DBD dyes as fluorescence lifetime probes to study conformational changes in proteins

Article abstract of DOI:10.1002/chem.201302368

Previously, [1,3]dioxolo[4,5-f][1,3]benzodioxole (DBD)-based fluorophores used as highly sensitive fluorescence lifetime probes reporting on their microenvironmental polarity have been described. Now, a new generation of DBD dyes has been developed. Although they are still sensitive to polarity, in contrast to the former DBD dyes, they have extraordinary spectroscopic properties even in aqueous surroundings. They are characterized by long fluorescence lifetimes (10-20 ns), large Stokes shifts (≈100 nm), high photostabilities, and high quantum yields (>0.56). Here, the spectroscopic properties and synthesis of functionalized derivatives for labeling biological targets are described. Furthermore, thio-reactive maleimido derivatives of both DBD generations show strong intramolecular fluorescence quenching. This mechanism has been investigated and is found to undergo a photoelectron transfer (PET) process. After reaction with a thiol group, this fluorescence quenching is prevented, indicating successful bonding. Being sensitive to their environmental polarity, these compounds have been used as powerful fluorescence lifetime probes for the investigation of conformational changes in the maltose ATP-binding cassette transporter through fluorescence lifetime spectroscopy. The differing tendencies of the fluorescence lifetime change for both DBD dye generations promote their combination as a powerful toolkit for studying microenvironments in proteins. Polarity probes: New fluorescent dyes based on the [1,3]dioxolo[4,5-f ][1,3]benzodioxole (DBD) skeleton were developed, with fluorescence lifetimes that exhibit a pronounced sensitivity to the polarity of the microenvironment. Several biocompatible derivatives were prepared, and their application was demonstrated with two biochemical examples (see figure; MBP=maltose binding protein).

Full text of DOI:10.1002/chem.201302368

FULL PAPER
DOI: 10.1002/chem.201302368
DBD Dyes as Fluorescence Lifetime Probes to Study Conformational
Changes in Proteins
[
a]
[b]
[b]
[a]
Robert Wawrzinek, Joanna Ziomkowska, Johanna Heuveling, Monique Mertens,
[
b]
[b]
[a]
Andreas Herrmann,* Erwin Schneider,* and Pablo Wessig*
Abstract: Previously, [1,3]dioxolo
A
H
U
G
R
N
U
G
erties and synthesis of functionalized
derivatives for labeling biological tar-
gets are described. Furthermore, thio-
reactive maleimido derivatives of both
DBD generations show strong intramo-
lecular fluorescence quenching. This
mechanism has been investigated and
is found to undergo a photoelectron
transfer (PET) process. After reaction
with a thiol group, this fluorescence
quenching is prevented, indicating suc-
cessful bonding. Being sensitive to
their environmental polarity, these
compounds have been used as powerful
fluorescence lifetime probes for the in-
vestigation of conformational changes
in the maltose ATP-binding cassette
transporter through fluorescence life-
time spectroscopy. The differing ten-
dencies of the fluorescence lifetime
change for both DBD dye generations
promote their combination as a power-
ful toolkit for studying microenviron-
ments in proteins.
[1,3]benzodioxole (DBD)-based fluoro-
phores used as highly sensitive fluores-
cence lifetime probes reporting on
their microenvironmental polarity have
been described. Now, a new generation
of DBD dyes has been developed. Al-
though they are still sensitive to polari-
ty, in contrast to the former DBD dyes,
they have extraordinary spectroscopic
properties even in aqueous surround-
ings. They are characterized by long
fluorescence lifetimes (10–20 ns), large
Stokes shifts (ꢀ100 nm), high photo-
stabilities, and high quantum yields
Keywords: dyes/pigments · electron
transfer · fluorescent probes · male-
imides · MalF · photoelectron trans-
fer
(
>0.56). Here, the spectroscopic prop-
Introduction
Time-correlated single-photon counting (TCSPC) spectros-
copy and fluorescence-lifetime imaging microscopy (FLIM)
detect the fluorescence decay of fluorophores at their emis-
sion maxima, giving high-contrast signals that are unaffected
by decreasing dye concentrations caused by photobleach-
Although there are many different commercially available
fluorescent dyes, the demand for new fluorophores is still
high. There is no universal dye that possesses all the desir-
[1–4]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
able characteristics (high quantum yield, long fluorescence
ing.
Therefore, the quantitative analysis of fluorescence
lifetime, high molar extinction coefficients, photostability,
low toxicity, etc.), so it is important to enrich the toolbox of
luminescent probes so that as many combinations as possi-
ble of those parameters are available, preferably for various
absorption and emission wavelengths. Fluorescence spec-
troscopy is developing continuously, and more and more
sensitive techniques require specialized probes to realize the
ever-higher resolution of small structures and fast processes.
lifetime spectroscopy is superior to and more accurate than
most steady-state techniques, which use fluorescence intensi-
ty as a measurement parameter.
