Development of a carbazole-based fluorescence probe for G-quadruplex DNA: The importance of side-group effect on binding specificity

Article abstract of DOI:10.1016/j.saa.2018.03.083

G-quadruplex DNAs are highly prevalent in the human genome and involved in many important biological processes. However, many aspects of their biological mechanism and significance still need to be elucidated. Therefore, the development of fluorescent probes for G-quadruplex detection is important for the basic research. We report here on the development of small molecular dyes designed on the basis of carbazole scaffold by introducing styrene-like substituents at its 9-position, for the purpose of G-quadruplex recognition. Results revealed that the side group on the carbazole scaffold was very important for their ability to selectively recognize G-quadruplex DNA structures. 1a with the pyridine side group displayed excellent fluorescence signal turn-on property for the specific discrimination of G-quadruplex DNAs against other nucleic acids. The characteristics of 1a were further investigated with UV–vis spectrophotometry, fluorescence, circular dichroism, FID assay and molecular docking to validate the selectivity, sensitivity and detailed binding mode toward G-quadruplex DNAs.

Full text of DOI:10.1016/j.saa.2018.03.083

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 199 (2018) 441–447
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular
Spectroscopy
journal homepage: www.elsevier.com/locate/saa
Short Communication
Development of a carbazole-based fluorescence probe for G-quadruplex
DNA: The importance of side-group effect on binding specificity
⁎
Ming-Qi Wang , Gui-Ying Ren, Shuang Zhao, Guang-Chang Lian, Ting-Ting Chen, Yang Ci, Hong-Yao Li
School of Pharmacy, Jiangsu University, Zhenjiang 212013, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
G-quadruplex DNAs are highly prevalent in the human genome and involved in many important biological pro-
cesses. However, many aspects of their biological mechanism and significance still need to be elucidated. There-
fore, the development of fluorescent probes for G-quadruplex detection is important for the basic research. We
report here on the development of small molecular dyes designed on the basis of carbazole scaffold by introduc-
ing styrene-like substituents at its 9-position, for the purpose of G-quadruplex recognition. Results revealed that
the side group on the carbazole scaffold was very important for their ability to selectively recognize G-
quadruplex DNA structures. 1a with the pyridine side group displayed excellent fluorescence signal turn-on
property for the specific discrimination of G-quadruplex DNAs against other nucleic acids. The characteristics
of 1a were further investigated with UV–vis spectrophotometry, fluorescence, circular dichroism, FID assay
and molecular docking to validate the selectivity, sensitivity and detailed binding mode toward G-quadruplex
DNAs.
Received 15 September 2017
Received in revised form 26 March 2018
Accepted 30 March 2018
Available online 05 April 2018
Keywords:
G-quadruplex DNA
Fluorescent probe
Carbazole
Side group
© 2018 Elsevier B.V. All rights reserved.
1. Introduction
quadruplex, such as thioflavin T [17], naphthalene diimide [18], thiazole
orange derivatives [19], triphenylamine [20], quinolinium [21], tetra-
Guanine-rich DNA sequences can fold into four-stranded secondary
structures known as DNA G-quadruplexes, which were different from
the regular double helix [1–4]. Based on mutual strand orientation,
they can adopt a wide range of topologies, i.e. parallel, antiparallel, hy-
brid of parallel and antiparallel [2]. DNA G-quadruplexes predominantly
exist in the chromosomal telomeric sequences and the promoter
regions of genes, such as c-myc, ckit, HRAS, VEGF [5–7]. In recently
years, DNA G-quadruplex structures have received immense attention
because their unique structural features are believed to be able to pro-
vide biological significance in transcriptional regulation, DNA replica-
tion, telomere maintenance and antitumor chemotherapy [8–12]. At
the same time, the research of DNA G-quadruplexes is in the center
stage and many aspects of their biological mechanism and significance
still need to be elucidated.
Given the importance of DNA G-quadruplexes, the ever-increasing
interest has promoted the development of the rapid and simple ap-
proaches for G-quadruplex detection. Fluorescence probes provide a
powerful and direct means for studying the folding, function, and local-
ization of biological macromolecules in vitro and in vivo [13]. Thus, fluo-
rescence probes that capable of reporting DNA G-quadruplex structures
are extremely attractive [14–16]. Recently, some small molecular based
fluorescence probes have been identified for analysis of DNA G-
arylimidazole [22]. Among the reported studies, carbazole is one of
the most frequently applied scaffolds for DNA G-quadruplex probe
modification probably due to its merit property in pharmacological ac-
tivities. Successful examples include BMVC [23], TO-CZ [24], BPBC [25]
and 9E PBIC [26]. Interestingly, the cationic charge groups at 3 and 6 po-
sitions of carbazole scaffold play essential roles in the enhancing their
binding potency on G-quadruplex DNA. For example, the molecular
3,6 bis(1 methyl 4 vinylpyridium)carbazole diiodide (BMVC) for recog-
nizing specific quadruplex structures has attractive much attention [23].
