Applying styryl quinolinium fluorescent probes for imaging of ribosomal RNA in living cells

Article abstract of DOI:10.1016/j.dyepig.2019.107986

The detection of subcellular domains in cells can be obtained by specific fluorescent markers. Here we report the use of styryl quinolinium dyes that selectively stain ribosomal RNA (rRNA) in nucleoli and in the cytoplasm of mammalian cells. Specifically,

Full text of DOI:10.1016/j.dyepig.2019.107986

Journal Pre-proof
Applying styryl quinolinium fluorescent probes for imaging of ribosomal RNA in living
cells
Abed Saady, Eli Varon, Avi Jacob, Yaron Shav-Tal, Bilha Fischer
PII:
S0143-7208(19)32054-6
DOI:
https://doi.org/10.1016/j.dyepig.2019.107986
DYPI 107986
Reference:
To appear in: Dyes and Pigments
Received Date: 30 August 2019
Revised Date: 7 October 2019
Accepted Date: 20 October 2019
Please cite this article as: Saady A, Varon E, Jacob A, Shav-Tal Y, Fischer B, Applying styryl
quinolinium fluorescent probes for imaging of ribosomal RNA in living cells, Dyes and Pigments (2019),
doi: https://doi.org/10.1016/j.dyepig.2019.107986.
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.
© 2019 Published by Elsevier Ltd.

Applying styryl quinolinium fluorescent probes for imaging
of ribosomal RNA in living cells
Abed Saadya, Eli Varonb, Avi Jacobb, Yaron Shav-Talb and Bilha Fischera*
a Department of Chemistry, Bar-Ilan University, Ramat-Gan 52900, Israel.
b Faculty of Life Sciences & Institute of Nanotechnology, Bar-Ilan University,
Ramat-Gan 52900, Israel.
* Corresponding author. Fax: 972-3-6354907. Tel: 972-3-5318303.
Bar-Ilan University, Ramat-Gan 52900, Israel
E-mail: Bilha.fischer@biu.ac.il
1

Abstract
The detection of subcellular domains in cells can be obtained by specific fluorescent markers.
Here we report the use of styryl quinolinium dyes that selectively stain ribosomal RNA (rRNA)
in nucleoli and in the cytoplasm of mammalian cells. Specifically, we synthesized a series of 1-
methyl-4-(substituted) styryl-quinolinium derivatives, 12a–l. We developed highly efficient
microwave-assisted synthesis which prevents the formation of side-products, leading to the
products in yields greater than 90%. Compounds 12c-f and 12i in various solvents exhibited
maximum absorbance at 500 - 660 nm, molar extinction coefficient of 25400 – 49000 M-1cm-1,
and emitted at 630 - 715 nm. Dyes 12a–l are highly photochemically stable. Dye 12e specifically
stained nucleoli and the cytoplasm and is non-toxic. Various tests showed the markedly higher
affinity of the dye for rRNA. We demonstrate that 12e is an attractive staining reagent for
visualizing and assessing rRNA in both fixed and living cells.
2

Introduction
Dyes emitting at long wavelengths, green to red regions have attracted the attention of scientists
for use in various applications,1 including visualization of biomolecules in living systems, real-
time tracking of cellular events, and fluorescent dye guidance during surgeries.2-9 These
applications are based on the minimal auto-fluorescence of cells at wavelengths above 600 nm,
the penetration depths of NIR- dyes, as well as their sensitivity and minimal photo-damage on
cells and tissues.10-12
Styryl quinolinium- and merocyanine- dye classes constitute scaffolds applied for the
development of NIR dyes.13 Specifically, a styryl quinolinium dye contains two non-identical
aryl groups usually joined by an ethylene moiety. One of the aryl groups is electron deficient
(e.g. pyridinium moiety) while the other one is electron rich, thus resulting in a push-pull
system.14 Unlike the positively charged styryl quinolinium dyes, merocyanine dyes are
zwitterions and hence are essentially neutral. Merocyanine push-pull dyes contain two ethylene-
bridged aryl groups, one of which is positively charged (e.g. pyridinium moiety) and the other
one is negatively charged (e.g. phenolate moiety) (Fig. 1).
Fig. 1 Resonance structures of anion of p-hydroxy-styryl quinolinium dye.
Merocyanine dyes are used for detection of DNA or RNA since they often exhibit enhanced
fluorescence intensity by binding to nucleic acids.15-17
3