The most established fluorophores are based on pyrene,
xanthene, coumarin, or oxazine, and are used successfully in
[5–8]
countless applications and experiments.
However, owing
to a lack of availability or even knowledge of more suitable
dyes, the resulting data are often not as convincing as de-
sired. In this article, we give an example of why it is still
worth investigating new types of fluorophores. In previous
work, we introduced new fluorescent dyes based on
[9,10]
[
a] R. Wawrzinek, M. Mertens, Prof. Dr. P. Wessig
[
1,3]dioxolo
A
H
U
T
E
N
N
[4,5-f]
A
H
U
G
R
N
U
G
We have
Institut fꢀr Chemie, Universitꢁt, Potsdam Karl-Liebknecht-Str. 24-25
shown that they could be used as sensitive fluorescence life-
time probes in cell-staining experiments reporting on the po-
1
4476 Potsdam (Germany)
Fax : (+49)331-977-5065
E-mail: Wessig@uni-potsdam.de
[11]
larity of their microenvironment.
Herein, we describe
[
b] J. Ziomkowska, Dr. J. Heuveling, Prof. Dr. A. Herrmann,
Prof. Dr. E. Schneider
a new generation of DBD dyes containing different elec-
tron-withdrawing groups (EWGs) at the 4- and 7-positions
of the fluorophore. Scheme 1 shows the general structures
of the fluorophores described in this article. They are pro-
vided with various functional groups that can be linked to
biomolecules. Furthermore, maleimido derivates of both
DBD dye generations were used as powerful probes to in-
Institut fꢀr Biologie, Humboldt-Universitꢁt zu Berlin
Invalidenstraße 42, 10115 Berlin (Germany)
E-mail: h1211dyz@rz.hu-berlin.de
erwin.schneider@rz.hu-berlin.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201302368.
Chem. Eur. J. 2013, 19, 17349 – 17357
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
17349

Table 1. Spectroscopic properties of 4 and 5.
Solvent
l
abs
l
em
t
F
e
F
F
e·F
F
ACHTUNGTRENNUNG[ m cm ]
ꢁ
1
ꢁ1
ꢁ1
ꢁ1
[
nm] [nm] [ns]
A[ m cm
H
U
G
R
N
N
]
4
5
4
5
4
5
4
5
4
5
449
425
444
385
452
384
452
384
–
561
512
561
495
596
510
611
509
–
24.4 4400
18.3 2300
20.3 4100
14.3 2800
11.6 4300
13.3 3800
0.35 1540
0.77 1800
0.57 2300
0.55 1500
0.11 473
0.63 2400
0.06 258
0.55 2500
dichloromethane
acetonitrile
ethanol
Scheme 1. General structure of DBD dyes.
5.9
4300
methanol
13.3 4400
[
a]
–
–
–
–
vestigate the conformational change of the maltose ATP-
binding cassette (ABC) transporter. In addition, we exam-
ined the quenching mechanism of DBD dyes caused by
2
H O
370
488
20.2 4600
0.82 3800
[12]
[
a] Insoluble in water.
a conjugated male ACHTUNGTRENNUNGi mide.
Because the previously described acyl-DBD dyes already
possess a large number of desired parameters, such as small
molecule size, non-toxicity, high photostability, large Stokes
shifts, and long fluorescence lifetimes (at least in non-polar
microenvironments), we aimed to add new items to the tool-
box by developing additional DBD dyes with other interest-
ing characteristics.
features a much higher fluorescence quantum yield and
longer fluorescence decay in polar solvents than its acyl-
bearing analog 4.
Furthermore, it was observed that the fluorescence quan-
tum yield and lifetime of compound 5 increases even further
in water, whereas these values decrease drastically for
[11]
water-soluble acyl-DBD derivates (ꢀ1.5 ns).
These results encouraged us to develop bioreactive deri-
vates that can be used as labels for proteins or other biologi-
cal targets. Because of their own reactivity, carboxylic
groups functioning as EWGs, these were found not to be
ideal residues for more complex compounds containing bio-
reactive groups. Therefore, we replaced them with esters, ex-
pecting these compounds to show spectroscopic properties
similar to those of 5, but to be less reactive at their EWGs.
We were able to synthesize compound 13a, which bears
ethyl esters as EWGs and a single reactive carboxylic func-
tion separated from its fluorophore (Scheme 3).
Results and Discussion
Synthesis and spectroscopy of new DBD dyes: Although
DBD dyes with acyl residues as EWGs are sensitive polarity
probes, we considered that different EWGs might reduce
the observed strong fluorescence intensity and lifetime
quenching in the presence of water to extend their applica-
tion field. The latter effect is probably caused by hydrogen
bridge bonds disturbing the push–pull system of the acyl-
DBD dye. Our first substitution of these acyl residues was
with carboxylic groups at positions 4 and 7 of the fluoro-
phore. Scheme 2 shows the synthesis of the prototype 5 and
Starting from sesamol 6, we optimized the synthetic route
to the methoxymethyl (MOM)-protected catechol 9 used
before for acyl-DBD dyes, and managed to introduce the
[9]
[13]
its acyl-bearing analog 4.
desired ethyl esters (10a).
