Highly sensitive and selective strategy for imaging Hg2+ using near-infrared squaraine dye in live cells and zebrafish

Article abstract of DOI:10.1016/j.dyepig.2018.05.016

This article describes the rational design and synthesis of a monothiosquaraine (MTSQ) fluorescent Hg²⁺ probe that exhibits near-infrared fluorescence, ultrahigh brightness, fast response, high sensitivity and excellent selectivity. Utilizing the strategy of the incorporation of the heavy atom effect and self-assembly caused quenching, the background fluorescence of MTSQ was greatly lowered and the sensitivity in PBS buffer solution was highly enhanced with a very low limit of detection (LOD = 1.2 nM). Systematic studies revealed excellent selectivity and specificity for Hg²⁺ over other thiophilic metal ions such as Ag⁺, Pb²⁺, and Cd²⁺. The mechanistic investigation indicates that self-assembly and desulfurization of MTSQ are responsible for Hg²⁺-triggered significantly higher fluorescence enhancement. With high sensitivity, optimal pH range of 3–7, weak background, good water-solubility and low cytotoxicity, MTSQ therefore can be applied for imaging Hg²⁺ in live cells and zebrafish. The very simple but effective strategy reported here can provide a more convenient approach for designing NIR probes.

Full text of DOI:10.1016/j.dyepig.2018.05.016

DyesandPigments157(2018)369–376
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Highly sensitive and selective strategy for imaging Hg2+ using near-infrared
squaraine dye in live cells and zebrafish
Guimei Wang, Wenjian Xu, Huanghao Yang, Nanyan Fu∗
MOE Key Laboratory for Analytical Science of Food Safety and Biology & Fujian Provincial Key Laboratory of Analysis and Detection Technology for Food Safety, State
Key Laboratory of Photocatalysis on Energy and Environment, College of Chemistry, Fuzhou University, Fuzhou 350116, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Squaraine
Self-assembly
Double quenching effect
Hg2+-triggered disaggregation
Fluorescent probe
Bioimaging
This article describes the rational design and synthesis of a monothiosquaraine (MTSQ) fluorescent Hg2+ probe
that exhibits near-infrared fluorescence, ultrahigh brightness, fast response, high sensitivity and excellent se-
lectivity. Utilizing the strategy of the incorporation of the heavy atom effect and self-assembly caused quenching,
the background fluorescence of MTSQ was greatly lowered and the sensitivity in PBS buffer solution was highly
enhanced with a very low limit of detection (LOD = 1.2 nM). Systematic studies revealed excellent selectivity
and specificity for Hg2+ over other thiophilic metal ions such as Ag+, Pb2+, and Cd2+. The mechanistic in-
vestigation indicates that self-assembly and desulfurization of MTSQ are responsible for Hg2+-triggered sig-
nificantly higher fluorescence enhancement. With high sensitivity, optimal pH range of 3–7, weak background,
good water-solubility and low cytotoxicity, MTSQ therefore can be applied for imaging Hg2+ in live cells and
zebrafish. The very simple but effective strategy reported here can provide a more convenient approach for
designing NIR probes.
1. Introduction
spontaneous self-assembly of different components giving rise to a
variety of nanostructures in solution and in the solid state has evolved
Due to their toxicity and tendency for bioaccumulation in organ-
isms, mercury ions have caused many severe human health problems,
including central nervous system defects and erythrism as well as ar-
rhythmias, cardiomyopathies, and kidney damage [1,2]. Development
of fluorescent mercury responsive probes with fast analysis, high sen-
sitivity and excellent selectivity has been an urgent priority. Inspirit-
ingly, many good small-molecular fluorescent probes for Hg2+ in live
cell have been reported [3–7]. However, the field of mercury-re-
sponsive fluorescent probes for living biological samples faces sig-
nificant challenges in overcoming issues of selectivity and specificity
as an attractive strategy for fabricating materials with tunable physical
and chemical properties. Among these, small molecules with planar,
conjugated, and rigid structures are attractive building blocks for the
construction of self-assembled materials. In addition, they have an at-
tractive potential for in vivo fluorescent probes because of their low
molecular weight, high chemical stability, and relatively good cell
permeability.
Squaraine dyes are a family of extensively studied zwitterionic dyes
possessing sharp and intense absorption and fluorescence in the visible
to near-infrared (NIR) spectral regions [19–27]. One striking property
of squaraine dyes is their high propensity to form aggregates in aqueous
solutions, with the driving forces of aggregation mainly from π−π
stacking, hydrophobic interactions [28,29]. The formation of ag-
gregated structures is sensitive to the structure of the dye molecule that
composes the supramolecular structure and other external stimuli. Till
now, aggregation behavior and self-assembled structures of squaraines
have been studied in mixed solvents [30–33], organized media [34,35]
and under external stimuli conditions [36,37]. Ajayaghosh and cow-
orkers [36] reported the self-assembly of tripodal squaraines in acet-
onitrile into hollow spherical structures that change to form linear
structures upon binding with Ca2+ or Mg2+. In another study, they
[8–10] over other thiophilic metal ions such as Ag+, Pb2+, and Cd2+
,
as well as poor aqueous solubility [11–13], meanwhile maintaining
biocompatibility. In addition, most of these probes show emission and
absorption within the ultraviolet or visible range (300–650 nm)
[14–17], which will hinder the applications in biological systems.