The Knoevenagel condensation is used for production of dyes,18 such as coumarin derivatives,19
and styryl quinolinium derivatives, e.g. 12a or 12b.20,21 The synthesis of the latter involves the
reaction of 1,4-dimethyl quinolinium salts with aromatic aldehydes in ethanol or toluene under
reflux for overnight, or in methanol at 50 °C for at least 3-8 h.22 Most of these processes are time
consuming and require purification of the final product, resulting in medium to good yields.
The Knoevenagel condensation is also applied for the synthesis of pharmaceuticals, such as
nifedipine hypertension drugs, 1-3,23,24 1-methyl-4-(substituted) styryl-quinolinium analogs 4-6
used for the treatment of tuberculosis,25 derivatives 7-8, used as choline acetyltransferase
inhibitors,26 and analog 9 which shows antitumor activity.27 Other applications of styryl
quinolinium dyes have been described.15, 28-30
Scheme 1. Previously reported styryl-quinolinium derivatives.
4

The demand for new fluorescent dyes that absorb and emit above 550 nm, encouraged us to
develop and characterize new dyes based on the scaffolds of merocyanine and styryl dyes.
Specifically, we report here on the microwave assisted Knoevenagel condensation31 of methyl
quinolinium salt, 10, and substituted benzaldehyde derivatives (11a-11l), to obtain (E)-2-
(substituted styryl)-1,4-dimethylpyridinium iodide analogs 12a – 12l with high efficiency.
Scheme 2. Synthetic scheme for the preparation of 12a – 12l.
In addition, we report on the photophysical properties and photostability of products 12a – 12l in
several solvents and pH values. The most promising dye termed 12e was further evaluated for its
photophysical properties and mode of interaction with nucleic acids in solution. Next, the safety
and photostability of the compounds was studied. Finally, we describe the application of dye 12e
for rapid and selective nucleoli staining in living cells.
5

Results
Synthesis of dyes 12a-l
Our strategy to design new dyes was based on increasing the polarization within styryl
quinolinium- and merocyanine push-pull scaffolds (Fig. 2). For instance, merocyanine analog
12a was substituted by one or two fluorine atoms at the ortho positions of the phenol moiety, 12c
and 12e, to stabilize the phenolate electron donating group. In contrast to 12c,e, the phenyl
moiety in 12i, was substituted by an electron donating, dimethylamino group (Scheme 2).
For synthesizing these dyes we targeted a facile and highly efficient method. Hence, we applied
microwave-assisted Knoevenagel condensation of 1,4-dimethyl quinolinium iodide, 10,21 and
several substituted benzaldehyde derivatives, 11a-l. Benzaldehyde derivatives 11a-d, 11f, and
11h-l, are commercially available. 3,5-Difluoro-p-hydroxybenzaldehyde, 11e, and 2,3,5,6-
tetrafluoro-p-hydroxybenzaldehyde, 11g, were synthesized from 2,6-difluorophenol,32 and 1-(di-
ethoxymethyl)-2,3,4,5,6-pentafluorobenzene,33 respectively.
All microwave-assisted reactions in ethanol were completed in 5 - 30 min at 70-80 °C. The
products were mostly obtained in a good to excellent yield (> 90%). The same reactions in
ethanol under reflux were completed in 4-16 h, with lower yields (27 – 85%)20 (Table 1).
We first explored microwave-assisted Knoevenagel condensation for the preparation of the
known compounds 12a, 12b and 12i.
The Knoevenagel reaction involving benzaldehyde derivatives having free hydroxyl group on the
aromatic ring (11a, 11c, 11e, 11g, 11h and 11j) occurs also in excellent yield and does not
require any protection of the phenolic moiety.
6

Table 1. Comparison of the reaction time and yield of products 12a- l obtained under microwave
vs. under normal conditions.
Microwave
condition
Normal condition
in EtOH
in EtOH
Product
Yield (%) Reaction time
(min)
Temperature
Yield (%)
Reaction time
Temperature
(°C)
(°C)
70
80
75
75
75
75
75
80
80
80
80
80
(h)
9
12a
12b
12c
12d
12e
12f
12g
12h
12i
94
96
91
95
99
90
99
95
98
94
91
90
15
25
10
15
5
65
53
27
55
85
73
45
37
81
64
60
59
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
6
12
12
9
10
5
14
10
16
8
15
5
20
30
30
12
18
20
12j
12k
12l
Dyes 12a–12l are barely soluble in nonpolar organic solvents but are readily soluble in polar
organic solvents. Dyes 12a, 12c, 12e, 12g, 12j and 12l that bear a hydroxyl group at the para
position of the styryl moiety, are soluble in a wide range of solvents giving colored solutions, the
color of which is solvent-dependent (Fig. 3).
7

