New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on long-wavelength 7-hydroxycoumarin scaffolds

Article abstract of DOI:10.1016/j.dyepig.2014.02.004

The synthesis and photophysical properties of novel water-soluble phenol-based fluorophores derived from 3-benzothiazolyl-7-hydroxycoumarin and emitting in the range 485-631 nm are described. Further conversion into thiol-sensitive fluorogenic probes thro

Full text of DOI:10.1016/j.dyepig.2014.02.004

Dyes and Pigments xxx (2014) 1e15
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
New insights into the water-solubilization of thiol-sensitive
fluorogenic probes based on long-wavelength 7-hydroxycoumarin
scaffolds
,c,
Benoît Roubinet a, Pierre-Yves Renard a, Anthony Romieu b
*
a Normandie Université, COBRA UMR 6014 & FR 3038; UNIV Rouen; INSA Rouen; CNRS, IRCOF, 1, Rue Tesnières, 76821 Mont-Saint-Aignan Cedex, France
b Institut de Chimie Moléculaire de l’Université de Bourgogne, UMR CNRS 6302, Université de Bourgogne, 9, Avenue Alain Savary, 21078 Dijon, France
c Institut Universitaire de France, 103 Boulevard Saint-Michel, 75005 Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and photophysical properties of novel water-soluble phenol-based fluorophores derived
from 3-benzothiazolyl-7-hydroxycoumarin and emitting in the range 485e631 nm are described. Further
conversion into thiol-sensitive fluorogenic probes through the chemical modification of their hydroxyl
group was next investigated. Depending on the type of thiol-reactive quenching moiety used (2,4-
dinitrobenzenesulfonyl ester, 2,4-dinitrophenyl ether or benzoquinone-type Michael acceptors) and
the water-solubilizing group(s) pre-introduced into the coumarin core, dramatic differences in the thiol-
induced fluorescence activation of these pro-fluorophores under physiological conditions were observed.
Results for this comparative study provide valuable informations for the selection of the most suitable
structural features for designing 7-hydroxycoumarin-based long-wavelength fluorescent probes for thiol
bioimaging.
Received 26 December 2013
Received in revised form
1 February 2014
Accepted 3 February 2014
Available online xxx
Keywords:
7-Hydroxycoumarins
Fluorogenic probe
Pro-fluorescence
Red-shift
Ó 2014 Elsevier Ltd. All rights reserved.
Thiols
Water-solubility
1. Introduction
monitoring and disease diagnostic assays. Consequently, during the
past decade, tremendous research efforts have been devoted to the
Thiols are important molecules in the environment and in bio-
logical processes. Cysteine (Cys), homocysteine (HCys), glutathione
(GSH) and the gasotransmitter hydrogen sulfide (H2S) play crucial
roles in a wide range of physiological and pathological processes
arising from their biological redox chemistry [1]. In contrast, aro-
matic thiols are versatile chemical intermediates currently used to
produce pesticides, polymers and pharmaceuticals [2], identified as
polluting compounds and highly toxic for human health causing
serious damages to the central nervous system and related injuries
[3]. Therefore, the design of reaction-based probes or related che-
mosensors for selective and quantitative detection of thiols by
simple spectroanalysis in complex environmental matrices or bio-
logical samples has been the focus of increasing attention. Owing to
their unique advantages, such as high sensitivity and operational
simplicity, thiol-sensitive fluorogenic probes are valuable (bio)
analytical tools for some applications in environmental pollution
development of “smart” reaction-based strategies for fluorescence
sensing and bioimaging of thiols [4e7]. Most of them are based
either on reductive cleavage reactions or nucleophilic reactions
(Michael addition and SNAr) and tandem processes, often imple-
mented on an aniline- or a phenol-based fluorophore whose
reversible chemical modification of amino/hydroxyl group causes
dramatic changes in its spectral properties [8e20]. Current im-
provements to these pro-fluorophores aim at developing pro-
fluorescent probes that can either (1) discriminate between
benzenethiols and aliphatic thiols or differentiate between physi-
ological thiols (e.g., effective discrimination of cysteine from ho-
mocysteine) [8,10,14,21] and/or (2) improve and facilitate thiol
detection in complex biological contexts (i.e., in cellulo or in vivo) by
red-shifting the spectral features of the released fluorescent aniline
or phenol [22,23]. Surprisingly, to the best of our knowledge, no
studies specifically focused on the optimization of physicochemical
properties of thiol-sensitive fluorogenic probes have been reported
to date. Since factors such as water-solubility and net electric
charge are known to strongly influence their cell permeability and
emission efficiency under physiological conditions of their
unmasked fluorescent label (directly related to their resistance to
* Corresponding author. Institut de Chimie Moléculaire de l’Université de Bour-
gogne, UMR CNRS 6302, Université de Bourgogne, 9, Avenue Alain Savary, 21078
Dijon, France. Tel.: þ33 3 80 39 36 24; fax: þ33 3 80 39 61 17.
E-mail address: anthony.romieu@u-bourgogne.fr (A. Romieu).
http://dx.doi.org/10.1016/j.dyepig.2014.02.004
0143-7208/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

2
B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
aggregation in aq. media), these are key parameters to be consid-
ered in the rational design of thiol-imaging agents [24]. Further-
more, structure and specific position of water-solubilizing moieties
onto the pro-fluorophore may complicate its synthesis (especially
the introduction of the selected thiol-reactive quenching moiety)
and affect its reactivity towards thiols under physiological condi-
tions, mainly due to adverse electrostatic and/or steric effects. In
this context, we have decided to make a comparative evaluation of
different water-solubilizing methodologies, implemented to a
family of thiol-sensitive pro-fluorophores derived from 3-
benzothiazolyl-7-hydroxycoumarin and exhibiting distinct red-
shifted emission in the range 485e631 nm, aimed at assessing
their effects on the thiol-mediated probes’ activation and on the
fluorescence efficiency of the released phenols. The ultimate goal of
the present work is to identify the best pair of molecular candidates
acting as water-solubilizing and thiol-reactive quenching moieties
respectively, to convert a specific hydrophobic long-wavelength 7-
chromatography system (P4000) equipped with a UVeVis 2000
detector. Automated flash purifications on RP-C18 cartridges were
performed with a Biotage IsoleraÔ One (ISO-1EW) system. Ion-
exchange chromatography (for desalting fluorophores purified
with TEAB as aq. mobile phase) was performed with an Econo-PacÒ
disposable chromatography column (Bio-Rad, #732e1010) filled
with an aq. solution of DowexÒ 50WX8-400 (Alfa Aesar, w5 g for
15 mg of dye, 15 ꢁ 50 mm bed), regenerated using aq. 10% HCl
solution and equilibrated with deionized water. Low-resolution
mass spectra were obtained with a Finnigan LCQ Advantage MAX
(ion trap) apparatus equipped with an electrospray source. UVe
visible absorption spectra were obtained on a Varian Cary 50 scan
spectrophotometer by using a rectangular quartz cell (Varian,
standard cell, Open Top, 10 ꢁ 10 mm, 3.5 mL). Fluorescence spec-
troscopic studies (emission/excitation spectra) were performed on
a Varian Cary Eclipse spectrophotometer with a semi-micro quartz
fluorescence cell (Hellma, 104F-QS, 10 ꢁ 4 mm, 1400
mL). The ab-
hydroxycoumarin derivative into
a
thiol-sensitive fluorogenic
sorption spectra of 7-hydroxycoumarin derivatives were recorded
(220e800 nm) in PB (100 mM, pH 7.5) at 25 ꢂC. Excitation/emission
spectra were recorded under the same conditions after emission/
excitation at the corresponding wavelength (390/470/510/550 nm,
excitation and emission filters: auto, excitation and emission slit:
5 nm). Fluorescence emission spectra of far-red emitting 7-
hydroxycoumarin-hemicyanine dyes (10, 11 and 13) were cor-
rected. Fluorescence quantum yields were measured at 25 ꢂC by a
relative method using 7-hydroxycoumarin (FF ¼ 76% in PB,
pH ¼ 7.4) [38], sulforhodamine 101 (SR101, FF ¼ 95% in EtOH),
fluorescein (Fluo, FF ¼ 91% in 0.1 N NaOH) or cresyl violet (CV,
FF ¼ 56% in EtOH) as a standard [39]. The following equation was
used to determine the relative fluorescence quantum yield:
probe fulfilling all requirements for the targeted biosensing
application.
2. Experimental
2.1. Chemicals and instruments
Flash column chromatography purifications were performed on
GeduranÒ Si 60 silica gel (40e63
m
m) from Merck. TLC were carried
out on Merck DC Kieselgel 60 F-254 aluminium sheets. The spots
were visualized by illumination with a UV lamp (
¼ 254/365 nm)
l
and/or staining with KMnO4 solution. Unless otherwise noted, all
chemicals were used as received from commercial sources without
further purification. All solvents were dried by standard procedures
(CH2Cl2: distillation over P2O5; pyridine: distillation over CaH2 and
stored over BaO; CH3CN: distillation over CaH2; absolute EtOH:
storage over anhydrous Na2SO4 and triethylamine (TEA): distillation
over KOH and storage over BaO). Peptide synthesis-grade NMP and
FFðxÞ ¼ ðAS=AXÞðFX=FSÞðnX=nSÞ2FFðsÞ
where A is the absorbance (in the range of 0.01e0.1 A.U.), F is the
area under the emission curve, n is the refractive index of the sol-
vents (at 25 ꢂC) used in measurements, and the subscripts s and x
represent standard and unknown, respectively. The following
refractive index values were used: 1.479 for DMSO, 1.362 for EtOH
and 1.337 for PB and PB þ 5% BSA.
ꢀ
anhydrous DMF were purchased from Carlo Erba and stored over 4A
molecular sieves. Peptide synthesis-grade N,N-diisopropylethyl-
amine (DIEA) was provided by Iris Biotech GmbH. HPLC gradient-
grade acetonitrile (CH3CN) was obtained from VWR. Phosphate
buffer (PB, 100 mM, pH 7.5) and aq. mobile phases for HPLC were
prepared with water purified by means of a MilliQ system (purified
Several chromatographic systems were used for the analytical
experiments and purification steps (by semi-preparative HPLC or
automated flash purification system): System A: RP-HPLC (Thermo
to 18.2 M
U
cm). Triethylammonium acetate (TEAA, 2.0 M) and
Hypersil GOLD C18 column, 5
mm, 2.1 ꢁ 100 mm) with CH3CN and
triethylammonium bicarbonate (TEAB,1.0 M) buffers were prepared
from distilled TEA and glacial acetic acid or CO2 gas, respectively. 3-
Benzothiazolyl-7-hydroxycoumarin (hydrochloride salt) and its 4-
cyano derivative, 3-benzothiazolyl-7-hydroxycoumarin-6-sulfonic
acid, di-tert-butyl iminodiacetate A, 2-aminoethane-1,1-disulfonic
acid (tetrabutylammonium salt, TBAþ salt) B, N-sulfopropyl-2-
methylbenzothiazole C, N-sulfopropyl-4-methylpyridine D,
1-chloromethyl-2,5-dimethoxy-3,4,6-trimethylbenzene E, amino-
trimethyl lock linker-functionalized quinone F and sodium thio-
phosphate (NaThioPi) were prepared according to literature
procedures [25e36].
0.1% trifluoroacetic acid (aq. TFA 0.1%, pH 2.0) as eluents [100% TFA
(5 min) then linear gradient from 0% to 100% (45 min) of CH3CN] at
a flow rate of 0.25 mL/min. Triple UVevis detection was achieved at
220, 260, and 380 nm and with the “Max Plot” (i.e., chromatogram
at absorbance maximum for each compound) mode (220e650 nm).
System B: system A with the following gradient [80% TFA (5 min)
then linear gradient from 20% to 100% (45 min) of CH3CN]. System
C: system A with TEAA buffer (25 mM, pH 7.0) as aq. mobile phase
[100% TEAA (5 min) then linear gradient from 0% to 100% (45 min)
of CH3CN]. System D: semi-preparative RP-HPLC (Varian Kromasil
C18 column, 10
m
m, 21.2 ꢁ 250 mm) with CH3CN and aq. TFA 0.1% as
1H and 13C spectra were recorded with either a Bruker DPX 300
or a Bruker Avance III 500 spectrometer (Bruker, Wissembourg,
France). Chemical shifts are expressed in parts per million (ppm)
using the residual solvent peak for calibration [37]. J values are
expressed in Hz. Infrared (IR) spectra were recorded with
a universal ATR sampling accessory on a Perkin Elmer FT-IR Spec-
trum 100 spectrometer. The bond vibration frequencies are
expressed in reciprocal centimetres (cmꢀ1). Analytical HPLC was
eluents [100% TFA (5 min) then linear gradient from 0% to 50%
(100 min) of CH3CN] at a flow rate of 20.0 mL/min. Visible detection
was achieved at 420 nm. System E: automated flash purification
(BiotageÒ SNAP cartridge KP-C18-HS, 60 g) with CH3CN and ultra-
pure water as eluents [100% H2O (5 min) then linear gradient from
0% to 100% (40 min) of CH3CN] at a flow rate of 35.0 mL/min. Dual
UV detection was achieved at 220 and 360 nm; System F: system E
with CH3CN and aq. TFA 0.1% as eluents [100% TFA (5 min) then
linear gradient from 0% to 10% (10 min) and 10%e60% (60 min) of
CH3CN] at a flow rate of 35.0 mL/min. System G: system D with the
following gradient [100% TFA 0.1% (5 min) then linear gradient from
performed on
equipped with a PDA detector. Semi-preparative HPLC was per-
formed on Thermo Scientific SPECTRASYSTEM liquid
a thermo Scientific Surveyor Plus instrument
a
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
3
0% to 15% (10 min) and 15%e55% (65 min) of CH3CN]. UV detection
was achieved at 360 nm. System H: system G with aq. TEAB
(50 mM, pH 7.5) as aq. mobile phase. System I: system G with the
following gradient [100% TEAB (5 min) then linear gradient from 0%
to 10% (10 min) and 10%e50% (45 min) of CH3CN]. System J: system
F with the following cartridge (BiotageÒ SNAP cartridge KP-C18-HS,
60 g) and gradient [100% TFA (5 min) then linear gradient from 0%
to 20% (10 min) and 20%e50% (45 min) of CH3CN] at a flow rate of
35.0 mL/min. System K: system D with the following gradient [100%
TFA (5 min) then linear gradient from 0% to 30% (10 min) and 30%e
60% (55 min) of CH3CN]. UV detection was achieved at 360 nm.