After cleavage, catechol 11a
The fluorescence properties of 5 were found to be rather
different from those of compound 4. As shown in Table 1, 5
was cyclized with benzyl-protected propiolic acid by using 4-
[9,14]
dimethylaminopyridine (DMAP).
Benzyl as a protecting
group turned out to be superior
to a methyl ester because the
subsequent cleavage, leading to
1
3a through hydrogenolysis,
gives much higher yields than
saponification. To compare
both generations of DBD dyes,
we also synthesized butyryl-
DBD 13b starting from the pre-
viously
described
catechol
[9]
1
1b.
Figure 1 shows the typical ab-
sorption and emission spectra
of DBD fluorophores. Similarly
to acyl-DBD dyes (dotted), the
n–p* absorption bands of ester-
DBD dyes (solid) are broad
and surprisingly intense. The
latter bands are slightly hypso-
Scheme 2. Preparation of 4 and 5 (p-TSA=p-toluenesulfonic acid, TMEDA=N,N,N’,N’-tetramethylethylene-
diamine).
17350
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17349 – 17357

DBD Dyes as Fluorescence Lifetime Probes
FULL PAPER
To shed light on the reasons
for the different photophysical
properties of acyl- and ester-
DBD dyes, we performed DFT
calculations with the model
compounds 14 and 15.
After geometry optimization
(
B3LYP/6-31G*), the 30 lowest-
energy excited singlet states of
4 and 15 were calculated by
1
using time-dependent (TD)
DFT theory (TD-B3LYP/6-
[15]
3
11+ +G**). The calculated
wavelengths of the lowest-
energy transition (14: 447.8 nm,
1
5: 425.8 nm) are in fairly good
agreement with the experimen-
tal UV data of 4, 5, and 13. In
particular, the considerably
shorter absorption wavelength
of ester-DBD compared with
acyl-DBD is reproduced cor-
rectly. In both cases, the transi-
tion dipole moment is located
in the p-plane of the molecules
and follows a line connecting
Scheme 3. Preparation of 13.
Table 2. Spectroscopic properties of 13a and 13b.
Solvent
l
abs
l
em
t
F
e
F
F
e·F
F
ACHTUNGTRENNUNG[ m cm ]
ꢁ
1
ꢁ1
ꢁ1
ꢁ1
[
nm] [nm] [ns]
3b 439 548 24.8 4300
3a 406 488 17.8 3105
3b 426 545 21.4 5683
3a 403 495 15.4 4680
3b 435 578 9.8 2183
3a 408 508 17.6 5109
3b 435 603 4.7 2110
3a 407 508 16.9 5050
A
H
U
G
R
N
U
G
]
1
1
1
1
1
1
1
1
0.35
0.75
0.41
0.56
0.18
0.63
0.08
0.59
<0.05
0.63
1505
2329
2330
2621
393
3219
169
2980
–
dichloromethane
acetonitrile
ethanol
methanol
1
1
3b 446 640 1.5
3a 414 528 20.5 4947
–
2
H O
3117
Figure 1. Absorption (black) and emission (gray) spectra of ester-DBD
the acetal C atoms (Figure 2).
1
3a (solid) and acyl-DBD 13b (dotted) in MeCN.
A remarkable difference between acyl- and ester-DBD
chromophores can be derived from the details of the lowest-
energy transitions. In the case of 14, this transition is affect-
ed solely by an excitation from the HOMO (No. 65) to the
LUMO (No. 66). This excitation is accompanied by a shift
chromically shifted compared with those of acyl-DBD dyes,
and give absorption maxima around 400 nm and emission
maxima with equally broad bands around 500 nm.
The spectroscopic properties of compound 13a and its
equivalent acyl-DBD dye 13b in various solvents (as shown
in Table 2) substantiate the assumption that highly polar en-
vironments do not strongly reduce the fluorescence lifetime
and intensity for this new generation of DBD dyes. In fact,
water even increases the fluorescence lifetime by a few
nanoseconds and preserves the high fluorescence quantum
yield.
of the electron density from the [1,3]dioxolo
[1,3]benzodioxole moiety to the acyl groups, which is clearly
discernible by visualizing these orbitals (Figure 3).
