Verification and biophysical characterization of a new three-color Foerster resonance-energy-transfer (FRET) system in DNA

Article abstract of DOI:10.1002/hlca.201100460

We report on a new three-color FRET system consisting of three fluorescent dyes, i.e., of a carbostyril (= quinolin-2(1H)-one)-derived donor D, a (bathophenanthroline)ruthenium complex as a relay chromophore A₁, and a Cy dye as A₂ (FRET = Foerster resonance-energy-transfer) (cf. Fig. 1). With their widely matching spectroscopic properties (cf. Fig. 2), the combination of these dyes yielded excellent FRET efficiencies. Furthermore, fluorescence lifetime measurements revealed that the long fluorescence lifetime of the Ru complex was transferred to the Cy dye offering the possibility to measure the whole system in a time-resolved mode. The FRET system was established on double-stranded DNA (cf. Fig. 3) but it should also be generally applicable to other biomolecules.

Full text of DOI:10.1002/hlca.201100460

Helvetica Chimica Acta – Vol. 95 (2012)
543
Verification and Biophysical Characterization of a New Three-Color Fçrster
Resonance-Energy-Transfer (FRET) System in DNA
by Andrea Altevogt ne´e Kienzlera), Roman Flehrb), Sçren Gehneb), Michael U. Kumkeb), and
Willi Bannwarth*a)
a) Albert-Ludwigs University of Freiburg, Institute of Organic Chemistry and Biochemistry,
Albertstr. 21, D-79104 Freiburg (phone: þ 49-761-2036073; fax: þ 49-761-2038705;
e-mail: willi.bannwarth@chemie.uni-freiburg.de)
b) University of Potsdam, Institute of Chemistry (Physical Chemistry), Karl-Liebknecht-Str. 24 – 25,
D-14476 Potsdam-Golm
We report on a new three-color FRET system consisting of three fluorescent dyes, i.e., of a
carbostyril (¼quinolin-2(1H)-one)-derived donor D, a (bathophenanthroline)ruthenium complex as a
relay chromophore A1, and a Cy dye as A2 (FRET¼ Fçrster resonance-energy-transfer) (cf. Fig. 1). With
their widely matching spectroscopic properties (cf. Fig. 2), the combination of these dyes yielded
excellent FRET efficiencies. Furthermore, fluorescence lifetime measurements revealed that the long
fluorescence lifetime of the Ru complex was transferred to the Cy dye offering the possibility to measure
the whole system in a time-resolved mode. The FRET system was established on double-stranded DNA
(cf. Fig. 3) but it should also be generally applicable to other biomolecules.
Introduction. – For the elucidation of distance-dependent biochemical events
Fçrster resonance-energy-transfer (FRET) systems have emerged as powerful tools,
since they allow to monitor distance changes on the nanometer scale and in real time
[1][2]. FRET Efficiencies are dependent on a number of parameters like the spectral
overlap of the donor (D) emission with the absorption of the acceptor dye (A), as well
as the distance from and orientation to each other [3].
Up to now, a whole plethora of suitable dye combinations for two-color systems has
been reported, but it has turned out that the development of so called three-color or
triple FRET (tc-FRET) is much more demanding. A tc-FRET consists of two coupled
donorꢀacceptor pairs sharing one chromophore which acts as a relay station A1 for the
energy transfer from the donor D to the acceptor A2 (Fig. 1).
Fig. 1. Principle of the three-color FRET system
Such systems offer the advantage that they yield information about the position of
the three chromophores relative to each other, they extend the working range
ꢀ 2012 Verlag Helvetica Chimica Acta AG, Zꢁrich

