1,8-naphthalimide
derivative (1) are described. The photophysical properties of 1 are shown to vary with the solvent polarity
and H-bonding ability. The fluorescence of 1 is enhanced and blue-shifted in its 1 : 1 complex with
5′-adenosine-monophosphate while it is partially quenched and red-shifted in its complex with
5′-guanosine-monophosphate. Linear and circular dichroism measurements show that 1 binds to
double-stranded DNA by intercalation. Comparative UV-visible and fluorescence studies with double
stranded synthetic polynucleotides poly(dA–dT)2 and poly(dG–dC)2 show that 1 binds much more
strongly to the AT polymer; 1 also has a strong preference for A–T rich sequences in natural DNA.
Thermal denaturation measurements also reveal a much greater stabilisation of the double-stranded poly
(dA–dT)2 than of natural DNA.
flash photolysis could be employed to observe the formation of
naphthalimide radical anion (NI ) in the presence of DNA.
Introduction
−
˙
The development of small molecules capable of binding to DNA
and exhibiting anticancer activities has received much attention
in recent times.1–3 1,8-Naphthalimide derivatives represent an
important family of DNA binders that have been shown to
exhibit anti-tumour activities in vitro and in vivo.4,5 Moreover,
the naphthalimides have rich photophysical properties; which are
dependent on the nature, as well as the location, of any aryl sub-
stituents. Photo-excitation of various naphthalimide derivatives
has been also shown to induce sequence selective DNA strand
cleavage.6–8 For example, Saito et al. showed that unsubstituted
as well as 4-nitro substituted L-lysine derived 1,8-naphthalimides
caused sequence selective cleavage at the 5′-guanine (G) residue
of 5′-GG sequence upon UV-irradiation and treatment with hot
piperidine. The cleavage ability of these naphthalimide systems
has been explained in terms of photoinduced electron transfer
(PET), occurring from the most oxidizable DNA base G, to the
triplet excited state of naphthalimide.6 In a related study, Kelly
and co-workers have reported the interaction of cationic unsub-
stituted naphthalimides functionalised with N-ethyl pyridinium
group, with DNA and mononucleotides8 and showed that laser
While these unsubstituted naphthalimides absorb only at
shorter wavelength, the 3- and 4-amino-1,8-naphthalimide
derivatives by contrast absorb within the visible region (λmax
∼
450 nm) and fluoresce at long (λmax ∼ 550 nm) wavelengths,
with reasonably high quantum yield of emission in various
organic and aqueous solvents.9 Because of this, they have been
extensively explored for the development of sensors for biologi-
cally relevant ions and molecules, etc.9–16 Their excited states
are characterized by a “push–pull” internal charge transfer (ICT)
character, arising from the presence of the electron-withdrawing
imide and the electron donating amino moiety; this being stron-
gest where the amino group is in the 4-position of the ring.9,15–19
Because of this they have also been employed as potential
cellular imaging and antitumor agents.20 In this context, we have
also reported development of DNA photocleavage agents based
on 4-amino-1,8-naphthalimide conjugated with ruthenium(II)
complex,19 and we have prepared Tröger’s base derivatives of
the 4-amino-1,8-naphthalimide structure and shown these to be
efficient DNA binders, and cytotoxic agents.20 In this article we
report the synthesis of 1 (a novel cationic pyridinium based
4-amino-1,8-naphthalimide derivative), and examine its inter-
action with natural and synthetic double-stranded DNAs using
UV-visible absorption, fluorescence, linear and circular dichro-
ism spectroscopic methods. These demonstrate that unlike its
unsubstituted analogue 2, the 4-amino-substituted compound
1 intercalates into double-stranded DNA and shows an apparent
preference for AT sites.
School of Chemistry, Centre of Synthesis and Chemical Biology,
University of Dublin, Trinity College Dublin, Dublin2, Ireland. E-mail:
jmkelly@tcd.ie, gunnlaut@tcd.ie; Tel: +3531 896 1947+353 1 896
3459
†Electronic supplementary information (ESI) available: Fig. S1–S3.
CCDC 843421. For ESI and crystallographic data in CIF or other elec-
tronic format see DOI: 10.1039/c2ob06898b
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3033–3043 | 3033

View Online
orientation of the pyridinium ring in Fig. 1a. The packing inter-
actions of 1 were found to be governed by N–H hydrogen,
anion⋯π interactions and π⋯π stacking. H-bonding interactions
between the amino proton on one molecule and the carbonyl
oxygen atom on an adjacent molecule give rise to a 1D poly-
meric chain in the crystallographic a-direction [N(1)⋯O(1) =
2.919(3) Å and <(N(1)–H(1X)⋯O(1)) = 154°]. π⋯π stacking
[centroid⋯centroid = 3.595 Å] between naphthalimide moieties
on an adjacent chain links the chains in an anti-parallel (head-to-
Results and discussion
Design, synthesis and solid state characterisation of 1
In designing the molecule we foresaw that in 1 the pyridinium
moiety would impart water solubility as well as favouring, elec-
trostatic interaction with the negative phosphate backbone of the
DNA. The substitution should also prevent protonation in the pH
range used for studies with the biomolecule, simplifying the
spectroscopic studies.
−
tail) fashion (as shown in Fig. 1b). The PF6 counter anion is
Compound 1 was synthesised in three steps as shown in
Scheme 1.8,21 The first step involved formation of 3, which was
achieved through a condensation reaction between the 4-nitro-
1,8-naphthalic anhydride and ethanolamine in dry ethanol under
an inert atmosphere and reflux.21 Product 3 was isolated as a
brown solid in 75% yield and fully characterised (see ESI
Fig. S1†). Compound 4 was then formed, as an orange solid in
90% yield by catalytic hydrogenation of 3 using 10% Pd/C in
DMF at 3 atm H2. Refluxing 4 in the presence of p-toluene sul-
fonyl chloride in pyridine, gave the desired product 1 as a tosy-
late salt. This was put into water, washed with CH2Cl2 and
purified using silica flash column chromatography (88 : 11 : 1
CH3CN : H2O : NaCl). The product was then precipitated as its
also involved in H-bonding [N(1)⋯F(5) = 3.304(3) Å and <(N
(1)–H(1Y)⋯F(5)) = 139°] and anion⋯π interactions (F⋯cen-
troid = 3.241 Å) further extending the packing.
UV-visible, fluorescence and NMR studies of 1
The absorption and emission spectra of 1 (8.3 μM) in 10 mM
phosphate buffer solution (pH 7.0) are shown in Fig. 2a. 1 exhi-
bits a broad absorption band 435 nm (ε = 13 200 M−1 cm−1),
which is, as discussed above, due to an ICT transition.9,23,24
Higher energy π–π* transitions were also observed at wavelength
ca. 250 nm (ε = 21 800 M−1 cm−1). In the same solvent (pH
7.0), 1 exhibits a broad ICT fluorescence emission band centred
at 552 nm, when excited at 435 nm. As anticipated the photo-
physical properties of 1 were found to be pH independent in the
range pH 3–11 (see ESI, Fig. S3†), which is useful for its appli-
cation as a DNA targeting molecule.
−
PF6 salt using NH4PF6 and converted to the chloride salt of 1
by the treatment with DOWEX-1 × 8–200 ion exchange resin.
All compounds were characterized by conventional 1H, 13C
NMR (see ESI Fig. S2†), HRMS and IR techniques.
