Turning on fluorescence with thiols-synthetic and computational studies on diaminoterephthalates and monitoring the switch of the Ca2+ sensor recoverin

Article abstract of DOI:10.1002/ejoc.201200879

The fluorescence of maleimide-functionalized diaminoterephthalate derivatives (NiWa Orange) is turned on by the conjugate addition of thiols. Three new representatives of this class of dyes with emission at 560 nm were synthesized from succinyl succinates. The mechanism of turning on the fluorescence was investigated by computational studies. Radiationless transition to an energetically low-lying excited state by a nonadiabatic interaction was the mechanism leading to absence of fluorescence prior to the reaction of the probe with its molecular target. This concept was proven by labeling cysteine-containing peptides and proteins. For example, the single cysteine residue at position 39 of the neuronal calcium sensor protein recoverin binds to NiWa Orange and turns on its fluorescence. The Ca²⁺-dependent Foerster resonance energy transfer (FRET) process from a tryptophan residue in proximity to Cys39-bound NiWa Orange was used for the determination of differences in Ca²⁺-binding affinities of myristoylated and nonmyristoylated recoverin.

Full text of DOI:10.1002/ejoc.201200879

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
FULL PAPER
DOI: 10.1002/ejoc.201200879
Turning On Fluorescence with Thiols – Synthetic and Computational Studies
on Diaminoterephthalates and Monitoring the Switch of the Ca2+ Sensor
Recoverin
Nina Wache,[a] Alexander Scholten,[b] Thorsten Klüner,[a] Karl-Wilhelm Koch,[b] and
Jens Christoffers*[a]
Keywords: Fluorescent probes / Sensors / Calcium
The fluorescence of maleimide-functionalized diamino-
terephthalate derivatives (NiWa Orange) is “turned on” by
the conjugate addition of thiols. Three new representatives
of this class of dyes with emission at 560 nm were synthe-
sized from succinyl succinates. The mechanism of turning on
the fluorescence was investigated by computational studies.
Radiationless transition to an energetically low-lying excited
state by a nonadiabatic interaction was the mechanism lead-
ing to absence of fluorescence prior to the reaction of the
probe with its molecular target. This concept was proven by
labeling cysteine-containing peptides and proteins. For ex-
ample, the single cysteine residue at position 39 of the neu-
ronal calcium sensor protein recoverin binds to NiWa Orange
and turns on its fluorescence. The Ca2+-dependent Förster
resonance energy transfer (FRET) process from a tryptophan
residue in proximity to Cys39-bound NiWa Orange was used
for the determination of differences in Ca2+-binding affinities
of myristoylated and nonmyristoylated recoverin.
Introduction
The fluorescent labeling of proteins is a powerful and
reliable tool in biochemistry, medicine, and biology, al-
lowing for the fast and precise investigation of the dynam-
ics, interactions, and localizations of proteins.[1] Several la-
beling techniques use fluorescent dyes with a reactive func-
tional group such as the maleimide moiety, which is well
known to react with protein surfaces holding thiol func-
tions.[2] A prominent example of such a dye is the fluores-
cein derivative 1 (Scheme 1).[3] There is a class of com-
pounds that are not luminescent themselves, but their fluo-
rescence is “turned on” upon the conjugate addition of a
thiol to the maleimide function of the dye.[4] The boron di-
pyrromethene (BODIPY) derivative 2 is an example of such
a so-called turn-on probe.[5]
Scheme 1. Thiol-reactive fluorescence probes with a maleimide
moiety. Whereas fluorescein derivative 1 shows luminescence even
without addition of a thiol, 2 and 3 are nonemissive; their fluores-
cence is turned on by the conjugate addition of a thiol, as is shown
for the formation of 4 from dye 3 (NiWa Blue).
The preparation of diaminoterephthalates, which have
been known for almost 100 years,[6] starts from succinyl suc-
cinates[7] and primary amines and is therefore relatively sim-
ple. However, these dyes are hardly found in the literature,[8]
although they are brilliant dyes with pronounced fluores-
[a] Institut für Chemie, Carl von Ossietzky Universität,
26111 Oldenburg, Germany
cence behavior. In continuation of our work on diaminote-
rephthalates as fluorescent dyes,[9] we have recently reported
on the preparation of the thiol-reactive turn-on fluorescent
probe 3 (NiWa Blue). Upon reaction with BnSH, the prod-
uct 4 showed fluorescence at 391 nm when excited at
338 nm (Scheme 1).[10] Its utility as a tool in biochemistry
was proven by staining of the neuronal calcium sensor re-
Fax: +49-441-798-3873
E-mail: jens.christoffers@uni-oldenburg.de
Homepage: www.christoffers.chemie.uni-oldenburg.de
[b] Institut für Biologie und Umweltwissenschaften, Carl von
Ossietzky Universität,
26111 Oldenburg, Germany
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201200879.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
J. Christoffers et al.
FULL PAPER
coverin, which plays a major role in the photoreceptor func- ethylenediamine[15] by a two-step procedure (Scheme 2).
tion in vertebrate retina.[11] Recoverin possesses four EF- First of all, the double enamine was formed, with removal
hand motifs as Ca2+ binding sites, of which only the second of water from the reaction mixture. Secondly, oxidation was
and third are functional. One Cys residue lies in position achieved with careful exclusion of moisture by using syn-
39 on the surface of this protein and is therefore susceptible thetic air and a solution of HCl in 2-propanol as an addi-
to reactions with maleimide.[12] The Ca2+-free and the Ca2+
-
tive. Product 6 (Scheme 2) was obtained (71% yield) as a
bound state of recoverin adopt two significantly different red crystalline solid (λabs = 469 nm, λem = 569 nm, Φ =
conformations that can be tracked by biochemical and bio- 0.26). The quantum yield was estimated by comparison
physical methods. For example, the thiol-sensitive fluores- with fluorescein as a standard.[16] Both carbamate protec-
cent dye 3 was recently employed to monitor Ca2+-induced tive groups were removed with trifluoroacetic acid (TFA),
conformational changes in recoverin.[10] This dye is turned either neat or diluted with CH2Cl2, and the material was
on by covalent attachment to recoverin and shows fluores- subsequently converted with maleic anhydride at elevated
cence emission at about 440 nm irrespective of the presence temperature in AcOH as solvent. In contrast to our earlier
of Ca2+. However, Ca2+-induced conformational changes in results with the preparation of 3, we were not able to pre-
recoverin lead to intensity changes in Trp fluorescence emis- vent double imide formation. Moreover, the mono- and bis-
sion, which by a Förster resonance energy transfer (FRET) imides could not be separated, not even after capping of
process can excite the bound NiWa Blue dye. In the absence the remaining amino group as an acetamide group. There-
of Ca2+, a Trp residue is in proximity to Cys39 and its con- fore, we used an excess of maleic anhydride to obtain the
jugated NiWa Blue. Upon excitation of the Trp chromo- bisimide 7 in 32% yield on a quarter-gram-scale if neat TFA
phore at 280 nm FRET occurs resulting in fluorescence was used for the deprotection. If TFA was diluted with
emission of NiWa Blue at ca. 440 nm. In the presence of CH2Cl2, the monoacetoxylated compound 8 (9% yield) was
Ca2+, recoverin has a protein conformation with a larger formed as a byproduct. Both products 7 and 8 could be
distance between the Trp and Cys residues. The FRET pro- separated by repeated chromatographies, therefore, the
cess is therefore significantly weakened and a Trp emission yields (15% of 7 and 9% of 8) were low. In contrast to their
at 340 nm becomes more pronounced.[10] The emission of starting material 6, compounds 7 (λabs = 462 nm) and 8
compound 4 is in the blue region of the visible spectrum (λabs = 446 nm) showed no fluorescence. The emission of
(λex = 338 nm, λem = 391 nm in CH2Cl2, λem = 440 nm on the fluorophores seemed to be quenched by the maleimide
the surface of the protein and in buffered solution) with a moieties.
matrix-dependent Stokes shift of up to 100 nm. In order
to avoid excitation of the NiWa dye (λex = 338 nm) when
irradiating into proximal wavelengths (e.g. at 280 nm for
Trp excitation or other aromatic biomolecules, like cofac-
tors, e.g. NADH),[1a] we were aiming for a more redshifted
dye and wish to report herein on our results of the develop-
ment of a NiWa Orange turn-on fluorescence probe for
thiols.