However, BMVC has similar binding affinities for G-quadruplex and du-
plex DNA. Seeking new probes that possessing better selectivity is im-
portant for further applications. With respect to the literatures, how a
small cationic charge pendant substituent applied to the 9-position of
carbazole core can induce selectivity fluorescence signal toward G-
quadruplex is unclear and has not been discussed. In the present
study, we reported two new fluorescence dyes, designed on the basis
of carbazole scaffold by introducing styrene-like substituents at its 9-
position, for the purpose of investigating the influence of side groups
on developing the effective fluorescence probes for G-quadruplex rec-
ognition. 1a was designed by linking a cationic pyridinium unit to the
carbazole core via a vinyl group, whereas 1b possesses a cationic
quinolinium compared to 1a. The characteristics of the dyes toward
DNA G-quadruplexes were investigated through both experimental
and modeling studies. Both of them are able to bind with G-
quadruplex DNAs, and excellent fluorescence discrimination with
⁎
Corresponding author.
E-mail address: wmq3415@163.com. (M.-Q. Wang).
https://doi.org/10.1016/j.saa.2018.03.083
1386-1425/© 2018 Elsevier B.V. All rights reserved.

442
M.-Q. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 199 (2018) 441–447
non-G-quadruplex and G-quadruplex structures is found to be deter-
mined by 1a. In addition, the optical properties and the binding mecha-
nism of 1a were further investigated and discussed.
for 1 min before fluorescence spectra recorded. The volume of the dye
was 2.5 mL and the changes in the dye concentration caused by dilution
at the end of each titration were negligible. For 1a and 1b, the fluores-
cence measurement was obtained at an excitation wavelength of 404
and 480 nm, respectively. The date from the fluorimetric titrations
were analyzed according to the independent-site model by nonlinear
fitting to equation [27], in which n is the putative number of 1a mole-
cules binding to a given DNA matrix, Q is the fluorescence enhancement
upon saturation, A = 1/[KaCdye]. The parameters Q and A were found by
Levenberg-Marquardt fitting routine in Origin 8.5 software, whereas n
was varied to obtain a better fit.
2. Experimental Sections
2.1. Materials
All commercially available chemicals used for synthesis were re-
agent grade and used without further purification. The 1H NMR and
13C NMR spectra measured on a Bruker AM400 NMR spectrometer
and the δ scale in ppm referenced to residual solvent peaks or internal
tetramethylsilane (TMS). Mass spectra (MS) were recorded on a
Shimazu LCMS-2010A instrument with an ESI detector. Oligonucleo-
tides (HPLC purified) were purchased from Sangon Biotechnology Co.
Ltd. (Shanghai, China) and the sequences were listed in Table S1. Oligo-
nucleotides were dissolved in 10 mM Tris-HCl buffer (pH 7.4, containing
60 mM KCl). Prior to use, all oligonucleotides were pre-treated by
heating at 95 °C for 5 min, followed by gradual cooling to room temper-
ature and kept at this temperature for 0.5 h. The ct-DNA, purchased
from Sigma Aldrich, was directly dissolved in water at a concentration
of 1 mg/mL. Its concentration was determined according to absorption
intensity at 260 nm with a molar extinction coefficient value of 6600.
Stock solutions of 2 (5 mM) were prepared in DMSO and stored at 4 °C.
ꢀ
ꢁ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
F
F0
Q−1
2
2
¼ 1 þ
A þ 1 þ x− ðA þ 1 þ xÞ −4x
2.3.2. UV–vis Titrations
The UV–vis spectra were recorded on an UV-2550 spectrophotome-
ter using a 1 cm path length quartz cuvette at room temperature. For the
titration experiments, small aliquots of a stock solution of the DNAs
were added to the solution containing 1a at a fixed concentration (8
μM) in aqueous buffer (10 mM Tris-HCl, pH 7.4, 60 mM KCl). After
each DNA addition, the solution was incubated for 1 min before absorp-
tion spectra recorded.
2.2. Synthesis and Characterization
2.2.1. 2 [4 (9H Carbazol 9 yl)styryl] 1 methylpyridinium Iodide 1a
A solution of intermediate 4 (9H carbazol 9 yl)benzaldehyde 3
(0.27 g, 1.0 mmol), 1 methyl 2 picoliniuiodide (0.2 g, 0.85 mmol) and
5 drops piperidine in anhydrous ethanol (50 mL) was refluxed for
12 h under nitrogen with stirring. After cooling to room temperature,
a precipitate was formed during the process of reaction. The reaction
mixture was filtered, and washed thoroughly with anhydrous ethanol.
