Closing the ring to bring up the light: Synthesis of a hexacyclic acridinium cyanine dye

Article abstract of DOI:10.1002/chem.201200802

The synthesis of a geometrically constrained and near-planar hexacyclic acridinium cyanine dye 9 is reported. When compared to its unlocked and non-fluorescent monomethine cyanine dye analogue 3, this photostable dye emits in the green area of the spectrum with a remarkable quantum yield close to unity in organic solvents and above 0.5 in water. A detailed steady-state and time-resolved spectroscopic study revealed that dye 9 forms emissive aggregates in water, which are responsible for a red-shifted and broadened emission band and longer emission lifetime, τ≈33 compared to 6.5-7.0-ns for the monomeric dye. Dye 9 also binds strongly to DNA (both duplex and quadruplex) in its monomeric form and is very efficiently taken up by cells, in which it accumulates primarily into the nucleus. Copyright

Full text of DOI:10.1002/chem.201200802

FULL PAPER
DOI: 10.1002/chem.201200802
Closing the Ring to Bring Up the Light: Synthesis of a Hexacyclic Acridinium
Cyanine Dye
Tariq Mahmood,[a] Yilei Wu,[b] Domitille Loriot,[c] Marina Kuimova,[b] and
Sylvain Ladame*[c]
Abstract: The synthesis of a geometri-
cally constrained and near-planar hexa-
cyclic acridinium cyanine dye 9 is re-
ported. When compared to its un-
locked and non-fluorescent monome-
thine cyanine dye analogue 3, this pho-
tostable dye emits in the green area of
the spectrum with a remarkable quan-
tum yield close to unity in organic sol-
vents and above 0.5 in water. A de-
tailed steady-state and time-resolved
spectroscopic study revealed that dye 9
forms emissive aggregates in water,
which are responsible for a red-shifted
and broadened emission band and
longer emission lifetime, tꢀ33 com-
pared to 6.5–7.0 ns for the monomeric
dye. Dye 9 also binds strongly to DNA
(both duplex and quadruplex) in its
monomeric form and is very efficiently
taken up by cells, in which it accumu-
lates primarily into the nucleus.
Keywords: acridines · aggregates ·
cyanine dyes · fluorescence · nucleic
acids
Introduction
orophores absorbing and emitting in the blue area of the
spectrum. Depending on their substitution pattern, these
pharmacophores also exhibit potent antibacterial or antitu-
mor activities.[6] To improve the spectroscopic and pharma-
cologic properties of acridines, extended and polysubstituted
acridines have been successfully developed. Representative
examples include 1) pentacyclic acridine RHPS4 and its ana-
logues,[7] 2) fused bispyrimidinoacridines[8] or 3) 3,6,9-trisub-
stituted acridines (Figure 1).[9] Very recently, we reported
Fluorescent small molecules are commonly used as labels or
probes in modern bioresearch. Efficient in vivo sensing or
imaging relies to a great extent on the physicochemical
properties of the chromophore used. Most suitable labels
for in vivo applications are biocompatible, stable and solu-
ble in relevant media or fluids and also conveniently excita-
ble (l>500 nm to avoid autofluorescence of biological tis-
sues and fluids).[1] Ideally, they will also be very bright (high
fluorescent quantum yield) under relevant conditions.
Within the wide selection of fluorescent dyes available,
those belonging to the cyanine dye (Cy)[2] and acridine[3]
families have been developed extensively, because of their
biocompatibility and high molar absorptivity.[4] They also
offer the advantage of covering a very broad spectral range,
from blue to near infrared. The family of cyanine dyes ge-
nerically consists of a conjugated system based on a polyme-
thine chain linking two nitrogen-containing heterocycles and
have found valuable applications in proteomics, imaging and
biomolecular labelling.[5] Acridines are tricyclic aromatic flu-
Figure 1. Structures of pentacyclic acridine RHPS4[7] (A) and a fused bis-
pyrimidinoacridine[8] (B), and general structures of mono- and polyme-
thine acridine-containing cyanine dyes[10] (C).
[a] Dr. T. Mahmood+
COMSATS, Institute of Information Technology
Abbottabad Campus (Pakistan)
[b] Y. Wu,+ Dr. M. Kuimova
the first family of acridine-based polymethine cyanine dyes
that covered a very broad spectral range from l=550 nm up
to the near infrared.[10] Among them, an acridine-containing
monomethine cyanine dye, analogue of the well-known thia-
zole orange, showed interesting fluorogenic properties, al-
though suffering from a very low fluorescence quantum
yield (Ff =0.20%) in water.
Department of Chemistry, Imperial College London
South Kensington Campus, London SW7 2AZ (UK)
[c] D. Loriot, Dr. S. Ladame
Department of Bioengineering, Imperial College London
South Kensington Campus, London SW7 2AZ (UK)
Fax : (+44)02075949817
E-mail: sladame@imperial.ac.uk
[+] These authors contributed equally to this work.
Herein, we describe the synthesis of a geometrically con-
strained and photostable hexacyclic acridine-containing
monomethine cyanine dye from 1-bromo-10-methylacridone
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201200802.
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

(4)[11] and demonstrate that a simple ring closure is sufficient
to convert a non-fluorescent molecule into a remarkably
bright dye. The potential of this class of compounds for cel-
lular imaging and nucleic acid staining is also investigated.