544
Helvetica Chimica Acta – Vol. 95 (2012)
compared to conventional FRET systems, and the spectral shift between the excitation
of the donor and the emission of A2 is much more pronounced. In addition, fewer
labeled sample molecules are required to measure relative distances [4][5]. Despite
these obvious advantages, relatively few systems have been hitherto reported [6 – 17].
The main reason for this is the difficulty to identify three chromophores with matching
spectroscopic properties as well as the development of suitable chemistry required for
their specific incorporation into biomolecules.
We have recently reported on a novel tc-FRET system which we had successfully
tested in peptides as well as in synthetic DNA [18]. The system consisted of a
carbostyril (¼quinolin-2(1H)-one)-derived donor D, a [RuII(bathophenanthroline)]
acceptor A1, and an anthraquinone (¼anthracene-9,10-dione) quencher A2 (batho-
phenanthroline ¼ 4,7-diphenyl-1,10-phenanthroline). All three chromophores were
introduced by a building-block approach avoiding postsynthetic labeling. Additional
features of our new system were robustness, outstanding well-matching spectroscopic
properties, large Fçrster distances, and high sensitivity. The full set of fundamental
photophysical properties was investigated in great detail. The experimental data were
in excellent agreement with theoretical calculations that were also performed for this
tc-FRET system.
An anthraquinone chromophore served as quencher entity (A2 in Fig. 1) for the Ru
complex, and no sensitized emission of the final FRET step was observed. This has, on
the one hand, the advantage that the photoluminescence of the Ru complex is not
obscured by an emission of the quencher, but it also has, on the other hand, the
consequence that no positive signal of A2 can be observed, which could be of additional
analytical benefit, e.g., by increasing the usable spectral shift further into the NIR
spectral region.
The aim of this work was to replace the anthraquinone quencher in our original tc-
FRET system by an emitting dye. This would lead to a positive, even more red-shifted
signal as a result of the second energy transfer. Moreover, if it was possible to transfer a
long decay time on this dye, it would have generally implications for such systems.
Insertion of a relay chromophore with a distinct longer luminescence emission
compared to conventional organic dyes would be enough to turn the system into a
position that it can be measured in a time resolved mode with the possibility to largely
eliminate any background fluorescence. Since the luminescence of the Ru complex
decays on a ms time scale, the decay time characteristics of the dye used as A2 should be
significantly increased, and the application of a time-gated detection scheme in
luminescence assays based on this tc-FRET system could be envisioned.
As a suitable chromophore, we have identified the cyanine dye 6 which was
synthesized according to the Scheme [19]. Thus, 2,3,3-trimethyl-3H-indol (1) was first
treated with 6-bromohexanoic acid (2) to form the indolium salt 3. Upon reaction
with the hydrochloride of the bis-imine 4 of malonaldehyde, the hemicyanine was
formed which was then treated with indolium salt 5 to yield finally the desired Cy5
derivative 6.
A comparison of the absorption/emission spectra of the carbostyril-derived donor
(D), the Ru complex (Ru), and the Cy5 derivative 6 (Cy) revealed a sufficiently strong
overlap between the emission of the Ru complex with the absorption of
6
demonstrating its potential suitability for the aspired tc-FRET system (Fig. 2).

Helvetica Chimica Acta – Vol. 95 (2012)
Scheme. Synthesis of Cyanine Dye 6 (Cy)
545
a) Microwave oven (1108, 15 bar max, 50 W), 35 min; 69%. b) Ac2O, 1208, 90 min. c) Pyridine, r.t.,
overnight; yield over 2 steps 85%.
Fig. 2. Spectral overlaps among the dyes
Synthetic double-stranded DNA decorated with the chromophores of a FRET
system is an ideal tool for the systematic investigation of the latter. It allows to place the
chromophores at will into defined positions within the DNA. Furthermore, due to the
rigid core structure, the distances between the dyes are controllable and can be varied
according to the requirements. In addition, influences based on ground-state
interactions hampering the estimation of reliable numerical values characterizing the
energy-transfer process (e.g., energy-transfer efficiencies, Fçrster radii) can be
minimized. For these obvious reasons, we aspired towards the verification of the new
tc-FRET system in DNA via the entities outlined in Fig. 3.

546
Helvetica Chimica Acta – Vol. 95 (2012)
Fig. 3. Chromophores attached to the DNA
Results and Discussion. – The incorporation of the carbostyril-derived donor and
the [RuII(bathophenanthroline)] complex was accomplished via their phosphorami-
dites and was described in detail in our previous publication [18]. Preliminary
experiments had revealed that cyanine dyes are not perfectly stable under the
conditions employed during DNA synthesis. This was unfortunately also true for the
Cy5 derivative 6, so that its incorporation by means of a phosphoramidite building
block was doomed to failure. Hence, we had to pursue a postsynthetic labeling. For this
purpose, 6 was coupled to an amino-modified synthetic DNA after its release from the
support and deprotection via an amide function. Purification of the desired DNA
fragment after this step was then performed by prep. HPLC on reversed-phase silica
gel.
For a detailed biophysical characterization of the new tc-FRET system, the DNA
sequences 7 – 17 were required (Fig. 4). All sequences containing the Cy5 chromo-
phore were prepared by a postsynthetic labeling procedure, the details of which are
outlined in [20]. The unlabeled sequences were the same as those used in our
previously reported system. This held also true for the DNA fragment containing the
[RuII(bathophenanthroline)] complex as label.

