Stable and selective formation of Hoogsteen-type triplexes and duplexes using twisted intercalating nucleic acids (TINA) prepared via postsynthetic sonogashira solid-phase coupling reactions

Article abstract of DOI:10.1021/ja053645d

Bulge insertions of (R)-1-O-[4-(1-pyrenylethynyl)phenylmethyl]glycerol (5) into the middle of homopyrimidine oligodeoxynucleotides (twisted intercalating nucleic acids, TINA) obtained via postsynthetic Sonogashira coupling reaction led to extraordinary hi

Full text of DOI:10.1021/ja053645d

Published on Web 10/01/2005
Stable and Selective Formation of Hoogsteen-Type Triplexes
and Duplexes Using Twisted Intercalating Nucleic Acids
(TINA) Prepared via Postsynthetic Sonogashira Solid-Phase
Coupling Reactions
Vyacheslav V. Filichev and Erik B. Pedersen*
Contribution from the Nucleic Acid Center, Department of Chemistry, UniVersity of Southern
Denmark, CampusVej 55, DK-5230 Odense M, Denmark
Received June 3, 2005; E-mail: EBP@chem.sdu.dk
Abstract: Bulge insertions of (R)-1-O-[4-(1-pyrenylethynyl)phenylmethyl]glycerol (5) into the middle of
homopyrimidine oligodeoxynucleotides (twisted intercalating nucleic acids, TINA) obtained via postsynthetic
Sonogashira coupling reaction led to extraordinary high thermal stability of Hoogsteen-type triplexes and
duplexes, whereas Watson-Crick-type duplexes of the same nucleotide content were destabilized. Modified
oligonucleotides were synthesized using the phosphoramidite of (S)-1-(4,4′-dimethoxytriphenylmethyloxy)-
3-(4-iodo-benzyloxy)-propan-2-ol followed by treatment of the oligonucleotide on a CPG-support with the
Sonogashira-coupling reaction mixture containing different ethynylaryls. Bulged insertion of the pyrene
derivative 5 into oligonucleotides was found to be the best among the tested modifications for binding to
the Hoogsteen-type triplexes and duplexes. Thus, at pH 7.2 an oligonucleotide with cytidine content of
36% possessing two bulged insertions of 5 separated by three bases formed a stable triplex (Tm ) 43.0
°C), whereas the native oligonucleotide was unable to bind to the target duplex. The corresponding Watson-
Crick-type duplex with the same oligonucleotide had Tm of 38.0 °C at pH 7.2, while the Tm of unmodified
dsDNA was 47.0 °C. Experiments with mismatched oligonucleotides, luminescent properties, and potential
applications of TINA technology is discussed.
nucleic acids such as peptide nucleic acids (PNAs),2 locked
nucleic acids (LNAs),3 2′-aminoethyl-oligoribonucleotides (2′-
Introduction
The sequence-specific recognition of double-stranded DNA
(dsDNA) is a topic of considerable interest in the development
of oligonucleotide-based tools in molecular biology, therapeu-
tics, and bionanotechnology. Triple helixes are formed when a
single-stranded triplex-forming oligonucleotide (TFO) binds to
dsDNA through specific major groove interactions, and this has
been the subject of intense research for gene targeting. This
approach allows transcriptional control, gene knock-out, and
sequence-selective treatment of genomic DNA aiming for
mutated or recombined genes.1
The third strand affinity of TFOs to their targets is generally
problematic due to their required recognition to homopurine
sequences of dsDNA and the disfavored formation of pH
sensitive C+‚G-C Hoogsteen base triples at physiological
conditions in the parallel (pyrimidine) binding motif. During
the past decade, many efforts have been devoted to modifying
TFOs to improve binding affinity to their targets along with
designing triplex nucleobases which could alleviate restriction
of the dsDNA sequence. Oligonucleotides possessing modified
AE-RNAs),4 and N3′fP5′ phosphoramidates5 inducing in-
creased binding affinity are among the most successful chemi-
cally modified TFOs. The stabilization of the triplex structures
has been also observed upon addition of heterocyclic compounds
(intercalators) sometimes possessing a positively charged side
chain to the aqueous solution containing all three oligonucleotide
sequences.6 It has been also shown that an intercalator covalently
linked to the 3′-end or the 5′-end of a TFO led to thermal
stabilization of parallel triplexes in the range +3.0 to +16.1
°C depending on linker length and type of intercalator.7
However, there has been limited attention paid to the covalently
attached intercalators inserted as a bulge in the middle of TFO.8
This design could have several advantages. First, in comparison
(2) (a) Nielsen, P. E.; Egholm, M.; Berg, R.; Buchardt, O. Science 1991, 254,
1497-1500. (b) Larsen, H. J.; Bentin, T.; Nielsen, P. E. Biochim. Biophys.
Acta 1999, 1489, 159-166. (c) Demidov, V.; Frank-Kamenetskii, M. D.
Methods 2001, 23, 108-122.
(3) (a) Obika, S.; Hari, Y.; Sugimoto, T.; Sekiguchi, M.; Imanishi, T.
Tetrahedron Lett. 2000, 41, 8923-8927. (b) Sun, B.-W.; Babu, B. R.;
Sørensen, M. D.; Zakrzewska, K.; Wengel, J.; Sun, J.-S. Biochemistry 2004,
43, 4160-4169.
(4) (a) Blommers, M. J. J.; Natt, F.; Jahnke, W.; Cuenoud, B. Biochemistry
1998, 37, 17714-17725. (b) Cuenoud, B.; Casset, F.; Hu¨sken, D.; Natt,
F.; Wolf, R. M.; Altmann, K.-H.; Martin, P.; Moser, H. E. Angew. Chem.
1998, 110, 1350-1353; Angew. Chem., Int. Ed. 1998, 37, 1288-1291.