The residue was purified by column chromatography on silica gel elut-
ing with CH2Cl2/CH3OH (10:1, v/v) to afford 1a (0.24 g, 57.9%) as a yel-
low solid. 1H NMR (400 MHz, DMSO d6) δ: 8.84–8.82 (m, 2H), 8.53 (d, J
= 7.68 Hz, 1H), 8.43 (t, J = 7.76 Hz, 1H), 8.31 (d, J = 7.64 Hz, 1H),
8.15–8.11 (br, 1H), 7.91–7.89 (m, 1H), 7.81–7.77 (m, 1H), 7.69–7.65
(m, 2H), 7.63–7.59 (br, 3H), 7.56–7.52 (m, 1H), 7.47–7.43 (m, 1H),
7.40–7.38 (br, 1H), 7.35–7.31 (br, 2H), 4.38 (s, 3H); 13C NMR
(100 MHz, DMSO d6) δ: 150.30, 146.22, 144.60, 144.32, 142.01, 141.38,
136.73, 130.80, 128.63, 127.77, 127.20, 121.99, 121.35, 115.05, 110.62,
46.67; LC-MS: (positive mode, m/z) calculated 361.1699, found
361.1687 for [M-I]+.
2.3.3. Fluorescence Intercalator Displacement (FID) Assay
The FID experiment was carried out with thiazole orange (TO) as a
fluorescence probe and the protocol was similar to those described else-
where [20]. The concentrations of TO and G-quadruplex 22AG were set
at 0.5 and 0.1 μM, respectively. 1a was added until no change was ob-
served in the fluorescence intensity indicating the binding saturation
has been achieved. The fluorescence spectra were measured using exci-
tation wavelength at 504 nm and the emission range was set between
515 and 700 nm.
2.3.4. Circular Dichroism (CD)
The concentration of G-quadruplex used for measurement was 4 μM
in 10 mM Tris-HCl buffer, pH 7.4 in the presence of 60 mM KCl and the
concentration of 1a was 1–2 folds. CD spectra (230–400 nm) were per-
formed on a JASCO-J815 circular dichroism spectrophotometer using a
10 mm path length quartz cuvette. The scanning speed of the instru-
ment was set to 400 nm min−1. Final analysis of the data was carried
out using Origin 8.5.
2.2.2. 2 [4 (9H Carbazol 9 yl)styryl] 1 methylquinolinuum Iodide 1b
Following the general procedure for 1a, the product 2a was obtained
as a brown solid (0.32, 53.8%). 1H NMR (400 MHz, DMSO d6) δ:
8.93–8.91 (m, 2H), 8.58 (d, J = 9.12 Hz, 1H), 8.46 (t, J = 8.8 Hz, 2H),
8.32 (d, J = 7.6 Hz, 1H), 8.23 (d, J = 8.0 Hz, 1H), 8.08–8.04 (m, 1H),
8.00–7.98 (m, 1H), 7.94–7.90 (m, 1H), 7.84 (t, J = 7.56 Hz, 1H),
7.72–7.69 (m, 2H), 7.63–7.56 (m, 3H), 7.49–7.45 (m, 1H), 7.40–7.33
(m, 3H), 4.55 (s, 3H); 13C NMR (100 MHz, DMSO d6) δ: 156.70, 148.96,
143.77, 142.44, 141.39, 139.57, 136.58, 135.02, 130.81, 130.35, 129.05,
128.95, 128.72, 127.84, 127.75, 127.63, 127.18, 123.91, 123.06, 122.69,
121.52, 121.35, 121.14, 119.56, 116.64, 110.70, 110.67, 41.92. LC-MS:
(positive mode, m/z) calculated 411.1856, found 411.1848 for [M-I]+.
2.3.5. KI Quenching
KI quenching was carried out in the presence and absence of 22AG. A
2.5 mL reaction vessel was prepared containing 1a (2 μM) and 10 mM
Tris-HCl (pH 7.4, containing 60 mM KCl). Emission fluorescence of mix-
ture was acquired after adding increasing concentration of KI (with the
concentration range of from 0 to 1 M) which is a known quencher. The
fluorescence quenching spectra were recorded and were fitted by using
the Stern–Volmer equation.
2.3.6. Molecular Docking
Molecular docking calculations were performed using the Autodock
Vina software, which has been reported to be of high accuracy of predic-
tion [28]. The crystal structure (PDB code: 1KF1, resolution: 2.10 Å) and
the NMR structure (PDB code: 143D) of G4 DNA were downloaded from
RCSB Protein Data Bank. The redundant solvent molecules and ions
were removed from the crystal structure while the first conformation
was retained from the NMR structure. Docked poses were visualised
by using UCSF Chimera.
2.3. Measurements and Methodology
2.3.1. Fluorimetric Titrations
Fluorescence spectra were measured on a Shimadzu RF-5301PCS
spectrofluorophotometer in a 10 mm quartz cell at room temperature.