Helvetica Chimica Acta – Vol. 95 (2012)
547
Fig. 4. Sequences used for the photophysical investigations
The tc-FRET system was then established by hybridizing the Ru-complex-modified
fragment to the complementary strands equipped with the donor and 6 (Cy) as depicted
in Fig. 5. The different hybrids were then investigated by photophysical measure-
ments.
The synthesized array of sequences would allow us to characterize double-stranded
samples with different relative distances between the donor (D), Ru (A1), and Cy (A2).
As outlined in the previous publication, the different dyes were in all hybrids well
separated, so that significant ground-state interactions, e.g., due to orbital overlap,
could be excluded. This also allowed for a nondisturbed investigation of both FRET
pairs involved. The first energy-transfer step from D to the Ru complex has already
been investigated in detail in our previously described tc-FRET system and compared
to other donor derivatives [21].
A comparison of the spectral data of free 6 and the sequences 10 and 11 indicated
only minor differences concerning the absorption coefficient as well as the wavelength
of the absorption maximum (only a slight bathochromic spectral shift Dl of 4 nm). This
indicated that between the DNA and incorporated 6 (Cy), no significant ground-state
interaction was operative, which has been reported for many other dyes before [22].

548
Helvetica Chimica Acta – Vol. 95 (2012)
Fig. 5. Principle of the three-color-FRET DNA arrangement
The combination between sequence 7 and 10 (Cy2) or 11 (Cy1) was used for the
study of the second FRET pair consisting of the Ru complex and the cyanine
chromophore 6 (Cy). Upon excitation at lex 430 nm, sequence 7 revealed the strong
emission at lem 620 nm. After hybridization to 11 (Cy1), this emission was strongly
reduced concomitantly with a strong emission at lem 665 nm indicative of an effective
FRET. This effect was as expected less pronounced for the combination of 7/10 due to
the greater distance between the two dyes (Fig. 6).
Fig. 6. Steady-state measurements (lex 430 nm) of the FRET pair Ru/Cy at different distances between
them in hybridized DNA strands (9/7: no Cy; 10/7: Cy2; 11/7: Cy1)
The combination between sequences 12 and 13 with the Ru-complex-labeled DNA
fragment 7 was then used to confirm the functionality of the tc-FRET system. This was
done in comparison with the combination of sequence 16 and 7 confirming the first

Helvetica Chimica Acta – Vol. 95 (2012)
549
FRET step between the carbostyril-derived donor and the [RuII(bathophenanthro-
line)] complex in the presence of Cy and with sequence 16 containing just the donor as
a control. Upon excitation at lex 360 nm, sample 16 showed a strong emission at lem ca.
440 nm. As expected, this emission was declined in the combination 16/7 due to the
energy transfer to the Ru complex resulting in the sensitized emission at lem ca. 620 nm.
In the tc-FRET arrangement 12/7, this emission was clearly reduced giving rise, at the
same time, to an emission at lem 665 nm (Fig. 7), clearly indicative of an active second
FRET step. As expected, the tc-FRET was reduced due to the larger distance between
the Ru complex and Cy in the sample 13/7.
Fig. 7. Steady-state emission spectra of the FRET constructs in hybridized DNA strands excited at lex
360 nm
For the determination of the FRET efficiency h, Eqn. 1 can be used, where I is the
luminescence intensity of D in the presence and I0 in the absence of A. The relationship
between the FRET efficiency h and the distance R between D and A can be seen in
Eqn. 2. With the Fçrster radius R0 of a given FRET pair and the calculated energy-
transfer efficiency, it is possible to determine the spectroscopic distance Rspec (¼ R)
between D and A.
I
I0
h ¼ 1 ꢀ
(1)
(2)
"
#
1
6
ꢀ1
ꢀ
ꢁ
ꢂꢀ ꢁ
ꢃ
6
R
R0
1
h
h ¼ 1 þ
$ R ¼ R0
ꢀ 1
In the Table, the FRET efficiencies are summarized together with the distances
between the dyes. The distances Rtheo obtained from accessible volume (AV)
simulations [18] and the R values calculated from FRET data of stationary-emission-
intensity measurements were in excellent agreement (see Table).
An interesting issue is the dependence of the fluorescence decay time of the Ru
complex and the Cy derivative. Changes in fluorescence decay times are of special

