Exciton-controlled hybridization-sensitive fluorescent probes: multicolor detection of nucleic acids

Full text of DOI:10.1002/anie.200902000

Communications
DOI: 10.1002/anie.200902000
Fluorescent Probes
Exciton-Controlled Hybridization-Sensitive Fluorescent Probes:
Multicolor Detection of Nucleic Acids**
Shuji Ikeda, Takeshi Kubota, Mizue Yuki, and Akimitsu Okamoto*
The design of a fluorescent probe
in which the fluorescence is
switched off when the probe
does not recognize the target
nucleic acid is very important for
the establishment of a nucleic acid
imaging method. The require-
ments for fluorescent probes for
nucleic acid detection are not only
sequence-selective emission and
the avoidance of nonspecific
emission, but it is also important
to ensure that the emission is
Figure 1. Structure of excitonic-interaction-controlled fluorescent probes for nucleic acid detection.
Doubly fluorescent dye labeled nucleotides are shown as Dnnn (“nnn” denotes the best wavelength for
polychrome for the simultaneous
monitoring of different targets. excitation).
The following have been used to
date for the molecular design of
highly functional fluorescent probes containing an “on–off”
switching system: photophysics and photochemistry (e.g.
excimer formation,[1] photoinduced charge transfer,[2] photo-
induced electron transfer,[3] and fluorescence resonance
energy transfer).[4] In the design of these probes, the
sensitivity of the dye to the environment, higher-ordered
probe conformations, and quenching by electron transfer
from/to nucleobases all have a strong influence on fluores-
cence intensity and sensitivity and often impair predictions.
The multicoloring of probes is also limited by a fluorescence-
switching mechanism, the probe conformation, or synthetic
processes.
Dissociation of the aggregates by hybridization with the
complementary strand results in disruption of the excitonic
interaction and strong emission from the hybrid. This clear
change in fluorescence intensity is rapid and reversible.[5,7]
Multicoloring of this exciton-controlled probe could result in
the simple simultaneous detection of plural target nucleic acid
sequences and could be useful in efforts to elucidate the
temporal correlation of gene expression and interactions
between nucleic acids. For this purpose, new nucleotides
doubly labeled with other dyes should be designed without
loss of the high on–off performance as hybridization-sensitive
fluorescent probes based on excitonic interactions.
We have focused on the excitonic interaction observed for
thiazole orange fluorescent dyes and designed a doubly
fluorescence labeled nucleotide for a new efficient nucleic
acid detection method.[5–7] The fluorescence of the probe in
which two thiazole orange dye molecules are attached to a
pyrimidine base is well-controlled by an excitonic interaction
(Figure 1). An excitonic interaction is produced by the
formation of an H aggregate between the dyes. As a result,
emission from the probe before hybridization is suppressed.
We herein report a new concept for the effective design of
hybridization-sensitive fluorescent DNA probes with differ-
ent colors. Our designed probes showed strong emission upon
hybridization with the target strand, whereas emission was
suppressed in the single-stranded state. Their fluorescence is
effective for the multicolor imaging of intracellular RNA, as it
is controlled by an excitonic interaction.
We synthesized a series of new fluorescent nucleotides in
which derivatives are substituted for the thiazole orange
moiety (Figure 2). The fluorescent dyes for incorporation
were obtained by coupling a methylquinoline, dimethylani-
line, or benzothiazole subunit with a carboxylate-terminated
alkyl linker containing benzothiazole, benzoxazole, or benzo-
selenazole derivatives.[8] The carboxylate end of the alkyl
linker of the dyes was activated by succinimidylation, and two
molecules of the activated dye were subsequently incorpo-
rated into an oligodeoxyribonucleotide (ODN) containing 2’-
deoxyuridine with two amino ends.[5]
[*] Dr. S. Ikeda, Dr. T. Kubota, M. Yuki, Dr. A. Okamoto
Advanced Science Institute, RIKEN
Wako, Saitama 351-0198 (Japan)
Fax: (+81)48-467-9205
E-mail: aki-okamoto@riken.jp
[**] We thank Dr. Takehiro Suzuki (Biomolecular Characterization Team,
RIKEN) for MALDI-TOF mass spectrometry and Dr. Yayoi Hongo
(Molecular Characterization Team, RIKEN) for ESI mass spec-
trometry.