Large orange, X-ray quality crystals of 1 were obtained as the
−
PF6 salt from acetone.† The low temperature (112 K) structure
of 1 is shown in Fig. 1a, where it can be seen that the pyridinium
ring is oriented towards the top face of the naphthalimide ring as
the ethylene linker adopts a gauche conformation, placing the
pyridinium moiety over one of the oxygen atoms of the imide
structure. This is probably a consequence of the electron
deficient nature of the pyridinium ring and the electron rich
imide portion of the naphthalimide ring. Indeed, Gao and
Marcus showed that this ‘push–pull’ character places a partial
positive charge on the 4-amino moiety and a partial negative
charge at the imide site causing the charge to be distributed
on the carbonyl portions of the imide and making the central
nitrogen moiety electron deficient.22 This might explain the
Fig. 1 (A) The X-ray crystal structure of 1 with thermal ellipsoids
shown at 50% probability. (B) The packing of 1 when viewed down the
crystallographic a-axis, showing the π–π interactions between the
stacked naphthalimide rings which adopt
orientation.
a ‘head-to-tail’ type
Scheme 1 Synthesis of 1.
3034 | Org. Biomol. Chem., 2012, 10, 3033–3043
This journal is © The Royal Society of Chemistry 2012

View Online
Table 1 UV-vis absorption and emission data for 1 in a range of
solvents
λmax
λmax
(Abs) (Flu)
kr(×107) knr(×108)
Solvent
(nm)a (nm)a ϕF
τ (ns)a (s−1
)
(s−1
)
Water
Ethanol
Propan-1-ol 444
Propan-2-ol 444
Butan-2-ol
436
440
552
536
534
532
530
0.08 2.25
3.6
6.1
7.2
7.43
7.8
4.0
3.2
2.4
1.8
1.2
0.16 2.60
0.23 3.21
0.29 3.90
0.40 5.10
445
a All spectra were recorded at optical density (OD) of ∼0.1 upon
excitation at λmax for each of these solvents.
the solvents, 1 exhibited a broad absorption band, which showed
a slight blue shift (ca. 8–9 nm) on moving from the less polar
butan-2-ol to highly polar water. The emission properties were
found to be much more strongly dependent on the polarity of the
medium, showing ca. 22 nm red shift on moving from butan-2-
ol to water and a significant decrease in the quantum yield of
emission and fluorescence lifetime. The radiative (kr) and non
radiative (knr) decay constants were estimated from the quantum
yield and lifetime values, respectively. This showed that for the
alcohols kr is not significantly altered with increase in solvent
polarity, while knr increased substantially between butan-2-ol and
ethanol. By contrast in water, kr was found to be significantly
lower compared to the alcohols.
Similar behaviour has previously been observed for related 4-
amino-1,8-naphthalimides, which suggests that the excited states
of these molecules have ICT character.23–25 The decrease of
quantum yield of emission and lifetime in more polar and H-
bonding solvents can be explained in terms of specific solvent–
solute interactions resulting in the increase in non radiative
decay rates. Since 1 exhibits significant solvatochromism and as
its emission quantum yield is strongly dependent on solvent
polarity we may expect to observe substantial changes in the
photophysical properties of 1 upon binding to nucleic acids.26
Fig. 2 (a) Normalised UV-vis (—), fluorescence spectra ( ) λex
=
435 nm and excitation spectra ( ) λemx = 550 nm of 1 in 10 mM phos-
phate buffer (pH 7). (b) The absorption and (c) the emission spectra of 1
in different solvent of varying polarity.
Compound 1 shows good water solubility, and obeys the
Beer–Lambert law in aqueous solution at concentrations up to
100 μM as expected if the molecule exists in its monomeric
form under these conditions. However, as we wish to study the
compound in higher concentrations in further studies (e.g. cellu-
lar studies), it was important to determine whether the compound
might aggregate under such conditions. To investigate this we
carried out 1H NMR studies (600 MHz, D2O) within the concen-
tration range of 1 μM → 10 mM. These showed (see ESI†) that
the naphthalimide ring protons were shifted upfield upon
increasing concentration above 100 μM (even though no new
band was observed in the UV/vis absorption spectrum). The
NMR results are consistent with the formation of small aggre-
gates at these higher concentrations. This is not too unexpected
given the ability of 1 to undergo π–π stacking as was observed in
the solid state structure in Fig. 1. Importantly, these results also
show that the compound is monomeric at the concentrations of 1
employed in the photophysical studies herein.
Interaction of 1 with mononucleotides
Ground state interaction of 1 with mononucleotides. We first
investigated its ability to bind mononucleotides. As can be seen
from Fig. 3a and 3b the UV-vis spectrum of 1 is significantly
altered in the presence of purine mononucleotides adenosine 5′-
monophosphate (AMP) (0 → 120 mM) and guanosine 5′-mono-
phosphate (GMP) (0 → 60 mM). Addition of both AMP and
GMP caused a decrease in the absorbance of 1, with a concomi-
tant red shift of the λmax. However, the extent of hypochromicity
was different with each mononucleotide (Table 2).
A single isosbestic point was observed with both AMP and
GMP, suggesting the formation of 1 : 1 supramolecular ground
state complex between 1 and either of the purine mononucleo-
tides. Similar behaviour has been previously observed for un-
substituted naphthalimides and has been attributed to
naphthalimide–nucleotide stacking interactions.8 In contrast, the
pyrimidine mononucleotides, cytidine 5′-monophosphate (CMP)
and thymidine 5′-monophosphate (TMP) did not cause any sig-
nificant change in the UV-vis spectrum of 1.
To investigate the effect of solvent polarity, the absorption and
fluorescence emission spectra of 1 were recorded in alcohols of
varying polarity as well as in water. Representative spectra are
shown in Fig. 2, and the results are summarised in Table 1. In all
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3033–3043 | 3035

View Online
Fig. 3 Changes in the absorption spectrum of 1 (7.2 μM) in the presence of increasing concentration of (a) AMP (0–120 mM) in 10 mM phosphate
buffer (pH 7.0), (b) GMP (0–60 mM). Inset: plot of ΔA/[NI] vs. concentration of AMP/GMP and fit (—) according to eqn (3) (ΔA = λmax of free 1).
Changes in the steady state emission spectra of 1 (7.2 μM) in the presence of increasing concentration of (c) AMP (0–120 mM) (λex = 463 nm) and
(d) GMP (0–60 mM) (λex = 470 nm) in 10 mM phosphate buffer (pH 7.0). Inset: plot of I/I0 vs. concentration of XMP.
(Table 2) indicates that 1 complexes more strongly to AMP than
to GMP. Furthermore, these are almost twice as high as that pre-
viously observed for the unsubstituted naphthalimide derivative
2,8b clearly demonstrating the importance of the ICT character,
Table 2 Summary of binding parameters obtained from absorption
spectra of 1 in presence of AMP and GMP
Binding constant
Binding constant
(Scatchard plot)27
(nonlinear model of
Deranleau)28
K1
K2
1 þ XMP O 1 : XMP O 1 : ðXMPÞ2
which makes 1 highly polar and as such more capable of π–π
ð1Þ
XMP
% Hypo
K1 (M−1
)
K2 (M−1
)
K1 (M−1
)
K2 (M−1
35
)
AMP
GMP
19
12
138 10
117 11
37
—
5
155
135
7
5
5
stacking with these mononucleotides.