(5) (a) Gryaznov, S. M.; Lloyd, D. H.; Chen, J. K.; Schulz, R. G.; Dedionisio,
L. A.; Ratmeyer, L.; Wilson, W. D. Proc. Natl. Acad. Sci. U.S.A. 1995,
92, 5798-5802. (b) Escude, C.; Giovannageli, C. Sun, J.-S.; Lloyd, D. H.;
Chen, J.-K.; Gryaznov, S. M.; Garestier, T.; Helene, C. Proc. Natl. Acad.
Sci. U.S.A. 1996, 93, 4365-4369.
(1) Reviews about application and development of triplex forming oligonucle-
otides: (a) Uil, T. G.; Haisma, H. J.; Rots, M. G. Nucleic Acids Res. 2003,
31, 6064-6078. (b) Knauert, M. P.; Glazer, P. Hum. Mol. Genet. 2001,
10, 2243-2251. (c) Guntaka, R. V.; Varma, B. R.; Weber, K. T. Int. J.
Biochem. Cell Biol. 2003, 35, 22-31. (d) Buchini, S.; Leumann, C. J. Curr.
Opin. Chem. Biol. 2003, 7, 717-726. (e) Faria, M.; Giovannageli, C. J.
Gene Med. 2001, 3, 299-310. (f) Gowers, D. M.; Fox, K. R. Nucleic Acids
Res. 1999, 27, 1569-1577.
9
10.1021/ja053645d CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 14849-14858
14849

A R T I C L E S
Scheme 1
Filichev and Pedersen
O-(1-pyrenylmethyl)glycerol in the middle of the oligodeoxy-
nucleotides (INA) resulted in significantly increased affinities
toward complementary ssDNA, whereas INA/RNA duplexes
and the Hoogsteen-type triplex and duplex were destabilized.9a,e
It has to also be mentioned that mismatch sensitivity on duplex
formation was maintained upon bulged insertions of intercalators
into the oligodeoxynucleotides.9b The unique combination of
the flexible, short glycerol linker which distorted the phosphate
backbone and the appropriate intercalator which stabilized INA/
DNA duplex by desolvation and by stacking with nucleobases
led to a valuable molecule which is now used in nucleic acid
chemical biology.10
We decided to explore this type of intercalator for the design
of TFO. To enhance the stability of the TFO using a short and
flexible linker, the aromatic structure of intercalators should be
long enough to place an intercalator into the dsDNA part of
the triple helix. Therefore, (R)-1-O-(4-polyaryl-phenyl)meth-
ylglycerol could be a good choice because phenyl could also
mimic a nucleobase in the TFO part of the triple helix. The
polyaryl intercalator can also be attached to this phenyl via an
acetylene bridge which provides the necessary structural rigidity
and twisting ability and still unites the aromatic structures. The
acetylene bond itself is believed to improve the intercalating
properties.11 According to the molecular modeling of (R)-1-O-
[4-(1-pyrenylethynyl)phenylmethyl]glycerol by MacroModel
8.0, there is a twisting of 1-pyrenyl and phenyl residues around
the triple bond with a torsion angle of 15.3°. It is believed that
this twisting ability can help the intercalator to adjust itself to
an appropriate position inside the dsDNA. Therefore, we refer
to this type of nucleic acid as a twisted intercalating nucleic
acid (TINA, Scheme 1). In a recent work, we deduced from
molecular modeling that twisting of aromatics around triple
bonds was a contributing factor to stabilize the intercalating
moiety in 5′-5′ alternate strand triplexes.7e Here, we report the
postsynthetic Sonogashira-type on-column derivatization of
oligodeoxynucleotides leading to different TINAs, which were
found to have extraordinarily high affinities in Hoogsteen-type
duplexes and triplexes. Thermal stability and fluorescence
studies of nucleic acid helixes with insertion of TINA monomers
as a bulge formed according to either the Watson-Crick or
Hoogsteen binding mode are also discussed.
with the synthesis of at least four nucleotide monomers needed
for sugar-modified nucleic acids, the synthesis of one phos-
phoramidite of intercalating pseudonucleotides is required.
Second, several bulged insertions of an intercalator monomer
into the sequence could considerably increase duplex and triplex
stabilities compared to the single insertion. Moreover, the
structural difference between Watson-Crick and Hoogsteen
binding modes along with the absence or presence of 2′-OH in
DNA and RNA gives rise to different properties for the various
types of helixes. As a result, bulged insertions of a linker and
the breaking up of the helix by intercalators are expected to
give unique properties for appropriately chosen helixes. This
has led to chemically modified oligonucleotides which could
discriminate between different types of single-stranded nucleic
acids. Recently, we have reported the synthesis and properties
of several intercalating nucleic acids designed for Watson-
Crick-type duplexes (Scheme 1).9 Bulged insertions of (R)-1-
(6) (a) Marchand, C.; Bailly, C.; Nguyen, C.-H.; Bisagni, E.; Garestier, T.;
Helene, C.; Waring, M. J. Biochemistry 1996, 35, 5022-5032. (b) Kukreti,
S.; Sun, J.-S.; Loakes, D.; Brown, D. M.; Nguyen, C.-H.; Bisagni, E.;
Garestier, T.; Helene, C. Nucleic Acids Res. 1998, 26, 2179-2183. (c)
Cassidy, S. A.; Strekowski, L.; Fox, K. R. Nucleic Acids Res. 1996, 24,
4133-4138. (d) Escude, C.; Sun, J.-S.; Nguyen, C.-H.; Bisagni, E.;
Garestier, T.; Helene, C. Biochemistry 1996, 35, 5735-5740. (e) Strekow-
ski, L.; Say, M.; Zegrocka, O.; Tanious, F. A.; Wilson, W. D.; Manzel, L.;
Macfarlane, D. E. Bioorg. Med. Chem. 2003, 11, 1079-1085. (f)
Strekowski, L.; Hojjat, M.; Wolinska, E.; Parker, A. N.; Paliakov, E.;
Gorecki, T.; Tanious, F. A.; Wilson, W. D. Bioorg. Med. Chem. Lett. 2005,
15, 1097-1100. (g) Sandstrom, K.; Warmlander, S.; Bergman, J.; Engqvist,
R.; Leijon, M.; Graslund, A. J. Mol. Recognit. 2004, 17, 277-285.