Results and Discussion
Effect of conformational freedom on fluorescence: By adap-
tation of a procedure recently reported by us,[10] we first syn-
thesised acridinium benzimidazole monomethine cyanine
dye 3. Briefly, N-methyl acridone (1) was converted into the
corresponding N-methyl-9-chloro-acridinium chloride (2),[12]
which was subsequently reacted with 1,2,3-trimethylbenzimi-
dazolium iodide[13] to afford the desired acridine-benzimida-
zole monomethine cyanine dye 3 (Scheme 1).
Figure 2. Energy-minimised (MM2 force field) structures of dyes 3 (left)
and 9 (right) produced by Chem3D. While hexacyclic dye 9 is locked in a
near-planar conformation, there is free rotation around the monomethine
bridge in compound 3 (white arrow) and steric clashes between the N-
methyl imidazole, and either H1 or H8 (represented as grey balls) of the
acridinium platform prevents co-planarity of both heterocycles.
lent bond between imidazole N1 and acridine C1 would lock
the molecule in a near-planar conformation by formation of
an additional six-membered ring, as suggested by MM2
energy minimisation studies (Figure 2).
Synthesis of a brightly emissive hexacyclic acridinium cya-
nine dye 9: To verify our model suggesting that bridging the
imidazole ring to the acridine scaffold would lock the struc-
ture of 3 into a near-planar and rigidified conformation—
thus resulting in increased conjugation and strongly altered
photophysical properties—we decided to synthesise hexacy-
clic acridinium dye 9 as shown in Scheme 2.
Scheme 1. Synthetic route to compound 3. Reagents and conditions: a)
1,2,4-triazole, POCl3, acetonitrile, reflux; b) 1,2,3-trimethylbenzimidazoli-
um iodide.[13]
We recently reported the fluorogenic properties of an ac-
ridine-benzothiazolium monomethine cyanine dye (Ff =
0.20% in water and Ff =5.6% in glycerol).[10] Similar to the
well-known molecule of thiazole orange (TO), free rotation
around the monomethine bridge makes this dye responsive
to viscosity and also to DNA binding.[10] However, unlike
TO, this dye cannot adopt a planar conformation, even
under conditions of restricted mobility, owing to steric clash-
es between H1 (and H8) from the acridine platform and the
second heterocycle.[10]
In comparison with that of its acridine-benzothiazolium
analogue (lmax =595 nm), the steady-state absorption spec-
trum of acridine-dimethylbenzimidazolium dye 3 has a
single band centred at around l=440 nm (see the Support-
ing Information). Particularly striking is the absence of de-
tectable fluorescence emitted by 3 in DMSO, methanol or
water and even in highly viscous glycerol solution. This may
be due to a larger interplanar angle between the acridine
and the heterocycle in the case of the benzimidazole dye
(when compared to the benzothiazole analogue). Indeed,
MM2 energy minimisation of compound 3 shows that bulky
N-methyl groups block the dye in a highly twisted confor-
mation (Figure 2). As demonstrated by Armitage et al., in-
terplanar angles exceeding 608 in unsymmetrical monome-
thine cyanine dyes correlate with a loss of conjugation be-
tween both heterocyclic platforms,[14] hence, a significantly
weaker (or the absence of) fluorescence. To restore the fluo-
rescence of the dye, we reasoned that formation of a cova-
Scheme 2. Synthetic route to the hexacyclic acridine dye 9. Reagents and
conditions: a) 2’-aminoacetanilide, PdACHTNUGTRENUNG(OAc)2, 2,2’-bis(diphenylphosphi-
no)-1,1’-binaphthyl (BINAP), Cs2CO3, toluene, 1008C; b) AcOH, 1408C,
24 h; c) MeI, 658C; d) SOCl2, DMF, 1008C, 10 min, then Et3N; e) NaPF6.
&
2
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis of a Hexacyclic Acridinium Cyanine Dye
FULL PAPER
Diarylamine 5 was first obtained by a palladium-catalysed
Buchwald–Hartwig coupling[15] between bromoacridone 4[11]
and commercially available N-acetyl-o-phenylene diamine.
Formation of 1-(N-methyl-benzimidazolium)-acridine 7 was
achieved by successive closure of the benzimidazole ring in
refluxing acetic acid[16] and quarternisation of the benzimi-
dazole in refluxing iodomethane. N-methylacridone 7 was
then converted into the corresponding 9-chloroacridinium 8
by treatment with a mixture of thionyl chloride and a few
drops of anhydrous DMF. Addition of triethylamine result-
ed in the in situ conversion of the unstable 9-chloroacridine
intermediate to the desired hexacyclic acridinium by nucleo-
philic attack of the 2-methylene-benzimidazole onto the ac-
ridineꢁs C-9 position. Dye 9 was finally isolated as a hexa-
fluorophosphate salt after treatment with sodium hexafluor-
ophosphate.
Figure 3. Electronic absorption (solid line) and emission (dashed line)
spectra of dye 9 (c=1.67 mm) in DMSO (black) and H2O (grey). Inset:
emission decay profiles recorded at lem =550 and lexc =467 nm.
Photophysical properties of dye 9: The electronic absorption
and emission spectra of hexacyclic acridine dye 9 were re-
corded in DMSO, MeOH, H2O and glycerol (Gly). The
main photophysical parameters of 9 in different solvents are
summarised in Table 1, and the electronic absorption and
mirror image of the absorption spectrum, with a maximum
emission measured at l=521 nm (Figure 3). Methanol (and,
to a lower extent, glycerol) causes only a slight blue-shift for
both absorption and emission spectra. In methanol, DMSO
and glycerol, dye 9 exhibits a remarkably high fluorescence-
emission quantum yield, comparable to that of fluorescein
in 0.1m NaOH used as a reference (Ff =0.95).[17] In all or-
ganic solvents, the excitation spectrum is identical to the ab-
sorption spectrum, and both the emission and excitation
spectra are independent of the detection/excitation wave-
length.