The absorption and emission of the ODN strands doubly
labeled
with
twelve
different
colors,
5’-d(TAC-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200902000.
CAGDnnnCACCAT)-3’ (where doubly fluorescent dye
6480
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 6480 –6484

Angewandte
Chemie
Figure 2. Absorption and emission of excitonic-interaction-controlled fluorescent probes before and after hybridization with the complementary
DNA. Absorption (shorter wavelength) and emission (longer wavelength) spectra are indicated by red and green lines for the nonhybridized probe
and the hybrid, respectively. The following nucleic acid sequences were used: 5’-d(TACCAGDnnnCACCAT)-3’/5’-d(ATGGTGACTGGTA)-3’. Photo-
physical data are given in Table S1 in the Supporting Information.
labeled nucleotides are shown as Dnnn), were investigated
before and after hybridization with the complementary DNA
strands (Figure 2; see Table S1 in the Supporting Informa-
tion). Fluorescence emission was observed immediately after
the addition of the labeled ODN to a solution of the
complementary nucleic acid. The emission was suppressed
in the nonhybridized state. The source of the fluorescence
behavior was confirmed by the absorption spectra. The
absorption band of the nonhybridized probe appears at a
shorter wavelength than that of the hybrid. This blue shift
suggested the splitting of the excited state because of
H aggregation of the dyes.[5,9] H aggregation allowed only
Angew. Chem. Int. Ed. 2009, 48, 6480 –6484
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6481

Communications
transitions to the upper excitonic level. The excited state was
Control of fluorescence emission by an excitonic inter-
action was also evident when the target was RNA. Several
colored nucleotides exhibited a shift of the absorption bands
and the switching of fluorescence intensity upon hybridization
with the complementary RNA strands. However, inefficient
exciton control was observed in some nucleotides. The shift of
the absorption bands was minimal for nucleotides with
hemicyanine dyes (D539, D570, and D590). The absorption
bands occurred at relatively short wavelengths for both the
hybrid and the nonhybrid form, in contrast to the absorption
behavior of nucleotides in Figure 2. The DNA-binding ability
of hemicyanines is known to be inherently lower than that of
thiazole orange dyes.[12] The small shift in the absorption band
suggests that the dissociation of the dye aggregate is
inefficient in the hybrid with RNA, and that it is difficult
for the dyes to bind independently to the hybrid structure.
Similarly, D640 and D660, which have relatively larger con-
jugated systems, showed low efficiency of dye-aggregate
dissociation upon hybrid formation with RNA. The ratio of
the fluorescence intensity of the hybrid to that of the
nonhybrid in these nucleotides was not as high as for other
nucleotides that were controlled well by an excitonic inter-
action.
rapidly transferred to the lower level, but the path from this
energy level to the ground state was not emissive. The
fluorescence from the fluorescent ODNs was suppressed by
an interdye excitonic interaction in the nonhybridized state.
Fluorescence quenching was much more effective than for
probes labeled with one dye unit.[5,10] Hybridization with the
target nucleic acid resulted in a great enhancement of
emission with dissociation of the aggregate. This exciton-
controlled fluorescence behavior was also observed for
doubly labeled ODNs with other sequences (see Table S2 in
the Supporting Information).
However, among the twelve colored nucleotides consid-
ered, only the D410-containing ODN showed relatively strong
emission, even when the ODN was in the nonhybridized state.
In this case, the blue-shifted absorption spectrum suggested
the existence of an interdye excitonic interaction. The
excitation spectra revealed two peaks, and the absorption
bands of the aggregate and the non-aggregate overlapped.
These spectra are quite different from the excitation spectra
of the D514 probe, for which only one signal corresponding to
the non-aggregate was observed (Figure 3). Furthermore, the
excitation and emission spectra showed that excitation at a
shorter wavelength (aggregate) corresponded to emission at a
longer wavelength, while excitation at a longer wavelength
(non-aggregate) corresponded to emission at a shorter wave-
length. This result suggested that the dyes of the D410
nucleotide were inclined with respect to one another, and
that the emission from the lower excitonic state after
excitation to the upper excitonic state was not completely
forbidden.[9,11]
The appropriate design of dye aggregation in the non-
hybridized state is the key point that determines the ability of
doubly labeled ODNs to function as hybridization-sensitive
fluorescent probes. We developed a series of hybridization-
sensitive fluorescent probes that cover the excitation-wave-
length range 400–700 nm. From this series, we can select the
probe that is most suitable for the nature and number of
nucleic acids, and the filter type of the detector.