ACHTUNGTRENNUNG[ 4,5-f]-
AHCTUNGTRENNUNG
In contrast to 14, the lowest-energy transition of ester-
DBD 15 is actually a mixture of the same HOMO–LUMO
excitation (73!74, CI coefficient 0.7233) and an excitation
from HOMOꢁ1 to LUMO+4 (72!78, CI coefficient
ꢁ0.10402). The latter excitation is characterized by a shift of
the electron density in the opposite direction, that is, from
Chem. Eur. J. 2013, 19, 17349 – 17357
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17351

A. Herrmann, E. Schneider, P. Wessig et al.
widely used fluorophores that show at least similar absorp-
tion and emission maxima, that is, 4-N-propargyl NBD 16a
and 5-propargylamido-carboxyfluorescein 16b, respectively.
Figure 5 illustrates the photobleaching of DBD dyes 13a
and 13b observed through the decrease in their absorption
maxima (403, 426 nm) compared with 16a and 16b at one of
their significant absorption maxima (330, 500 nm), upon ex-
posure to a 500 W mercury arc lamp combined with
a 290 nm edge filter (see the Supporting Information for ex-
perimental details). It is seen that acyl DBD 13b shows
a higher photostability than NBD dye 16a and a similar be-
havior to the carboxyfluorescein derivative 16b. Ester DBD
Figure 2. Transition dipole moment (arrows), calculated absorption wave-
length and oscillator strength of the lowest-energy transition of 14 and
1
5.
1
3a is even less photolabile under the experimental condi-
tions used.
Figure 3. HOMO (left) and LUMO (right) of acyl-DBD dye 14.
the ester groups to the [1,3]dioxolo CATHUNGTRNEUNNG[ 4,5-f] CAHTUNGTRNEUNNG[ 1,3]benzodioxole
moiety (Figure 4).
We hypothesize that the different nature of the lowest-
energy transitions of acyl- and ester-DBD dyes is responsi-
ble for the differences between their photophysical proper-
ties.
Figure 5. Photobleaching of 13a (
&
), 13b (!), 16a (N), and 16b (*) in
acetonitrile (dye concentration: 1ꢃ10 m).
ꢁ4
Biofunctionalized ester-DBD dyes: We then decided to take
advantage of these positive parameters of DBD esters. As
mentioned above, the combination of photostability and
high fluorescence quantum yields in addition to long fluores-
cence lifetimes and large Stokes shifts is rare, and should be
interesting for new probes for bioanalytical issues. Thus, var-
ious functional derivatives of 13a bearing bioreactive resi-
dues were synthesized successfully (Scheme 4). Because of
their small molecular size, these derivatives should be able
to approach small protein pockets and other less accessible
[16]
domains.
DBD dyes bearing maleimido (17), NHS ester (21), alcohol
25), azide (27), alkyne (23), sulfonate ester (26), and amino
Scheme 4 shows the synthetic routes to ester-
(
groups (19, 28). The synthesis of most bioreactive com-
Figure 4. HOMO (top left), LUMO (top right), HOMOꢁ1 (bottom left),
and LUMO+4 (bottom right) of ester-DBD dye 15.
pounds is based mainly on classic peptide-coupling tech-
[17]
niques.
Through optimization of the reaction conditions
and the testing of different coupling agents, it was possible
to obtain all the compounds in good to excellent yields. Al-
Photostability: Another important parameter for fluorescent
dyes is their photostability. Owing to the unusual combina-
tion of large Stokes shifts, long fluorescence lifetimes, and
high fluorescence quantum yields, it is difficult to compare
DBD dyes with other commonly used fluorophores. Howev-
er, we try to give a reference point in terms of the behavior
of DBD dyes concerning the issue of photostability. There-
fore, we chose to compare them with two derivatives of
cohol 25 was accessible by using a NaBH -provided reduc-
4
[18]
tion under mild conditions, as described by Kokotos et al.
These functionalized derivatives could react with corre-
sponding groups such as thiols, amines, carboxylic acids, al-
kynes, and azides, which are often targets in biological struc-
[19]
tures.
This was proven through sample reactions with
simple compounds containing such target groups.
17352
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17349 – 17357

DBD Dyes as Fluorescence Lifetime Probes
FULL PAPER
Scheme 4. Bioreactive derivatives of 13a (PyBOP=(benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate, DIPEA=diisopropylethyla-
mine, Fmoc-EDA=N1-(9-fluorenylmethyloxycarbonyl)ethylenediamine, NHS=N-hydroxysuccinimide, DCC=dicyclohexylcarbodiimide, Tos=4-tolu-
ene-sulfonyl).
[
29]
Fluorescence quenching of DBD dyes caused by male-
ACHTUNGTRENNUNGi mides: Along with the other biofunctionalized compounds,
maleimide 17 drew particular attention because its C=C
not attached directly to the probe. Starting from tert-bu-
tylcarbamate (Boc)-protected ethanol-amine, we substituted
the hydroxyl group by a maleimido residue by using a Mitsu-
nobu reaction.
ACHTUNGTRENNUNG
[30]
double bond can undergo selective addition reactions with
After deprotection, the resulting 2-male-
imido-ethylamine was reacted with 13a by using a peptide-
[20]
thiols.
