Near-infrared pH responsive heptamethine cyanine platforms: Modulating the proton acceptor

Article abstract of DOI:10.1016/j.dyepig.2020.108611

In this report, we wish to present the rational design, synthesis, and study of optical characteristics of near-infrared (NIR) molecular probes emitting over 800 nm. The title pH sensing platforms consist of a piperazine moiety serving as a hydronium ion

Full text of DOI:10.1016/j.dyepig.2020.108611

Journal Pre-proof
Near-infrared pH responsive heptamethine cyanine platforms: Modulating the proton
acceptor
Atanas Kurutos, Yutaka Shindo, Yuki Hiruta, Kotaro Oka, Daniel Citterio
PII:
S0143-7208(20)31027-5
DOI:
https://doi.org/10.1016/j.dyepig.2020.108611
DYPI 108611
Reference:
To appear in: Dyes and Pigments
Received Date: 27 April 2020
Revised Date: 28 May 2020
Accepted Date: 29 May 2020
Please cite this article as: Kurutos A, Shindo Y, Hiruta Y, Oka K, Citterio D, Near-infrared pH responsive
heptamethine cyanine platforms: Modulating the proton acceptor, Dyes and Pigments (2020), doi:
https://doi.org/10.1016/j.dyepig.2020.108611.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Ltd.

Atanas Kurutos: Conceptualization, Formal analysis, Investigation, Writing – Original Draft, Visualization,
Funding acquisition. Yutaka Shindo: Investigation, Validation, Formal analysis, Writing – Review &
Editing. Yuki Hiruta: Resources, Supervision. Kotaro Oka: Resources, Supervision. Daniel Citterio:
Resources, Writing – Review & Editing, Supervision, Project administration, Funding acquisition.

Near-infrared pH responsive heptamethine cyanine platforms:
Modulating the proton acceptor
a,b,*
c
a
c, d, e
a, **
Atanas Kurutos
, Yutaka Shindo , Yuki Hiruta , Kotaro Oka
, Daniel Citterio
a
Department of Applied Chemistry, Faculty of Science and Technology, Keio University, 3-
4-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa, 223-8522, Japan
1
b
Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of
Sciences, Acad. G. Bonchev str., bl. 9, 1113 Sofia, Bulgaria
c
Department of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Kohoku-ku,
Yokohama, Kanagawa, 223-8522, Japan
d
Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University,
Kaohsiung, Taiwan
e
Waseda Research Institute for Science and Engineering, 2-2 Wakamatsucho, Shinjuku,
Tokyo, Japan.
*
*
corresponding author e-mail address: akurutos@keio.jp
* corresponding author e-mail address: citterio@applc.keio.ac.jp
Keywords: cyanine dyes; NIR molecular probes; pH response; optical properties; cellular
imaging; fluorescence microscopy.
1

Abstract
In this report, we wish to present the rational design, synthesis, and study of optical
characteristics of near-infrared (NIR) molecular probes emitting over 800 nm. The title pH
sensing platforms consist of a piperazine moiety serving as a hydronium ion receptor,
coupled with a rigidified symmetric heptamethine backbone. The spectrally fine-tuned
response of the fluorophores was studied in both organic solvents and aqueous buffer
solutions, employing combined UV-Vis and steady state fluorimetric techniques. In parallel,
computational approach and optical spectroscopy were employed aiming to ascribe the origin
of spectral behavior of the amino-substituted heptamethines. Herein, we also investigated the
halochromic effect of the differently substituted piperazines on the spectral properties of the
NIR cyanine dyes. Under acidic conditions, the protonation of the piperazine moiety
promoted a gradual bathochromic shift in the absorption spectra. The Δλ between the neutral
and the protonated forms are indirectly related to the modifications applied to the proton
acceptor. The calculated pK values (5.13-7.06) were found to be dependent on the dye
a
chemical structure. A good linear correlation was established between the pK of the newly
a
synthesized dyes and that of the reported piperazine moieties. This aspect allows the future
expansion of the current facile strategy since sensing chromophores with predicted pH
response could readily be designed based on pK databases for substituted piperazine
a
derivatives. A detailed literature study along with NMR and computational methods were
found to be in agreement with our assumption, that both nitrogen atoms of the 1,4-
disubstituted amine are involved in the sensing process. The covered pK range allows us to
a
recommend the title fluorophores as promising potential candidates for qualitative analysis of
biological samples. A cytotoxicity study by the MTT assay showed that the NIR dyes are safe
and applicable for bioanalytical purposes at dye concentrations up to 10 μM. Fluorescence
2

imaging and cellular applications, including staining and visualization of HeLa cervical
cancer cells was successfully demonstrated.
Introduction
The hydronium ion is one of the most important species, known to play a major role in
+
biological systems. Monitoring processes with acid/base sensors at variable H concentration
is of a significant importance in the field of life sciences [1]. Abnormal pH values in tissues
and cells have a great impact on the activities and normal functions of our body. Thus,
intracellular pH is vital in modulating organelles’ functions. The pH of intracellular vesicles
including lysosomes, endosomes and phagosomes typically ranges between 4.5-6.5 [2,3].
Acidic environments are also known to be associated with a wide variety of renal conditions,
solid tumors, asthma, and cystic fibrosis [4–6]. For instance, the acidosis strategy relies on
+
the acidic extracellular metabolic environment of tumors due to an H /lactate co-transporter
+
+
system and an activated Na /H exchanger (NHE1) [7]. Therefore, the development of novel
sensors capable of rapid and precise monitoring of pH changes inside tissues and cells is a
highly demanding assignment.
Near-infrared probes emitting between 650-900 nm are considered as powerful tools and
have been widely utilized for the detection of acidic tumor environments and monitoring pH
changes in vivo. The advantages of operating in the NIR region are: i) low background
fluorescence from endogenous molecules, ii) deep tissue propagation, and iii) low risk of
photodamage upon irradiation [8]. Another distinctive advantage is the relatively low energy
required to promote molecular excitation. Fluorescence-based detection methods are known
for their simplicity and high sensitivity, hence inexpensive diode laser sources can effectively
be used to excite NIR fluorogenic compounds [12]. Due to their excellent spectral and
physical properties, cyanine dyes have attracted immense interest and have been employed as
fluorescent labels and biosensors for in vivo imaging [10,11], theranostic agents for
3