Therefore, the development of effective fluorescent probes capable of
imaging in vivo, with fast response, ultrasensitivity, excellent se-
lectivity, and near infrared optical windows at excitation and emission
wavelengths remains a huge challenge.
Very recently, self-assembly has gained increasing interest in the
field of supramolecular chemistry and materials chemistry [18]. The
∗
Corresponding author.
E-mail address: nanyan_fu@fzu.edu.cn (N. Fu).
https://doi.org/10.1016/j.dyepig.2018.05.016
Received 21 March 2018; Received in revised form 10 May 2018; Accepted 10 May 2018
0143-7208/©2018ElsevierLtd.Allrightsreserved.

G. Wang et al.
DyesandPigments157(2018)369–376
reported the ultrasound-triggered self-assembly and gelation of a
squaraine dye tethered with dodecyloxy galloyl diamide unit in n-bu-
tanol, forming fibrous structures [37]. Ramaiah and coworkers eluci-
dated the aggregation behavior of a cholesterol-appended squaraine
dye in solution and thin films, and the dependence of aggregation on
solvent composition, temperature and concentration [34]. However,
self-assembly structures of squaraine dyes in water and their applica-
tion in molecular recognition have received less attention.
Herein, we would like to report for the first time a novel mono-
thiosquaraine (MTSQ) which was designed and synthesized based on
the strategy of the incorporation of the heavy atom effect and self-as-
sembly. MTSQ exhibits ultrahighly sensitive and selective fluorescence
response for Hg2+ and improving imaging in live cells and zebrafish.
Furthermore, the unique self-assembly structures of squaraine dyes in
PBS solution have been observed using freeze-fracture TEM imaging
techniques and the assembly-disassembly behavior of MTSQ has been
investigated by DLS measurement and UV–Vis spectra.
2. Experimental
2.1. Materials and instrumentations
Scheme 1. The synthetic procedure of MTSQ.
All chemicals used in this paper were obtained from commercial
suppliers and used without further purification. Solvents were dried
according to standard procedures.
3456, 2875, 1685, 1611, 1396, 1343, 1282, 1223, 1192, 1114, 808,
720, 648 cm−1 1H NMR (400 MHz, CDCl3) δ 11.78 (s, 0.6H), 11.72 (s,
1H NMR (400 MHz) and 13C NMR (100 MHz) spectra were recorded
;
0.6H), 11.11 (s, 0.8H), 9.23 (d, J = 9.4 Hz, 0.8H), 8.95 (d, J = 9.4 Hz,
0.6H), 8.18 (d, J = 9.3 Hz, 0.6H), 6.55–6.35 (m, 2H), 6.17 (d,
J = 7.6 Hz, 2H), 3.72 (dd, J = 14.7, 4.7 Hz, 16H), 3.60 (dd, J = 5.5,
3.3 Hz, 8H), 3.52 (dd, J = 6.0, 3.0 Hz, 8H), 3.38 (s, 12H); 13C NMR
(100 MHz, CDCl3) δ 202.04, 197.31, 185.31, 184.97, 180.60, 178.30,
176.00, 164.80, 164.55, 164.12, 157.38, 157.34, 157.18, 134.40,
132.43, 131.51, 111.15, 111.01, 110.53, 107.81, 107.71, 107.62,
99.66, 98.91, 98.82, 71.92, 70.79, 68.56, 59.10, 51.76; ESI-MS: m/z
721.7 ([M+H]+), 743.6 ([M+Na]+); HRMS(ESI): calcd for
on
a Bruker AV-400 spectrometer (TMS as internal standard).
Absorption spectra were measured on a Perkin Elmer Lambda 750
UV–Vis spectrophotometer. Fluorescent emission spectra were mea-
sured using a Cary Eclipse fluorescence spectrophotometer.
2.2. Synthetic procedure
2.2.1. Synthesis of 2
The compound 2 was synthesized according to the literature pro-
cedure [38]. Details of the synthesis of 2 are given in the Supporting
Information (SI) together with full characterization by 1H NMR, 13C
NMR.
C
C
36H53N2O11
S ) 721.3370, found 721.3373; calcd for
([M+H]+
36H52N2O11NaS ([M+Na]+) 743.3190, found 743.3192.
2.3. Spectral measurements
2.2.2. Synthesis of 3
Stock solutions of MTSQ in DMSO were prepared in advance. The
Squaric acid (38.8 mg, 0.34 mmol) was dissolved in a mixture of n-
butanol (30 mL) and benzene (15 mL) in a 100 mL round bottom flask
equipped with a Dean-Stark trap and refluxed under nitrogen with
stirring for 1 h. Then aniline 2 (281 mg, 0.90 mmol) dissolved in 2 mL of
n-butanol was added dropwise and refluxed for further 4 h. Most of the
solvent was removed under reduced pressure. The crude product was
washed with petroleum ether three times and the product was obtained
as a blue solid 180 mg, yield 75%, m.p. 153–154 °C. FRIR (KBr): νmax
2874, 2811, 1623, 1568, 1530, 1431, 1402, 1344, 1267, 1234, 1183,
solutions of metal ions (Hg2+, Fe3+, Ni2+, Zn2+, Cd2+, Ag+, Sr2+
,
Al3+, Mn2+, Pb2+, Co2+, Cu2+, Li+, Na+, K+, Mg2+, Ca2+, Ba2+) and
anions (SCN−, S2O32−, S2−, SO42−, I−, ClO4−, CH3COO−, NO3
,
−
HCO3−, CO32−, NO2−) were prepared in twice distilled water. One μL
of the dye stock solution was added to 2 mL of PBS buffer solution
(10 mM, pH 7.0) in a quartz cuvette of 1 cm path length to acquire
2.0 μM dye solutions. Different volumes of the ions solutions were
added into the above solution respectively and the spectra were re-
corded after 20 min.