550
Helvetica Chimica Acta – Vol. 95 (2012)
Table 1. FRET Efficiencies Based on Emission Intensity Data and Distances R Between the Different Dyes.
Compared are the theoretical and the experimentally determined distances.
theo
theo
spec
spec
Sample
hD/Ru
hRu/Cy
RD/Ru
RRu/Cy
RD/Ru
RRu/Cy
[nm]a)
[nm]a)
[nm]b)
[nm]c)
17/7
14/7
15/7
16/7
13/7
12/7
10/7
11/7
D2-Ru-
0.22 ꢁ 0.08
0.23 ꢁ 0.07
0.26 ꢁ 0.03
0.52 ꢁ 0.10
0.53 ꢁ 0.10
0.44 ꢁ 0.15
–
–
4.6 ꢁ 0.7
4.6 ꢁ 0.7
4.6 ꢁ 0.7
3.2 ꢁ 0.5
3.2 ꢁ 0.5
3.2 ꢁ 0.5
–
–
4.7 ꢁ 0.5
4.7 ꢁ 0.4
4.5 ꢁ 0.1
3.8 ꢁ 0.3
3.7 ꢁ 0.3
4.0 ꢁ 0.5
–
–
D2-Ru-Cy2
D2-Ru-Cy1
D1-Ru-
D1-Ru-Cy2
D1-Ru-Cy1
-Ru-Cy2
0.56 ꢁ 0.06
0.61 ꢁ 0.05
–
0.53 ꢁ 0.08
0.66 ꢁ 0.05
0.51 ꢁ 0.07
0.68 ꢁ 0.09
4.3 ꢁ 0.8
3.8 ꢁ 0.9
–
4.3 ꢁ 0.8
3.8 ꢁ 0.9
4.3 ꢁ 0.8
3.8 ꢁ 0.9
4.8 ꢁ 0.2
4.7 ꢁ 0.2
–
4.9 ꢁ 0.2
4.5 ꢁ 0.2
5.0 ꢁ 0.2
4.4 ꢁ 0.3
-Ru-Cy1
–
–
–
a
d
theo
spec
Sample
hD/Cy
RD/Cy
[nm] )
RD/Cy [nm] )
14/8
15/8
13/8
12/8
D2- -Cy2
D2- -Cy1
D1- -Cy2
D1- -Cy1
0.11 ꢁ 0.09
0.13 ꢁ 0.08
0.14 ꢁ 0.08
0.21 ꢁ 0.04
7.4 ꢁ 0.7
6.4 ꢁ 0.6
6.4 ꢁ 0.6
4.1 ꢁ 0.7
4.7 ꢁ 0.7
4.5 ꢁ 0.7
4.5 ꢁ 0.7
4.1 ꢁ 0.8
a) Calculated from AV simulations. b) With R0(D/Ru) ¼ 38.2 ꢂ. c) With R0(Ru/Cy) ¼ 50.0 ꢂ. d) With R0(D/Cy)
¼
33.3 ꢂ.
interest since they are basically independent on the concentration of the fluorophore
and, therefore, not prone to any artefacts induced by fluctuations of the concentrations
of donor and acceptor. This makes a data evaluation based on decay-time measure-
ments more robust, which is especially beneficial in life-science applications. In Fig. 8,
the fluorescence decay time of the samples 9/7, 10/7, and 11/7 are shown. In the presence
of Cy, the fluorescence decay kinetics of the Ru complex were distinctly altered. With
decreasing distance between Cy and the Ru complex, an increase in the observed
quenching was found (first part of the decay curves of samples 10/7 and 11/7, resp.). In
Fig. 8. Changes in the luminescence decay of the emission at lem 620 nm of the Ru complex due to the
FRET between the Ru complex and Cy dye in hybridized DNA strands (lex 430 nm)

Helvetica Chimica Acta – Vol. 95 (2012)
551
the absence of Cy, a simple luminescence decay of the Ru complex (according to a first-
order decay law) was observed. The presence of Cy induced a more complex
fluorescence decay kinetics. A tentative explanation for the observed time dependence
is the presence of slightly different distances between donor (Ru complex) and
acceptor (Cy) during the liftetime of the electronically excited Ru complex, resulting in
a distribution of energy transfer efficiencies. Here, work is in progress to analyse the
experimental data with more sophisticated decay models considering such a
distribution in the distances between the Ru complex and Cy.
A further issue of interest is the influence of the long decay time of the emission of
the Ru complex on the fluorescence decay time of Cy. For this investigation, we used
again the combination of the sequences 10 and 11 with the Ru-complex-labeled
sequence 7. In Fig. 9, the fluorescence decay kinetics of Cy upon indirect excitation (lex
430 nm) via the Ru complex (and subsequent transfer of energy to Cy by FRET) as
well as upon direct excitation (lex 590 nm) are shown for the samples 10/7 and 11/7. In
case of direct excitation of Cy, no difference in the decay kinetics were observed for the
two samples. In contrast, for the sensitized excitation, i) a distinct change compared to
direct excitation, and ii) a clear relation to the relative distance between D and A was
found. Due to the energy transfer, the long decay kinetics of the Ru complex was
ꢃtransferredꢄ to the Cy dye. Like in the well-known systems in which lanthanide
complexes (often containing europium or terbium as central ions) are used as donor
and the energy is transferred to organic dyes (e.g., in homogeneous fluorescence
immuno assays (FIA) [23]), the decay kinetics of the Ru complex was ꢃpunchedꢄ on the
Cy dye. Based on the information gained in time-resolved fluorescence experiments,
contributions from direct vs. indirect (equal to successful FRET) could be readily
distinguished. Moreover, any fluorescence arising from matrix components that may
obscure the measurement could now also be excluded, since those signals are expected
to decay on a lower nanosecond timescale. Compared to the lanthanide-based systems
commonly used in FIA applications, Ru complexes show higher thermodynamic and
(even more important) kinetic stability, which make such FRET systems highly
attractive as an alternative for FIA-based diagnostic applications.
Fig. 9. Lifetime extension caused by energy transfer (lem 670 nm) in hybridized DNA strands