We believed that these doubly labeled ODNs
might be useful for multicolor RNA imaging. The
development of chemical methods for imaging the
dynamic and static behavior of RNA in a living cell is
essential for increasing our understanding of cell life:
a key goal of life scientists.[13] Although many
methods for RNA detection have been developed,
such as molecular beacons,[14] MS2-GFP fusion
proteins,[15] GFP reconstitution,[16] quenched autoli-
gation probes,[17] and dye-binding aptamers,[18] there
are various associated problems, including the low
availability of many color probes, limitations in
sequence design, slow response in terms of reactivity
or conformation change, high background fluores-
cence, or low fluorescence reversibility. Our new
hybridization-sensitive probes may offer solutions to
these issues and provide advantages.
We next designed a model experiment with a
series of probes for RNA imaging in a living cell. The
probes capable of binding to the polyA tail of
mRNA, 5’-d(T6DnnnT6)-3’, were transfected to HeLa
cells with the common transfection reagent Lip-
ofectamine 2000. After incubation for 1 h and wash-
ing, fluorescence emission from the cells was
observed at the wavelength expected for each Dnnn
(Figure 4a). The fluorescence from the cells
decreased upon competitive hybridization of a dT
70-mer (dT70) DNA molecule. This fluorescence
Figure 3. Excitation and emission spectra of the single-stranded D410 and D514
probes 5’-d(TACCAGDnnnCACCAT)-3’, and a model of molecule orientation and
transition dipole interactions in H aggregates. D410: excitation spectra for emis-
sion at 426 (red) and 480 nm (black), and emission spectra at 410 (red) and
360 nm (black); D514: excitation spectra for emission at 537 (red) and 600 nm
(black), and emission spectra at 516 (red) and 480 nm (black). M is the transition
dipole moment; m+ and mÀ are the excitonic states.
6482 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 6480 –6484

Angewandte
Chemie
appeared immediately in the nucleus of each cell, and the
emission wavelength corresponded to the type of intracellular
miRNA. The target-specific emission indicated that the probe
recognized the target miRNA in the cell and emitted
fluorescence with a characteristic color. The fluorescence in
the cells was identified as emission from the dyes of the
probes on the basis of emission spectra obtained with a
multichannel spectrum detector. Three colors were observed
simultaneously for the cell containing the three miRNA
strands.
In conclusion, we have designed a series of fluorescent
probes based on the concept of quenching caused by an
excitonic interaction. The hybridization-sensitive, quencher-
free fluorescent probes with various colors facilitate the
imaging of intracellular RNA. Although further aspects need
to be examined, such as the detection of a very small amount
of RNA, we anticipate that this new concept of hybridization-
sensitive probes based on photochemical techniques will be
the starting point for the development of a practical assay for
the simultaneous imaging of plural RNA molecules in living
cells.
Received: April 15, 2009
Published online: July 27, 2009
Keywords: bioorganic chemistry · excitonic interaction ·
.
fluorescent probes · nucleic acids · RNA
[1] a) I. Aoki, H. Kawabata, K. Nakashima, S. Shinkai, J. Chem. Soc.
Chem. Commun. 1991, 1771 – 1773; b) S. Nishizawa, Y. Kato, N.
Teramae, J. Am. Chem. Soc. 1999, 121, 9463 – 9464; c) A.
Okamoto, T. Ichiba, I. Saito, J. Am. Chem. Soc. 2004, 126,
8364 – 8365.
[2] a) G. Weber, F. J. Farris, Biochemistry 1979, 18, 3075 – 3078; b) G.
Grynkiewicz, M. Poenie, R. Y. Tsien, J. Biol. Chem. 1985, 260,
3440 – 3450; c) D. W. Pierce, S. G. Boxer, J. Phys. Chem. 1992, 96,
5560 – 5566; d) B. E. Cohen, T. B. McAnaney, E. S. Park, Y. N.