photothermal therapy (PTT) [12–16], detection of analytes [17], amyloid lipid aggregation
18,19], protein staining [20], DNA and RNA labeling [21,22], sensing of viruses, [23],
[
photosensitizers [24], sequencing assays [25] and flow cytometry [26]. Among several
classes of organic compounds operating in the NIR region, cyanines are the preferred tools
for cancer imaging due to their high molar extinction coefficients and great photothermal
and/or photodynamic effect. What is more, a wide range of available synthetic strategies and
structural modification tailoring tunable absorption and fluorescence emission wavelengths
can be found reported in the art [27]. Appropriate substituents can be introduced to improve
the optical properties and provide chemical affinity to specific biotargets. The design and
synthesis of extended chromophoric systems coupled with complex recognition units is often
a challenging task. Therefore, the development of alternative sensing protocols is currently
gaining tremendous popularity.
Heptamethine cyanine dyes with an incorporated cyclohexenyl ring along the polymethine
chain provide a reactive site for chemical modifications. More appropriately, a nucleophilic
substitution of the chlorine atom at meso position is an example of a facile approach to the
development of functional chromophores absorbing in the wavelength region > 650 nm. To
date, several indole based bimodal fluorescent substrates including norcyanines and
tricarbocyanines are known [28–31]. However, to the best of our knowledge, no systematic
correlation of chemical structure and pK fine-tuning can be found in the literature. As a
a
matter of fact, the development of probes bearing sharp response is subject of delicate
molecular design and chromophore selection. In this aspect, our efforts were emphasized on
the design of NIR fluorescent probes responding at a well-defined pH range. The reactive site
of the heptamethine scaffold has been exploited to introduce a series of substituents on the
central cyclohexenyl ring via nucleophilic substitution of the chlorine atom with various
amines. In this paper, we report on new strategy enabling the flexible control of the pK range
a
4

by introducing several piperazines serving as the modulating site (Fig. 1). Last but certainly
not least, our present research is directed towards the demonstration of bio-applicability for
cellular staining using the current NIR heptamethine cyanine platforms in the visualization of
HeLa cervical cancer cells as a proof-of-concept.
Figure 1. NIR cyanine platforms bearing modulated proton acceptor.
2
. Materials and methods
2
.1. Analysis methods and equipment
All starting materials and solvents required for the synthesis were purchased from
Sigma-Aldrich, Tokyo Chemical Industry Co. Ltd., Merck KGaA, BLD Pharmatech
Ltd. or Oakwood Products Inc. and used as supplied without any further purification.
The spectral grade solvents used for the UV-Vis and fluorimetric analyses were
purchased from Wako Pure Chemical Industries Ltd. and Kanto Chemical Co. Inc.
Deuterated solvents were obtained from Kanto Chemical Co. Inc. and Merck KGaA
and used as supplied. All other starting materials and solvents were commercial
products of analytical grade. Purification of the target pH responsive dyes was
performed by flash column chromatography using silica gel 60 (0.063-0.200 mm) and
a mobile phase DCM by increasing the methanol gradient (0 - 8 % v/v). Thin-layer
chromatography (TLC) was used for monitoring the reaction outcome using Merck
silica gel 60 F254 glass plates, Merck silica gel 60 RP-18 F254s glass plates and
5

aluminium oxide 60 F254 neutral plastic plates. Spots were visualised under a UVGL-
5 Compact UV Lamp Dual wavelength 254/365 nm and power of 4 Watt. The
2
identity of the compounds was confirmed by various spectroscopic techniques
1
including NMR spectroscopy and high-resolution mass spectrometry. The H-NMR
13
and C-NMR spectra were recorded using 5 mm tubes on JEOL JNM-ECA500 at 500
MHz / 125 MHz and JEOL JNM-ECZ400S at 400 MHz / 100 MHz, respectively, in
o
DMSO-d and CDCl at 25 C. The signals corresponding to the residual protons of
6
3
DMSO-d at 2.50 ppm were used as an internal reference. All chemical shifts (δ) are
6
quoted in ppm with an accuracy of 0.01 ppm. Coupling constants
J are expressed in
Hz and determined with an accuracy of 0.1 Hz. The following abbreviations are
1
quoted to describe spin multiplicity in H-NMR:
s
= singlet,
d
= doublet,
t
= triplet,
q
=
quartet, dd = doublet of doublets, ddd = doublet of doublets of doublets,
m
=
multiplet. Spectral processing was done using MestreNova v. 14.1.1 by Mestrelab
Research S.L. HRMS of all compounds were acquired on a Waters Xevo G2-SQTof
MS - Waters LCT Premier XE (ESI-TOF).
2
.2. Optical properties of the NIR cyanines in organic solvent
The UV-visible absorption spectra were recorded at room temperature using 10-mm
path-length quartz cells on a Shimazu UV-3600 Plus UV-VIS-NIR spectrophotometer.
Steady state fluorescence measurements were performed at room temperature using
1
0-mm path-length quartz cells on a Shimazu RF-6000 spectrofluorometer. Scientific
data analysis and spectral processing were conducted on OriginPro 2019 graphing and
analysis software - v. 9.6.0172 by OriginLab Corporation (www.originlab.com). The
relative fluorescence quantum yield of the studied functionalised NIR heptamethine
cyanine dyes was calculated in spectroscopic grade acetonitrile according to the
formula expressed in Equation 1 [32]:
6

