Coherent quenching of a fluorophore for the design of a highly sensitive in-stem molecular beacon

Article abstract of DOI:10.1002/anie.201001459

(Figure Presented) Excitonic interaction was utilized to design a highly sensitive in-stem molecular beacon (ISMB) in which both a fluorophore and a quencher on D-threoninols are incorporated as a pseudo base pair (see scheme; optimized combination with Cy3 and modified Methyl Red). Minimization of the difference between λ_max of the fluorophore and quencher maximized quenching efficiency.

Full text of DOI:10.1002/anie.201001459

Communications
DOI: 10.1002/anie.201001459
Coherent Quenching
Coherent Quenching of a Fluorophore for the Design of a Highly
Sensitive In-Stem Molecular Beacon**
Yuichi Hara, Taiga Fujii, Hiromu Kashida, Koji Sekiguchi, Xingguo Liang, Kosuke Niwa,
Tomokazu Takase, Yasuko Yoshida, and Hiroyuki Asanuma*
Fluorescence-labeled oligonucleotides provide powerful tools
for highly sensitive, sequence-specific detection of target
DNA/RNA.[1] A variety of fluorescent probes have been
designed for purposes such as genotyping of individuals,
identification of pathogens, and real-time monitoring of
mRNA within cells.[2] One practical problem with the use of
these fluorescent probes is background emission intensity in
the absence of the target, because background emission, as
well as scattered light of excitation due to a small Stokes shift,
critically affects probe sensitivity.[3] Suppression of back-
ground emission is particularly important for the design of a
highly sensitive molecular beacon (MB); that is, a hairpin
oligonucleotide dual-labeled at both 5’- and 3’-termini with a
fluorophore and a quencher.[4] Therefore, efforts have been
made to efficiently quench a fluorophore in a closed MB.[5]
There are two types of energy transfer that will efficiently
quench fluorescence: Fꢀrster (dynamic) quenching and con-
tact (static) quenching.[6] The former requires a fluorophore–
quencher pair that has sufficient spectral overlap between the
emission spectrum of the fluorophore and the absorption
spectrum of the quencher, and allows relatively long range
(10–100 ꢁ) quenching.[7] The latter is based on the formation
of a ground state complex of a fluorophore and a dark
quencher by their direct contact,[8] which is favorable for an
MB. In this case, a change in their absorption spectra often
occurs due to excitonic interaction (coherency). However,
Marras also pointed out that any non-fluorescent quencher
can be an efficient acceptor for contact quenching.[3] We
previously established a novel method for the preparation of a
heterodimer or cluster by hybridization of two oligonucleo-
tides, each of which had dyes tethered to d-threoninols
(termed threoninol nucleotides).[9] The use of d-threoninol as
a scaffold facilitated dimerization or clustering of the dyes in
the duplex. We have utilized this base surrogate (threoninol
nucleotide) to develop a new in-stem molecular beacon
(ISMB) in which both fluorophore (perylene) and quencher
(anthraquinone) on d-threoninols are incorporated into the
stem region as a pseudo base pair (Scheme 1a).[10] Direct and
[*] Y. Hara, T. Fujii, Prof. Dr. H. Kashida, K. Sekiguchi,
Prof. Dr. X. G. Liang, Prof. Dr. H. Asanuma
Graduate School of Engineering, Nagoya University
Furo-cho, Chikusa-ku, Nagoya 464-8603 (Japan)
Fax: (+81)52-789-2488
Scheme 1. a) An in-stem molecular beacon (ISMB), and b) the chem-
ical structure of the dyes and sequences of the modified DNAs. In the
MB sequences, the bases in the stem part are underlined.
E-mail: asanuma@mol.nagoya-u.ac.jp
Prof. Dr. H. Asanuma
CREST (Japan) Science and Technology Agency
4-1-8 Honcho, Kawaguchi, Saitama 332-0012 (Japan)
strong contact of perylene with the quencher in the stem
efficiently removed background emission in the closed ISMB
and allowed discrimination of fully matched and one-base
deletion mutant pairs. In this case, the absorption spectrum of
perylene showed only a slight change with any quencher that
we tested, showing that coherency (excitonic interaction)
between the perylene and the quenchers was poor. Never-
theless, quenching was very strong, indicating that strong
contact by pseudo base-pairing dominated the efficient
quenching. However, as Johansson et al. proposed previ-
ously,[6,11] stronger coherency (that is, excitonic interaction or
ground-state complexation) should enhance the quenching
and further increase the sensitivity of the probe.