The concentration of 1a/1b was fixed at 2 μM in buffer, to which the
DNA solution was added step by step. Dye-DNA solution was incubated

M.-Q. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 199 (2018) 441–447
443
3. Results and Discussion
addition of telomeric G-quadruplex DNA 22AG. This turn-on fluores-
cence was also observed when 1a was treated with the G-
quadruplexes Ckit1, CM22, HRAS, Htg-21, C-myc, G3T3, Hras-2 and
Ckit3. Almost no changes in the fluorescence spectra were observed
when other oligonucleotides including single-stranded DNA (ss26), du-
plex DNA (ds26, ctDNA, Polyd(A-T)9 and Polyd(G-C)9)were added
(Fig. S1). These results suggest that 1a exhibits a high selectively fluo-
rescence response toward G-quadruplex DNAs.
3.1. Synthesis of New Fluorescence Dyes
As shown in Scheme 1, intermediate 3 (4 (9H carbazol 9 yl)benzal-
dehyde) was obtained by the reaction of 2 (N phenylcarbazol) with
DMF/POCl3 (Vilsmeier reagent). The targeted fluorescence dyes 1a
and 1b with different styryl substituents were synthesized by the
reaction of
3
with N-methylated aromatic groups including
To ascertain the selective turn-on fluorescent signal is due to the
specific binding of 1a with G-quadruplex DNA structure, a competitive
titration was conducted in the presence of a large excess of a duplex
DNA (ctDNA, 50 μM) with 22AG G-quadruplex (Fig. 2B). At the high
concentration of ctDNA, a weak background fluorescent signal of 1a
was observed. The addition of 22AG led to a significant enhancement
of the fluorescence intensity, although the fluorescence of 1a was some-
what affected by the presence of duplex DNA. The results revealed that
1a possessed excellent specificity for G-quadruplex sensing in the com-
petitive biological environment.
1 methyl 2 picoliniuiodide (1a) and 1 methyl 2 quinoliniumiodide
(1b). The structures of these two new dyes were characterized by 1H
NMR, 13C NMR and MS (see the Supplementary information).
3.2. Selectivity Study of New Dyes Toward G-quadruplex DNAs Against
Other Nucleic Acids
The selectivity of these new dyes to G-quadruplex DNAs was first in-
vestigated and evaluated by using the fluorescence spectra of the dyes
in the presence of different DNA forms including G-quadruplexes,
single- and double-stranded DNAs and the protein BSA. As shown in
Fig. 1, the enhanced fluorescence intensity of 1a was found to be re-
markable upon binding with different G-quadruplex DNAs (parallel,
anti-parallel and hybrid-type) at saturated concentration, while
single- and double-stranded DNAs and the protein BSA showed negligi-
ble fluorescence enhancement, suggesting that 1a could discriminate of
G-quadruplexes from single-, double-stranded DNAs and BSA. In con-
trast, under the same experimental conditions, although 1b responded
to most of these DNAs and BSA, the enhanced intensity gave poor selec-
tivity. These results validate the importance of side group of carbazole
scaffold in achieving effective fluorescence probing of G-quadruplexes.
Generally, 1a with the pyridine side group has better performance
than 1b with the quinoline group for utilizing as a G-quadruplex selec-
tive fluorescent probe in bioassays. As a result, 1a was chosen for further
detailed studies.
3.4. Time-dependent Response
The response time of the probe was a key parameter for the biolog-
ical application. We then proceeded to study the time-dependent re-
sponse characteristic of the probe 1a in the presence of 22AG was
performed by using fluorescence spectra. As shown in Fig. 3, once
22AG was added, the fluorescence intensity of 1a exhibited a sharp in-
crease and reached the maximum within 10 s, meaning that the binding
performance between 1a and 22AG could accomplished within 10 s at
room temperature.
3.5. UV–vis Absorption Spectroscopy
The ability of 1a to interact with G-quadruplex DNA was also studied
by spectrophotometric titration experiments. In aqueous buffer solu-
tion, 1a shows a wide absorption band with a λmax at 402 nm. Fig. 4
shows that the maximum absorption wavelength of the 1a in the 1a-
22AG system has very slight shift in λmax but its absorbance decreases
significantly as compared with the free 1a system. The similar changes
in the spectral profiles of other G-quadruplex DNAs during titrations
were shown in Fig. S2. It is a generally accepted concept that unimpor-
tant (or small) shift in maximum wavelength of absorption is the most
3.3. Fluorescence Properties of 1a With DNA Forms
The detail fluorescence properties of 1a with the DNA forms were
studied by using fluorescence titration assay. As shown in Fig. 2A, 1a
alone in buffer displayed weak emission with a maximum wavelength
at approximately 545 nm, which was remarkably enhanced by the
Scheme 1. Synthesis route to 1a–b and their molecular structuresa.