There are two key issues combined in our diamino-
terephthalate system, which are not fulfilled by so far
known dyes: (1) the turn on of fluorescence by reaction with
a thiol and (2) the chromophore itself defines a molecular
scaffold with four points of diversification ready for further
functionalization. Besides the labeling of recoverin, which
we report herein, we are therefore aiming to use NiWa dyes
for other applications in the future, e.g. for probing the
switch of another Ca2+ sensor, GCAP2.[13]
Results and Discussion
Organic Synthesis
Scheme 2. Preparation of NiWa Orange I 7 and NiWa Orange II
8. Reagents and conditions: (a) 1. 4 equiv. BocHNCH2CH2NH2,
cat. AcOH, toluene, Dean–Stark trap, 110 °C, 16 h; 2. O2 (synthetic
air), cat. HCl, iPrOH, N,N-dimethylformamide (DMF), 50 °C, 5 h;
(b) 1. TFA, 23 °C, 0.5 h; 2. 4.0 equiv. maleic anhydride, AcOH,
95 °C, 3 d; 32% of product 7, no compound 8 obtained; (c) 1. TFA,
CH2Cl2, 23 °C, 1 h; 2. 4.0 equiv. maleic anhydride, AcOH, 90 °C,
In order to generate redshifted fluorescence wavelengths
by electronic decoupling of the donor and acceptor, we de-
cided to introduce an ethylene spacer between the di-
aminoterephthalate and maleimide part of the compound.
For this reason, diketone 5[14] was converted with N-Boc- 2 d; 15% of product 7, 9% of product 8.
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
Turning On Fluorescence with Thiols
In order to test its utility as a turn-on probe for thiols,
NiWa Orange II 8 was converted with the cysteine-contain-
ing peptide glutathione (GSH) on an analytical scale in di-
methyl sulfoxide (DMSO, Scheme 3). Whereas 8 itself
showed no fluorescence, an orange emission (λem = 568 nm)
appears after a couple of minutes, when exciting at λex
=
469 nm. The experiment was performed on an analytical
scale, therefore, product 9 was not isolated, but its [M +
H]+ signal could be clearly detected by ESI-HRMS.
Scheme 3. Turn on of NiWa Orange II 8 with glutathione (GSH).
Experiment on an analytical scale.
When treating NiWa Orange I 7 with an excess of BnSH,
an orange fluorescence is turned on (Scheme 4). The double
addition product 10 was formed and isolated on a prepara-
tive scale and spectroscopically fully characterized (λex
=
467 nm, λem = 559 nm, Φ = 0.80; Table 1). In order to ob-
tain a compound that is turned on by only one equivalent
of thiol, we aimed to prepare the mono addition product
with BnSH, however, we failed, because the mono- and bis-
addition products were not separable. Therefore, we applied
hydrophobic n-hexadecanethiol for formation of the mono-
addition product 11, and indeed, it was now separable by
column chromatography from remaining starting material
7 and bis-addition product 12. In a typical experiment with
0.5 equiv. n-C16H33SH, products 11 and 12 were isolated
in 11% and 8% yield, respectively, and compound 7 was
reisolated in 19% yield. Double addition product 12 again
showed strong fluorescence (λex = 467 nm, λem = 561 nm,
Φ = 0.77). The monothiol addition product 11 showed no
fluorescence, as one maleimide moiety remained un-
touched. It is therefore an ideal probe for detecting sub-
stoichiometric quantities of thiols and was named NiWa
Scheme 4. Derivatization and turn on of NiWa Orange I 7. Rea-
gents and conditions: (a) Excess BnSH, NEt3, CH2Cl2, 23 °C, 1 d,
87%; (b) 0.5 equiv. nC16H33SH, 4 equiv. NEt3, CH2Cl2, 23 °C, 1 d;
19% starting material 7, 11% mono adduct 11 (NiWa Orange III),
8% bis adduct 12; (c) Excess GSH, DMSO, 23 °C, 20 min (analyti-
Orange III. The fluorescence of 11 can be turned on with cal scale).
glutathione (GSH) and was again investigated on an analyt-
ical scale in DMSO. As observed for NiWa Orange II, an
Computational Chemistry
orange emission (λem = 565 nm) appears after a couple of
minutes, when exiting at λex = 469 nm. Product 13 was not
In order to shed light on the mechanistic details of the
isolated, but it was detectable by ESI-HRMS ([M + H]+ fluorescence properties of NiWa Blue and NiWa Orange,
signal). Compounds 10–12 were obtained on a preparative in particular on the process of fluorescence quenching of
scale and fully characterized. Although possessing two compounds with a maleimide moiety, we performed a theo-
stereocenters and therefore existing as two diastereoiso- retical analysis based on time-dependent density functional
mers, both 10 and 12 show only single signal sets of reso- theory
nances in their NMR spectra.
(TDDFT)
using
the
program
package
Gaussian09.[17] Throughout the study the B3LYP func-
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
J. Christoffers et al.
FULL PAPER
Table 1. Spectroscopic properties of compounds 6–13.
Compound
Solvent
c [10–4 moldm–3]
λmax [nm]
ε [dm3 mol–1 cm–1]
λem [nm]
λex [nm]
Φ
6
7
CH2Cl2
2.73
2.34
42
3.20
2.45
2.23
1.65
1.52
1.25
469
462
445
446
469
467
468
467
471
5600
290
569
–
–
469
–
–
0.26
–
–
CH2Cl2
aqueous buffer[a]
CH2Cl2
1800
6700
6700
3400
2500
4200
2100
8
9
10
11
12
13
–
–
–
[b]
DMSO
CH2Cl2
CH2Cl2
CH2Cl2
568
559
–
561
565
469
467
–
467
469
0.80
–
0.77
[b]
DMSO
[a] Water/DMSO, 50:1, phosphate-buffered saline (PBS; 130 nM NaCl, 70 mm Na2HPO4, 30 mm NaH2PO4; pH = 7.4–7.5). [b] Experi-
ments were performed on an analytical scale, no quantum yield was determined.
tional for exchange and correlation was chosen, and a 6-
31+G* basis set turned out to be appropriate for suffi-
ciently accurate calculations. Firstly, a full geometry optimi-
zation of all compounds was performed in the electronic
ground state. Using the optimized geometries, we calculated
vertical excitation energies and oscillator strengths for the
three lowest excited states by TDDFT. The geometry of the
excited state with the largest oscillator strength was opti-
mized subsequently, and the vertical transition energy to
the ground state using the excited state optimized geometry
was calculated in order to obtain the fluorescence wave-
length.
Firstly, we investigated the properties of 14,[10] which ex-
perimentally showed an orange-red fluorescence (λex
=
434 nm, λem = 535 nm) as a benchmark test for our compu-
tations. Within the computational strategy outlined above,
the lowest excited state shows a strong highest occupied
molecular orbital (HOMO) Ǟ lowest unoccupied molecu-
lar orbital (LUMO) transition with an oscillator strength
of f = 0.1034. The calculated absorption and emission wave-
lengths (λex = 454 nm, λem = 544 nm, see Table 2 for all
calculated data) were in very good agreement with the ex-
perimental values; this demonstrates the reliability of our
computational approach (TD-B3LYP/6-31+G*) (Figure 1).
Figure 1. Compounds subjected to theoretical calculations. Dye 14
is the benchmark test system for the computational methods. Com-
pounds 15 and 16 are structurally simplified models for 3 and its
thiol addition product 4. Compounds 17 and 18 are structurally
simplified models for the NiWa Orange dyes 7, 8, and 11 and their
thiol addition products.
