Pentamethinium fluorescent probes: The impact of molecular structure on photophysical properties and subcellular localization

Article abstract of DOI:10.1016/j.dyepig.2013.12.021

The performance of fluorescent probes arises from their structural basis. In this study, we report design and synthesis of a series of novel symmetrical γ-substituted pentamethinium fluorescent probes. Relationship between structure, photophysical propert

Full text of DOI:10.1016/j.dyepig.2013.12.021

Dyes and Pigments 107 (2014) 51e59
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Pentamethinium fluorescent probes: The impact of molecular
structure on photophysical properties and subcellular localization
a
,
b
,
c
d
a
a,
*
ꢀ
T. Bríza
, S. Rimpelová , J. Králová , K. Záruba , Z. Kejík b, T. Ruml c, P. Martásek b,
,e,
V. Král a
**
a Department of Analytical Chemistry, Institute of Chemical Technology, Prague, Technická 5, Prague 6 166 28, Czech Republic
b
ꢀ
Department of Pediatrics and Adolescent Medicine, First Faculty of Medicine, Charles University in Prague, Katerinská 32, Prague 2 121 08, Czech Republic
c Department of Biochemistry and Microbiology, Institute of Chemical Technology, Prague, Technická 5, Prague 6 166 28, Czech Republic
d
ꢀ
Institute of Molecular Genetics, Academy of Science of the Czech Republic, Prague, Vídenská 1083, Prague 4 142 20, Czech Republic
e Zentiva Development (Part of Sanofi-Aventis Group), U Kabelovny 130, Prague 10 102 37, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
The performance of fluorescent probes arises from their structural basis. In this study, we report design
and synthesis of a series of novel symmetrical g-substituted pentamethinium fluorescent probes. Rela-
Received 23 September 2013
Received in revised form
16 December 2013
Accepted 23 December 2013
Available online 1 January 2014
tionship between structure, photophysical properties and intracellular localization was explored and
structural variations resulting in high photostabillity and selectivity for subcellular localization were
obtained. Substitution of benzothiazolium moieties on both sides of the pentamethinium chain for
indolium units led to red shift in the absorption and fluorescence emission maxima of the probe and
increased photostability. The major advantage of the probes with side indolium units is their mito-
chondrial specificity, high photostability and synthetic accessibility.
Keywords:
Pentamethinium salts
Fluorescent probes
Mitochondria
Ó 2014 Elsevier Ltd. All rights reserved.
Cardiolipin
Photostability
Organelle imaging
1. Introduction
fluorescence quantum yields and broad application possibility.
Much effort has been made to improve the photostability of
Fluorescent probes are standardly used in many biological and
sensor applications [1,2]. Their popularity has continuously risen
due to their sensitivity, specificity and versatility. First of all, choice
of a probe depends on its selectivity. However, there are another
significant factors (e.g. photostability, phototoxicity, dark-toxicity,
brightness, solubility, pH sensitivity) affecting applicability of the
selected probe for microscopic studies under given conditions. One
of the main drawbacks of many probes is their low photostability
strongly related to photobleaching. The more photostable the
probe, the longer one can monitor its fluorescence emission inside
cells or tissues by means of fluorescence microscopy.
cyanine dyes by structural modifications or by admixtures [3].
Fluorescent probes based on pentamethinium chain with side
indolium structural motifs are widely used dyes (Cy5). In the case of
indolium methinium salts, many variants has been prepared to
date, for example trimethinium salts [4], penta- and heptamethi-
nium salts [5] including water soluble indolium pentamethines [6].
To the best of our knowledge, all the reported pentamethinium
indolium salts lack the
structureeactivity relationship (e.g. photophysical properties,
intracellular localization, cytotoxicity) have been performed on
substituted indolium pentamethines, so far. Therefore, our first
approach was to prepare probes based on symmetrical
g-aryl substitution. No detailed studies on
g-
Recently, polymethinium dyes have drawn extensive interest
due to their excellently high molar extinction coefficients, high
g-
substituted pentamethines with side benzothiazolium units [7]. In
that study, photostability of benzothiazolium salts was compared
to commercially widely-used fluorescent probe Cy5. The intention
* Corresponding author. Department of Analytical Chemistry, Institute of
was to reveal the effect of
stability of the dyes. The general formula of the developed pen-
tamethines and Cy5 dye is depicted in Fig. 1.
For proper evaluation of the structureefunction relationship in a
group comprised of our methinium salts and commercial Cy5, it is
g-aryl substitution on overall photo-
Chemical Technology, Prague, Technická 5, Prague
6 166 28, Czech Republic.
Tel.: þ420 731756799.
** Corresponding author. Department of Analytical Chemistry, Institute of
Chemical Technology, Prague, Technická 5, Prague 6 166 28, Czech Republic.
ꢀ
E-mail address: brizat@vscht.cz (T. Bríza).
http://dx.doi.org/10.1016/j.dyepig.2013.12.021
0143-7208/Ó 2014 Elsevier Ltd. All rights reserved.

ꢀ
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
52
Fig. 1. General scheme of g-substituted pentamethinium systems (A) and Cy5 dye (B).
important that Cy5 has been utilized as a fluorescent probe for
mitochondrial staining when conjugated to oligonucleotide with a
certain, accidentally discovered, targeting sequence [8]. Because of
the negatively charged nature of the oligonucleotide, such probe
has to be delivered inside cells via transfection, which is not a
convenient method, especially because of usually low transfection
efficiency. Lorenz et al. tested the staining pattern of unconjugated
Cy5 with active NHS ester (Cy5-NHS) transfected in the same way
as the Cy5-oligonucleotide conjugate. Such approach did not result
in mitochondrial staining. These results may be somewhat
misleading as they may be affected by the method used. Because of
the hydrophobic nature of Cy5, the dye appeared to be associated
with residues of the transfection reagent and was dispersed inside
the cells. Summing up, according to this paper, Cy5 alone, does not
stain mitochondria. However, the authors have not tried to stain
cells solely with Cy5-NHS or with Cy5 omitting transfection.
We intended to compare Cy5 as a representative of indolium
them with Cy5 and to investigate their potential employment as
mitochondrial probes.