Because accessible thiol groups in proteins are relatively
rare and of great interest, it was promising to attempt the
use of DBD-maleimides as probes working in thiol-contain-
coupling strategy involving (benzotriazol-1-yloxy)tripyrroli-
dinophosphonium hexafluorophosphate (PyBOP) as a cou-
pling agent.
[21–24]
ing domains of proteins.
We first investigated the spectroscopic properties of 17 in
detail. It is known that maleimides often quench the lumi-
Thiols, also known as sulfohydryls, often exist in proteins
under reducing conditions, but are more often found in
native proteins in their oxidized form as disulfide bridges.
The latter play an important role in the tertiary structure of,
for example, antibodies. Labeling of these protein domains
requires cleavage of the disulfide bonds by reducing agents
[23,31,32]
nescence of chromophores,
and as compound 17
shows almost no fluorescence at all, it is indicated that such
quenching is observed here. The fluorescence intensity and
lifetime of 17 increase strongly upon reaction of the double
bond of the maleimide with a thiol (e.g., cysteine) to form
the corresponding thioether 18. Because of the interruption
of the former conjugated p-system, quenching is now pre-
vented (Figure 6).
It is important to stress that the detected fluorescence of
maleimide 17 is actually the signal of the already reacted
maleimide with some impurities. We assume that intramo-
lecular quenching of pure 17 is so efficient as to be close to
one hundred percent. Otherwise, the detected fluorescence
decay of 17 seen in Figure 6 would not show such a sharp
edge but monoexponential behavior. As described by Guy
et al. and Maeda et al., the quenching by maleimides does
not involve singlet deactivation through intersystem crossing
[25]
to obtain the free thiol groups. This technique is common-
ly followed by labeling of the hinge region of antibodies.
Another field of interest is the labeling of artificially intro-
[26–28]
duced cysteine residues inside peptide chains.
Here, the
aim is to place a reactive thiol close to a domain where
a maleimide-functionalized probe could gather information
about its surroundings. Therefore, we intended to investigate
whether DBD dyes have the potential to fulfill such a task.
To obtain 17, we needed to synthesize a maleimide deriva-
tive bearing a short linker with a nucleophilic amine at its
end. This linker is important because the often buried thiol
groups in proteins are more accessible if the maleimide is
Chem. Eur. J. 2013, 19, 17349 – 17357
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17353

A. Herrmann, E. Schneider, P. Wessig et al.
Figure 7. Cyclic voltammogram of 13a in MeCN: reversible oxidation
stage at 1.03 V; quasi-reversible reduction stage at ꢁ1.03 V.
to give information about the quenching kinetics, and
[36]
should confirm the assumed PET mechanism. As Figure 8
illustrates, the fluorescence intensity of 13a is reduced with
increasing concentrations of NMM. We then determined the
Stern–Volmer quenching rate KSV by plotting the decreasing
fluorescence lifetime against the NMM concentration (inset
of Figure 8).
Figure 6. Fluorescence intensity and lifetime of maleimide 17 and its cor-
responding Cys-thioether 18 in water/acetone (1:1).
[23,33]
(
ISC).
Because of the lack of matching absorption and
emission spectra, Fçrster resonance energy transfer (FRET)
[34,35a]
between 13a and a maleimide can also be ruled out.
In-
stead, deactivation through photoelectron transfer (PET) is
assumed.
For a consideration of whether the fluorophore and a mal-
eimide (we chose N-methylmaleimide, NMM) are compati-
ble for PET, their electron donor and acceptor levels must
be compared by using the Rehm–Weller equation [Eq. (1)].
þ
ꢁ
2
DG ¼ EðD =DÞꢁEðA=A ÞꢁDG ꢁðe =edÞ
ð1Þ
0
0
+
Here, E
phore, E(A/A ) the reduction potential of the quencher,
ACHTUNGTRENNUNG( D /D) is the oxidation potential of the fluoro-
ꢁ
ACHTUNGTRENNUNG
Figure 8. Titration of 13a with various amounts of N-methyl-maleimide
DG00 the energy difference between the ground and excited
(
3
NMM) in MeCN: 0.0 mm (black), 1.0 mm (red), 2.7 mm (red, dotted),
.5 mm (green), 4.5 mm (green, dotted), 7.2 mm (blue), 9.9 mm (blue,
dotted).
2
state, and (e /ed) the reorganization energy. The latter cou-
ꢁ
1
lomb term can be simplified to a value of 1.3 kcalmol for
acetonitrile, because it has a minor influence compared with
[35b]
other thermodynamic parameters.