Our calculations clearly indicate that 15 and 16 exhibit
very similar excitation spectra. Compound 15 shows a
strong excitation (λex = 360 nm, f = 0.0896) very similar
to that of 16 (λex = 357 nm, f = 0.0889) and close to the
experimental values of 3 and 4 (both with λex = 333 nm).
However, regardless of this striking similarity of the ab-
sorption spectra, a crucial difference between the excited
states manifolds of 15 and 16 was observed. In 16 the lowest
excited state shows the largest oscillator strength, whereas
in 15 a dark state exists, which is lower in energy as com-
pared to the transition reported in the last paragraph but
exhibits a very low oscillator strength (λex = 426 nm, f =
0.0008). Therefore, we expect a similar excitation spectrum
for both 15 and 16, but the fate of the optically excited
species will be entirely different.
Table 2. Calculated excitation wavelength, emission wavelength,
and oscillator strengths.
Transition
λex [nm]
λem [nm]
f
14
15
HOMO Ǟ LUMO
(πǞπ*)
HOMO Ǟ LUMO
HOMO Ǟ LUMO+1
(πǞπ*)
454 (434)[a] 544 (535) 0.1034
[a]
426
360
–
–
0.0008
0.0896
16
17
HOMO Ǟ LUMO
(πǞπ*)
HOMO Ǟ LUMO
HOMO Ǟ LUMO+1
(πǞπ*)
357
400
0.0889
807
470
–
–
0.0001
0.1207
18
HOMO Ǟ LUMO
(πǞπ*)
472
558
0.1245
After Franck–Condon excitation, 16 will geometrically
rearrange because the nuclear wavefunction is not an eigen-
[a] Experimental value in brackets.
In order to understand the absence of fluorescence of dye state of the electronically excited potential energy surface
3 (NiWa Blue) and the blue fluorescence of 4 after conju- (PES). Following the gradients of the excited state PES, the
gate addition of a thiol, we performed corresponding calcu- global minimum of the excited state might be reached
lations using 15 and 16 as simplified models.
eventually, which results in the expected fluorescence at this
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
Turning On Fluorescence with Thiols
geometry. Our calculations predict a rather strong emission aminoterephthalate moiety to the maleimide unit. The cor-
with a wavelength of λem = 400 nm and an oscillator responding orbitals are displayed in Figure 3.
strength of f = 0.0769 in excellent agreement with experi-
mental data for compound 4 (λem = 391 nm).
On the other hand, in compound 15 the energetically
low-lying dark excited state will open an alternative route
for the relaxation of the electronic excitation, which is much
faster than the fluorescence of species 16 and therefore
more efficient. Due to nonadiabatic coupling through a
conical intersection, the system will undergo a radiationless
transition to the dark state, from which an electronic relax-
ation is optically forbidden. Therefore, no fluorescence
should be observed for NiWa Blue compound 15, which is
in perfect agreement with experimental observations.
In order to understand the differences in the excitation
spectra of 15 and 16 on an atomistic level, we investigated
the nature of the excited states involved in a one-particle
(orbital) picture. Nevertheless, a complete mechanistic un-
derstanding would also require us to locate and to charac-
terize the conical intersection responsible for the radiation-
less decay of 15 after optical excitation. However, this kind
of theoretical investigation was beyond the scope of the
present study, which predominantly focuses on a basic
mechanistic understanding of the experimental results.
Along these lines, we investigated the nature of the lowest
excited states for 15 and 16 to illustrate their different fluo-
rescence behavior. In 16, the lowest electronically excited
state can be characterized as an optically allowed πǞπ*
transition from the HOMO to the LUMO predominantly
localized at the diaminoterephthalate moiety. Optical exci-
tation as well as fluorescence involves a transition between
these orbitals within the limits of a one-particle approxi-
mation. The corresponding molecular orbitals are displayed
in Figure 2.
Figure 3. HOMO (left) and LUMO (right) of compound 15. The
lowest excited state corresponds to an optically forbidden charge-
transfer transition from the HOMO to the LUMO.
As illustrated in Figure 3, the energetically low-lying
charge-transfer excitation of 15 is not accessible by optical
excitation due to a negligible HOMO–LUMO Franck–Con-
don orbital overlap, which results in a very small oscillator
strength (f = 0.0008). However, this state plays the crucial
role in quenching the fluorescence of 15 by nonadiabatic
relaxation.
A similar conclusion can be drawn concerning the
fluorescence properties of 17 and 18 as simplified models
for the NiWa Orange dyes 7, 8, 11, and their thiol addition
products. Again, the absorption spectra of 17 and 18 turn
out to be very similar. Compound 17 shows a strong exci-
tation (λex = 470 nm, f = 0.1207) very similar to that of 18
(λex = 472 nm, f = 0.1245) and close to the experimental
values of 6–13 (λex = 445–471 nm, cf. Table 1). The excited
states involved can again be characterized as optically al-
lowed πǞπ* transitions predominantly localized at the di-
aminoterephthalate moiety. The corresponding orbitals of
18 are displayed in Figure 4.
Figure 2. HOMO (left) and LUMO (right) of compound 16. The
lowest excited state corresponds to an optically allowed πǞπ* tran-
sition from the HOMO to the LUMO.
Figure 4. HOMO (left) and LUMO (right) of 18. The lowest excited
state corresponds to an optically allowed πǞπ* transition from the
HOMO to the LUMO.
Similar to 16, the NiWa Blue compound 15 exhibits an
optically allowed πǞπ* transition from the HOMO to the
LUMO+1 as the second electronically excited state of the
For 18, this πǞπ* transition is involved in the corre-
system. However, the first electronically excited, dark state sponding fluorescence, which is predicted at a wavelength
responsible for a radiationless decay in 15 can be charac- of λem = 558 nm with an oscillator strength of f = 0.1163
terized as a HOMO to LUMO transition to a charge-trans- in excellent agreement with the experimental values (λem
=
fer state, in which an electron is promoted from the di- 559–569 nm, cf. Table 1).
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
J. Christoffers et al.
FULL PAPER
Similar to 18, the simplified NiWa Orange model 17 ex- Monitoring the Ca2+ Switch of Recoverin with NiWa
hibits an optically allowed πǞπ* transition from the Orange II
HOMO to the LUMO+1 as the second electronically ex-
cited state of the system. However, the first electronically
excited, dark state responsible for a radiationless decay in
NiWa Orange and 17 can be characterized as a HOMO to
LUMO transition to a charge-transfer state, in which an
electron is promoted from the diaminoterephthalate moiety
to the maleimide unit. The corresponding orbitals are dis-
played in Figure 5.
Site-directed labeling of proteins at selected sites is a
valuable approach in structure–function studies of proteins.
Based on our previous experience with the neuronal Ca2+
sensor recoverin as a benchmark model for studying confor-
mational changes by fluorescent probes,[10,12b] we investi-
gated whether the dye NiWa Orange II 8 is suitable for de-
tecting Ca2+-induced conformational changes in recoverin,
in particular, whether differences between the myristoylated
and nonmyristoylated recoverin forms can be detected by
this approach. Labeling of recoverin was performed suc-
cessfully and yielded a 1:1 stoichiometry, which indicates
free accessibility of the cysteine in the 39-position. No prin-
ciple difference was observed, if we used a borate or car-
bonate buffer for the labeling procedure. Once conjugated
to recoverin, NiWa Orange II exhibited a large increase in
fluorescence emission at 580 nm, which was not sensitive to
changes in free Ca2+ concentration (Figure 6, A). However,
the spectrum also showed a shoulder at ca. 500 nm, the in-
tensity of which increased with higher free Ca2+ concentra-
tion (compare spectra obtained at 58 nm and 500 μm free
Ca2+ in Figure 6). Guided by our previous studies with
NiWa-Blue-labeled recoverin,[10] we also observed that exci-
tation of intrinsic Trp residues at 280 nm and emission be-
tween 320–360 nm triggered a FRET process and excited
the covalently attached NiWa Orange II dye (Figure 6, B).