—
Singlet excited state interactions with mononucleotides. Hav-
ing established a strong association of 1 with AMP and GMP in
the ground state, we next evaluated changes in the excited state
of 1. A significant blue shift (7 nm) in the emission maximum is
observed for 1 in the presence of AMP (120 mM) (Fig. 3c) and
concurrently the quantum yield of emission of 1 (Φfree = 0.08) is
approximately doubled (Φbound = 0.14) when excited at the iso-
sbestic point (λex = 463 nm). This behaviour was further sup-
ported by time resolved fluorescence measurements, as the
fluorescence lifetime increased from 2.2 ns for free 1 to 4.8 ns in
the presence of AMP (70 mM), Table 3. The increase in relative
quantum yield of emission and singlet state lifetime of 1 in its
AMP complex can be attributed primarily to a decrease of the
Assuming a 1 : 1 association between 1 and the nucleotide
monophosphates (XMP), the association constants for their for-
mation were first determined using linear Scatchard analysis (see
ESI† for details).27 Good linearity was observed for the associ-
ation of 1 with GMP. However, for the interaction of 1 with
AMP, deviation from linearity was observed at higher nucleotide
concentration, which possibly indicates that higher order supra-
molecular complexes also exist under these conditions. The
binding data were therefore also analysed using the nonlinear
expression developed by Deranleau (see eqn (3), ESI†) consider-
ing the formation of 1 : 1 and 1 : 2 complexes between 1 and
XMP as shown in eqn (1).28 The derived binding constants
3036 | Org. Biomol. Chem., 2012, 10, 3033–3043
This journal is © The Royal Society of Chemistry 2012

View Online
Table 3 Fluorescence lifetime data as a function of mononucleotide
concentration in 10 mM phosphate buffer (pH 7)
[XMP]
(mM)
λmax (Flu)
(nm)
ϕF
τ1 (ns)
τ2 (ns)
1
0
0.08 552
0.02 556
0.14 545
2.2
(100%)
1.9 (33%) 0.7
(67%)
2.6 (29%) 4.8
(71%)
—
1 +
GMP
1 +
66
71
AMP
rate of non radiative decay processes as it is less accessible to
solvent molecules and is in a relatively more rigid orientation.
Similar changes in non radiative decay processes have been
reported for the interaction of various porphyrin2a and ethidium
bromide derivatives with nucleobases and DNA.29–31
Fig. 4 The LD spectrum of free st-DNA (400 μM) and st-DNA in
presence of 1 at varying P/D in 10 mM phosphate buffer (pH 7.0).
In contrast to what is found for AMP, addition of GMP (up to
60 mM) caused a red shift of 4 nm in λmax and about 60%
quenching of the emission intensity of 1 when excited at 470 nm
(Fig. 3). This may be attributed to PET from guanine to the
photo-excited naphthalimide moiety in the preformed complex,
as has been for other naphthalimides.8b,c Quenching of the emis-
sion intensity of 1 at lower concentration of GMP occurs in a
1 : 1 complex. The residual fluorescence at very high concen-
tration of GMP, where 1 is completely bound can be explained
by considering fluorescence emission from the NI : GMP com-
plexes. At very high concentration of GMP, formation of higher
order complexes cannot be ruled out.
The nature of the interaction of 1 with GMP was also sup-
ported by time correlated single photon counting measurements.
In aerated 10 mM phosphate buffer the fluorescence of 1 decays
by single exponential kinetics with a lifetime of 2.2 0.2 ns,
while in the presence of GMP, two components are observed.
The first one has a lifetime of ca. 2 ns and the second one has a
lifetime of 0.7 0.1 ns. The relative contribution of this short-
lived component increased with increasing concentration of
GMP up to 66 mM. This short-lived component is presumed to
arise from a 1 : 1 excited state complex between 1 and GMP.
However, it is intriguing to note that at the highest concentration
used, the long-lived component (1.9 ns) still represents 33% of
the emitting species, even though calculations from the UV-vis
absorption titrations suggest that the concentration of free 1
should be much lower. A possible explanation is that this emit-
Interaction of 1 with st-DNA
Having explored the ability of 1 to interact with mononucleo-
tides, we next explored the binding of 1 to double stranded
salmon testes DNA (st-DNA) using various spectroscopic tech-
niques, including LD, CD, UV-vis absorption and fluorescence
spectroscopy, as well as thermal denaturation.
Linear dichroism measurements. LD spectroscopy, which is
increasingly being found to be a powerful method for probing
the binding of small molecules to nucleic acids,33 was used to
investigate the mode of binding of 1 to st-DNA. The DNA was
flow-oriented in a Couette cell at 2000 rpm. The spectrum of
free st-DNA (400 μM) exhibits a negative LD signal around
260 nm characteristic of B-DNA, which arises from the orien-
tation of transition dipole of the base pairs approximately per-
pendicular to the helical axis of DNA. The reduced linear
dichroism {LDr = LD/(isotropic absorbance)} was calculated to
be −0.027 0.003 (a value similar to that obtained for B-DNA
samples by other researchers33). The LD was also measured for
st-DNA in the presence of 1 at nucleotide/drug ratio P/D = 2.5, 5
and 10, Fig. 4. Most significantly, a new negative band appears
at 440 nm corresponding to the ICT absorption band of 1. The
reduced LD for this feature for the P/D = 5 sample is 0.033
0.003, which is similar to that of the DNA sample, indicating
that the naphthalimide is at approximately the same angle to the
direction of flow as the base-pairs. This is as would be expected
for an intercalative mode of binding by the naphthalimide (This
calculation assumes that the orientation factor S is the same for
both DNA and the naphthalimide–DNA sample. It is possible
that the interaction of 1 slightly stiffens the biomolecule and
hence improves the orientation of the DNA molecules along the
direction of flow, which would lead to a slightly higher LDr). At
P/D = 2.5 unbound ligand makes a significant contribution to
the overall absorbance value, thereby making LDr for the
440 nm band quite low compared to the nucleobase.
ting species is 1 : (GMP)2*.
A longer lifetime for the
1 : (GMP)2* would be similar to the behaviour recently reported
for Pt(II)(meso-tetrakis(4-N-methyl-pyridyl))porphyrin, where
the Por(GMP)2* has a lifetime approximately five times longer
than that for the 1 : 1 Por(GMP)* excited state.2a
The fluorescence enhancement of 1 in the presence of AMP is
in contrast to the behaviour of the corresponding unsubstituted
naphthalimide 2, where the steady state emission of the naphtha-
limide derivative was quenched in the presence of both AMP
and GMP.8b This was suggested to be due to PET from the
nucleotide to the photoexcited naphthalimide chromophore. The
failure to observe fluorescence quenching of 1 with AMP is pre-
sumably a consequence of the lower reduction potential of the
excited state of the 4-amino derivative compared to the unsubsti-
tuted analogue, 2.32
Circular dichroism measurements. CD spectroscopy33 was
also used to investigate the mode of interaction of 1 with st-
DNA (150 μM) (Fig. 5) by monitoring the changes in the CD
spectrum upon addition of increasing amounts of 1 (P/D = 0 →
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3033–3043 | 3037

View Online
20). The ellipticity of the positive band at 275 nm band doubled,
while that of the negative band at 240 nm decreased slightly.
These changes are consistent with the alteration of the helical
structure of B-DNA in the presence of 1. Significantly, no
induced CD signal was observed for the bound naphthalimide.
An induced signal would be expected if it were groove-bound to
the DNA as has previously been reported for its unsubstituted
analogue 2.8e While these CD results do not explicitly demon-
strate that 1 binds to DNA via intercalation, they clearly show
that the interaction (particularly in light of the results obtained in
the LD-measurements) alters the secondary structure of DNA
significantly. These results might also support our objective that
the pyridinium moiety is also essential to achieving high binding
affinity for DNA via electrostatic interactions.
