Two-photon fluorescence imaging of RNA in nucleoli and cytoplasm in living cells based on low molecular weight probes

Article abstract of DOI:10.1016/j.dyepig.2013.12.005

Two low molecular weight, indole-based, mono-cationic probes, were designed and synthesized. According to their spectral response to ribonucleic acid in vitro and direct fluorescence imaging in living cell lines, the two compounds were identified as ribon

Full text of DOI:10.1016/j.dyepig.2013.12.005

Dyes and Pigments 103 (2014) 191e201
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Two-photon fluorescence imaging of RNA in nucleoli and cytoplasm
in living cells based on low molecular weight probes
,
,
Yong Liu a 1, Weijia Zhang a 1, Yuming Sun b, Guofen Song a, Fang Miao a, Fuqiang Guo a,
Minggang Tian a, Xiaoqiang Yu a , Jing Zhi Sun
,
c,
**
*
a Center of Bio & Micro/Nano Functional Materials, State Key Laboratory of Crystal Materials, Shandong University, Jinan 250100, PR China
b School of Information Science and Engineering, Shandong University, Jinan 250100, PR China
c MoE Key Laboratory of Macromolecule Synthesis and Functionalization, Department of Polymer Science and Engineering, Institute of Biomedical
Macromolecules, Zhejiang University, Hangzhou 310027, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Two low molecular weight, indole-based, mono-cationic probes, were designed and synthesized. Ac-
cording to their spectral response to ribonucleic acid in vitro and direct fluorescence imaging in living cell
lines, the two compounds were identified as ribonucleic acid-selective fluorescent turn-on probes with
large two-photon excited fluorescence action absorption cross-sections when binding to ribonucleic acid.
Moreover, both dyes with good membrane permeability have been successfully used to image ribonu-
cleic acid in nucleoli and cytoplasm in living cells by confocal and two-photon fluorescence microscopy.
Furthermore, the dyes possess good counterstain compatibility with Hoechst 33342, important deoxy-
ribonucleic acid-staining dyes for ribonucleic acidedeoxyribonucleic acid colocalization.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 21 August 2013
Received in revised form
28 November 2013
Accepted 6 December 2013
Available online 21 December 2013
Keywords:
Ribonucleic acid
Probes
Living cells
Two-photon fluorescence
Counterstain compatibility
Low molecular weight
1. Introduction
autoligation probes [9]; and (f) probes using fluorescent proteins as
reporters [10,11]. However, because most of the RNA fluorescent
In-vivo, fluorescent imaging of ribonucleic acid (RNA) in nucleoli
and cytoplasm is of great significance in biochemistry and
biomedicine [1,2]. For example, Knowles et al., used nucleic acid
(NA)stain SYTO 14 to visualize the translocation of endogenous RNA
in living cells, and they found that labeled RNA was distributed
nonrandomly as discrete granules in the neuronal processes [3].
Fluorescent imaging of RNA in parasites increases parasite detec-
tion, improves the spatial and temporal resolution of the parasite
under drug treatments, and resolves the problems of morpholog-
ical changing in an individual cell [4]. Previously, several classes of
molecular probes have been developed for RNA detection in living
cells, including (a) oligodeoxyribonucleotides (ODN) (Chart. S1)
probes [5]; (b) linear fluorescence resonance energy transfer (FRET)
probes [6]; (c) dual-labeled oligonucleotide hairpin probes (e.g.,
molecular beacons) [7]; (d) dual FRET molecular beacons [8]; (e)
probes have no membrane-permeability, they have to be injected
into a target living cell in a typical in-vivo fluorescent imaging
process [12e14]. Such a microinjection technology is destructive to
living cells and may cause biological malfunctions [12,15], it is
necessary to develop a low-molecular-weight fluorescent probe for
imaging RNA in living cells. Even though there are many
commercially available dyes with membrane-permeability for
various organelles such as the nucleus [16] and mitochondria [17],
preparation of RNA fluorescent probes with membrane perme-
ability is still a challenge. Some of the facts that restricted its
development include the diminished affinity to RNA than to
double-stranded deoxyribonucleic acid (DNA) [18], the non-
specificity in binding with proteins due to the hydrophobicity of
probe molecules. Moreover, there is limited knowledge of the
interaction mechanisms between RNA and fluorescent probes,
including outside, groove and intercalative binding [19], compared
with the wealth of papers on DNA biosensors. All increased the
difficulties to develop an RNA biosensor, which also promotes the
demand for a library of compounds. Li and coworkers screened a
large combinatorial library of 1336 fluorescent styryl molecules but
* Corresponding author. Tel.: þ86 0531 88366418; fax: þ86 0531 88364263.
** Corresponding author. Tel./fax: þ86 0571 87953734.
E-mail addresses: yuxq@sdu.edu.cn (X. Yu), sunjz@zju.edu.cn (J.Z. Sun).
Liu and Zhang made equal contributions to this work.
1
0143-7208/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.dyepig.2013.12.005

192
Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
ultimately obtained only three RNA probes [20]. In another work
just three dyes displaying high affinities to RNA were sifted from
125 fluorescent molecules [4]. Even though Molecular Probes Co.
offers a commercially available RNA probe “SYTO RNA-Select” [21]
for RNA imaging in living cells, its chemical structure has not been
described.
Ltd (Shanghai, China). Palladium (II) acetate and tri-o-tolylphos-
phine were purchased from J&K Chemical (Beijing, China). Tris and
PBS were purchased from Seikagaku Corporation (Japan). The sol-
vents used in the spectral measurement are of chromatographic
grade. The double-stranded DNA-specific dye 40,6-diamidino-2-
phenylindole (DAPI)/Hoechst 33342 and SYTO RNA-Select were
purchased from Molecular Probes. Calf thymus DNA and torula
yeast RNA, which were used as the model of DNA and RNA, and
ribonuclease A (RNase) were obtained from Sigma. Spectroscopic
measurements of were performed in acetonitrile/TriseHCl buffer
solution (tris: 10 mM, KCl: 100 mM) with pH 7.2. TLC analyses were
performed on silica gel plates and column chromatography was
conducted over silica gel (mesh 200e300), both of which were
obtained from Qingdao Ocean Chemicals.
Two-photon fluorescence microscopy (TPM) has advantages
over traditional techniques such as lower photodamage, reduced
photobleaching, higher detection sensitivity, as well as diminished
image distortion [22e27]. TPM has an intrinsically high axial res-
olution without the need of a confocal pinhole in the detection
path. Also the technique will reduce cell damage by using either
longer wavelength excitation light, which avoids damaging ultra-
violet (UV) [28], or blue excitation light. It also reduced out-of-focus
irradiations [29,30]. Therefore, TPM can be used for repetitive im-
aging of living cells without severely damaging the cellular vitality.