Conversely, in water, the shape of the absorption spec-
trum of dye 9 shows two absorption bands of similar ampli-
tude and centred at about l=470 and 496 nm, which is char-
acteristic of an aggregation/dimerisation process.[18] In com-
parison to organic solvents, the emission spectrum of 9 in
water is red-shifted, broadened and no longer excitation in-
dependent, suggesting the presence of more than one emis-
sive species.
Table 1. Photophysical properties of hexacyclic acridinium dye 9 in vari-
ous solvents[a]
.
Absorption
Emission
Fem
Solvent
DMSO
lmax [nm]
e [MÀ1 cmÀ1
]
lmax [nm]
t [ns]
496
468
492
464
498
471
497
468
8.1E+04
4.9E+04
9.0E+04
5.2E+04
4.3E+04
4.2E+04
6.9E+04
3.9E+04
521
0.93
6.5
MeOH
H2O
513
0.85
7.1
526[b]
517
0.56[b]
0.94
7.0/33
7.0
Gly
[a] The concentration of 9 in all solvents was 2 mm. [b] In water, dye 9
shows multiple emissive species; therefore, the reported values are the
apparent values measured at a dye concentration of 2 mm on a freshly
prepared solution by using l=467 nm excitation.
To further investigate this hypothesis, time-resolved emis-
sion spectroscopy (TRES) experiments of 9 in water were
carried out. While in DMSO, methanol and glycerol the ex-
cited-state decays are mono-exponential with lifetimes of
t=6.5, 7.1 and 7.0 ns, respectively, a bi-exponential decay
was obtained in water, corresponding to the co-existence of
two distinct emissive species. One is emitting at shorter
wavelengths and is characterised by a shorter lifetime (ca.
t=7.0 ns, Figure 4, black spectrum). This spectrum strongly
resembles that of the dye in organic solvent, in both the
spectral shape and lifetime. The second spectrum is red-
shifted (lmax =531 nm) and longer-lived (t=33 ns, Figure 4,
grey spectrum).
emission spectra in both DMSO and water are shown in
Figure 3. All the values reported in Table 1 and the spectra
shown in figures 3 and 4 were obtained from freshly pre-
pared solutions of dye 9. It is worth noting that in the case
of diluted aqueous solutions (<0.2 mm), time-dependent ag-
gregation occurs that has an influence on the spectroscopic
properties of 9. This effect will be discussed in detail in the
next section.
In DMSO, locked hexacyclic dye 9 has two close absorp-
tion maxima. The strong absorption maximum of the lowest
energy transition is located at l=496 nm (e=
8.2E4 MÀ1 cmÀ1), accompanied by a weaker shoulder at
about l=468 nm. Notably, the extinction coefficient at the
maximum is about tenfold higher than that of the non-cy-
clised analogue 3, probably owing to the smaller degree of
conjugation in 3. The emission spectrum of 9 in DMSO is a
Aggregation properties of dye 9 in water: Aggregation (for-
mation of dimers or larger multimeric complexes) of hydro-
phobic and polyaromatic small molecules in aqueous solu-
tions and its effect on their spectroscopic properties have
been extensively investigated in the literature.[19]
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

S. Ladame et al.
Figure 4. Time-resolved emission spectroscopy (TRES) experiments car-
ried out on a solution of dye 9 at a 2 mm concentration in water. Global
data analysis resolved two distinct emission spectra (shown in black and
grey; lexc =467 nm, acquisition time: 120 s for each decay). Inset: decay
profiles at lem =480 (black) and 680 nm (grey).
As demonstrated in the previous section, a solution of 9
in water (2 mm) displays emission from monomeric and ag-
gregated species. The contribution of the monomer/aggre-
gate emission can be estimated from time-resolved emission
spectra (Figure 4) as about 60 versus 40% (this was derived
from the area of emission peaksꢂexcited-state lifetime).
The overall apparent quantum yield for both of these spe-
cies is 0.56 (Table 1), although this figure, as well as the rela-
tive contribution from the two species, will be dependent on
both the concentration of 9 and on the excitation wave-
length. We have measured the apparent quantum yield of 9
in aqueous solutions at different concentrations ranging
from 0.2 mm to 0.12 mm (see Figure S2 in the Supporting In-
formation). It could be seen that the quantum yield decreas-
es dramatically from almost unity (at c=0.2 mm) to approxi-
mately 0.4 (at c=0.12 mm). At low concentrations (ca.
0.2 mm), fluorescent spectra of freshly prepared samples in
water resemble those in organic solvents, both in shape and
in quantum yield.
Figure 5. Top: excitation (black) and emission (grey) spectra recorded
after dilution of dye 9 in water (c=0.2 mm, 0.004 vol% DMSO). Inset:
plot of the fluorescence intensity ratio I/I0 (lexc =470and lem =515 nm)
versus time. Bottom: excitation (black) and emission (grey) spectra re-
corded following addition of increasing vol% of DMSO. Inset: plot of
the fluorescence intensity ratio I/I0 (lexc =470 and lem =515 nm, corrected
for dilution factor) versus vol% DMSO (lexc =470 and lem =550 nm).