K. Niwa, T. Takase, Prof. Dr. Y. Yoshida
NGK INSULATORS, Nagoya (Japan)
Prof. Dr. Y. Yoshida
MEXT Innovative Research Center for PME, Nagoya (Japan)
[**] This work was supported by the Core Research for Evolution Science
and Technology (CREST) (Japan) Science and Technology Agency
(JST). Partial support by a Grant-in-Aid for Scientific Research from
the Ministry of Education, Culture, Sports, Science and Technology
(Japan) is also acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201001459.
5502
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5502 –5506

Angewandte
Chemie
In the present study, which works towards the design of a
highly sensitive ISMB, the effect of coherency (excitonic
interaction) on the quenching of a fluorophore was system-
atically examined using azobenzene derivatives as quenchers.
We have previously experimentally verified the molecular
exciton theory using threoninol nucleotides involving azo
dyes that have different absorption maxima, and found that
heterodimerization induced hyperchromism of the band of
shorter wavelength (in-phase transition),[9b] but hypochrom-
ism of the band of longer wavelength (out-of-phase transition;
Scheme 2). Herein, we used Thiazole Orange (TO) and
transition moment is partially cancelled due to its antiparallel
orientation. The above excitonic interaction (namely coher-
ency) is enhanced when the gap between the transition
energies of quencher and fluorophore, this is, Dlmax
,
decreases. To evaluate coherency, the absorption spectrum
of Flu(X)/Qnt(Y) duplexes was compared to that of the
individual single strands. The degree of coherency can be
estimated from the hypochromicity (decrease in absorbance)
of the absorption band of TO (516 nm at 208C) or of Cy3
(550 nm), which is at longer wavelengths than that of the
quenchers.[9b,13] Figure 1A shows the UV/Vis spectra of
Flu(TO)/Qnt(Y) duplexes, Flu(TO) and Qnt(Y) single
strands, and a spectrum calculated as the simple sum of the
two single strands. When Azo was used as a quencher, the
spectral change between the summation spectrum and the
Flu(TO)/Qnt(Azo) duplex was small (orange and black lines
in Figure 1A(a)); that is, hypochromicity induced by hybrid-
ization was small. However, hypochromicity increased as the
Dlmax between the TO and the quencher decreased.[9b] In the
case of the TO and NR combination, the UV/Vis spectrum of
the Flu(TO)/Qnt(NR) duplex was entirely different from that
of the summation spectrum (orange and black lines in
Figure 1A(d)):[14] the band of Flu(TO) at 516 nm and that
of Qnt(NR) at 513 nm almost disappeared in Flu(TO)/
Qnt(NR), and a new strong band and weak shoulder band
appeared at 492 and 590 nm, respectively.[15] This large change
in the UV/Vis spectrum, namely the hypochromicity of the
TO band, was attributed to the strong coherency (excitonic
interaction or ground-state complexation) of TO and NR,
because the Dlmax between the two dyes was as small as 3 nm.
Similar results were obtained using Cy3: strong hypochrom-
ism occurred with MR and NR (Figure 1B(c) and 1B(d),
respectively), although this hypochromism was smaller than
that of TO due to the larger Dlmax (37 nm for NR and 70 nm
for MR). No spectral change was induced when Flu(Cy3) was
combined with Qnt(Azo), because the Dlmax was as large as
216 nm (note that the Flu(Cy3)/Qnt(Azo) spectrum almost
completely overlapped with that of the summation spectrum;
Figure 1B(a)). This spectroscopic behavior agreed with our
previous results and above expectation based on the model
shown in Scheme 2.
Scheme 2. Energy diagram of the heterodimer of fluorophore and
quencher depicted on the basis of the exciton model. Each arrow,
outlined by an ellipse or an oblong, designates the transition dipole
moment.
pentamethylindocarbocyanine (Cy3) as fluorophores (Sche-
me 1b; see also the Supporting Information, Scheme S1,S2
for their syntheses), because these are of known practical use
in the labeling of biomolecules,[5b,12] and are expected to have
stronger coherency with azo dyes than perylene. By system-
atically varying the absorption maxima of azo dyes with a
similar structure as shown in the Supporting Information,
Figure S1 (see Scheme 1b for the structures), the quenching
efficiency was maximized, resulting in a more sensitive ISMB.