The fast development of two-photon excited fluorescence (TPEF)
probes to different targets in living cells has attracted a lot of
attention and many researchers [31e34], including us [35], have
contributed to this area. Amongst all of these, the only one on live-
cell RNA TPEF probes was given by Ohulchanskyy et al., who found
that the conventional NA dye 4-((1-methylbenzothiazolyliliden-2)
methyl)-1,2,6-trimethylpyridinium perchlorate (cyan 40) (Chart.
S2) could be used to image RNA in nucleoli in TPM [36]. However,
its two-photon excited fluorescence action absorption cross-
2.2. Measurements
Nuclear magnetic resonance spectra (1H and 13C) were obtained
on a Bruker Avanace 300/400 spectrometer. The HRMS spectra
were recorded on Agilent Technologies 6510 Q-TOF LC/MS or
ThermoFisher LCQ FLEET. The elemental analyses were performed
on a Vario EI III instrument. The UVevisibleenear-IR absorption
spectra of dilute solutions were recorded on a Cary 50 spectro-
photometer using a quartz cuvette having 1 cm path length. One-
photon fluorescence spectra were obtained on a HITACHI F-2700
spectrofluorimeter equipped with a 450-W Xe lamp. Two-photon
fluorescence spectra were recorded on a SpectroPro300i and the
pump laser beam comes from a mode-locked Ti: sapphire laser
system at the pulse duration of 200 fs, with a repetition rate of
76 MHz (Coherent Mira900-D). Confocal microscopic photos of
photostability were obtained with Carl Zeiss Microscopy LSM780,
The confocal microscopic image and differential interference
contrast (DIC) image were taken with a 488 nm laser, CD Spec-
trometer Jasco J-810.
Wide-field fluorescence microscopy images were acquired with
an Olympus IX71 inverted microscope coupling with a CCD and
display controller software. The fluorescence of four dyes and DAPI/
Hoechst 33342 were excited and collected through U-MNIBA3 and
U-MWU2, respectively. The confocal microscopic image and dif-
ferential interference contrast (DIC) image were taken with a
488 nm Arion laser. Fluorescence of INR1/INR2 were collected with
a beam splitter DM570 and BA510-540 nm bandpass emission filter
combination. All of TPM microscopic photos were obtained with
Olympus FV 300 Laser Confocal System with a 60ꢀ water objective
(N.A. ¼ 1.25) and photomultiplier tubes and Ti: sapphire laser
(Coherent) was used to excite the specimen at 800 nm. The total
power provided by laser source can be maintained stable and the
incident power was examined with Power Monitor (Coherent)
directly. A multiphoton emission filter (FF01-750; Semrock) was
used to block the IR laser.
sections (
F
ꢀ
d), a crucial parameter for TPEF probes, was not
given. Given the limited number of suitable dyes we were keen to
explore new RNA-selective biosensors that can be used for both
confocal microscopy and TPM in living cells.
In our previous work, we have focused on small-molecule
probes, carbazole derivatives, for selective imaging NA, cysteine
and lysosome in live-cells and deep tissues [35,37,38]. These dyes
possesses large
F
ꢀ
d, thus they can luminescence more intensely
under the some incident power. Despite the numerous desirable
properties of small-molecule NA probes have been designed by us
[37], many of them work well with DNA in plant tissue but are not
suitable for RNA TP imaging in living human cells. Among them,
possessing V-shaped conjugated structures inspired us to survey
RNA markers, especially TPEF live-cell RNA sensors among various
cationic salts with the particular chemical structure. The results
indicate that this idea should be rational. In this ongoing study we
synthesized and examined V-shape structures, namely 3,5-bis((E)-
2-(pyridin-4-yl)vinyl)-1H-indole
monoiodides
(INR1/INR2)
(Scheme 1).
2. Material and methods
2.1. Material
All chemicals used are of analytical grade, and 4-picoline, 1-
iodobutane, 1-iodohexane and 5-Bromo-1H-indole-3-carbal
dehyde were purchased from Sinopharm Chemical Reagent Co.,
TPA cross-sections have been measured using the two-photon
induced fluorescence method [40]. Fluorescein (pH ¼ 13, cyan
Scheme 1. The synthetic routes to INR1/INR2.

Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
193
diamond) in aqueous NaOH was used as the standard, whose two-
photon properties have been well characterized in the literature
[41], and thus cross-sections can be calculated by means of equa-
tion (1)
2.3.3. Synthesis of INR1/INR2
The condensation between 3 (0.36 g, 1.45 mmol) and N-(2-
hexyl)-4-methylpyridinium iodide (0.49 g, 1.57 mmol) or N-(2-
butyl)-4-methylpyridinium iodide (0.43 g, 1.57 mmol) happened
in methanol with about four drops of piperidine. The mixture was
stirred at 70 ꢁC for 4 h and then a precipitate was formed. This
precipitate was filtered and washed with dichloromethane and
little methanol as a red powder, which was then recrystallized from
methanol twice. The title product was obtained as a red solid.
INR1 (n ¼ 5): The yield: 0.50 g (68%), mp: 280e281 ꢁC. FT-IR
Fr cr nr Fs
Fs cs ns Fr
ds
¼
dr
(1)
where the subscripts s and r refer to the sample and the reference
materials, respectively. is the two-photon absorption (TPA) cross
d
(KBr)
968 (]CeH). 1H NMR (400 MHz, DMSO-d6),
g
(cmꢂ1): 3128 (NeH), 2927 (CH, aliphatic), 1593 (C]C),
sectional value, c is the concentration of the solution, n is the
refractive index of the solution, F is the two-photon excited fluo-
d
(ppm): 12.05 (d,
rescence integral intensity and
F is the fluorescence quantum yield.
J ¼ 2.0 Hz, 1H), 8.81 (d, J ¼ 6.8 Hz, 2H), 8.58 (d, J ¼ 5.5 Hz, 2H), 8.40
(s, 1H), 8.28 (d, J ¼ 16.0 Hz, 1H), 8.17 (d, J ¼ 6.8 Hz, 2H), 8.00 (d,
J ¼ 2.8 Hz, 1H), 7.61 (m, 5H), 7.34 (dd, J ¼ 23.3, 16.3 Hz, 2H), 4.43 (t,
TPA quantum yield measurement is quite difficult compared to the
well-established quantum yield measurement of one-photon
fluorescence. So, one might suppose that two quantum yields are
coincidental.