Fig 2. Styryl-quinolinium derivatives 12a-l synthesized and evaluated in this study.
Fig. 3. Compound 12e in various solvents, from left to right: acetic acid, water, methanol,
ethanol, isopropanol, DMSO, DMF, acetone and dichloromethane.
In polar solvent such as water, the zwitterion form is dominates, while in nonpolar solvents such
as chloroform, the neutral form prevails.15
8

Photophysical properties of dyes 12a-l
The photophysical properties of compounds 12a – 12l were evaluated in various solvents at a
range of relative polarity values; between 1 34 to 0.355 (acetone)35 (Table 2).
Table 2. Photophysical properties of derivatives 12a- 12l.
Dye
λAbs
ϵ
(M-1cm-1)
λem
Dye
λAbs
ϵ
(M-1cm-1)
λem
(nm)
(nm)
(nm)
(nm)
Water
12a
12b
12c
MeOH
12a
12b
12c
460
21200
20700
26300
24900
45800
41200
23300
22900
29800
19200
16400
13300
515
501
560
555
605
580
518
508
575
503
495
499
475
440
560
508
535
520
505
490
530
478
445
480
22100
19900
27200
24900
46100
40900
23800
22900
30300
18600
17400
12200
525
490
610
600
670
648
591
580
665
500
495
500
438
508
502
520
515
503
478
510
470
440
475
12d
12e
12d
12e
12f
12f
12g
12h
12i
12g
12h
12i
12j
12j
12k
12l
12k
12l
EtOH
12a
12b
12c
DMSO
12a
12b
12c
480
450
565
21900
21300
27100
525
495
640
490
465
566
23500
20600
25700
528
501
660
9

510
556
548
509
506
544
480
460
485
28200
49000
41200
23900
22700
31000
17900
16900
13900
605
685
665
625
608
675
495
495
506
520
600
591
499
500
590
480
450
485
23600
44700
42100
22950
20900
30100
18400
17900
12200
610
715
675
603
595
703
500
501
509
12d
12e
12f
12d
12e
12f
12g
12h
12i
12g
12h
12i
12j
12j
12k
12l
12k
12l
DMF
12a
12b
12c
12d
12e
12f
Acetone
12a
12b
12c
12d
12e
12f
490
460
555
519
599
590
495
500
588
482
450
485
21500
21100
28100
25000
44900
41100
22100
23400
29800
19200
16400
11100
520
500
660
610
715
675
603
595
705
503
500
505
490
455
510
490
650
590
500
505
580
460
490
480
19900
18100
27600
26300
47100
44500
23200
22900
28300
18700
15800
12200
520
505
628
605
702
685
590
590
695
495
515
508
12g
12h
12i
12g
12h
12i
12j
12j
12k
12l
12k
12l
Compound 12a exhibits λabs at 490 nm in DMSO, while substitution of one ortho-hydrogen atom
of the phenol with fluorine atom, 12c, results in 76 nm red shift. Moreover, substitution of both
10

ortho positions of the phenol by fluorine atoms, 12e, results in a 110 nm red shift (in DMSO).
Similarly, molar extinction coefficient values (in water) are affected by fluorine atom(s) at ortho-
position: 12a, ϵ 21200 M-1cm-1, 12c, ϵ 26300 M-1cm-1, 12e, ϵ 45800 M-1cm-1. Furthermore, we
explored the effect of substitution of 12a phenol group by four fluorine atoms, 12g, on its
photophysical properties. Surprisingly, all measured parameters, λabs, λem, and ϵ, decreased for 12g
as compared to 12e. For instance, while 12e in EtOH exhibited for λabs, λem, and ϵ, values of 556
nm, 49000 M-1cm-1, and 685 nm, respectively; 12g showed 509 nm, 23900 M-1cm-1, and 625 nm,
respectively.
Compounds 12d and 12f, both contain a methoxy group in the phenol para position, and one
fluorine atom, 12d, or two fluorine atoms, 12f, in the ortho position. We found that ϵ and λabs
values of 12f are higher than those for 12d in all solvents. Yet, substitution of the para position
by methoxy vs. hydroxy group, resulted in decrease of all measured parameters, λabs, λem, and ϵ, in
all solvents, for 12f as compared to 12e, as well as for 12a as compared to 12b. These data
indicate the importance of the hydroxyl group at the phenolic para position for improving the
photophysical properties of these styryl quinolinium dyes.
We compared the effect of electron withdrawing groups at ortho position, e.g. nitro vs. fluoro
group, on
ϵ in 12l vs. 12c. We noticed that in most solvents ϵ values of 12l decreased as
compared to those of 12c due to the stronger electron withdrawing effect of the nitro group.
Compound 12i bearing a strong electron donating group at the p-position, N(CH3)2, shows also
high λabs, λem, and ϵ values, about the same as for 12c, in all solvents.
Compounds 12g, 12h, 12j and 12k bearing meta substituents – F or OH – showed poor
photophysical properties.
11