Jan, S. G. Boxer, L. Y. Jan, Science 2002, 296, 1700 – 1703; e) A.
Okamoto, K. Tainaka, K.-i. Nishiza, I. Saito, J. Am. Chem. Soc.
2005, 127, 13128 – 13129; f) K. Tainaka, K. Tanaka, S. Ikeda, K.-i.
Nishiza, T. Unzai, Y. Fujiwara, I. Saito, A. Okamoto, J. Am.
Chem. Soc. 2007, 129, 4776 – 4784.
Figure 4. Multicolored living cells. a) Transfection of color probes
5’-d(T6DnnnT6)-3’ to HeLa cells. Top: Fluorescence images of probe-
transfected cells. Bottom: Fluorescence images of the cells in which
probe binding to the mRNA polyA tail is blocked by dT70 DNA. Scale
bar: 10 mm. b) Fluorescence images of miRNA-rich HeLa cells, into
which the probe mixture was injected. DIC: differential interference
contrast images; Ch G: lex =514 nm (Ar), lem =520–555 nm; Ch O:
lex =543 nm (He–Ne), lem =560–615 nm; Ch R: lex =633 nm (He–
Ne), lem =656–731 nm. I) Control, II) miR-20-rich cell, III) miR-17-rich
cell, IV) miR-30-rich cell, V) miR-17-, miR-20-, and miR-30-rich cell.
Scale bar: 10 mm.
[3] a) R. A. Bissell, A. P. de Silva, H. Q. N. Gunaratne, P. L. M.
Lynch, G. E. M. Maguire, C. P. McCoy, K. R. A. S. Sandanayake,
Top. Curr. Chem. 1993, 168, 223 – 264; b) T. D. James, K. R. A. S.
Sandanayake, R. Iguchi, S. Shinkai, J. Am. Chem. Soc. 1995, 117,
8982 – 8987; c) R. Bergonzi, L. Fabbrizzi, M. Lichelli, C.
Mangano, Coord. Chem. Rev. 1998, 170, 31 – 46; d) Y. Urano,
M. Kamiya, K. Kanda, T. Ueno, K. Hirose, T. Nagano, J. Am.
Chem. Soc. 2005, 127, 4888 – 4894.
[4] a) P. Wu, K. G. Rice, L. Brand, Y. C. Lee, Proc. Natl. Acad. Sci.
USA 1991, 88, 9355 – 9359; b) P. Wu, L. Brand, Anal. Biochem.
1994, 218, 1 – 13; c) D. Klostermeier, D. P. Millar, Biopolymers
2002, 61, 159 – 179.
[5] S. Ikeda, A. Okamoto, Chem. Asian J. 2008, 3, 958 – 968.
[6] a) S. Ikeda, T. Kubota, K. Kino, A. Okamoto, Bioconjugate
Chem. 2008, 19, 1719 – 1725; b) T. Kubota, S. Ikeda, H.
Yanagisawa, M. Yuki, A. Okamoto, Bioconjugate Chem. 2009,
20, 1256 – 1261.
decrease suggests that the probe to be hybridized with the
polyA tail of mRNA was replaced with a dT70 DNA unit to
form a more stable duplex with polyA RNA (Tm = 618C).[7]
The transfection efficiency of the probes labeled with hemi-
cyanine dyes (D539, D570, and D590) was very low (< 5%), and
fluorescence from the cells was not observed.
The model experiment of multicolor live-cell RNA
imaging was also carried out with HeLa cells. Cells containing
one or all of three different small noncoding microRNA
strands (miRNA) were prepared by microinjection into nuclei
(Figure 4b).[19] We synthesized three probes, each of which
corresponded to a different miRNA strand. The probes each
had one of three different fluorescence-labeled nucleotides
(D514, D543, or D640), which were selected from the series of
Dnnn because their fluorescence wavelengths were widely
separated. The three probes were mixed and microinjected
into the nuclei of living HeLa cells. The fluorescence
[7] T. Kubota, S. Ikeda, A. Okamoto, Bull. Chem. Soc. Jpn. 2009, 82,
110 – 117.