444
M.-Q. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 199 (2018) 441–447
Fig. 1. Comparison of enhancement of emission intensity of 1a and 1b with BSA; single-, and double-stranded DNAs; G-quadruplexes DNAs. The concentration of the dyes was 2 μM in
10 mM Tris-HCl buffer (pH 7.4) containing 60 mM KCl.
possible outcome of groove binding [29]. Thus, the recorded spectra in-
dicated the binding between 1a and G-quadruplex DNA was in the
mode of groove binding rather than that of intercalation binding [30].
Encouraged by the fluorescence emission enhancement of 1a with
G-quadruplexes, we then investigated the detection limits (LOD) asso-
ciated with using 1a to G-quadruplexes in solution. The LOD values
were calculated according to the equation 3s/k [31]. The s value repre-
sents the standard deviation for multiple measurements of blank solu-
tion. The k value is the slope derived from the linear range of the 1a
fluorescence titration curve with different G-quadruplexes. The linear
ranges of the fluorescence titration curves for 1a with most of G-
quadruplex DNAs ranged from 0.1 to 1.0 μM. Thus, the corresponding
values of the LODs were shown in Table 1 and (Fig. S5). Clearly, the
LODs of 1a for different G-quadruplexes in solution were below 1 μM,
which meant it was able to detect G-quadruplex DNAs in the micromo-
lar range.
3.6. Binding Affinities and the Detection Limits (LOD) of 1a With
G-quadruplex DNAs
The binding properties of 1a with G-quadruplex DNAs were also
studied by fluorescence titrations. To estimate the binding constant
(Ka) and stoichiometry (n) of 1a to G-quadruplex DNA, the fluorescence
titration curve was fitted to an independent-site model (see Experimen-
tal sections, Fig. S3) and the results were further confirmed by Job plot
experiments (Fig. S4). The results presented in Table 1 revealed several
regularities: 1) 1a exhibited higher binding affinity to anti-parallel G-
quadruplex DNAs (Ckit3, HRAS and G3 T3) and lower affinity to parallel
(CM22, C-myc, Hras-2, Ckit1) and hybrid-type (22AG, Htg-21) G-
quadruplex DNAs; 2) 1a showed different binding stoichiometries to
the tested G-quadruplex DNAs; 3) The binding affinities and fluores-
cence enhancements are not directly correlated, may due to the differ-
ent microenvironments in the binding sites, which avoid the non-
radiative relaxation process to different extents.
3.7. Effect of Potassium Ionic Strengths of G-quadruplex DNA on the Interac-
tions With 1a
G-quadruplex structures are inherently stabilized by the presence of
alkali-metal cations which coordinated to O6 atoms of guanines in the
quartet [9]. In order to study the effect of buffer cationic strength on
this dye 1a binding to G-quadruplex DNA, fluorescence titrations were
Fig. 2. (A) Fluorescence titration of 2 μM 1a with stepwise addition of the G-quadruplex DNA 22AG (0–1.4 μM) in 10 mM Tris-HCl buffer (pH 7.4) containing 60 mM KCl. (B) Fluorescence
titration of 2 μM 1a with stepwise addition of the G-quadruplex DNA 22AG in the presence/absence of 50 μM ctDNA in 10 mM Tris-HCl buffer (pH 7.4) containing 60 mM KCl.

M.-Q. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 199 (2018) 441–447
445
Table 1
Apparent binding constants (Ka), binding stoichiometry (n) and the detection limits (LOD)
of 1a to different G-quadruplex DNAs, determined from fluorimetric titrations. [a]Experi-
mental conditions: 1a = [2 μM] in 10 mM Tris-HCl buffer (pH 7.4) containing 60 mM
KCl; [b]Linear detection range.
G-quadruplex
DNA[a]
Stoichiometry
(1a:DNA)
Ka
Lrd
(μM)
LOD
(μM)
[106 M−1
]
22AG
CM22
C-myc
Ckit1
3:1
1:1
3:1
3:1
3:1
3:1
3:1
4:1
1:1
0.33
1.45
0.36
0.96
2.43
0.33
6.58
3.88
0.28
0.1–0.7
0.1–0.4
0.2–0.6
0.2–0.7
0.1–0.6
0.1–1.0
0.2–0.6
0.1–0.4
0.1–1.0
0.11
0.065
0.11
0.24
0.22
0.32
0.24
0.10
0.27
Ckit3
Htg-21
HRAS
G3 T3
Hras-2
peak at ~290 nm and a negative peak at ~255 nm. The telemetric
hybrid-type G-quadruplex 22AG showed a strong positive peak at
~290 nm, a minor shoulder positive peak at ~270 nm and a negative
peak at ~235 nm [33]. The addition of 1a to these three G-
quadruplexes had negligible impacts on the characteristic CD peaks, in-
dicating that this dye can sense G-quadruplexes without affecting the
G-quadruplex topologies.