DG00 can be derived
ꢁ
1
from the point of intersection between the normalized ab-
The derived slope is KSV (144m ), which, divided by the
monoexponential fluorescence lifetime of 13a (15.4 ns),
sorption and emission spectra of the fluorophore. For 13a,
ꢁ
1
this value was found to be 460 nm (62.2 kcalmol ). The
redox potentials of 13a were determined through cyclic vol-
tammetry (Figure 7, vs. NHE), which gave an oxidation po-
tential of 1.03 V (23.8 kcal). The reduction potential of
NMM was determined to be ꢁ1.185 V (ꢁ27.21 kcal). It was
calculated that an electron transfer from 13a to NMM is en-
gives the second-order quenching rate k . The determined
q
9
ꢁ1 ꢁ1
value for 13a (9.35ꢃ10 m s ) is near the diffusion limit of
the solvent, which is consistent with a fast quenching mecha-
nism such as PET. The fact that the emission maxima did
not shift at any given NMM concentration but that the fluo-
rescence lifetime decreased strongly indicates that the
ꢁ
1
[37]
ergetically preferred (DG=ꢁ12.51 kcalmol ).
quenching must be a dynamic process.
To confirm these results further, we investigated whether
the steady-state fluorescence and fluorescence lifetime de-
crease when the quencher (NMM) is added to a solution of
Besides the ester-DBD 13a, identical measurements and
calculations were made for the equivalent acyl-DBD 13b,
and similar results were obtained (see the Supporting Infor-
mation for details).
13a in MeCN. This Stern–Volmer experiment was expected
17354
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17349 – 17357

DBD Dyes as Fluorescence Lifetime Probes
FULL PAPER
Hence, we were able to confirm PET as the preferred
quenching mechanism for ester- and acyl-DBD fluorophores
in the presence of maleimides.
Maleimido-DBD dyes as fluorescence lifetime probes for
monitoring conformational changes in proteins: This very
effective fluorescence quenching of DBD-maleimides al-
lowed these probes to report clearly when any bonding to
corresponding thiols actually occurred, because only then is
[21]
the fluorescence detectable. We expected that maleimides
of both generations would show distinct fluorescence life-
time changes upon a change in their microenvironment cre-
ating a new surrounding polarity. To probe this, we chose
the well-characterized maltose ATP-binding cassette trans-
porter MalFGK as a biological test system, taking 17 and
2
Figure 9. Simulation of the location of 17 in MalF
2
ACHTUNGTNERNGNU( S252C)GK (pdb
32 (Scheme 5) as promising representative labels. MalFGK2
3PV0). The transporter is trapped in an intermediate prehydrolytic
form.
[
38]
Scheme 5. Synthesis of maleimide 32.
is a worthwhile target protein, because studies on the con-
formational change during the transition from the cofactor-
less state (apo-state) to an intermediate ATP/ADP-bound
state (trap-state), which is acquired by adding vanadate to
the ATP hydrolyzing transport complex (see the Supporting
Information), have been performed previously by limiting
proteolysis and using steady-state fluorescence spectrosco-
[
28]
py. The transport complex to which the substrate is deliv-
ered by maltose binding protein (MBP) consists of one copy
of each of the transmembrane subunits (MalF and MalG)
and two copies of the nucleotide-binding subunit (MalK).
The binding of the nucleotide to MalK causes a conforma-
tional change in two tryptic cleavage sites in the periplas-
matic loops (P2 of MalF and P1 of MalG). The previously
Figure 10. Changes in fluorescence lifetimes and emission maxima of 32
and 17 located in MalF(S252C)GK (apo and trap forms).
A
H
U
G
R
N
U
G
2
state conformation. In line with previous data, this confor-
mation presumably creates a more apolar microenvironment
2
52
[28]
constructed variant Ser Cys of MalF located ten residues
to which the probe responds. As expected, the opposite
2
62
apart from the putative tryptic cleavage site; Lys was used
signal-change tendency is observed for 17. Here, the fluores-
cence lifetime is reduced by 1.6 ns in the trap-state of MalF.
However, the value is still very high (10.4 ns) and far from
any background fluorescence decay. The according behavior
of both probes is also observed in the shift of the emission
maxima, as expected from the spectroscopic measurements
of the original compounds (Table 2). Whereas 32 shows a sig-
nificant blueshift of 9 nm in the trap-state, which again con-
[28]
for labeling this subdomain with DBD-maleimides. Exem-
plarily for both probes, Figure 9 illustrates schematically the
location of 17 in the maltose ABC transporter.
Figure 10 presents the fluorescence lifetime signals detect-
ed from both probes as well as the shifts of the emission
maxima. The average fluorescence lifetime of 32 increases
by more than 1 ns up to 6.3 ns when MalF is in the trap-
Chem. Eur. J. 2013, 19, 17349 – 17357
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17355

A. Herrmann, E. Schneider, P. Wessig et al.
firms its movement to a more apolar environment, 17 does
not display such behavior, as the spectrum is insensitive to
changes in polarity. These findings correlate perfectly with
previous observations proving that the orthogonal behaviors
of acyl- and ester-DBDs concerning fluorescence lifetime
changes make them a very handy tool, as they deliver a sen-
sitive positive/negative control.