1115, 815, 781, 746 cm−1 1H NMR (400 MHz, CDCl3): δ 8.02 (d,
;
J = 7.7 Hz, 0.6H), 7.87 (d, J = 9.2 Hz, 1.4H), 6.42 (d, J = 10.4 Hz, 2H),
6.17 (s, 2H), 3.71 (dd, J = 11.5, 4.5 Hz, 16H), 3.62–3.59 (m, 8H),
3.54–3.51 (m, 8H), 3.38 (s, 12H); 13C NMR δ (100 MHz, CDCl3):
182.92, 173.33, 164.68, 156.93, 132.49, 110.29, 107.96, 98.94, 71.94,
70.78, 68.56, 59.11, 51.69; ESI-MS: m/z 727.6 ([M+Na]+); HRMS
(ESI): calcd for C36H53N2O12 ([M+H]+) 705.3599, found 705.3646;
2.4. DLS measurement
Solution of MTSQ and Hg(NO3)2 were all filtered using syringe fil-
ters with pore size 0.4 μm. Dynamic light scattering (DLS) measure-
ments were performed on a Zetasizer Nano-ZS90 (Malvern).
calcd for C36H52N2O12Na ([M+Na+ +) 727.3418, found: 727.3464.
]
2.5. FF-TEM
2.2.3. Synthesis of MTSQ
Squaraine 3 (105 mg, 0.15 mmol) and Lawesson's reagent (121 mg,
0.30 mmol) were dissolved in 30 mL of dry toluene and refluxed under
nitrogen with stirring for 15 h. After cooling down, most of solvent was
removed under reduced pressure, and then the crude product was
purified by column chromatography on silica gel. Elution of the column
with a mixture of dichloromethane and methanol (100:1, v/v) to afford
the green solid 57 mg, yield 53%, m.p. 125–127 °C. FTIR (KBr): νmax
The samples were freeze-cooled in liquid nitrogen and the solid
samples were ruptured by external force in
a vacuum sprayer
(JEOLJEE-4X Vacuum Evaporator, high vacuum freeze-etching
a
system, Balzers BAF-400D). The resulting fresh surface was sprayed
with platinum and carbon, and then dissolved out the original sample,
giving a platinum-carbon complex film. The samples were observed
using a JEOL JEM-100CX II electron microscope with an accelerating
370

G. Wang et al.
DyesandPigments157(2018)369–376
Scheme 2. The response mechanism of probe MTSQ
to Hg2+
.
Fig. 1. Changes in absorption spectra of MTSQ (2.0 μM) with increase of water
Fig. 3. Changes in fluorescence spectra of MTSQ (2.0 μM) with increasing ad-
dition of Hg2+ (0–2.0 μM) in pH 7.0 PBS buffer solution (λex = 630 nm). Inset:
Fluorescence detection of Hg2+, MTSQ alone and upon addition of Hg2+ ir-
radiated by a hand-held lamp (630 nm channel).
percentage in EtOH/water solution.
voltage of 120 kV.
2.6. Determination of absolute quantum yields
with different concentrations of the MTSQ for 24 h. The control groups
were cultivated under the same conditions without MTSQ. Then, the
medium was changed by DMEM (100 μL per well) containing MTT
(0.5 mg/mL), followed by incubation for another 4 h. Subsequently, the
supernatant in the wells were removed, and frozen crystals were dis-
solved with freshly prepared DMSO (100 μL). Before the cytotoxicity
measurement, the plate was agitated gently for 15 min, and then the
absorbance intensity at 450 nm was recorded by a microplate reader.
The relative cell viability (%) for each sample was finally calculated.
The absolute quantum yields of MTSQ and compound 3 in ethanol
and PBS buffer solution were determined using a spectrofluorometer
(FLSP920, Edinburgh Instruments LTD) equipped with an integrating
sphere. The integrating sphere consists of a 120 mm inside diameter
spherical cavity, which is machined from BENFLEC block. A 3 mL
sample of MTSQ solution was sealed in a quartz cell (10 mm × 10 mm),
and its maximal absorption intensity was measured in the range of
0.15–0.40. The same volume of ethanol or PBS buffer solution was used
as the blank sample. The excitation wavelength was set at 630 nm, and
the emission spectral range was from 640 to 800 nm.