In the presence of 1 (P/D = 10), stabilisation of st-DNA was
indicated by a 5° increase of the DNA melting temperature of st-
DNA in 10 mM buffer solution (Fig. 6). It is perhaps significant
that the extent of stabilisation is smaller than the effect observed
for well-established intercalators such as ethidium bromide.
However such moderate stabilization has been observed for other
mono-naphthalimide based DNA intercalators previously and
perhaps suggests a weaker intercalation interaction of 1.5,20,34
The thermal denaturation profile of the synthetic polynucleo-
tide poly(dA–dT)2 recorded in the presence of 1 shows that 1
stabilises poly(dA–dT)2 (ΔTm = 12 °C) to a much greater extent
than it does st-DNA, which contains 58% AT basepairs. This is
indicative of a higher affinity of 1 towards AT rich sequences.
(Unfortunately due to the high melting temperature of poly(dG–
dC)2 (∼90 °C) in 10 mM phosphate buffer, it was not possible to
study the thermal melting of this polynucleotide in the presence
of 1.)
Thermal denaturation studies. The influence of 1 on the
thermal stability of st-DNA was next investigated by monitoring
the temperature induced change in absorbance at 260 nm arising
from the conversion of double stranded DNA to the single
stranded DNA.
UV-visible measurements. The binding of 1 with st-DNA was
investigated from the changes in absorption of 1 in 10 mM phos-
phate buffer solution at pH 7.0 (Fig. 7a). In the presence of st-
DNA, a large hypochromicity and ca.10 nm red shift is observed
in the ICT absorption band, accompanied by a single isosbestic
point at ca. 480 nm for all DNA/l (P/D) concentrations
suggesting the presence of only two spectroscopically distinct
species.
The absorbance data were fitted to the non-cooperative
binding model of McGhee–von Hippel to determine the intrinsic
binding constant (K) and the binding site size (n) (inset Fig. 7a
Fig. 5 The CD spectra of st-DNA (150 μM) in 10 mM phosphate
buffer in presence of 1 at varying P/D ratios: 0 ( ), 2.5 (—), 10 ( ), 20
Fig. 6 Melting profile of (a) st-DNA and (b) poly(dA–dT)2 in absence
of 1 ( ) and in presence of 1 ( ) (P/D = 10), ( ) (P/D = 5) in 10 mM
phosphate buffer (pH 7.0).
(
).
Fig. 7 (a) UV-vis spectra of 1 (7.2 μM) in the presence of increasing concentration of st-DNA (0–780 μM) in 10 mM phosphate buffer (pH 7.0);
Inset: plot of r/Cf vs. r (■) and the best fit of the data (─) using the McGhee–von Hippel model. (b) Steady state emission of 1 (7.2 μM) in presence
of increasing concentration of st-DNA (0–780 μM) (λex = 480 nm) in 10 mM phosphate buffer (pH 7.0); Inset: plot of I/I0 vs. DNA nucleotide phos-
phate/ligand (P/D).
3038 | Org. Biomol. Chem., 2012, 10, 3033–3043
This journal is © The Royal Society of Chemistry 2012

View Online
Table 4 Summary of binding parameters obtained from the absorption
spectra of 1 in presence of st-DNA in 10 mM phosphate buffer solution
containing 0, 50, 100 mM NaCl
Table 5 Summary of binding parameters obtained from the absorption
and fluorescence spectra of 1 in the presence of st-DNA, poly(dA–dT)2
and poly(dG–dC)2 in 10 mM phosphate buffer solution (pH 7)
NaCl (mM)
% Hypo
Δλmax (nm)
K (×105 M−1
)
n (bp)
st-DNA
Poly(dA–dT)2
Poly(dG–dC)2
0
50
100
44
41
36
12
12
12
1.85 0.05
0.64 0.08
0.26 0.01
2.65 0.22
3.83 0.32
4.49 0.12
λmax (Abs) (nm)
Δλmax (nm)
% Hypo
447
+12
44
1.85 0.05
2.65 0.22
547
−5
2
447
+12
49
5.15 0.07
2.54 0.12
547
−5
2
445
+10
52
0.82 0.02
3.62 0.09
544
K (×105 M−1
n (bp)
)
λmax (Flu) (nm)
Δλmax (nm)
I/I0
τ1 (ns)
τ2 (ns)
and Table 4).35 In the 10 mM phosphate buffer solution the
intrinsic binding constant for the association of 1 with st-DNA
was found to be ca. 1.85 × 105 M−1, which is comparable to the
affinity of phenothiazine dyes for DNA.36 Notably this is an
order of magnitude higher than that seen for the unsubstituted
analogue 2, demonstrating once more the importance of the 4-
amino moiety and ICT character of 1.8 Similar DNA binding
affinity has recently been reported by Qian and co-workers for
thio-heterocyclic rings fused naphthalimides.5
To investigate the role of electrostatic interaction in the
binding process, the binding of 1 with st-DNA was investigated
in the presence of increasing NaCl concentration. An overall
decrease in the magnitude of binding constant was observed
with increasing ionic strength of the buffer, indicating significant
contribution of electrostatic interaction in the ligand binding
process (Table 4). This suggests that in the addition of the inter-
calation binding mode of the naphthalimide moiety, the pyridi-
nium part plays a significant role in the overall binding affinity
of 1 for DNA.
−8
0.7a
3.7 (30%)
9.2 (70%)
3.1 (6%)
9.9 (94%)
1.5 (51%)
3.7 (49%)
a Fluorescence intensity after correcting for the absorbance decrease at
435 nm.
either polynucleotide results in large hypochromicity and a sig-
nificant red shift in the ICT absorption band of 1. In the case of
poly(dA–dT)2, a single isosbestic point was observed at 480 nm
for the entire polynucleotide concentration range (Fig. 8a). By
contrast for poly(dG–dC)2 (Fig. 8b) an isosbestic point was
observed (at 480 nm) only at low concentration of the polynu-
cleotide ([poly(dG–dC)2/[ligand] ≈ 3) indicating the possible
existence of more than one bound form of ligand at higher con-
centration. The equilibrium binding constants values for the
association of 1 with polynucleotides were determined as
described above, from the fit of the absorbance data to non-coop-
erative model of McGhee and von Hippel and are summarized in
Table 5. This shows that 1 exhibits a very much higher affinity
for poly(dA–dT)2. It may be noted that this behaviour contrasts
with the lack of sequence preference for the unsubstituted com-
pound 2.8e
Fluorescence studies. In parallel with the absorption spectra,
the steady state emission spectra of 1 were measured upon
increasing concentration of st-DNA (Fig. 7b). It was found that
the fluorescence emission of 1 was enhanced in the presence of
st-DNA by a factor of two with a concomitant blue shift in λmax
of ca. 5 nm. This enhancement in fluorescence of 1 probably
results from a decreased rate of non-radiative decay of the
excited state due both to increased rigidity and the less polar
nature of the binding site.
Time resolved emission measurements were also undertaken
and these studies showed that in the presence of st-DNA, the
emission decay of 1 is consistent with a bi-exponential process
with a short component of τ1 = 3.73 ns (30%) and a longer lived
component of τ2 = 9.23 ns (70%) at a P/D of 100, Table 5,
which presumably represents two different localization of 1
within the DNA helix.