(7) (a) Puri, N.; Zamaratski, E.; Sund, C.; Chattopadhyaya, J. Tetrahedron 1997,
53, 10409-10432. (b) Mouscadet, J.-F.; Ketterle, C.; Goulaouic, H.;
Carteau, S.; Subra, F.; Le Bret, M.; Auclair, C. Biochemistry 1994, 33,
4187-4196. (c) Mohammadi, S.; Slama-Schwok, A.; Leger, A.; El
Manouni, D.; Shchyolkina, A.; Leroux, Y.; Taillandier, E. Biochemistry
1997, 36, 14836-14844. (d) Cheng, E.; Asseline, U. Tetrahedron Lett.
2001, 42, 9005-9010. (e) Jessen, C. H.; Pedersen, E. B. HelV. Chim. Acta
2004, 87, 2465-2471.
(8) (a) Zhou, B.-W.; Puga, E.; Sun, J.-S.; Garestier, T.; Helene, C. J. Am. Chem.
Soc. 1995, 117, 10425-10428. (b) Kukreti, S.; Sun, J.-S.; Garestier, T.;
Helene, C. Nucleic Acids Res. 1997, 25, 4264-4270. (c) Ossipov, D.;
Zamaratski, E.; Chattopadhyaya, J. HelV. Chim. Acta 1999, 82, 2186-
2200. (d) Abdel-Rahman, A. A.-H.; Ali, O. M.; Pedersen, E. B. Tetrahedron
1996, 52, 15311-15324. (e) Aubert, Y.; Asseline, U. Org. Biomol. Chem.
2004, 2, 3496-3503.
(9) (a) Christensen, U. B.; Pedersen, E. B. Nucl. Acids Res. 2002, 30, 4918-
4925. (b) Christensen, U. B.; Pedersen, E. B. HelV. Chim. Acta 2003, 86,
2090-2097. (c) Filichev, V. V.; Pedersen, E. B. Org. Biomol. Chem. 2003,
1, 100-103. (d) Filichev, V. V.; Hilmy, K. M. H.; Christensen, U. B.;
Pedersen, E. B. Tetrahedron Lett. 2004, 45, 4907-4910. (e) Christensen,
U. B.; Wamberg, M.; El-Essawy, F. A. G.; Ismail, A. E.-H.; Nielsen, C.
B.; Filichev, V. V.; Jessen, C. H.; Petersen, M.; Pedersen, E. B. Nucleosides,
Nucleotides Nucleic Acids 2004, 23, 207-225. (f) Nielsen, C. B.; Petersen,
M.; Pedersen, E. B.; Hansen, P. E.; Christensen, U. B. Bioconjugate Chem.
2004, 15, 260-269.
Results and Discussion
The postsynthetic oligonucleotide modification is a better
alternative to the routine and time-consuming preparation of
several pseudonucleoside phosphoramidites, which are required
for the selection of the right candidate for TINA. There have
been several reports devoted to the palladium(0)-catalyzed
modification of oligonucleotides during solid-phase synthesis.12
Sonogashira coupling conditions were found to be compatible
(10) (a) Filichev, V. V.; Vester, B.; Hansen, L. K.; Abdel Aal, M. T.; Babu, B.
R.; Wengel, J.; Pedersen, E. B. ChemBioChem 2005, 6, 1181-1184. (b)
Filichev, V. V.; Christensen, U. B.; Pedersen, E. B.; Babu, B. R.; Wengel,
J. ChemBioChem 2004, 5, 1673-1679. (c) Loue¨t, S. Bioentrepreneur 23
June 2003, http://dx.doi.org/10.1038/bioent748. INA is commercially
available from Human Genetic Signatures Pty Ltd., Sydney, Australia.
(11) (a) Froehler, B. C.; Wadwani, S.; Terhorst, T. J.; Gerrard, S. R. Tetrahedron
Lett. 1992, 33, 5307-5310. (b) Sa´gi, J.; Szemzo¨, A.; EÄ binger, K.; Szabolcs,
A.; Sa´gi, G.; Ruff, E.; O¨ tvo¨s, L. Tetrahedron Lett. 1993, 34, 2191-2194.
(c) Colocci, N.; Dervan, P. B. J. Am. Chem. Soc. 1994, 116, 785-786.
(12) (a) Khan, S. I.; Grinstaff, M. W. J. Am. Chem. Soc. 1999, 121, 4704-
4705. (b) Beilstein, A. E.; Grinstaff, M. W. Chem. Commun. 2000, 509-
510. (c) Rist, M.; Amann, N.; Wagenknecht, H.-A. Eur. J. Org. Chem.
2003, 2498-2504. (d) Mayer, E.; Valis, L.; Wagner, C.; Rist, M.; Amann,
N.; Wagenknecht, H.-A. ChemBioChem 2004, 5, 865-868.