2. Results and discussion
2.1. Photophysical properties
In our previous work [7], we described photostability of mito-
chondrial probes based on
g-substituted pentamethinium salts
with side benzothiazolium units, which are related to the structure
of salt 7. We observed that some derivatives showed comparable
photostability to widely used Cy5 dye, despite the fact, that Cy5 is
not derivatized in
g-position. The comparable photostability of Cy5
to some of the -aryl substituted pentamethines with side benzo-
g
thiazolium units [7] might be positively influenced by the presence
of indolium heteroaromates in side units of the pentamethinium
chain. In some cases, an increased photostability of indolium salts
over benzothiazolium pentamethinium salts was described [9].
mitotrackers without
g-substitution with g-substituted benzo-
thiazolium and indolium pentamethinium salts and to find rela-
tionship among structure, photophysical properties and
intracellular localization of the dyes.
However, these analyzed derivatives lacked
of the used medium was not considered.
Therefore, one of our goals was to reveal the influence of ben-
zothiazolium or indolium subunits and -substituents on the
g-substitution and type
Our first goal was to prepare a series of pentamethine systems
g
with side indolium units and g-substitution. Our next goal was to
find a relationship between the nature of the side units (indolium
overall photostability of pentamethinium probes in various media
and pH.
vs. benzothiazolium) and the photostability of the probes. The
presence of g-substitution and used medium has been also taken
into consideration. Another scope was to explore the intracellular
localization of the indolium pentamethine probes, to compare
We prepared four types of compounds, symmetrical, unsym-
metrical pentamethinium salts all with
without substitution in -position. More specifically, the aim of our
study was to compare symmetrical derivatives with indolium units
g-substituents and the salt
g
Fig. 2. Comparison of pentamethinium probes structures differing either by side unit or by substitution in g-position.

ꢀ
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
53
on both ends (compounds 1, 2, 4 and 5), symmetrical derivative 7
with benzothiazolium units on both sides of the chain Ref. [7],
unsymmetrical derivative 3 containing one indolium and one
benzothiazolium unit on each side and symmetrical derivative 6
side indolium units. Compounds 2, 3 and 7 have two, one and none
indolium units, respectively. As mentioned above, the photo
stability of the compounds in DMSO did not differ significantly. The
decrease of photostability of compounds 2, 3 and 7 in PBS corre-
lated with the decreasing number of indolium units, under dark
conditions. When the solutions of pentamethine salts were irra-
diated, the differences in photostability disappeared and the pho-
tostability of all three compounds seemed to be of the same order,
Table 1.
without g-substitution, Fig. 2.
In contrast to experiments carried out in DMSO, where almost
all the tested compounds were fully stable, in aqueous medium
photostability of the compounds varied dependently on structure
of the probe and pH of the medium. The least stable compound 5
had still 80% and 77% of the original absorbance after 30 min of light
exposure and in dark, respectively. Experiments carried out in PBS
(amount of DMSO less than 5%; v/v) provided significant differ-
ences in photostability of the individual compounds tested. These
differences were pH-dependent. During decomposition, absor-
bance at all wavelengths decreased according to LamberteBeer’s
law with the exception of compound 2. The absorption spectrum
profile of compound 2 was changed in PBS at pH 4.8 and 7.0, (see
Supplementary material). Initially, there were three absorption
maxima at ca 602, 639 and 685 nm. The maxima gradually merged
during 30 min light exposure. The intensity of 602 and 685 nm
maxima decreased and appeared like shoulders of the absorption
maximum at 639 nm. After 10 min of light exposure, the shoulder at
685 nm (never observed in DMSO or PBS at pH 9.3) disappeared
and the absorbance at 639 nm was slightly increased (data not
shown). Then, no further changes of wavelength maxima were
observed and absorbance at all wavelengths followed the
LamberteBeer’s law.
The next issue, which we wanted to address, was the influence
of
indicated in the introduction, the photostability of the widely used
Cy5 (probe without -substitution) was comparable to our
substituted probes based on benzothiazolium pentamethinium
salts [7]. Therefore we compared symmetrical -substituted pen-
tamethinium indolium probes with a probe of similar structure, but
without -substitution. Structures of the former (4-pyridyl- (1), 4-
g-substitution on the photostability of a dye. As it has been
g
g-
g
g
nitro- (2), 2-chinolinyl- (4) and N-methyl-4-pyridinium- (5)) and
the later probes (6, Cy5) are depicted in Fig. 1.
Generally, slightly positive influence of
increase of photostability was evident in all tested pH values for
probes 1, 4 and 5 in comparison to probes 6 and Cy5 without
g-substitution on the
g-
substitution, which showed comparatively the same photostability,
Table 1.
In acidified PBS (pH 4.8), salts 1, 4 and 5 with aromatic substi-
tution in
g-position exhibited photostability in range of 90e95%,
whereas the salt 6 and Cy5 lacking the
g-substituent showed 85%
PBS solutions of compounds 3 and 7 showed the lowest stability
of all compounds under the dark conditions. Percentage of the
original absorbance value dropped down to 92% for PBS solutions of
compound 3 at both pH 4.8 and 7.0 and to 91% at pH 9.3. Analogical
measurement was performed for compound 7, the values were 83%,
93% and 85%, respectively. Obviously, this instability was man-
ifested in larger extent upon exposition of the solutions to light
illumination. This phenomenon was especially apparent in PBS at
pH 4.8 where only 35% of the original absorbance of compound 3
remained after 30 min of light exposition. On the contrary side of
the “stability spectrum”, there was compound 1, which was very
stable under all the studied conditions. This compound was as
photostable as Cy5 in DMSO and even more stable in PBS. Also
compound 5 exhibited improved photostability in PBS when
compared to Cy5, but not so in DMSO.
and 88% photostability, respectively. We observed surprising
behavior of salt 2, which showed only 60% photostability. This could
be explained by worse accuracy of the method due to the already
mentioned change of spectral pattern. Observed photolability
cannot be probably ascribed to strong electron withdrawing effect
of the para-nitrophenyl substituent in g-position of 2 because the
salt 5 having also strong electron withdrawing pyridinium sub-
stituent exhibited the highest photostability from all the com-
pounds, in acidic medium, Table 1.
In conclusion, the effect of g-substitution was also affected by
the type of medium as well as side units, resulting in the changes of
the photostability of a dye.
2.2. Determination of the quantum yields (F)
New compounds 2 and 3, together with previously published
compound 7 [7], allowed us to correlate the photostability-
structure relationship. All these three compounds have the same
Further point to the fully characterized and evaluated our new
potential probes 1e6 was determination of their quantum yields
). The calculated values are summarized in Table 2.