552
Helvetica Chimica Acta – Vol. 95 (2012)
Conclusion and Outlook. – We successfully established a novel three-color FRET
system consisting of a carbostyril-derived donor, a [RuII(bathophenanthroline)]
complex, which served as a relay chromophore, and a Cy5 derivative as a second
acceptor (A2). With their widely matching spectroscopic properties, the combination of
these three chromophores yielded excellent FRET efficiencies with respect to the
single as well as the overall transfer step. In contrast to our earlier published tc-FRET
system, where a nonfluorescent anthraquinone quencher was employed as A2, we used
in this approach three emitting chromophores. The benefit of such a system is that the
FRET efficiency could now also be determined via the emission of the third
chromophore. Furthermore, the long emission luminescence kinetics of the Ru-
complex was transferred to the Cy dye so that the decay time of the Cy chromophore
was significantly increased, and the whole tc-FRET system could be measured in a
time-resolved mode, virtually free of background, with high sensitivity. The system was
established in synthetic DNA as a show case, but it should find general application in
interaction studies of biomolecules.
Experimental Part
General. Column chromatography (CC): silica gel 60 (SiO2; ACC 35 – 70 mm). UV Spectra: Perkin-
Elmer-Lambda-35 UV/VIS spectrometer. 1H-, 13C-, and 31P-NMR Spectra: AC 250, AM 400, and DRX
500 (Bruker), and Mercury VX 300 (Varian); CDCl3, CD3OD, CD3CN, or (D6)DMSO as solvent; d in
ppm, J in Hz. APCI-MS: LCQ Advantage (Thermo); APCI ¼ atmospheric-pressure chemical ionization;
in m/z (rel. %).
General Procedures for the Synthesis and Characterization of Oligonucleotides. Oligonucleotide
syntheses were carried out on an ExpediteTM 8909 nucleic acid synthesis system on a 1 mmol scale by
means of phosphoramidite chemistry and standard protocols. Phosphoramidites as well as nucleosides
coupled to solid support (CPG) were obtained from Proligo/SigmaꢀAldrich. Cleavage of the modified
and unmodified oligonucleotides from the solid support and concomitant deprotection was performed by
adding a 25% NH3 soln. followed by vortexing overnight at r.t. After exchanging the counter ion NHþ4 by
Kþ, the oligonucleotide samples were desalted by size-exclusion chromatography (NAP-10 column).
Purification of the donor- and Ru-complex-labeled oligonucleotides was performed by prep. poly-
acrylamide-gel electrophoresis (PAGE). Pre-electrophoresis was performed overnight at 400 V with
Tris/borate buffer. Oligonucleotide (5 ml, 1 OD/ml in H2O) and bromophenol blue/xylenecyanol soln.
(5 ml) were heated to 908 for 2 min and rapidly cooled to 08 before being loaded onto the gel.
Electrophoretic separation was performed for 18 h at 400 V. The gel was visualized at 366 nm, and the
bands containing the desired DNAwere isolated and cut into small slices. Together with twice the volume
of MilliQ water, they were extracted overnight. After centrifugation, the supernatant was collected and
concentrated. The resulting donor- and Ru-complex-labeled oligonucleotide samples were desalted by
size-exclusion chromatography (NAP-10 column). Purification of the Cy-labeled oligonucleotides was
performed by prep. HPLC (Merck/Hitachi system, anal. EC-125/4-Nucleosil-100-5-C18 column). The
mobile phase was a gradient of solvent A (0.1m Et3NH(OAc) buffer at pH 7.0) and solvent B (MeCN).
All oligonucleotides were analyzed by electrophoresis on polyacrylamide gels (20%) of 0.4 mm
thickness. Pre-electrophoresis was performed for 2 h at 500 V with Tris/borate running buffer.
Oligonucleotides (1 ml, 0.1 OD/ml in H2O) and bromophenol blue/xylenecyanol soln. (2 ml) were heated
to 908 for 2 min and rapidly cooled to 08 before being loaded onto the gel. Electrophoretic separations
were performed for 2 h at 500 V. Oligonucleotide bands were stained with a soln. of 3,3’-diethyl-9-
methyl-4,5,4’,5’-dibenzothiacarbocyanine bromide (ꢃStains Allꢄ, Fluka).
1-(5-Carboxypentyl)-2,3,3-trimethyl-3H-indolium Bromide (3 · HBr) [19]. A mixture of 6-bromo-
hexanoic acid (2; 670 mg, 3.4 mmol, 1 equiv.) and 2,3,3-trimethyl-3H-indol (1; 541 mg, 3.4 mmol,
1 equiv.) was irradiated for 35 min. in a microwave oven (1108, 15 bar max, 50 W). Then the mixture was