9
14850 J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005

Hoogsteen-Type Triplexes and Duplexes
A R T I C L E S
Scheme 2 a
Sonogashira reaction mixture as a large portion for several
coupling reactions. After the coupling reaction (3-4 h), the
CPGs were flushed with DMF (2 × 0.5 mL) and CH3CN (2 ×
1 mL), and dried. Then, the oligonucleotides were cleaved off
from the CPG-support with 32% NH4OH (2 h), and deprotected
at 55 °C (overnight). Due to the different lypophilic ability, the
unreacted oligomer and the target TINA were separated by
semipreparative HPLC on a C18 column. In the case of the
overlapping peaks (structure 2), a longer HPLC program was
applied (see Supporting Information). After the first separation,
DMT-on oligonucleotides were treated with 10% AcOH,
purified again on HPLC, and precipitated from ethanol. The
purity of the final TFOs was found to be over 90% for pure
pyrimidine-containing oligodeoxynucleotides and 85-88% for
oligodeoxynucleotides with purines as judged by ion-exchange
HPLC. The composition was verified by MALDI-TOF.
The conversion during the Sonogashira coupling depended
on the reactivity of acetylenes and on the oligo sequence. As
can be judged from a number of experiments with 1-eth-
ynylpyrene, one more treatment with the fresh reaction mixture
was more efficient than the prolonged reaction time (4-16 h).
Smaller amounts of sparingly soluble Glazer byproducts were
formed and better oligo derivatization was observed for Pd-
(PPh3)4 than for Pd(PPh3)2Cl2 as the catalyst in the case of
1-ethynylpyrene. The presence of purines in the sequence
resulted in lower conversion (50-60%) to the target TINA even
after double treatment of the support with the oligonucleotide
by the Sonogashira-coupling reagent mixture containing 1-eth-
ynylpyrene compared to the homopyrimidine sequences (80-
85%) after a single treatment. This also seems true for other
aromatic acetylenes, because in a purine-containing sequence
no target oligonucleotides were obtained using 4-ethynylbiphen-
yl, which was found to be the least reactive compound among
the tested acetylenes. In the synthesis of ON14, we experienced
that the treatment of a complete oligonucleotide with a Sono-
gashira reaction mixture with 1-ethynylpyrene gave a more pure
oligomer than interruption of the DNA synthesis after the second
insertion of 8 followed by Sonogashira reaction and continued
DNA synthesis. In the latter case, short oligomers possessing
pyrenes contaminated the final TINA as judged by ion-exchange
HPLC.
a Reagents and conditions: (a) 4-iodobenzylbromide, KOH, toluene; (b)
80% aq CF3COOH, room temperature, 100% over two steps; (c) DMTCl,
pyridine, room temperature, 70%; (d) NC(CH2)2OP(NPri2)2, diisopropyl-
ammonium tetrazolide, CH2Cl2, 0 °C to room temperature, overnight, 67%.
with the DNA synthesis, and no side reactions were observed
for nucleobases possessing protective groups.12a According to
the known protocols,12 the DNA synthesis is stopped after the
incorporation of 5′-O-DMT-2′-deoxy-5-iodouridine at the 5′-
end of the sequence followed by treatment of the oligonucleotide
support under Sonogashira conditions. Afterward, the oligo-
nucleotide support is attached to the DNA synthesizer, and oligo
synthesis is continued automatically. There is a risk that the
coupling efficiency for the standard phosphoramidites drops after
on-column derivatization, which we have also observed in our
experiments described below. Despite the fact that some
organometallic couplings were applied for postsynthetic oligo-
nucleotide modification,13 the postsynthetic Sonogoshira-type
reactions on the convertible nucleoside 2′-deoxy-5-iodouridine
located in the middle of the sequence were reported to be
unsuccessful.12b Instead, we took a chance to use (R)-1-O-(4-
iodobenzyl)glycerol in Sonogoshira-type reactions after its
incorporation into the middle of the oligos. A number of
aromatic structures with the terminal triple bond (2-5) were
used in this context (Scheme 1).
The required phosphoramidite 8 was synthesized in four steps
from 4-iodobenzylbromide and (S)-(+)-2,2-dimethyl-1,3-diox-
olane-4-methanol in 47% overall yield (Scheme 2). The coupling
efficiency of compound 8 during DNA synthesis on a 0.2 µmol
scale using standard nucleotide coupling conditions (2 min
coupling, 4,5-dicyanoimidazole as an activator) and increased
deprotection time (100 s) was estimated to be more than 99%.
After the DNA synthesis, the CPG-supports with DMT-on
oligonucleotides possessing 4-iodophenyl moieties were treated
with a Sonogashira-coupling reagent mixture containing Pd-
(PPh3)4 or Pd(PPh3)2Cl2 (7.5 mM), an aromatic structure
possessing a terminal acetylene (22.5 mM), and CuI (7.5 mM)
in dry DMF/Et3N (3.5/1.5, 500 µL) in 1 mL syringes under dry
conditions at room temperature. It was important to flush
supports and syringes with argon instead of nitrogen prior to
the coupling reaction in order to avoid Glazer oxidative
dimerization. The conversion was found to be better when the
Sonogashira reaction mixture was prepared directly in the plastic
syringe for each individual oligo instead of preparing the
Very recently, B. Liang et al.14 have reported the copper-
free Sonogashira coupling reaction with PdCl2 in water in the
presence of pyrrolidine. The compatibility with water, aerobic
conditions, and traces of homocoupling products are the very
big advantages of this method. We applied the analogous
conditions on the fully deprotected ON2. However, after
treatment of ON2 with 1-ethynylnaphthalene and PdCl2 in water/
pyrrolidine (1:1) at 50 or 20 °C overnight, no trace of the desired
nucleic acids was observed after HPLC purification.
The thermal stability of triplexes and DNA/DNA and DNA/
RNA duplexes with the synthesized oligonucleotides was
assessed by thermal denaturation experiments. The melting
temperatures (Tm’s, °C) determined as first derivatives of melting
curves are listed in Tables 1-4. The sequences possessing
different TINAs were studied in pH-dependent Hoogsteen-type
base pairing, both in the parallel triplex15a toward the duplex
(13) (a) Gartner, Z. J.; Kanan, M. W.; Liu, D. R. Angew. Chem. 2002, 114,
1874-1878; Angew. Chem., Int. Ed. 2002, 41, 1796-1800. (b) Gartner,
Z. J.; Grubina, R.; Calderone, C. T.; Liu, D. R. Angew. Chem. 2003, 115,
1408-1413; Angew. Chem., Int. Ed. 2003, 42, 1370-1375.