[8] See the Supporting Information (Schemes S1–S7).
Angew. Chem. Int. Ed. 2009, 48, 6480 –6484
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6483

Communications
[9] M. Kasha, Radiat. Res. 1963, 20, 55 – 70.
2003, 31, 1319 – 1330; d) J. F. Hopkins, S. A. Woodson, Nucleic
Acids Res. 2005, 33, 5763 – 5770.
[10] a) T. Ishiguro, J. Saitoh, H. Yawata, M. Otsuka, T. Inoue, Y.
Sugiura, Nucleic Acids Res. 1996, 24, 4992 – 4997; b) N. Svanvik,
J. Nygren, G. Westman, M. Kubista, J. Am. Chem. Soc. 2001, 123,
803 – 809; c) O. Kꢀhler, D. V. Jarikote, O. Seitz, Chem. Commun.
2004, 2674 – 2675; d) V. L. Marin, B. A. Armitage, Biochemistry
2006, 45, 1745 – 1754.
[11] a) A. S. Davydov, Theory of Molecular Excitons, McGraw-Hill,
New York, 1962; b) C. Peyratout, L. Daehne, Phys. Chem.
Chem. Phys. 2002, 4, 3032 – 3039.
[12] a) D. L. Boger, W. C. Tse, Bioorg. Med. Chem. 2001, 9, 2511 –
2518; b) N. Akbay, M. Y. Losytskyy, V. B. Kovalska, A. O.
Balanda, S. M. Yarmoluk, J. Fluoresc. 2008, 18, 139 – 147.
[13] a) S. Paillasson, M. van de Corput, R. W. Dirks, H. J. Tanke, M.
Robert-Nicoud, X. Ronot, Exp. Cell Res. 1997, 231, 226 – 233;
b) J. Pernthaler, F. O. Glockner, W. Schonhuber, R. Amann,
Methods Microbiol. 2001, 30, 207 – 226; c) A. J. Rodriguez, J.
Condeelis, R. H. Singer, J. B. Dictenberg, Semin. Cell Dev. Biol.
2007, 18, 202 – 208.
[15] a) E. Bertrand, P. Chartrand, M. Schaefer, S. M. Shenoy, R. H.
Singer, R. M. Long, Mol. Cell 1998, 2, 437 – 445; b) D. Fusco, N.
Accornero, B. Lavoie, S. M. Shenoy, J.-M. Blanchard, R. H.
Singer, E. Bertrand, Curr. Biol. 2003, 13, 161 – 167; c) T. T. Weil,
K. M. Forrest, E. R. Gavis, Dev. Cell 2006, 11, 251 – 262.
[16] T. Ozawa, Y. Natori, M. Sato, Y. Umezawa, Nat. Methods 2007,
4, 413 – 419.
[17] a) S. Sando, E. T. Kool, J. Am. Chem. Soc. 2002, 124, 2096 – 2097;
b) S. Sando, H. Abe, E. T. Kool, J. Am. Chem. Soc. 2004, 126,
1081 – 1087.
[18] a) J. R. Babendure, S. R. Adams, R. Y. Tsien, J. Am. Chem. Soc.
2003, 125, 14716 – 14717; b) S. Sando, A. Narita, M. Hayami, Y.
Aoyama, Chem. Commun. 2008, 3858 – 3860; c) T. P. Constantin,
G. L. Silva, K. L. Robertson, T. P. Hamilton, K. Fague, A. S.
Waggoner, B. A. Armitage, Org. Lett. 2008, 10, 1561 – 1564.
[19] Approximately 100 fL of a solution of miRNA strands (6–
48 mm), estimated to be 102–103 times the inherent amount of
intracellular miRNA strands, was microinjected. Fluorescence
from the original miRNA is therefore negligible in this model
experiment.
[14] a) S. Tyagi, F. R. Kramer, Nat. Biotechnol. 1996, 14, 303 – 308;
b) S. Bernacchi, Y. Mꢁly, Nucleic Acids Res. 2001, 29, e62; c) A.
Tsourkas, M. A. Behlke, S. D. Rose, G. Bao, Nucleic Acids Res.
6484 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 6480 –6484