2.8. Cell incubation and fluorescence microscopy imaging
The HeLa cells were grown in RPMI-1640 culture medium supple-
mented with 1.0% penicillins and treptomycin and 10% fetalbovine
serum (FBS) in an atmosphere of 5% CO2 at 37 °C. After incubation for
24 h, the cells were treated with 4.0 μM solutions of MTSQ in culture
medium and incubated at 37 °C for 1.5 h, the culture medium was then
removed and the treated cells were washed with PBS solutions three
2.7. MTT assay
The HeLa cells were seeded in a 96-well plate at a concentration of
1.0 × 104 cells/well, and maintained overnight at 37 °C in an incubator
containing a 5% CO2 humidified environment. The cells were treated
Fig. 2. (a) The fluorescence enhancement ratio (I/I0) at 662 nm of MTSQ (2.0 μM) after the addition of Hg2+ in different ethanol/water mixtures. (b) Time-
dependent normalized fluorescence emission intensity of MTSQ at 662 nm after the addition of Hg2+ (40 μM) in different ethanol/water mixtures (λex = 630 nm).
371

G. Wang et al.
DyesandPigments157(2018)369–376
Fig. 4. (a) The fluorescence spectra of MTSQ
(2.0 μM) in the presence of different metal ions
(40 μM)
in
pH
7.0 PBS
buffer
solution
(λex = 630 nm). Inset: the color changes of MTSQ
with the addition of different metal ions. (b) The
competition experiments of Hg2+ to different metal
ions at 657 nm in pH 7.0 PBS buffer solution. Black
bar: The fluorescence intensity of MTSQ (2.0 μM)
with the addition of the respective competing metals
(40 μM). Striated bar: The fluorescence intensity of
subsequent introduction of Hg2+ (40 μM) to the
system (λex = 630 nm). (For interpretation of the
references to color in this figure legend, the reader is
referred to the Web version of this article.)
IR, ESI-MS, and HRMS(ESI) (Figs. S1–S12, Supporting Information). It
was anticipated that reaction with Hg2+ would lead to formation of an
intramolecular hydrogen bond driving this desulfurization reaction to
completion more rapidly and leading to the formation of a more
fluorescent compound 3. At the same time, the addition of Hg2+ would
induce the disassembly of the aggregated structure, responsible for the
enhanced fluorescence. The response mechanism of probe MTSQ to
Hg2+ is depicted in Scheme 2.
3.2. Effect of self-assembly on sensing of Hg2+
Squaraine dyes are prone to self-assembly in aqueous solution, and
the polarity and the composition of solvents have a great influence on
the self-assembly and spectral properties of the dyes. We studied the
self-assembly of MTSQ in different ethanol-water mixtures utilizing the
UV absorption spectra technique. As shown in Fig. 1, aggregation is
induced above a critical water concentration of ∼90% vol. Above this
threshold value, the monomer band at 652 nm greatly reduced to
∼10% of its initial intensity, along with the appearance of a new blue-
shifted broad band at 562 nm and a tail in the region between 700 and
800 nm, indicating that there is a strong π−π stacking interaction be-
tween the molecules in the aggregates.
Fig. 5. The absorption spectra change of MTSQ (2.0 μM) upon addition of Hg2+
(0–2.0 μM) in pH 7.0 PBS buffer solution.
times and incubated with Hg2+ in culture medium for 1 h for imaging.
(excitation wavelength: 630 nm; emission wavelength: 660 nm).
To test our hypothesis that the self-assembling properties would
promote the performance of the dye in sensing Hg2+, we have sys-
tematically studied the fluorescence of MTSQ before and after the ad-
dition of mercury ions in different ethanol/water mixtures with an
ethanol content of 0, 20%, 40%, 60%, 80% and 100%, respectively. As
anticipated, MTSQ displayed a weak fluorescence background in pure
water solution, while it exhibits high fluorescence intensity in ethanol
solutions, which is consistent with the UV–Vis absorption spectra. Upon
the addition of Hg2+, the fluorescence intensities of the solutions all
had a certain degree of enhancement, and the fluorescence enhance-
ment ratio in pure water was obviously higher than those of ethanol/
water mixtures (Fig. 2a). These comparisons show the key role of the
self-assembling process that could indeed promote the sensitivity and
performance of the dye in sensing Hg2+ by lowering the background
fluorescence intensity.
We also investigated the kinetics of interaction between MTSQ and
Hg2+ in different ethanol/water mixtures by the fluorescence emission
technique. As can be seen in Fig. 2b, when Hg2+ is added to a mixture
with an ethanol content of more than 20%, the fluorescence is clearly
enhanced and became essentially stable within 10 min, whereas it took
about 25 min in pure water to acquire fluorescence stability. The results
suggested that the formation of aggregates would result in steric
crowding, decreasing the reaction rate.
2.9. Imaging of exogenous Hg2+ in vivo
Zebrafish were maintained at 28 °C in E3 embryo media. The 5-day
old zebrafish were incubated with 4.0 μM MTSQ in PBS buffer for
20 min. After washing with PBS to remove the remaining MTSQ, the
zebrafish was further incubated with 10 μM Hg2+ in PBS buffer for
15 min. The zebrafish was imaged using a confocal microscope (ex-
citation wavelength: 630 nm; emission wavelength: 660 nm).
3. Results and discussion
3.1. Design and synthesis of MTSQ
The chemical structure and synthetic procedure of MTSQ are out-
lined in Scheme 1. The design of probe MTSQ is based on the following
considerations. Firstly, the substitution of the oxygen atom in the
squaraine with sulfur atom was designed to get good selectivity based
on thiophilic nature of mercury ion and a weak fluorescent emission as
compared to its all-oxygen analogue due to the heavy atom effect.