(14) Liang, B.; Dai, M.; Chen, J.; Yang, Z. J. Org. Chem. 2005, 70, 391-393.
(15) (a) Felsenfeld, G.; Miles, H. T. Annu. ReV. Biochem. 1967, 36, 407-449.
(b) Liu, L.; Miles, H. T.; Frazier, J.; Sasisekharan, V. Biochemistry 1993,
32, 11802-11809.
9
J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005 14851

A R T I C L E S
Filichev and Pedersen
Table 1. Tm [°C] Data for Triplex and Duplex Melting, Taken from UV Melting Curves (λ ) 260 nm)
triplexa
3
′
-CTGCCCCTTTCTTTTTT
parallel duplexb
-GACGGGGAAAGAAAAAA
(ON15)
antiparallel duplexc
-GGGGAAAGAAAAAA
(ON16)
5
′
-GACGGGGAAAGAAAAAA
5′
3′
(D1)
pH 5.0
pH 6.0
pH 7.2
pH 5.0
pH 6.0
pH 5.0
pH 6.0
pH 7.2
ON1
5′-CCCCTTTCTTTTTT
5′-CCCCTT1TCTTTTTT
5′-CCCCTT2TCTTTTTT
5′-CCCCTT3TCTTTTTT
5′-CCCCTT4TCTTTTTT
5′-CCCCTT5TCTTTTTT
5′-CCCCTTTC5TTTTTT
5′-CCCCTTT5CTTTTTT
5′-CCC5CTTTCTTTTTT
5′-5CCCCTTTCTTTTTT
5′-5CCCCTT5TCTTTTTT
5′-CCCCTT5T5CTTTTTT
5′-CCCCT5TT5CTTTTTT
5′-CCCCTT5TCT5TTTTT
55.0e
i
i
i
57.0
59.0e
i
i
i
61.0
65.5e
55.5e
59.5e
63.0e
27.0
15.0
26.0
26.0
35.0
46.0
39.5
42.5
41.0
44.5
57.0d
40.0
56.5e
56.5e
<5.0
<5.0
<5.0
<5.0
13.5
28.0
21.5
26.0
24.0
20.5
35.5
<5.0
40.0
43.0
29.5
i
i
i
33.5
42.0
i
i
i
46.0
53.5
37.0
41.0
45.5
19
47.0
i
i
i
44.5
44.0
i
i
i
49.5
46.5
37.5
44.5
42.5
48.0
40.5
42.0
40.0
45.0
46.5
44.5
45.0
45.5
53.0
47.0
41.0
45.0
41.0
47.0
i
i
i
46.0
45.5
i
i
i
52.0
46.5
41.0
42.0
38.0
ON2
<5.0
<5.0
17.0
22.0
33.5
30.0
28.0
31.5
36.0
45.5
26.5
38.0
38.0
ON3f
ON4g
ON5f
ON6f,g
ON7f
ON8f
ON9g
ON10f
ON11g
ON12h
ON13h
ON14h
a C ) 1.5 µM of ON1-14 and 1.0 µM of each strand of dsDNA (D1) in 20 mM sodium cacodylate, 100 mM NaCl, 10 mM MgCl2, pH 5.0, 6.0 and 7.2;
duplex Tm ) 56.5 °C (pH 5.0), 58.5 °C (pH 6.0) and 57.0 °C (pH 7.2). b C ) 1.0 µM of each strand in 20 mM sodium cacodylate, 100 mM NaCl, 10 mM
MgCl2, pH 6.0 or pH 5.0. c C ) 1.0 µM of each strand in 20 mM sodium cacodylate, 100 mM NaCl, 10 mM MgCl2, pH 5.0, 6.0 and pH 7.2. d Third strand
and duplex melting overlaid. Transition with Tm ) 54.5 °C was determined at 373 nm. e Third strand and duplex melting overlaid. f Prepared by Sonogashira
reaction mixture: Pd(PPh3)2Cl2 (7.5 mM), corresponding acetylene (22.5 mM), CuI (7.5 mM), dry DMF/Et3N (3.5/1.5, 500 µL), 3 h. g Prepared by Sonogashira
reaction mixture: Pd(PPh3)4 (7.5 mM), corresponding acetylene (22.5 mM), CuI (7.5 mM), dry DMF/Et3N (3.5/1.5, 500 µL), 3 h. h Prepared by double
treatment with Sonogashira reaction mixture: Pd(PPh3)4 (7.5 mM), 1-ethynylpyrene (22.5 mM), CuI (7.5 mM), dry DMF/Et3N (3.5/1.5, 500 µL), 3 h. i Not
determined.
Table 2. Tm [°C] Data for Mismatched Parallel Triplexa and
pH 6.0 which corresponds to ∆Tm ) 19.0 °C when compared
Parallel Duplexb Melting Taken from UV Melting Curves (λ ) 260
to the wild-type triplex. Even at pH 7.2 a single incorporation
nm) at pH 6.0
of 5 led to a considerable stabilization of the triplex (ON6/
D1), despite a high cytosine content (36%). At this pH, no
sequence
3′-CTGCCCCTTXCTTTTTT
hybridization could be detected for the wild-type triplex (Figure
1). For the parallel duplexes at pH 6.0, the stabilization of 3.0
and 14.5 °C per modification was detected for 1-naphthalenyl-
ethynyl (ON5) and 1-pyrenylethynyl (ON6), respectively. As
expected, at lower pH (pH ) 5.0) parallel duplexes were found
to be more stable due to protonation of cytosine. It could be
concluded that attaching the aromatic structures at the 4-position
of the phenyl ring in (R)-1-O-(phenylmethyl)glycerol resulted
in increasing hybridization affinity in Hoogsteen-type helixes.