Figure 5. HOMO (left) and LUMO (right) of 17. The lowest excited
state corresponds to an optically forbidden charge-transfer transi-
tion from the HOMO to the LUMO.
As illustrated in Figure 5, the energetically low-lying Surprisingly, the large fluorescence emission peak at 580 nm
charge-transfer excitation of 17 is not accessible by optical disappeared and the shoulder at ca. 500 nm was blueshifted
excitation due to a negligible HOMO–LUMO Franck–Con- to 475 nm. Apparently, the fluorescence emission at 580 nm
don orbital overlap, which results in a very small oscillator was completely quenched when endogenous Trp residues in
strength (λex = 807 nm, f = 0.0001). However, this state recoverin were excited. Thus, the FRET recordings showed
plays the crucial role in quenching the fluorescence of 17 two maxima at 330 nm and at 475 nm. The signals were
by nonadiabatic relaxation through a conical intersection.
also Ca2+ dependent, exhibiting an increase at 330 nm and
Therefore, the general mechanistic picture of the fluores- a decrease in intensity at 475 nm when the free Ca2+ con-
cence behavior is the same for NiWa Blue and NiWa centration was lowered (Figure 6, B). The emission intensity
Orange.
changed with incremental changes of free Ca2+ concentra-
Figure 6. Emission spectra of myristoylated recoverin [2 μm in 80 mm buffer 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-
KOH, pH 7.5, 40 mm KCl, 1 mm dithiothreitol (DTT)] labeled with NiWa Orange II. Part A (left): emission scan in the presence of high
Ca2+ concentration (500 μm, dotted line) and low Ca2+ concentration (58 nM, solid line); excitation of NiWa Orange II at λex = 440 nm.
Part B (right): FRET study of 2 μm NiWa-Orange-labeled recoverin. Intrinsic FRET was measured by excitation of Trp at 280 nm and
the emission spectra were recorded between 300 and 550 nm. Buffer composition and Ca2+ concentration was as in Figure 6 A.
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
Turning On Fluorescence with Thiols
tion between 0.37 μm and 1000 μm, indicating a Ca2+-de- with disabled EF-hand Ca2+ binding sites showed that EF-
pendent movement of one or more Trp residues in the vicin- hand 3 binds Ca2+ with a KD of 0.1–0.26 μm and EF-hand
ity of the thiol-attached dye. Superimposed FRET spectra 2 exhibits a lower affinity with a KD of 6–7 μm.[19,20] The
recorded at different Ca2+ concentrations are shown in Fig- myristoyl group in recoverin serves as an allosteric modula-
ure 7 for myristoylated recoverin. Plotting the ratio of tor and lowers the affinity for Ca2+ resulting in an apparent
fluorescence emission at 475 nm and 330 nm (F475/F330) as KD of 17–18 μm for the whole binding process.[18,19] Our
a function of the Ca2+ concentration gave a dose response fluorescence data on myristoylated recoverin matched ex-
curve with an apparent KD value of 18.4 μm (Figure 8, A). actly the affinity that was previously determined by direct
The same set of fluorescence recordings were performed binding studies with 45Ca2+. Interestingly, for nonmyristoyl-
with labeled nonmyristoylated recoverin and gave a similar ated recoverin we determined by FRET analysis that the
general picture with respect to emission maxima and Ca2+ conformational change is half maximal at 0.3 μm Ca2+ and
dependency (not shown). Analyzing the data by the same equal to the apparent KD of the high affinity Ca2+ binding
plot of F475/F330 vs. Ca2+ yielded the curve shown in part B site in EF-hand 3. Previous NMR and X-ray crystallo-
of Figure 8. The change in emission ratio was half maximal graphic studies on wildtype and mutant recoverin forms
at 0.3 μm Ca2+. These data are in perfect agreement with indicate that binding of Ca2+ to the high affinity site in
previous estimates of the dissociation constants of Ca2+ EF-hand 3 triggers a major conformational change, which
binding to recoverin.[18–20] Direct Ca2+-binding studies on includes a swiveling of the two domains in recoverin by
nonmyristoylated wildtype and mutant recoverin forms 45°.[20,21] We conclude from our data that the distance be-
tween the FRET pair of Trp and the dye at the 39-position
changes significantly during this domain rearrangement.
Conclusions
We had recently introduced NiWa Blue dyes, the fluores-
cence of which is turned on by conjugated addition of thiols
(λem = 400 nm, λex = 338 nm). We now have prepared NiWa
Orange dyes with redshifted emission (λem = 560 nm, λex
=
470 nm). The diaminoterephthalate chromophore was elec-
tronically decoupled from the thiol-reactive maleimide
group by introduction of an ethylenediamine spacer.
Whereas the parent compound NiWa Orange I (7) pos-
sesses two maleimide functions, derived NiWa Orange II (8)
and III (11) have a single thiol-reactive group.
In order to shine light on the mechanism of turning on
the fluorescence of NiWa Blue and Orange dyes by thiol
addition, we have performed computational studies. The
thiol addition products in both, the blue and orange series,
turned out to be normal fluorescence dyes with HOMO–
LUMO transitions located at the aromatic chromophores
of these compounds. In contrast, the nonfluorescent malei-
Figure 7. FRET emission scan of myristoylated recoverin labeled
with NiWa Orange II. The spectra were recorded with 2 μm labeled
recoverin at different Ca2+ concentrations as indicated in the inset.
The excitation of intrinsic Trp was at 280 nm.
Figure 8. Ratio of fluorescence emission at 475 nm and 330 nm plotted as a function of the free Ca2+ concentration for myristoylated
(A) and nonmyristoylated labeled recoverin (B). Sets of spectra as shown in Figure 7 were evaluated. The change was half maximal at 18
μm for myristoylated and at 0.3 μm for nonmyristoylated recoverin.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
J. Christoffers et al.
FULL PAPER
System). Compound 5[14] and N-Boc-1,2-diaminoethane[15] were
prepared according to literature procedures. All other starting ma-
terials were commercially available.
mides possess a different, lower LUMO located at the imide
moiety of the molecule, but the excitation from the aromatic
HOMO to the LUMO is optically forbidden because of a
vanishing orbital overlap. Therefore, excitation of the elec- Dimethyl
2,5-Bis[(2-tert-butoxycarbonylamino)ethylamino]tere-
phthalate (6): A suspension of succinyl succinate 5 (367 mg,
1.61 mmol), N-Boc-ethylendiamine (1.03 g, 6.43 mmol) and glacial
acetic acid (0.8 mL) in toluene (8 mL) was heated to reflux in a
Dean–Stark apparatus for 16 h. The solution was diluted with
EtOAc (75 mL) and washed with saturated aqueous NaHCO3 solu-
tion (75 mL). The aqueous layer was extracted with EtOAc
(75 mL), and the combined organic layers were dried (MgSO4). Af-
ter filtration and evaporation, the residue was chromatographed
(SiO2, hexane/EtOAc 2:1 with 2 vol.-% NEt3). Fractions with Rf =
0.25 and 0.13 were collected, and the solvents were evaporated.