J ¼ 7.3 Hz, 2H), 1.91 (m, 2H), 1.31 (s, 6H), 0.87 (t, J ¼ 6.7 Hz, 3H). 13
C
NMR (300 MHz, DMSO-d6), d (ppm): 155.36, 150.49, 146.28, 144.37,
138.73, 137.05, 135.69, 133.82, 130.58, 126.33, 124.66, 123.00, 121.66,
121.07, 118.27, 115.01, 114.00, 60.05, 31.50, 31.30, 26.07, 22.83, 14.77.
HRMS: m/z (C28H30IN3), found: 408.2465[M ꢂ I]þ. Anal. calcd. for
2.3. Synthesis and characterization
C28H30IN3 (%): C 62.81, H 5.65, N 7.81; found: C 62.60, H 5.41, N 7.34.
2.3.1. N-(2-butyl/hexyl)-4-methylpyridinium iodide (1)
INR2 (n ¼ 3): The yield: 0.50 g (65%), mp: 295e296 ꢁC. FT-IR
4-picoline (4.65 g, 0.05 mol) and 1-iodobutane (11.04 g,
0.06 mol) or 1-iodohexane (12.72 g, 0.06 mol) were mixed and
stirred for 4 h, thereafter heated under refluxed for 2 h. After
cooling and filtrating, the final products were washed with ether.
The title product was obtained as a red viscous liquid.
(KBr)
g
(cmꢂ1): 3146 (NeH), 3029 (AreH); 2962 (CH, aliphatic),
1593e1639 (C]C), 971 (]CeH). 1H NMR (400 MHz, DMSO-d6),
d
(ppm): 12.06 (s, 1H), 8.82 (d, J ¼ 6.9 Hz, 2H), 8.56 (d, J ¼ 6.0 Hz,
2H), 8.41 (s, 1H), 8.29 (d, J ¼ 16.2 Hz, 1H), 8.18 (d, J ¼ 6.9 Hz, 2H),
8.00 (s, 1H), 7.71 (d, J ¼ 16.4 Hz, 1H), 7.64 (dd, J ¼ 8.5, 1.0 Hz, 1H),
7.57 (m, 3H), 7.38 (d, J ¼ 16.2 Hz,1H), 7.31 (d, J ¼ 16.4 Hz,1H), 4.45 (t,
N-(2-butyl)-4-methylpyridinium iodide (n ¼ 3): The yield:
13.16 g (95.0%), boiling point (bp) (760 mmHg): 214e215 ꢁC. FT-IR
J ¼ 7.3 Hz, 2H), 1.89 (m, 2H), 1.33 (m, 2H), 0.97 (t, J ¼ 6.7 Hz, 3H). 13
C
(KBr)
1517 (C]C, phenyl). 1H NMR (DMSO-d6, 400 MHz),
g
(cmꢂ1): 2873e2933 (CH, aliphatic), 1973 (C]N); 1470e
NMR (300 MHz, DMSO-d6), d (ppm): 154.87, 150.50, 145.23, 143.90,
d
(ppm): 8.99 (d,
138.20, 136.58, 134.75, 133.35, 130.14, 125.83, 124.30, 122.49, 121.06,
120.49, 117.75, 114.50, 113.51, 59.33, 32.95, 19.30, 13.86. HRMS: m/z
(C26H26IN3), found: 380.2119 [M ꢂ I]þ. Anal. calcd. for C26H26IN3
(%): C 61.53, H 5.16, N 8.28; found: C 61.39, H 5.25, N 8.31.
J ¼ 6.6 Hz, 2H), 8.00 (d, J ¼ 6.4 Hz, 2H), 4.57 (t, J ¼ 7.4 Hz, 2H), 2.61 (s,
3H), 1.86 (m, 2H), 1.26 (m, 2H), 0.89 (t, J ¼ 7.4 Hz, 3H). 13C NMR
(400 MHz, DMSO-d6),
d (ppm): 159.19, 144.07, 128.81, 60.02, 32.99,
22.01, 19.12, 13.81. HRMS: m/z (C10H16IN), found: 150.1309 [M ꢂ I]þ.
N-(2-hexyl)-4-methylpyridinium iodide (n ¼ 5): The yield:
2.4. Cell culture
13.72 g (90.0%), bp (760 mmHg): 215e216 ꢁC. FT-IR (KBr)
g
(cmꢂ1):
2857e2952 (CH, aliphatic), 1973 (C]N); 1469e1517 (C]C, phenyl).
Cancer cells (SiHa and HeLa) were cultured in Dulbecco’s
modified Eagle’s medium supplemented with penicillin/strepto-
mycin and 10% bovine calf serum in a 5% CO2 incubator at 37 ꢁC.
Cells PC3 and MDA-MB-231 were grown in Dulbecco’s modified
Eagle’s medium and 10% bovine calf serum in a 5% CO2 incubator at
37 ꢁC.
1H NMR (DMSO-d6, 400 MHz),
d
(ppm): 9.05 (d, J ¼ 6.6 Hz, 2H), 8.05
(d, J ¼ 6.4 Hz, 2H), 4.60 (t, J ¼ 7.4 Hz, 2H), 2.64 (s, 3H), 1.92 (m, 2H),
1.28 (s, 6H), 0.86 (t, J ¼ 6.8 Hz, 3H). 13C NMR (400 MHz, DMSO-d6),
d
(ppm): 159.15, 144.11, 128.81, 60.21, 31.03, 30.98, 25.40, 22.30,
21.98, 14.28. HRMS: m/z (C12H20IN), found: 178.1579 [M ꢂ I]þ.
2.3.2. Synthesis of (E)-5-(2-(pyridin-4-yl) vinyl)-1H-indole-3-
carbaldehyde (3)
2.5. Live-cell imaging
5-Bromo-1H-indole-3-carbaldehyde 2 (1.12 g, 5.00 mmol) was
added to a high pressure tube with a mixture of palladium(II)
acetate (0.11 g, 0.50 mmol), tri-o-tolylphosphine (0.46 g,
1.50 mmol) and then to which was added the solvent pair trie-
thylamine (14 mL, 0.10 mol)/tetrahydrofuran (42 mL, 0.52 mol)
and 4-vinylpyridine (2 mL, 0.02 mol). The tube was sealed after
bubbling for 15 min with nitrogen. After keeping the system under
w108 ꢁC for three days, the precipitate was collected and washed
with water and dichloromethane. The title product was obtained
as a yellow solid. The yield: 0.62 g (50.0%), melting point (mp):
For living cells imaging experiment of INR1/INR2, cells were
incubated with 10 m
M INR1/INR2 in PBS (pH 7.2) for 1.5 h at 37 ꢁC.