The azobenzene derivatives (Y= Azo, S-Azo, MR, NR)
used for Qnt showed a lmax at 334, 394, 480, and 513 nm,
respectively, whereas TO and Cy3 in Flu showed a lmax at 516
and 550 nm in single-stranded DNAs at 208C (Supporting
Information, Figure S1). What we expect for the UV/Vis
spectra of Flu/Qnt duplexes based on the molecular exciton
theory (Scheme 2) is as follows:[9b] when the quencher that
has a higher transition energy is excitonically coupled with a
fluorophore of lower transition energy, the excited state splits
into two energy levels. The higher energy level corresponds to
the in-phase transition in which the transition dipole moments
of the quencher and fluorophore have the same direction. In
this case, the energy level becomes higher than that of
quencher due to the repulsive interaction between the two
transition dipoles. The absorption coefficient of the in-phase
transition is expected to increase due to the sum of the
transitions quencher and fluorophore. In contrast, the energy
level of the out-of-phase transition becomes lower than that
of fluorophore due to the attractive interaction. However, the
absorption coefficient of this transition decreases because the
Next, the quenching efficiency of this model Flu/Qnt
duplex was evaluated. As shown in Figures 2a,b, all the
quenchers that were tested dramatically quenched emission
from TO or Cy3 in the Flu/Qnt duplex, demonstrating that
close stacking of a fluorophore and a quencher on d-
threoninols facilitates quenching. However, the quenching
efficiency obviously depended on the type of quencher that
was used. The efficiency of TO quenching was in the order:
NR ꢀ MR > S-Azo @ Azo, which correlated fairly well with
the degree of hypochromicity (coherency) as depicted in the
Supporting Information, Figure S2a.[13,16] Cy3 was quenched
by azo dyes in a similar order: NR ꢀ MR @ S-Azo ꢀ Azo (See
the Supporting Information, Figure S2b, for the correlation of
coherency and quenching efficiency).[13]
To examine the effect of duplex stability on the quenching
efficiency, the melting temperature of the Flu/Qnt duplex was
measured (Supporting Information, Table S2). In the case of
Flu(TO)/Qnt(Y), the Tm decreased with Y in the order NR >
Angew. Chem. Int. Ed. 2010, 49, 5502 –5506
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5503

Communications
Figure 2. Fluorescence emission spectra of a) Flu(TO) excited at
516 nm and b) Flu(Cy3) excited at 546 nm, and their duplexes with
Qnt(Y). Solution conditions: [Flu(X)]=0.2 mm, [Qnt(Y)]=0.4 mm,
[NaCl]=100 mm, pH 7.0 (10 mm phosphate buffer); 208C.
fluorophore and a quencher because their close contact in the
duplex facilitates ground state complexation (that is, coher-
ency or excitonic interaction) to allow rapid electron (or hole)
transfer to the quencher. It can now be concluded that to
maximize the quenching efficiency Dlmax, the difference in the
absorption maxima between the fluorophore and the
quencher should be minimized to enhance the coherency.[17]
Based on the above experiments, we chose NR or MR as a
quencher of Cy3 to design an ISMB (Scheme 1) targeting the
survivin gene.[18] We designed the ISMB as sharing NR or MR
strand of the stem region with the target Surv, a synthetic
18 nucleotide-long DNA (see Scheme 1b for the sequence),
because sharing of Cy3 or TO strand (that is, intercalation of
Cy3 or TO) reduced its fluorescence (data not shown). As the
Tm of MBNR was determined to be 63.18C,[19] MBNR was
available below 608C, where the MB was closed in the
absence of the target. Figure 3a depicts the fluorescence
emission spectra of MBNR involving a Cy3 and NR pair at the
stem region in the presence and absence of the target at 208C.
The ratio of the fluorescence intensities of MBNR with and
without the target (Iopen/Iclose) at 564 nm was as high as 70
under the conditions employed.[20,21] This high sensitivity was
attributed to effective quenching of the Cy3 emission. We also
Figure 1. UV/Vis spectra of Flu(X)/Qnt(Y) with A) X=TO, B) X=Cy3.
Sum: a simple sum of the spectra. Solution conditions:
[Flu(X)]=4 mm, [Qnt(Y)]=4 mm, [NaCl]=100 mm, pH 7.0 (10 mm
phosphate buffer); 208C.