Helvetica Chimica Acta – Vol. 95 (2012)
553
dissolved in MeCN (3 ml) and the crude product precipitated with Et2O (50 ml). The obtained crystals
were filtered off and purified by CC (SiO2; 2 ! 10% MeOH/CH2Cl2): 3 · HBr (739 mg, 69%). Pink
crystals. M.p. 122 – 1258 ([19]: 127 – 1298). Spectroscopic data: in accordance with literature data.
1H-NMR ((D6)DMSO, 400 MHz): 1.41 – 1.46 (m, COCH2CH2CH2); 1.51 – 1.58 (s, 2 MeꢀC(3),
COCH2CH2); 1.80 – 1.86 (m, NCH2CH2), 2.22 (t, J(CH2,CH2) ¼ 7.2, COCH2CH2); 2.85 (s, MeꢀC(2));
4.46 (t, J(CH2,CH2) ¼ 7.7, NCH2); 7.61 – 7.63 (m, HꢀC(5), HꢀC(7)); 7.84 – 7.86 (m, HꢀC(6)); 7.96 – 7.99
(m, HꢀC(4)); 12.00 (br. s, NH). 13C-NMR ((D6)DMSO, 100 MHz): 13.98 (Me); 21.97 (CH2CH2COOH);
23.98 (Me); 25.38 (Me); 26.90 (NCH2CH2); 33.33 (CH2COOH); 38.87 (C(3)); 54.12 (NCH2); 115.46
(C(7)); 123.47 (C(4)); 128.90 (C(5)); 129.36 (C(6)); 141.01 (C(8)); 141.84 (C(9)); 174.24 (COOH);
196.48 (C(2)). APCI-MS: 544.9 (8), 272.3 (100, [M ꢀ Brꢀ]þ).
2-[5-(1-Ethyl-1,3-dihydro-3,3-dimethyl-2H-indol-2-ylidene)penta-1,3-dien-1-yl]-3,3-dimethyl-1-(5-
carboxypentyl)-3H-indolium Chloride (6 · HCl) [19]. Compound 3 · HBr (200 mg, 0.53 mmol, 1.00 equiv.)
and malonaldehyde bis(phenylimine) monohydrochloride (¼ N,N’-propane-1,3-diylidenebis[benzen-
amine] hydrochloride (1:1); 4 · HCl; 165 mg, 0.57 mmol, 1.10 quiv.) were dissolved in Ac2O (10 l) and
stirred for 30 min at 1208. After cooling to r.t., 1-ethyl-2,3,3-trimethyl-3H-indolium iodide (5; 234 mg,
0.74 mmol, 1.40 equiv.) in pyridine (10 ml) was added, and the mixture was stirred overnight at r.t. The
solvent was evaporated, and the residue dissolved in CHCl3 (8 ml), and the crude product precipitated
with hexane (40 ml). The obtained blue crystals were filtered off and then dissolved in CHCl3 (30 ml).
The CHCl3 soln. was washed with H2O (2 ꢂ 30 ml) and 0.1m HCl (30 ml), dried (MgSO4), and
concentrated and the residue purified by CC (SiO2, 0 ! 10% MeOH/CH2Cl2): 6 · HCl (258 mg, 85%).
1
Blue solid. Spectroscopic data: in accordance with literature data. H-NMR (CD3OD, 400 MHz; Ha to
He ¼ protons of pentadienyl): 1.38 (t, J(Me,CH2) ¼ 7.2, MeCH2); 1.45 – 1.53 (m, COCH2CH2CH2); 1.64 –
1.70 (m, COCH2CH2); 1.71 – 1.74 (2s, 2 MeꢀC(3), 2 MeꢀC(3’)); 1.78 – 1.88 (m, NCH2CH2); 2.31 (t,
J(CH2,CH2) ¼ 7.3, COCH2CH2); 4.08 – 4.19 (m, 2 NCH2); 6.29 (d, Ja,b ¼ 14, Ha); 6.30 (d, Je,d ¼ 14, He);
6.60 – 6.68 (m, Hc); 7.22 – 7.31 (m, HꢀC(5), HꢀC(5’), HꢀC(7), HꢀC(7’)); 7.38 – 7.43 (m, HꢀC(6),
HꢀC(6’)); 7.47 – 7.50 (m, HꢀC(4), HꢀC(4’)); 8.20 – 8.29 (m, Hb, Ha). 13C-NMR (CD3OD, 100 MHz):