System L: system K with the following gradient [100% TFA (5 min)
then linear gradient from 0% to 30% (10 min) and 30%e70% (65 min)
of CH3CN]. System M: system K with the following gradient [100%
TFA (5 min) then linear gradient from 0% to 25% (15 min) and 25%e
70% (70 min) of CH3CN]. System N: system K with the following
gradient [100% TFA (5 min) then linear gradient from 0% to 25%
(10 min) and 25%e60% (55 min) of CH3CN]. System O: system K
with the following gradient [100% TFA 0.1% (5 min) then linear
gradient from 0% to 20% (10 min) and 20%e70% (65 min) of CH3CN].
System P: system H with the following gradient [100% TEAB (5 min)
then linear gradient from 0% to 15% (10 min) and 15%e45% (40 min)
of CH3CN]. Visible detection at 400 nm. System Q: system G with
the following gradient [100% TFA (5 min) then linear gradient from
0% to 20% (10 min) and 20%e60% (65 min) of CH3CN].
residue was subjected to an automated flash purification on a RP-
C18 cartridge (system E). The product-containing fractions were
lyophilized to give compound 3a as an amorphous yellow solid
(80 mg, yield 24%). 1H NMR (300 MHz, CDCl3):
d 8.95 (s, 1H), 8.03
(d,1H, J ¼ 6.0 Hz), 7.93 (d, 1H, J ¼ 9.0 Hz), 7.51 (m, 2H), 7.36 (t, 1H,
J ¼ 9.0 Hz), 6.91 (d, 1H, J ¼ 6.0 Hz), 4.26 (s, 2H), 3.44 (s, 4H), 1.50 (s,
18H). 13C NMR (75 MHz, CDCl3):
d 169.9, 164.0, 160.8, 160.0, 153.8,
152.6,142.5,136.6,130.3,126.4,125.0,122.6,121.7,115.6,155.5,111.7,
108.5, 82.6, 55.5, 48.1, 28.2. MS (ESI, positive mode): m/z ¼ 552.93
[M þ H]þ, calcd for C29H32N2O7S 552.19.
2.2.3. Iminodiacetic acid derivative (3b)
Di-tert-butyl ester 3a (59 mg, 0.11 mmol, 1 equiv) was dissolved
in a mixture of TFA/CH2Cl2 (1:1, v/v, 20 mL) and the resulting re-
action mixture was stirred at reflux for 2 h. Thereafter, the depro-
tection mixture was concentrated to dryness and triturated with
Et2O. The solid was recovered and dried under vacuum to give the
TFA salt of compound 3b as a yellow amorphous solid (40 mg, yield
66%). IR:
NMR (300 MHz, DMSO-d6):
n
1720, 1606, 1573, 1431, 1384, 1298, 1196, 1087, 1002. 1H
d
12.24 (s, 2H, 2 ꢁ CO2H), 9.16 (s, 1H),
8.16 (d, 1H, J ¼ 6.0 Hz), 8.05 (d, 1H, J ¼ 9.0 Hz), 7.89 (d, 1H,
J ¼ 9.0 Hz), 7.57 (t, 1H, J ¼ 6.0 Hz), 7.52 (t, 1H, J ¼ 6.0 Hz), 6.93 (d, 1H,
J ¼ 9.0 Hz), 4.17 (s, 2H), 3.53 (s, 4H). 13C NMR (75 MHz, DMSO-d6):
d
172.4, 163.1, 160.4, 159.6, 153.8, 152.0, 143.0, 135.6, 130.9, 126.5,
125.0, 122.2, 122.2, 114.50, 114.3, 111.4, 109.4, 53.8, 46.7. MS (ESI,
negative mode): m/z ¼ 439.00 [M ꢀ H]ꢀ, calcd for C21H16N2O7S
440.07. HPLC (system A): tR ¼ 24.8 min, purity ¼ 97%.
For the detailed synthetic procedures of compounds 4a, 4b, 6, 8,
17a and 17b, see supplementary material.
Most of 7-hydroxycoumarins and related fluorogenic probes
were found to be soluble in D2O but bad quality spectra were ob-
tained (i.e., broad and poor-resolved peaks). Thus, all NMR spectra
of water-soluble derivatives were recorded in DMSO-d6.
2.2.4. Sarcosine derivative (5)
Paraformaldehyde (65 mg, 2.18 mmol, 7.2 equiv) was dissolved
in dry EtOH (2 mL) and solid KOH (122 mg, 2.18 mmol, 7.2 equiv)
was added at rt and the resulting solution was stirred for 5 min.
Then, sarcosine tert-butyl ester hydrochloride (184 mg, 1.09 mmol,
3.6 equiv) was added and the mixture was stirred at reflux for 3 h.
The iminium salt formed in situ was slowly added to a solution of 3-
benzothiazolyl-7-hydroxycoumarin (100 mg, 0.3 mmol, 1 equiv) in
dry EtOH (5 mL) and the resulting reaction mixture was stirred
under reflux for 4 h. Premature cleavage of tert-butyl ester occurred
during this Mannich-type reaction and the free sarcosinee
coumarin conjugate was isolated as the major product as follows:
after removal of volatiles under reduced pressure, the crude residue
was subjected to an automated flash purification on a RP-C18 car-
tridge (system F). The product-containing fractions were lyophi-
lized to give the TFA salt of compound 5 as an amorphous yellow
2.2. Synthesis of water-soluble 3-benzothiazolyl-7-
hydroxycoumarin derivatives
2.2.1. 4-Cyano 6-sulfonated derivative (2)
To
sulfonic acid 1 (0.1 g, 0.24 mmol, 1 equiv) in DMF (2 mL), an
aqueous solution of KCN (250 L, 35 mg, 0.54 mmol, 2.3 equiv) was
a
solution of 3-benzothiazolyl-7-hydroxycoumarin-6-
m
added and the resulting reaction mixture was stirred at 50 ꢂC for
3 h. Thereafter, the mixture was cooled to 0 ꢂC, DDQ (61 mg,
0.27 mmol, 1.1 equiv) was added and the mixture stirred at room
temperature for a further 2 h. The crude solution was concentrated
and the resulting residue was purified by semi-preparative RP-
HPLC (System D). The product-containing fractions were lyophi-
lized to give the TFA salt of compound 2 as an amorphous red solid
solid (73 mg, yield 48%). IR:
n
1696, 1635, 1570, 1407, 1291, 1255,
9.21 (s, 1H), 8.18
1191, 1130, 1027. 1H NMR (300 MHz, DMSO-d6):
d
(56 mg, yield 45%). IR:
n
2198 (CN), 1725, 1606, 1538, 1351, 1228,
8.23 (d, 1H,
(d, 1H, J ¼ 6.0 Hz), 8.03 (m, 2H), 7.56 (t, 1H, J ¼ 9.0 Hz), 7.46 (t, 1H,
1172, 1084, 1020. 1H NMR (300 MHz, DMSO-d6):
d
J ¼ 9.0 Hz), 7.05 (d, 1H, J ¼ 6.0 Hz), 4.46 (s, 2H), 3.98 (s, 2H), 2.72 (s,
J ¼ 9.0 Hz), 8.15 (d,1H, J ¼ 6.0 Hz), 8.11 (s,1H), 7.63 (t,1H, J ¼ 9.0 Hz),
3H). 13C NMR (75 MHz, DMSO-d6):
d 168.2,163.3,160.2,159.2,154.8,
7.56 (t, 1H, J ¼ 9.0 Hz), 7.04 (s,1H). 13C NMR (75 MHz, DMSO-d6):
152.0, 142.8, 135.6, 132.9, 126.6, 125.1, 122.2, 122.2, 114.7, 113.9, 11.5,
104.1, 55.5, 47.8, 41.1. MS (ESI, positive mode): m/z ¼ 396.87
[M þ H]þ, calcd for C20H16N2O5S 396.08. HPLC (system A):
tR ¼ 24.5 min, purity ¼ 95%.
d
159.3, 158.7, 157.5, 154.3, 151.4, 136.5, 130.4, 127.1, 126.5, 125.9,
123.3, 122.2, 121.4, 120.8, 113.9, 109.8, 103.8. MS (ESI, negative
mode): m/z ¼ 399.07 [M ꢀ H]-, calcd for C17H8N2O6S2 399.98. HPLC
(system A): tR ¼ 26.3 min, purity ¼ 95%.
2.2.5. Di-sulfonated sarcosine derivative (7)
2.2.2. Di-tert-butyl ester of 8-aminomethyl derivative (3a)
To a solution of 7-hydroxycoumarin carboxylic acid 5 (20 mg,
Paraformaldehyde (132 mg, 4.1 mmol, 6.8 equiv) was dissolved
in dry EtOH (2 mL) and solid KOH (17 mg, 0.30 mmol, 0.5 equiv) was
added at 0 ꢂC and the resulting solution was stirred for 5 min. Then,
di-tert-butyl iminodiacetate A (530 mg, 2.0 mmol, 3.4 equiv) was
added and the mixture was stirred at rt for 1 h. The iminium salt
formed in situ was slowly added to a solution of 3-benzothiazolyl-
7-hydroxycoumarin (200 mg, 0.6 mmol,1 equiv) in dry EtOH (5 mL)
and the resulting reaction mixture was stirred under reflux for 24 h.
Volatiles were removed under reduced pressure and the resulting
39.3
NMP (200
DIEA in NMP (56
m
mol, 1 equiv) in NMP (200
L, 35 mg, 75.6 mol, 1.5 equiv) and a 2.0 M solution of
L, 111 mol, 2.2 equiv) were sequentially added
mL), a 0.38 M solution of PyBrOP in
m
m
m
m
and the resulting reaction mixture was stirred at rt for 30 min.
Then, this crude acyl bromide coumarin derivative was added
dropwise to a pre-cooled (0 ꢂC) 0.5 M solution of 2-aminoethane-
1,1-disulfonic acid B (TBAþ salt) in NMP (0.3 mL, 150
m
mol, 3 equiv)
L of a 2.0 M solution in NMP), over
a period of 15 min. The resulting reaction mixture was stirred at rt
containing 5 equiv of DIEA (125
m
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

4
B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
for 2 h. Then, a further amount of PyBrOP (23 mg, 50.4
75 L of NMP) was added and the mixture was stirred for a further
3 h. This amidification reaction was checked for completion by RP-
HPLC (system A), quenched by adding acetic acid (50 L) and
m
mol, in
2.2.8. 7-Hydroxycoumarin-hemicyanine dye (11)
To a solution of aldehyde 9 (30 mg, 93 mol,1 equiv) in dry EtOH
(3 mL), N-sulfopropyl-4-methylpyridine D (60 mg, 279 mol,
3 equiv) and pyrrolidine (24 L, 279 mol, 3 equiv) were sequen-
m
m
m
m
m
m
dilution with aq. TFA-CH3CN (4:1, v:v, 5 mL) and finally purified by
RP-HPLC (system G). The product-containing fractions were
lyophilized to give the TFA salt of compound 7 in mixture with
TBAþ salts. Desalting by ion-exchange chromatography (followed
by lyophilization) afforded compound 7 as a yellow amorphous
tially added and the resulting reaction mixture was stirred at rt
under an Ar atmosphere for 2 h. The reaction was checked for
completion by RP-HPLC (system C), the mixture was concentrated
and the resulting residue was purified by semi-preparative RP-
HPLC (system P). The product-containing fractions were thrice
lyophilized to give the TEA salt of 11 as a red amorphous solid
powder (7 mg, yield 31%). IR:
n
1730, 1680, 1607, 1570, 1498, 1250,
11.99 (s, 1H), 9.82
1193, 1063, 1016. 1H NMR (300 MHz, DMSO-d6):
d
(25 mg, yield 44%). IR:
n
1720, 1641, 1603, 1565, 1499, 1328, 1185,
8.98 (s, 1H), 8.84 (d,
(s, 1H), 9.17 (s, 1H), 8.16 (d, 1H, J ¼ 6.0 Hz), 8.09 (s, 1H), 8.03 (m, 2H),
7.54 (t, 1H, J ¼ 9.0 Hz), 7.44 (t, 1H, J ¼ 9.0 Hz), 7.06 (d, 1H, J ¼ 6.0 Hz),
4.53 (s, 2H), 4.10 (s, 2H), 3.75 (s, 2H), 3.58 (t, 1H, J ¼ 6.0 Hz), 2.81 (s,
1037. 1H NMR (300 MHz, DMSO-d6):
d
2H, J ¼ 9.0 Hz), 8.13 (m, 3H), 8.05 (d, 2H, J ¼ 9.0 Hz), 7.98 (d, 1H,
J ¼ 9.0 Hz), 7.74 (d, 1H, J ¼ 9.0 Hz), 7.51 (t, 1H, J ¼ 9.0 Hz), 7.40 (t, 1H,
J ¼ 9.0 Hz), 6.76 (d,1H, J ¼ 9.0 Hz), 4.62 (t, 2H, J ¼ 9.0 Hz), 3.10 (q, 2H,
1 ꢁ NeCH2eCH3, 0.33 ꢁ TEA), 2.50 (m, 2H, masked by DMSO
3H). 13C NMR (75 MHz, DMSO-d6):
d 163.4, 162.8, 160.2, 159.1, 155.0,
152.0, 142.6, 135.7, 133.3,126.6,125.1, 122.3,115.0, 113.8,111.6, 103.2,
74.2, 56.3, 47.4, 41.2, 38.6. MS (ESI, negative mode): m/z ¼ 581.67
[M ꢀ H]ꢀ, calcd for C22H21N3O10S3 583.04. HPLC (system A):
tR ¼ 22.6 min, purity ¼ 98%.
signal), 2.23 (m, 2H) 1.19 (t, 3H, 1 ꢁ NeCH2eCH3, 0.33 ꢁ TEA). 13
C
NMR (126 MHz, DMSO-d6): d 162.0, 160.3, 157.5, 155.2, 152.4, 143.6,
142.3, 135.1, 133.2, 132.1, 126.1, 124.0, 122.5, 121.9, 121.3, 109.8, 57.9,
47.1, 45.8 (NeCH2eCH3, TEA), 27.3, 8.73 (NeCH2eCH3, TEA), five
carbons are missing despite long acquisition time on a 500 MHz
spectrometer. MS (ESI, negative mode): m/z ¼ 519.00 [M ꢀ H]-,
calcd for C26H20N2O6S2 520.08. HPLC (system B): tR ¼ 25.7 min,
purity ¼ 98%.