ꢈ
ꢃꢄ
ꢆꢅ
ꢇ
ꢄ
ꢀꢁ
= ꢀ × × ꢆꢄ × ꢇ (Eq.1)
ꢂ
ꢈ
ꢃꢅ
ꢅ
where,
Φ
is the fluorescence quantum yield,
I
represent the integrated fluorescence
band of the relative intensity upon excitation at wavelength λ , A is the absorption
abs
intensity, and
n is the refractive index of the solvent (1.3414 for acetonitrile [33]). The
superscripts x and R refers to the unknown sample and the used reference – the
heptamethine cyanine dye IR-780 (Sigma Aldrich) of a reported fluorescence quantum
yield (Φr = 7.6 % in methanol [34]). Acetonitrile was employed for the current
investigation since the nature of the polar protic methanol could potentially result in
artefacts due to the possibility of hydrogen bonding with the dye. The photochemical
stability was studied in acetonitrile solutions by recording the intensity of the
fluorescence emission peak at the λem on continuous irradiation with a 150-Watt Xe
lamp at the corresponding λabs over a duration of 3 hours. The measured fluorescence
intensity at the maxima is expressed in percentage relative to that of the original
signal.
2
.3. UV-Vis and spectrofluorimetric pH response measurements in buffer solutions
The stock solution of the dyes were prepared in DMSO and used for further dilution
with aqueous buffer for the preparation of the working solutions at 2 μM (10 %
DMSO as co-solvent [30]). Deionized water was used from a Waters Milli-Q system
(
Bedford, MA) with a resistivity of 18.2 MΩ at 298 K. Buffer solutions (pH = 3.05,
.04, 5.03, 5.51, 6.03, 6.53, 7.28, 8.028, 9.04, 10.05): Citric acid (20 mM) was titrated
4
with sodium citrate (20 mM) in order to prepare solutions with pH 3-5. For the mid-
range, MES buffer solutions (20 mM) of pH 5.5-7.3 were prepared by adjusting of 2-
(
N-morpholino)ethanesulfonic acid with NaOH (1M) to the desired acidity. For the
upper range (pH 8-10), a solution of boric acid (20mM) was alkalized with NaOH
7

(
1M) [35]. The pH of the aqueous buffer solutions was evaluated on a Horiba D-51 or
a DKK-TOA corporation HM-30R pH meter. The spectrofluorimetric data of the
varying pH sample solutions were fitted to the Hill sigmoid equation [36,37] (Origin
Pro – Category: Growth/Sigmoidal, Function: DoseResp, Iteration Algorithm:
Levenberg Marquardt) to evaluate the pKa according to Equation 2:
ꢃꢍꢎꢄꢏꢃꢍꢐꢑ
ꢉ = ꢊꢋꢌꢇ
+
(Eq. 2)
ꢒꢓꢒꢔꢕꢖꢗꢘꢙꢚꢎꢛꢙꢜ
ꢝ×ꢞ
where, y is the fluorescence response of the NIR cyanine dye in aqueous buffer
solution, and I max represent the corresponding minimum and maximum
I
min
fluorescence intensity, pKa of the Hill-type component and h is the Hill coefficient to
the slope obtained from the plot measured at the inflection point α=0.5.
2
.4. Computational Methods - Density Functional Theory (DFT) studies
All calculations were implemented using Gaussian 16 B.01 program suite [38]. The
Becke’s three-parameter hybrid exchange functional with Lee-Yang-Parr gradient
correlation (B3LYP level theory) and basis set 6-31G+(d p) were selected for the
current calculations [39–42]. The ground and excited state geometry optimization of
the cyanine dyes was carried out in both gas phase and water using the polarizable
conductor calculation model (PCM) [43,44]. No symmetry constrains, bond
length/angles or dihedral angles were applied in the geometry optimisation process.
GaussView v. 6.0 [45], Chemcraft (v. 1.8) - graphical software for visualization of
quantum chemistry computations (https://www.chemcraftprog.com) and UCSF
Chimera v. 1.13.1 were employed for the graphical representations [46].
2
.5. Cell culture
8

HeLa cells were cultured in DMEM (Thermo Fisher Scientific, Waltham, MA, USA)
supplemented with 10% (v/v; Thermo Fisher Scientific) FBS, 50 U/mL penicillin and 50
μg/mL streptomycin (Nacalai Tesque, Kyoto, Japan) in an incubator maintained at 37°C and
a humidified atmosphere of 5% CO . Cells were plated on glass bottom dishes (Iwaki, Tokyo,
2
Japan) for fluorescence imaging and on 96 well plates (Thermo Fisher Scientific) for the
MTT assay.
2
.6. Measurement of cell viability
Cell viability was measured by the MTT assay. The cells plated on 96 well plate (0.1 mL
5
of 1 x 10 cells/ mL for each well) were incubated in culture medium or Hank’s balanced salt
solution (HBSS; Thermo Fisher Scientific, pH was buffered by 10 mM HEPES and adjusted
to 7.4 with NaOH) containing various concentration of the cyanines 6a-6e for 1, 2, 4 or 24
hours in the incubator. After the treatment with the dye, the medium was replaced with
culture medium containing 0.5 mg/mL MTT (Nacalai Tesque), and cells were incubated for 2
hours in the incubator. Then, the culture medium was removed and 100 μL DMSO was added
in each well to dissolve the precipitation, and the absorbance at 570 nm was measured by a
microplate reader, Varioscan (Thermo Fisher Scientific). Cell viability was calculated as a
percent to the control culture value.
2
.7. Fluorescence Imaging
For staining, HeLa cells plated on glass bottom dishes (0.1 mL of 1×10 cells/ mL for each
5
well) were incubated in HBSS with 10 μM dye and 2% (v/v) DMSO for 30 min in the
incubator. Then, the cells were washed with HBSS to remove the dye remaining outside the
cells. To estimate fluorescence responses to various intracellular pH values, the cells stained
with the cyanine dye 6b were incubated with nigericin (5 μg/mL; Wako, Osaka, Japan) in
+
high-concentration K HBSS at various pH (15 mM NaCl; 125 mM KCl; 1.3 mM CaCl ; 0.5
2
mM MgCl ; 0.4 mM MgSO ; 0.3 mM Na HPO ; 0.4 mM KH PO ; 4.2 mM NaHCO ; 5.6
2
4
2
4
2
4
3
9