After rinsing with PBS twice, cells were imaged immediately. INR1/
INR2 were dissolved in DMSO at a concentration of 1 mM and
Hoechst 33342 was prepared as 1 mM aqueous solution.
For wide-field fluorescence microscopy, confocal fluorescence
microscopy and TPM imaging of cells stained with INR1/INR2:
Living SiHa cells were stained with 10 mM INR1/INR2 for 1.5 h at
ambient temperature and then imaged with fluorescence micro-
scopy. For cells counterstain experiment with the dyes and Hoechst
285e286 ꢁC. FT-IR (KBr)
(eC]O), 1596e1641 (C]C), 958 (]CeH). 1H NMR (DMSO-d6,
300 MHz),
g
(cmꢂ1): 3247 (NeH), 3022 (AreH), 1640
33342: Living SiHa cells were stained with 10
After rinsing with PBS twice, the same sample was stained with
5 mM Hoechst 33342 for 30 min and then imaged with wide-field
mM dyes for 1.5 h.
d
(ppm): 12.21 (s, 1H), 9.97 (s, 1H), 8.53 (d, J ¼ 6.0 Hz,
2H), 8.33 (d, J ¼ 12.9 Hz, 2H), 7.70 (d, J ¼ 16.5 Hz, 1H), 7.64 (dd,
J ¼ 8.4 Hz, 1.5 Hz, 1H), 7.60 (d, J ¼ 6.0 Hz, 2H), 7.54 (d, J ¼ 8.4 Hz,
1H), 7.19 (d, J ¼ 16.2 Hz, 1H). 13C NMR (400 MHz, DMSO-d6),
fluorescence microscopy.
2.6. RNase digest test of fixed cells
d
(ppm): 185.52, 150.40, 145.13, 139.69, 137.74, 134.58, 131.03,
125.01, 124.47, 122.86, 121.19, 120.93, 118.87, 113.36. HRMS: m/z
Before cells digest test, cultured cells grown on glass coverslips
were pretreated according to the following procedure: cells were
(C16H12N2O), found: 249.1041 [M þ H]þ.

194
Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
first fixed by 4% paraformaldehyde for 30 min and then per-
meabilized by 0.5% Trition X-100 for 2 min at ambient temperature.
DAPI was prepared as 1 mM aqueous solution. For RNase digest test,
([NA]:[dye] < 250:1) and decrease regularly until stable ([NA]:[dye]
ranges from 700:1 to 1050:1) (Fig. 2a). Furthermore, TPEF titrations
of dyes were carried out at 800 nm which is an optimal output
wavelength from a commercial mode-locked Ti: sapphire laser
source. In the two-photon fluorescence titration, it was found that
the fluorescence intensity of INR1 increases in the presence of RNA
or DNA, and that the fluorescence in the presence of RNA is stronger
than that in the presence of DNA (Fig. 2b). A similar trend is
observed in the one- and two-photon fluorescence of INR2 with
addition of NA (Fig. 2c and d). The dyes preferring RNA to DNA
should be superior to those, such as 2,7-bis(1-hydroxyethyl-4-
vinylpyridium iodine)-N-ethylcarbazole (2,7-9E-BHVC) (Chart. S3)
[39], with identical response to RNA and to DNA for RNA-selective
probes.
two sets of pretreated SiHa cells were stained with 10 mM INR1/
INR2 for 30 min. After rinsing with PBS twice, a total 1 mL PBS (as
control experiment) was added into a set of cells and 25 mg/mL
DNase-Free RNase (GE) was added into the other set of cells, and
then two sets of cells were incubated at 37 ꢁC in 5% CO2 for 2 h. After
rinsing with PBS twice, both two sets of cells were imaged with
wide-field fluorescence microscopy. In addition, the RNase digest
test of cells stained with 1
comparison.
mM DAPI was also carried out for
2.7. Cell viability evaluated by MTT assay
Two factors can be used to explain the mechanism of the light-
up effect upon binding to RNA or DNA. Firstly, dyes that are in a
twist intramolecular charge transfer (TICT) state whose non-
radiative process quenches the fluorescence will be non-emissive;
Viability of the cells was assayed using cell proliferation Kit I
with the absorbance of 492 nm being detected using a Perkine
Elmer Victor plate reader. Five thousand cells were seeded per well
in a 96-well plate. After overnight culture, various concentrations of
INR1/INR2 were added into the 96-well plate. After 2 h treatment,
inversely, dyes for which the formation of
a TICT state is
restricted become brightly luminescent [42]. For indole derivatives
reported here, the intramolecular charge transfer (ICT) state could
occur between the electron donor of the indole moiety and the
electron acceptor of the pyridinium cation. Once in water, the larger
dipole moment in the excited state than that in the ground state
interacts strongly with the polar solvent, which could lead to
charge separation resulting in the formation of a TICT state. How-
ever, the formation of a TICT state is restricted when the dyes are
protected from water in the grooves of RNA or DNA, and finally the
fluorescence of the dyes will be restored. On the other hand, the
weak emission of the dyes in water can be attributed to the rapid
non-radiative decay that results from the torsional motion of flu-
orophore responsible for the low quantum yield and from colli-
sional interactions [43]. When binding to nucleic acids, the
torsional motion of vinyl groups can be restricted by the steric ef-
fect derived from the static interactions between the cationic N-
pyridinium and the anionic phosphate of nucleic acids, thus the
fluorescence intensity can increase. Their one- and two-photon
fluorescence intensities are high in organic solvents and very low
in triseHCl buffer and aqueous solution (Fig. S1). The appreciable
solvent effect implies that the ICT state is predominant in the
photophysical behavior of INR1/INR2, which is similar to our other
carbazole derivatives [44]. It is well known that the concentration-
quenching effect is common for many aromatic compounds due to
the formation of sandwich-shaped excimers and exciplexes aided
by the collisional interactions between the aromatic molecules in
the excited and ground states [45]. However, the formation of a
TICT state is restricted when the dyes are protected from water in
the grooves of RNA or DNA, and finally the fluorescence of dyes will
be restored.
20
bromide (MTT) solution (5 mg/mL in phosphate buffer solution)
was added into the each well. After 4 h incubation at 37 ꢁC, 200
DMSO was added to dissolve the purple crystals. After 20 min in-
cubation, the optical density readings at 492 nm were taken using a
plate reader.
mL of 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium
mL
3. Results and discussion
3.1. Preparation of probes
INR1/INR2 were synthesized by Knoevenagel condensation re-
actions between N-(2-butyl/hexyl)-4-methylpyridinium iodide and
a substituted indole-3-carboxaldehyde using piperidine as catalyst.