Secondly, squaraines have a high tendency to form aggregates in aqu-
eous solution, which results in the fluorescence quenching. Moreover,
hydroxyl substitutions on the squaraine phenyl rings would not only
increase the yield of squaraine but also result in greater resistance to
nucleophilic attack by water [28].
3.3. Spectroscopic properties of probe MTSQ toward Hg2+ in PBS solution
Accordingly, we synthesized the compound MTSQ by condensation
of the corresponding aniline derivatives and squaric acid in butanol-
benzene mixtures and subsequent sulfurization with Lawesson's re-
agent. The obtained MTSQ was characterized by 1H NMR, 13C NMR, FT-
Inspired by the above observation, we subsequently investigated the
sensitivity of the probe for monitoring Hg2+. The emission variation of
MTSQ in PBS solution was monitored during titration experiments with
372

G. Wang et al.
DyesandPigments157(2018)369–376
Fig. 6. Dynamic light scattering analysis (a, b), FF-TEM images (c, d) of MTSQ (2.0 μM) in pH 7.0 PBS buffer solution before and after the addition of Hg2+ (40 μM),
respectively.
3.4. Selectivity of probe MTSQ toward Hg2+
The selectivity evaluations of MTSQ toward Hg2+ against poten-
tially interfering ions (Fe3+, Ni2+, Zn2+, Cd2+, Ag+, Sr2+, Al3+
,
Mn2+, Pb2+, Co2+, Cu2+, Li+, Na+, K+, Mg2+, Ca2+, Ba2+) and an-
ions such as SCN−, S2O32−, S2−, SO42−, I−, ClO4−, CH3COO−, NO3
,
−
−
2−
−
HCO3
,
CO3
, NO2 were performed (Fig. 4a, and Fig. S15,
Supporting Information). The fluorescence intensity of MTSQ increased
remarkably in the presence of Hg2+, while little increase in fluores-
cence were observed after addition of other metal ions (40 μM). No-
tably, Ag+, Pb2+, and Cd2+, which easily competes with Hg2+ detec-
tion, caused negligible fluorescence intensity increase in this work. The
competition experiments showed that the response of MTSQ to Hg2+
Scheme 3. Schematic illustration of Hg2+-triggered morphological change.
various concentrations of Hg2+. The fluorescence intensity of the probe
gradually increased with increasing Hg2+ concentration and reached a
plateau at ca. 1.0 equivalent per repeat unit of Hg2+ (Fig. 3). Upon
irradiation by a UV lamp at 330 nm, the solution of MTSQ was nearly
dark while upon addition of Hg2+ the solution emitted strong red
fluorescence (Fig. 3, inset). A good linear relation was found for the
titration experiments between the fluorescence intensity and the Hg2+
concentration from 0 to 1.0 μM (R2 = 0.996) (Fig. S13, Supporting
Information). A very low detection limit of MTSQ towards Hg2+ was
measured as 1.2 nM with a signal to noise ratio of 3, which is much
lower than the limit of Hg2+ in drinking water (10 nM) set by the
United States Environmental Protection Agency. Furthermore, pH ti-
tration experiments indicated the response capability of MTSQ to Hg2+
at physiological pH 3.0 to 7.0 (Fig. S14, Supporting Information). The
spectral response clearly showed that MTSQ is a highly sensitive off-on
sensor for Hg2+ ions.
was unaffected in the background of each of the metal ions and anions
2−
mentioned above except for S2O3
and I−, which had some inter-
ference with the fluorescence response (Fig. 4b, and Fig. S16,
Supporting Information). The unique specificity of MTSQ towards
Hg2+ in PBS buffer solution indicated that MTSQ can be used as an
excellent NIR Hg2+ probe for practical detection.
3.5. Response mechanism
To demonstrate that compound 3 is the product of Hg2+-induced
desulfurization, the ESI-MS spectra of MTSQ were recorded in the ab-
sence and presence of Hg2+ (Figs. S17 and S18, Supporting
Information). The peaks at m/z 721.7 (calcd as 721.3370) and 743.6
(calcd as 743.3190) corresponded to [MTSQ+H]+ and [MTSQ+Na]+
respectively. After addition of Hg2+, peaks appeared at 705.4 (calcd as
705.3599) and 727.3 (calcd as 727.3418), were assigned to the de-
sulfurized products [3+H]+ and [3+Na]+. The results of ESI-MS ex-
periments verified the mechanism proposed in Scheme 2, in which
373

G. Wang et al.
DyesandPigments157(2018)369–376
Fig. 7. Images of living HeLa cells. (a) Fluorescent images of HeLa cells incubated with MTSQ (4.0 μM) for 1.5 h, (b) Fluorescence image of living cells pretreated
with MTSQ (4.0 μM) for 1.5 h and further incubated with Hg2+ (10 μM) for 1 h (c, d) bright field, (e, f) Merged images.
Fig. 8. Images of live zebrafish. (a) Fluorescent images of live zebrafish incubated with MTSQ (4.0 μM) for 20 min, (b) Fluorescence image of live zebrafish pretreated
with MTSQ (4.0 μM) for 20 min and further incubated with Hg2+ (10 μM) for 15 min (c, d) bright field, (e, f) Merged images.