tron occurs from the aromatic HOMO to the LUMO+1,
which is also located at the six-membered ring. Interest-
ingly, due to nonadiabatic coupling of the potential surfaces
of the two excited states, the system will undergo a radia-
tionless transition from LUMO+1 to LUMO, from which
a further electronic relaxation to the HOMO is again op-
tically forbidden. This mechanism is even applicable to the
NiWa Orange systems, where the diaminoterephthalate and
the maleimide parts of the molecule are quite distal. This
scenario is significantly different from the fluorescence The materials (662 mg) were redissolved in abs. DMF (10 mL), a
solution of HCl in iPrOH (0.5 mL, 5–6 molL–1) was added, and
under exclusion of moisture, synthetic air was passed through the
quenching mechanism postulated for the BODIPY deriva-
tive 2 (cf. Figure 1), where a strongly distance-dependent
solution for 5 h while heating the mixture to 50 °C. After cooling
donor-excited photoinduced electron transfer was as-
to ambient temperature, EtOAc (50 mL) and saturated aqueous
sumed.[5]
NaHCO3 solution (50 mL) were added to the mixture, and the
Recoverin is a neuronal calcium sensor protein, which
aqueous layer was extracted with EtOAc (50 mL). The combined
plays a major role in the photoreceptor function in ver-
organic layers were dried (MgSO4). After filtration and evapora-
tebrate retina. The protein adopts two significantly different
tion, the residue was chromatographed [SiO2, gradient elution hex-
conformations in the absence and presence of Ca2+. More-
ane/EtOAc 3:1 Ǟ 2:1 with 2 vol.-% NEt3, Rf (hexane/EtOAc 2:1)
over, it has a single Cys residue at the 39-position, which
= 0.25] to yield compound 6 (581 mg, 1.14 mmol, 71%) as a red
covalently binds to NiWa Orange II 8. A fluorescence emis- solid, m.p. 154 °C. UV/Vis (CH2Cl2): λmax (ε, mol–1 dm3 cm–1) =
469 nm (5600); fluorescence (CH2Cl2): λex = 469 nm; λem = 569 nm.
1H NMR (500 MHz, CDCl3): δ = 1.43 (s, 18 H), 3.32 (t, J = 5.6 Hz,
4 H), 3.38–3.42 (m, 4 H), 3.89 (s, 6 H), 4.80 (br. s, 2 H), 6.90 (br.
s, 2 H), 7.33 (s, 2 H) ppm. 13C{1H} NMR (125 MHz, CDCl3): δ =
28.32 (6 CH3), 39.88 (2 CH2), 43.61 (2 CH2), 51.85 (2 CH3), 79.28
(2 C), 114.15 (2 CH), 116.91 (2 C), 140.99 (2 C), 155.96 (2 C),
sion is turned on by this conjugate addition. Excitation of a
Trp residue in proximity to Cys-39 leads to FRET to NiWa
Orange. The intensity of this FRET process is distance de-
pendent and therefore sensitive to conformational changes
of the protein induced by Ca2+. An internal FRET study
using excitation of Trp allowed us to monitor the critical
Ca2+-binding step that triggered the major conformational
change in recoverin. Moreover, the Ca2+-binding properties
of myristoylated and nonmyristoylated recoverin could be
168.16 (2 C) ppm. IR (ATR): ν = 3362 (s), 2968 (m), 2948 (m),
˜
1687 (vs), 1522 (vs), 1467 (m), 1439 (m), 1420 (m), 1392 (w), 1366
(w), 1332 (w), 1291 (m), 1272 (m), 1207 (vs), 1160 (vs), 1119 (vs),
1092 (s), 1037 (m), 998 (m), 969 (m), 907 (w), 872 (m), 849 (w),
clearly distinguished. The nonmyristoylated recoverin 788 (s), 759 (w) cm–1. HRMS (ESI, pos. mode): calcd. for
C24H39N4O8 [M + H]+ 511.2768; found 511.2779. C24H38N4O8
(510.58): calcd. C 56.46, H 7.50, N 10.97; found C 56.18, H 7.81,
N 11.17.
showed an apparent KD of 0.3 μm Ca2+. The myristoyl
group in recoverin serves as an allosteric modulator and
lowers the affinity for Ca2+, therefore, the FRET study re-
sulted in an apparent KD of 18.4 μm Ca2+ in this case.
Dimethyl 2,5-Bis[2-(2,5-dioxo-2,5-dihydropyrrol-1-yl)ethylamino]-
terephthalate (NiWa Orange I) (7): A solution of biscarbamate 6
(894 mg, 1.75 mmol) in TFA (5 mL) was stirred at 23 °C for 30 min.
Glacial acetic acid (6.4 mL) and maleic anhydride (687 mg,
7.01 mmol) were added, and the mixture was further heated at
95 °C for 3 d. The solution was then diluted with EtOAc (75 mL)
and saturated NaHCO3 solution (75 mL), and the layers were sepa-
rated. The aqueous layer was extracted with EtOAc (2ϫ75 mL),
and the combined organic layers were dried (MgSO4). After fil-
tration and evaporation, the residue was chromatographed (SiO2,
hexane/EtOAc 2:1 Ǟ 1:1 Ǟ 1:2 Ǟ EtOAc). The first fraction [Rf
(hexane/EtOAc 1:2) = 0.43] contained the title compound 7
(262 mg, 0.557 mmol, 32%) as an orange solid; m.p. 223 °C. No
byproduct 8 was isolated in this case. UV/Vis (CH2Cl2): λmax (ε,
mol–1 dm3 cm–1) = 462 nm (290). 1H NMR (500 MHz, CDCl3): δ =
3.40–3.65 (m, 4 H), 3.80 (t, J = 6.2 Hz, 4 H), 3.90 (s, 6 H), 6.69 (s,
4 H), 6.89 (br. s, 2 H), 7.34 (s, 2 H) ppm. 13C{1H} NMR (125 MHz,
CDCl3): δ = 37.11 (2 CH2), 41.94 (2 CH2), 51.96 (2 CH3), 114.20
(2 CH), 117.26 (2 C), 134.15 (4 CH), 140.67 (2 C), 168.08 (2 C),
Experimental Section
General: Preparative column chromatography was carried out using
Merck SiO2 (0.035–0.070 mm, type 60 A) with hexane, CH2Cl2,
and ethyl acetate (EtOAc) as eluents (for mixtures volume fractions
are given). TLC was performed with Merck SiO2 F254 plates on
1
aluminum sheets. H and 13C NMR spectra were recorded with a
Bruker Avance DRX 500. The multiplicities of carbon signals were
determined with DEPT experiments. MS and HRMS spectra were
obtained with a Finnigan MAT 95 (EI and CI) and a Waters Q-
TOF Premier (ESI) spectrometer. IR spectra were recorded with a
Bruker Tensor 27 spectrometer equipped with a GoldenGate dia-
mond attenuated total reflectance (ATR) unit. Elemental analyses
were measured with a Euro EA-CHNS instrument from HEKA-
tech. UV/Vis spectra were recorded with Perkin–Elmer LS 50 and
Analytik Jena SPECORD 205 spectrometers. Fluorescence spectra
were recorded with a Shimadzu RF-5301PC and a fluorescence
spectrometer from Photon Technology International (Lamp Power
supply Model LPS-220B and 810/814 Photomultiplier Detection
170.62 (4 C) ppm. IR (ATR): ν = 3354 (m), 3118 (w), 2953 (w),
˜
2922 (m), 2849 (m), 1690 (vs), 1592 (w), 1533 (m), 1467 (w), 1439
(m), 1406 (m), 1360 (m), 1333 (m), 1258 (m), 1209 (vs), 1158 (m),
1110 (vs), 1089 (m), 1057 (m), 998 (m), 962 (m), 912 (m), 874 (w),
8
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
Turning On Fluorescence with Thiols
853 (w), 834 (vs), 787 (s), 721 (w), 696 (vs) cm–1. HRMS (EI, (w), 1698 (vs), 1604 (w), 1578 (w), 1532 (s), 1497 (w), 1456 (m),
70 eV): calcd. for C22H22N4O8 [M]+ 470.1438; found 470.1429.
1437 (m), 1420 (m), 1397 (s), 1359 (m), 1336 (m), 1216 (vs), 1175
(s), 1116 (s), 1074 (m), 1031 (w), 991 (w), 920 (w), 874 (w) cm–1.
HRMS (EI, 70 eV): calcd. for C36H38N4O8S2 [M]+ 718.2131; found
718.2148.