The intermediate 3 was obtained by Heck coupling reaction be-
tween 5-bromo-1H-indole-3-carbaldehyde with 4-vinylpyridine.
The separated INR1/INR2 were characterized by various spectro-
scopic methods, from which satisfactory analysis data were ob-
tained (Supporting Information).
3.2. RNA and DNA titrations
In the absorption spectrum of INR1 (Fig. 1), the single peak at
425 nm, appears as a double-peak (370 nm and 467 nm for RNA,
370 nm and 464 nm for DNA) when it is mixed with NA. This
suggests that strong interactions between dye and NA. Fluores-
cence titration experiments were then performed to study the af-
finity of probes to RNA and DNA (Fig. 2). The fluorescence
intensity of INR1 increase significantly in the low ratio region
Fig. 1. The UVevis absorption (a and b) of INR1/INR2 in buffer (black), RNA (green) and DNA (red). Buffer, 10 mM TriseHCl, 100 mM KCl, pH 7.2. Compound concentration: 2.0
mM.
Ratio (phosphate of NA/dye): 1000:1 for INR1/INR2. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
195
Fig. 2. One- (a and c) and two-photon (b and d) excited fluorometric titration of dyes with the addition of RNA and DNA. Each point shown in b and d has an error of 8%. Inset: the
corresponding fluorescent spectra of the dyes with RNA or DNA. Compound concentration: 2.0 mM. Excitation wavelength (Ex): a and c, 465 nm; b and d, 800 nm.
3.3. Circular dichroism (CD)
may affect the molecular conformation and bring about an asym-
metric chiral center in INR1/INR2, resulting in optical activity [47].
Compared with RNA, the induction of INR1/INR2 DNA should be
weaker (Fig. 3 inset). The results show that the affinity between
INR1/INR2 and RNA is stronger than that of DNA, thus INR1/INR2
may preferentially interact with RNA within cells.
To further explain the mechanism of the light-up effect upon
binding to RNA, the interactions between INR1/INR2 and RNA/DNA
were also investigated by induced circular dichroism (ICD) spectra.
As shown in Fig. 3, CD signal of RNA and DNA appear in range of
200e300 nm, and plus and minus two signals confirm spiral chain
characteristics of RNA [46]. Due to induction of RNA, INR1/INR2,
two compounds without optical activity, present optically activity.
Moreover, there are negative Cottons effect in the electronic ab-
sorption (410e500 nm) region of INR1 and INR2, implying that the
electron transition easily absorbs right-circularly polarized light.
Meanwhile Cotton peaks at 465 (e) nm corresponded to UV ab-
sorption peaks at 467/465 nm of INR1/INR2 bound with RNA. It can
be assumed that the groove binding between INR1/INR2 and RNA
3.4. One-photon fluorescent imaging
One-photon fluorescent spectra of INR1/INR2 in buffer, RNA and
DNA solutions were studied before imaging. INR1 is weakly fluo-
rescent in the triseHCl buffer, whereas the fluorescence intensity
shows an approximately 15-fold enhancement in the presence of
RNA and 6-fold in the presence of DNA (Fig. 4a). A similar trend is
observed in the case of INR2 (see Fig. 4b). This indicates that these
Fig. 3. (a) and (b): CD spectra of the RNA, DNA, INR1/INR2, INR1/INR2 þ RNA and INR1/INR2 þ DNA; Inset: CD spectrum of RNA and INR1/INR2 þ DNA. Concentration: 40
Ratio (phosphate of NA/dye): 400:1 for INR1, 150:1for INR2.
mM.

196
Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
Fig. 4. One-photon excited fluorescence (a and b) spectra of INR1/INR2 in buffer (black), RNA (green) and DNA (red). Buffer, 10 mM TriseHCl, 100 mM KCl, pH 7.2. Compound
concentration: 2.0 M. Ratio (phosphate of NA/dye): 1000:1 for INR1/INR2. (c): confocal fluorescence images of live-cells (SiHa, HeLa, PC3 and MDA-MB-231) stained with 10 mM
m
dyes for 1.5 h. Ex: 488 nm; Detection wavelength: 510e540 nm; Bar ¼ 20
mm. (For interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article.)
Fig. 5. Wide-field fluorescence images of SiHa cells stained with 10
INR2, respectively. b and e: fluorescence of Hoechst 33342. c and f: merged picture. Ex (INR1/INR2): 460e490 nm; Detection wavelength: >510 nm; Ex (Hoechst 33342): 330e
385 nm; Detection wavelength: >410 nm; Bar ¼ 20 m.
mM dyes for 1.5 h followed by staining with 5 mM Hoechst 33342 for 20 min. a and d: fluorescence of dyes INR1/
m

Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
197
Fig. 6. The RNase digest experiment. Wide-field fluorescence imaging of SiHa cells stained with DAPI (a and b), INR1 (c and d) and INR2 (e and f) before (a, c and e) or after (b, d and
f) treatment with RNase (30 M, dyes; 1 M, DAPI; Incubation time: 30 min. Ex (INR1/INR2): 460e490 nm, Detection wavelength: >510 nm;
g/ml) for 2 h at 37 ꢁC. Conditions: 10
Ex (DAPI): 330e385 nm, Detection wavelength: >410 nm; Bar ¼ 20 m.
m
m
m
m
dyes have higher one-photon fluorescence responses to RNA than
to DNA, and therefore have potential in imaging the RNA distri-
bution in living cells.
and RNA distribution simultaneously. Fig. 5 shows the counterstain
result of dyes and Hoechst 33342. The green fluorescence from
nucleoli and cytoplasm is easily discriminated from the nuclear
zone with blue fluorescence in live SiHa cells, which indicates a
good counterstain compatibility with Hoechst 33342. The charac-
teristics of these dyes could help to image RNA distribution in
relation to DNA in living cells and probably to reveal different
patterns of RNAeDNA colocalization which are cell type dependent.
To further confirm the selectivity of dyes to RNA, the digest test
of ribonuclease (RNase) was performed (Fig. 6), in which RNA was
hydrolyzed while DNA kept intact. Fixed-permeabilized HeLa cells
were used in this experiment, with DAPI (Chart. S4) in the control
group. After treatment with RNase, the fluorescence in cytoplasm
and nucleoli dramatically diminished and tended to redistribute to
the nucleus (Fig. 6d and f) in contrast to the untreated sample
(Fig. 6c and e). However, there was no apparent effect on the
staining result of DAPI (Fig. 6a and b), considering that the double-
stranded DNA is a more preferable target for DAPI than RNA [20].
This result indicates that INR1/INR2 prefer RNA to DNA in the
complex internal environment of cells.