Notably, the color of the solution of dyes containing p-hydroxyl group, e.g. 12a, 12c and 12e,
(merocyanine push-pull systems) changes depending on the solvent and its polarity. In general,
the more polar solvent is, the shorter is the absorption wavelength of the dye (bathochromic
shift). This effect stems in part from the stabilization of the ground state of the merocyanine
molecule in polar solvents, which increases the energy gap between the ground state and excited
states, and corresponds to shorter wavelengths (increased energy) of the absorbed light.
Next, we evaluated the effect of pH on the absorption and emission wavelengths of 12a-l in basic
vs. acidic solutions. Dyes containing p-OH/NMe2 groups exhibited red-shift of the wavelengths
of absorption and emission, up to 11 nm, in basic vs. acidic medium (Table 3).
Table 3. Absorption and emission wavelengths of derivatives 12a- 12l under acidic (pH 2) and
basic conditions (pH 10).
Dye
λabs (nm)
pH 2
460
λabs (nm)
pH 10
467
λem (nm)
pH 2
515
λem (nm)
pH 10
522
12a
12b
12c
12d
12e
12f
12g
12h
12i
438
438
501
500
508
512
560
564
502
502
555
556
520
531
605
608
515
515
580
580
503
501
518
520
478
478
508
508
510
508
575
578
12

470
440
475
475
447
477
503
495
499
500
499
502
12j
12k
12l
The quantum yields of dyes 12a-l in all the solvents were rather low (0.0011- 0.0177). However,
in glycerol ca. 3-fold increase of the quantum yield was measured. Table 4 summarizes the
quantum yields of dyes 12a- 12l in water and glycerol.
Table 4 Quantum yields of the dyes in water and glycerol.
Dye
12a
12b
12c
12d
12e
12f
12g
12h
12i
Φa (%)
0.69
0.61
0.94
0.87
1.77
1.61
0.62
0.48
1.07
0.79
0.11
0.28
Φa (glycerol) (%)
2.13
1.74
2.84
3.47
5.55
4.89
1.68
1.09
3.74
2.91
0.31
0.92
12j
12k
12l
13

a The fluorescence quantum yields of the dyes (2 µM) were determined in ethanol at 25 °C using
rhodamine B as the standard.
The same quantum yield values were observed in MeOH, EtOH, DMF, DMSO and acetone. The
high viscosity of glycerol (950 cP)36 reduces degrees of freedom and free rotations in the dye
molecules,37 and hence allows planarity and enhanced light emission.
Photostability of dyes 12a-l
Photostability is an important requirement for the application of a fluorescent dye in biological
studies. The photostability of dyes 12a-l was tested by exposure to a light source of a Cary
Eclipse Fluorescence Spectrophotometer (λex 500–650 nm), with the exclusion of ambient light.
2µM Dye solutions in PBS were irradiated for 200 min, with an assessment of their fluorescence
intensity at five-minute intervals (Fig. 5).
90
12a
12b
12c
12d
12e
12f
12g
12h
12i
80
70
60
50
40
30
20
10
0
12j
12k
12l
0
50
100
150
200
Time (min)
Fig. 5 Photostability of dyes 12a-l measured by the loss of fluorescence intensity (2 µM dye, in
PBS buffer).
14

After a 200-minute irradiation in PBS buffer, it was clearly observed that the fluorescence
intensity of dyes 12a- l was not affected by the irradiation. The remaining emission percentages
of the dyes after 200 min were: 12a (89%), 12b (93%), 12c (91%), 12d (88%), 12e (96%), 12f
(92%), 12g (95%), 12h (93%), 12i (95%), 12j (88%), 12k (94%) and 12l (87%). These data
indicate that all the dyes have excellent photostability.
Dyes 12a-l are not toxic to cells
We evaluated the toxicity of the promising dyes on T lymphocytes by flow cytometry.38 We
measured cell death by fluorescence-activated cell sorting (FACS) which is a powerful and
precise tool for measurement of a single cell within a population. We checked the cell death
before adding the dyes and 3 hours after the addition of the dyes. Dyes 12c, 12i and 12e were not
toxic for a period of 3 hours (Fig. S1).
Dye 12e binds RNA in solution
The photostability and the most promising photophysical properties of compound 12e
encouraged us to evaluate it as a probe or staining agent for nucleic acids (RNA). This was tested
by adding this dye to an RNA solution extracted from human cells and evaluating the increase in
fluorescence intensity. Notably, a significant, 3.1-fold increase in fluorescence intensity was
observed once compound 12e interacted with total RNA extract (Fig. 7). Dye 12e alone (black),
RNA solution+ dye 12e (blue), dye 12e+ RNA solution+ RNASE, control, (green), water +
RNASE, control, (red).
15