mM D-glucose; pH was buffered by 10 mM MES for pH4.0, 4.5, 5.0, 5.5, 6.0 and by 10 mM
HEPES for pH6.5, 7.0, 8.0) for 10 min, respectively. Fluorescence images were acquired on a
fluorescence microscope, BZ-X800 (Keyence, Tokyo, Japan) with a 100x objective. The dye
was excited by a metal halide lamp through a 710/75 excitation filter and a 760 nm dichroic
mirror, and the fluorescence was observed through an 810/90 nm emission filter. The
fluorescence intensity within the cell was determined as an averaged fluorescence intensity in
the whole cell body. Fluorescence intensity at each pH condition was calculated as averaged
fluorescence intensity of all the cells observed at that pH, respectively.
3
. Results and discussion
3
.1. Molecular design and synthesis of the pH responsive cyanine platforms
The overall synthetic scheme includes four reaction steps starting from commercially
available materials as outlined in Scheme 1. The N-quaternary heterocyclic starting material 2
was obtained in 71% via an alkylation of 2-methylbenzothiazole 1 with a 10% molar excess
of benzyl bromide. Several studies disclosing the good cell membrane permeability and
relatively low cytotoxicity of benzothiazole incorporating compounds account for the
selection of this particular building block in the current design of pH responsive fluorophores
[
47–53]. Benzyl groups were introduced to the heterocycles targeting enhanced photostability
of the cyanine backbone [54–57]. The dianilide linker 4 was prepared by a Vilsmeier-Haack
formylation of 4-tertbutyl cyclohexanone 3 with phosphoryl chloride and N,N-
dimethylformamide. The crude product was then treated in situ with aniline to stabilize the
dialdehyde intermediate. Subsequent condensation reaction between bimolar ratios of the N-
quaternary heterocycle 2 and 1 equivalent of the polymethine linker 4 in the presence of
anhydrous sodium acetate afforded the precursor cyanine 5. Purification was achieved by
recrystallization from ethanol. The preparation of the target heptamethines 6a-6e was readily
achieved in moderate to excellent yields by a nucleophilic substitution of the chlorine atom at
1
0

meso-position of the precursor dye 5. Analytical samples of the dyes 6a-6e were obtained
after purification by flash column chromatography on silica gel. Structural identification of
all compounds was achieved employing NMR and high-resolution ESI mass spectrometry
(HRMS) (Fig. S1-S26). The detailed synthetic protocol for the preparation of the dyes and the
corresponding intermediate products can be found in the Supplementary Information file.
Scheme 1. Synthetic approach to the target near-infrared pH responsive cyanine platforms.
The piperazine substructure known to possess versatile binding properties, is largely found
in pharmaceutical and biologically important substances [58]. Recently, numerous papers
report the use of piperazine derivatives in the development of sensing platforms [59–64]. The
high sensitivity and exceptionally rapid rates for proton transfer in acid-base equilibria, are
what classifies these amino moieties as excellent platforms for pH sensing. In this context
and taking into consideration the above benefits, we facilitated a cyanine scaffold coupled
with various N-substituted piperazines of known pK values aiming to create a small library
a
of fine-tuned pH-responsive dyes emitting in the range over 800 nm.
3
.2. Optical characteristics of the NIR dye series in organic media
1
1

Absorption characteristics of the synthesised cyanine dyes were studied and
determined through measuring their visible spectra. Comparing the spectra of the
cyanine dye
5 with the derivatives 6a-6e, the introduction of the piperazine fragment
at meso position resulted in broadening of the absorption band. In addition, a
hypsochromic shift (Δλ ~ 93-105 nm) was observed in comparison with the parental
chlorinated dye
5 (λabs = 803 nm).
Figure 2. Normalized absorption and fluorescence spectra for the dyes 5 (panel A), and 6b
panel B). Excitation wavelength = 803 nm for dye 5 and 700 nm for dye 6b, respectively.
(
1
2

The longest wavelength absorption maxima of the heptamethines 6a-6e were
localized between 698-710 nm. The positions of these bands were barely affected by
the types of the piperazine residue. The molar extinction coefficients - ε were
-3
-1
evaluated as 65,785-85,032 mol dm cm in acetonitrile. Despite their huge array of
functions, traditional cyanines, rhodamines and BODIPY dyes suffer from the
limitation of rather small Stokes shifts (~ 20-30 nm), which somewhat hinder their use
as fluorescent bio-probes [65–71]. The emitted light re-absorption by the dye scaffold
is a major drawback, since it is limiting imaging related applications. Small Stokes
shifts usually lead to self-quenching and detection errors arising from the excitation
light [72]. Overcoming these limitations, the current modified fluorophores exhibit
diminished spectral overlap and potentially less pronounced secondary inner-filter
effect. The fluorescence emission spectra of the cyanines 6a-6e exhibit a nearly
constant maxima – λem detected between 804-807 nm. As shown in Fig. 2, the
fluorescence peaks exhibit analogous consistent tendency to the UV-Vis, being blue
shifted by approximately 25-28 nm compared to dye
5. The increase in the Stokes shift
is accounted as an important advantage of the amino-substituted cyanines for their
applicability in fluorimetric analysis. Optical characteristics for the dyes 6a-6e were
evaluated and summarized in Table 1.
1
3