2.2.6. 3-Benzothiazolyl-8-formyl-7-hydroxycoumarin (9)
To a solution of 3-benzothiazolyl-7-hydroxycoumarin (500 g,
1.5 mmol, 1 equiv) in TFA (15 mL), hexamine (474 mg, 3.40 mmol,
2.3 equiv) was added and the resulting reaction mixture was stirred
at reflux for 24 h. Then, aq. 1.0 M HCl (30 mL) was added and the
mixture was stirred for a further 3 h. Thereafter, this aq. mixture
was extracted with ethyl acetate (EtOAc) and the combined organic
phases were washed with deionized water, dried over anhydrous
Na2SO4 and finally evaporated to dryness. The resulting residue was
purified by flash column chromatography on silica gel (petroleum
ether/EtOAc with a step gradient from 1:0 to 3:2) in order to give
2.2.9. 3-Benzothiazolyl-8-formyl-7-hydroxycoumarin 6-sulfonic
acid (12)
To a suspension of 3-benzothiazolyl-7-hydroxycoumarin 6-
sulfonic acid 1 (100 mg, 0.24 mmol, 1 equiv) in CH2Cl2 (50 mL),
5 mL of an aq. solution of tetrabutylammonium hydroxide (40%,
1.5 M) was added and the resulting mixture was stirred at rt for
5 min. TBAþ salt of 1 was extracted with CH2Cl2 and the resulting
organic phase was dried over anhydrous Na2SO4 and evaporated to
dryness. The TBAþ salt of 1 was next dissolved in TFA (2 mL),
hexamine (74 mg, 0.54 mmol, 2.2 equiv) was added and the
resulting reaction mixture was stirred at reflux for 48 h. Thereafter,
aq. 1.0 M HCl (30 mL) was added and the mixture was stirred for a
further 3 h. Then, the mixture was concentrated to dryness and the
crude product was subjected to an automated flash purification on
a RP-C18 cartridge (system J). The product-containing fractions
were lyophilized to give the TFA salt of compound 12 (in mixture
with TBAþ salts) as a yellow amorphous solid (32 mg). For spec-
troscopic characterizations, a sample (17 mg) was desalted by ion-
exchange chromatography (followed by lyophilization) to afford
aldehyde 9 as a yellow solid (280 mg, yield 58%). IR:
1586, 1473, 1313, 1290, 1188, 1166, 1069, 1008. 1H NMR (300 MHz,
DMSO-d6):
10.48 (s, 1H), 9.19 (s, 1H), 8.18 (m, 2H, J ¼ 9.0 Hz), 8.04
n 1722, 1658,
d
(d, 1H, J ¼ 9.0 Hz), 7.56 (t, 1H, J ¼ 9.0 Hz), 7.47 (t, 1H, J ¼ 9.0 Hz), 7.07
(d, 1H, J ¼ 9.0 Hz). 13C NMR (75 MHz, DMSO-d6):
d 190.0, 165.6,
159.8, 158.8, 155.5, 152.0, 142.4, 137.7, 135.7, 126.7, 125.3, 122.3,
122.2, 115.8, 115.2, 111.4, 109.4. MS (ESI, positive mode): m/
z ¼ 321.93 [M þ H]þ, calcd for C23H9NO4S 323.03.
2.2.7. 7-Hydroxycoumarin-Hemicyanine dye (10)
To a solution of aldehyde 9 (50 mg, 0.16 mmol, 1 equiv) in dry
EtOH (2 mL), N-sulfopropyl-2-methylbenzothiazole C (84 mg,
aldehyde 12 as a yellow amorphous powder (4 mg, yield 7%). IR:
1736, 1647, 1596, 1484, 1383, 1192, 1138, 1045. 1H NMR (300 MHz,
DMSO-d6): 10.47 (s, 1H), 9.25 (s, 1H), 8.50 (s, 1H), 8.16 (d, 1H,
n
0.31 mmol, 2 equiv) and pyrrolidine (30 mL, 0.31 mmol, 2 equiv)
were sequentially added and the resulting reaction mixture was
stirred at rt under an Ar atmosphere for 2 h. The reaction was
checked for completion by RP-HPLC (system C) and purified by
semi-preparative RP-HPLC (System I). The product-containing
fractions were lyophilized to give the TEA salt of compound 10 in
mixture with TEAB salt. Desalting by ion-exchange chromatog-
raphy (followed by lyophilization) afforded compound 10 as a
d
J ¼ 9.0 Hz), 8.08 (d, 1H, J ¼ 9.0 Hz), 7.56 (t, 1H, J ¼ 9.0 Hz), 7.47 (t, 1H,
J ¼ 9.0 Hz). 13C NMR (75 MHz, DMSO-d6):
d 186.9, 160.9, 159.8,
158.9, 154.8, 151.0, 142.4, 135.8, 134.3, 130.0, 126.7, 125.3, 122.5,
122.2, 116.5, 111.1, 110.6. MS (ESI, negative mode): m/z ¼ 402.07
[M
ꢀ
H]-, calcd for C17H9NO7S2 402.98. HPLC (system A):
tR ¼ 24.5 min, purity ¼ 95%.
brown amorphous powder (46 mg, yield 50%). IR:
1557, 1495, 1330, 1190, 1149, 1102, 1033. 1H NMR (300 MHz, DMSO-
d6): 9.23 (s, 1H), 8.45 (m, 3H), 8.33 (s, 1H, OH), 8.18 (d, 1H,
n 1727, 1588,
2.2.10. 7-Hydroxycoumarin-hemicyanine dye (13)
d
To a solution of sulfonated aldehyde 12 (15 mg, 29.1 mmol,
J ¼ 9.0 Hz), 8.11 (d, 1H, J ¼ 9.0 Hz), 8.10 (m, 2H), 7.91 (t, 1H,
J ¼ 9.0 Hz), 7.82 (t, 1H, J ¼ 9.0 Hz), 7.57 (t, 1H, J ¼ 9.0 Hz), 7.47 (t, 1H,
J ¼ 9.0 Hz), 7.16 (d, 1H, J ¼ 9.0 Hz), 5.02 (m, 2H), 2.68 (m, 2H), 2.50
(m, 2H, masked by DMSO signal). Twisting and bending molecular
motions of such hemicyanine dye in solution prevent the recording of a
good quality 13C NMR spectrum even for a reasonable sample con-
centration of 20 mg/mL. MS (ESI, positive mode): m/z ¼ 577.20
1 equiv, contaminated with TBAþ salts, vide supra) in dry EtOH
(1 mL), N-sulfopropyl-2-methylbenzothiazole C (21 mg, 74.4
mmol,
2.6 equiv) and pyrrolidine (6 L, 74.4 mol, 2.6 equiv) were
m
m
sequentially added and the resulting reaction mixture was stirred
at rt under an Ar atmosphere for 2 h. The reaction was checked for
completion by RP-HPLC (system C), the mixture was concentrated
and the resulting residue was purified by semi-preparative RP-
HPLC (system I). The product-containing fractions were twice
lyophilized to give the TEA salt of compound 13 as a black
[M
þ
H]þ, calcd for C17H9NO4S 576.05. HPLC (system C):
tR ¼ 26.4 min, 99%.
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
5
amorphous solid (20 mg, overall yield for the two steps 23%). IR:
1704, 1599, 1555, 1511, 1482, 1421, 1323, 1270, 1159, 1119, 1016. 1H
NMR (300 MHz, DMSO-d6):
8.85 (s, 1H), 8.74 (d, 1H, J ¼ 15.0 Hz),
n
dryness. The resulting residue was dissolved in EtOAc, washed with
deionized water (50 mL), dried over anhydrous Na2SO4 and
concentrated. The crude product was subjected to an automated
flash purification on a BiotageÒ SNAP KP-SIL (10 g) cartridge and
using a mixture of cyclohexane/EtOAc as eluents (step gradient
from 1:0 to 3:2) to afford DNBS ester 16a as a yellow solid (30 mg,
d
8.38 (d, 1H, J ¼ 15.0 Hz), 8.27 (m, 2H), 8.10 (d, 1H, J ¼ 9.0 Hz), 7.99
(m, 2H), 7.76 (t, 1H, J ¼ 9.0 Hz), 7.65(t, 1H, J ¼ 9.0 Hz), 7.50 (t, 1H,
J ¼ 9.0 Hz), 7.38 (t, 1H, J ¼ 9.0 Hz), 4.82 (m, 2H), 3.10 (q, 12H, 6 ꢁ Ne
CH2eCH3, 2 ꢁ TEA), 2.72 (m, 2H), 2.22 (m, 2H), 1.19 (t, 18H, 6 ꢁ Ne
CH2eCH3, 2 ꢁ TEA). Twisting and bending molecular motions of such
hemicyanine dye in solution prevent the recording of a good quality
13C NMR spectrum even for a reasonable sample concentration of
20 mg/mL. MS (ESI, negative mode): m/z ¼ 654.73 [M ꢀ H]- calcd for
yield 46%). 1H NMR (300 MHz, CDCl3):
d 9.01(s, 1H), 8.71 (s, 1H),
8.62 (s, 1H), 8.08 (d, 1H, J ¼ 9.0 Hz), 7.98 (d, 1H, J ¼ 9.0 Hz), 7.69 (d,
1H, J ¼ 9.0 Hz), 7.54 (t, 1H, J ¼ 9.0 Hz), 7.42 (m, 2H), 7.26 (s, 1H), 4.12
(s, 2H), 3.46 (s, 4H), 1.38 (s, 18H). 13C NMR (75 MHz, CDCl3):
d 170.4,
159.2, 158.8, 153.7, 152.6, 151.2, 151.1, 149.1, 140.5, 137.0, 134.3, 134.2,
129.6, 127.2, 126.8, 125. 8, 123.2, 121.9, 120.7, 118.9, 118.3, 81.3, 55.9,
47.2, 28.2. MS (ESI, positive mode): m/z ¼ 805.07 [M þ Na]þ, calcd
for C35H34N4O13S2: 782.16.
C
17H9NO4S 656.01. HPLC (system C): tR ¼ 24.8 min, purity ¼ 96%.
2.3. Synthesis of water-soluble thiol-sensitive fluorogenic probes
2.3.1. 2,4-Dinitrobenzenesulfonyl ester (14)
2.3.3.2. Acid removal of tert-butyl esters. Di-tert-butyl ester 16a
To a solution of 3-benzothiazolyl-7-hydroxycoumarin (120 mg,
(46 mg, 58.8 mmol,1 equiv) was dissolved in a mixture of TFA/CH2Cl2
0.39 mmol, 1 equiv) in dry DMF (20 mL), 2,6-lutidine (140
m
L,
(1:1, v/v, 20 mL) and the resulting reaction mixture was stirred at
reflux for 2 h. Thereafter, the deprotection mixture was concen-
trated to dryness to give the TFA salt of compound 16b as a brown
1.22 mmol, 3.1 equiv) and 2,4-dinitrobenzenesulfonyl chloride
(DNBS-Cl, 163 mg, 6.10 mmol, 1.6 equiv) were sequentially added
and was the resulting reaction mixture was stirred at 120 ꢂC for
18 h. Thereafter, the mixture was evaporated to dryness. The
resulting residue was dissolved in EtOAc, washed with deionized
water (50 mL), dried over anhydrous Na2SO4 and concentrated. The
crude product was purified by flash column chromatography on
silica gel using CH2Cl2 as the eluent to afford the desired DNBS ester
amorphous solid (39 mg, yield 85%). IR:
n 1733, 1602, 1542 (NO2),
1477, 1348 (NO2), 1185, 1067, 1000. 1H NMR (300 MHz, acetone-d6):
d
9.50 (s, 2H, 2 ꢁ CO2H), 9.16 (s, 1H), 9.01(s, 1H), 8.80 (d, 1H,
J ¼ 9.0 Hz), 8.66 (d,1H, J ¼ 9.0 Hz), 8.10 (m, 3H), 7.58 (t,1H, J ¼ 9.0 Hz),
7.47 (m, 2H), 4.33 (s, 2H), 3.73 (s, 4H). 13C NMR (75 MHz, acetone-
d6): d 172.3, 160.2,159.4,154.6,153.5,152.8,152.0,149.7, 141.8,137.6,
14 as a yellow solid (23.5 mg, yield 12%). IR:
(NO2), 1405, 1388, 1347 (NO2), 1119, 1111. 1H NMR (300 MHz, DMSO-
d6):
n
1722, 1609, 1534
134.9,133.6,131.8,128.7,127.5,126.5,123.8,122.0,121.2,120.1,119.6,
119.6, 114.3, 55.9, 47.9. MS (ESI, negative mode): m/z ¼ 668.67 [M
ꢀ H]- and 782.53 [M þ TFA ꢀ H]ꢀ, calcd for C27H18N4O13S2: 670.03.
HPLC (system B): tR ¼ 20.0 min, purity ¼ 97%.
d
9.28 (s, 1H), 9.14 (s, 1H), 8.63 (d, 1H, J ¼ 9.0 Hz), 8.35 (d, 1H,
J ¼ 9.0 Hz), 8.18 (m, 2H), 8.10 (d, 1H, J ¼ 9.0 Hz), 7.60 (t, 1H,
J ¼ 9.0 Hz), 7.53 (m, 2H), 7.31 (d, 1H, J ¼ 9.0 Hz). 13C NMR (75 MHz,
DMSO-d6):
d
159.4, 158.9, 153.9, 151.9, 151.7, 150.9, 148.1, 140.9,
2.3.4. 2,4-Dinitrophenyl ether (18)
136.0, 133.6, 132.1, 130.7, 127.7, 126.8, 125.7, 122.6, 122.3, 121.3,
120.2, 119.1, 118.8, 110.5. MS (ESI, positive mode): m/z ¼ 526.00
[M þ H]þ, calcd for C22H11N3O9S2 524.99. HPLC (system A):
tR ¼ 36.5 min, purity ¼ 99%.