Interestingly, naphthalene and pyrene rings gave considerably
better stabilization than 4-biphenyl and benzene. This supports
the idea that aromatic structures with a large surface such as
pyrene are preferred for attachment to (R)-1-O-(4-substituted
phenylmethyl)glycerol over small aromatic structures to achieve
good binding in Hoogsteen-type helixes.
Destabilization of antiparallel dsDNA was observed for all
studied modified oligodeoxynucleotides except when the inter-
calator 5 was placed at the 5′-end (ON10/ON16) as compared
with the wild-type dsDNA (ON1/ON16, Table 1). The stabiliz-
ing effect in the latter case could be ascribed to stacking of an
aromatic polycyclic system on the adjacent nucleobase (the
effect as a lid), while the effect of the acyclic linker is
marginal.16 Hybridization affinity was also dependent on the
structure of TINA. The least destabilized duplexes were formed
with 4 and 5, whereas the destabilization of dsDNA was larger
for structures 1-3 incorporated as a bulge in the middle of the
sequence. Already at this stage it can be concluded that TINA
incorporated as a bulge into helixes is improving the stability
of Hoogsteen-type helixes and not Watson-Crick type duplexes.
5′-GACGGGGAAYGAAAAAA
D1,
D2,
D3,
D4,
X‚
Y
)
T‚A X‚Y) A‚T X‚Y) C‚G X‚Y) G‚C
ON1 5′-CCCCTTTCTTTTTT
ON6 5′-CCCCTT5TCTTTTTT
ON8 5′-CCCCTTT5CTTTTTT
ON10 5′-5CCCCTTTCTTTTTT
27.0
46.0
42.5
44.5
<5.0
27.0
28.5
22.5
40.5
<5.0
34.5
26.5
27.0
45.5
<5.0
28.5
26.5
28.0
42.0
ON11 5′-5CCCCTT5TCTTTTTT 57.0
5′-GACGGGGAAYGAAAAAA
ON15,
ON17,
ON18,
ON19,
Y
)A
Y
)
T
Y
)
G
Y) C
ON1
ON6
ON8
5′-CCCCTTTCTTTTTT
5′-CCCCTT5TCTTTTTT
5′-CCCCTTT5CTTTTTT
19.0
33.5
28.0
<5.0
21.5
20.0
10.0
20.5
18.5
<5.0
20.5
20.0
a C ) 1.5 µM of ON1-14 and 1.0 µM of each strand of dsDNA in 20
mM sodium cacodylate, 100 mM NaCl, 10 mM MgCl2, pH 6.0. b C ) 1.0
µM of ON1-14 and 1.0 µM of purine strand.
D1 and in parallel dsDNA15b toward ON15 (Table 1). The same
sequences (ON1-14) were used for Watson-Crick DNA/DNA
antiparallel duplexes toward ON16. For the latter type of
duplexes, mixed pyrimidine/purine sequences similar to those
described earlier for INA9a,d were also used for TINA oligo-
nucleotides for hybridization with ssDNA and ssRNA (Table
4).
As can be seen from the Tm data in Table 1, considerable
destabilization of the Hoogsteen-type triplex and duplex was
observed for ON2 with (R)-1-O-(4-iodophenylmethyl)glycerol
as a bulge in the middle of the sequence compared to the wild-
type complexes at pH 6.0 (ON1 toward D1 and ON15).
Substitution of the iodine with aryl substituents gave more stable
triplexes (ON3-6 toward D1, pH 6.0). The highest Tm value
46.0 °C was observed for the 1-pyrenylethynyl substituent at
(16) (a) Ren, R. X.-F.; Chaudhuri, N. C.; Paris, P. L.; Rumney, I. S.; Kool, E.
T. J. Am. Chem. Soc. 1996, 118, 7671-7678. (b) Guckian, K. M.;
Schweitzer, B. A.; Ren, R. X.-F.; Sheilis, C. J.; Paris, P. L.; Tahmassebi,
D. C.; Kool, E. T. J. Am. Chem. Soc. 1996, 118, 8182-8183.
9
14852 J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005

Hoogsteen-Type Triplexes and Duplexes
A R T I C L E S
Table 3. Tm [°C] Data for Parallel Triplex Meltingsa for Insertions of 5 in the Sequence of the Watson-Crick Duplex Taken from UV Melting
Curves (λ ) 260 nm) at pH 6.0b
triplex
parallel duplex
D5, 3
′
-CTGCCCCTT5TCTTTTTT (ON20)c
D6, 3′-CTGCCCCTTTCTTTTTT
-GACGGGGAA5AGAAAAAA (ON21)d
sequence
D1
5′
-GACGGGGAAAGAAAAAA
5′
D7, ON20/ON21
ON15
ON21
ON1
ON6
ON9
5′-CCCCTTTCTTTTTT
5′-CCCCTT5TCTTTTTT
5′-CCC5CTTTCTTTTTT
27.0
46.0
41.0
38.0
38.0
52.5e
24.0
27.5
41.5
27.0
31.5
43.5
19.0
33.5
31.5
14.0
26.5
29.0
a C ) 1.5 µM of ON1, ON6, ON9, and 1.0 µM of each strand of dsDNA in 20 mM sodium cacodylate, 100 mM NaCl, 10 mM MgCl2, pH 6.0, duplex
Tm ) 55.0 °C (D5), 56.0 °C (D6), 57.0 °C (D7). b The meltings are also given for parallel duplexes (C ) 1.0 µM of ON1, ON6, ON9, ON15 and ON21
in 20 mM sodium cacodylate, 100 mM NaCl, 10 mM MgCl2, pH 6.0.) with insertion of 5 in the purine stretch. c Prepared by Sonogashira reaction mixture:
Pd(PPh3)4 (7.5 mM), 1-ethynylpyrene (22.5 mM), CuI (7.5 mM), dry DMF/Et3N (3.5/1.5, 500 µL), 3 h. d Prepared by double treatment with Sonogashira
reaction mixture: Pd(PPh3)4 (7.5 mM), 1-ethynylpyrene (22.5 mM), CuI (7.5 mM), dry DMF/Et3N (3.5/1.5, 500 µL), 3 h. e Third strand and duplex melting
overlaid.