Table 1 Optical characteristics of the near-infrared cyanine dyes 6a-6e in acetonitrile.
kobs
a
b
λabs
λem
Δλ
Δν
ε
t1/2
d
-2
Dye
R
Φ
(×10 )
-
1
−1
−1
(
nm) (nm) (nm) (cm ) (M cm )
(min)
-
1
(
min )
c
IR-780
5
Cl
Cl
780 802
22
23
98
352
303
268,300
188,700
72,500
65,800
85,000
81,300
80,400
0.078
0.118
0.128
0.172
0.152
0.104
n.d.
n.d.
n.d.
71
803 823
706 804
n.d.
6
a
H
1726
7.11
8.81
9.10
7.11
8.38
6
b
CH3
700 804 104 1848
59
6
c
(C H )CH 698 805 107 1904
50
6
5
6
d
C H
6
706 807 101 1773
710 806 96 1678
76
5
6
e
p-CF C H
4
0.156
66
3
6
a
b
c
−1
−1
longest wavelength absorption maximum; fluorescence maximum - excitation at λabs
;
ε = 274,000 M cm in
d
methanol [73]; fluorescence quantum yield relative to the IR 780 - 7.6% [34]; n.d. = not determined.
Several research groups proposed an interplay between localized excitation and
intermolecular charge transfer (ICT) as an explanation to the enhanced Stokes shift of
amino-substituted type of cyanines [74,75]. However clear experimental evidence for
this observation was absent from the literature. According to a recent work, the origin
of the large Stokes shift seems to be related to a geometry phenomenon [76]. In the
above paper, the authors report a strong effect on the conjugation of the polymethine
chain by the nitrogen atom linked to the cyanine scaffold at meso position.
Consequently, a broken symmetry was observed and reported for the ground state. In
turn, the symmetry appears to be partially restored upon relaxation in the first excited
state, leading to a large structural re-arrangement of the molecule
.
Following this assumption, similar behavior was observed for the current piperazine
derived series based on the outcome of the performed computational investigation. The
dihedral angles and the bond length alternations provide information about the planarity of
the molecule, hence we adopted DFT and TD-DFT methods using the B3LYP functional and
1
4

6
-31G+(d p) basis set. Our results were found consistent with the above report. Analogous
remarkable change in the geometry between the ground and the excited state of cyanine 6b
was observed. The ground state of the cyanine 5 resembles a nearly planar structure, with δ1
o
o
o
and δ close to 180 (~177 and 174 respectively as outlined in Fig. S27). The symmetry of
2
the molecule can also be supported by the bond lengths of the polymethine chain shown in
the same graphical representation. On the opposite, the amino substituted dye 6b appears to
be distinctly non-planar in the ground state, with both dihedral angles appreciably different
o
o
(
~156 and 160 respectively as outlined in Fig. S28). Minor changes of the bond length can
also be observed compared to the parental chloro-substituted dye. However, the excited state
geometry of the dye molecule resembles almost a restored planar geometry with both δ and
1
o
δ of 177 as presented in Fig. S29.
2
3
.3. Investigating the probability of solvatochromism and intramolecular charge transfer
As previously mentioned, numerous reports attribute the blue shift of the absorption
maximum and the increment of the Stokes shift of amino substituted heptamethines to excited
state intramolecular charge transfer (ICT) [74,77–80]. A widely used criterion to identify an
ICT process is whether the dyes exhibit solvatochromic properties. It is known that varying
the polarity of the environment, classical D-π-A type of solvatochromic fluorophores
undergoing ICT, change their absorption and/or fluorescence emission wavelength and
intensities. The push−pull structure ensures that the charge is transferred from the donor
group to the acceptor upon light absorption [81]. Generally, solvatochromic dyes can exhibit
either positive or negative solvatochromism in solvents of variable polarity. The
solvatochromism is caused by differential solvation of the ground and Franck-Condon
excited states. If the ground state is stabilized more than the excited state by increasing
solvent polarity, negative solvatochromism is exhibited and vice versa. To get a clear insight
and to assess the probability of solvatochromism and internal charge transfer from the
1
5