MR > S-Azo > Azo, which coincided with the order of
quenching efficiency. However, combination of these quench-
ers with Flu(Cy3) completely reversed the order of the Tm
decrease with Y to Azo > S-Azo > MR > NR. Therefore, a
contribution of the stability of the Flu/Qnt duplex to the
quenching efficiency appears to be unlikely. These results
strongly suggest that coherency between the fluorophore and
the quencher and also direct contact of the fluorophore–
quencher pair significantly contribute to effective quenching.
Thus d-threoninol may be an ideal scaffold for the pairing of a
5504
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5502 –5506

Angewandte
Chemie
[1] a) E. T. Kool, Acc. Chem. Res. 2002, 35, 936; b) A. A. Martꢂ, S.
Jockusch, N. Stevens, J. Ju, N. J. Turro, Acc. Chem. Res. 2007, 40,
402.
[2] W. Tan, K. Wang, T. J. Drake, Curr. Opin. Chem. Biol. 2004, 8,
547.
[3] S. A. E. Marras, Mol. Biotechnol. 2008, 38, 247 – 255.
[4] S. Tyagi, F. R. Kramer, Nat. Biotechnol. 1996, 14, 303.
[5] a) O. Seitz, Angew. Chem. 2000, 112, 3389; Angew. Chem. Int.
Ed. 2000, 39, 3249; b) S. Tyagi, S. A. E. Marras, F. R. Kramer,
Nat. Biotechnol. 2000, 18, 1191; c) B. Dubertret, M. Calame, A. J.
Libchaber, Nat. Biotechnol. 2001, 19, 365; d) H. Du, M. D.
Disney, B. L. Miller, T. D. Krauss, J. Am. Chem. Soc. 2003, 125,
4012; e) C. J. Yang, H. Lin, W. Tan, J. Am. Chem. Soc. 2005, 127,
12772; f) C. J. Yang, M. Pinto, K. Schanze, W. Tan, Angew. Chem.
2005, 117, 2628; Angew. Chem. Int. Ed. 2005, 44, 2572; g) L.
Wang, C. J. Yang, C. D. Medley, S. A. Benner, W. Tan, J. Am.
Chem. Soc. 2005, 127, 15664; h) C. Crey-Desbiolles, D.-R. Ahn,
C. J. Leumann, Nucleic Acids Res. 2005, 33, e77; i) D. M.
Kolpashchikov, J. Am. Chem. Soc. 2006, 128, 10625; j) Y. Kim,
C. J. Yang, W. Tan, Nucleic Acids Res. 2007, 35, 7279; k) T. N.
Grossmann, L. Rꢀglin, O. Seitz, Angew. Chem. 2007, 119, 5315;
Angew. Chem. Int. Ed. 2007, 46, 5223; l) E. Socher, L. Bethge, A.
Knoll, N. Jungnick, A. Herrmann, O. Seitz, Angew. Chem. 2008,
120, 9697; Angew. Chem. Int. Ed. 2008, 47, 9555; m) S.
Nakayama, L. Yan, H. O. Sintim, J. Am. Chem. Soc. 2008, 130,
12560; n) E. Socher, D. V. Jarikote, A. Knoll, L. Rꢀglin, J.
Burmeister, O. Seitz, Anal. Biochem. 2008, 375, 318; o) S.
Berndl, H.-A. Wagenknecht, Angew. Chem. 2009, 121, 2454;
Angew. Chem. Int. Ed. 2009, 48, 2418.
[6] M. K. Johansson, R. M. Cook, Chem. Eur. J. 2003, 9, 2466.
[7] T. Fꢀrster, Ann. Phys. 1948, 2, 55.
[8] J. Lakowicz, Principles of Fluorescence Spectroscopy, 2nd, ed.,
Plenum, New York, 1999.
[9] a) H. Kashida, T. Fujii, H. Asanuma, Org. Biomol. Chem. 2008,
6, 2892; b) T. Fujii, H. Kashida, H. Asanuma, Chem. Eur. J. 2009,
15, 10092.
[10] H. Kashida, T. Takatsu, T. Fujii, K. Sekiguchi, X. Liang, T.
Takase, Y. Yoshida, H. Asanuma, Angew. Chem. 2009, 121, 7178;
Angew. Chem. Int. Ed. 2009, 48, 7044.