(F
g-substitution, para-nitrophenyl, but they differ in the number of
The fluorescence quantum yields of salts 1e6 are summarized in
Table 2. Our results revealed that the
tamethine salts affected the value, when comparing salts 1e5 and
salt 6 possessing -substitution. Similarly, it was reported that the
-aryl substitution can strongly reduce the of Cy5 derivatives
g-aryl substitution of pen-
Table 1
F
Photostabilities of the probes in various pH of PBS and in DMSO. The percentages of
the original absorbance of the probes after 30 min of light exposure (“light”) or
maintenance in a black box (“dark”).
g
g
F
(one third of origin of Cy5) [10]. The authors found, that this
reduction of fluorescence intensity was caused by rotation of the
substituted groups. The rotation or vibration of bulky substituents
can change the geometry of the molecule in the excited state. These
phenomena permit the molecule in the excited state to relax by a
nonradiative process, thus reducing its quantum yield. Significant
Amax(30 min)/Amax(0 min)
PBS (pH 4.8)
PBS (pH 7.0)
PBS (pH 9.3)
DMSO
Light
Light
Dark
Light
Dark
Light
Dark
Dark
MS
1
2
3
4
5
6
7
Cy5
91%
60%a
35%
90%
95%
85%
75%b
88%
95%
107%a
92%
94%
89%a
76%
89%
93%
88%
83%b
86%
96%
113%a
92%
93%
75%
76%
91%
96%
87%
85%b
90%
100%
101%
91%
94%
99%
100%
85%b
99%
97%
94%
97%
93%
80%
95%
97%
98%
103%
98%
102%
96%
77%
96%
differences among the
F of the salts 1e5 and 6 were probably
caused by rotation of their aromatic substituents at the meso-po-
sition of the pentamethine chain. In the case of salt 4, an influence
97%
93%
101%
96%
100%
94%
of bulkiness of the substituent was apparent in DMSO, because its
F
value almost reached the one of salt 6. The lowest fluorescence
emission intensity of 2 and 3 in DMSO was possibly caused by its
fluorescence quenching due to the presence of substituted nitro
group. On the other hand, in MeOH, it seemed that the bulkiness of
83%b
98%
93%b
94%
100%
99%
a
Accuracy of the presented values was strongly affected by a change of spectral
profile of 2 in time (details in Supplementary information).
b
Values estimated after delayed start (details in Supplementary information).
g-substituents did not play an important role in F, either. Since F of

ꢀ
54
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
Table 2
Table 3
Photophysical properties of methinium salts 1e7 and Cy5 measured in DMSO, MeOH
Logarithmic values of binding constants and stoichiometry of cardiolipin (CL) and
and PB at pH 7.0.
phosphatidylserine (PS)/pentamethine salts 1e7 (MS) and Cy5.
qe
MS
CL/MSa
Log(Ksb)
PS/MSa Log(Ksb)
Phosphatidylserine
a
c
d
MS
Solvent
lmax
3
b/1 000 000
lemmax
lemmax
max
[nm]
[nm]
[nm]
Cardiolipin
1
DMSO
MeOH
PB
DMSO
MeOH
PB
DMSO
MeOH
PB
DMSO
MeOH
PB
DMSO
MeOH
PB
DMSO
MeOH
PB
DMSO
MeOH
PB
DMSO
MeOH
PB
643
634
633
643
635
635
643
634
631
641
629
633
637
627
626
645
637
633
654
645
642
644
636
637
0.134
0.227
0.134
0.175
0.188
0.102
0.189
0.241
0.116
0.195
0.234
0.198
0.146
0.227
0.172
0.176
0.190
0.161
0.092
0.220
0.088
0.144
0.152
0.137
655
648
658
664
656
663
666
659
662
664
652
653
657
658
658
660
659
652
673
663
669
667
658
661
641
629
643
647
640
647
648
636
636
650
642
649
648
629
643
650
647
639
656
643
650
649
645
649
0.15
0.24
0.07
0.07
0.17
0.05
0.03
0.03
0.06
0.37
0.18
0.01
0.12
0.14
0.02
0.55
0.27
0.05
0.02
0.05
0.02
0.56
0.37
0.55
1
1:1
1:2
1:2
7.2 ꢀ 0.6
12.6 ꢀ 0.6
12.0 ꢀ 0.5
1:1
2:1
1:1
1:2
1:1
2:1
1:1
2:1
2:1
4.6 ꢀ 0.1
9.6 ꢀ 0.5
5.5 ꢀ 0.1
10.9 ꢀ 0.2
5.4 ꢀ 0.1
10.2 ꢀ 0.2
4.8 ꢀ 0.1
9.6 ꢀ 0.1
10.5 ꢀ 0.1
2
2
3
1:1
1:2
1:1
1:2
1:1
1:2
1:1
1:2
1:1
6.0 ꢀ 0.1
12.0 ꢀ 0.2
5.7 ꢀ 0.1
11.5 ꢀ 0.1
7.1 ꢀ 0.5
12.5 ꢀ 0.6
6.7 ꢀ 0.2
12.6 ꢀ 0.4
5.3 ꢀ 0.1
3
4
5
4
6
1:1
2:1
1:1
2:1
1:1
1:2
4.9 ꢀ 0.1
9.8 ꢀ 0.2
4.1 ꢀ 0.3
8.5 ꢀ 0.4
5.2 ꢀ 0.2
10.4 ꢀ 0.6
5
7c
Cy5c
1:1
5.3 ꢀ 0.2
6
a
Complex stoichiometry (analyte:methine salt).
Conditional constant (Ks) determined in PBS, pH 7.0.
The values were taken from Rimpelova et al. [7].
7f
Cy5
b
c
partners via charge complementarity (phosphate group and
carboxylate group in the case of PS) or through nonspecific hydro-
phobic interactions and hydrogen bonds [12e15].
We have observed high affinity of 1e6 to CL, summarized in
Table 3 and shown in Fig. 3 for salt 1. Salts 1e6 displayed higher
selectivity for CL than the Cy5 probe, Table 3. Salts 1e6 bound to PS
with lower affinity than CL, suggesting that they might be prom-
ising mitochondrial probes.
a
lmax e wavelength maximum of absorption spectra.
b
c
d
e
f
lmax e absorption coefficient of the maximum of absorption spectra.
lemmax e wavelength maximum of fluorescence emission spectra.
lexmax ewavelength maximum of fluorescence excitation spectra.
e quantum yield of fluorescence.
The values were taken from Rimpelova et al. [7].
q
Further step was to prove the applicability of the salts 1e6 in live
cells of various cells lines as mitochondrial probes, which was
strongly indicated by the in vitro interaction studies with CL.