piperazine (donor) towards the cyanine backbone (acceptor), we next studied the visible
absorption and the fluorescence emission spectra of the dyes 6b and 6d in a series of 10
organic solvents with a wide polarity range. The corresponding main spectrophotometric and
fluorimetric data of dye 6b are listed in Table 2.
Table 2 Optical characteristics of the NIR cyanine dye 6b in various organic solvents.
a
b
c
ET (30)
λabs
λem
fwhm
(nm)
53
Δλ
(nm)
91
Δν
Solvent
Type
-
1
-1
(
kcalmol )
(nm)
713
716
721
699
724
719
709
719
(nm)
804
811
809
804
823
818
808
814
(cm )
1587
1636
1509
1868
1661
1683
1728
1623
MeOH
EtOH
55.4
polar protic
51.9
54
95
i-PrOH
AcN
48.4
54
88
45.6
55
105
99
DMSO
DMF
45.1
54
polar aprotic
43.2
56
99
Acetone
CHCl3
42.2
54
99
39.1
57
95
moderate
EtOAc
polar aprotic
38.1
33.9
707
717
806
818
58
70
99
1737
1722
Toluene
non-polar
101
a
b
c
longest wavelength absorption maximum; excited at λ ; full width at half maximum.
abs
The cyanines were characterized by a single absorption band in the visible range (550-750
nm) with the longest wavelength maxima being centered around 699-721 nm. By analogy,
the fluorescence emission band maxima were allocated between 804-823 nm. As shown in
Fig. 3, negligible solvent polarity dependence on the spectral shifts could be observed with
respect to both the absorption and the fluorescence maxima. The shape and the full width at
half maximum (fwhm) of the fluorescence emission spectra also remained relatively
unaffected in all studied solutions. The N-phenyl substituted heptamethine 6d (Fig. S30,
Table S1) revealed similar behavior to compound 6b. The Stokes shifts - Δν were evaluated
-1
ranging from 1509 to1868 cm in the different organic media. Among all studied solvents,
1
6

the largest difference between the absorption and fluorescence emission maxima was
observed in acetonitrile (Δλ = 105 nm), and the smallest change was found for isopropanol (~
8
8 nm). However, no polarity-controlled trend in the Stokes shift could be addressed. It is
obvious, that the spectral behavior most likely reflects a rather limited change of the dipole
moments of the ground state and the excited state.
Figure 3. Absorption (panel A) and fluorescence emission spectra (panel B) of the NIR
heptamethine cyanine dye 6b in organic solvents of varying polarity. Stokes shift plotted as a
function of the solvent polarity parameters E (30) and Δf (panels C and D respectively).
T
LM
In order to investigate the influence of the solvent on the optical properties of the dyes, the
calculated Δν values were plotted as a function of both the Dimroth-Reichardt E (30)
T
parameter and the ΔfLM function of orientation polarizability (Lippert–Mataga function). The
1
7

data points correlating to the different solvents are indicated in Fig. 3C and Fig. 3D,
respectively. Solvent induced spectral shifts are often interpreted in terms of the Lippert–
Mataga equation, which describes the Stokes shifts in terms of changes in the dipole moment
of the fluorophore and the dependence of the energy of the dipole on the dielectric constant
and the refractive index of the solvent as expressed in Equation 3 [82]:
ꢣꢤꢏꢒ
ꢇꢈꢏꢒ
ꢟꢠꢡꢢ
=
−
(Eq. 3)
ꢈ
ꢥꢣ ꢏꢒ
ꢤ
ꢥꢇ ꢏꢒ
where, ε and n represent the dielectric constant and the refractive index of the solvent.
r
No linear correlation was observed between the Δ
ν and the corresponding solvent
parameters E (30) and Δf . Scrutinising the two plots led us to assume, that
T
LM
intramolecular charge transfer is not a dominant factor in the case of the investigated
cyanines. Apparently, the meso-functionalized cyanines do not possess significant
properties of typical D-π-A type solvatochromic chromophores, which further supports
the assignment of the enhanced Stokes shift to a geometry related aspect [72,83].
3
.4. Photostability of the probes
Next of interest to us was to assess the resistance against photooxidative
degradation. The photostability studies of the compounds 6a-6e were undertaken in
acetonitrile solutions. All samples were excited at the corresponding wavelength of
maximum absorbance - λabs, while the fluorescence intensity was monitored at the
fluorescence emission maximum - λem
.
1
8

Figure 4. Photobleaching decays of the NIR dyes representing % of initial emission signal
over irradiation time (panel A), pseudo-first order linear variation of ln I /I as a function of
0
t
irradiation time for the photocatalytic degradation, where I represents the initial emission
0
intensity and I is the emission intensity at a given time of measurement (panel B). The
t
fluorescence intensity was monitored over time at the λ of each individual compound.
em
After 3 hours of continuous irradiation, approximately 16-26 % of the initial
fluorescence intensity could still be detected. The kinetic curves revealed ~ 50 % of
the original intensity after a period of 50-71 minutes of constant irradiation (Fig. 4A).
The studied dyes 6a-6e revealed comparable photostability behaviour upon irradiation.
Highest resistance to photobleaching was observed for the 6d, and lowest for the 6c
cyanine (Fig. 4B). We also performed a photostability test of a previously reported
indole derived cyanine dye [30]. Surprisingly, the benzothiazolic backbone was found
more labile to photocatalytic degradation. Over the same time of irradiation, minor
changes could be observed with respect to emission response of the indole derived
fluorophore (data not shown). In addition, the effect of the counter anion on the
photostability of the heptamethines 6a
exchange from bromide to iodide on dye 6b yielded an enhancement of approximately
5 % (t = 59 min for the bromide to t = 74 min in the case of iodide counter
-6e was investigated. Intriguingly, an anion
2
1/2
1/2
1
9

anion). The different polarizing effect of the counterion exhibiting superior
-
photostabilization in favour of I is presumed as a plausible explanation to the above
finding On the basis of detailed literature studies and our experimental observation,
.
we anticipated iodides to express a quenching effect on singlet oxygen - responsible
for the photocatalytic degradation of organic materials [84–86].
3
.5. An overview of the protonation site
For practical application, we further studied the halochromic behavior of the
cyanine dyes varying the pH conditions. A drastic color change was observed after the
addition of hydrochloric acid. The deep-blue colored turned green on acidification of
the neutral dye solution (Scheme 2). Analogous colorimetric response for a piperazine-
substituted cyanine was previously reported [30]. Nevertheless, the exact allocation of
the proton is ambiguous, since extensive discussion on the protonation sequence is
alleged.
Scheme 2. Colorimetric response upon acidification of the solution - pH-switching concept.
The NMR chemical shift on certain atoms including hydrogen, carbon and fluorine
can strongly be influenced by the total amount of charge distributed on them. Thus,
nuclear magnetic resonance spectroscopy can be employed as a useful analytical tool
in the qualitative description of various processes such as protonation and hydrogen
bonding [87]. In order to shed more light to the protonation and obtain fundamental
1
information on the switching process, we performed H-NMR studies on the neutral
2
0