These trends are caused by changes in both the absorption values and
the emission quantum yield of 9.
Additional complication occurs owing to the fact that the
aggregation process is time dependent. A solution of dye 9
in water (0.2 mm) was prepared from a stock solution in
DMSO (final sample contains 0.004 vol% DMSO), and the
fluorescence intensity was monitored at regular intervals
from mixing till 80 min (Figure 5, top). It is clear that a
time-dependent decrease in fluorescence intensity was ob-
served, with only about 10% of the initial fluorescence re-
maining after 60 min. This quenching could be due to the
slow aggregation of the hydrophobic dye in water. We estab-
lished that this aggregation can be reversed by successive
additions of DMSO. The decrease in fluorescence intensity
is therefore clearly reversible and is not caused by decompo-
sition and/or photobleaching of the sample. As shown in
Figure 5 (bottom), addition of DMSO to the previously
quenched solution leads to a significant increase of the fluo-
rescence intensity (increase of ca. 30-fold at 50 vol%
DMSO). It is important that this dramatic recovery of fluo-
rescence cannot be solely explained by the solvatochromic
properties of the dye, but instead must be due to aggrega-
tion of 9 in water and a consequent decrease in both the ab-
sorbance and the emission quantum yield. The maximum
fluorescence of the dye is achieved at about 70–80 vol%
DMSO (see Figure 6, inset), when the complete disaggrega-
tion is observed.
The time-resolved fluorescence experiments are even
more revealing. As shown in Figure 6, the emission decays
depend on the content of DMSO present in these binary
mixtures. In pure water, the emission of dye 9 is character-
ised by a bi-exponential decay profile, with the long-lived
component (tꢀ33 ns) observed in addition to the monomer
lifetime of about 7 ns. If the detection emission wavelength
is fixed to the red edge of the emission band (e.g., at l=
&
4
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis of a Hexacyclic Acridinium Cyanine Dye
FULL PAPER
emission band (lmax ꢀ535 nm, tꢀ33 ns), as seen in
a
DMSO/H2O binary-mixture experiment. At lower dye con-
centrations (<0.2 mm), aggregation still occurs, but the proc-
ess is somewhat more complex. While when freshly dis-
solved in water the absorption and emission spectra of the
dye resemble those observed in organic solvents (monomer-
like emission), aggregation, however, slowly occurs (>
60 min), as can be monitored by absorption and fluores-
cence spectroscopy (see the Supporting Information).
Thus, the aggregation of 9 in water is a function of the
dye concentration, the time after mixing and the composi-
tion of the solvent mixture. At very low concentrations of
the dye (ꢁ0.2 mm), the monomer-like emission is initially ob-
served almost exclusively with a single fluorescent decay
component of about t=7 ns, although aggregation then
slowly occurs, thus, reducing the fluorescence intensity by
near to 90%.
Figure 6. Decay profiles in DMSO/H2O binary mixtures at a constant
concentration of dye 9 (2 mm; lexc =467 and lem =650 nm, acquisition
time: 278 s for each decay). Inset: plot of the amplitudes versus volume
fraction of DMSO. The amplitudes are calculated by global fitting of the
decays using a bi-exponential function. Empty circles: the amplitude of
the longer-lived component (tꢀ33 ns); empty squares: the amplitude of
the shorter-lived component (tꢀ7 ns). Note: the relative contribution is
wavelength dependent and does not necessarily reflect the absolute con-
centrations of emitting species.
DNA-binding properties of dye 9: Since flat aromatic mole-
cules belonging to the families of acridines or cyanine dyes
are well-known for their ability to bind to DNA by either
intercalation[20] or groove binding,[21] the DNA-binding prop-
erties of dye 9 were also examined. First, the interaction be-
tween compound 9 and double-stranded DNA was moni-
tored by using UV/Vis absorption spectroscopy. Titration
experiments were carried out by adding increasing amounts
of a 17 bp fragment of duplex DNA to a solution of 9 in
TRIS.HCl buffer (c=10 mm, TRIS=tris(hydroxymethyl)a-
minomethane) also containing 100 mm KCl.[22] A strong in-
crease of the absorption band at longer wavelength (mono-
mer) at the expense of the band at shorter wavelength
(dimer) was observed upon addition of double-stranded
DNA, with two isosbestic points at l=521 and 457 nm (see
the Supporting Information). This result suggests the disas-
sembly of aggregates (dimers and maybe also larger aggre-
gates) of 9 upon binding to DNA.
650 nm), the long lived component (tꢀ33 ns) becomes pre-
dominant.
Figure 6 shows time-resolved fluorescence traces recorded
at l=650 nm, in which red-shifted long-lived dimer emission
dominates, especially at low DMSO concentrations. It can
be clearly seen that increased concentration of DMSO of up
to 70% causes the decays to become mono-exponential,
with the lifetime of 7 ns characteristic of the monomeric
emission. We note that at concentrations of >0.2 mm, the ag-
gregation process is very fast, and time-dependent changes
cannot be observed.
All these experiments demonstrate interconversion be-
tween two emissive species, which are characterised by dis-
tinct absorption and emission spectral features: a monomer-
ic form and an aggregated form. Aggregation (or dimerisa-
tion) is promoted in water and decreases as the percentage
of DMSO in water increases, whilst dye 9 exists almost ex-
clusively as a monomer in binary solutions containing up to
30% of water.