Therefore, an uptake of salts 1e6 by live cells was performed to
determine their subcellular localization.
salt 1 was almost of the same value. In PB, salt 1 performed the best
results. The value of of 1 was even higher than for salt 6 without
substitution in meso-position, see Table 2.
F
We tested the ability of our red-emitting pentamethine dyes
to cross the plasma membrane of U-2 OS, HeLa, PC-3 and MCF-7
cell lines and to selectively localize inside the cells. We followed
rapid entry of the dyes into the cells (ca 5 min, strong and stable
fluorescent signal observed after 10 min) of 1, 2e4 and 6
regardless on the cell line used in 10e35 nM concentrations. We
have observed highly specific localization of the dyes in a
network-like shape organelles, resembling mitochondria, Fig. 4.
To further determine the organelle nature, we conducted
counter-staining of the tested pentamethines with various
organelle trackers. We have found major co-localization pattern
with the green-emitting mitochondrial marker, MTG, as shown in
Fig. 4. Thus, we have confirmed mitochondrial localization of the
2.3. Evaluation of the salts as mitochondrial probes
The next step in characterization of our salts 1e6 was determi-
nation of the binding constants with cardiolipin (CL) and phospha-
tidylserine (PS). CL is present almost solely in inner mitochondrial
membrane. Very low amount of CL was detected in membranes of
endoplasmic reticulum, but this residual portion might be caused
only by imperfect cell fractionation or by mitochondria-associated
endoplasmic reticulum membrane. PS occurs in all eukaryotic
membranes; nevertheless, it is mainly present in the inner side of cell
plasma membrane [11]. We have expected higher affinity of the
probes to CL, as we have observed for resembling pentamethine-
based mitochondrial probes [7]. Numerous studies have demon-
strated the interaction of methinium salts with membrane binding
tested dyes. Surprisingly, doubly positively charged g-substituted
Fig. 3. A) Absorption spectra of salt 1 (4.7 m mmol) in the presence of cardiolipin
M) with cardiolipin in a solution of PB [H2O:DMSO, 98:2, v/v], pH 7.0. B) Titration curve of salt 1 (4.7
at 633 and 640 nm.

ꢀ
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
55
pentamethine 5 did not localize in any of the tested cell lines, not
even after prolonged period (up to 16 h) and increased concen-
trations (up to 2 mM), as shown in Fig. 4. This behavior might be
caused by strong hydrophilicity of salt 5. We found, that this one
displayed significantly enhanced water solubility and hydrophi-
licity than other tested compounds. It is well known, that more
hydrophilic compounds can have poor membrane permeability
[16]. This fact is very likely to explain, why we did not observe
any cellular uptake of MS 5. Moreover, strong fluorescent signal
of compound 5 outside of the cells (probably in the extracellular
matrix) was detected, Figs. 4 and 5D. Since salt 5 cannot cross
cellular membrane, its selectivity for mitochondrial phospholipid
derivative cardiolipin, cannot control its cellular distribution. It is
indicated, that incorporation of another charge into the structural
Fig. 4. Intracellular localization of pentamethinium salts in live U-2 OS cells (human osteosarcoma cell line). A. Bright field of the cells. B. Fluorescent images of mitochondria (in
red) stained with pentamethinium dyes 1e4 and 6 (concentration of 1 was 10 nM, for 2e4 and 6 it was 35 nM) and extracellular space stained by 5 (35 nM). C. MTG stained
mitochondria (in green). D. Merge of the images B and C. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

ꢀ
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
56
Fig. 5. Pentamethinium salts compartmentalization with mtDNA in live U-2 OS cells. A. Bright field of the cells. B. Fluorescent images of mitochondria (in red) stained with
pentamethinium dyes 1 and 4 (concentration of 1 was 10 nM, 4 was 35 nM). C. Nuclear and mtDNA (in blue) stained with DAPI. D. Merge of the images B and C. E., F. Region of
interest of D. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
motif of indolium salt can significantly influence properties of
indolium probes based on methinium salts, such as suppression
of their cellular uptake.
This ability of pentamethine dyes (with the exception of 5) to
selectively and rapidly localize in mitochondria of living cells in low
nanomolar concentrations was in concordance with our previous
for antibodies and oligonucleotide coupling. Bare Cy5 (without
active NHS ester, azide or other derivatization) differs from the salt
6 only in the length of the alkyl substituents on nitrogens (methyl
and butyl chain).
Salt 6 exhibited high affinity to CL. This, together with its
specificity for CL surpassing PS in living cells, suggests that this Cy5-
like probe 6 has all aspects to become an efficient mitochondrial
probe. The large resemblance of salt 6 and Cy5 is a strong reason to
use Cy5 also as a valid mitochondrial probe. It is clear that selec-
tivity of mere Cy5 molecule is not dependent on its conjugation to
peptides (unless real mitochondria-targeting sequence is used at
protein amino terminus [17] or oligonucleotides as reported by
Lorenz [8]). Using mere Cy5 for mitochondria labeling would vastly
speed up and facilitated the staining process, summed in 10 min, in
comparison to a time-consuming and financially demanding
transfection (Cy5-oligonucleotide, mitochondria-targeted proteins)
and immunofluorescence procedures using Cy5 (or by other fluo-
rophore) conjugated antibodies, which could be avoided.
findings regarding pentamethine structures based on
g-aryl
substituted pentamethine salts [7]. Salts 1, 2e4 and 6 were retained
in mitochondria even after numerous washings. Further proof of
localization of 1, 2e4 and 6 in mitochondria was done by mtDNA
staining with DAPI (binding to A-T rich regions). In Fig. 5, one can
see that both the pentamethine salts and mtDNA reside in the same
compartment, mitochondria. This evidence gave us another proof
that salts 1, 2e4 and 6 localize in mitochondria.
The bio-stability hand in hand with photostability of the dyes
was tested by recordings of stained mitochondria and monitoring
the bleaching kinetics of the individual dyes (except for 5, which
did not localize inside cells). We found that salts 1, 2e4 and 6
exhibited unexpectedly high stability up to at least 1000 frames (1
frame per 1 s) of video recordings in U-2 OS and HeLa cells. Com-
pound 1 performed well during up to 3000 frames Supplementary
Video 1. This compound was also the most photostable in in vitro
assay when measured in near physiological conditions, in PBS at pH
7.0. Compounds 2e4 lasted up to 2000 frames, then gradual
bleaching occurred. Pentamethine 6, which performed comparable
(88%), but slightly better, in in vitro photostability tests when
compared to compound 7 (83%) and Cy5 (86%), was bleached after
ca 1000 frames, Supplementary Video 2. MTD acted similarly,
though with inferior photostability, this dye was bleached after ca
500 frames, Supplementary Video 3.