and the protonated form of dye 6b in CDCl (Fig. S31). Under neutral conditions, the
3
piperazine protons closer to the cyanine dye backbone resemble a multiplet at δ 3.68
ppm. In turn, the broad singlet peak observed at δ 2.67 ppm accounts for the
methylene groups next to the terminal nitrogen atom of the cyclic amino moiety
[(CH2) -N-CH )]. A distinct colorimetric response was observed in the deuterated
2 3
solvent, which was provoked by the addition of 1 equivalent of trifluoroacetic acid -
+
TFA (Fig. S32). The presence of H also resulted in minor shifts for the methine
groups with trans coupling constants
.31 ppm of the neutral form were shifted to δ 5.85 ppm and 7.55 ppm, respectively, in
the case of the protonated form (data not shown). The same deshielding tendency from
.5 ppm to 3.36 ppm was observed for the singlet (Ha) corresponding to the methyl
J ~ 13 Hz. The CH signals at δ 5.87 ppm and
7
2
group. At first sight it seems, that the protonation process mainly involves the terminal
nitrogen atom. However, the parallel reduction of shielding of all methylene protons
(
Hb and Hc) on acidification, led us to consider the probability of proton allocation
between the two nitrogen atoms. Such a competitive protonation behaviour for 1,4-
disubstituted piperazines is attributed to a proton transfer between the two nitrogen
atoms, and it is repeatedly reported in the literature [64,88,89]. According to the same
work, a chair to boat conformation interconversion and the inversion of the nitrogen
centre can be slowed down due to the stabilization of the boat conformer by hydrogen
bond interaction. As a matter of fact, intramolecular proton relocation requires a
relatively large activation energy. Nevertheless, the presence of protic solvents can
successfully catalyse this process and remarkably affect the activation free energy of
the system by compensating the energy barrier between conformers [90].
Prominently, the piperazine ring preferentially adopts chair conformation due to
-1
lower energy compared to that of the half-boat (ΔE = 6.33 kcal mol ), boat (ΔE = 5.34
2
1

-1
-1
kcal mol ) and twist-boat (ΔE = 1.05 kcal mol ) [91]. However, several reports
suggest that bidentate systems with two nitrogen atoms readily undergo a series of
conformational changes in order to provide the equatorial position to substituents.
These changes result in reduction of the total energy of the molecule in the case of
twist-boat conformation [92]. According to another investigation, an N-H···N distance
of ~2.4-2.5Å is responsible for a considerable reduction in energy and adoption of a
twist-boat conformation rather than chair for the 1,4-disubstituted piperazines ring
[
93]. Our DFT computational study appears to be in agreement, since the calculated
distance of the proton and the second nitrogen atom of dyes 6b is approximately 2.6Å
Fig. S33).
(
3
.6. Spectroscopy studies on the pH response behaviour and evaluation of pKa
In the current work, we intended to build a small library of fine-tuned pH responsive dyes
by coupling several N-substituted piperazines with an NIR cyanine dye scaffold. Envisioning
the potent of the title cyanines to be applied as pH indicators in analytical assays, we
proceeded to examine their halochromic properties. The spectra were recorded in the range
5
50-900 nm immediately, in order to eliminate the effect of time. Halochromic evaluation of
the synthesized heptamethine dyes was carried out by studying their visible absorption and
fluorescence emission spectra in aqueous buffer solutions having varied pH values. A
positive halochromism was observed at decreasing the pH, which was accompanied by the
gradual increase of the newly emerged absorption band (Table 3).
Table 3 Optical characteristics of the NIR cyanine dyes 6a-6e under neutral and acidic
conditions evaluated in aqueous buffer solutions (Cdye = 2 μM).
a
Dye
terminal R
Absorption - λabs (nm) Emission - λem (nm) pK ± error
a
2
2

(piperazine)
Neutral
pH<7
Neutral
pH<7
6
a
H
676
677
649
653
558
735
749
663
661
658
802
803
803
803
815
816
804
804
804
7.06 ± 0.04
6.35 ± 0.06
6.08 ± 0.19
5.79 ± 0.03
6b
CH3
6c
CH(C H )
6
5 2
6d
C H
6
5
*
6e
C H CF
3
803
5.13 ± 0.03
6
4
a
*
determined from fitting of the spectrofluorimetric data; the pK was evaluated based
a
on the UV-Vis results by fitting the data from the UV-Vis experiments.
As the pH of the environment was diminished, the absorption peak was progressively red-
shifted. The largest bathochromic displacement was observed for the cyanines 6a and 6b (Δλ
~
59 nm and 72 nm respectively). This remarkable change allowed the use of the two NIR
absorbing chromophores as ratiometric pH responsive probes. Self-calibrating ratiometric
probes exhibit improved signal stability and reproducibility, which is a distinctive advantage
over conventional single wavelength. In the case of dye 6b, the new absorption maximum at
7
49 nm gradually increased, while that at 677 nm decreased on acidification of the buffer
solution (Fig. 5A). The exact same observation was reported earlier by Miyochin et. al.,
where the longest wavelength absorption maxima of the neutral and protonated form were
allocated at 667 nm and 750 nm, respectively [30]. Тhe protonation constants of the cyanines
can be evaluated from the pH dependence of dye spectra providing an equilibrium between a
single dye species and acidified product. On plotting the absorbance intensity ratios at fixed
wavelength (protonated/neutral dye) of the dyes 6a and 6b against pH, the Hill-type
sigmoidal curves were obtained and delineated in Fig. 5B and Fig. 5D respectively.
2
3