Hg2+ induces the desulfurization of MTSQ and the formation of com-
pound 3.
with a mean diameter of 358 nm in PBS solution, while after addition of
Hg2+ the mean diameter decreased to 220 nm (Fig. 6a and b). The
results confirm that MTSQ can self-assemble into aggregates in PBS
buffer solution and that the addition of Hg2+ results in induced dis-
assembly of the aggregates.
The absolute quantum yield of MTSQ and compound 3 in ethanol
were determined to be 0.163 and 0.647, respectively. It can be clearly
observed that the introduction of sulfur atoms in the squaraine drasti-
cally decreased the quantum yield. In addition, the absolute quantum
yield of MTSQ in PBS buffer solution was also examined, while it was
too low to be measured, suggesting that aggregation-caused quenching
effectively eliminated the fluorescence of MTSQ.
Freeze-fracture TEM (FF-TEM) imaging techniques were applied to
gain
a deeper insight into the morphologies of the aggregates.
Observation by FF-TFM, as shown in Fig. 6c, clearly indicates that
MTSQ in PBS solutions forms ordered and close-stacked aggregates
with a width of ∼30 nm. The wormlike micelles are several hundred
nanometers long, and some of them even extend to more than 1 μm.
After the addition of Hg2+, the morphology changed from the wormlike
micelles structures into a scattered distribution of globular aggregates
with a mean diameter of 100 nm (Fig. 6d). Observations of the changes
of morphology from extended structures to nanoparticles also reflects
Hg2+-triggered significantly disassembly of aggregates.
To investigate the dynamic process of Hg2+-triggered disassembly,
we monitored the change in absorption of MTSQ in PBS solution with
addition of Hg2+ by UV–Vis. As shown in Fig. 5, MTSQ exhibited a very
weak monomeric absorption band at 652 nm accompanied by a blue-
shifted band at 562 nm and a tail in the region between 700 and
800 nm, which is characteristic of the H-aggregation state of the dye.
With the addition of Hg2+, the monomer absorption band at 652 nm
gradually increased and blue-shifted to 638 nm, and concomitantly, the
blue-shifted band at 562 nm and the tail band observed between 700
and 800 nm disappeared. Lastly, the absorption spectra show only a
single sharp band with λmax at 638 nm. This result supported the pro-
posed sensing mechanism of Hg2+-triggered disaggregation.
To further verify the Hg2+ induced disassembly of the aggregated
structure, dynamic light scattering analysis (DLS) was carried out. The
DLS measurements revealed the presence of nanoaggregates of MTSQ
A graphic model of Hg2+-triggered morphological change of MTSQ
in PBS solution is shown in Scheme 3. We propose that MTSQ self-
assembles in PBS solution to form ordered and close-stacked wormlike
micelles due to the π−π stacking between the electron-rich benzene
and the electron-poor four-membered ring. When Hg2+ was added to
the aqueous solution, MTSQ underwent desulfurization, resulting in the
formation of a bola-type amphiphile [39]. The close-packed arrange-
ment of MTSQ was broken down gradually. Meanwhile, bola-type
374

G. Wang et al.
DyesandPigments157(2018)369–376
amphiphiles further self-assembled to spherical micelles.
Biochem 2006;43:257–68.
[2] Zahir F, Rizwi SJ, Haq SK, Khan RH. Low dose mercury toxicity and human health.
Environ Toxicol Pharmacol 2005;20:351–60.
3.6. Fluorescence imaging in live cells
[3] Lin W, Cao X, Ding Y, Yuan L, Long L. A highly selective and sensitive fluorescent
probe for Hg2+ imaging in live cells based on a rhodamine-thioamide-alkyne
scaffold. Chem Commun 2010;46:3529–31.
[4] Du J, Fan J, Peng X, Sun P, Wang J, Li H, Sun S. A new fluorescent chemodosimeter
for Hg2+: selectivity, sensitivity, and resistance to Cys and GSH. Org Lett
2010;12:476–9.
To evaluate the biocompatibility of MTSQ, an MTT assay was per-
formed and indicated little cytotoxicity on cell viability at the con-
centration of 1–100 μM with the HeLa cells after incubation for 24 h
(Fig. S19, Supporting Information). The low cytotoxicity and good so-
lubility in aqueous conditions as well as the retained NIR photophysical
properties encouraged us to further investigate the imaging capability
of MTSQ both in vitro and in vivo.
[5] Chen Y, Zhu C, Yang Z, Li J, Jiao Y, He W, Chen J, Guo Z. A new “turn-on” che-
modosimeter for Hg2+: ICT fluorophore formation via Hg2+-induced carbaldehyde
recovery from 1,3-dithiane. Chem Commun 2012;48:5094–6.
[6] Guo Z, Zhu Wei, Zhu M, Wu X, Tian H. Near-infrared cell-permeable Hg2+-selective
ratiometric fluorescent chemodosimeters and fast indicator paper for MeHg+ based
on tricarbocyanines. Chem Eur J 2010;16:14424–32.
Firstly, we employed MTSQ for imaging of intracellular Hg2+ by a
confocal laser scanning microscopy with HeLa cells. The HeLa cells
were first cultured with MTSQ (4.0 μM) for 1.5 h at 37 °C, then washed
with PBS buffer for three times and incubated with Hg2+ in culture
medium for 1 h. As shown in Fig. 7a, a weak fluorescence was observed
when HeLa cells were treated only with MTSQ, while ultrabright in-
tracellular fluorescence from HeLa cells stained further with Hg2+ was
clearly observed (Fig. 7b). The results reveal that MTSQ can be used to
[7] Sklavounos AA, Pefkianakis EK, Toubanaki DK, Vougioukalakis GC, Calokerinos AC.