3. Experimental section
3.1. Chemicals
Unless otherwise specified, common reagents or materials were
obtained from commercial source and used without further puri-
fication. Chemicals for preparation of compounds 1e6 were pur-
chased from SigmaeAldrich. Aromatic malondialdehydes were
purchased from Acros Organics. Cy5 NHS ester (Cy5) was purchased
from Lumiprobe. Bovine heart solution of cardiolipin in ethanol was
purchased from SigmaeAldrich.
Videos from live-cell fluorescence microscopy of pentamethine
labeled mitochondrial network, toxicities of the probes, UVeVis
study of interactions with cardiolipin and phosphatidyl serine can
be found online at http://dx.doi.org/10.1016/j.dyepig.2013.12.021.
3.2. Syntheses and characterization of indolium fluorescent probes
The synthetic method for symmetrical indolium salts is based
on our previously published work [7], describing synthesis and
in vitro application of symmetrical g-substituted benzothiazolium
Intracellular localization study confirmed that the
g-aryl
substituted salts 1e5 can be utilized as selective mitochondrial
probes. Their mitochondrial localization can be explained by a
strong affinity to CL in inner mitochondrial membrane.
salts.
We prepared four types of compounds. The first type comprised
symmetrical pentamethinium salts 1, 2 and 4 with side indolium
units. Their synthesis was based on condensation reactions of
corresponding malondialdehyde with indolium salt [7,18], see
Scheme 1.
Secondly, we prepared unsymmetrical salt 3, which was
necessary for comparison studies. In the structure of salt 3, one of
the indolium units of the pentamethinium chain was replaced with
benzothiazolium heteroaromate unit. This compound was
2.4. Aspects of Cy5 applicability as an efficient mitochondrial probe
Finally, we clarified possible applicability of Cy5 as mitochon-
drial probe. We have tested the properties of salt 6, a pentam-
ethinium salt without g-substituents. It is the most resembling one
to Cy5, which is a commercially available probe widely used in
fluorescence microscopy, especially in a form of active NHS-ester

ꢀ
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
57
Scheme 1. Synthetic approach to compounds 1e6.
prepared by a previously published method based on condensation
of corresponding merocyanine with indolium salt [19], see
Scheme 1.
separated as a deep blue band. The separated fraction was evapo-
rated to dryness and the rest was macerated with diethylether and
sonicated for 2 min. The product was filtered off and dried in vac-
uum. Yield: (112 mg, 38%) of shiny green metallic powder. 1H NMR
(300 MHz, DMSO-d6, 25 ꢁC): 8.52 (2H, d, J ¼ 14.3 Hz), 8.42 (2H, d,
J ¼ 7.7 Hz), 7.67 (2H, d, J ¼ 7.13 Hz), 7.59 (2H, d, J ¼ 7.7 Hz), 7.42 (4H,
m), 7.28 (2H, m), 5.59 (2H, d, J ¼ 14.3 Hz), 3.78 (4H, bs), 1.76 (12H, s),
1.54 (4H, m), 0.74 (6H, bs); 13C NMR (126 MHz, DMSO-d6,
25 ꢁC):173.4, 151.9, 146.9, 143.1, 142.0, 141.2, 131.8, 131.6, 128.5,
125.2, 124.2, 122.6, 111.5, 100.8, 49.2, 44.9, 27.0, 20.3, 11.0.; HRMS
[M]þ (m/z) for C37H42N3O2 calculated: 560.3272, found: 560.3274.
As a third type, pentamethinium salt 6 without substitution in
g-position was prepared. Its synthesis was based on condensation
of indolium salt with corresponding malondialdehyde dianil hy-
drochloride as shown in Scheme 1.
The fourth type of compound represented by doubly charged
salt 5, was synthesized by quarternization using metyliodide of salt
1 in dimethylformamide (DMF), Scheme 1.
All compounds were fully characterized by NMR techniques on
Varian Gemini 300HC and by high resolution mass spectrometry
(HRMS) on Hewlett Packard-GC HP5890.
3.2.3. Compound 3
A flask was charged with quarternary salt (50 mg, 0.15 mmol),
merocyanine (50 mg, 0.14 mmol) and dry butanol (7 ml). The
mixture was heated to 110 ꢁC for 18 h. After this period, the mixture
was cooled to room temperature. The product in a form of green
powder was filtered off and dried in vacuum, the yield was 44 mg.
The rest of the product was separated from the filtrate by column
chromatography. The filtrate was evaporated to dryness. By the
means of chromatography (eluent: dichlormethan/methanol 10:1,
silica gel 5$40 cm), deep blue band was separated. The fraction was
evaporated to dryness and the rest was macerated with dieth-
ylether and sonicated for 2 min. The product was filtered off and
dried in vacuum, the yield was 12 mg. The total yield: (56 mg, 54%)
of shiny green metallic powder. 1H NMR (300 MHz, DMSO-d6,
25 ꢁC): 8.50e7.10 (16H, m), 6.01 (2H, d, J ¼ 13,9 Hz), 5.51 (2H, d,
J ¼ 13,4 Hz), 4.27 (2H, bs), 3.70 (2H, bs), 1.71 (6H, s), 1.66 (2H, m),
3.2.1. Compound 1
A flask was charged with quarternary salt (1600 mg, 4.86 mmol),
malondialdehyde (353 mg, 2.37 mmol), dry butanol (30 ml) and
three drops of triethylamine. The mixture was heated to 110 ꢁC for
18 h. After this period, the mixture was cooled to room temperature
and evaporated to dryness. The product was separated by column
chromatography (eluent: chloroform/methanol 10:1, silica gel
5$30 cm) as a deep blue band. The separated fraction was evapo-
rated to dryness and to the rest, ethylacetate was added and whole
mixture was sonicated for 2 min. The pure product was separated
by filtration. Yield: (347 mg, 80%) of deep green powder. 1H NMR
(300 MHz, DMSO-d6, 25 ꢁC): 8.78 (2H, d, J ¼ 5.6 Hz), 8.50 (2H, d,
J ¼ 14.3 Hz), 7.67 (2H, d, J ¼ 7.4 Hz), 7.46e7.24 (8H, m), 5.59 (2H, d,
J ¼ 14.4 Hz), 3.77 (4H, t, J ¼ 6.7 Hz), 1.75 (12H, s), 1.55 (4H, sextet,
J ¼ 7.4 Hz), 0.74 (6H, t, J ¼ 7.3 Hz).; 13C NMR (126 MHz, DMSO-d6,
25 ꢁC): 173.3, 151.7, 150.4, 144.0, 142.0, 141.2, 131.1, 128.5, 125.3,
125.2, 122.6, 111.5, 100.7, 49.2, 45.0, 27.0, 20.2, 11.0.; HRMS [M]þ (m/
z) for C36H42N3 calculated: 516.3373, found: 516.3377.