Table 4. Tm [°C] Data for Antiparallel Duplexa Melting Taken from UV Melting Curves (λ ) 260 nm)
ON25
DNA
ON26b
DNA
ON27
RNA
5′
-AGCTTGCTTGAG
5′
-AGCTTG5CTTGAG
5
′
-AGCUUGCUUGAG
ON22
ON23b
ON24c
3′-TCGAACGAACTC
3′-TCGAAC5GAACTC
3′-TCGAAC5G5AACTC
47.5
39.5
34.0
32.0
36.0
22.5
40.5
30.5
25.0
a C ) 1.0 µM of each oligonucleotide in 140 mM NaCl, 10 mM sodium phosphate buffer, 1 mM EDTA, pH 7.0. b Sonogashira reaction mixture:
Pd(PPh3)4 (7.5 mM), 1-ethynylpyrene (22.5 mM), CuI (7.5 mM), dry DMF/Et3N (3.5/1.5, 500 µL), 3 h. c Double treatment with Sonogashira reaction
mixture: Pd(PPh3)4 (7.5 mM), 1-ethynylpyrene (22.5 mM), CuI (7.5 mM), dry DMF/Et3N (3.5/1.5, 500 µL), 3 h.
(ON10) among the tested TFOs with single insertion of
1-pyrenylethynyl (ON6-10). It was a surprise that the lid
effect16 was absent here. This could be a consequence of
generally lower stability of C-rich regions of TFO with the target
dsDNA under physiological conditions.17 However, it is an
important observation that efficient hybridization affinity could
be achieved by placing 5 in the middle of the C-rich region
(ON9) in neutral media. One can speculate whether intercalation
will make protonation more likely in the triplex structure at
physiological pH because the intercalator is separating two
positively charged triples.
The dependence of the distance between multiple bulged
insertions of the pyrene intercalator 5 on thermal stability was
investigated using ON11-14 (Table 1). In the case of overlap-
ping triplex and duplex transitions, melting experiments were
performed at 373 nm. However, sometimes not very well-
defined transitions were observed at 373 nm. In these cases,
Figure 1. First derivative plots of triplex melting recorded at 260 nm versus
increasing temperature (1 °C/min) in 20 mM sodium cacodylate, 100 mM
NaCl, 10 mM MgCl2, pH 7.2.
the meltings at pH 6.0 of the triplexes at temperatures close to
those of the duplexes were based on comparison with meltings
at pH 7.2 which were measured at 260 nm. When the intercalator
5 was inserted as a next nearest neighbor (ON12), the Hoogsteen
triplex and duplex were stabilized compared to the unmodified
ON1 at pH 6.0. However, the stabilities in both cases were lower
than for the single insertion of 5 (ON6), and no triplex formation
was observed at pH 7.2. This could be due to the large
interruption of the double and triple helixes by two bulged (R)-
1-O-methylglycerol linkers positioned very close to each other.
When the two insertions of 5 were separated by two or three
nucleobases (ON13 and ON14, respectively), the complexes
with D1 and ON15 were more stable than those with single
insertions. At pH 7.2, the Tm for the triplexes was even higher
than the physiological temperature 37 °C (see ON14/D1 in
Figure 1). Like double insertions of 5 in the middle, double
insertions with one insertion at the 5′-end with six base-pairs
between the insertions (ON11) considerably stabilized the
Thus, a single insertion of (R)-1-O-[4-(1-pyrenylethynyl)-
phenylmethyl]glycerol as a bulge in a less stable parallel triplex
(ON6/D1) at pH 6.0 stabilized the triplex to the level of a
Watson-Crick duplex (ON6/ON16) with the same nucleotide
content. The thermal stability for different TINAs prompted us
to investigate the properties of the 1-pyrenylethynyl-containing
TINA more closely.
Some fluctuation in the thermal stability of Hoogsteen’s
triplexes and duplexes was seen on placing 1-pyrenylethynyl
at different positions in the TFO. When cytosine was neighbor-
ing either the 5′-side or the 3′-side (ON7-9), both the parallel
triplex and the parallel duplex were less stabilized than when 5
was placed between two thymidines at pH 6.0 (ON6). This could
be a result of the interaction of the aromatic structure with the
positively charged pair of C+‚G. Interestingly, at pH 7.2, when
cytosine was not protonated, the lowest triplex hybridization
affinity was detected for TFO with 5 at the 5′-dangling end
(17) Hildbrand, S.; Blaser, A.; Parel, S. P.; Leumann, C. J. J. Am. Chem. Soc.
1997, 119, 5499-5511.
9
J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005 14853

A R T I C L E S
Filichev and Pedersen
Hoogsteen-type duplex and triplex at pH 6.0. Opposite to the
Hoogsteen helixes, antiparallel duplexes with double insertions
of 5 (ON12-14/ON16) showed decreased stabilities when
compared with the wild-type duplex ON1/ON16, especially
when one or three nucleobases were between the two insertions.