To a solution of 3-benzothiazolyl-7-hydroxycoumarin (30 mg,
0.09 mmol, 1 equiv) in dry DMF (1 mL), anhydrous K2CO3 (35 mg,
0.254 mmol, 2.8 equiv) was added and the mixture was stirred at rt
under an Ar atmosphere for
5 min. Then, 1-fluoro-2,4-
dinitrobenzene (DNBF, 51 L, 0.407 mmol, 4.5 equiv) was added
m
2.3.2. 6-Sulfonated 2,4-dinitrobenzenesulfonyl ester (15)
and the reaction mixture was stirred at rt overnight. Thereafter, the
crude mixture was evaporated to dryness. The resulting residue
was dissolved in EtOAc and washed with deionized water, dried
over anhydrous Na2SO4 and evaporated to dryness. The resulting
crude product was purified by flash column chromatography on
silica gel using a mixture of cyclohexane and CH2Cl2 as eluents (step
gradient from 1:0 to 0:1) to give DNP ether 18 as a yellow solid
To a solution of 3-benzothiazolyl-7-hydroxycoumarin 6-sulfonic
acid 1 (50 mg, 0.12 mmol, 1 equiv) in dry pyridine (2 mL), DMAP
(10 mg, 0.066 mmol, 0.55 equiv) and DNBS-Cl (107 mg, 0.40 mmol,
3.3 equiv) were sequentially added and the resulting reaction
mixture was stirred at 50 ꢂC for 24 h. Thereafter, the crude mixture
was concentrated and purified by semi-preparative RP-HPLC (sys-
tem K). The product-containing fractions were lyophilized to give
the TFA salt of sulfonated DNBS ester 15 as a yellow amorphous
(25 mg, yield 48%). IR:
n
1724, 1594, 1524 (NO2), 1474, 1342 (NO2),
9.31 (s, 1H), 8.96
1276, 1197, 1119. 1H NMR (300 MHz, DMSO-d6):
d
solid (13 mg, yield 15%). IR:
n
1725, 1599, 1538 (NO2), 1478, 1346
(d, 1H, J ¼ 9.0 Hz), 8.55 (dd, 1H, J ¼ 9.0 Hz, J ¼ 3.0 Hz), 8.22 (m, 2H),
(NO2), 1272, 1245, 1191, 1150, 1065. 1H NMR (300 MHz, DMSO-d6):
8.10 (d, 1H, J ¼ 9.0 Hz), 7.60 (t, 1H, J ¼ 9.0 Hz), 7.51 (m, 3H), 7.37 (dd,
d
9.33 (s, 1H), 8.87 (s, 1H), 8.52 (s, 1H), 8.43 (d, 1H, J ¼ 9.0 Hz), 8.20
1H, J ¼ 9.0 Hz, J ¼ 3.0 Hz). 13C NMR (75 MHz, DMSO-d6):
d 159.7,
(d, 1H, J ¼ 9.0 Hz), 8.13 (d, 1H, J ¼ 9.0 Hz), 7.60 (t,1H, J ¼ 9.0 Hz), 7.53
159.2, 158.4, 154.8, 153.0, 151.9, 142.9, 141.5, 140.4, 135.9, 132.5,
130.0, 126.8, 125.5, 122.5, 122.3, 122.1, 122.0, 118.5, 116.6, 116.4,
107.0. MS (ESI, positive mode): m/z ¼ 462.13 [M þ H]þ, calcd for
(m, 2H), 7.08 (d, 1H, J ¼ 9.0 Hz). 13C NMR (75 MHz, DMSO-d6):
d
159.6, 159.1, 155.6, 154.5, 153.7, 152.0, 141.7, 141.2, 139.1, 137.9,
136.0, 130.8, 129.1, 126.8, 125.5, 122.7, 122.3, 121.5, 120.3, 119.3,
116.6, 110.45. MS (ESI, negative mode): m/z ¼ 540.07 [M ꢀ H]- and
1080.53 [2M ꢀ H]ꢀ, calcd for C22H11N3O12S3: 604.95, a non-
identified side-reaction occurred within the ESI probe and led to a
loss of 64 Da. HPLC (system A): tR ¼ 28.7 min, purity ¼ 98%.
C22H11N3O7S 461.03. HPLC (system A): tR ¼ 34.5 min, purity ¼ 96%.
2.3.5. 6-Sulfonated 2,4-dinitrophenyl ether (19)
To a solution of 3-benzothiazolyl-7-hydroxycoumarin 6-sulfonic
acid 1 (30 mg, 73
K2CO3 (30 mg, 200
m
m
mol, 1 equiv) in dry DMF (1 mL), anhydrous
mol, 2.8 equiv) was added and the mixture was
L,
mol, 4.5 equiv) was added and the reaction mixture was
2.3.3. Iminodiacetic acid 2,4-dinitrobenzenesulfonyl ester (16b)
2.3.3.1. Di-tert-butyl iminodiacetate 2,4-dinitrobenzenesulfonyl ester
(16a). To a solution of 7-hydroxycoumarin derivative 3a (46 mg,
stirred at rt under an Ar atmosphere for 5 min. Then, DNBF (40
320
m
m
stirred at 50 ꢂC for 2 h. Thereafter, the crude mixture was
concentrated and the resulting residue was directly purified by
semi-preparative RP-HPLC (system O). The product-containing
fractions were lyophilized to give the TFA salt of sulfonated DNP
83
mmol, 1 equiv) in dry CH2Cl2 (10 mL), 2,6-lutidine (100
mL,
830
mmol, 10 equiv) and DNBS-Cl (77 mg, 290 mol, 3.5 equiv) were
m
sequentially added and the resulting reaction mixture was stirred
at rt for 18 h. Thereafter, the crude mixture was evaporated to
ether 19 as a yellow amorphous solid (14 mg, yield 30%). IR: n 1732,
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

6
B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
1600, 1536 (NO2), 1470, 1409, 1345 (NO2), 1253, 1120, 1076. 1H NMR
(300 MHz, DMSO-d6):
of benzyl ether 22a as a yellow amorphous solid (19 mg, yield 23%).
1H NMR (300 MHz, DMSO-d6):
9.16 (s, 1H), 8.31 (s, 1H), 8.16 (d, 1H,
d
9.31(s, 1H), 8.86 (d, 1H, J ¼ 3.0 Hz), 8.51 (s,
d
1H), 8.42 (dd, 1H, J ¼ 9.0 Hz, J ¼ 3.0 Hz), 8.19 (d, 1H, J ¼ 9.0 Hz), 8.11
J ¼ 9.0 Hz), 8.07 (d, 1H, J ¼ 9.0 Hz), 7.56 (t, 1H, J ¼ 9.0 Hz), 7.45 (m,
(t, 1H, J ¼ 9.0 Hz), 7.54 (m, 3H), 7.10 (d, 1H, J ¼ 9.0 Hz). 13C NMR
2H), 5.20 (s, 2H), 3.68 (s, 3H), 3.59 (s, 3H), 2.27 (s, 3H), 2.17 (s, 6H).
(75 MHz, DMSO-d6):
d
159.6, 159.2, 155.6, 154.6, 153.7, 152.0, 141.7,
13C NMR (75 MHz, DMSO-d6):
d 160.8, 160.3, 159.8, 156.0, 153.4,
141.3, 139.1, 137.8, 136.0, 130.9, 129.2, 126.8, 125.6, 122.7, 122.3,
152.6, 152.0, 142.8, 135.7, 134.5, 131.4, 130.3, 130.0, 127.4, 126.6,
125.2, 125.4, 122.1, 115.7, 110.7, 100.1, 64.0, 62.1, 59.9, 12.8, 12.4, 11.7.
MS (ESI, positive mode): m/z ¼ 567.87 [M þ H]þ, calcd for
121.6, 120.4, 119.4, 116.6, 110.5. MS (ESI, negative mode): m/
z
C
¼
540.00 [M
ꢀ
H]ꢀ and 1080.53 [2MꢀH]ꢀ, calcd for
22H11N3O10S2 540.99. HPLC (system A): tR
¼
28.1 min,
C
28H25NO8S2: 567.10. HPLC (system A): tR ¼ 29.7 min, purity ¼ 98%.
purity ¼ 98%.
2.3.8.2. CAN-mediated removal of methoxy groups. To a solution of
dimethoxy-benzene derivative 22a (25 mg, 37 mol, 1 equiv) in a
mixture of CH2Cl2/CH3CN (2:1, v/v, 0.75 mL), cerium ammonium
nitrate (CAN, 50 mg, 85 mol, 2.3 equiv) was added and the resulting
2.3.6. 6-Sulfonated 4-cyano 2,4-dinitrophenyl ether (20)
m
To a solution of 3-benzothiazolyl-4-cyano-7-hydroxycoumarin-
6-sulfonic acid 2 (30 mg, 59
mmol, 1 equiv) in dry DMF (1 mL),
m
anhydrous K2CO3 (31 mg, 225
mixture was stirred at rt under an Ar atmosphere for 5 min. DNBF
m
mol, 3.8 equiv) was added and the
reaction mixture was stirred at rt for 2 h. Further amounts of CAN
(2 ꢁ 50 mg) was added after 2 h and 4 h of stirring. The reaction was
checked for completion by RP-HPC (system A) and volatiles were
evaporated to dryness. The crude product was purified by semi-
preparative RP-HPLC (system M). The product-containing fractions
were lyophilized to give the TFA salt of 22b as a yellow amorphous
(37 mL, 300 mmol, 5.1 equiv) was added and the reaction mixture
was stirred at 50 ꢂC for 5 h. Thereafter, the crude mixture was
concentrated and directly purified by semi-preparative RP-HPLC
(system Q). The product-containing fractions were lyophilized to
give the TFA salt of DNP ether 20 as a yellow amorphous solid (7 mg,
solid (7 mg, yield 30%). IR:
n
1720, 1607, 1555 (NO2), 1365 (NO2),
9.17 (s, 1H),
yield 18%). IR:
n
2118 (CN), 1738, 1596, 1538 (NO2), 1476, 1348 (NO2),
8.89 (d, 1H,
1292, 1253, 1226, 1173. 1H NMR (300 MHz, DMSO-d6):
d
1267, 1202, 1090, 1021.1H NMR (300 MHz, DMSO-d6):
d
8.30 (s,1H), 8.17 (d,1H, J ¼ 9.0 Hz), 8.08 (d,1H, J ¼ 9.0 Hz), 7.57 (t, 2H,
J ¼ 3.0 Hz), 8.48 (m, 2H), 8.28 (d, 1H, J ¼ 9.0 Hz), 8.21(d, 1H,
J ¼ 9.0 Hz), 7.46 (t, 2H, J ¼ 9.0 Hz), 5.13 (s, 2H), 2.15 (s, 3H), 2.02 (s,
J ¼ 9.0 Hz), 7.63 (m, 3H), 7.11 (d, 1H, J ¼ 9.0 Hz). 13C NMR (75 MHz,
3H), 2.00 (s, 3H). 13C NMR (75 MHz, DMSO-d6):
d 187.0, 185.4, 160.2,
DMSO-d6):
d
158.2,157.0,155.1, 154.4,153.3,151.4,141.7,139.4,138.1,
160.2,159.8,155.7,152.0,146.1,142.7,140.7,139.9,136.0,135.7,134.7,
130.0, 126.6, 125.2, 122.4, 122.1, 116.1, 111.1, 100.3, 61.6, 12.3, 12.2,
12.2. MS (ESI, negative mode): m/z ¼ 535.80 [M ꢀ H]ꢀ and 649.33
[M þ TFA ꢀ H]ꢀ, calcd for C26H19NO8S2:537.06. HPLC (system A):
tR ¼ 28.4 min, purity ¼ 97%.
136.8, 129.4, 127.6, 127.3, 126.9, 125.1, 123.6, 122.4, 121.7, 120.6,
120.0, 114.6, 113.7, 110.8. MS (ESI, negative mode): m/z ¼ 565.00
[M ꢀ H]- and 1130.73 [2M ꢀ H]ꢀ, calcd for C23H10N4O10S2 565.98.
HPLC (system A): tR ¼ 28.8 min, purity ¼ 99%.
2.3.7. N-Sulfopropylpyridinium 2,4-dinitrophenyl ether (21)
2.3.9. 6-Sulfonated quinone “trimethyl lock” carbamate (23)
To a pre-cooled mixture (0 ꢂC) of phosgene (w20% w/v solution
To a solution of TEA salt of 7-hydroxycoumarin-hemicyanine
dye 11 (20 mg, 32
K2CO3 (15 mg, 113
stirred at rt under an Ar atmosphere for 5 min. DNBF (24
129 mol, 4.0 equiv) was added and the reaction mixture was
m
mol, 1 equiv) in dry DMF (1 mL), anhydrous
mol, 3.5 equiv) was added and the mixture was
L,
in toluene, 150
mL, 280 mmol, 3.3 equiv) and TEA (40 mL, 280 mmol,
m
3.3 equiv), a solution of N-methyl secondary amine F (90 mg,
m
280 mmol, 3.3 equiv) in dry toluene (1 mL) was added dropwise at
m
0 ꢂC for 30 min and the resulting reaction mixture was stirred at rt
for 24 h. Thereafter, this chlorocarbonylation mixture was
concentrated and the residue was re-dissolved in dry pyridine
(2 mL). Then, a solution of 3-benzothiazolyl-7-hydroxycoumarin-6-
stirred at rt for 5 h. Thereafter, the crude mixture was concentrated
and directly purified by semi-preparative RP-HPLC (system Q). The
product-containing fractions were lyophilized to give the TFA salt
of DNP ether 21 as a yellow amorphous solid (10 mg, yield 39%). IR:
sulfonic acid 1 (35 mg, 85 mmol, 1 equiv) in dry pyridine (1 mL) was
n
1735, 1612, 1586, 1536 (NO2) 1474, 1347 (NO2), 1258, 1197, 1035. 1H
added dropwise at rt. The resulting reaction mixture was stirred for
24 h. After evaporation to dryness, the crude product was purified
by semi-preparative RP-HPLC (system N). The product-containing
fractions were lyophilized to give the TFA salt of carbamate 23 as
NMR (300 MHz, DMSO-d6): d 9.35 (s, 1H), 9.00 (m, 3H), 8.57 (d, 1H,
J ¼ 9.0 Hz), 8.33 (d, 2H, J ¼ 9.0 Hz), 8.24 (m, 2H), 8.12 (d, 1H,
J ¼ 9.0 Hz), 8.05 (d, 1H, J ¼ 16.0 Hz), 7.86 (d, 1H, J ¼ 16.0 Hz), 7.63 (t,
1H, J ¼ 9.0 Hz), 7.49 (m, 2H), 7.35 (d, 1H, J ¼ 9.0 Hz), 4.68 (m, 2H),
2.50 (m, 2H, masked by DMSO signal), 2.22 (m, 2H). Twisting and
bending molecular motions of such hemicyanine dye in solution pre-
vent the recording of a good quality 13C NMR spectrum even for a
reasonable sample concentration of 10 mg/mL. MS (ESI, positive
mode): m/z ¼ 687.20 [M þ H]þ, calcd for C32H22N4O10S2 686.08.
HPLC (system A): tR ¼ 29.4 min, purity ¼ 94%.
a yellow amorphous solid (7 mg, yield 10%). IR:
1555 (NO2), 1405, 1238 (NO2), 1162, 1081, 1022. 1H NMR (300 MHz,
DMSO-d6):
9.25 (s, 1H), 8.37 (s, 1H), 8.20 (d, 1H, J ¼ 9.0 Hz), 8.12 (d,
n 1721, 1644, 1611,
d
1H, J ¼ 9.0 Hz), 7.59 (t, 2H, J ¼ 9.0 Hz), 7.49 (t, 2H, J ¼ 9.0 Hz), 7.35
(s,1H), 3.70 (s, 2H), 3.57 (s, 2H), 3.42 (m, 1H), 3.32 (m, 1H), 3.18 (s,
1H), 3.02 (m, 3H), 2.87 (s, 2H), 2.79 (s, 1H), 2.73 (s, 1H), 2.05 (m, 3H),
1.86 (m, 6H), 1.33 (m, 6H). Not enough product to record a good
quality 13C NMR spectrum. MS (ESI, negative mode): m/z ¼ 720.13
[M ꢀ H]- and 1440.47 [2M ꢀ H]ꢀ, calcd for C35H35N3O10S2: 721.18.