The cytotoxicity of INR1/INR2 were evaluated using a 3-(4,5-
dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium bromide (MTT)
assay. The results indicate that the probes generally present low
toxicity for living cell imaging (e.g., HeLa cell) under the conditions
applied (Fig. S2). Four live-cell lines (SiHa, HeLa, PC3 and MDA-MB-
231) were incubated. The fluorescent imaging was carried out with
confocal microscopy and wide-field fluorescence microscopy
(Fig. S3), using 488 nm Arþ laser and mercury lamp (U-MNIBA3),
respectively. The differential interference contrast (DIC) images
served as proof that the cells were viable throughout the imaging
experiments. According to Fig. 4c, the amounts of nucleoli in SiHa
and HeLa cells are more than that in PC3 and MD-MBA-231 cells
and the distributions of nucleoli in the four kinds of cells are
significantly different, indicating that there is diverse transcrip-
tional activity in different cell lines. The green fluorescence mainly
localizes in nucleoli, a dense region of rRNA, and cytoplasm where
mRNA is likely to be present. Moreover, it is evident that the po-
sition, shape and amount of the distinguishable nucleoli labeled by
INR1/INR2 are consistent with the dense, dark-phase region of the
nucleus in DIC photos which definitely represents nucleoli.
A good compatibility of RNA sensors with the classic DNA-
staining dye Hoechst 33342 would be helpful in imaging DNA
3.5. Two-photon fluorescent imaging
The influence of DNA and RNA on their selectivity in TPEF
properties was examined before imaging RNA in living cells with

198
Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
Fig. 7. Two-photon photophysical properties and fluorescent images of INR1/INR2. a and c: Two-photon excited fluorescence spectra of dyes in the absence and in the presence of
DNA or RNA. Error limit: 8%. b and d: Two-photon action absorption spectra of dyes in DNA and RNA; Error limit: 20%. f: TPM fluorescence images of live-cells (SiHa and HeLa)
stained with 10
mM dyes for 1.5 h. Compound concentration: 2.0
m
M. Ratio (phosphate of NA/dye): 1000:1 for INR1/INR2. Ex: 800 nm, Bar ¼ 20
mm.
TPM. As was revealed from the absorption experiments, the linear
absorption wavelengths of INR1 þ RNA and INR2 þ DNA are in the
range of 350e500 nm, hence their two-photon absorption experi-
ments can be excited by a sapphire laser source, Ti, which emits
lights with the wavelength of 760e910 nm. Moreover, the TPEF of
INR1 þ RNA is far higher than that of INR1, and is significantly
stronger than INR1 þ DNA. The intensity of INR1 þ RNA at 540 nm
is 9 times as strong as that of INR1 at the same wavelength (Fig. 7a).
The corresponding ratio of INR2 þ RNA is 11 (Fig. 7c). From Figs. 7b
and RNA are 7.11 GM and 12.53 GM at 810 nm, much larger
comparison with commercial NA dyes, such as DAPI-DNA (2.18 GM)
[22] and EBDNA (0.91 GM) [43]. According to our data, the of
INR1 at 800 nm is 0.66, 5.71 and 8.93 GM in triseHCl buffer, DNA
and RNA, respectively (Fig. 7e) which suggests a ca. 8- and 13-fold
enhancement compared with that in triseHCl buffer. Similarly, the
F
ꢀ
d in
F
ꢀ
d
F
ꢀ
d of INR2 in presence of RNA shows ca. 14-fold enhancement.
These all serve as evidence that the dyes can be used as two-photon
fluorescence “light-switch” for RNA.
and d, the maximum
F
ꢀ
d
of INR1/INR2 in DNA and RNA solutions
of INR1 in the presence of DNA
The TPEF imaging of INR1/INR2 (Fig. 7f) was completed at
excitation wavelength 800 nm, and the bright fluorescence was
are at approximate 800 nm.
F
ꢀ
d
Fig. 8. Two-photon I and II (aec) and confocal fluorescence I and II (def) images of Prefixed cells (HeLa). Cells stained with 5
fluorescence): 488 nm; Detection wavelength: 510e540 nm; Bar ¼ 20 m.
mM dyes for 30 min. Ex (TP): 800 nm; Ex (confocal
m

Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
199
Fig. 9. The left picture: the optical section from video S1. A1 and B1: two areas of nucleoli; A2 and B2: two areas of dark nuclear; A3 and B3: two areas of cytoplasm. A and B stand
for the two cells, respectively. The right picture: changes of fluorescence intensity of INR1 in cell A and B with depths at the same incident power. Red, green and black lines
correspond to nuclei nucleoli and cytoplasm, respectively. A1, A2, A3: A; B1, B2, B3: B. (For interpretation of the references to color in this figure legend, the reader is referred to the
web version of this article.)
found localized in nucleoli and cytoplasm in different cell types,
although different areas exhibit disparity of fluorescent intensity.
Consistent with results in confocal microscopy and TPM, INR1 ex-
hibits better imaging performances than INR2. Additionally, ac-
cording to one- and two-photon bio-imaging (Fig. 8), fixed cells
show more nucleoli with significantly different distribution
compared to that of living cells, which reflect the difference of RNA
dynamics between live-cells and prefixed cells. Furthermore, a
video of confocal microscopic imaging of INR1 including 12 optical
sections under the same imaging conditions at different depths has
been recorded (see Video S1). In the 12 sections, we randomly
selected two nucleoli areas (A1 and B1), two dark nuclear areas (A2
and B2) and two cytoplasm areas (A3 and B3) (Fig. 9). For identical
measurement area, the relationship between the fluorescent in-
tensity of INR1 and imaging depths were plotted in the right pic-
increases dramatically with the increase of depth and then de-
creases to nearly zero in deeper section, and there is hardly any
fluorescence in nuclear area. Fluorescence signal may reveal RNA
distribution in nucleoli, the sites of rRNA synthesis, and cytoplasm
(rRNA, mRNA and tRNA).
Supplementary video related to this article can be found at
http://dx.doi.org/10.1016/j.dyepig.2013.12.005.