Dimethyl 2-[2-(3-Acetoxy-2,5-dioxopyrrolidin-1-yl)ethylamino]-5-[2-
(2,5-dioxo-2,5-dihydropyrrol-1-yl)ethylamino]terephthalate (NiWa
Orange II) (8): A solution of biscarbamate 6 (410 mg, 0.803 mmol)
in CH2Cl2 (4 mL) and TFA (2 mL) was stirred at 23 °C for 1 h.
Glacial acetic acid (3 mL) and maleic anhydride (315 mg,
3.21 mmol) were added, and the resulting mixture was further
heated at 90 °C for 2 d. The solution was then diluted with EtOAc
(90 mL) and H2O (40 mL), and the layers were separated. The or-
ganic layer was extracted with H2O (40 mL) and dried (MgSO4).
After filtration and evaporation, the residue was chromatographed
(SiO2, hexane/EtOAc 2:1 Ǟ 1:1 Ǟ 1:2). The first fraction [Rf (hex-
ane/EtOAc 1:2) = 0.43] contained compound 7 (56 mg, 0.12 mmol,
15%) as an orange solid. From the second fraction [Rf (hexane/
EtOAc 1:2) = 0.33], compound 8 (40 mg, 75 μmol, 9%) was isolated
Conversion of Bisimide 7 with Hexadecanethiol: A solution of 1-
hexadecanethiol (0.12 mmol, 8.1 mL, 0.015 molL–1 in CH2Cl2) was
added (cooling with an ice-water bath) to a solution of bismalein-
imide 7 (114 mg, 0.242 mmol) and NEt3 (100 mg, 0.988 mmol) in
CH2Cl2 (4 mL). The mixture was warmed to 23 °C and then stirred
for 24 h. The solvent was evaporated under reduced pressure, and
the crude solid was purified by chromatography (SiO2, gradient
elution with hexane/CH2Cl2/EtOAc 2:1:1 Ǟ 1:1:1 Ǟ CH2Cl2/
EtOAc 1:1) to yield bis adduct 12 (18 mg, 18 μmol, 8%) in the first
fraction [Rf(SiO2, CH2Cl2/EtOAc 1:1) = 0.53]. The second fraction
[Rf (SiO2, CH2Cl2/EtOAc 1:1) = 0.30] was mono adduct 11 (20 mg,
27 μmol, 11%; NiWa Orange III). Finally, starting material 7
(22 mg, 47 μmol, 19%) was obtained as the third fraction [Rf (SiO2,
CH2Cl2/EtOAc 1:1) = 0.11].
as
a byproduct; m.p. 128 °C. UV/Vis (CH2Cl2): λmax (ε,
mol–1 dm3 cm–1) = 446 nm (6700). H NMR (500 MHz, CDCl3): δ
= 2.14 (s, 3 H), 2.64 (dd, J = 4.8, J = 18.2 Hz, 1 H), 3.15 (dd, J =
8.7, J = 18.3 Hz, 1 H), 3.40–3.51 (m, 4 H), 3.78–3.82 (m, 2 H),
3.82–3.86 (m, 2 H), 3.90 (s, 3 H), 3.91 (s, 3 H), 5.43 (dd, J = 4.8,
J = 8.6 Hz, 1 H), 6.70 (s, 2 H), 7.39 (s, 1 H), 7.40 (s, 1 H) ppm;
the two NH protons are not detectable. 13C{1H} NMR (125 MHz,
CDCl3, 330 K): δ = 20.38 (CH3), 35.72 (CH2), 37.23 (CH2), 38.47
(CH2), 41.11 (CH2), 42.02 (CH2), 51.91 (2 CH3), 67.50 (CH),
114.46 (CH), 114.60 (CH), 117.46 (C), 117.74 (C), 134.16 (2 CH),
140.50 (C), 140.93 (C), 168.02 (C), 168.11 (C), 169.64 (C), 170.54
1
Dimethyl 2-[2-(2,5-Dioxo-2,5-dihydropyrrol-1-yl)ethylamino]-5-[2-
(3-hexadecylsulfanyl-2,5-dioxopyrrolidin-1-yl)ethylamino]terephthal-
ate (NiWa Orange III) (11): M.p. 186 °C. UV/Vis (CH2Cl2): λmax
(ε, mol–1 dm3 cm–1) = 468 nm (2500). 1H NMR (500 MHz, CDCl3):
δ = 0.88 (t, J = 6.9 Hz, 3 H), 1.22–1.30 (m, 23 H), 1.32–1.39 (m, 2
H), 1.54–1.66 (m, 3 H), 2.52 (dd, J = 3.6, J = 18.6 Hz, 1 H), 2.71
(ddd, J = 6.8, J = 8.3, J = 15.0 Hz, 1 H), 2.85 (ddd, J = 6.3, J =
8.7, J = 14.4 Hz, 1 H), 3.14 (dd, J = 8.0, J = 18.8 Hz, 1 H), 3.43
(q, J = 6.5 Hz, 4 H), 3.72 (dd, J = 3.6, J = 8.1 Hz, 1 H), 3.79–3.83
(m, 4 H), 3.90 (s, 3 H), 3.91 (s, 3 H), 6.70 (s, 2 H), 7.37 (s, 1 H),
7.42 (s, 1 H) ppm; the two NH protons are not detectable. 13C{1H}
NMR (125 MHz, CDCl3): δ = 14.05 (CH3), 22.67 (CH2), 28.83
(CH2), 29.10 (CH2), 29.20 (CH2), 29.35 (CH2), 29.50 (CH2), 29.59
(2 CH2), 29.66 (2 CH2), 29.69 (4 CH2), 31.77 (CH2), 31.93 (CH2),
36.27 (CH2), 37.11 (CH2), 38.23 (CH2), 39.33 (CH), 42.11 (CH2),
51.98 (CH3), 52.03 (CH3), 114.52 (2 CH), 117.39 (2 C), 134.18 (2
CH), 141.05 (2 C), 168.00 (2 C), 170.56 (2 C), 174.76 (C), 176.59
(2 C), 173.05 (C), 173.31 (C) ppm. IR (ATR): ν = 3355 (w), 3111
˜
(w), 2953 (w), 2848 (w), 1693 (vs), 1601 (w), 1533 (m), 1438 (m),
1405 (m), 1358 (m), 1335 (m), 1213 (vs), 1110 (s), 1040 (m), 997
(w), 964 (w), 913 (m), 873 (w), 830 (m), 788 (m), 728 (m), 696 (s),
648 (w) cm–1. HRMS (EI, 70 eV): calcd. for C24H26N4O10 [M]+
530.1649; found 530.1641.
Turn On of NiWa Orange II (8) with Glutathione (GSH), Formation
of 9: DMSO (1 mL) was added to a mixture of acetoxy adduct 8
(1.3 mg, 2.5 μmol) and GSH (10 mg, 33 μmol), and the resulting
mixture was stirred for 20 min at 23 °C. It was then diluted with
DMSO (9 mL) and further analyzed. UV/Vis (DMSO): λmax (ε,
(C) ppm. IR (ATR): ν = 3357 (w), 2919 (m), 2850 (m), 1692 (vs),
˜
1537 (m), 1439 (m), 1404 (m), 1360 (m), 1336 (m), 1220 (vs), 1112
(vs), 963 (w), 911 (w), 874 (w), 838 (m), 788 (s), 721 (w), 697 (s)
cm–1. HRMS (EI, 70 eV): calcd. for C38H56N4O8S [M]+ 728.3819;
found 728.3806.
mol–1 dm3 cm–1) = 469 nm (6700); fluorescence (DMSO): λex
=
469 nm; λem = 568 nm. HRMS (ESI, pos. mode): calcd. for
C34H44N7O16S [M + H]+ 838.2565; found 838.2551.