12.55; 21.29; 27.36; 27.86; 27.97; 28.15; 34.72; 39.99; 44.80; 50.56; 50.61; 104.05; 104.28; 108.98; 111.81;
112.02; 122.97; 123.41; 123.46; 126.21; 126.28; 126.62; 129.75; 129.78; 142.66; 142.81; 143.08; 143.57;
155.47; 155.63; 174.40; 174.66. ESI-MS: 618.1 (10), 499.2 (6), 498.2 (36), 497.1 (100, [M ꢀ Clꢀ]þ), 483.2
(11), 482.1 (33), 468.2 (11), 467.2 (35).
Oligonucleotides 7 – 17. The modified carbostyril donor and the Ru complex were incorporated via a
pre-synthetic labeling strategy by means of labeled phosphoramidites [18]. For the incorporation of the
Cy5 derivative 6 a postsynthetic strategy was applied by using an amino-modified phosphoramidite. All
modified building blocks (67 mmol/ml in MeCN) were incorporated into the corresponding oligonucleo-
tide sequences during the automated synthesis by using standard DNA-synthesis cycles.
The coupling of the hydroxysuccinimide (NHS) ester of the Cy5 derivative 6 to the free amino
function of the oligonucleotide was carried out in a buffer soln. A stock soln. of the Cy5-NHS ester
(10 mg) in DMF (420 ml) was prepared. To the amino-modified DNA (20 OD, 60 nmol, in 200 ml of
H2O), 1m NaHCO3 (26 ml, pH 8.5) and 52.5 ml of the Cy5-NHS ester stock soln. were added and
incubated for 24 h at r.t. Then 3m AcONa (30 ml) and EtOH (825 ml) were added, and the mixture was
incubated on dry ice for 1 h. After centrifugation, the supernatant was decanted, and the residue was
washed with cold EtOH (3 ꢂ 1 ml) and purified by prep. HPLC (0 – 100% B in 48 min).
Hybridization. To the DNA strands 9 – 17 (0.1 OD in 1 ml of H2O), 1.0 equiv. of the corresponding
complementary DNA strand 7 or 8 (0.1 OD in 1 ml of H2O) was added, and the solvent was evaporated.
The precipitate was dissolved in phosphate buffer (5 mm, pH 7), and thermal hybridization was
performed by heating the sample to 908 for 2 min which was followed by slow cooling to r.t.
Photophysical Characterization. The steady-state absorption and fluorescence measurements were
performed on a Cary-500 UV/VIS spectrometer (Varian) and a Fluoromax3 fluorescence spectrometer
(Jobin Yvon), resp. The decay curves were measured on a FL920 fluorescence spectrometer (Edinburgh
Instruments) in the time-correlated single-photon-counting mode. For the excitation of the RuII entity,
the frequency-doubled output of a titaniumꢀsapphire laser (lex 430 nm) was used. The repetition rate of
80.2 MHz was reduced to 50 kHz by a Pulse Picker (Pulse Select, APE). The donor entity was excited
with a picosecond-pulsed laser diode EPL-375 (Edinburgh Instruments) at lex 375 nm and a repetition