Fluorescence studies. The variation in the steady state emis-
sion intensity of 1 in the presence of increasing concentration of
poly(dA–dT)2 is shown in Fig. 8c. At high concentration of poly
(dA–dT)2, the emission intensity of 1 was approximately
doubled, with a 5 nm blue shift similar to that observed for the
titration of 1 with st-DNA. Time resolved emission measure-
ments revealed that in the presence of poly(dA–dT)2 at a P/D =
30 a long lived component with τ = 9.9 ns (94%) is observed
corresponding to the bound form of 1 (Table 5).
As can be seen in Fig. 8d, the excited state behaviour of 1
with poly(dG–dC)2 is strikingly different from that seen for poly
(dA–dT)2 or st-DNA, as the fluorescence is partially quenched
rather than enhanced. Thus in the presence of poly(dG–dC)2, the
emission intensity of 1 is reduced by ca. 30% after correcting for
the absorbance decrease at 435 nm. The fluorescence intensity
plot (inset on Fig. 8d) again reveals the marked difference in the
binding affinities for the two synthetic polynucleotides (note the
different P/D scales in Fig. 8c and 8d). Time resolved measure-
ments in the presence of excess poly(dG–dC)2 (at P/D = 105)
showed that the fluorescence of 1 decays by a bi-exponential
process with τ1 = 1.5 ns (51%) and τ2 = 3.8 ns (49%), so that the
observed average singlet state lifetime of 1 bound to poly(dG–
dC)2 remained practically unchanged from the free ligand. This
behaviour may be contrasted with what was found for 5′-GMP
Interaction of 1 with synthetic polynucleotides
UV-visible absorption studies. It has been demonstrated above
that 1 has a high affinity for DNA and that binding causes sig-
nificant changes in its photophysical properties. As we showed
earlier that complexing of 1 to AMP leads to increased excited
state lifetime while GMP causes a partial quenching, it was of
interest to study the behaviour of 1 in the presence of the syn-
thetic polynucleotides poly(dA–dT)2 and poly(dG–dC)2.
As before the changes in the UV-vis absorption spectra of 1
with increasing concentration of poly(dA–dT)2 and poly(dG–
dC)2 were monitored (Fig. 8a and 8b, respectively). Addition of
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3033–3043 | 3039

View Online
Fig. 8 The UV-vis absorption spectra of (a) 1 (7.0 μM) in presence of increasing concentration of poly(dA–dT)2 (0–265 μM); (b) 1 (4.0 μM) in pres-
ence of increasing concentration of poly(dG–dC)2 (0–410 μM); Inset a and b: plot of r/Cf vs. r and best fit ( ) to McGhee–von Hippel model for poly
(dA–dT)2 and poly(dG–dC)2 respectively; (c) 1 (7.0 μM) in presence of poly(dA–dT)2 (0–265 μM) (λex = 480 nm); (d) 1 (4.0 μM) in presence of
increasing concentration of poly(dG–dC)2 (0–410 μM). (λex = 435 nm) in 10 mM phosphate buffer (pH 7.0). Inset c and d: Inset c: plot of I/I0 vs. P/D
for poly(dA–dT)2; and poly(dG–dC)2 respectively for poly(dG–dC)2: ( ) fluorescence intensity not corrected for absorbance decrease at λex
435 nm; (■) the fluorescence intensity decrease after correcting for the absorbance decrease at 435 nm.
=
(Table 3) where no longer-lived emission was observed and with
that of poly(dA–dT)2, where a much longer-lived component
was detected. It is probable that two conflicting trends are operat-
ive for the singlet excited state of 1 bound to poly(dG–dC)2.
Firstly it is expected that there should be an increase in lifetime
due to location within the non-aqueous and more rigid environ-
ment of the polynucleotide. However this will be balanced by a
reduction in the excited state lifetime because of photo-induced
electron transfer (PET) from guanine to the excited naphthali-
mide, as found with 5′-GMP.
naphthalimide ring and the purine bases is further demonstrated
by the complexation with the mononucleotides 5′-AMP and 5′-
GMP.
1 shows a strong preference to bind to A–T regions of DNA,
as is demonstrated by the six times larger binding constant deter-
mined for poly(dA–dT)2 than for poly(dG–dC)2. The similar
emission enhancement in both double-stranded st-DNA and in
poly(dA–dT)2 strongly suggests that in the natural DNA the
intercalator also seeks out AT sequences. It is also notable that
thermal denaturation experiments show that binding of 1 to poly
(dA–dT)2 causes a significantly higher stabilisation of the
double-stranded form than is found for natural DNA, indicative
of a thermodynamically stronger interaction with such alternating
AT sequences. This could indicate that compounds such as 1
might therefore have a medicinal role for targeting AT rich DNA,
as is for example found in the genome of the malaria parasite P.
falciparum.37
Conclusions
This study shows that the cationic 4-amino-1,8-naphthalimide
derivative 1, which possesses a pyridinium moiety, binds
strongly to DNA and synthetic polynucleotides. Linear dichro-
ism and the other spectroscopic studies are consistent with the
compound binding by intercalating between the base-pairs of
polynucleotide. It is also likely that the pyridinium moiety binds
electrostatically within the groove and possibly with the phos-
phate group of the DNA. The strong interaction between the
It is interesting to note that 1 behaves quite differently in
many respects to the analogous unsubstituted compound 4,
which McMasters and Kelly have shown locates in the minor
groove of double stranded DNA with the binding predominantly
3040 | Org. Biomol. Chem., 2012, 10, 3033–3043
This journal is © The Royal Society of Chemistry 2012

View Online
governed by the electrostatic interactions between negatively
charged phosphate backbone of DNA and positively charged
pyridinium group of the ligand.8e Surprisingly that compound
does not show a preference for AT sequences, as might be
expected for a groove binder. Another difference between the
two compounds is that the singlet excited state of the unsubsti-
tuted compound 2 is quenched by both adenine and guanine,
presumably indicative of the more oxidising nature of its excited
state (due at least in part to its higher excited state energy). In the
case of the amino-derivative 1 its singlet excited state clearly
must have a reduction potential too low to oxidise adenine but
sufficient to form the guanine radical cation (or possibly guanine
radical). We have carried out preliminary picosecond transient
absorption measurements to look for evidence for this electron
transfer process for 1 in the presence of 20 mM GMP.38
Although bleaching and transient absorption bands characteristic
of the singlet excited state of 1 were observed and found to
decay with a lifetime of 0.7 0.1 ps, no new transient species
were observed suggesting that back electron transfer is very
rapid. Further studies are required and will be the subject of a
later paper. We are also currently commencing a full biological
evaluation of 1 and related compounds, which includes evaluat-
ing their cellular uptake, localisation studies, where the green
emission of 1 will be capitalised on, as well as carrying out
detail cytotoxic and mechanistic studies.
were carried out in quartz cuvettes (10 mm × 10 mm). 10 mM
phosphate buffer of pH 7.0 was made up by mixing portions of
1 M stock solutions of Na2HPO4 and NaH2PO4. All solutions
were prepared fresh prior to measurement. The association con-
stants, K for nucleotides was determined from the slope of
Scatchard plot (see ESI† for details).27
For the thermal denaturation studies, semi-micro UV-vis cuv-
ettes were used (path length of 1 cm and window width of
4 mm, from Starna). The temperature-based measurements were
obtained using a Cary temperature controller in conjunction with
the Varian Cary 300 UV-vis spectrometer. The temperature was
ramped from 30–90 °C at a rate of 0.5 °C min−1. All solutions
were thoroughly degassed before the experiment.