HPLC (system A): tR ¼ 30.8 min, purity ¼ 92%.
2.3.8. 6-Sulfonated quinone methyl ether (22b)
2.3.8.1. 6-Sulfonated 2,5-dimethoxy-3,4,6-trimethylbenzyl ether
(22a). To a solution of 3-benzothiazolyl-7-hydroxycoumarin-6-
sulfonic acid 1 (50 mg, 0.12 mmol, 1 equiv) in dry DMF (3 mL),
anhydrous K2CO3 (55 mg, 0.40 mmol, 3.3 equiv) was added and the
mixture was stirred at rt for 20 min. Thereafter, 1-chloromethyl-
2.4. General procedure for in vitro thiolysis of fluorogenic probes e
fluorescence assay
2,5-dimethoxy-3,4,6-trimethylbenzene
E
(36 mg, 0.32 mmol,
Stock solutions (1.0 mg/mL) of water-soluble pro-fluorophores
15, 16b, 17b and 19-23 were prepared in H2O/DMSO (9:1, v/v)
whereas reference thiol probes 14 and 18 were dissolved in DMSO
(1.0 mg/mL). Stock solutions (10 mg/mL) of analytes (Cys, NaThioPi,
K3PO4 and 4-chlorothiophenol) were prepared in ultrapure water
except for 4-chlorothiophenol (in CH3CN). A micromolar solution
2.7 equiv) and KI (33 mg, 0.20 mmol, 1.7 equiv) were sequentially
added and the resulting reaction mixture was stirred at rt for 24 h.
Thereafter, the crude mixture was evaporated to dryness and
directly purified by semi-preparative RP-HPLC (system L). The
product-containing fractions were lyophilized to give the TFA salt
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
7
(for each thiol probe) was obtained by dilution of stock solution
with PB or PB/DMSO. Depending on the reactivity of the probe (i.e.,
thiolysis kinetics) and the fluorescence efficiency of the released
3-benzothiazolyl-7-hydroxycoumarin, this latter approach will lead
to a dramatic red-shift of its spectral features due to the extension of
the aromatic
p-system. For the two first water-solubilizing meth-
fluorophore, a concentration in the range from 2.2
m
M to 8.6
m
M
odologies, a further chemical modification of the coumarin core,
namely cyanation of 4-position, was also considered in order to
obtain a further 100 nm red-shift in emission maximum of 3-
benzothiazolyl-7-hydroxycoumarin. The practical implementation
of these functionalization methods has led to eleven different
water-soluble 3-benzothiazolyl-7-hydroxycoumarin derivatives,
and is summarized in Scheme 1 (for aromatic sulfonation and
Mannich approach) and Scheme 2 (for Knoevenagel approach).
First, 6-sulfonated derivative 1 was readily synthesized from 5-
formyl-2,4-dihydroxybenzenesulfonic acid and benzothiazole-2-
acetonitrile according to a two-step protocol previously reported
in the literature [27,28,33]. Compound 1 was next used for the
preparation of water-soluble 4-cyano derivative 2 through Michael-
type addition of cyanide anion on C3eC4 double bond and subse-
quent re-aromatization with DDQ [52]. The 6-sulfonated coumarin
2 was readily purified by semi-preparative RP-HPLC and recovered
in a pure form with a satisfying 45% yield. Concerning the synthesis
of 8-substituted derivatives 3b and 4b, aminomethylation reactions
were carried out according to a protocol initially developed for the
grafting of the bis(carboxymethyl)aminomethyl moiety onto 3-
unsubstituted 7-hydroxycoumarins [50]. Thus, 3-benzothiazolyl-
7-hydroxycoumarin and its 4-cyano derivative were treated with an
excess of freshly prepared Mannich reagent (i.e., iminium salt
resulting from the reaction between paraformaldehyde and di-tert-
butyl iminodiacetate with a catalytic amount of KOH) in refluxing
EtOH. The reactions were found not to be complete even after a
prolonged time of heating and all attempts to isolate the resulting 8-
N,N-dialkylamino coumarins 3a and 4a by silica or alumina column
chromatography failed. Alternative purification over reversed-
phase silica gel provided 3a and 4a in a pure form but with
modest yields (24% and 18% respectively), partly explained by losses
of material through non-specific adsorption phenomena over the
chromatographic stationary phase. Subsequent treatment of 3a and
4a with a 50% solution of TFA in CH2Cl2 to remove tert-butyl groups,
afforded the targeted 8-iminodiacetic acid derivatives 3b and 4b.
Much better yields were obtained for the Mannich reactions
involving iminium salts derived from sarcosine tert-butyl ester and
ethyl isonipecotate. Furthermore, the alkaline reaction conditions
led to the premature cleavage of alkyl ester moieties of these un-
usual amino acids and has enabled us to directly recover the free
carboxylic acid derivatives 5 and 6 in 48% and 38% yield respectively.
In order to further increase the water-solubility of these latter hy-
drophilic coumarins, we next considered the amidification of their
added carboxylic acid functionality with a di-sulfonated amino
linker derived from taurine [32,35]. Our first attempts involving the
aminolysis of N-hydroxysuccinimidyl (NHS) esters of 5 and 6 with
an excess of 2-aminoethane-1,1-disulfonic acid failed. Thus, we have
chosen to convert the carboxylic acid of 5 and 6 into the more
reactive acyl bromide derivatives by using PyBrOP/DIEA. Upon this
acid pre-activation step, subsequent reaction with 2-aminoethane-
1,1-disulfonic acid (tetrabutylammonium, TBAþ salt) led to the
desired di-sulfonated coumarins 7 and 8 which were purified by
semi-preparative RP-HPLC (31% and 20% yields respectively).
The synthetic access to the last three targeted water-soluble 3-
benzothiazolyl-7-hydroxycoumarin derivatives 10, 11 and 13 has
required the prior preparation of the 8-formyl derivatives 9 and 12
using the Duff reaction [51] (Scheme 2). This aromatic formylation
was carried out under standard conditions which are not really
effective for the 6-sulfonated derivative 1 (isolated yield7% vs. 58% for
3-benzothiazolyl-7-hydroxycoumarin) even after improving its sol-
ubility in TFA/hexamine mixture by a counter-ion exchange process
(TBAþ instead of Hþ). However, this latter sulfonated benzaldehyde
was used. 3 mL of this solution was transferred into a quartz fluo-
rescence cell (Varian, fluorescence cell, Open Top, 10 ꢁ 10 mm,
3.5 mL) and thermostated at 25 ꢂC. The required number of
equivalents of analyte (2, 5, 50 or 250 equiv) was added and the
resulting mixture was homogenized through magnetic stirring for
2 min. The fluorescence emission of the released fluorophore was
monitored at the suitable wavelength (
l
¼ 490, 600 or 655 nm,
emission slit ¼ 5 nm, upon excitation at
l
¼ 390, 510 or 500,
excitation slit ¼ 5 nm; excitation/emission filters: auto) over time
with measurements recorded every 1 s.
3. Results and discussion
3.1. Synthesis of water-soluble 3-benzothiazolyl-7-
hydroxycoumarin derivatives
In the early 1980s, the Wolfbeis group was interested in the
synthesis and photophysical characterization of a series of 3-
substituted 7-hydroxycoumarin derivatives with the aim of shift-
ing the excitation and emission maxima of umbelliferone to longer
wavelengths and simultaneously lowering the pKa value of its
phenol moiety [25]. Of all the electron-withdrawing substituents
explored, 2-benzothiazolyl group was identified as a promising
molecular unit to achieve these ambitious goals (as inferred from
the spectral features under simulated physiological conditions: Abs/
Em max. 431/488 nm, FF ¼ 43% [40], pKa ¼ 7.0 vs. Abs/Em max. 326/
452 nm, FF ¼ 76%, pKa ¼ 7.7 for umbelliferone [38]). Consequently,
3-benzothiazolyl-7-hydroxycoumarin has emerged as a valuable
alternative to 7-hydroxycoumarin, and is frequently used as a flu-
orogenic dye in various (bio)sensing applications [29,30,41e44].
Although the pKa value of this phenolic fluorophore allows a high
solubility in various alkaline aq. buffers (pH 8e10), it is surprising
that little attention has been paid to the synthesis of hydrophilic
analogues which exhibit a much higher water-solubility even if
their 7-OH group is masked. This is particularly interesting to
facilitate the design of phenol-based fluorogenic probes (e.g., thiol-
imaging agents derived from 7-hydroxycoumarins) readily soluble
in biological media whatever the pH and without using an organic
co-solvent. Contrary to 7-N,N-dialkylaminocoumarin derivatives,
few synthetic methods aimed at introducing various polar or
ionizable groups at specific positions of a 7-hydroxycoumarin
scaffold have been reported in the literature. Particularly note-
worthy, some umbelliferones whose 3-position is functionalized
with a carboxyl moiety or the 4-position is substituted by an acetic
acid arm or a
a-sulfo-b-alanyl linker have been synthesized but
these strategies cannot be applied to 3-heteroaryl-7-
hydroxycoumarins [45e48]. There is clearly a need to explore
alternative synthetic accesses to decorate the 3-substituted
coumarin ring systems with one or several hydrophilic sub-
stituents. Thus, we decided to explore three different approaches to
achieve this: (1) sulfonation of the 6-position through electrophilic
aromatic substitution [33], (2) Mannich-type reaction to function-
alize the 8-position with an aminomethyl arm derived from
a hydrophilic secondary amine (i.e., iminodiacetic acid, sarcosine,
and isonipecotic acid) [49,50] optionally followed by a post-
amidification reaction of the carboxylic acid function with
2-aminoethane-1,1-disulfonic acid B [32,35], and (3) Knoevenagel-
type reaction between an aldehyde functionality pre-introduced
in the 8-position and a sulfobetain derivative of 2- or 4-methyl-
azaheterocyle (N-sulfopropyl-2-methylbenzothiazole
methylpyridine D) [51]. In addition to impart water-solubility of
C or -4-
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

8
B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
Scheme 1. Synthesis of water-soluble 3-benzothiazolyl-7-hydroxycoumarin derivatives through aromatic sulfonation and Mannich approach. Reagents and conditions: (i) oleum
(6 equiv), 2,4-dihydroxybenzaldehyde (1 equiv), 0 ꢂC, 30 min, then rt, 2 h, 54%; (ii) (a) benzothiazole-2-acetonitrile (1 equiv), piperidine (10 equiv), rt, 18 h, (b) HCl (6 M), reflux,
24 h, 71% (for the two steps); (iii) KCN (2.3 equiv), H2O, DMF, 50 ꢂC, 3 h, then DDQ (1.1 equiv), rt, 2 h, 48%; (iv) paraformaldehyde (6.8 equiv), A (3.4 equiv), KOH (0.5 equiv), EtOH, rt,
1 h, then reflux, 24 h, 24% (3a) and 18% (4b); (v) TFA, CH2Cl2, reflux, 2 h, 66% (3b) and 60% (4b); (vi) paraformaldehyde (7.2 equiv), sarcosine tert-butyl ester HCl salt (3.6 equiv), KOH
(7.2 equiv), EtOH, reflux, 4 h, 48%; (vii) paraformaldehyde (6.8 equiv), ethyl isonipecotate (3.4 equiv), KOH (6.8 equiv), EtOH, reflux, 4 h, 43%; (viii) PyBrOP (3.2 equiv), DIEA
(2.8 equiv), NMP, rt, 30 min, then TBAþ salt of disulfonated amine B (3.8 equiv), DIEA (6.3 equiv), NMP, 0 ꢂC then rt, 5 h, 31% (7) and 20% (8). 7-Hydroxycoumarin derivatives were
isolated as HCl (3-benzothiazolyl-7-hydroxycoumarin and 1) or TFA (2, 3b, 4b, 5 and 6) salts, except for 7 and 8 (acid form, after Dowex Hþ desalting).
was obtained in a sufficient amount to achieve the Konevenagel-type
condensation reaction with N-sulfopropyl-2-methylbenzothiazole C.
Experimental conditions currently used for the preparation of 7-N,N-
dialkylamino- or 7-hydroxycoumarin-hemicyanine dyes (pyrrolidine
in dry EtOH at rt) [53,54,51] were employed and led to water-soluble
extended conjugated coumarins 10 and 13 readily purified by semi-
preprative RP-HPLC (50% and 23% yields respectively). The less hy-
drophobic sulfobetain N-sulfopropyl-4-methylpyridine D was also
subjected to the same base-catalysed condensation reaction with the
8-formyl derivative 9 to give a further water-soluble derivative of 3-
benzothiazolyl-7-hydroxycoumarin (compound 11, yield 44%).
Structures of these novel hydrophilic phenol-based fluo-
rophores were confirmed by detailed measurements including ESI
mass spectrometry and NMR analyses. Furthermore, the purity of
each compound (determined through RP-HPLC analyses) was
found to be equal or above to 95%, suitable for an accurate and
reliable determination of their photophysical properties.
phosphate buffer (pH 7.5) and compiled in Table 1. For some
compounds, further measurements were achieved in DMSO,
especially for assessing the differences in fluorescence quantum
yield compared to an aq. medium, when the use of this organic co-
solvent is required for fluorescence-based thiol detection assays
involving the release of a coumarinic dye (vide infra). The first thing
to note is that the introduction of the water-solubilizing moieties
namely sulfonic acid and N,N-disubstituted aminomethyl onto the
6- or 8-position of the 3-benzothiazolyl-7-hydroxycoumarin scaf-
fold has nearly no influence on absorption and (fluorescence)
emission band positions (see entries 3, 7 and 9e12 compared to
entry 1, and Fig. 1). Conversely, and as already reported for other 7-
hydroxycoumarin derivatives [25,51], cyanation of the 4-position or
grafting of a N-sulfopropyl azaheterocycle to the 8-position and
through a dimethine chain, gave rise to dramatic red-shifts in ab-
sorption (63e109 nm) and emission (105e143 nm) maxima in aq.
media (see entries 5, 8, 13, 15, 17 compared to entry 1, and Fig. 1).