3.6. The advantages of INR1/INR2
First, we have confirmed that the dyes INR1/INR2 can achieve
one- and two-photon fluorescence imaging by means of 488 and
800 nm through the above one- and two-photon imaging experi-
ments. Compared to INR1/INR2, commercially available RNA probe
SYTO RNA-Select cannot realize TP fluorescence imaging in living
and fixed cells at 800 nm (Fig. S4), and one-photon images with
ture of Fig. 9 (cell A and B), in which from 0 to 11
mm, fluorescence
intensity in nucleoli (A1 and B1) and cytoplasm (A3 and B3)
Fig. 10. Comparison of photobleaching of INR1/INR2 and SYTO RNA-Select in confocal fluorescence microscopy imaging. Cells stained with 10
M dyes for 30 min (SYTO RNA-Select); Ex: 488 nm; Detection wavelength: 510e540 nm; Bar ¼ 20 m.
mM dyes for 1.5 h (INR1/INR2) and
1
m
m

200
Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
[5] Honcharenko D, Zhou C, Chattopadhyaya J. Modulation of pyrene fluorescence
in DNA probes depends upon the nature of the conformationally restricted
nucleotide. J Org Chem 2008;73:2829e42.
[6] Alabi CA, Love KT, Sahay G, Stutzman T, Young WT, Langer R, et al. FRET-
labeled siRNA probes for tracking assembly and disassembly of siRNA nano-
complexes. ACS Nano 2012;6:6133e41.
[7] Kam Y, Rubinstein A, Nissan A, Halle D, Yavin E. Detection of endogenous K-
ras mRNA in living cells at a single base resolution by a PNA molecular beacon.
Mol Pharm 2012;9:685e93.
[8] Santangelo PJ, Nix B, Tsourkas A, Bao G. Dual FRET molecular beacons for
mRNA detection in living cells. Nucleic Acids Res 2004;32:e57.
[9] Sando S, Kool ET. Imaging of RNA in bacteria with self-ligating quenched
probes. J Am Chem Soc 2002;124:9686e7.
[10] Paige JS, Wu KY, Jaffrey SR. RNA mimics of green fluorescent protein. Science
2011;333:642e5.
[11] Bann DV, Parent LJ. Application of live-cell RNA imaging techniques to the
study of retroviral RNA trafficking. Viruses 2012;4:963e79.
[12] Bao G, Rhee1 WJ, Tsourkas A. Fluorescent probes for live-cell RNA detection.
Annu Rev Biomed Eng 2009;11:25e47.
[13] Kubota T, Ikeda1 S, Yanagisawa H, Yuki M, Okamoto A. Sets of RNA repeated
tags and hybridization-sensitive fluorescent probes for distinct images of RNA
in a living cell. PLoS One 2010;9:e13003.
INR1/INR2 are more distinct than that with SYTO RNA-Select in
living cells (Fig. 10a).
Second, under the high-intensity illumination conditions used
for fluorescence microscopy, the irreversible destruction of the
excited fluorophore (photobleaching) often becomes the factor
limiting fluorescence detectability. Thus, photostability is one of
the most important criteria for developing fluorescent imaging
agents. Continuous scanning by confocal microscope (Zeiss laser
scanning confocal microscope LSM780) was used to quantitatively
investigate the photostability of INR1/INR2 and SYTO RNA-Select.
Two dishes of SiHa cells subcultured from the same source were
stained with 10 mM INR1/INR2 for 1.5 h and 1 mM SYTO RNA-Select
for 30 min (According to technology of Molecular Probes Co),
respectively. With the help of a power meter, excitation power from
488 nm laser line of the microscope was unified at 60 mW and used
to irradiate the INR1/INR2 and SYTO RNA-Select. Within 300 s, the
fluorescence signals of INR1 and INR2 decreased w8% and 28%,
respectively. In contrast, SYTO RNA-Select exhibited significant
photobleaching with only w45% signal intensity remaining
(Fig. 10b). The results indicate that under the applied conditions
INR1/INR2 display a much better photostability than SYTO RNA-
Select.
[14] Catrina IE, Marras SAE, Bratu DP. Tiny molecular beacons: LNA/2-O-methyl
RNA chimeric probes for imaging dynamic mRNA processes in living cells. ACS
Chem Biol 2012;7:1586e95.
[15] Santangelo P, Nitin N, Bao G. Nanostructured probes for RNA detection in
living cells. Ann Biomed Eng 2006;34:39e50.
[16] Krishan A, Dandekar PD. DAPI fluorescence in nuclei isolated from tumors.
J Histochem Cytochem 2005;53:1033e6.
[17] Pendergrass W, Wolf N, Poot M. Efficacy of MitoTracker GreenÔ and
CMXrosamine to measure changes in mitochondrial membrane potentials in
living cells and tissues. Cytom A 2004;61:162e9.
4. Conclusions
[18] Yu J, Parker D, Pal R, Poole RA, Cann MJ. A europium complex that selectively
stains nucleoli of cells. J Am Chem Soc 2006;128:2294e9.
In summary, two low-molecular-weight fluorescent probes
(INR1/INR2) for detection of RNA in living cells are presented. The
[19] Kumar CV, Asuncion EH. DNA binding studies and site selective fluorescence
sensitization of an anthryl probe. J Am Chem Soc 1993;115:8541e53.
[20] Li Q, Kim Y, Namm J, Kulkarni A, Rosania GR, Ahn YH, et al. RNA-selective, live
cell imaging probes for studying nuclear structure and function. Chem Biol
2006;13:615e23.
[21] Haugland RP. In: Spence MTZ, editor. A guide to fluorescent and labeling
technologies: the handbook. 10th ed. 2005. pp. 710e1.
[22] Li L, Shen X, Xu Q-H, Yao SQ. A switchable two-photon membrane tracer
capable of imaging membrane-associated protein tyrosine phosphatase ac-
tivities. Angew Chem Int Ed Engl 2013;52:424e8.
[23] Oheim M, Michael DJ, Geisbauer M, Madsen D, Chow RH. Principles of two-
photon excitation fluorescence microscopy and other nonlinear imaging ap-
proaches. Adv Drug Deliv Rev 2006;58:788e808.
[24] Dedov VN, Coxb GC, Roufogalis BD. Visualisation of mitochondria in living
neurons with single- and two-photon fluorescence laser microscopy. Micron
2001;32:653e60.
[25] Cho EE, Drazic J, Ganguly M, Stefanovic B, Hynynen K. Two-photon fluores-
cence microscopy study of cerebrovascular dynamics in ultrasound-induced
blood-brain barrier opening 2011;31:1852e62.
[26] Hoptand A, Neher E. Highly nonlinear photodamage in two-photon fluores-
cence microscopy. Biophys J 2001;80:2029e36.
[27] Potter SM. Vital imaging: two photons are better than one. Curr Biol 1996;6:
1595e8.
[28] Li C, Liu Y, Wu Y, Sun Y, Li F. The cellular uptake and localization of non-
emissive iridium(III) complexes as cellular reaction-based luminescence
probes. Biomaterials 2013;34:1223e34.