A squaraine derivative for cost-effective, quick, and highly sensitive determination
of mercury and thiols and pH sensing. ChemPlusChem 2016;81:913–6.
[8] Koteeswari R, Ashokkumar P, Malar EJP, Ramakrishnan VT, Ramamurthy P. Highly
selective, sensitive and quantitative detection of Hg2+ in aqueous medium under
broad pH range. Chem Commun 2011;47:7695–7.
[9] Maity SB, Banerjee S, Sunwoo K, Kim JS, Bharadwaj PK. A fluorescent chemosensor
for Hg2+ and Cd2+ ions in aqueous medium under physiological pH and its ap-
plications in imaging living cells. Inorg Chem 2015;54:3929–36.
[10] Zhang X, Xiao Y, Qian X. A ratiometric fluorescent probe based on FRET for imaging
visualize the cellular Hg2+
.
Hg2+ ions in living cells. Angew Chem 2008;47:8025–9.
[11] Bera K, Das AK, Nag M, Basak S. Development of a rhodamine-rhodanine-based
fluorescent mercury sensor and its use to monitor real-time uptake and distribution
of inorganic mercury in live zebrafish larvae. Anal Chem 2014;86:2740–6.
[12] Chen C, Wang R, Guo L, Fu N, Dong H, Yuan Y. A squaraine-based colorimetric and
“turn on” fluorescent sensor for selective detection of Hg2+ in an aqueous medium.
Org Lett 2011;13:1162–5.
[13] Ru J, Chen X, Guan L, Tang X, Wang C, Meng Y, Zhang G, Liu W. Ratiometric
iridium(III) complex-based phosphorescent chemodosimeter for Hg2+ applicable in
time-resolved luminescence assay and live cell imaging. Anal Chem
2015;87:3255–62.
3.7. Fluorescence imaging in live zebrafish
In vivo fluorescence imaging was then carried out using a live 5-
day-old zebrafish. The fluorescence images were acquired from the
dorsal side of the zebrafish. As shown in Fig. 8a, after incubation with
4.0 μM MTSQ for 20 min at 21
1 °C, the zebrafish showed only
negligible luminescence. However, after further treatment with 10 μM
Hg2+, the luminescence in the zebrafish was remarkably enhanced
(Fig. 8b). Both in vitro and in vivo fluorescence imaging experiments
demonstrate the high stability and potential of MTSQ as an NIR ima-
ging probe for fluorescence microscopy applications.
[14] Arivazhagan C, Borthakur R, Ghosh S. Ferrocene and triazole-appended rhodamine
based multisignaling sensors for Hg2+ and their application in live cell imaging.
Organometallics 2015;34:1147–55.
[15] Zang J, Li C, Zhou K, Dong H, Chen B, Wang F, Zhao G. Nanomolar Hg2+ detection
using β-lactoglobulin-stabilized fluorescent gold nanoclusters in beverage and
biological media. Anal chem 2016;88:10275–83.
[16] Makam P, Shilpa R, Kandjani AE, Periasamy SR, Sabri YM, Madhu C, Bhargava SK,
Govindaraju T. SERS and fluorescence-based ultrasensitive detection of mercury in
water. Biosens Bioelectron 2018;100:556–64.
4. Conclusion
[17] Gao Y, Ma T, Ou Z, Cai W, Yang G, Li Y, Xua M, Li Q. Highly sensitive and selective
turn-on fluorescent chemosensors for Hg2+ based on thioacetal modified pyrene.
Talanta 2018;178:663–9.
[18] Zhai D, Xu W, Zhang L, Chang YT. The role of “disaggregation” in optical probe
development. Chem Soc Rev 2014;43:2402–11.
[19] Jisha VS, Arun KT, Hariharan M, Ramaiah D. Site-selective binding and dual mode
recognition of serum albumin by a squaraine dye. J Am Chem Soc 2006;28:6024–5.
[20] Sreejith S, Divya KP, Ajayaghosh A. A near-infrared squaraine dye as a latent ra-
tiometric fluorophore for the detection of aminothiol content in blood plasma.
Angew Chem Int Ed 2008;47:7883–7.
[21] Xu Y, Malkovskiy A, Pang Y. A graphene binding-promoted fluorescence enhance-
ment for bovine serum albumin recognition. Chem Commun 2011;47:6662–4.
[22] Xu Y, Liu Q, Li X, Wesdemiotis C, Pang Y. A zwitterionic squaraine dye with a large
Stokes shift for in vivo and site-selective protein sensing. Chem Commun
2012;48:11313–5.
In summary, we have developed a sensitive and selective probe for
detecting and imaging of Hg2+ in live cells and zebrafish, utilizing the
strategy of the incorporation of the heavy atom effect and self-assembly
caused quenching. This Hg2+ probe exhibits high selectivity against
metal ions and a low detection limit of 1.2 nM. The remarkable ana-
lytical performance, together with the unique properties of MTSQ,
makes the probe successfully employed in live cells and zebrafish
imaging. This methodology may also be extended to design and develop
biosensors of other biological species and provides practicability for
further research of physiological and pathological processes.