Finally, the effect of concentration of the dye on aggrega-
tion was recorded in water. The absorption titration experi-
ment shows a non-linear relationship between dye concen-
tration and absorbance intensity. At dye concentrations
>0.2 mm, aggegration occurs instantaneously. As a conse-
quence, the absorbance and emission no longer increase lin-
early with the dye concentration. Instead, the blue-shifted
band (characteristic of aggregated species) becomes pre-
dominant on the absorption spectrum, while fluorescence
red shifts and the intensity decreases. The broadening and
red-shifting of the emission spectrum are caused by the si-
multaneous decrease of the green monomeric emission band
(lmax ꢀ515 nm, t=7.0 ns) and the rise of a lower energy
To gain further insight into the DNA binding properties
of hexacyclic dye 9, fluorescence titration experiments were
carried out. For these experiments, the dye concentration
was set to 0.2 mm to prevent fast aggregation, as explained
above. According to the fluorescence-lifetime measure-
ments, at such low concentration the fluorescence decay of
the dye is initially mono-exponential and thus, dye 9 is pres-
ent almost exclusively in its monomeric form for the dura-
tion of the experiment (see the Supporting Information).
The excitation maximum is centred at l=491 nm, with a
weaker shoulder at about l=465 nm, while the emission
maximum is centred at l=516 nm (t=7.0 ns). Upon addi-
tion of duplex DNA, a slight red shift of the excitation (to
l=502 nm) and the emission maxima (to l=520 nm) was
observed (Figure 7 and the Supporting Information). More-
over, a noticeable narrowing of the spectral bandwidths sug-
gests a reduced vibrational freedom of the acridine cyanine
framework, which is associated with the binding process. As
the concentration of the DNA increases (up to 3.6 equiva-
lents), the integrated fluorescence intensity (lexc =470 nm)
increased up to 1.9-fold. The emission lifetime of the dye
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

S. Ladame et al.
Figure 7. Emission spectra recorded during titration of an aqueous solu-
tion of dye 9 (c=0.2 mm in 10 mm TRIS.HCl of pH 7.4 also containing
100 mm KCl) with a solution of 17 bp duplex DNA. The DNA concentra-
tion varies from 0 to 3.6 equivalents. Inset: plot of the integrated fluores-
cence spectra ratio Area/Area0 (lexc =470 nm) versus [DNA]/ACHTUNTRGNEUNG[dye 9]
ratio.
Figure 8. Results from confocal fluorescence imaging of dye 9 internal-
ised in MDA 231 cells (incubation time: 25 min) before (a,b) and after
(c,d) illumination (lexc =488 nm).
bound to the duplex DNA was also found to be 12 ns,
longer than the monomer lifetime in water (7 ns).
Considering the high structural similarity of dye 9 with
pentacyclic acridine RHPS4,[7] its potential as a G-quadru-
plex binding platform was also examined. A fluorescence ti-
tration experiment was carried out by using a solution of
parallel-stranded intramolecular DNA quadruplex.[23] As for
the titration with double-stranded DNA, a gradual increase
in fluorescence was observed upon addition of quadruplex
DNA (see the Supporting Information). However, no signif-
icant difference was detectable when compared to duplex
DNA, the increase in fluorescence being of comparable am-
plitude in both cases. These results suggest that 9 can also
bind to quadruplex DNA, but with no selectivity with
regard to the double-stranded form. Attempts to further
functionalise this hexacyclic heteroaromatic platform to
induce quadruplex versus duplex discrimination are current-
ly underway in our laboratory.
The absence of differences between the fluorescent spectra
of the dye localised in the nucleus and of those located in
the cytoplasm is consistent with a very low sensitivity of dye
9 to its environment (see Table 1).
Moreover, it is worth noting that the fluorescent spectra
recorded from all locations in live cells are sharp and well
defined, rather than broadened and red shifted (e.g., com-
pare spectra of 9 in water and DMSO, Figure 3). This is
characteristic of the monomer-like emission, which is ob-
served in organic solvents when bound to DNA or, perhaps,
bound to proteins. Thus, on the basis of these data, we can
exclude the possibility of dye 9 diffusing freely in the cytosol
of the cell. The presence of fluorescent hot spots within the
nucleus suggests the preferred accumulation of 9 in specific
organelles, which remain to be characterised, probably nu-
cleoli. We did not detect any differences between fluores-
cent spectra of 9 recorded from bright spots and the rest of
the nucleus, suggesting the similar mode of binding of 9 in
both of these environments. Interestingly, this molecule also
proved highly cytotoxic towards MDA 231 human breast
cancer cells. Further investigation to determine the mecha-
nism and specificity of this toxicity are underway.
Cellular uptake of dye 9: The remarkable fluorescence prop-
erties (quantum yield of almost one in organic solvents and
above 0.5 in water) of 9 and its ability to interact with DNA
led us to investigate the potential of this new dye as an in-
tracellular DNA stain. Human breast cancer MDA 231 cells
were stained for 25 min with a solution of dye 9 (c=2 mm in
media), washed and imaged by confocal microscopy. A fast
and efficient intracellular uptake of the fluorescent dye was
observed, with both the cytoplasm and the nucleus of the
cells being stained (Figure 8a–b). Under illumination, an ap-
parent increase of fluorescence intensity was observed, ac-
companied by an accumulation of the dye in the nucleus
(Figure 8c–d). The emission spectra of dye 9 (lexc =488 nm)
recorded in different regions of the living cells were similar
to that of the dye in aqueous solution (emission maximum
at about lmax =525Æ5 nm, see the Supporting Information).