These results encouraged us to adapt the approach to an-
other system that performs larger and well-studied move-
ments: the maltose-binding protein MBP. Figure 11 illus-
Figure 12. Changes in fluorescence lifetimes and emission maxima of 32
1
3
and 17 located in MalE ACHUTNGRNENUG( G C) (maltose ligated and unligated form).
ronment changes from apolar to polar as the lobes of MBP
close, because the fluorescence lifetime of 32 decreases by
1
.3 ns and that of 17 increases by about 3 ns after the addi-
tion of maltose. The reciprocal behavior of the two male-
AHCTUNGTREGiNNUN mide labels provides internal confirmation of the consisten-
cy of these data, which is consistent with the shift of the
emission maxima to longer wavelengths by 4 nm with 32 in
the closed conformation, whereas those of 17 remain con-
stant. However, the opposite result could have been pro-
posed with a more packed yet more apolar neighborhood of
this position in the ligand form of MBP.
Apparently, the Gly Cys position bends towards a more
polar ambit. The fluorophore is probably oriented outward,
toward the interface between MBP and the transmembrane
subunits, thus preventing productive interaction. Together,
both test systems confirm the suitability of the developed
Figure 11. The N- and C-lobes of maltose-binding protein reside in an
“
open” conformation in the absence of a ligand (top) and change confor-
mation to a “closed” form upon trapping of maltose (blue, “Venus fly-
trap” model; bottom). The location of Gly13 residues is indicated and
maltose is shown in space-filling representation (in blue). Pdb files 1N3X
1
3
[
39,40]
(
open) and 1ANF (closed).
trates the conformational change of MBP resulting from the
1
3
binding of maltose and the location of Gly at the entrance
maleimido-DBD dyes for monitoring microenvi AHCTUNGTERNNUNGr on ACHTUNGTRENNUNGm ental
1
3
of the maltose-binding pocket. Gly changes its relative po-
sition significantly during the transition from the open to
the closed conformation. Substitution of this glycine with
changes in biological samples, providing an internal control
through their juxtaposing behavior and giving precise and
reliable information thanks to their long but sensitive fluo-
rescence lifetimes.
13
cysteine (Gly Cys) provides a protein that can be labeled
with 17 and 32, respectively, monitoring the changes in the
specific microenvironment. The labeled variant retained its
ability to bind maltose, as verified through its binding to an
amylose matrix and elution with maltose (see the Support-
ing Information).
Conclusion
We have presented the synthesis and spectroscopic proper-
1
3
The 17- and 32-labelled Gly Cys variant displays fluores-
ties of novel fluorophores containing a [1,3]dioxolo
ACHTUNGTRENNUNG[ 4,5-f]-
cence lifetime changes (Figure 12) that suggest that the envi-
ACHTUNGTREN[NGNU 1,3]benzodioxole (DBD) backbone. In comparison with
17356
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17349 – 17357

DBD Dyes as Fluorescence Lifetime Probes
FULL PAPER
previously described DBD dyes, they show long fluores-
cence decay and high quantum yields, even in aqueous solu-
tions. Furthermore, they can be functionalized for reaction
with biomolecular targets. Maleimido derivatives of both
generations were shown to undergo a photoelectron transfer
[15] M. A. L. Marques, C. A. Ullrich, F. Nogueira, A. Rubio, K. Burke,
E. K. U. Gross, Time-Dependent Density Functional Theory, Spring-
er, Heidelberg, 2006.
[
[
16] T. Kawabata, N. Go, Proteins Struct. Funct. Bioinf. 2007, 68, 516.
17] A. El-Faham, F. Albericio, Chem. Rev. 2011, 111, 6557–6602.
[18] G. Kokotos, Synthesis 1990, 4, 299–301.
[
19] G. T. Hermanson, Bioconjugate Techniques, 2nd ed., Elsevier, Am-
sterdam 2008, p. 169.
(
PET) quenching manifold. We labeled these dyes to pro-
tein pockets containing a central thiol group and investigat-
ed their spectroscopic behaviors during conformational
changes in the proteins. In combination, ester- and acyl-
DBD dyes are a sensitive toolkit for monitoring microenvir-
onmental changes in biological samples, providing an inter-
nal control due to the juxtaposition of their fluorescence
lifetime behaviors.
[
[
20] C. F. Brewer, J. P. Riehm, Anal. Biochem. 1967, 18, 248–255.
21] K. Huang, W. Bulika, A. Marti, Chem. Commun. 2012, 48, 11760–
11762.
[22] X. Wang, S. Wang, H. Ma, Analyst 2008, 133, 478–484.
[
[
[
[
23] J. Guy, K. Caron, S. Dufresne, S. W. Michnick, W. G. Skene, J. W. J.
Keillor, J. Am. Chem. Soc. 2007, 129, 11969–11977.