[23] Sreejith S, Ma X, Zhao Y. Graphene oxide wrapping on squaraine-loaded meso-
porous silica nanoparticles for bioimaging. J Am Chem Soc 2012;134:17346–9.
[24] Xu W, Zhu X, Wang G, Sun C, Zheng Q, Yang H, Fu N. Host–guest assembly of
adamantyl tethered squaraine in β-cyclodextrin for monitoring pH in living cells.
RSC Adv 2014;4:52690–3.
[25] Gao FP, Lin YX, Li LL, Liu Y, Mayerhöffer U, Spenst P, Su JG, Li JY, Würthner F,
Wang H. Supramolecular adducts of squaraine and protein for noninvasive tumor
imaging and photothermal therapy in vivo. Biomaterials 2014;35:1004–14.
[26] Sreejith S, Joseph J, Lin M, Menon NV, Borah P, Ng HJ, Loong YX, Kang Y, Yu SWK,
Zhao Y. Near-infrared squaraine dye encapsulated micelles for in vivo fluorescence
and photoacoustic bimodal imaging. ACS Nano 2015;9:5695–704.
[27] Grande V, Doria F, Freccero M, et al. An aggregating amphiphilic squaraine: alight-
up probe that discriminates parallel G-quadruplexes. Angew Chem Int Ed
2017;56(26):7520–4.
[28] Law KY. Squaraine chemistry: effects of structural changes on the absorption and
multiple fluorescence emission of bis[4-(dimethylamino)phenyl]squaraine and its
derivatives. J Phys Chem 1987;91:5184–93.
[29] Das S, Thomas KG, Thomas KJ, Madhavan V, Liu D, Kamat PV, George MV.
Aggregation behaviour of water soluble bis(benzothiazolydene)squaraine deriva-
tives in aqueous media. J Phys Chem 1996;100:17310–5.
Acknowledgements
This work was financially supported by the Program for Chang
Jiang Scholars and Innovative Research Team in University of Ministry
of Education of China (IRT15R11), the Fujian Provincial Natural
Science Foundation, P.R. China (No. 2017J01577), the Open Project
Program of the State Key Laboratory of Photocatalysis on Energy and
Environment, Fuzhou University (No. SKLPEE-KF201724, SKLPEE-
KF201727), and Funding (Type A) from Fujian Education Department,
P.R. China (No. JAT170104). Dedicated to Prof. Thomas T. Tidwell on
the occasion of his eightieth birthday.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://dx.
doi.org/10.1016/j.dyepig.2018.05.016.
[30] Chen H, Farahat MS, Law KY, Whitten DG. Aggregation of surfactant squaraine dyes
in aqueous solution and microheterogeneous media: correlation of aggregation
behavior with molecular structure. J Am Chem Soc 1996;118:2584–94.
[31] Geiger C, Stanescu M, Chen L, Whitten DG. Organogels resulting from competing
self-assembly units in the gelator: structure, dynamics, and photophysical behavior
of gels formed from cholesterol-stilbene and cholesterol-squaraine gelators.
References
[1] Magos L, Clarkson TW. Overview of the clinical toxicity of mercury. Ann Clin
375

G. Wang et al.
DyesandPigments157(2018)369–376
Ajayaghosh A, Armaroli N. Ultrasound stimulated nucleation and growth of a dye
assembly into extended gel nanostructures. Chem Eur J 2013;19:12991–3001.
[37] Ajayaghosh A, Chithra P, Varghese R. Self-assembly of tripodal squaraines: cation-
assisted expression of molecular chirality and change from spherical to helical
morphology. Angew Chem 2007;119:234–7.
Langmuir 1999;15:2241–5.
[32] Mayerhöffer U, Würthner F. Cooperative self-assembly of squaraine dyes. Chem Sci
2012;3:1215–20.
[33] Ohsedo Y, Miyamoto M, Oono M, Shikii K, Tanaka A. Hydrogel formed by a simple
squaric acid derivative. RSC Adv 2013;3:3844–7.
[34] Jyothish K, Hariharan M, Ramaiah D. Chiral supramolecular assemblies of a
squaraine dye in solution and thin films: concentration-, temperature-, and solvent-
induced chirality inversion. Chem Eur J 2007;13:5944–51.
[35] Zhang C, Zhang X, Zhang X, Fan X, Jie J, Chang JC, Lee CS, Zhang W, Lee ST. Facile
one-step growth and patterning of aligned squaraine nanowires via evaporation-
induced self-assembly. Adv Mater 2008;20:1716–20.
[38] Wang G, Xu W, Guo Y, Fu N. Near-infrared squaraine dye as a selective protein
sensor based on self-assemble. Sensor Actuator B Chem 2017;245:932–7.
[39] Chi XD, Zhang HC, Vargas-Zúñiga GI, Peters GM, Sessler JL. A dual-responsive bola-
type supra-amphiphile constructed from a water-soluble calix[4]pyrrole and a tet-
raphenylethene-containing pyridine bis-N-oxide. J Am Chem Soc
2016;138:5829–32.
[36] Malicka JM, Sandeep A, Monti F, Bandini E, Gazzano M, Ranjith C, Praveen VK,
376