Fluorescence measurements. Steady-state fluorescence spectra
were recorded on a Varian Cary Eclipse spectrofluorimeter. Flu-
orescence titrations, were carried out using optically dilute sol-
utions (absorbance < 0.1) following the same procedure as
described for UV-vis titrations using slit widths 5 nm for both
excitation and emissions. Fluorescence quantum yields were
measured using dilute solutions (absorbance <0.05) against flu-
orescein in 0.1 N NaOH solution (Φ = 0.9).39 Fluorescence life-
times were measured and analysed using Fluorolog 3 time
correlated single photon counting apparatus with a Nanoled (λex
= 458 nm) used as the excitation source. Emission decays were
recorded at 550 nm using slit width of 2 nm. All the decay traces
were corrected for instrument response using a dilute solution of
colloidal silica. In the absence of any mononucleotides, DNA or
polynucleotides, decay kinetic of 1 was analysed by monoexpo-
nential model. In the presence of mononucleotides or polynu-
cleotides, decay traces were best described by biexponential
functions. Goodness of the fit was determined from the χ2 value
(1.0–1.2) and the average residual distributions.
Experimental details
All the solvents used for photophysical studies were spectro-
scopic grade. Freshly distilled acetonitrile was used for quantum
yield measurement. Deuterated solvents for NMR were pur-
chased from Apollo Ltd. Millipore-purified water was used in
DNA related work. The mononucleotides AMP, GMP, CMP and
TMP, st-DNA and the synthetic polynucleotides poly(dA–dT)2
and poly(dG–dC)2 were obtained from Sigma Aldrich as their
sodium salts. The DNA concentration (in nucleotide base) was
determined spectrophotometrically using ε260 of 6.0 × 103 M−1
cm−1. The concentrations of the polynucleotide stock solutions
were determined using ε254 of 8.4 × 103 M−1 cm−1 and ε260 of
6.6 × 103 M−1 cm−1 for poly(dG–dC)2 and poly(dA–dT)2,
respectively (per mole of nucleotide). All other chemicals were
obtained from Sigma-Aldrich or Fluka and were used without
further purification.
Linear dichroism measurements. Flow oriented linear dichro-
ism studies of were performed on a Jasco J-815 CD spectrometer
fitted with a Dioptica Scientific Ltd Linear Dichroism 2 appar-
atus. For an oriented sample, linear dichroism is defined as the
differential absorption of light polarised parallel (Apar) and per-
pendicular (Aper) to a reference axis (eqn (2))33
LD ¼ Apar ꢀ Aper
ð2Þ
Long flexible polymers such as peptides or DNA are usually
oriented by the technique of flow orientation. In this technique,
the molecules are oriented by shear force generated by a flow
gradient in the solution flowing between the narrow gap between
a spinning cell and a stationary cylindrical rod.
Quantitative information about the orientation of a chromo-
phore with respect to the reference axis can be obtained from
reduced LD (LDr) as described in eqn (3) below,
Infrared spectra were recorded on a Perkin Elmer Spectrum
100 FT-IR spectrophotometer. Melting points were determined
using an Electrothermal IA9000 digital melting point apparatus.
NMR spectra were recorded using a Bruker Advance III spec-
1
trometer, operating at 400 MHz for H NMR and 100 MHz for
13C NMR. Shifts were referenced relative to the internal solvent
signals. Electrospray mass spectra were recorded on a Mass
Lynx NT V 3.4 on a Waters 600 controller connected to a 996
photodiode array detector with HPLC grade methanol, water or
acetonitrile as carrier solvents. All accurate masses are quoted to
≤5 ppm.
LD Apar ꢀ Aper
3
2
LDr ¼
¼
¼
Sð3cos2α ꢀ 1Þ
ð3Þ
A
A
where, A is the absorbance of the sample under isotropic con-
dition (i.e. no orientation), S refers to the orientation factor (S =
1 for perfectly oriented sample, S = 0 for random orientation)
and α is the angle between the transition dipole of the chromo-
phore and the reference axis. The magnitude of S provides
Spectroscopic measurements
UV-vis absorption spectra. These were recorded on a Varian
Carry 50 spectrophotometer. All the spectroscopic measurements
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3033–3043 | 3041

View Online
information about structural changes in the macromolecule such
as lengthening, stiffening, bending etc.
filtered and washed with DCM followed by purification using
silica flash column chromatography using CH3CN : H2O : NaCl
(sat.) (88 : 11 : 1). The product was precipitated as PF6 salt
−
Circular dichroism measurements. CD spectra were recorded
on a Jasco J810 spectropolarimeter. CD titrations were carried
out by adding aliquots of solution of 1 to solution of st-DNA
(150 μM) in 10 mM phosphate buffer at varying P/D.
using ammonium hexafluorophosphate. The solid was dissolved
in minimum amount of MeOH and treated with
DOWEX-1×8–200 ion exchange resin to convert the product
into the chloride form. The product was obtained as an orange-
yellow solid after removal of excess MeOH under reduced
pressure (0.60 g, 60%). m.p. degraded at T > 250 °C; HRMS:
319.1233 ([M + H]+. C19H16N2O3 requires 319.1243); δH
(600 MHz, d6-DMSO) 9.13 (2H, d, J = 5.6 Hz, H15), 8.67 (1H,
d, J = 8.2 Hz, Ar–H5), 8.59 (1H t, J = 7.6 Hz, H17), 8.33 (1H,
d, J = 7.3Hz, Ar–H7), 8.07 (3H, m, H16, Ar–H2), 7.63 (1H, t, J
= 8.0, Ar–H6), 7.59 (2H, s, NH2), 6.83 (1H, d, J = 8.6 Hz, Ar–
H3), 4.94 (2H, t, J = 4.9 Hz, CH2), 4.57 (2H, t, J = 4.5 Hz,
CH2); δC (150 MHz, d6-DMSO), 164.4 (CvO), 163.3 (CvO),
153.5 (C), 146.2 (CH), 145.7 (CH), 134.5 (CH), 131.6 (CH),
130.3 (C), 130.0 (CH), 128.2 (CH), 124.3 (CH), 121.6 (C),
119.7 (C), 108.6 (CH), 107.0 (C), 60.2 (CH2), 40.6 (CH2).
νmax(neat sample)/cm−1: 3375.42 (w, N–H stretch), 2920.58 (w,
C–H asym. stretch), 2872.49 (w, C–H sym. stretch), 1679.93 (m,
CvO stretch).
Synthesis of N-(2-hydroxy ethyl)-4-nitro-1,8-naphthalimide (3)21
4-Nitro-1,8-naphthalic anhydride (2.034 g, 8.365 mmol) was
suspended in dry ethanol and ethanol amine (0.5 ml) was added
dropwise. The mixture was refluxed for 24 h under an argon
atmosphere and cooled to room temperature. The desired product
was separated by filtration, washed with cold ethanol and
obtained as a dark brown solid (1.8 g, 75%). m.p. 157–160 °C
(ref. 21, 157–160 °C); HRMS: 309.0488 ([M
+
Na]+.