Fig. 3. The fluorescence intensity at 545 nm of 1a (2 μM) in the presence of 22AG (2 μM).
carried out in different Tris-HCl buffer with variation in K+ concentra-
tions (0–100 mM). Result emphasized that the value of enhancement
in the fluorescence intensity of 1a with 22AG was increased with in-
creasing of buffer salt concentration, with maximal value occurring
near 10 mM (Fig. 5). Above this concentration, however, it displayed a
slight decrease. Perhaps because K+ ions preferentially stabilized the
G-quadruplex with folding structures and only a low concentration of
3.9. Binding Mode Studies
K
+ ions is required for the structural stabilization. In the case of higher
The main binding modes of ligands with G-quadruplexes are end-
stacking and grooves binding. To obtain the details on the interactions
of 1a with G-quadruplexes, an indirect fluorescent intercalator displace-
ment (FID) assay by using thiazole orange (TO) was performed [34].
This method is based on the competitive displacement of fluorescence
probe TO/G-quadruplex complex. TO is considered to bind G-
quadruplex DNA mainly by an end-stacking mode and it consequently
exhibits high fluorescence intensity (~500- to 3000-fold). Dyes having
the same binding site will compete with TO for G-quadruplex binding,
resulting in the large decrease of the emission. As shown in Fig. 7A
and B, in the presence of large (20 equiv) excess of 1a, partial displace-
ment of TO was detected and the 50% threshold was not reached. Since
TO should be easier to displace by an end-stacking dye, this partial dis-
placement may due to the groove binding mode between 1a and G-
quadruplex 22AG [35]. For comparison, 1b with a quinolinium side
chain was also studied. FID experiment of 1b showed that fluorescence
emission of TO was gradually reduced and almost completely quenched
(Fig. S6). Therefore, one can conclude that 1b exerted similar effect on
concentration of K+ ions, it demonstrates that K+ ions occupy the avail-
able binding sites on G-quadruplex surface and reduces the possibility
for ligand to bind exterior of G-quadruplex [32]. The positively charged
side group would enhance electrostatic interaction strength with the
negatively charged DNA phosphate backbone. The dependence of the
fluorescence enhancements on the ionic strength of the buffer indicates
that the electrostatic interactions play an important role in the binding
process.
3.8. Effect of 1a on the Topologies of G-quadruplexes
To understand the effect of the dye on the topologies of G-
quadruplexes, we deployed circular dichroism (CD) spectroscopy to
monitor the formation of the G-quadruplex structure in the absence
and presence of this dye. As shown in Fig.6, the parallel G-quadruplex
CM22 exhibited a characteristic positive peak at ~263 nm and a negative
peak at ~242 nm. The anti-parallel G-quadruplex HRAS had a positive
Fig. 4. Spectrophotometric titration of 1a with 22AG in Tris-HCl buffer (10 mM, pH 7.4)
containing 60 mM KCl. Conditions: [1a] = 8 μM; [22AG] = 0–2.2 μM. Arrows indicate
the change in absorbance upon increasing the 22AG concentrations.
Fig. 5. Fluorescence enhancements of 1a (2 μM) at 545 nm with 22AG (1 μM) in 10 mM
Tris-HCl, pH 7.4, containing 0–100 mM KCl.

446
M.-Q. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 199 (2018) 441–447
Fig. 6. CD spectra of 4 μM G-quadruplex-forming oligonucleotides CM22 (A), HRAS (B), and 22AG (C) in 10 mM Tris-HCl buffer, 60 mM KCl, pH 7.4 with and without 1a.
the TO, consisting in gradual release of TO from G-quadruplex structure.
In general, this may be caused the quinolinium moiety may interact
with G-quartet plane by π–π stacking binding as previous reported
[36]. From the FID experiments, in fact, different side group conjugating
with the same scaffold can generate distinct G-quadruplex binding
properties. Although the majority of dyes reported so far recognize G-
quadruplexes by end-stacking binding mode. The groove binding may
be the more promising binding mode, which is favorable for the
enhancement of selectivity. G-quadruplex topologies endow grooves
with specific dimensions and accessibilities, which are different from
various other structures. In order to further confirm the binding mode
of 1a to G-quadruplex DNA, its fluorescence quenching in the absence
and presence of 22AG were studied by using a well-known quencher
(potassium iodide). As reported in many studies, fluorescence intensity
of an intercalating small molecule is well protected from being
quenched as the approach of quencher e.g. iodide toward the
Fig. 7. (A) Fluorescence displacement of TO bound G-quadruplex 22AG by 1a in 10 mM Tris-HCl buffer (pH 7.4, containing 60 mM KCl). Arrow shows the intensity changes upon increasing
the concentration of 1a. (B) Plot of TO displacement vs. ratio of 1a/TO. (C) Molecular modeling for 1a and parallel G-quadruplex 1kf1. (D) Molecular modeling for 1a and anti-parallel G-
quadruplex 143d.