Dimethyl 2,5-Bis{2-[3-(hexadecylsulfanyl)-2,5-dioxopyrrolidin-1-yl]-
Dimethyl 2,5-Bis{2-[3-(benzylsulfanyl)-2,5-dioxopyrrolidin-1-yl]eth-
ylamino}terephthalate (10): A solution of bismaleinimide 7 (25 mg,
53 μmol), BnSH (32 mg, 0.26 mmol) and NEt3 (19 mg, 0.19 mmol) 467 nm; λem = 561 nm. H NMR (500 MHz, CDCl3, 328 K): δ =
ethylamino}terephthalate (12): M.p. 165 °C. UV/Vis (CH2Cl2): λmax
(ε, mol–1 dm3 cm–1) = 467 nm (4200); fluorescence (CH2Cl2): λex
=
1
in CH2Cl2 (2 mL) was stirred at 23 °C for 1 d. The solvent was
removed under reduced pressure, and the residue was chromato-
graphed [SiO2, gradient elution hexane/EtOAc 2:1 Ǟ 1:1 Ǟ EtOAc,
Rf (hexane/EtOAc 2:1) = 0.09] to yield the compound 10 (33 mg,
46 μmol, 87%) as a red solid, m.p. 70 °C. UV/Vis (CH2Cl2): λmax
0.89 (t, J = 6.8 Hz, 6 H), 1.24–1.33 (m, 48 H), 1.33–1.45 (m, 4 H),
1.56–1.69 (m, 4 H), 2.51 (dd, J = 3.6, J = 18.5 Hz, 2 H), 2.70–2.76
(m, 2 H), 2.85 (ddd, J = 6.3, J = 8.1, J = 18.7 Hz, 2 H), 3.09 (dd,
J = 9.1, J = 18.5 Hz, 2 H), 3.45 (t, J = 6.3 Hz, 4 H), 3.68 (dd, J =
3.7, J = 9.0 Hz, 2 H), 3.77–3.82 (m, 4 H), 3.91 (s, 6 H), 7.37 (s, 2
H) ppm; the two NH protons are not detectable. 13C{1H} NMR
(125 MHz, CDCl3, 328 K): δ = 14.00 (2 CH3), 22.67 (2 CH2), 28.83
(2 CH2), 29.15 (2 CH2), 29.20 (2 CH2), 29.34 (2 CH2), 29.50 (2
(ε, mol–1 dm3 cm–1) = 467 nm (3400); fluorescence (CH2Cl2): λex
=
1
467 nm; λem = 559 nm. H NMR (500 MHz, CDCl3, 330 K): δ =
2.43 (dd, J = 3.9, J = 18.6 Hz, 2 H), 2.96 (dd, J = 9.2, J = 18.6 Hz,
2 H), 3.47 (t, J = 6.1 Hz, 4 H), 3.54 (dd, J = 3.9, J = 9.2 Hz, 2 H), CH2), 29.59 (2 CH2), 29.69 (4 CH2), 29.69 (6 CH2), 31.80 (2 CH2),
3.77–3.84 (m, 4 H), 3.88 (d, J = 13.4 Hz, 2 H), 3.91 (s, 6 H), 4.19 31.93 (2 CH2), 36.29 (2 CH2), 38.39 (2 CH2), 39.39 (2 CH), 42.22
(d, J = 13.5 Hz, 2 H), 6.87 (br. s, 2 H), 7.27–7.30 (m, 2 H), 7.31– (2 CH2), 51.90 (2 CH3), 114.53 (2 CH), 117.60 (2 C), 140.75 (2 C),
7.36 (m, 4 H), 7.37–7.41 (m, 6 H) ppm. 13C{1H} NMR (125 MHz, 168.07 (2 C), 174.62 (2 C), 176.57 (2 C) ppm; signals for two CH2
CDCl3, 330 K): δ = 35.66 (2 CH2), 36.00 (2 CH2), 37.95 (2 CH), groups at approx. 37 ppm are not detectable. IR (ATR): ν = 3372
˜
38.41 (2 CH2), 41.13 (2 CH2), 51.87 (2 CH3), 114.43 (2 CH), 117.56
(w), 2917 (s), 2850 (m), 1769 (w), 1692 (vs), 1539 (m), 1467 (m),
(2 C), 127.53 (2 CH), 128.69 (4 CH), 129.16 (4 CH), 136.94 (2 C), 1439 (m), 1401 (m), 1342 (m), 1220 (s), 1203 (s), 1116 (s), 949 (w),
140.79 (2 C), 168.07 (2 C), 174.48 (2 C), 176.52 (2 C) ppm. IR
873 (w), 788 (m), 720 (w), 694 (w) cm–1. HRMS (EI, 70 eV): calcd.
for C54H90N4O8S2 [M]+ 986.6200; found 986.6209.
(ATR): ν = 3373 (m), 3060 (w), 3030 (w), 2951 (m), 2859 (w), 1177
˜
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
9

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
J. Christoffers et al.
FULL PAPER
Turn On of NiWa Orange III (11) with Glutathione (GSH), Forma-
tion of Product 13: DMSO (1 mL) was added to a mixture of thiol
adduct 11 (0.9 mg, 1.3 μmol) and GSH (10 mg, 33 μmol) and the
resulting mixture was stirred for 20 min at 23 °C. It was then di-
luted with DMSO (9 mL) and further analyzed. UV/Vis (DMSO):
λmax (ε, mol–1 dm3 cm–1) = 469 nm (2100); fluorescence (DMSO):
λem = 565 nm; λex = 469 nm. HRMS (ESI, pos. mode): calcd. for
C48H74N7O14S2 [M + H]+ 1036.4735; found 1036.4720.
[1]
For reviews, see: a) A. Mayer, S. Neuenhofer, Angew. Chem.
1994, 106, 1097–1126; Angew. Chem. Int. Ed. Engl. 1994, 33,
1044–1072; b) X. Chen, Y. Zhou, X. Peng, J. Yoon, Chem. Soc.
Rev. 2010, 39, 2120–2135; c) T. Nagano, Proc. Jpn. Acad., Ser.
B 2010, 86, 837–847.
[2] a) K. K.-W. Lo, W.-K. Hui, D. C.-M. Ng, K.-K. Cheung, Inorg.
Chem. 2002, 41, 40–46; b) P. Ge, P. R. Selvin, Bioconjugate
Chem. 2003, 14, 870–876; c) S. Girouard, M.-H. Houle, A.
Grandbois, J. W. Keillor, S. W. Michnick, J. Am. Chem. Soc.
2005, 127, 559–566; d) J. Weh, A. Duerkop, O. S. Wolfbeis,
ChemBioChem 2007, 8, 122–128; e) G.-L. Li, K.-Y. Kung, L.
Zou, H.-C. Chong, Y.-C. Leung, K.-H. Wong, M.-K. Wong,
Chem. Commun. 2012, 48, 3527–3529.
Preparation and Labeling of Recoverin: Bovine recoverin was
heterologously expressed in E. coli (BL21 strain) exactly as de-
scribed previously.[10,12] In order to express myristoylated recoverin
in E. coli cells these were cotransformed with the plasmid carrying
the gene for yeast N-myristoyl-transferase (pBB131). Myristoylated
recoverin was expressed, purified, and analyzed by HPLC as de-
scribed before.[18] For labeling with NiWa Orange II, we dissolved
0.5 mg of lyophilized recoverin in 0.45 mL carbonate or borate cou-
pling buffer (carbonate buffer: 20 mm NaHCO3/NaOH pH 10.0,
100 mm NaCl; borate buffer: 20 mm H3BO3/NaOH pH 10.0,
100 mm NaCl). The dye NiWa Orange II (8) (50 μL of a 1.9 mm
solution in DMSO) was added to the recoverin solution and the
mixture was incubated under gentle shaking for 1.5 h at room tem-
perature in the dark. Unreacted dye was removed by passing the
solution over a short gel filtration column (PD10 containing Se-
[3] P. Y. Reddy, S. Kondo, S. Fujita, T. Toru, Synthesis 1998, 999–
1002.
[4] a) L. Yi, H. Li, L. Sun, L. Liu, C. Zhang, Z. Xi, Angew. Chem.