554
Helvetica Chimica Acta – Vol. 95 (2012)
rate of 5 MHz. The excitation of the Cy entity was done with a supercontinuum laser source SC-400-PP
(Fianium) at lex 590 nm with a repetition rate of 20 MHz. The sample luminescence was measured in a
right-angle configuration to the incoming beam and detected with a multichannel plate (Europhoton). In
the data evaluation, the fluorescence spectra of RuII were corrected for the bleed through contribution of
the Cy dye.
Calculation of Fçrster radii R0 was based on the spectroscopic properties of the dyes and performed
with Eqn. 3, where FD is the quantum yield of the donor, n the refractive index of the solvent, NA the
Avogadro number, FD(l) the area normalized donor emission spectra, eA(l) the extinction coefficient of
the acceptor, k2 the orientation of the transition dipols of D and A to each other, and l the wavelength.
1
Z
ð9 ꢃ ln 10Þk2FD
R60 ¼
FDðlÞeAðlÞl4dl
(3)
128p5n4NA
0
Theoretical Calculations. The theoretical interdye distances were calulated by accessible volume
(AV) simulations based on the Model Satelite Prior algorithm implemented in the FRETnpsTools
program, as discribed in [24][25]. The pdb file of double-stranded DNA was built with the Nucleic Acid
Builder (NAB) software provided by AmberTools [26].
REFERENCES
[1] K. E. Sapsford, L. Berti, I. L. Menditz, Angew. Chem., Int. Ed. 2006, 45, 4562.
[2] P. Tinnefeld, M. Sauer, Angew. Chem., Int. Ed. 2005, 44, 2642.
[3] T. Fçrster, Ann. Phys. 1948, 2, 55.
[4] H. M. Watrob, C.-P. Pan, M. D. Barkley, J. Am. Chem. Soc. 2003, 125, 7336.
[5] Z.-J. Jiang, W. A. Goedel, Phys. Chem. Chem. Phys 2008, 10, 4584.
[6] E. Galperin, V. V. Verkhusha, A. Sorkin, Nat. Methods 2004, 1, 209.
[7] L. He, X. Wu, J. Simone, D. Hewgill, P. E. Lipsky, Nucleic Acids Res. 2005, 33, e61.
[8] S. Kawahara, T. Uchimaru, S. Murata, Chem. Commun. 1999, 563.
[9] A. K. Tong, S. Jokusch, Z. Li, H.-R. Zhu, D. L. Akins, N. J. Turro, J. Ju, J. Am. Chem. Soc. 2001, 123,
12923.
[10] A. K. Tong, Z. Li, G. S. Jones, J. J. Russo, J. Ju, Nat. Biotechnol. 2001, 19, 756.
[11] J. Liu; Y. Lu, J. Am. Chem. Soc. 2002, 124, 15208.
[12] Y. Ohya, K. Yabuki, M. Hashimoto, A. Nakajima, T. Ouchi, Bioconjugate Chem. 2003, 14, 1057.
[13] E. Haustein, M. Jahnz, P. Schwille, ChemPhysChem 2003, 4, 745.
[14] S. Hohng, C. Joo, T. Ha, Biophys. J. 2004, 87, 1328.
[15] J.-P. Clamme, A. A. Deniz, ChemPhysChem 2005, 6, 74.
[16] P. Tinnefeld, M. Heilemann, M. Sauer, ChemPhysChem 2005, 6, 217.
[17] I. H. Stein, V. Schꢁller, P. Bçhm, P. Tinnefeld, T. Liedl, ChemPhysChem 2011, 12, 689.
[18] A. Kienzler, R. Flehr, R. A. Kramer, S. Gehne, M. U. Kumke, W. Bannwarth, Bioconjugate Chem.
2011, 22, 1852.
[19] M. V. Kvach, A. V. Ustinov, I. A. Stepanova, A. D. Malakhov, M. V. Skorobogatyi, V. V. Shmanai,
V. A. Korshun, Eur. J. Org. Chem. 2008, 2107.
[20] L. Clima, C. Hirtz-Haag, A. Kienzler, W. Bannwarth, Helv. Chim. Acta 2007, 90, 1082.
[21] R. A. Kramer, R. Flehr, M. Lay, M. U. Kumke, W. Bannwarth, Helv. Chim. Acta 2009, 92, 1933.
[22] A. Kupstat, T. Ritschel, M. U. Kumke, Bioconjugate Chem. 2011, in press.
`
[23] N. Hildebrandt, L. J. Charbonniere, M. Beck, R. F. Ziessel, H.-G. Lçhmannsrçben, Angew. Chem.,
Int. Ed. 2005, 44, 7612.
[24] A. Muschielok, J. Andrecka, A. Jawhari, F. Brꢁckner, P. Cramer. and J. Michaelis, Nat. Methods
2008, 5, 965; A. Muschielok, J. Michaelis, FRETnps Tool 2.6, 2008, Ludwig-Maximilians-Universitꢅt
Munich.

Helvetica Chimica Acta – Vol. 95 (2012)
555
[25] S. Sindbert, S. Kalinin, H. Nguyen, A. Kienzler, L. Clima, W. Bannwarth, B. Appel, S. Mꢁller,
C. A. M. Seidel, J. Am. Chem. Soc. 2011, 133, 2463.
[26] D. A. Case, T. A. Darden, I. T. E. Cheatham, C. L. Simmerling, J. Wang, R. E. Duke, R. Luo, R. C.
´
Walker, W. Zhang, K. M. Merz, B. Roberts, B. Wang, S. Hayik, A. Roitberg, G. Seabra, I. Kolossvary,
K. F. Wong, F. Paesani, J. Vanicek, X. Wu, S. R. Brozell, T. Steinbrecher, H. Gohlke, Q. Cai, X. Ye,
M. J. Hsieh, G. Cui, D. R. Roe, D. H. Mathews, M. G. Seetin, C. Sagui, V. Babin, T. Luchko, S.
Gusarov, A. Kovalenko, P. A. Kollman, AmberTools Version 1.4, 2010, University of California, San
Francisco.
Received November 21, 2011