M.-Q. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 199 (2018) 441–447
447
fluorophore is restricted. However, groove binding affords little shield
for the bounded molecule, so the fluorophore is exposed to the external
environment and iodide anions can quench their fluorescence without
difficulty even in the presence of DNA [37]. The obtained data were plot-
ted according to the Stern-Volmer equation (Fig. S7) [38]. The values of
Ksv of 1a by I− ion in the absence and presence of 22AG were calculated
to be 1.51 M−1 and 1.66 M−1, respectively. The Ksv value was higher
than the quenching of 1a by KI in the absence of 22AG and these find-
ings are good proof of groove binding mode.
Province (17KJB150009), the China Postdoctoral Science Foundation
Funded Project (Grant No. 2017M611704).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.saa.2018.03.083.
To gain further insight into the interactions of 1a and G-
quadruplexes, molecular docking studies were carried out by using
AutoDock Vina modeling tool. The anti-parallel basket NMR G-
quadruplex structure (PDB 143D) and a parallel-type crystal G-
quadruplex structure (PDB 1KF1) were used as the templates for the
docking studies. As displayed in Fig. 7C and D, the docking studies sug-
gested that 1a bound to these two G-quadruplexes in the groove region.
The side group provided steric hindrance, and the molecule was non-
planar and inserted into shallow groove of G-quadruplex DNA. More-
over, the modeling study revealed that 1a bound to the anti-parallel
G-quadruplex and parallel G-quadruplex with a calculated binding en-
ergy of ~−7.2 and −8.2 kcal/mol, respectively. Therefore, 1a exhibited
a superior binding energy to anti-parallel G-quadruplex to that of paral-
lel G-quadruplex. This finding was consistent with the trend for binding
constants of the G-quadruplexes as observed in the fluorescence titra-
tion assay.
References
[1] M. Gellert, D.R. Davies, Proc. Natl. Acad. Sci. U. S. A. 48 (1962) 2013–2018.
[2] S. Burge, G.N. Parkinson, P. Hazel, A.K. Todd, S. Neidle, Nucleic Acids Res. 34 (2006)
5402–5415.
[3] H.J. Lipps, D. Rhodes, Trends Cell Biol. 19 (2009) 414–422.
[4] P. Murat, S. Balasubramanian, Curr. Opin. Genet. Dev. 25 (2014) 22–29.
[5] J.L. Huppert, FEBS J. 277 (2010) 3452–3458.
[6] J.L. Huppert, S. Balasubramanian, Nucleic Acids Res. 33 (2005) 2908–2916.
[7] S. Rankin, A.P. Reszka, J. Huppert, M. Zloh, G.N. Parkinson, A.K. Todd, S. Ladame, S.
Balasubramanian, S. Neidle, J. Am. Chem. Soc. 127 (2005) 10584–10589.
[8] D. Rhodes, H.J. Lipps, Nucleic Acids Res. 43 (2015) 8627–8637.
[9] Q. Cao, Y. Li, E. Freisinger, P.Z. Qin, R.K.O. Sigel, Z.W. Mao, Inorg. Chem. Front. 4
(2017) 10–32.
[10] G. Cimino-Reale, N. Zaffaroni, M. Folini, Curr. Pharm. Des. 22 (2016) 6612–6662.
[11] T. Tian, H. Xiao, X. Zhou, Curr. Top. Med. Chem. 15 (2015) 1988–2001.
[12] Z. Tan, J. Tang, Z.Y. Kan, Y.H. Hao, Curr. Top. Med. Chem. 15 (2015) 1940–1946.
[13] T. Terai, T. Nagano, Pflugers Arch. - Eur. J. Physiol. 465 (2013) 347–359.
[14] R. Balayeshwanth, V.J. Alzeer, N.W. Luedtke, Chembiochem 14 (2013) 540–558.
[15] D.L. Ma, Z.H. Zhang, M.D. Wang, L.H. Lu, H.J. Zhong, C.H. Leung, Chem. Biol. 22 (2015)
812–828.
[16] A.C. Bhasikuttan, J. Mohanty, Chem. Commun. 51 (2015) 7581–7597.
[17] J. Mohanty, N. Barooah, V. Dhamodharan, S. Harikrishna, P.I. Pradeepkumar, A.C.
Bhasikuttan, J. Am. Chem. Soc. 135 (2013) 367–376.
4. Conclusions
[18] M. Zuffo, S. Ladame, F. Doria, M. Freccero, Sensors Actuators B Chem. 245 (2017)
In conclusion, we have designed and investigated the specific G-
quadruplex recognition ability of carbzole derivatives with different
styrene-like substituents. To this end, two new dyes 1a and 1b were
synthesized and structurally characterized. Result showed that different
side group conjugating with the same scaffold can generate distinct G-
quadruplex binding properties. 1a with pyridine side group can dis-
criminate G-quadruplex from duplex and single-stranded DNA based
on its significant fluorescence enhancement in the presence of G-
quadruplexes, while 1b with quinolinium group gave poor selectivity.