Conclusion
In conclusion, we have reported the synthesis of a hexacyclic
acridinium cyanine dye 9, locked into a near-planar confor-
mation and absorbing in the green area of the spectrum.
Unlike its acridine-based monomethine cyanine dye ana-
&
6
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis of a Hexacyclic Acridinium Cyanine Dye
FULL PAPER
(Sigma) and 10% fetal bovine serum (Sigma). The cells were maintained
at 378C in a humidified 5% CO2 atmosphere. For the imaging experi-
ments, the cells were seeded at 104 cells per well in 0.2 mL of culture
medium in untreated 8-well coverglass chambers (Lab-TekTM, Nunc) and
allowed to grow to confluence for 24 h. The culture media was replaced
with the medium containing the dye and incubated for 5–25 min. Follow-
ing incubation, the chambers were washed twice with PBS and images
were collected at 208C.
logue 3, which is blocked into a twisted conformation and
non-fluorescent, compound 9 exhibits a high molar absorp-
tivity and remarkable fluorescence properties: a fluores-
cence quantum yield approaching unity in organic solvents
and remaining above 0.5 in water. While it exists mainly in
its monomeric form in organic solvents, dye 9 forms aggre-
gates in water or in binary aqueous-based mixtures. Interest-
ingly, these aggregates are characterised by a longer fluores-
cence lifetime (33 vs. 7 ns) but weaker fluorescence intensity
than those of the monomeric dye. The aggregation process
is both time and concentration dependent, but can, however,
be reversed by addition of DNA, thus, resulting in a twofold
increase in fluorescence. As anticipated considering its struc-
tural similarity with the well-known quadruplex binding
ligand RHPS4, dye 9 also recognises these alternative DNA
secondary structures, and studies aiming at designing quad-
ruplex-specific analogues based on this scaffold are currently
underway in our group. Taken altogether, the unique bright-
ness and DNA-binding properties of this hexacyclic mole-
cule make it perfectly suited for in vivo DNA sensing or
staining applications. This was confirmed by preliminary
studies demonstrating its fast and efficient cellular uptake.
Finally, the observed cytotoxicity of dye 9 on a line of
human breast cancer cells suggests that molecules based on
this hexacyclic scaffold and belonging to the family of acri-
dines (pharmacophores with potent antibacterial and/or an-
titumor activities)[6] could also find valuable applications as
anti-cancer agents.
General experimental procedures: Solvents were of HPLC or reagent
quality and purchased commercially. Starting materials were purchased
commercially and used without further purification. Compounds were
characterised by using 1H and 13C NMR spectroscopy. The spectra were
recorded on
a Bruker Avance DRX 400 spectrometer at 400 and
100.6 MHz, respectively. Chemical shifts are reported as d values (ppm)
with reference to the residual solvent peaks.
Acridinium monomethine cyanine dye (3): Phosphoryl chloride (130 mL,
1.39 mmol) was added dropwise to a solution of 1,2,4-triazole (0.320 g,
4.62 mmol) in dry acetonitrile (8 mL) at 08C (ice bath). Then, triethyla-
mine (0.775 mL, 5.55 mmol) was added and the mixture was stirred for
30 min. A solution of 10-methyl-9ACTHNUTRGNE(UNG 10H)-acridone (100 mg, 0.48 mmol) in
CH2Cl2 (5 mL) was added to the phosphoryl tristriazolide and the mix-
ture was refluxed for 48 h. A solution of freshly prepared 1,2,3-trimethyl-
benzimidazolium iodide[13] (150 mg, 0.5 mmol) in CH2Cl2 (5 mL) and trie-
thylamine (1 mL) was added to the above mixture and stirring was con-
tinued for further 48 h at 608C before the solvents were removed under
reduced pressure. The residue was then dissolved in CH2Cl2, washed with
water (100 mL), dried over magnesium sulphate and purified by flash
column chromatography (CH2Cl2/MeOH 98:2). The desired compound 3
was obtained pure as a dark-brown amorphous solid (14 mg, 4% yield).
1H NMR (CD3OD): d=8.29 (1H, d, J=8 Hz; ArH), 7.88–7.84 (2H, m;
ArH), 7.71–7.65 (3H, m; ArH), 7.60–7.59 (2H, m; ArH), 7.55 (1H, d, J=
8 Hz; ArH), 6.91–6.86 (2H, m; ArH), 6.70 (1H, s; CH), 3.83 (3H, s;
CH3), 3.34 ppm (6H, s; CH3); 13C NMR (CD3OD): d=151.5 (C), 147.4
(C), 141.2 (C), 139.6 (C), 132.2 (CH), 131.8 (CH), 131.2 (CH), 126.2
(CH), 125.7 (CH), 124.1 (CH), 122.1 (CH), 121.7 (C), 121.5 (CH), 118.7
(C), 114.9 (CH), 114.2 (CH), 112.1 (CH), 93.1 (CH), 33.5 (CH3),
31.4 ppm (CH3); HRMS (ESI): m/z calcd for C24H22N3: 352.181 [M]+;
found 352.179.