Moreover, and as foreseen, 6-sulfonic acid substituent and the N,N-
disubstituted aminomethyl arms improve significantly the fluo-
rescence efficiency of 3-benzothiazolyl-7-hydroxycoumarin
scaffold in aq. media. Indeed, a 63% increase in fluorescence
quantum yield was obtained for the 6-sulfonated fluorophore 1
whereas for the Mannich derivatives 3b and 5e8, increases ranged
3.2. Photophysical properties of water-soluble 3-benzothiazolyl-7-
hydroxycoumarin derivatives
Due to their excellent water-solubility (greater than 10 mM), the
optical properties of these novel fluorophores were evaluated in
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
9
Scheme 2. Synthesis of water-soluble 3-benzothiazolyl-7-hydroxycoumarin derivatives through Knoevenagel approach. Reagents and conditions: (i) (a) hexamine (2.3 equiv), TFA,
reflux, 24 h, (b) HCl (1 M), 3 h, rt, 58%; (ii) C or D (2e3 equiv), pyrrolidine (2e3 equiv), EtOH, rt, 2 h, 50% (10), 44% (11) and 23% (13, for the two steps iieiii); (iii) (a) Bu4NOH (1 equiv),
H2O, CH2Cl2, (b) hexamine (2.2 equiv), TFA, reflux, 48 h, (c) HCl (1 M), rt, 3 h. 7-Hydroxycoumarin derivatives were isolated as TEA salts (11 and 13) except for 12 (acid form, after Dowex
Hþ desalting).
from 44% to 151% were observed. For these latter 7-
hydroxycoumarins, lowering the pKa of their phenol moiety
through the formation of a strong hydrogen bond with the adjacent
tertiary amino group that leads to an increase in the molar fraction
of the emitting phenolate form, is generally accepted to assess their
higher fluorescence intensity compared to the 8-unsubstituted
parent compounds [49]. Thus, the highest quantum yield is ob-
tained with the 7-hydroxycoumarin-isonipecotic acid conjugate 6
that bears the most basic N,N-dialkylaminomethyl moiety, as
inferred from the comparison between the pKa values for the
secondary amine of isonipecotic acid (10.85 ꢃ 0.1), sarcosine
(10.20 ꢃ 0.1) and iminodiacetic acid (9.30 ꢃ 0.5) [55]. A similar
assumption could be claimed to partially explain the good quantum
yield of 1 (compared to 3-benzothiazolyl-7-hydroxycoumarin)
because ortho-sulfonation of a phenol moiety is known to decrease
its pKa by ca. 0.5 units [56]. Furthermore, the presence of two
negative charges at physiological pH within the coumarin scaffold
(i.e., sulfonate and phenolate) promotes dyeedye repulsion and
resistance to aggregation-induced fluorescence quenching, as
confirmed by the perfect match between the absorption and exci-
tation spectra of 1 recorded in aq. buffer (Fig S2.1.1). For the
hemicyanine dyes 10 and 11, sulfobetain moiety of the grafted aza-
heterocycle is not sufficient to prevent dyeedye aggregation in
phosphate buffer and the free fluorescent monomers of 10 and 11
with satisfactory quantum yields (entries 13 and 15) are obtained
by adding 5% (w/v) of bovine serum albumin (BSA), an additive
often used in buffers for mimicking body fluids. Indeed, this protein
is known to enhance the emission of many fluorophores due to a
combination of rigidization, reduction in the polarity of the dye’s
microenvironment (binding in the hydrophobic BSA pocket), and
deaggregation [57]. Surprisingly, the introduction of a further sul-
fonate group on the 6-position of coumarin scaffold of bis-
benzothiazolyl derivative 13 has no beneficial effects to disrupt
aggregates in phosphate buffer. Again, there is a need to add BSA to
determine a significant value of fluorescence quantum yield for 13
under simulated physiological conditions (entry 17). For the 4-
cyano derivatives 2 and 4b that emit in the orange spectral re-
gion (maximal emission peaks at 599 and 593 nm respectively), no
evidence of dyeedye aggregation is observed (see ESI for the cor-
responding absorption/excitation spectra, Fig S2.1.2 and S2.1.4) and
quantum yields close to 20% are obtained (see entries 5 and 8). All
these results clearly show that the chosen water-solubilizing
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

10
B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
Table 1
Photophysical properties of water-soluble 3-benzothiazolyl-7-hydroxycoumarins at 25 ꢂC.
Dye
Solvent
PB
lmax, abs (nm)a
lmax, em (nm)
Stokes shift (cmꢀ1
2710
)
FF (%)
Stdb/lexc (nm)
3-benzothiazolyl-
431
488
43
7OH/390
7-OH
3-benzothiazolyl-7-OH
1
1
2
DMSO
PB
DMSO
PB
DMSO
PB
PB
PB
397
441
395
502
416
434
494
436
436
e
485
485
475
599
590
485
593
485
485
e
4570
2057
4264
3226
7089
2423
3379
2317
2317
e
81
70
70
19
29
68
24
72
7OH/390
7-OH/390
7-OH/390
SR101/510
Fluo/470
7-OH/390
SR101/510
7-OH/390
7-OH/390
Fluo/440
7-OH/390
7-OH/390
CV/550
2
3b
4b
5
6
PB
e
z100c
z100c
62
90
24
4
26
4
18
7
7
8
PB
PB
436
436
535
598
502
558
540
550
485
485
631
704
615
705
626
669
2317
2317
2844
2518
3660
3737
2544
3234
10
10
11
11
13
13
PB þ 5% BSAd
DMSO
CV/550
SR101/510
CV/550
CV/550
CV/550
PB þ 5% BSAd
DMSO
PB þ 5% BSAd
DMSO
a
Most of the water-soluble 3-benzothiazolyl-7-hydroxycoumarin derivatives were not obtained in sufficient amounts for highly accurate measurements of absorption
coefficients. Molar absorptivity of 3-benzothiazolyl-7-hydroxycoumarin and 6-sulfonated derivative 1 in EtOH have been already reported in the literature: 30 050 Mꢀ1 cmꢀ1
(at 398 nm) [28] and 42 400 Mꢀ1 cmꢀ1 (at 400 nm) [33] respectively.
b
7-OH ¼ 7-hydroxycoumarin (FF ¼ 76% in PB, lex ¼ 390 nm), SR101 ¼ sulforhodamine 101 (FF ¼ 95% in EtOH, lex ¼ 510 nm), Fluo ¼ fluorescein (FF ¼ 91% in 0.1 N NaOH,
lex ¼ 470 nm), CV ¼ cresyl violet (FF ¼ 56% in EtOH, lex ¼ 550 nm) [39].
c
Relative quantum yield of 6 is slightly greater than 100% (107e108%) because this compound was found to be more fluorescent than the standards 7-hydroxycoumarin
and fluorescein; no suitable standard is available to determine a quantum yield lower than 100%.
d
BSA was added to the phosphate buffer to disrupt aggregates that prevent the determination of relative quantum yield (a non-linear relationship between fluorescence
emission and absorbance at the excitation wavelength was obtained in PB).
strategies are particularly effective to get 7-hydroxycoumarin-
based fluorophores with strong long-wavelength emission in aq.
media.
2,4-dinitrobenzenesulfonyl ester 15 [58]. Thus, the treatment of
phenol 1 with 3.3 equiv of DNBS-Cl and 0.5 equiv of DMAP in dry
pyridine at 50 ꢂC has enabled us to obtain 15 which was recovered
by semi-preparative RP-HPLC (yield 15%). The synthesis of 7-O-
DNBS ester of 3-benzothiazolyl-7-hydroxycoumarin (compound
14) was also achieved in order to provide a reference thiol probe. By
contrast, all attempts to obtain the 7-O-DNBS ester derived from
the 4-cyano derivative 2 in a significant amount have failed, high-
lighting the lower nucleophilicity of its phenol group. Not
3.3. Synthesis of water-soluble fluorogenic probes and optical
responses to thiols
Among the numerous quenching moieties commonly found in
fluorescent turn-on thiol probes, 2,4-dinitrobenzenesulfonyl
(DNBS) group is probably one of the most heavily used, through
either the esterification of a phenol or sulfonylation of an aniline
moiety of the selected fluorescent organic dye [1]. Due to the high
level of electron-deficiency on its phenyl ring, DNBS moiety can act
as an electron sink when attached to a fluorophore and may incur
photoinduced electron transfer (PeT) or strongly impacts the in-
ternal charge transfer (ICT) state of the molecule, both leading to
the quenching of its native fluorescence. In the presence of thiols,
DNBS arylesters or arylamides can easily undergo a SNAr desulfo-
nylation reaction (addition-elimination mechanism) whose driving
force is the release of SO2 gas, and thus resulting in a dramatic
fluorescence increase. We have decided to implement this strategy
to the water-soluble 3-benzothiazolyl-7-hydroxycoumarin de-
rivatives previously synthesized (Scheme 3). Positive effects on the
water-solubility of the resulting thiol probes are expected because
polar substituents such as sulfonate, sulfobetain or N,N-dia-
lkylaminomethyl should readily offset the hydrophobic character of
DNBS moiety. This O-sulfonylation reaction was first conducted
with the 6-sulfonated derivative 1 and was found not to work
properly, probably due to the steric hindrance and electronic effects
of the sulfonic acid substituent located close to the 7-OH group. A
wide range of experimental conditions including the use of
different bases (pyridine, DMAP, 2,6-lutidine, tBuOK and K2CO3),
solvents (pyridine and DMF), temperatures (rt, 50 ꢂC, 80 ꢂC or
reflux) and variable amounts of DNBS chloride (DNBS-Cl) has
been evaluated aimed at optimizing the synthesis of fluorogenic
Fig. 1. Normalized fluorescence emission spectra of 7-hydroxycoumarin derivatives in
PB (or in PB þ 5% BSA) buffer (blue: 7-hydroxycoumarin, lex ¼ 390 nm; green: 3-
benzothiazolyl-7-hydroxycoumarin-6-sulfonic acid, lex
¼
390 nm; orange: 3-
510 nm; red:
7-hydroxycoumarin-hemicyanine 11, lex ¼ 510 nm; dark red: 7-hydroxycoumarin-
hemicyanine 13, lex 550 nm and purple: 7-hydroxycoumarin-hemicyanine 10,
benzothiazolyl-4-cyano-7-hydroxycoumarin-6-sulfonic acid 2, lex
¼
¼
lex ¼ 550 nm). (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
11
Scheme 3. Synthesis of water-soluble thiol-sensitive fluorogenic probes. Reagents and conditions: (i) DNBS-Cl (1.6 equiv), 2,6 lutidine (3.1 equiv), DMF, 120 ꢂC, 18 h, 12% (14) or
DNBS-Cl (3.3 equiv), DMAP (0.55 equiv), pyridine, reflux, 24 h, 15% (15) or (a) DNBS-Cl (3.5 equiv), 2,6 lutidine (10 equiv), CH2Cl2, rt, 18 h, (b) TFA, CH2Cl2, reflux, 2 h, 39% (16b, for the
two steps) and 10% (17b, for the two steps); (ii) DNBF (4.5 equiv), K2CO3 (2.8 equiv), DMF, rt, overnight, 48% (18) and 39% (21) or DNBF (4.5 or 5.1 equiv), K2CO3 (2.8 or 3.8 equiv),
DMF, 50 ꢂC, 2 h or 5 h, 30% (19) and 18% (20); (iii) (a) K2CO3 (3.3 equiv), DMF, rt, 20 min, (b) E (2.7 equiv), KI (1.7 equiv), rt, 24 h, 23% (22a), (c) CAN (6.9 equiv), CH2Cl2/CH3CN (2 : 1, v/
v), rt, 6 h, 30%; (iv) (a) F (3.3 equiv), phosgene (3.3 equiv), TEA (3.3 equiv), toluene, rt, 24 h, quant. yield (b) F carbamoyl chloride (3.3 equiv), pyridine, rt, 24 h, 10%. All fluorogenic
probes were isolated as TFA salts except for 14 and 18.
surprisingly, sulfonylation conditions are not compatible with the
moderate stability of hemicyanine dyes such as 10, 11 and 13 and
our initial attempts have led to the complete degradation of their
dimethine chain. Finally, we have explored the O-sulfonylation of 7-
hydroxycoumarins bearing an N,N-disubstitued aminomethyl arm
on the 8-position. All reactions carried out with the free (di)car-
boxylic or di-sulfonic acid derivatives 5e8 failed to afford the tar-
geted DNBS esters. Degradation of their 8-substituent in particular
through decarboxylation and/or substitution (by pyridine at the
benzylic position) reactions was obtained. To circumvent this issue,
we have considered the sulfonylation of 8-substituted 7-
hydroxycoumarins whose two carboxylic acids are protected as
tert-butyl esters (compounds 3a and 4a). This then allows (1) to
avoid the premature decarboxylation of the hydrophilic arm and
(2) to use a less polar and unreactive solvent (i.e., CH2Cl2). Com-
pounds 3a and 4a were thus reacted with DNBS-Cl in the presence
of excess 2,6-lutidine and in dry CH2Cl2. After purification by con-
ventional column chromatography or preparative TLC, a further TFA
treatment provided the DNBS esters 16b and 17b in 39% and 10%
yield respectively. All spectroscopic data were in agreement with
the structures assigned for 14, 15, 16b and 17b. Furthermore, the
lack of free 7-OH parent fluorophore (as a minor impurity) was
undoubtedly confirmed by RP-HPLC analyses and purity above 97%
was found for all DNBS arylesters except for the 4-cyano derivative
17b (purity ¼ 78%). This latter compound is easily prone to hy-
drolysis in aq. solution, since electron-withdrawing properties of its
cyano group may enhance the electrophilicity of sulfonate ester.
The ability of DNBS group to effectively quench the blue-green
fluorescence of water-soluble 7-hydroxycoumarins 14, 15 and 16b
was confirmed through the determination of fluorescence quantum
yields of these fluorogenic probes that do not exceed 5% in phos-
phate buffer (ESI, Table S2.3.1).
The sensing response of the three stable fluorogenic DNBS
probes toward two different thiols namely Cys and thiophosphate
anion (ThioPi) and the unreactive phosphate anion (Pi) was next
examined. The time-dependant fluorescence intensity changes of
probes 14, 15 and 16b with these analytes were studied and the
results are shown in Fig. 2. Upon addition of Cys (5 equiv), the so-
lution of non-sulfonated probe 14 showed an initial fast fluores-
cence increase followed by
a gradual further increase in
fluorescence intensity at 490 nm to reach a plateau after 200 min
(Fig. 2a). A slower fluorescence response was obtained for the less
nucleophilic ThioPi whereas the addition of Pi anion did not induce
obvious variations in fluorescence intensity, suggesting that probe
14 was stable in the assay conditions. In sharp contrast, a much
slower but steady fluorescence increase was observed for the 6-
sulfonated derivative 15, yet only in the presence of 250 equiv. of
Cys (Fig. 2C1). No plateau has been reached after 16 h of incubation
and a fluorescence level three times lower than that noted for 14
was obtained. This different behaviour can be explained by the
steric effects of sulfonic acid substituent leading to a reduction in
the accessibility of the thiol-reactive electrophilic centre of the
probe, and by the fact that the fluorogenic SNAr reaction involves a
build-up of negative charge in the transition state (Meisenheimer
complex) which can be destabilized by the negative electrostatic
field induced by this charged water-solubilizing group [10].