25
20
15
10
5
12e
RNA+ 12e
RNA+ RNASE+ 12e
UPW+RNASE+ 12e
0
500
550
600
650
700
Wavelength (nm)
Fig. 7 Emission spectra of probe 12e with RNA in water. Dye 12e alone (black), RNA solution+
dye 12e (blue), dye 12e+ RNA solution+ RNASE, control, (green), water + RNASE, control,
(red).
Dye 12e stains fixed and living cells
To analyze the fluorescent properties of compound 12e in cells and to determine whether the dye
can stably interact with RNA within the cells, we first determined the fluorescent properties of
the dye in cells. We applied the dye to fixed HeLa cells and using a scanning confocal
microscope equipped with a supercontinuum laser and tunable detection, we performed a
lambda-lambda scan to determine the excitation-emission spectrum of the compound. We found
that the peak fluorescent intensity was at the emission range of 647-667nm and optimal
excitation at 570 nm, clearly showing the detection of the cells (Fig. 8A). We then measured the
intensity of the signal within the images to show the spectrum as shown in Fig. 8B. Figure 8C
shows a 2D heat map representing the full scan as exported from the acquisition software and
16

shows the peak excitation vs peak emission, as above. The signal obtained in the images at 570
nm (Fig. 8A) showed staining of several substructures in the nucleus of the cells that resembled
nucleoli, while the nucleoplasm did not show signal. The cytoplasm also showed staining. This
staining was observed in other cell lines as well (Fig. S2). To confirm that nucleoli were stained
we performed immunofluorescent staining with a specific antibody to the nucleolar protein
fibrillarin. Indeed, compound 12e specifically stained nucleoli and the cytoplasm (Fig. 8D).
Nuclei were detected with the Hoechst dye.
17

Fig. 8: Spectrum scan shows peak of excitation/emission. (A) Dye 12e was added to HeLa cells
and these were scanned as described in the Methods section. The strongest excitation, at the
emission range of 647-667 nm was found to be 570 nm. Bar = 10 µm. (B) Images were analyzed
as described in the Methods and the image emission intensities were plotted. The excitation that
resulted in the strongest emission was found to be at 570 nm. (C) 2D heat map exported from
acquisition software showing the excitation vs. the emission profiles. (D) Dye 12e localizes to
nucleoli in treated cells. Nucleoli are identified by staining with an antibody to the nucleolar
marker fibrillarin (green). Hoechst DNA stain is in cyan. No 12e signal is detected in untreated
cells. Bar = 10 µm.
Since the dye stained subcellular structures in fixed cells we wanted to examine if it can be
applied to living cells as well. The dye rapidly entered the cells showing that it is highly cell
permeable. Within 30 seconds the compound was detected in the cytoplasm, and within 150
seconds the compound stained nucleoli (Fig. 9A). Additionally, the compound could be applied
to living cells for over 24 hours without affecting viability, cell morphology, or proliferation as
cell divisions continuously occurred indicating again that compound 12e is not toxic to cells
(Fig. 9B and Supplemental Movie S1). Notably, the compound remained fluorescently stable,
bright and specifically bound to nucleoli (Fig. 9B), with a high signal-to-noise ratio (Fig. 9C).
18

Fig. 9. Dye 12e rapidly stains living HeLa cells and does not affect viability. (A) Cells were
imaged every 5 sec for 5 min. The compound entered the cells within 30 sec and stained nucleoli
within 150 sec. (B) Cells were then imaged every 15 min for 26 hrs and the compound intensity
remained strong and specific to the nucleoli. The cells remained viable. Bar = 20 µm. (C) Signal-
19