[29] Patterson GH, Piston DW. Photobleaching in two-photon excitation micro-
scopy. Biophys J 2000;78:2159e62.
[30] van Zandvoort MAMJ, de Grauw CJ, Gerritsen HC, Broers JLV, oude
Egbrink MGA, Ramaekers FCS, et al. Discrimination of DNA and RNA in cells by
a vital fluorescent probe: life time imaging of SYTO13 in healthy and apoptotic
cells. Cytometry 2002;47:226e35.
two dyes with reasonably large
F
ꢀ
d when binding to RNA could be
applied to imaging studies with confocal microscopy and TPM.
Moreover, they have good counterstain compatibility with Hoechst
33342, which is suitable for RNAeDNA colocalization. Especially,
INR1 has tangible potential as an effective one- and two-photon
RNA probes for living cells and we hope this would help the
exploitation of novel TPEF probes for RNA. In this study, an
important unresolved issue we are hammering at is the lack of
specificity of our probe system to specifically monitor particular
RNA (e.g., mRNA or rRNA), this area remains under active
investigation.
Acknowledgments
For financial support, we thank the National Natural Science
Foundation of China (51273107, 30771091, 50990061 and
50721002), Natural Science Foundation of Shandong Province,
China (ZR2012EMZ001), Shandong University 2011JC006, Open
Project of State Key Laboratory for Supramolecular Structure and
Materials (SKLSSM201207).
Appendix A. Supplementary data
[31] Zhang M, Yu M, Li F, Zhu M, Li M, Gao Y, et al. A highly selective fluorescence
turn-on sensor for cysteine/homocysteine and its application in bioimaging.
J Am Chem Soc 2007;129:10322e3.
Supplementary data related to this article can be found online at
http://dx.doi.org/10.1016/j.dyepig.2013.12.005.
[32] Lee JH, Lim CS, Tian YS, Han JH, Cho BR. A two-photon fluorescent probe for
thiols in live cells and tissues. J Am Chem Soc 2010;132:1216e7.
[33] Koo CK, Wong KL, Man CWY, Tam HL, Tsao SW, Cheah KW, et al. Two-photon
plasma membrane imaging in live cells by an amphiphilic, water-soluble
cyctometalated platinum(II) complex. Inorg Chem 2009;48:7501e3.
[34] Gaus K, Gratton E, Kable EPW, Jones AS, Gelissen I, Kritharides L, et al. Visu-
alizing lipid structure and raft domains in living cells with two-photon mi-
croscopy. Proc Natl Acad Sci U S A 2003;100:15554e9.
References
[1] Zheng M, Librizzi D, Kılıç A, Liu Y, Renz H, Merkel OM, et al. Enhancing in vivo
circulation and siRNA delivery with biodegradable polyethylenimine-
graftpolycaprolactone-block-poly(ethylene glycol) copolymers. Biomaterials
2012;33:6551e8.
[35] Yang Z, Zhao N, Sun Y, Miao F, Liu Y, Yu X, et al. Highly selective red- and
green-emitting two-photon fluorescent probes for cysteine detection and
their bio-imaging in living cells. Chem Commun 2012;48:3442e4.
[36] Ohulchanskyy TY, Pudavar HE, Yarmoluk SM, Yashchuk VM, Bergey EJ,
Prasad PN. A monomethine cyanine dye cyan 40 for two-photon-excited
fluorescence detection of nucleic acids and their visualization in live cells.
Photochem Photobiol 2003;77:138e45.
[2] Braasch DA, Jensen S, Liu Y, Kaur K, Arar K, White MA, et al. RNA interference
in mammalian cells by chemically-modified RNA. Biochemistry 2003;42:
7967e75.
[3] Knowles RB, Sabry JH, Martone ME, Deerinck TJ, Ellisman MH, Bassell GJ, et al.
Translocation of RNA granules in living neurons. J Neurosci 1996;16:7812e20.
[4] Cervantes S, Prudhomme J, Carter D, Gopi KG, Li Q, Chang YT, et al. High-
content live cell imaging with RNA probes: advancements in high throughput
antimalarial drug discovery. BMC Cell Biol 2009;10:45e54.

Y. Liu et al. / Dyes and Pigments 103 (2014) 191e201
201
[37] Zhang Y, Wang J, Jia P, Yu X, Liu H, Liu X, et al. Two-photon fluorescence
imaging of DNA in living plant turbid tissue with carbazole dicationic salt. Org
Biomol Chem 2010;8:4582e8.
[42] Chang C, Chu JF, Kuo H, Kang CC, Lin S, Chang T. Solvent effect on photo-
physical properties of a fluorescence probe: BMVC. J Luminesc 2006;119:84e
90.
[38] Miao F, Song G, Sun Y, Liu Y, Guo F, Zhang W, et al. Fluorescent imaging of
acidic compartments in living cells with a high selective novel one-photon
ratiometric and two-photon acidic pH probe. Biosens Bioelectron 2013;50:
42e9.
[39] Liu X, Sun Y, Zhang Y, Miao F, Wang G, Yu X, et al. A 2, 7-carbazole-based
dicationic salt for fluorescence detection of nucleic acids and two-photon
fluorescence imaging of RNA in nucleoli and cytoplasm. Org Biomol Chem
2011;9:3615e8.
[40] Xu C, Webb WW. Measurement of two-photon excitation cross sections of
molecular fluorophores with data from 690 to 1050 nm. J Opt Soc Am B
1996;13:481e91.
[41] Albota MA, Xu C, Webb WW. Two-photon fluorescence excitation cross sec-
tions of biomolecular probes from 690 to 960 nm. Appl Opt 1998;37:7352e6.
[43] Malak H, Castellano FN, Gryczynski I, Lakowicz JR. Two photon excitation of
ethidium bromide labeled DNA. Biophys Chem 1997;67:35e41.
[44] Liu X, Sun Y, Zhang Y, Zhao N, Zhao H, Yu X, et al. A series of carbazole cationic
compounds with large two-photon absorption cross sections for imaging
mitochondria in living cells with two-photon fluorescence microscopy.
J Fluoresc 2011;21:497e506.
[45] Hong Y, Lama JWY, Tang BZ. Aggregation-induced emission: phenomenon,
mechanism and applications. Chem Commun 2009;29:4332e53.
[46] Lyng R, Hard T, Norden B. Induced CD of DNA intercalators: electric dipole
allowed transitions. Biopolymers 1987;26:1327e45.
[47] Larsson A, Carlsson C, Jonsson M, Albinsson B. Characterization of the binding
of the fluorescent dyes YO and YOYO to DNA by polarized light spectroscopy.
J Am Chem Soc 1994;116:8459e65.