Although the DNBS probe 15 is perfectly soluble in phosphate
buffer alone, we found that the use of DMSO as a co-solvent
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

12
B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
Fig. 2. Time-dependant fluorescence intensity of DNBS probes at 490 nm (lex ¼ 390 nm) in the presence of sulfhydryl analytes (Cys or ThioPi) and Pi anion in PB buffer (100 mM, pH
7.5) at 25 ꢂC. (A) probe 14 (6.3
M) with Cys (5 equiv) in blue, ThioPi (5 equiv) in red, and Pi (5 equiv) in green; (B) probe 16b (3.4 M) with Cys (5 equiv) in blue, ThioPi (5 equiv) in
red, and Pi (5 equiv) in green; (C1) probe 15 (8.6 M) with Cys (250 equiv) in blue, ThioPi (250 equiv) in red, and Pi (250 equiv) in green; (C2) probe 15 (8.6 M) with Cys (50 equiv)
m
m
m
m
in blue, ThioPi (250 equiv) in red, and Pi (250 equiv) in green recorded in PB (100 mM) þ 45% DMSO (pH 8.3). A lower concentration of probe 16b was used to avoid exceeding the upper
limit of fluorescence detection (1000 units). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
promoted the kinetics of this thiolysis process with lower amounts
of Cys (50 equiv) and with ThioPi (250 equiv) (Fig. 2C2). However, to
avoid the use of such additives, that are scarcely or not at all
compatible for in cellulo or in vivo applications, we next assessed
the fluorogenic reactivity of 8-substituted DNBS probe 16b. As
depicted in Fig. 2b, in the presence of non-nucleophilic Pi anion,
this probe displayed no significant changes in fluorescence in-
tensity at 490 nm. Conversely, when being treated with 5 equiv of
Cys, the solution exhibited a dramatic increase in the emission
intensity which reached a maximum in less than 50 min. Again, a
slower fluorescence response was observed for ThioPi anion.
However, for both sulfhydryl analytes, a large rate acceleration of
the fluorogenic SNAr process was observed, compared to thiolysis of
DNBS probes 14 and 15. Thus, 8-functionalization of 3-
RP-HPLC (30% and 18% yields respectively). A similar protocol was
applied to 3-benzothiazolyl-7-hydroxycoumarin and hemicyanine
dye 11 and the lack of 6-SO3H substituent allowed the reaction to
work at rt. These milder conditions prevented the degradation of
the dimethine chain of 11 previously observed during the synthesis
of DNBS arylesters (vide supra). Thus, the DNP probes 18 and 21
were readily recovered by conventional column chromatography
and semi-preparative RP-HPLC respectively. As expected, these four
DNP ethers are essentially non-fluorescent in both phosphate
buffer and DMSO (Table S2.3.1) and their sensing response to thiols
were also studied through time-dependant analyses (Fig. 3). No
significant changes in emission intensity of these DNP probes were
observed in phosphate buffer upon addition of thiols (Cys and
ThioPi) whatever the number of equivalents used (Fig. 3A1eC1).
These results can be related to the study of Lin et al. which high-
lights the excellent selectivity of DNP probe 18 for benzenethiols
over aliphatic thiols (such as Cys) at neutral pH (i.e., phosphate
buffer, pH 7.0) [41]. This property was attributed to the distinct pKa
values of benzenethiols (pKa ¼ 6.5) and aliphatic thiols (pKa ¼ 8.5),
and to the thiolysis of dinitrophenyl ethers proceeding via SNAr by
the nucleophilic thiolate. In our case, the addition of DMSO to
phosphate buffer (45% v/v) seriously affects the pH of the medium
(8.3 against 7.5 for the buffer alone) and a fluorescence turn-on
response was then obtained for all DNP probes except for the 4-
cyano derivative 20 (Fig. 3A2eC2). For this latter fluorogenic
probe, thiolysis of DNP moiety and subsequent fluorescence spec-
tral changes were observed only in DMSO and with 250 equiv. of
thiol analytes (data not shown). The fastest response time was
obtained with the non-sulfonated DNP ether 18 (Fig. 3A2), thereby
demonstrating once again that ortho-sulfonation should not be the
preferred approach to make water-soluble fluorogenic phenolic
dyes. The small far-red fluorescence enhancement observed for the
benzothiazolyl-7-hydroxycoumarin
by
Mannich
reaction
involving the use of a hydrophilic secondary amine, should be
preferred in order to obtain blue-green emitting thiol-reactive
DNBS probes soluble in aq. media.
In order to find alternative 7-O-quenching moieties whose thiol-
reactivity is less or not affected by the presence of 6-SO3H sub-
stituent onto the 7-hydroxycoumarin scaffold, we next explored
the synthesis of water-soluble fluorogenic probes whose DNBS
arylester is replaced either by the 2,4-dinitrophenyl ether (DNP) or
a benzoquinone-type Michael acceptor grafted to the 7-OH group
through a benzyl ether or a self-immolative carbamate linkage.
For 3-benzothiazolyl-7-hydroxycoumarin and its three water-
soluble derivatives 1, 2 and 11, the O-etherification reaction was
performed with Sanger reagent (2,4-dinitrofluorobenzene, DNBF)
according to the procedure reported by Lin et al. [41]. The 6-
sulfonated derivatives 1 and 2 were reacted with an excess
of DNBF and K2CO3 in dry DMF at 50 ꢂC. The resulting 7-O-DNP
ethers 19 and 20 were isolated in a pure form by semi-preparative
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
13
Fig. 3. Time-dependant fluorescence intensity of DNP probes in the presence of sulfhydryl analytes (Cys or ThioPi) and Pi anion in PB buffer (100 mM, pH 7.5) or in PB þ 45% DMSO
(pH 8.3) at 25 ꢂC. (A1) probe 18 (8.6
M) with Cys (50 equiv) in blue, ThioPi (250 equiv) in, lex ¼ 390 nm, lem ¼ 490 nm; (A2) ibid. in PB þ 45% DMSO, a further kinetic curve for Pi
(50 equiv) in green; (B1) probe 19 (8.6
M) with Cys (50 equiv) in blue, ThioPi (250 equiv) in, lex ¼ 390 nm, lem ¼ 490 nm; (B2) ibid. in PB þ 45% DMSO, a further kinetic curve for Pi
(50 equiv) in green; (C1) probe 21 (8.6
m
m
m
M) with Cys (50 equiv) in blue, ThioPi (250 equiv) in red, lex ¼ 390 nm, lem ¼ 490 nm; (C2) ibid. in PB þ 45% DMSO except for ThioPi
(50 equiv.), a further kinetic curve for Pi (50 equiv) in green, lex ¼ 500 nm, lem ¼ 655 nm. Thiolysis of 21 leads to the release of 7-hydroxycoumarin-hemicyanine dye 11 whose
fluorescence emission maximum is highly dependent on the solvent used: 490 nm in PB, 615 nm in PB þ 5% BSA, 655 nm in PB þ 45% DMSO and 705 nm in DMSO. (For inter-
pretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
fluorogenic thiolysis of 21 can be linked to the poor quantum yield
(less than 5%) of the released hemicyanine dye 11 in the mixture
PB þ 45% DMSO (Fig. 3C2). More generally, it is very obvious that
the combined use of our water-solubilizing methodologies and
thiol-reactive quenching moiety DNP is not a relevant approach to
convert 7-hydroxycoumarin scaffolds into water-soluble fluores-
cent probes for detection of biological thiols (such as Cys) in aq.
media but a very promising way to obtain water-soluble “OFFeON”
chemosensors for benzenethiol pollutants whose thiol pKa value is
lower than physiological pH. Indeed, as shown in Fig. 4, a remark-
able enhancement of the rate of DNP ether 18 thiolysis was ob-
tained with 4-chlorothiophenol (pKa ¼ 6.12) [55] compared to
cysteine.
Alternative thiol sensing mechanisms based on quinone-
methide-type rearrangement or tandem cyclization reactions
(“trimethyl lock” cyclization and intramolecular cyclic urea for-
mation) both induced by thiol-Michael addition on a benzoquinone
moiety, were finally implemented through the preparation of flu-
orogenic probes 22b and 23 from the 6-sulfonated fluorophore 1
(Scheme 3). We thus seeked to take advantage of the substantial
distance between the thiol-reactive centre and the bulky fluores-
cent dye to minimize the steric interference and electronic effects
of the water-solubilizing substituent, in order to improve the thi-
olysis kinetics. First, alkylation of phenol 1 with 1-chloromethyl-
2,5-dimethoxy-3,4,6-trimethylbenzene E and subsequent removal
of the two methoxy protecting groups with CAN, were achieved
using literature protocols [30]. Due to the presence of 6-SO3H
substituent, purification of quinone-based probe 22b was carried
out by semi-preprative RP-HPLC but complicated by its limited
stability in all aq. mobile phases tested (i.e., TFA 0.1%, ammonium
formate 50 mM, pH 6.4, ultrapure water and TEAB 50 mM, pH 7.5).
A small amount of 22b was recovered, sufficient for NMR analyses,
but its gradual decomposition giving back the parent fluorescent
phenol 1 prevented us to perform reliable measurements of
quenching efficiency and fluorescence response to thiols. To in-
crease the overall stability of such fluorogenic Michael addition-
based probe, a “cloak trimethyl-lock benzoquinone” unit [59] was
chosen as an alternative to previous benzyl-like group and grafted
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

14
B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
the following main conclusions: (1) DNBS ester is a valuable thiol-
responsive trigger group for long wavelength 7-hydroxycoumarins
bearing a water-solubilizing moiety (other than eSO3H) onto the 8-
position and provide fluorescent probes with favourable features
(i.e., fast response time, significant fluorescence enhancement,
solubility and stability at physiological pH) for biological thiol
detection, (2) DNP ether is a more versatile thiol-reactive quench-
ing moiety than DNBS ester, due to its easier introduction onto the
phenol moiety of
a wide range of 6- or 8-substituted
7-hydroxycoumarins. However, the sensing ability of the water-
soluble DNP pro-fluorophores is limited to thiols completely con-
verted into their thiolate forms in slightly alkaline media (i.e.,
benzenethiols in the pH range 7.5e8.5), and (3) a sulfonic acid
moiety located ortho to the phenol functionality of long-
wavelength 7-hydroxycoumarins hampers the construction of
water-soluble fluorogenic probes that possess both high hydrolytic
stability and reactivity toward biological thiols. Thus, the results
of this study provides the rationale for the design of
7-hydroxycoumarin-based pro-fluorophores with improved pho-
tophysical (i.e., red-shifted absorption/emission and enhanced
quantum yield in water) and physicochemical (i.e., water-solubility)
properties suitable for the sensing and bioimaging of thiol species.
This could be also useful for developing fluorescent probes tar-
geting other nucleophilic (bio)analytes [61e63].
Fig. 4. Time-dependant fluorescence intensity of DNP probe 18 (8.6 mM) in the pres-
ence of sulfhydryl analytes (Cys or 4-chlorothiophenol) in PB þ 45% DMSO (pH 8.3) at
25 ꢂC (lex ¼ 390 nm, lem ¼ 490 nm). (red) Cys (50 equiv); (brown) 4-chlorothiophenol
(50 equiv) and (orange) 4-chlorothiophenol (5 equiv). Thiolysis of probe 18 with
50 equiv of 4-chlorothiophenol leads to fluorescence signal saturation (>1000 units). (For
interpretation of the references to colour in this figure legend, the reader is referred to
the web version of this article.)
to the phenol moiety of 1 through a diamine spacer involving N-
methylcarbamate linkages. The O-acylation of 1 with carbamoyl
chloride derivative of F (pre-formed by reaction of N-methyl sec-
ondary amine F and phosgene) was achieved under conditions
previously reported by us for the synthesis of protease-sensitive
fluorogenic probes using 7-hydroxycoumarin as reporter group
[60]. The resulting quinone-based probe 23 was purified by semi-
preparative RP-HPLC but again, the loss of quenching moiety
leading to the premature release of fluorophore 1 partially occurred
in aq. mobile phase. Full characterization of this water-soluble
fluorogenic probe by mass spectrometry and NMR spectroscopy
has however been made but its limited stability in aq. buffers
prevented us to reliably assess its thiol sensing ability through
fluorescence measurements in phosphate buffer. It seems therefore
that the ortho-substituent SO3H favours the hydrolysis of O-aryl
benzyl ether and carbamate linkages and prevents the design of
latent fluorophores that possess a unique combination of chemical
stability and thiol reactivity. This may be related to a better leaving
group ability of 6-sulfonated coumarin 1 as a result of lower pKa of
its phenol moiety compared to parent 3-benzothiazolyl-7-
hydroxycoumarin (vide supra).
Acknowledgements
This work was supported by the “ Région Haute-Normandie”
(especially for the Ph. D. grant of Benoît Roubinet), the CRUNCh
program (CPER 2007e2013) and “ Institut Universitaire de France”
(IUF). We thank Dr. Michel Picquet (University of Burgundy, ICMUB,
UMR CNRS 6302) for its helpful assistance for achieving NMR an-
alyses on Bruker Avance III 500 spectrometer and Patricia Martel
(University of Rouen, laboratory COBRA, UMR CNRS 6014) for the
recording of IR spectra. Dr. Nicolas Maindron is also greatly
acknowledged for the synthesis of di-tert-butyl-iminodiacetate.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.dyepig.2014.02.004.
References
[1] Peng H, Chen W, Cheng Y, Hakuna L, Strongin R, Wang B. Thiol reactive probes
and chemosensors. Sensors 2012;12:15907e46.
[2] Roy K-M. Thiols and organic sulfides. In: Ullmann’s encyclopedia of industrial
chemistry. Weinheim: Wiley-VCH; 2012. pp. 629e55.