to-noise ratio pixel intensity graph of nucleoli/nucleus and nucleoli/outside of the cell derived
from ImageJ software.
Dye 12e stains fixed and living cells
The in vitro data suggested that compound 12e binds to RNA. Since the nucleolus is the region
in the cell where ribosomal RNA is transcribed by RNA polymerase I, we further investigated
the properties of the dye to binding nucleic acids within cells. To this end, fixed and stained cells
were treated with either deoxyribonuclease (DNase) or ribonuclease (RNase) to remove DNA or
RNA from cells, respectively (Fig. 10). While with the treatment with DNase no considerable
loss of fluorescence signal was observed (Fig. 10A), after RNase treatment, cytoplasmic and
nucleolar fluorescence signal disappeared in comparison to untreated cells (Fig. 10B). These
digestion experiments validated the preference of compound 12e to attach to RNA since cells in
which the RNA was removed by RNase, did not stain anymore. When the cells were treated with
actinomycin D, an inhibitor of RNA polymerase II that transcribes mRNA (and not rRNA), there
was not much change in the intensity of the dye in the cells, suggesting that the dye preferably
binds to rRNA in the nucleolus and the cytoplasm.
20

Fig. 10. Compound 12e binds to RNA in cells. (A) Wide-field fluorescence images of untreated
or DNase treated cells stained with the compound 12e. Hoechst DNA stain is in cyan. (B)
Untreated cells or cells treated with RNase or with the actinomycin D transcriptional inhibitor.
DIC images are in grey. Bar = 10 µm.
To directly test the interaction of compound 12e with rRNA, we inhibited nucleolar rRNA
transcription by inhibiting RNA polymerase I activity with two specific inhibitors33, 34, 39, 40 (Fig.
11A). Indeed, these treatments significantly reduced the fluorescent signal throughout the
21

nucleoli and the cytoplasm, confirming the preference of compound 12e to rRNA. To further
examine whether the compound is specific to rRNA and not mRNA, we induced the assembly of
cytoplasmic stress granules that contain mRNAs but not rRNAs.41 This way foci containing
mRNAs were formed in the cytoplasm, which might be detectable by the compound if it binds
mRNA. Stress granules were formed by the addition of arsenite to the cells. However, the
compound did not localize with stress granules, implying that 12e is not binding to cytoplasmic
mRNA (Fig. 11B).
Fig. 11. Compound 12e binds rRNA in cells. (A) Transcriptional inhibition of RNA polymerase
I by CX-5461 and BMH-21 diminishes the fluorescence signal of compound 12e in cells. (B)
22

Stress granules formed by arsenite treatment and labeled with anti-G3BP1 (green) do not contain
compound 12e (white). Hoechst DNA stain is in cyan. DIC is in grey. Bar = 10µm.
A main difference between rRNA and mRNA molecules is that rRNA is highly structured and
contains many complex secondary structures. In order to prove that compound 12e has
preference to secondary structures in the rRNA, we purified rRNA and mRNA from total cell
RNA extracts, and examined the binding of the compound to these RNAs under two conditions.
At room temperature the secondary structures are preserved, and strong binding to rRNA was
observed (Fig. 12). Once the secondary structures of the RNA were destroyed by heating the
RNAs to 90°C, the binding to the rRNA was lost, indicating that indeed compound 12e binds the
complex secondary structures of rRNA.
7
6
5
Dye 12e
RNA+ 12e
4
Boiling RNA+ 12e
3
rRNA+ 12e
Boiling rRNA+ 12e
2
mRNA+ 12e
1
Boiling mRNA+ 12e
0
500
550
600
650
700
Wavelength (nm)
Fig. 12. Compound 12e binds to secondary structures in rRNA. The binding of compound 12e to
rRNA and mRNA in solution at room temperature and at 90°C. In the presence of structured
rRNA there is an increase in the fluorescent signal; however, when the samples were heated and
23

the secondary structures of the rRNA are dismantled, then no increase in fluorescence was
observed.
Dye 12e binds histones-free DNA extract
Dye 12e binds secondary structures in the rRNA that are probably double stranded (ds) however
nuclear dsDNA was not stained as well. We postulated that the histones forming the
nucleosomes on the DNA in cells might mask the binding of the dye. We therefore extracted
DNA from cells, with or without histones, and found that dye 12e did not bind to DNA with
histones and could bind only to DNA extract in which the histones were removed (Fig. 13).
12
10
Baseline
8
12e
12e + DNA with Histones
6
12e + DNA without Histones
4
2
0
550
600
650
700
-2
Wavelength (nm)
Fig. 13. Compound 12e binds to DNA without histones
24