[11] M. K. Johansson, H. Fidder, D. Dick, R. M. Cook, J. Am. Chem.
Soc. 2002, 124, 6950.
Figure 3. Fluorescence emission spectra of a) MBNR and b) MBMR
excited at 546 nm with (g) and without (c) the presence of Surv.
Solution conditions: [MBMR]=[MBNR]=0.2 mm, [Surv]=0.4 mm,
[NaCl]=100 mm, pH 7.0 (10 mm phosphate buffer); 208C.
[12] a) W. R. Algar, M. Massey, U. J. Krull, J. Fluoresc. 2006, 16, 555;
b) R. B. Mujumdar, L. A. Ernst, S. R. Mujumdar, C. J. Lewis,
A. S. Waggoner, Bioconjugate Chem. 1993, 4, 105; c) G. A.
Korbel, G. Lalic, M. D. Shair, J. Am. Chem. Soc. 2001, 123, 361;
d) O. Seitz, F. Bergmann, D. Heindl, Angew. Chem. 1999, 111,
2340; Angew. Chem. Int. Ed. 1999, 38, 2203; e) F. Menacher, M.
Rubner, S. Berndl, H.-A. Wagenknecht, J. Org. Chem. 2008, 73,
4263.
tested an MBMR bearing MR (Dabcyl) that is used as a
conventional quencher, and found that Iopen/Iclose was as high
as 30 (Figure 3b). However, this value was still less than half
[22]
that obtained with MBNR
.
Undoubtedly, strong coherency
(excitonic interaction) facilitated the quenching that strongly
enhanced the sensitivity of the ISMB.[23]
[13] Quantitatively evaluated degree of hypochromicity (coherency)
and quenching efficiency for the Flu(X)/Qnt(Y) duplex are
summarized in Table S1 of the Supporting Information.
[14] Pseudo pair of TO (or Cy3) and Y is stacked in the duplex; see
Ref. [9b].
[15] This large spectral change is explained from the molecular
exciton theory. In the case of the Flu(TO)/Qnt(NR) or Flu(TO)/
Qnt(MR) combinations, strong excitonic interaction due to the
small Dlmax induced large blue shift of TO (in-phase) band and
red shift of NR (out-of-phase) band, which resulted in the
appearance of new bands. However, as Dlmax of the Flu(TO)/
Qnt(S-Azo) or Flu(TO)/Qnt(Azo) combination was much
larger, such large shift did not occur. See: T. Kobayashi, J-
aggregates, World Scientific, Singapore, 1996, pp. 1 – 40.
[16] Excitation spectra of the Flu(TO)/Q(NR) duplex was very
similar in shape to the excitation and absorption spectra of
Flu(TO) (Supporting Information, Figure S3), demonstrating
In conclusion, excitonic interaction (coherency) was
utilized to design a highly sensitive ISMB. Based on the
results of a systematic study, we conclude that minimizing the
Dlmax maximizes quenching efficiency due to maximization of
coherency. We are currently spotting this ISMB on a solid
surface and testing its potential as a capture probe on DNA
chip.
Received: March 11, 2010
Revised: April 22, 2010
Published online: July 2, 2010
Keywords: Cy3 · DNA · excitonic interactions ·
.
fluorescent probes · molecular beacons
Angew. Chem. Int. Ed. 2010, 49, 5502 –5506
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5505

Communications
that the heterodimer did not emit fluorescence at all; see
Ref. [11].
[20] The response profile of MBNR under these conditions is depicted
in Figure S4 of the Supporting Information.
[17] In other words, spectral overlap between the quencher and
fluorophore should be maximized.
[21] MBNR showed sufficient sequence-specificity; see Figure S5 of
the Supporting Information.
[18] X.-H. Peng, Z.-H. Cao, J.-T. Xia, G. W. Carlson, M. M. Lewis,
W. C. Wood, L. Yang, Cancer Res. 2005, 65, 1909.
[19] Melting temperatures of MBMR and MBNR in the presence and
absence of the target Surv are listed in Table S3 of the
Supporting Information.
[22] MBMR and MBNR were available below 608C, at which temper-
ature the molecular beacon was closed in the absence of a target.
See Figure S6 of the Supporting Information.
[23] Note that Iopen/Iclose of previous ISMB, composed of a single pair
of perylene and anthraquinone, was 26 under similar conditions.
5506
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5502 –5506