C14H10N2O5 requires 309.0487); δH (400 MHz, d6-DMSO),
8.65 (1H, d, J = 8.5 Hz, Ar–H7), 8.56 (1H, d, J = 8.0 Hz, Ar–
H5), 8.52 (2H, m. Ar–H2, Ar–H3), 8.05 (1H, t, J = 8.0 Hz, Ar–
H6), 4.85 (1H, t, J = 6.0 Hz, OH), 4.14 (2H, t, J = 6.5 Hz, CH2),
3.64 (2H, q, J = 6.0 Hz, CH2); δC (100 MHz, d6-DMSO), 166.2
(CvO), 164.4 (CvO), 151.16 (C), 133.8 (CH), 132.2 (CH),
131.7 (CH), 130.8 (CH), 130.4 (C), 128.8 (C), 126.4 (CH),
124.9 (C), 124.8 (C), 59.8 (CH2), 44.37 (CH2). νmax(neat
sample)/cm−1: 3516.46 (m, O–H stretch), 3011.6 (w, C–H asym.
stretch), 2943.2 (w, C–H sym. stretch), 1691.52 (s, CvO
stretch), 1579.99 (s, NO2 asym. stretch), 1321.88 (s, NO2 sym
stretch).
Acknowledgements
Financial support of Science Foundation Ireland (SFI) for RFP
2006, RFP2007, RFP 2009 and PRTLI for Cycle 4 funding to
CSCB is acknowledged. JAK thanks IRCSET for a postdoctoral
fellowship. We would like to thank particularly Dr John O’Brien
and Dr Martin Feeney for their help with NMR and MS, respect-
ively and Professor Mike Towrie, Dr Ian Clark and Dr Greg
Greetham of the Ultrafast Laboratory of the STFC Rutherford
Appleton Laboratory for the picosecond transient absorption
measurements.
Synthesis of N-(2-hydroxy ethyl)-4-amino-1,8-naphthalimide (4)
Compound 4 was synthesised by catalytic hydrogenation of 3
(1.00 g, 3.494 mmol) in DMF using a Parr hydrogen shaker
apparatus at 3 atm pressure in the presence of 10% Pd/C catalyst
until the hydrogen consumption ceased. The reaction mixture
was filtered through celite and washed with DMF. The filtrate
and washings were concentrated under reduced pressure and the
product was precipitated as an orange solid from ice (0.812 g,
90%). m.p. 260–262 °C (ref. 70 261–263 °C); HRMS: 279.0748
([M + Na]+ C14H12N2O3 requires 279.0746). δH (400 MHz, d6-
DMSO), 8.62 (1H, d, J = 8.5 Hz, Ar–H7), 8.43 (1H, d, J = 7.5
Hz, Ar–H2), 8.20 (1H, d, J = 8.5 Hz, Ar–H5), 7.65 (1H, dd, J =
7.5 Hz and 8.0 Hz, Ar–H6), 7.44 (2H, s, –NH2), 6.85 (1H, d, J
= 8.0 Hz, Ar–H3), 4.80 (1H, t, J = 6.0 Hz, OH), 4.11 (2H, t, J =
6.5 Hz, CH2), 3.57 (2H, q, J = 6.5 Hz, CH2); δC (100 MHz, d6-
DMSO), 163.9 (CvO), 163.0 (CvO), 152.6 (C), 133.9 (CH),
130.9 (CH), 129.7 (C), 129.2 (CH), 123.9 (CH), 121.8 (C),
119.3 (C), 108.1 (CH), 107.6 (C), 57.9 (CH2), 41.3 (CH2).
νmax(neat sample)/cm−1: 3437.71 (m, N–H stretch), 3349.47 (m,
O–H stretch), 2925.20 (w, C–H asym. stretch), 2895.73 (w, C–H
sym. stretch), 1667.78 (m, CvO stretch).
References
1 (a) L. H. Hurley, Nat. Rev. Cancer, 2002, 2, 188; (b) S. Neidle and D.
E. Thurston, Nat. Rev. Cancer, 2005, 5, 285.
2 (a) P. A. Keane and J. M. Kelly, Photochem. Photobiol. Sci., 2011, 10,
1578; (b) J. G. Vos and J. M. Kelly, Dalton Trans., 2006, 4869; (c) E.
M. Tuite and J. M. Kelly, J. Photochem. Photobiol., B, 1993, 21, 103.
3 (a) R. B. P. Elmes, M. Erby, S. M. Cloonan, S. J. Quinn, C. D. Williams
and T. Gunnlaugsson, Chem. Commun., 2011, 47, 686; (b) D. Griffith,
M. P. Morgan and C. J. Marmion, Chem. Commun., 2009, 6735;
(c) J. Andersson, S. Li, P. Lincoln and J. Andréasson, J. Am. Chem. Soc.,
2008, 130, 11836.
4 (a) P. F. Bousquet, M. F. Brana, D. Conlon, K. M. Fitzgerald, D. Perron,
C. Cocchiaro, R. Miller, M. Moran, J. George, X. D. Qian, G. Keilhauer
and C. A. Romerdahl, Cancer Res., 1995, 55, 1176; (b) M. F. Brana,
M. Cacho, A. Ramos, M. T. Dominguez, J. M. Pozuelo, C. Abradelo, M.
F. Rey-Stolle, M. Yuste, C. Carrasco and C. Bailly, Org. Biomol. Chem.,
2003, 1, 648; (c) M. F. Brana, M. Cacho, M. A. Garcia, B. de Pascual-
Teresa, A. Ramos, M. T. Dominguez, J. M. Pozuelo, C. Abradelo, M.
F. Rey-Stolle, M. Yuste, M. Banez-Coronel and J. C. Lacal, J. Med.
Chem., 2004, 47, 1391; (d) M. Lv and H. Xu, Curr. Med. Chem., 2009,
16, 4797; (e) L. Ingrassia, F. Lefranc, R. Kiss and T. Mijatovic, Curr.
Med. Chem., 2009, 16, 1192.
5 (a) X. Qian, Y. Li, Y. Xu, Y. Liu and B. Qu, Bioorg. Med. Chem. Lett.,
2004, 14, 2665; (b) Y. Li, Y. Xu, X. Qian and B. Qu, Tetrahedron Lett.,
2004, 45, 1247; (c) I. Ott, Y. Xu, J. Liu, M. Kokoschka, M. Harlos, W.
S. Sheldrick and X. Qian, Bioorg. Med. Chem., 2008, 16, 7107;
(d) Q. Yang, P. Yang, X. Qian and L. Tong, Bioorg. Med. Chem. Lett.,
2008, 18, 6210; (e) Z. Chen, X. Liang, H. Zhang, H. Xie, J. Liu, Y. Xu,
Synthesis-(2-(N-pyridinium)-ethyl)-4-amino-1,8-naphthalimide (1)
A mixture of 4 (0.8 g, 3.12 mmol) and p-toluene sulfonyl chlor-
ide (0.65 g, 3.91 mmol) were stirred in pyridine at 60 °C for
20 h and then refluxed for 4 days. The resulting precipitate was
3042 | Org. Biomol. Chem., 2012, 10, 3033–3043
This journal is © The Royal Society of Chemistry 2012

View Online
W. Zhu, Y. Wang, X. Wang, S. Tan, D. Kuang and X. Qian, J. Med.
Chem., 2010, 53, 2589; (f) Z. Li, Q. Yang and X. Qian, Bioorg. Med.
Chem. Lett., 2005, 15, 3143.
6 (a) I. Saito, M. Takayama, H. Sugiyama, K. Nakatani, A. Tsuchida and
M. Yamamoto, J. Am. Chem. Soc., 1995, 117, 6406; (b) I. Saito,
M. Takayama and S. Kawanishi, J. Am.Chem. Soc., 1995, 117, 5590;
(c) S. M. T. Matsugo, K. Yamaguchi, T. Mori and I. Saito, Nucleic Acids
Res. Symp. Ser., 1991, 21, 109.