Figure 5. UV-Vis spectra of the heptamethine cyanine dye 6a and 6b measured at Cdye = 2
μM (panels A and C), and Hill-type sigmoidal plot of the relative absorption intensities at the
corresponding longest wavelength maxima of the neutral/protonated form against the pH of
the solution (panels B and D).
On the other hand, compounds 6c and 6d bearing a benzhydryl and a phenyl substituted
piperazines revealed only a minor bathochromic shift of 14 nm and 8 nm respectively (Fig.
S34A and S34D). Due to the considerable overlap between the absorption bands of the
neutral and the protonated form of those two heptamethines, attempts to plot a ratiometric
response of the UV-Vis data was hampered. Intriguingly, cyanine 6e featuring a CF group
3
revealed an even stronger, but opposite blue shift of approximately 100 nm (558 nm – neutral
buffer solution / 658 nm – acidified buffer solution – Fig. S35). The distinctive narrow
absorption maximum centered at 558 nm, could presumably be ascribed to either instability
2
4

of the chromophore over the upper pH range, or strong aggregation due to π-π stacking being
disrupted on acidification [94,95].
Next, we performed fluorimetric investigation of the pH response in order to evaluate the
pK values of the two dyes (6c and 6d – Fig. S34 / panels B, C, E and F). The exact same
a
solutions were used in both the absorption and fluorescence emission measurements. To
investigate how the emission properties of the rest of the series changed upon varying the
environment, we analyzed the fluorescence emission spectra at alternating pH. Samples 6a
and 6b were excited at the two distinctive wavelengths - λ , while a single wavelength of
abs
6
60 nm was employed for the dyes 6c and 6d. As shown in Fig. S34 and Fig. S35, the
spectroscopic data revealed an increase of the fluorescence intensity with lowering the pH of
the environment (Fig. S36). Dyes 6a and 6b also exhibited ratiometric fluorescence response,
reciprocal to the UV-Vis studies. Based on the Hill equation, the sigmoidal plots of the
heptamethines 6a, 6b, 6c and 6d afforded pK values of 7.06, 6.35, 6.08 and 5.79,
a
respectively. Prominently, no fluorescence could be detected upon excitation at the maximum
wavelengths of the two bands of dye 6e. Therefore, the protonation constant (5.13) was
determined after processing the UV-Vis data. Notably, a plot of the experimentally evaluated
pK values of the cyanine dyes against reported pK values for substituted piperazines,
a
a
2
revealed a good linear correlation with a regression coefficient R = 0.96 as presented in Fig.
6
.
2
5

Figure 6. Plot of experimentally evaluated protonation constants - pK of the NIR cyanine
a
dyes 6a-6e against the pK of 1,4-disubstituted piperazine derivatives reported in the
a
literature.
3
.7.
Cytotoxicity investigation by MTT assay
The HeLa cells were initially treated with DMSO only with variable quantity (0.1-10%).
The results were monitored over a period of 1 to 24 hours (Fig. S37). The preliminary
experiment revealed that co-solvent concentration over 5% induced approximately 20% cell
death. However, concentration of up to 1% did not induce a toxic effect even after a longer
incubation time (24 hours). On the other hand, dye concentrations up to 10 μM showed
almost no cytotoxic effect after 2 hours, while a minor decrease of the cell viability was
observed at increasing the exposure time to 4 hours (Fig. 7A). Cytotoxicity of all NIR dyes
6
a-6e including the parent meso-chloro substituted cyanine 5 was also studied over 4 hours
incubation. As shown in Fig. 7B, the comparative study of the fluorophores 6a-6e bearing a
piperazine moiety exhibit significantly lower cytotoxicity as opposed to the parental
fluorophore 5. We assume, that this could be due to the presence of halogen in the given
molecule at this position of the backbone. Apparently, this is the only structural difference
between dye 5 and the title compounds 6a-6e. The NIR heptamethine derivatives 6a, 6b, 6c
2
6

and 6e revealed comparable behavior, while the phenyl-substituted 6d showed amplification
of the cell death at higher concentrations.
Figure 7. Cell viability measured by the MTT assay. The cells were incubated in HBSS
containing 10 μM of the cyanine dye 6b and measured after 1, 2 and 4 hours (panel A), and
various concentrations of the cyanine dyes 5 and 6a-6e for 4 hours in the incubator (panel B).
Error bars represent mean ± SEM with n = 6 replicates. Statistical differences were estimated
by analysis of variance (ANOVA) and Student's T-test. P < 0.01 was considered as
significantly different compared to the value in the cells not treated with the dye and
indicated with * above the bars.
3
.8. Fluorescence Imaging application
Having established their capability for fine-tuned pH response, the importance of the
chemosensing platforms was also evaluated by studying their potential inside a more complex
environment, such as living cells. In the view of practical application of the title NIR cyanine
dyes as an efficient proton sensing platform, we performed studies on cellular imaging with
HeLa cells. For the purpose of staining, 10 μM concentration of the cyanine dye was applied
to the cell culture. Cyanine dye 6b was selected for the imaging experiments, due to the
relatively acceptable cytotoxicity (100 % survival over 1 and 2 hours, and approximately
7
9% survival over 4 hours respectively for 10 μM) and the evaluated pK (6.35 ± 0.06 -
a
spectrofluorimetric data), which was found to be more suitable for monitoring the acidic
2
7