Experimental Section
Hexacyclic acridinium dye (9): Two to three drops of N,N-dimethylfor-
mamide were added to a solution of compound 7 (110 mg, 0.31 mmol) in
thionyl chloride (10 mL) and the reaction mixture was refluxed at 1008C
for 30 min. Solvents were then removed under reduced pressure, the un-
stable 9-choloroacridinium was dissolved in a mixture of anhydrous
CH2Cl2 (10 mL) and freshly distilled Et3N (1 mL) and the mixture was
stirred overnight at room temperature. The solvents were removed under
reduced pressure and the crude mixture was purified by flash column
chromatography (CH2Cl2, then CH2Cl2/MeOH 95:5). The residue was
then dissolved in a minimal volume of methanol and one equivalent of
sodium hexafluorophosphate was added. A bright yellow solid formed,
which was collected by filtration. Thus, the desired hexacyclic acridinium
was obtained pure as a hexafluorophosphate salt (35 mg, 24% yield).
1H NMR (CD3OD): d=8.14 (1H, d, J=8 Hz; ArH), 7.96 (1H, d, J=
8 Hz; ArH), 7.69–7.62 (2H, m; ArH), 7.57–7.52 (2H, m; ArH), 7.38 (1H,
t, J=8 Hz; ArH), 7.33 (1H, t, J=8 Hz; ArH), 7.15–7.11(2H, m; ArH),
7.04–6.97 ppm (2H, m; ArH); 13C NMR (CD3OD): d=143.8 (C), 140.9
(C), 139.9 (C), 139.2 (C), 133.9 (C), 133.9 (CH), 133.8 (CH), 133.0 (C),
127.3 (CH), 127.2 (C), 124.9 (CH), 124.7 (CH), 122.7 (CH), 116.2 (C),
115.3 (CH), 115.0 (CH), 112.2 (C), 110.9 (CH), 108.8 (CH), 105.6 (CH),
87.4 (CH), 33.6 (CH3), 29.2 ppm (CH3); HRMS (ESI): m/z calcd for
C23H18N3: 336.150 [M]+, found 336.148.
Spectroscopy: The photophysical studies were carried out in dimethyl
sulfoxide (DMSO), methanol (MeOH), glycerol (Gly) and H2O (all of
spectrofluorimetric grade). The samples were prepared by dilution of a
stock solution of dye 9 (c=5 mm in DMSO). Absorption spectra were re-
corded with an Agilent 8453 Diode Array Spectrophotometer. Molar ab-
sorption values (e) were calculated by applying the Lambert–Beer law to
the absorbance spectra (Amax <1.5) of the compounds.
Steady-state photoluminescence spectra were measured in right angle
mode with a HORIBA FluoroMaxꢃ-4. The concentration of the air-equi-
librated sample solutions was adjusted to obtain absorption values A<
0.1 at the excitation wavelength. Emission quantum yields were deter-
mined according to the approach described by Demas and Crosby[24] by
using fluorescein (Ff =0.95[17]) as standard. Ten and two millimeter path
length square optical Suprasil Quartz (QS) cuvettes were used for meas-
urements at room temperature of dilute and concentrated solutions, re-
spectively.
Fluorescence lifetimes were measured with HORIBA Jobin Yvon IBH
5000F time-correlated single-photon counting device by using pulsed
NanoLED excitation sources at l=404 and 467 nm. Analysis of the lumi-
nescence decay profiles against time was accomplished with the Decay
Analysis Software DAS6 provided by the manufacturer.
Fluorescence microscopy: Imaging was performed by using a confocal
laser scanning microscope (Leica TCS SP5), coupled to a CW argon-ion
laser (l=488 nm). The fluorescence emission of the dye in the cells was
spectrally dispersed by using a prism and detected by using a photomulti-
plier tube. Water immersion 63ꢂ objectives (NA=1.23) were used to
image. The human breast carcinoma cells (MD 231, ECACC) were
grown in Dulbeccoꢁs modified Eagleꢁs medium (DMEM, Gibco), supple-
mented with 100 UmLÀ1 penicillin (Sigma), 100 mgmLÀ1 streptomycin
Acknowledgements
The authors thank the Welcome Trust and EPSRC for funding, including
the EPSRC Career Acceleration Fellowship to MKK (U.K.).
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

S. Ladame et al.
[13] J. A. Zoltewicz, J. K. OꢁHalloran, J. Org. Chem. 1978, 43, 1713–
1718.
[14] G. L. Silva, V. Ediz, D. Yaron, B. A. Armitage, J. Am. Chem. Soc.
2007, 129, 5710–5718.
[15] I. C. F. R. Ferreira, M. J. R. P. Queiroz, G. Kirsch, Tetrahedron 2003,
59, 975–981.
[1] a) Fluorescent and Luminescent Probes for Biological Activity, 2nd
ed. (Ed.: W. T. Mason), Academic Press, London, 1999; b) L. D.
Lavis, R. T. Raines, ACS Chem. Biol. 2008, 3, 142–155; c) U. Resch-
Genger, M. Grabolle, S. Cavaliere-Jaricot, R. Nitschke, T. Nann,
Nat. Methods 2008, 5, 763–775.
[2] a) Chemistry of Heterocyclic Compounds, Vol 18: The Cyanine Dyes
and Related Compounds (Ed: F. M. Hamer), Wiley, London, 1964;
b) A. Mishra, R. K. Behera, P. K. Behera, B. K. Mishra, G. B.
Behera, Chem. Rev. 2000, 100, 1973–2011.
[3] P. Belmont, J. Bosson, T. Godet, M. Tiano, Anticancer Agents Med.
Chem. 2007, 7, 139–169.
[16] A. Gazit, K. Yee, A. Uecker, F. D. Bçhmer, T. Sjçblom, A. ꢄstman,
J. Waltenberger, G. Golomb, S. Banai, M. C. Heinrich, A. Levitzki,
Bioorg. Med. Chem. 2003, 11, 2007–2018.