13
1.53 (2H, m), 0.77 (6H, m); C NMR (126 MHz, DMSO-d6, 25 ꢁC):
170.5, 167.3, 151.7, 148.6, 146.7, 143.2, 142.4, 141.3, 140.5,131.6, 129.3,
128.5, 128.3, 126.2, 126.0, 124.2, 123.9, 123.5, 122.4, 114.7, 110.4,
101.5, 97.7, 48.3, 48.1, 44.3, 27.3, 21.1, 20.0, 11.1, 10.9.; HRMS [M]þ
(m/z) for C34H36N3O2S calculated: 550.2523, found: 550.2528.
3.2.2. Compound 2
A flask was charged with quarternary salt (300 mg, 0.91 mmol),
malondialdehyde (83 mg, 0.43 mmol) and dry butanol (14 ml). The
mixture was heated to 110 ꢁC for 18 h. After this period, the mixture
was cooled to room temperature and evaporated to dryness. The
product was separated by column chromatography (eluent:
dichlormethan/methanol 10:1, silica gel 5$40 cm). The product was
3.2.4. Compound 4
A flask was charged with quarternary salt (135 mg, 0.41 mmol),
malondialdehyde (40 mg, 0.20 mmol) and dry butanol (7 ml). The
mixture was heated to 110 ꢁC for 18 h. After this period, the mixture
was cooled to room temperature and evaporated to dryness. The
product was separated by column chromatography (eluent:

ꢀ
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
58
dichlormethan/methanol 10:1, silica gel 5$40 cm) as a deep blue
band. The separated fraction was evaporated to dryness and to the
rest, macerated with diethylether and sonicated for 2 min. The
product was filtered off and dried in vacuum. Yield: (63 mg, 45%) of
brown-red metallic shiny powder. 1H NMR (300 MHz, DMSO-d6,
25 ꢁC): 8.56 (1H, s), 8.52 (2H, d, J ¼ 4.8 Hz), 8.11 (1H, d, J ¼ 7.9 Hz),
8.07 (1H, d, J ¼ 13.3 Hz), 7.84 (1H, t, J ¼ 7.04 Hz), 7.69 (3H, m), 7.60
(1H, d, J ¼ 8.4 Hz), 7.40 (4H, m), 7.27 (2H, t, J ¼ 8.0), 6.00 (2H, d,
J ¼ 14.1 Hz), 3.70 (4H, bs), 1.79 (12H, s), 1.56 (4H, m), 0.66 (6H, t,
(illuminated and in the dark) were collected (Cintra 404, GBC Sci-
entific) after 10, 20 and 30 min (corresponding total light doses
were 3.6 and 9 J cmꢂ2). Decrease of absorbance of solutions of 1e7
and Cy5 was evaluated in dark and after exposure to the light of
wavelengths longer than 500 nm. The amounts of the original
absorbance after 30 min light exposure are expressed in percentage
(Table 1) with the only exception of compound 7, of which the
photostability was measured after 1 h. At that time the stabilization
of the absorbance in the PBS was reached. Afterwards, the solution
was subjected to photolysis (or was maintained in dark) according
to the common procedure used for the other compounds. The
percentage of original absorbance at absorption maximum was
calculated as a mean of two replicates.
13
J ¼ 7.4); C NMR (126 MHz, DMSO-d6, 25 ꢁC): 173.4, 155.5, 152.3,
147.9, 142.0, 141.2, 136.8, 132.4, 129.9, 128.7, 128.4, 128.1, 126.9,
126.7, 125.0, 123.6, 122.5, 111.4, 100.9, 49.1, 44.9, 27.1, 20.2, 10.9;
HRMS [M]þ (m/z) for C40H44N3 calculated: 566.3530, found:
566.3533.
3.4. Determination of the absorption spectra and absorption
3.2.5. Compound 5
coefficients of pentamethinium salts
A flask was charged with pentamethinium salt 1 (100 mg,
0.015 mmol), dry DMF (7 ml) and excess of methyliodide (1 ml). The
mixture was heated to 40 ꢁC for 18 h. After this period, the mixture
was cooled to room temperature and diluted with diethylether
(50 ml). The precipitate was filtered off and washed three times
with diethylether (5 ml). The product was dried in vacuum. Yield:
(130 mg, 100%) of shiny green metallic powder. 1H NMR (300 MHz,
DMSO-d6, 25 ꢁC): 9.11(2H, d, J ¼ 6.3 Hz), 8.48 (2H, d, J ¼ 14.6 Hz),
8.12 (2H, d, J ¼ 6.4 Hz), 7.70 (2H, d, J ¼ 7.4 Hz), 7.48 (2H, d,
J ¼ 8.0 Hz), 7.43 (2H, t, J ¼ 7.9 Hz), 7.32 (2H, t, J ¼ 7.5 Hz), 5.74 (2H, d,
J ¼ 14.0 Hz), 4.43 (3H, s), 3.98 (4H, t, J ¼ 6.7 Hz), 1.77 (12H, s), 1.63
(4H, sextet, J ¼ 7.1 Hz), 0.83 (6H, t, J ¼ 7.4 Hz); 13C NMR (126 MHz,
DMSO-d6, 25 ꢁC): 174.4, 152.9, 146.0, 141.8, 141.4, 129.3, 128.5, 125.5,
122.6, 111.8, 100.1, 49.4, 47.6, 44.9, 26.9, 20.6, 11.0; HRMS [M]þ (m/z)
for C37H45N3 calculated: 265.6801, found: 265.6804.