Discussion
In this study, we have synthesized and established the structure-photoactivity relationship of new
and known merocyanine dyes, 12a-l (Fig. 14).
Generally, the photophysical properties- λabs, λem, and ϵ- of 12a-l were in the range of 440-658
nm; 485-715 nm; and 12,200-49,000 M-1cm-1, respectively. These values were solvent-
dependent. Specifically, the more polar the solvent is, the shorter is the absorption wavelength of
the dye.
The following structure-photoactivity relationships were observed: Substitution of the phenyl
para-position with OH group enhanced photoactivity, whereas OMe group reduced it.
Substitution of one or two fluorine atoms at ortho positions to OH/OMe groups enhanced
photoactivity, due to the 1000-fold greater acidity of 2,6-difluoro-phenol than that of phenol (pKa
values of phenol and 2,6-difluoro-phenol are 9.99 and 6.88, respectively) whereas additional
fluorine atoms at meta-position reduced overall photoactivity.
X = OH or Me N increases
2
photoactivity
X = OMe decreases photoactivity
Y / Y+Y’ = F increases photoactivity
Y=Y’=F Z=Z’= F X=OMe/OH
decrease photoactivity
Y=H Z=OH X=OH decrease
photoactivity
X = NO decreases photoactivity
2
Y=NO Y’=OMe Z’=OH
2
decrease photoactivity
Fig. 14 Structure-photoactivity relationship in styryl-quinolinium dyes 12a-l.
25

A similar observation was made for nitro or hydroxyl groups at meta-position. The photophysical
properties of the dyes were not only polarity-dependent, but also pH-dependent (for -OH
p
substituted dyes) and viscosity-dependent. Thus, at pH 10 a bathochromic shift of maximum
absorption of up to 11 nm was observed vs. that at pH 2. In glycerol the quantum yield of the
compounds increased ca. 3-fold as compared to that in water. Notably, compound 12e exhibited
quantum yield of 5.5% in glycerol.
Furthermore, dyes 12a–l demonstrated excellent photostability for 200 min irradiation; in
particular 12e (96% remaining).
When compound 12e, the most promising dye, was applied to cells it showed a distinct staining
pattern in the nucleoli and in the cytoplasm. Experiments within cells and in vitro with purified
nucleic acids showed preferential binding of the dye to rRNA in the nucleolus and the cytoplasm.
Moreover, we could not detect any obvious binding to mRNA in vitro. mRNA resides in the
cytoplasm and in the nucleoplasm. Since the nucleoplasm remained unstained as were
cytoplasmic stress granules that contain mRNAs, we could conclude that the dye does not bind to
mRNA. The difference between rRNA and mRNA that might affect the differential binding of
the dye could be the different structure of these two types of RNAs. rRNA is a highly structured
molecule and indeed when its structure was disrupted at high temperature, the binding of the dye
to rRNA was lost.
Dye 12e does not stain nuclear dsDNA. We found that dye 12e did not bind to DNA with
histones and could bind only to DNA extract in which the histones were removed.
Further characterization of the dye showed its high photostability, and capability to be retained
within cells for long periods with a high signal-to-noise ratio without affecting cell viability.
26

Various dyes have been reported for selective nucleoli staining.42, 43 The advantages of dye 12e
over the known dyes include rapid and facile 1-step synthesis, easy purification, and rapid
nucleoli staining (150 seconds).
Conclusions
In summary, we have developed here a new far-red styryl-quinolinium dye, 12e, as a rapid and
selective stain of nucleoli in both fixed and living cells. The advantages of the promising dye
include: easy, rapid, efficient, and inexpensive synthetic access; water solubility; selective
fluorescence enhancement in the presence of rRNA; no effect on cell viability; and high photo-
stability.
Experimental
Materials and Methods
Reagents and solvents were purchased from commercial sources and were used without further
purification. The quaternary salts were prepared as previously reported in the literature by reflux
in 1,4-dioxane of the starting compounds with iodomethane.44 All reactants in moisture sensitive
reactions were dried overnight in a vacuum oven. Progress of reactions was monitored by TLC
on precoated Merck silica gel plates (60F-254). Visualization was accomplished by UV light.
Compounds were characterized by nuclear magnetic resonance using Bruker, DPX-300 and
1
DMX-600 spectrometers. H, 13C, DEPT, 19F NMR spectra were measured. Compounds were
analyzed under ESI (electron spray ionization) conditions on a Q-TOF micro-instrument
(Waters, UK). Absorption spectra were measured on a UV-2401PC UV-VIS recording
spectrophotometer (Shimadzu, Kyoto, Japan). Emission spectra were measured using Cary
27