The result hints that side group plays an important role in G-
quadruplex binding process and it is possible to obtain probes with
high selectivity.
In addition, the detailed interactions of 1a with G-quadruplex DNAs
using various experiments and molecular docking were illustrated. 1a
exhibited sensitive fluorescence detection of G-quadruplex DNAs with
the LOD values below 1 μM in solution. Additionally, the CD experi-
ments demonstrated that 1a is able to sense G-quadruple structures
without affecting their topologies. Further studies showed that 1a was
inserted into the groove region of G-quadruplex structure and exhibited
higher binding affinity to anti-parallel G-quadruplex structures. These
results of this study gave some crucial factors on developing of effective
probes for G-quadruplex DNA applications.
780–788.
[19] P. Yang, A.D. Cian, M.-P. Teulade-Fichou, J.-L. Mergny, D. Monchaud, Angew. Chem.
Int. Ed. 48 (2009) 2188–2191.
[20] M.-Q. Wang, Y. Wu, Z.-Y. Wang, Q.-Y. Chen, F.-Y. Xiao, Y.-C. Jiang, S. Amy, Dyes Pig-
ments 145 (2017) 1–6.
[21] Y.-J. Lu, X.-L. Guo, M.-H. Xu, W.-W. Chen, W.-L. Wong, K. Zhang, C.-F. Chow, Dyes
Pigments 143 (2017) 331–341.
[22] M.-H. Hu, S.-B. Chen, Y.-Q. Wang, Y.-M. Zeng, T.-M. Ou, D. Li, L.-Q. Gu, Z.-S. Huang, J.-
H. Tan, Biosens. Bioelectron. 83 (2016) 77–84.
[23] C.C. Chang, I.C. Kuo, I.F. Ling, C.T. Chen, H.C. Chen, P.J. Lou, J.J. Lin, T.C. Chang, Anal.
Chem. 76 (2004) 4490–4494.
[24] Y.-J. Lu, Q. Deng, J.-Q. Hou, D.-P. Hu, Z.-Y. Wang, K. Zhang, L.G. Luyt, W.L. Wong, C.F.
Chow, ACS Chem. Biol. 11 (2016) 1019–1029.
[25] Y. Wei, X. Zhang, L. Wang, Y. Liu, T. Bing, X. Liu, D. Shangguan, RSC Adv. 5 (2015)
75911–75917.
[26] D. Lin, X. Fei, Y. Gu, C. Wang, Y. Tang, R. Li, J. Zhou, Analyst 140 (2015) 5772–5780.
[27] F.H. Stootman, D.M. Fisher, A. Rodger, J.R. Aldrich-Wright, Analyst 131 (2006)
1145–1151.
[28] O. Trott, A.J. Olson, J. Comput. Chem. 31 (2010) 455–461.
[29] D. Sahoo, P. Bhattacharya, S. Chakravorti, J. Phys. Chem. B 114 (2010) 2044–2050.
[30] C. Kumar, E.H. Asuncion, J. Am. Chem. Soc. 115 (1993) 8547–8553.
[31] M. Shortreed, R. Kopelman, M. Kuhn, B. Hoyland, Anal. Chem. 68 (1996) 1414–1418.
[32] N. Nagesh, V.K. Sharma, A.G. Kumar, E.A. Lewis, J. Nucleic Acids 2010 (2010) 146418.
[33] Y. Hong, H. Xiong, J.W. Lam, M. Häußler, J. Liu, Y. Yu, Y. Zhong, H.H. Sung, I.D.
Williams, K.S. Wong, B.Z. Tang, Chem. Eur. J. 16 (2010) 1232–1245.
[34] B.T. Nguyen, E.V. Anslyn, Coord. Chem. Rev. 250 (2006) 3118–3127.
[35] P.L.T. Tran, E. Largy, F. Hamon, M.P. Teulade-Fichou, J.L. Mergny, Biochimie 93
(2011) 1288–1296.
[36] Z.-Q. Liu, S.-T. Zhuo, J.-H. Tan, T.-M. Ou, D. Li, L.-Q. Gu, Z.-S. Huang, Tetrahedron 69
(2013) 4922–4932.
[37] J. Li, D. Li, Y. Han, S. Shuang, C. Dong, Spectrochim. Acta A 73 (2009) 221–225.
[38] J.R. Lakowicz, G. Webber, Biochemist 12 (1973) 4161–4170.
Acknowledgements
This work was financially supported by the General Program of Nat-
ural Science Foundation of the Higher Education Institutions of Jiangsu