Absorption dependence on concentration of the tested probes
1e6 and Cy5 was measured in DMSO, methanol (MeOH) and
phosphate buffer [PB] (H2O:DMSO, 98:2, v/v) at pH 7.0. Plastic cu-
vettes with optic path of 1 cm were used for absorption spectra
measurements. Concentrations of the salts varied in the range of 0e
10
m
mol lꢂ1. Absorption spectra were obtained using UVeVis
spectrophotometer (Cary 400, Varian; USA), measured at room
temperature. Molar absorption coefficients were calculated from
their absorbance maxima by linear regression with Microsoft Office
Excel 2010 software.
3.5. Determination of the conditional constants of the salts titrated
with cardiolipin and phosphatidylserine
The association of the salts 1e6 with the tested analytes, car-
diolipin (CL) and phosphatidylserine (PS), was studied by means of
UVeVis spectroscopy in PB (H2O:DMSO, 98:2, v/v) at pH 7.0. Con-
ditional constants (Ks) were calculated from the absorbance
changes of the salts using absorbance maximum of pure salts and
3.2.6. Compound 6
A flask was charged with indolium salt (1320 mg, 4.01 mmol),
malondialdehyde dianil hydrochloride (520 mg, 2.01 mmol) and
dry pyridine (25 ml). The mixture was heated to 90 ꢁC for 18 h. After
this period, the mixture was cooled to room temperature and
evaporated to dryness. The crude product was separated by column
chromatography (eluent: dichlormethane/methanol 10:1, silica gel
5$40 cm). The crude product was separated as a deep blue band.
Second purification was done by column chromatography (eluent:
diethylether/methanol 10:1, silica gel 5$25 cm). The separated deep
blue fraction was evaporated to dryness and to the rest, dieth-
ylether (5 ml) was added and whole mixture was sonicated for
1 min. The pure product was separated by filtration. Yield: (158 mg,
7%) of deep green powder. 1H NMR (300 MHz, DMSO-d6, 25 ꢁC):
8.34 (2H, t, J ¼ 12,9 Hz), 7.63 (2H, d, J ¼ 7.35 Hz), 7.41 (4H, m), 7.24
(2H, m), 6.59 (1H, t, J ¼ 12.3 Hz), 6.32 (2H, d, J ¼ 13.8 Hz), 4.08 (4H, t,
J ¼ 7.0 Hz), 1.72 (4H, m), 1.69 (12H, s), 0.95 (6H, t, J ¼ 7.3 Hz); 13C
NMR (126 MHz, DMSO-d6, 25 ꢁC): 172.7, 154.0, 142.1, 141.1, 128.4,
125.5, 124.7, 122.4, 111.2, 103.2, 48.9, 44.7, 27.2, 20.4, 11.0; HRMS
[M]þ (m/z) for C31H39N2 calculated: 439.3108, found: 439.3109.
their complexes (DA) by nonlinear regression using Letagrop Spefo
2005 software. Error of the measurements was expressed as the
standard deviation of the three times measured data and calculated
curve. Concentration of the salts used was 4.7
m
mol lꢂ1, concen-
tration of the studied analytes was in the range of 0e0.5 mmol lꢂ1
.
3.6. Fluorescence emission spectra measurement of
pentamethinium salts
Fluorescence emission spectra of 1e6 and Cy5 were collected
using a Fluoromax-2 spectrofluorometer in the DMSO, MeOH and
PB (H2O:DMSO, 98:2, v/v) at pH 7.0 at room temperature. The
whole fluorescence emission spectra of compounds 1e6 could not
be measured because the wavelengths of their emission and exci-
tation maxima partially overlap. Therefore, we measured only parts
of their spectra.
3.3. Photostability measurements
3.7. Determination of quantum yields of pentamethinium salts 1e6
Photostability of compounds 1e7 and Cy5 was studied in
dimethylsufoxide (DMSO) and in phosphate buffer saline (PBS) at
pH equal to 7.0. Concentrated HCl or NaOH have been used to adjust
pH of PBS to 4.8 and 9.3 to determine if the pH of PBS influences
photostability of the examined dyes. Two solutions were prepared
in each experiment: one solution was kept in dark, the second was
illuminated with a 150 W halogen lamp with an edge-pass filter
(Panchromar, Germany) that transmitted light at wavelengths
500 nm and longer. The fluency rate at the level of the solution in
cuvette was 5 mW cmꢂ2. Absorption spectra of both solutions
The quantum yields (F) of pentamethinium salts 1e6 and Cy5
were measured using thiadicarbocyanine, as a standard, likewise in
our previous study [7]. Samples of the pentamethinium salts were
prepared from fresh stock solutions to reach the absorbance lower
than 0.04 at lmax (to prevent the inner filter effect in fluorescence
measurement). The fluorescence emission spectra of the salts 1e6
were measured at their excitation maxima. For measurement of the
emission spectra of the standard, the excitation maxima of the
studied salts 1e6 were applied. For both sets of salts, the area under
each of the fluorescence emission curve was calculated and

ꢀ
T. Bríza et al. / Dyes and Pigments 107 (2014) 51e59
59
corrected for the Rayleigh peak area (if necessary). The average
fluorescence emission peak areas were determined for each sam-
ple. The relative fluorescence quantum yields were calculated using
structural factors influence the overall dye photostability. All con-
clusions regarding to relations between structure and photo-
physical properties of the selected probe must be done with
consideration to the used medium, as it is evident from the pho-
tostabilities shown in Table 1.
Probes 1e4 and 6 showed excellent mitochondrial selectivity,
Fig. 4 based on cardiolipin affinity, Table 3. The selectivity was
strongly governed by their charge density of the probe. Introduc-
tion of the second positive charge in probe 5 resulted in total loss of
the probe selectivity, Figs. 4 and 5D. The probes 1e4 and 6 can be
used for live mitochondrial imaging in very low, nanomolar, con-
centrations which are by two orders of magnitude lower than their
cytotoxic concentrations, see Supplementary materials.
The widely commercially used fluorescent probe Cy5 could be
used in a bare form (without active NHS-ester or likewise) as a
selective mitochondrial probe as well without need of conjugation
to any targeting agent.
the following equation: Fx
¼
Fs∙Fx/Fs/Ax∙As/ns∙nx, where F repre-
sents quantum yield; F stands for integrated area under the cor-
rected fluorescence emission spectrum; A is absorbance at the
excitation wavelength; n is refractive index of the used solvent and
the subscripts “x” and “s” refer to the unknown sample and to the
standard compound, respectively. Quantum yields of salts 1e6
were determined in DMSO, MeOH and in water environment rep-
resented by PB, pH 7.0.