2009, 121, 4094–4097; Angew. Chem. Int. Ed. 2009, 48, 4034–
4037; b) X. Zhang, X. Ren, Q.-H. Xu, K. P. Loh, Z.-K. Chen,
Org. Lett. 2009, 11, 1257–1260; c) L. Peng, G. Zhang, D.
Zhang, J. Xiang, R. Zhao, Y. Wang, D. Zhu, Org. Lett. 2009,
11, 4014–4017; d) H. S. Jung, K. C. Ko, G.-H. Kim, A.-R. Lee,
Y.-C. Na, C. Kang, J. Y. Lee, J. S. Kim, Org. Lett. 2011, 13,
1498–1501; e) S.-Y. Lim, S. Lee, S. B. Park, H.-J. Kim, Tetrahe-
dron Lett. 2011, 52, 3902–3904; f) X.-D. Jiang, J. Zhang, X.
Shao, W. Zhao, Org. Biomol. Chem. 2012, 10, 1966–1968.
phadex G25 material with
a
void volume of 2.5 mL, GE
[5]
T. Matsumoto, Y. Urano, T. Shoda, H. Kojima, T. Nagano,
Org. Lett. 2007, 9, 3375–3377.
a) H. Liebermann, Justus Liebigs Ann. Chem. 1914, 404, 272–
321; b) For a review, see: S. S. Labana, L. L. Labana, Chem.
Rev. 1967, 67, 1–18.
Y. Zhang, J. Christoffers, Synthesis 2007, 3061–3067.
a) S. D. Ohmura, T. Moriuchi, T. Hirao, Tetrahedron Lett.
2010, 51, 3190–3192; b) M. Shimizu, Y. Asai, Y. Takeda, A.
Yamatani, T. Hiyama, Tetrahedron Lett. 2011, 52, 4084–4089.
a) Y. Zhang, P. Starynowicz, J. Christoffers, Eur. J. Org. Chem.
2008, 3488–3495; b) R. Pflantz, J. Christoffers, Chem. Eur. J.
2009, 15, 2200–2209.
N. Wache, C. Schröder, K.-W. Koch, J. Christoffers, Chem-
BioChem 2012, 13, 993–998.
I. I. Senin, K.-W. Koch, M. Akhtar, P. P. Philippov, Adv. Exp.
Med. Biol. 2003, 514, 69–99.
a) O. H. Weiergräber, I. I. Senin, E. Y. Zernii, V. A. Churu-
mova, N. A. Kovaleva, A. A. Nazipova, S. E. Permyakov, E. A.
Permyakov, P. P. Philippov, J. Granzin, K.-W. Koch, J. Biol.
Chem. 2006, 281, 37594–37602; b) T. Gensch, K. E. Komolov,
I. I. Senin, P. P. Philippov, K.-W. Koch, Proteins Struct. Funct.
Bioinf. 2007, 66, 492–499.
H. Kollmann, S. F. Becker, J. Shirdel, A. Scholten, A. Os-
tendorp, C. Lienau, K.-W. Koch, ACS Chem. Biol. 2012, 7,
1006–1014.
Healthcare). The PD10 column was equilibrated in fluorescence
buffer (80 mm HEPES-KOH pH 7.5, 40 mm KCl, 1 mm DTT) when
labeled recoverin was further used for fluorescence studies. For this
purpose the recoverin solution after passing over the PD 10 column
was diluted with 10 mL of fluorescence buffer to yield a final vol-
ume of 13.5 mL. The stoichiometry of bound dye was determined
by measuring the extinction of the bound dye at 440 nm using a
UV/Vis spectrophotometer (SPECORD 205, Analytik Jena). The
protein concentration was determined with the Coomassie blue dye
method of Bradford[22] yielding a stoichiometry of approx. 1:1 for
nonmyristoylated and myristoylated recoverin.
[6]
[7]
[8]
[9]
[10]
[11]
[12]
Fluorescence Recordings: For each recording 0.96 mL of labeled re-
coverin stock was mixed with 0.04 mL of Ca2+-EGTA buffer
(2 mm) or CaCl2 to yield final free Ca2+ concentrations from
0.37 μm to 1 mm. Those concentrations of free Ca2+ were adjusted
by mixtures of K2H2EGTA (EGTA = ethylene glycol bis(2-aminoe-
thyl ether) tetraacetic acid, 50 mm) and CaH2EGTA (50 mm) as de-
scribed previously.[23] Concentrations of free Ca2+ above 2 μm were
adjusted by adding CaCl2 at the desired final concentration. The
fluorescence emission spectra of 2 μm labeled recoverin were mea-
sured with a fluorescence spectrometer from Photon Technology
International (Lamp Power supply Model LPS-220B and 810/814
Photomulitplier Detection System). The sample was excited at
440 nm and the emission spectra were recorded from 450–650 nm.
FRET from endogenous Trp residues to the conjugated dye was
measured by setting the excitation wavelength to 280 nm and re-
cording the emission spectra from 300 to 550 nm. Recording and
data analysis was performed with the photon Technology Inter-
national software package FELIX32.
[13]
[14]
[15]
[16]
D. Seebach, T. Hoffmann, F. N. M. Kühnle, J. N. Kinkel, M.
Schulte, Helv. Chim. Acta 1995, 78, 1525–1540.
T. L. Foley, A. Yasgar, C. J. Garcia, A. Jadhav, A. Simeonov,
M. D. Burkart, Org. Biomol. Chem. 2010, 8, 4601–4606.
a) J. R. Lakowicz, Principles of Fluorescence Spectroscopy, 3rd
ed. Springer, New York, 2006, p. 54; b) For fluorescein as stan-
dard, see: G. Weber, F. W. J. Teale, Trans. Faraday Soc. 1957,
53, 646–655.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starov-
erov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Ad-
[17]
Supporting Information (see footnote on the first page of this arti-
cle): Additional fluorescence spectra and 1H and 13C NMR spectra
of all reported compounds.
Acknowledgments
We gratefully acknowledge support from the Deutsche Forschungs-
gemeinschaft (DFG).
10
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
Turning On Fluorescence with Thiols
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
revision A.1, Gaussian, Inc., Wallingford CT, 2009.
[20]
[21]
O. H. Weiergräber, I. I. Senin, P. P. Philippov, J. Granzin, K.-
W. Koch, J. Biol. Chem. 2003, 278, 22972–22979.
J. B. Ames, N. Hamasaki, T. Molchanova, Biochemistry 2002,
41, 5776–5787.
[22] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[23] P. Behnen, A. Scholten, N. Rätscho, K.-W. Koch, J. Biol. Inorg.
Chem. 2009, 14, 89–99.
[18] I. I. Senin, T. Fischer, K. E. Komolov, D. V. Zinchenko, P. P.
Philippov, K.-W. Koch, J. Biol. Chem. 2002, 277, 50365–50372.
[19] J. B. Ames, T. Porumb, T. Tanaka, M. Ikura, L. Stryer, J. Biol.
Chem. 1995, 270, 4526–4533.
Received: July 3, 2012
Published Online: ᭿
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
11

Job/Unit: O20879
/KAP1
Date: 27-08-12 18:56:11
Pages: 12
J. Christoffers et al.
FULL PAPER
Fluorescent Probes
A Click Turns It On: Maleimide-func-
tionalized diaminoterephthalates (NiWa
Orange) are thiol-reactive fluorescence pro-
bes. Whereas the dye itself shows no fluo-
rescence, it is turned on by the conjugate ad-
dition of the thiol. NiWa Orange was used
to label recoverin. Its Ca2+-dependent con-
formational change was traced by Förster
resonance energy transfer of a tryptophan
residue to NiWa Orange.
N. Wache, A. Scholten, T. Klüner,
K.-W. Koch, J. Christoffers* ........... 1–12
Turning On Fluorescence with Thiols –
Synthetic and Computational Studies on
Diaminoterephthalates and Monitoring the
Switch of the Ca2+ Sensor Recoverin
Keywords: Fluorescent probes / Sensors /
Calcium
12
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0