24] K. Tyagarajan, E. Pretzler, J. E. Wiktorowicz, Electrophoresis 2003,
2
4, 2348–2358.
25] R. Singh, G. V. Lamoureux, W. J. Lees, G. M. Whitesides, Methods
Enzymol. 1995, 251, 167–173.
26] T. Kurpiers, H. Mootz, Angew. Chem. 2007, 119, 5327–5330; Angew.
Chem. Int. Ed. 2007, 46, 5234–5237.
Acknowledgements
[
[
27] T. Kurpiers, H. Mootz, ChemBioChem 2008, 9, 2317–2325.
28] M. L. Daus, H. Landmesser, A. Schlosser, P. Mꢀller, A. Herrmann,
E. Schneider, J. Biol. Chem. 2006, 281, 3856–3865.
We thank Dr. W. Mickler for performing the cyclovoltammometric meas-
urements and Katlen Brennenstuhl for performing the fluorescence
quantum yield measurements.
[
[
[
[
[
[
[
29] M. Link, X. Li, J. Kleim, O. S. Wolfbeis, Eur. J. Org. Chem. 2010,
6
922–6927.
30] C. Antczak, B. Bauvois, C. Monneret, J.-C. Florent, Bioorg. Med.
Chem. 2001, 9, 2843–2848.
31] X. Chen, Y. Zhou, X. Peng, J. Yoon, Chem. Soc. Rev. 2010, 39,
[1] M. Y. Berezin, S. Achilefu, Chem. Rev. 2010, 110, 2641–2684.
[2] J. W. Borst, A. Visser, Meas. Sci. Technol. 2010, 21, 102002.
[3] R. N. Dsouza, U. Pischel, W. M. Nau, Chem. Rev. 2011, 111, 7941–
2
120–2135.
32] K. Caron, V. Lachapelle, J. W. Keillor, Org. Biomol. Chem. 2011, 9,
85–197.
7
980.
4] S. C. Mondal, K. Sahu, K. Bhattacharyya, in Reviews in Fluorescence
007 (Ed.: C. D. Geddes), Springer, New York, 2009, pp. 157–177
5] G. Bains, A. B. Patel, V. Narayanaswami, Molecules 2011, 16, 7909–
935.
6] M. Beija, C. A. M. Alfonso, J. M. G. Martinho, Chem. Soc. Rev.
009, 38, 2410–2433.
[
[
[
1
2
33] H. Maeda, T. Maeda, K. Mizuno, K. Fujimoto, H. Shimizu, M.
Inouye, Chem. Eur. J. 2006, 12, 824–831.
34] C. W. Miller, E. S. Jçnsson, C. E. Hoyle, K. Viswanathan, E. J. Val-
ente, J. Phys. Chem. B 2001, 105, 2707–2717.
35] J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer,
New York, 2006; a) p. 443, b) p. 337.
36] O. Stern, M. Z. Volmer, Phys. Z. 1919, 20, 183–188.
37] L. K. Fraiji, D. M. Hayes, T. C. Werner, J. Chem. Educ. 1992, 69,
7
2
[
[
[
7] B. D. Wagner, Molecules 2009, 14, 210–237.
8] C. Sun, J. Yang, L. Li, S. Lui, J. Chromatogr. B 2004, 803, 173–190.
9] P. Wessig, R. Wawrzinek, K. Mçllnitz, E. Feldbusch, U. Schilde, Tet-
rahedron Lett. 2011, 52, 6192–6195.
[
[
4
24–428.
[
[
10] P. Wessig, R. Wawrzinek, K. Mçllnitz, EP2399913 A1, 2010.
11] R. Wawrzinek, P. Wessig, K. Mçllnitz, J. Nikolaus, R. Schwarzer, P.
Mꢀller, A. Herrmann, Bioorg. Med. Chem. Lett. 2012, 22, 5367–
[
[
[
38] M. L. Oldham, J. Chen, Science 2011, 332, 1202–1205.
39] P. G. Telmer, B. H. Shilton, J. Biol. Chem. 2003, 278, 34555–34567.
40] F. A. Quiocho, J. C. Spurlino, L. E. Rodseth, Structure 1997, 5, 997–
5
371.
12] E. Bordignon, M. Grote, E. Schneider, Mol. Microbiol. 2010, 77,
354–1366.
13] Y. Q. Ye, H. Koshino, J. Onose, K. Yoshikawa, N. Abe, S. Takahashi,
Org. Lett. 2007, 9, 4131–4134.
1
015.
[
[
[
1
Received: June 20, 2013
Revised: September 6, 2013
14] M.-J. Fan, G.-Q. Li, L.-H. Li, S.-D. Yang, Y.-M. Liang, Synthesis
2
006, 14, 2286–2292
Published online: November 8, 2013
Chem. Eur. J. 2013, 19, 17349 – 17357
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17357