[17] G. Weber, F. W. J. Teale, Trans. Faraday Soc. 1957, 53, 646–655.
[18] W. West, S. Pearce, J. Phys. Chem. 1965, 69, 1894–1903.
[19] a) B. A. Armitage, Top. Curr. Chem. 2005, 253, 55–76; b) U. Rçsch,
S. Yao, R. Wortmann, F. Wꢅrthner, Angew. Chem. 2006, 118, 7184–
7188; Angew. Chem. Int. Ed. 2006, 45, 7026–7030; c) T. Biver, A.
Boggioni, F. Secco, E. Turriani, M. Venturini, S. Yarmoluk, Arch. Bi-
ochem. Biophys. 2007, 465, 90–100; d) S. Ghosh, X. Q. Li, V. Stepa-
nenko, F. Wꢅrthner, Chem. Eur. J. 2008, 14, 11343–11357.
[20] a) L. R. Ferguson, W. A. Denny, Mutat. Res. 2007, 623, 14–23;
b) D. R. Boer, A. Canals, M. Coll, Dalton Trans. 2009, 399–414.
[21] a) G. Y. Guralchuk, A. V. Sorokin, I. K. Katrunov, S. L. Yefimova,
A. N. Lebedenko, Y. V. Malyukin, S. M. Yarmoluk, J. Fluoresc. 2007,
17, 370–376; b) A. L. Stadler, B. R. Renikuntla, D. Yaron, A. S.
Fang, B. A. Armitage, Langmuir 2011, 27, 1472–1479.
[4] M. S. GonÅalves, Chem. Rev. 2009, 109, 190–212.
[5] a) N. A. Karp, K. S. Lilley, Proteomics 2005, 5, 3105–3115; b) A.
Fegan, P. S. Shirude, S. Balasubramanian, Chem. Commun. 2008,
2004–2006; c) X. Peng, Z. Yang, J. Wang, J. Fang, Y. He, F. Song, B.
Wang, S. Sun, J. Qu, J. Qi, M. Yan, J. Am. Chem. Soc. 2011, 133,
6626–6635; d) K. J. Zanotti, G. L. Silva, Y. Creeger, K. L. Robert-
son, A. S. Waggoner, P. B. Berget, B. A. Armitage, Org. Biomol.
Chem. 2011, 9, 1012–1020.
[6] a) M. Wainwright, J. Antimicrob. Chemother. 2001, 47, 1–13;
b) W. A. Denny, Curr. Med. Chem. 2002, 9, 1655–1665.
[7] a) S. M. Gowan, R. Heald, M. F. G. Stevens, L. R. Kelland, Mol.
Pharmacol. 2001, 60, 981–988; b) M. K. Cheng, C. Modi, J. C. Cook-
son, I. Hutchinson, R. A. Heald, A. J. McCarroll, S. Missailidis, F.
Tanious, W. D. Wilson, J. L. Mergny, C. A. Laughton, M. F. G. Ste-
vens, J. Med. Chem. 2008, 51, 963–975.
[22] A stock solution (100 mm) of duplex DNA was prepared by anneal-
ing a DNA strand of sequence 5’-CCAGTTCGTAGTAACCC-3ꢁ
with its complementary strand in TRIS.HCl buffer (10 mm, pH 7.4)
also containing 100 mm KCl.
[8] J. Debray, W. Zeghida, M. Jourdan, D. Monchaud, M. L. Dheu-An-
dries, P. Dumy, M. P. Teulade-Fichou, M. Demeunynck, Org.
Biomol. Chem. 2009, 7, 5219–5228.
[9] M. J. Moore, C. M. Schultes, J. Cuesta, F. Cuenca, M. Gunaratnam,
F. A. Tanious, W. D. Wilson, S. Neidle, J. Med. Chem. 2006, 49, 582–
599.
[23] A stock solution (50 mm) of the c-kit quadruplex (5’-
CGGGCGGGCGCGAGGGAGGGG-3’) in TRIS.HCl buffer
(10 mm, pH 7.4) also containing 100 mm KCl was heated at 958C for
5 min and slowly cooled to room temperature over 8 h.
[24] G. A. Crosby, J. N. Demas, J. Phys. Chem. 1971, 75, 991–1024.
[10] T. Mahmood, A. Paul, S. Ladame, J. Org. Chem. 2010, 75, 204–207.
[11] R. A. Heald, M. F. G. Stevens, Org. Biomol. Chem. 2003, 1, 3377–
3389.
Received: March 9, 2012
[12] C. B. Carlson, P. A. Beal, Org. Lett. 2000, 2, 1465–1468.
Published online: && &&, 0000
&
8
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis of a Hexacyclic Acridinium Cyanine Dye
FULL PAPER
Organic Dyes
T. Mahmood, Y. Wu, D. Loriot,
M. Kuimova, S. Ladame* . . &&&& — &&&&
Closing the Ring to Bring Up the
Light: Synthesis of a Hexacyclic Acri-
dinium Cyanine Dye
A ring of light: The synthesis of a geo-
metrically constrained and near-planar
hexacyclic acridinium cyanine dye is
reported (see figure). Its remarkable
quantum yield close to unity in organic
solvents and above 0.5 in water com-
bined with efficient cellular uptake
and high affinity for DNA make it a
perfectly suited molecule for in vivo
sensing and imaging.
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!