4. Conclusion
In summary, a series of water-soluble fluorophores based on the
3-benzothiazolyl-7-hydroxycoumarin scaffold were successfully
synthesized by means of electrophilic aromatic sulfonation or
Mannich- and Knoevenagel-type reactions with hydrophilic
building blocks. Unprecedented high fluorescence quantum yields
in physiological conditions were obtained for the majority of
compounds, especially those emitting in the blue-green spectral
range. To the best of our knowledge, the present work is the first
study devoted to the water-solubilization of long-wavelength 7-
hydroxycoumarins, now regarded as valuable tools in the field of
fluorogenic probes. The masking of 7-OH group of these fluorescent
organic dyes with various thiol-sensitive protecting groups acting
as quenching moieties, was also investigated and ten novel water-
soluble pro-fluorophores potentially usable for thiol detection in
the visible region (485e631 nm), were synthesized. A comparative
study focused on their ease of synthesis, chemical stability and thiol
sensing ability under physiological conditions has led us to make
[3] Hathaway GJ, Proctor NH. Proctor and Hugues’ chemical hazards of the
workplace. 5th ed. Hoboken: John Wiley & Sons, Inc.; 2004.
[4] Chen X, Zhou Y, Peng X, Yoon J. Fluorescent and colorimetric probes for
detection of thiols. Chem Soc Rev 2010;39:2120e35.
[5] Li H, Cai L, Chen Z. Coumarin-derived fluorescent chemosensors. In: Advances
in chemical sensors. InTech; 2012. pp. 121e50.
[6] Jung HS, Chen X, Kim JS, Yoon J. Recent progress in luminescent and colori-
metric chemosensors for detection of thiols. Chem Soc Rev 2013;42:6019e31.
[7] Yin C, Huo F, Zhang J, Martinez-Manez R, Yang Y, Lv H, et al. Thiol-addition
reactions and their applications in thiol recognition. Chem Soc Rev 2013;42:
6032e59.
[8] Wang L, Zhou Q, Zhu B, Yan L, Ma Z, Du B, et al. A colorimetric and fluorescent
chemodosimeter for discriminative and simultaneous quantification of
cysteine and homocysteine. Dyes Pigm 2012;95:275e9.
[9] Zeng X, Zhang X, Zhu B, Jia H, Li Y. A highly selective wavelength-ratiometric
and colorimetric probe for cysteine. Dyes Pigm 2012;94:10e5.
[10] Dong W, Wen H, Yang X-F, Li H. Highly selective and sensitive fluorescent
sensing of N-acetylcysteine: effective discrimination of N-acetylcysteine from
cysteine. Dyes Pigm 2013;96:653e8.
[11] Huang Q, Yang X-F, Li H. A ratiometric fluorescent probe for hydrogen sulfide
based on an excited-state intramolecular proton transfer mechanism. Dyes
Pigm 2013;99:871e7.
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004

B. Roubinet et al. / Dyes and Pigments xxx (2014) 1e15
15
[12] Li J, Zhang C-F, Ming Z-Z, Yang W-C, Yang G-F. Novel coumarin-based sensi-
tive and selective fluorescent probes for biothiols in aqueous solution and in
living cells. RSC Adv 2013;3:26059e65.
[13] Long L, Zhou L, Wang L, Meng S, Gong A, Du F, et al. A coumarin-based
fluorescent probe for biological thiols and its application for living cell im-
aging. Org Biomol Chem 2013;11:8214e20.
[14] Shiraishi Y, Yamamoto K, Sumiya S, Hirai T. Selective fluorometric detection of
aromatic thiols by a chemosensor containing two electrophilic sites with
different local softness. Chem Commun 2013;49:11680e2.
[15] Wang H, Zhou G, Mao C, Chen X. A fluorescent sensor bearing nitroolefin
moiety for the detection of thiols and its biological imaging. Dyes Pigm
2013;96:232e6.
[16] Liu T, Xu Z, Spring DR, Cui J. A lysosome-targetable fluorescent probe for
imaging hydrogen sulfide in living cells. Org Lett 2013;15:2310e3.
[17] Liu T, Zhang X, Qiao Q, Zou C, Feng L, Cui J, et al. A two-photon fluo-
rescent probe for imaging hydrogen sulfide in living cells. Dyes Pigm
2013;99:537e42.
[18] Liu Y, Feng G. A visible light excitable colorimetric and fluorescent ESIPT probe
for rapid and selective detection of hydrogen sulfide. Org Biomol Chem
2014;12:438e45.
[37] Fulmer GR, Miller AJM, Sherden NH, Gottlieb HE, Nudelman A, Stoltz BM, et al.
NMR chemical shifts of trace impurities: common laboratory solvents, or-
ganics, and gases in deuterated solvents relevant to the organometallic
chemist. Organometallics 2010;29:2176e9.
[38] Setsukinai K-I, Urano Y, Kikuchi K, Higuchi T, Nagano T. Fluorescence
switching by O-dearylation of 7-aryloxycoumarins. Development of novel
fluorescence probes to detect reactive oxygen species with high selectivity.
J Chem Soc Perkin Trans 2000;2:2453e7.
[39] Brouwer AM. Standards for photoluminescence quantum yield measurements
in solution (IUPAC technical report). Pure Appl Chem 2011;83:2213e28.
[40] Determined by us in PB at 25 ꢂC using 7-hydroxycoumarin as a standard.
[41] Lin W, Long L, Tan W. A highly sensitive fluorescent probe for detection of
benzenethiols in environmental samples and living cells. Chem Commun
2010;46:1503e5.
[42] Huang HC, Wang KL, Huang ST, Lin HY, Lin CM. Development of a sensitive
long-wavelength fluorogenic probe for nitroreductase: a new fluorimetric
indictor for analyte determination by dehydrogenase-coupled biosensors.
Biosens Bioelectron 2011;26:3511e6.
[43] Moon JO, Lee JW, Choi MG, Ahn S, Chang S-K. Dual signaling of hypochlorous
acid by desulfurization of thiocoumarin. Tetrahedron Lett 2012;53:6594e7.
[44] Mahoney KM, Goswami PP, Winter AH. Self-immolative aryl phthalate esters.
J Org Chem 2013;78:702e5.
[19] Shi J, Wang Y, Tang X, Liu W, Jiang H, Dou W, et al. A colorimetric and fluo-
rescent probe for thiols based on 1, 8-naphthalimide and its application for
bioimaging. Dyes Pigm 2014;100:255e60.
[45] Abrams B, Diwu Z, Guryev O, Aleshkov S, Hingorani R, Edinger M, et al. 3-
[20] Youziel J, Akhbar AR, Aziz Q, Smith MEB, Caddick S, Tinker A, et al. Bromo- and
thiomaleimides as a new class of thiol-mediated fluorescence ‘turn-on’ re-
agents. Org Biomol Chem 2014;12:557e60.
[21] Liu J, Sun Y-Q, Huo Y, Zhang H, Wang L, Zhang P, et al. Simultaneous fluo-
rescence sensing of cys and GSH from different emission channels. J Am Chem
Soc 2014;136:574e7.
Carboxy-6-chloro-7-hydroxycoumarin: a highly fluorescent, water-soluble
violet-excitable dye for cell analysis. Anal Biochem 2009;386:262e9.
[46] Louise-Leriche L, Paunescu E, Saint-Andre G, Baati R, Romieu A, Wagner A,
et al. A HTS assay for the detection of organophosphorus nerve agent scav-
engers. Chem Eur J 2010;16:3510e23.
[47] Romieu A, Bruckdorfer T, Clavé G, Grandclaude V, Massif C, Renard P-Y. N-
[22] Yuan L, Lin W, Zheng K, He L, Huang W. Far-red to near infrared analyte-
responsive fluorescent probes based on organic fluorophore platforms for
fluorescence imaging. Chem Soc Rev 2013;42:622e61.
Fmoc-a-sulfo-b-alanine: a versatile building block for the water solubilization
of chromophores and fluorophores by solid-phase strategy. Org Biomol Chem
2011;9:5337e42.
[23] Guo Z, Park S, Yoon J, Shin I. Recent progress in the development of near-
infrared fluorescent probes for bioimaging applications. Chem Soc Rev
2014;43:16e29.
[48] Massif C, Dautrey S, Haefele A, Ziessel R, Renard P-Y, Romieu A. New insights
into the water-solubilisation of fluorophores by post-synthetic “click” and
Sonogashira reactions. Org Biomol Chem 2012;10:4330e6.
[24] Li X, Gao X, Shi W, Ma H. Design strategies for water-soluble small molecular
chromogenic and fluorogenic probes. Chem Rev 2014;114:590e659.
[25] Wolfbeis OS, Koller E, Hochmuth P. The unusually strong effect of a 4-cyano
group upon electronic spectra and dissociation constants of 3 substituted 7-
Hydroxy coumarins. Bull Chem Soc Jpn 1985;58:731e4.
[26] Achilefu SI, Srinivasan A. Methods for incorporating metal chelators at
carboxyl-terminal site of peptides, Mallinckrodt Inc., WO2001052898.
[27] Khilya O,V, Frasinyuk MS, Turov AV, Khilya VP. Chemistry of 3-
hetarylcoumarins. 1. 3-(2-benzazolyl)coumarins. Chem Heterocycl Comp
2001;37:1029e37.
[49] Adamczyk M, Cornwell M, Huff J, Rege S, Rao SVT. Novel 7-hydroxycoumarin
based fluorescent labels. Bioorg Med Chem Lett 1997;7:1989e98.
[50] Hagen V, Kilic F, Schaal J, Dekowski B, Schmidt R, Kotzur N. [8-[Bis(carbox-
ymethyl)aminomethyl]-6-bromo-7-hydroxycoumarin-4-yl]methyl moieties
as photoremovable protecting groups for compounds with COOH, NH2, OH,
and C¼O functions. J Org Chem 2010;75:2790e7.
[51] Kisin-Finfer E, Shabat D. New repertoire of ‘donor-two-acceptor’ NIR fluoro-
genic dyes. Bioorg Med Chem 2013;21:3602e8.
[52] Bharate SB. 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ). Synlett; 2006:
496e7.
[28] Deligeorgiev T, Tsvetkova T, Ivanova D, Timtcheva I. Synthesis and electronic
spectra of 3-hetaryl substituted coumarin derivatives 7-hydroxy-2H-chro-
men-2-one and 9-hydroxy-2H-benzo[f]chromen-2-one. Color Technol
2008;124:195e203.
[29] Huang ST, Peng YX, Wang KL. Synthesis of a new long-wavelength latent
fluorimetric indicator for analytes determination in the DT-diaphorase
coupling dehydrogenase assay system. Biosens Bioelectron 2008;23:
1793e8.
[53] Richard J-A, Massonneau M, Renard P-Y, Romieu A. 7-Hydroxycoumarin-
hemicyanine hybrids: a new class of far-red emitting fluorogenic dyes. Org
Lett 2008;10:4175e8.
[54] Yang Z, Liu Z, Chen Y, Wang X, He W, Lu Y. A new ratiometric and colorimetric
chemosensor for cyanide anion based on coumarin-hemicyanine hybrid. Org
Biomol Chem 2012;10:5073e6.
[55] Calculated using Advanced Chemistry Development (ACD/Labs) Software
V11.02 (Ó 1994-2013 ACD/Labs).
[30] Huang ST, Ting KN, Wang KL. Development of a long-wavelength fluorescent
probe based on quinone-methide-type reaction to detect physiologically
significant thiols. Anal Chim Acta 2008;620:120e6.
[56] Okuda H, Fujimoto T, Kawamura M, Hirose S. Application of complementary
tristimulus colorimetry to the simultaneous determination of phenol de-
rivatives. Bunseki Kagaku 1974;23:644e9.
[31] Gromov SP, Vedernikov AI, Kuz’mina LG, Kondratuk DV, Sazonov SK,
Strelenko YA, et al. Photocontrolled molecular assembler based on cucurbit[8]
uril: [2þ2]-Autophotocycloaddition of styryl dyes in the solid state and in
water. Eur J Org Chem 2010;2010:2587e99.
[57] Pauli J, Grabolle M, Brehm R, Spieles M, Hamann FM, Wenzel M, et al. Suitable
labels for molecular imaging - influence of dye structure and hydrophilicity on
the spectroscopic properties of IgG conjugates. Bioconjug Chem 2011;22:
1298e308.
[32] Romieu A, Tavernier-Lohr D, Pellet-Rostaing S, Lemaire M, Renard P-Y. Water
solubilization of xanthene dyes by post-synthetic sulfonation in organic me-
dia. Tetrahedron Lett 2010;51:3304e8.
[58] Li M, Wu X, Wang Y, Li Y, Zhu W, James TD. A near-infrared colorimetric
fluorescent chemodosimeter for the detection of glutathione in living cells.
Chem Commun 2014;50:1751e3.
[33] Deligeorgiev T, Tsvetkova T, Stanimirov S. Synthesis of novel water-soluble
fluorescent coumarins bearing sulphonic and hydroxyl groups. Color Tech-
nol 2011;127:434e9.
[34] Maki H, Ueda Y, Nariai H. Protonation equilibria and stepwise hydrolysis
behavior of a series of thiomonophosphate anions. J Phys Chem B 2011;115:
3571e7.
[59] Levine MN, Raines RT. Trimethyl lock: a trigger for molecular release in
chemistry, biology, and pharmacology. Chem Sci 2012;3:2412e20.
[60] Meyer Y, Richard J-A, Delest B, Noack P, Renard P-Y, Romieu A. A comparative
study of the self-immolation of para-aminobenzylalcohol and hemithioaminal-
based linkers in the context of protease-sensitive fluorogenic probes. Org Bio-
mol Chem 2010;8:1777e80.
[35] Romieu A, Massif C, Rihn S, Ulrich G, Ziessel R, Renard P-Y. The first
comparative study of the ability of different hydrophilic groups to water-
solubilise fluorescent BODIPY dyes. New J Chem 2013;37:1016e27.
[36] Shafeekh KM, Rahim MKA, Basheer MC, Suresh CH, Das S. Highly selective and
sensitive colourimetric detection of Hg2þ ions by unsymmetrical squaraine
dyes. Dyes Pigm 2013;96:714e21.
[61] Eun Jun M, Roy B, Han Ahn K. “Turn-on” fluorescent sensing with “reactive”
probes. Chem Commun 2011;47:7583e601.
[62] Chan J, Dodani SC, Chang CJ. Reaction-based small-molecule fluorescent
probes for chemoselective bioimaging. Nat Chem 2012;4:973e84.
[63] Shi W, Ma H. Spectroscopic probes with changeable
p-conjugated systems.
Chem Commun 2012;48:8732e44.
Please cite this article in press as: Roubinet B, et al., New insights into the water-solubilization of thiol-sensitive fluorogenic probes based on
long-wavelength 7-hydroxycoumarin scaffolds, Dyes and Pigments (2014), http://dx.doi.org/10.1016/j.dyepig.2014.02.004