Eclipse Fluorescence Spectrophotometer. The fluorescence quantum yields of the dyes were
determined relative to rhodamine B in ethanol at 25 °C. The quantum yield was calculated
according to the following equation:
ΦF
Φ and ΦR are the fluorescence quantum yield of the sample and the reference, respectively,
= ΦR I ⁄ IR * ODR ⁄
OD * η2 ⁄ ηR
Here,
I and IR are areas under the fluorescence spectra of the sample and of the reference, respectively,
OD and ODR are the absorption values of the sample and the reference at the excitation
wavelength, and
and the reference. The photostability of all the dyes was tested by exposure to a light source of a
Cary Eclipse Fluorescence Spectrophotometer, with the exclusion of ambient light. 2 M Dye
η and ηR are the refractive index for the respective solvents used for the sample
µ
solutions in PBS were irradiated for 200 min, with an assessment of their fluorescence intensity
at five-minute intervals
General synthetic procedure
To a 10 mL microwave reaction vessel equipped with a magnetic stirring bar were added 1,4-
dimethyl quinolinium iodide (0.5 mmol), absolute ethanol (5 mL), piperidine (cat. amount), and
substituted benzaldehyde (0.5 mmol). The vessel was sealed, and the mixture was irradiated in a
microwave oven (CEM Focused Microwave type Discover) for 5 to 30 min at 70–80 °C (as
indicated in Table 1). Then the reaction mixture was cooled to RT, and diethyl ether (10 mL)
was added to precipitate the products as black crystals. The latter were vacuum-filtered and
washed 3x3 mL with diethyl ether to yield pure products. Products 12a-l were >95% pure as
determined by 1H NMR, 13C NMR, 19F NMR, DEPT, UV and HRMS.
28

Selected Data
3,5-Difluoro-4-hydroxybenzaldehyde. 1H-NMR (400 MHz, CDCl3): 9.82 (t, J = 1.8 Hz, 1H),
7.49 (d, J = 6.4 Hz, 2H). 13C NMR (ppm): 190.93, 152.32, 137.15, 132.74, 111.53. 19F NMR
(376 MHz, CDCl3)
158.03024.
δ -135.14 ppm. HRMS Calcd for C7H4F2O2 m/z 158.03035, Found
2,3,5,6-Tetrafluoro-4-hydroxybenzaldehyde. 1H NMR (DMSO, 400 MHz):
3.36 (bs, 1H). 13C NMR (DMSO 75.45 MHz):
δ 10.22 (s, 1 H),
δ
91.5, 98.7, 136.1, 139.5, 143.2, 146.5, 182.3.
Calcd for C7H2F4O2 m/z 194.09183, Found 193.09189.
1
(E)-2-(4-hydroxystyryl)-1,4-dimethylpyridinium iodide (12a). H-NMR (400 MHz, DMSO
(d6)): 8.97 (d, 5.89 Hz, 1H, Ar), 8.68 (d, 8.5 Hz, 1H, Ar), 7.93 (d, 16.1 Hz, 1H, CH=CH), 7.91
(d, 16.1 Hz, 1H, CH=CH), 7.8 (d, 7.8 Hz, 1H, Ar), 7.66 (dd, 8.5, 7.8 Hz, 1H, Ar), 7.57 (d, 5.8
Hz, 1H, Ar), 7.31 (d, 8.72 Hz, 2H, Ar), 7.16 (dd, 8.5, 7.8 Hz, 1H, Ar), 6.95 (d, 8.72 Hz, 2H, Ar),
+
13
4.31 (s, 3H, NCH3). C NMR (ppm): 157.71, 157.32, 142.85, 132.54, 130.83, 130.52, 129.98,
129.54, 128.14, 126.91, 126.71, 125.02, 123.82, 119.87, 116.16, 115.95, 46.92. HRMS Calcd for
C18H16NOI m/z 389.23025, Found 389.23024.
1
(E)-2-(4-Methoxystyryl)-1,4-dimethylpyridinium iodide (12b). H-NMR (400 MHz, DMSO
(d6)): 8.96 (d, 5.89 Hz, 1H, Ar), 8.68 (d, 8.5 Hz, 1H, Ar), 7.93 (d, 16.1 Hz, 1H, CH=CH), 7.91
(d, 16.1 Hz, 1H, CH=CH), 7.8 (d, 7.8 Hz, 1H, Ar), 7.66 (dd, 8.5, 7.8 Hz, 1H, Ar), 7.57 (d, 5.8
Hz, 1H, Ar), 7.31 (d, 8.72 Hz, 2H, Ar), 7.16 (dd, 8.5, 7.8 Hz, 1H, Ar), 6.95 (d, 8.72 Hz, 2H, Ar),
+
4.31 (s, 3H, NCH3), 3.92 (s, 3H, OCH3). 13C NMR (ppm): 157.71, 157.32, 142.85, 132.54,
29