3.8. Cell uptake
U-2 OS (human osteosarcoma) and HeLa (human cervix carci-
noma) cells (1∙105 wellꢂ1) were seeded on 35 mm glass bottom
dishes for live-cell imaging (MatTek Corporation, USA) in high
glucose DMEM media (PAA, Austria) supplemented with 10% fetal
bovine serum (FBS, Invitrogen, USA), 1% MEM vitamins solution
(Invitrogen, USA) and 2 mM L-Glutamine (SigmaeAldrich, USA).
Cells were left to adhere overnight (16 h). The attached cells were
washed twice with PBS (pH 7.4, 37 ꢁC) and incubated with the
tested pentamethine dye (10e1000 nM) dissolved in complete cell
culture medium without phenol red at 37 ꢁC for 0e60 min. After
the incubation period, the cells were washed twice with PBS and
fresh medium without phenol red was added. To assess the exact
localization of the used cyanine dyes, we used green-emitting
mitochondrial marker MitoTracker Green (MTG). After the pen-
tamethine dye uptake, the cells were rinsed twice with PBS and
incubated with 50 nM solution of MTG in phenol red free complete
medium for 12 min and rinsed with PBS again. Staining of mito-
chondrial DNA (as well as nuclear) was performed in live cells using
These dyes, suitable for fluorescence microscopy imaging of
mitochondria, might be used not only for studies of mitochondrial
morphology, but also for special applications including mitochon-
drial physiology, pathophysiology, activity, and determination of
mitochondrial mass in a cell and for monitoring of effect of new
drugs on mitochondria.
Acknowledgments
This work was financially supported by Grant Agency of the
Academy of Sciences of the Czech Republic (KAN200100801), Grant
Agency of the Czech Republic (P303/11/1291), BIOMEDREG
(CZ.01.05/2.1.00/01.00.30), Charles University (UNCE 204011/2012
and P24/LF1/3) and PROPMEDCHEM CZ.1.07/2.300/30.0060.
1 m
g mlꢂ1 of DAPI for 5 min as described in our previous study [7].
Appendix A. Supplementary data
3.9. Fluorescence microscopy in living and fixed cells
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.dyepig.2013.12.021.
The intracellular localization of the methine compounds was
studied by real-time live-cell fluorescence microscopy at 37 ꢁC and
in 5% CO2 atmosphere. The images were acquired by an inverse
fluorescent microscope Olympus IX-81, Cell^R System using high-
stability 150 W xenon arc burner and EM-CCD camera C9100-02
(Hamamatsu, Germany). Living cells were analyzed under physio-
logical conditions by a 60ꢃ oil immersion objective (Olympus,
Japan) with numerical aperture of 1.4 using sets of Olympus filters:
U-MWU-2, U-MNB2, U-MNG2, U-MWIY2 and U-DM-CY5. All im-
ages were deconvolved using Cell^R basic deconvolution module.
Fluorescence microscopy video recordings of moving mito-
chondria stained with the tested cyanines were recorded in the
same module as described for capturing images. The videos were
performed 10 min after staining (35 nM concentration of the tested
compounds) using 57% light intensity, exposure time 33 ms, in-
terval between individual frames 1 s and the total count of the
recorded frames 2000 (ca 33 min long video). Commercial red-
emitting mitochondrial tracker for live-cell imaging, MitoTracker
Deep (MTD; 50 nM), was used for comparison.
References
[1] Stephens DJ, Allan VJ. Science 2003;300:82.
[2] a) Zhang J, Campbell RE, Ting AY, Tsien RY. Nat Rev Mol Cell Biol 2002;3:906;
b) Wu X, Chang S, Sun X, Guo Z, Li Y, Tang J, et al. Chem Sci 2013;4:1221;
c) Guo Z, Zhu W, Zhu M, Wu X, Tian H. Chem Eur J 2010;16:14424.
[3] Hongyuan Z, Lele L, Chao G, Ruyi S, Qiaochun W. Dyes Pigment 2012;94:266.
[4] Chen G, Song F, Wang X, Sun S, Fan X, Peng X. Dyes Pigment 2012;93:1532.
[5] Strekowski L, Gupta RR. In: Heterocyclic polymethine dyes. Heidelberg:
Springer-Verlag Berlin Heidelberg; 2008.
[6] Markova IL, Fedyunyayeva IA, Povrozin YA, Semenova OM, Khabuseva SU,
Terpetschnig EA, et al. Dyes Pigment 2013;96:535.
ꢀ
ꢀ
[7] Rimpelová S, Bríza T, Králová J, Záruba K, Kejík Z, Císarová I, et al. Bioconjug
Chem 2013;24:1445.
[8] Lorenz S, Tomcin S, Mailänder V. Microsc Microanal 2011;17:440.
[9] Henary M, Mojzych M. Stability and reactivity of polymethine dyes in solution.
In: Strekowski L, editor. Heterocyclic polymethine dyes. Heidelberg: Springer-
Verlag Berlin Heidelberg; 2008. pp. 221e38.
[10] Peng X, Yang Z, Wang J, Fan J, He Y, Song F, et al. J Am Chem Soc 2011;133:
6626.
[11] van Meer G, de Kroon AIPM. J Cell Sci 2011;124:5.
[12] Underhaug Gjerde A, Holmsen H, Nerdal W. Biochim Biophys Acta 2004;1682:28.
[13] Shao C, Novakovic VA, Head JF, Seaton BA, Gilbert GE. J Biol Chem 2008;283:
7230.
4. Conclusions
[14] Dahlberg M, Marini A, Mennucci B, Maliniak A. J Phys Chem A 2010;114:4375.
[15] Nury H, Dahout-Gonzalez C, Trézéguet V, Lauquin G, Brandolin G, Pebay-
Peyroula E. FEBS Lett 2005;579:6031.
[16] Szachowicz-Petelska B, Figaszewski Z, Lewandowski W. Int J Pharm 2001;222:169.
[17] Omura T. J Biochem 1998;123:1010.
We successfully designed, synthesized and carefully character-
ized a series of novel symmetric
salts 1e5. We have studied the influence of the side heteroaromatic
units and -aryl substitution on the photophysical properties of the
probes. The effect of side unit is influenced by the nature of used
medium, like in the case of -substitution, and both of these
g-substituted pentamethinium
ꢀ
ꢀ
[18] Bríza T, Kejík Z, Císarová I, Králová J, Martásek P, Král V. Chem Commun
(Camb) 2008;28:1901.
g
ꢀ
[19] Bríza T, Kejík Z, Králová J, Martásek P, Král V. Cellect Czech Chem C 2009;74:
g
1081.