7 (a) T. Takada, K. Kawai, S. Tojo and T. J. Majima, J. Phys. Chem. B,
2004, 108, 761; (b) K. Kawai, Y. Osakada, E. Matsutani and T. Majima,
J. Phys. Chem. B, 2010, 114, 10195.
8 (a) J. E. Rogers, B. Abraham, A. Rostkowski and L. A. Kelly, Photo-
chem. Photobiol., 2001, 74, 521; (b) S. McMasters and L. A. Kelly, J.
Phys. Chem. B, 2006, 110, 1046; (c) J. E. Rogers and L. A. Kelly, J. Am.
Chem. Soc., 1999, 121, 3854; (d) J. E. Rogers, S. Weiss and L. A. Kelly,
J. Am. Chem. Soc., 2000, 122, 427; (e) S. McMasters and L. A. Kelly,
Photochem. Photobiol., 2007, 83, 889.
17 (a) F. M. Pfeffer, A. M. Buschgens, N. W. Barnett, T. Gunnlaugsson and
P. E. Kruger, Tetrahedron Lett., 2005, 46, 6579; (b) F. M. Pfeffer,
M. Seter, N. Lewcenko and N. W. Barnett, Tetrahedron Lett., 2006, 47,
5241.
18 Y. Li, L. Cao and H. Tian, J. Org. Chem., 2006, 71, 8279.
19 G. J. Ryan, S. Quinn and T. Gunnlaugsson, Inorg. Chem., 2007, 47, 401.
20 (a) E. B. Veale and T. Gunnlaugsson, J. Org. Chem., 2010, 75, 5513;
(b) E. B. Veale, D. O. Frimannsson, M. Lawler and T. Gunnlaugsson,
Org. Lett., 2009, 11, 4040.
21 S. Chatterjee, S. Pramanik, S. U. Hossain, S. Bhattacharya and S.
C. Bhattacharya, J. Photochem. Photobiol., A, 2007, 187, 64.
22 Y. Q. Gao and R. A. Marcus, J. Phys. Chem. A, 2002, 106, 1956.
23 D. Yuan and R. G. Brown, J. Phys. Chem. A, 1997, 101, 3461.
24 (a) A. Pardo, J. M. L. Poyato, E. Martin, J. J. Camacho, D. Reyman, M.
F. Brana and J. M. Castellano, J. Photochem. Photobiol., A, 1989, 46,
323; (b) A. Pardo, J. M. L. Poyato, E. Martin, J. J. Camacho and
D. Reyman, J. Lumin., 1990, 46, 381.
9 R. M. Duke, E. B. Veale, F. M. Pfeffer, P. E. Kruger and
T. Gunnlaugsson, Chem. Soc. Rev., 2010, 39, 3936.
25 S. Saha and A. Samanta, J. Phys. Chem. A, 2002, 106, 4763.
26 (a) R. Jin and K. J. Breslauer, Proc. Natl. Acad. Sci. U. S. A., 1988, 85,
8939; (b) D. A. Barawkar and K. N. Ganesh, Nucleic Acids Res., 1995,
23, 159.
27 G. Scatchard, Ann. N. Y. Acad. Sci., 1949, 51, 660.
28 (a) D. A. Deranleau, J. Am. Chem. Soc., 1969, 91, 4044; (b) D.
A. Deranleau, J. Am. Chem. Soc., 1969, 91, 4050.
29 I. V. Sazanovich, E. P. Petrov and D. A. Deranleau, J. Am. Chem. Soc.,
1969, 91, 4044.
30 V. S. Chirvony, V. A. Galievsky, N. N. Kruk, B. M. Dzhagarov and P.
Y. Turpin, J. Photochem. Photobiol., B, 1997, 40, 154.
31 C. Prunkl, M. Pichlmaier, R. Winter, V. Kharlanov, W. Rettig and H.
A. Wagenknecht, Chem.–Eur. J., 2010, 16, 3392.
10 (a) T. Gunnlaugsson, P. E. Kruger, T. C. Lee, R. Parkesh, F. M. Pfeffer
and G. M. Hussey, Tetrahedron Lett., 2003, 44, 6575;
(b) T. Gunnlaugsson, T. C. Lee and R. Parkesh, Org. Biomol. Chem.,
2003, 1, 3265; (c) R. M. Duke and T. Gunnlaugsson, Tetrahedron Lett.,
2007, 48, 8043; (d) R. Parkesh, T. C. Lee and T. Gunnlaugsson, Org.
Biomol. Chem., 2007, 5, 310; (e) E. B. Veale and T. Gunnlaugsson, J.
Org. Chem., 2008, 73, 8073; (f) E. B. Veale, G. M. Tocci, F. M. Pfeffer,
P. E. Kruger and T. Gunnlaugsson, Org. Biomol. Chem., 2009, 7, 3447.
11 (a) Z. Xu, X. Qian and J. Cui, Org. Lett., 2005, 7, 3029; (b) Z. Xu, K.
H. Baek, H. N. Kim, J. Cui, X. Qian, D. R. Spring, I. Shin and J. Yoon,
J. Am. Chem. Soc., 2009, 132, 601.
12 Z. Xu, S. Kim, H. N. Kim, S. J. Han, C. Lee, J. S. Kim, X. Qian and
J. Yoon, Tetrahedron Lett., 2007, 48, 9151.
13 Z. Xu, J. Yoon and D. R. Spring, Chem. Commun., 2010, 46, 2563.
14 D. Wang, X. Zhang, C. He and C. Duan, Org. Biomol. Chem., 2010, 8,
2923.
32 R. Badugu, J. Fluoresc., 2005, 15, 71.
33 B. Nordén, A. Rodger and T. Dafforn. Linear and Circular Dichroism,
RSC Publishing, 2010.
34 F. Li, J. Cui, L. Guo, X. Qian, W. Ren, K. Wang and F. Liu, Bioorg. Med.
Chem., 2007, 15, 5114.
15 (a) A. P. d. Silva, H. Q. N. Gunaratne, J. L. Habib-Jiwan, C. P. McCoy, T.
E. Rice and J. P. Soumillion, Angew. Chem., Int. Ed. Engl., 1995, 34,
1728; (b) A. P. de Silva, H. Q. Nimal Gunaratne and T. Gunnlaugsson,
Tetrahedron Lett., 1998, 39, 5077.
16 (a) E. Tamanini, A. Katewa, L. M. Sedger, M. H. Todd and
M. Watkinson, Inorg. Chem., 2009, 48, 319; (b) E. Tamanini, K. Flavin,
M. Motevalli, S. Piperno, L. A. Gheber, M. H. Todd and M. Watkinson,
Inorg. Chem., 2010, 49, 3789.
35 J. D. McGhee and P. H. von Hippel, J. Mol. Biol., 1974, 86, 469.
36 E. Tuite and J. M. Kelly, Biopolymers, 1995, 35, 419.
37 M. J. Gardner, N. Hall, E. Fung, O. White, M. Berriman, R. W. Hyman,
J. M. Carlton, A. Pain, K. E. Nelson and S. Bowman, et al., Nature,
2002, 419, 49.
38 S. Banerjee, X.-Z. Sun, M. W. George, I. Clark, G. Greetham, M. Towrie,
T. Gunnlaugsson and J. M. Kelly, unpublished results.
39 J. Olmsted, J. Phys. Chem., 1979, 83, 2581.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3033–3043 | 3043