When comparing thermal stabilities of parallel and antiparallel
duplexes with double insertions of 5 at pH 5.0, Hoogsteen
duplexes ON11/ON15 and ON14/ON15 were even more stable
than the corresponding Watson-Crick duplexes (ON11/ON16
and ON14/ON16). The stabilization of parallel triplexes and
parallel duplexes upon addition of an intercalator was first
reported by C. Escude et al.18 for benzopyridoindole (BPI)
derivatives. The reorganization of a nonperfectly matched
Watson-Crick DNA duplex into a perfectly matched Hoogsteen
paired DNA duplex has been detected when BPI was added to
the aqueous solution of the oligodeoxynucleotides. In cases of
multiple insertions of 5 into oligodeoxynucleotides, the similar
reorganization to a fully matched parallel duplex is also
anticipated depending on pH and salt concentration.
strand on the 3′-site of the intercalator (ON6/D3 and ON11/
D3, Table 2). In all other cases, the dropping of Tm was in the
range 14.0-22.0 °C. In the work of Zhou et al.,8a the bulged
insertion of 2-methoxy-6-chloro-9-aminoacridine via a flexible
linker in the middle of TFO was used to stabilize triplexes with
mismatches to the level of the perfectly matched triplexes. The
∆Tm values were within 10 °C. In our case, the ∆Tm values are
higher, which is believed to be sufficient to distinguish between
matched and single mismatched triplexes. To reduce the false
positives which could come from binding of the TFO possessing
5 to mismatched or shorter dsDNAs, the structured probes
should be used.19 The benefit of structured probes over linear
nucleic acids has been already presented for specific DNA
triplex formation.19c Mismatched parallel duplexes with a single
insertion of 5 were destabilized in the range 8.0-13.0 °C which
was in the same range as mismatched wild type parallel duplexs.
For comparison, the least sensitive mismatched unmodified
duplex showed a ∆Tm of 9 °C at pH 6.0 (Tm(ON1/ON15)
-
Tm(ON1/ON18)).
The extraordinary stabilization of parallel triplexes was
observed at pH 5.0. High content of cytosines in the TFO shifted
the melting of the unmodified triplex (Tm ) 55.0 °C) close to
the melting of the duplex. However, this value was still lower
than that of the duplex melting at pH 5.0 (Tm(D1) ) 56.5 °C).
Single bulged insertion of 1-naphthalenylethynyl derivative
We studied the luminescent characteristics of the TFO
possessing (R)-1-O-[4-(1-pyrenylethynyl)phenylmethyl]glycerol
(5) which was the most effective TINA to form triplexes and
to discriminate mismatches to the duplex. The introduction of
5 into oligonucleotides resulted in a characteristic monomeric
fluorescence spectrum, with maxima at 400 and 421 nm upon
excitation at 373 nm (Figure 2A, black curve), which was similar
to previously published data for 4-[4-(1-pyrenylethynyl)phenyl]-
1,3-butanediol inserted into DNA.20 In all cases, a 4 nm shift
of monomeric fluorescence was detected upon formation of
triplexes or duplexes. Formation of the fully matched triplex
led to approximately 1.5-fold increased monomeric fluorescence
(Figure 2A, ON6/D1) compared to the single-stranded ON6.
For nonperfectly matched triplexes, the fluorescence intensity
depended on the sequence of dsDNA. Thus, an almost 2-fold
increase was detected for a TA inversion site (ON6/D2)
compared to ON6. On the contrary, when a cytosine or a
guanine base was mismatching in dsDNA to the TFO near the
insertion of 5 (D3 and D4), a decrease of monomeric fluores-
cence was seen in comparison with the perfectly matched triplex
(Figure 2A). Especially guanine gave a large effect with 2-fold
lower fluorescence intensity for the mismatched triplex ON6/
D3.
Interestingly, a considerable increase in monomer fluores-
cence was detected upon formation of the antiparallel duplex
(ON6/ON16, Figure 2C), whereas the formation of the parallel
duplex (ON6/ON15) resulted in only a slightly increased
fluorescence when compared with the single-strand fluorescence
(Figure 2C). When a second 4-(1-pyrenylethynyl)phenyl residue
appeared as a next-nearest neighbor in ON12, the monomeric
fluorescence of the single-strand decreased approximately 3-fold
(Figure 2C, comparison of ON6 and ON12), and an excimer
fluorescence with a maximum at 500 nm and with an intensity
half of that of the monomeric intensity could be observed (Figure
2B). A considerable decrease of the monomeric fluorescence
and disappearance of the excimer band was observed for the
4
in the TFO slightly increased the triplex stability
(∆Tm(ON5/D1-ON1/D1) ) 2.0 °C). However, bulged insertion of 5
in all cases led the dissociation of the whole complex at
temperatures which were higher than Tm for the dsDNA (D1).
The clear transition for ON11 was observed at 373 nm at the
same temperature as at 260 nm, which confirmed that the triplex
and the duplex melted together. The same dependence of thermal
stability for double insertion of 5 in TFO as at pH 6.0 was
observed at pH 5.0. Thus, the double insertion of 5 as next-
nearest neighbors (ON12) and insertions of 5 in the middle and
at the 5-end (ON11) were the least and the most stabilized
triplexes, respectively. These results support the idea of placing
the intercalator into the dsDNA part of the triple helix which
under appropriate conditions can increase the thermal stability
of the duplex itself. At pH 5.0, the triplex ON14/D1 was 16.5
and 20.5 °C more stable than the corresponding parallel and
antiparallel duplexes, respectively. Importantly, even at pH 7.2
oligonucleotide ON14 forms more stable Hoogsteen-type triplex
(Tm ) 43.0 °C, ON14/D1) than the corresponding Watson Crick
dsDNA (Tm ) 38.0 °C, ON14/ON16). At pH 7.2, the melting
temperature for the parallel duplex (ON14/ON15) is supposed
to be lower than 38.0 °C observed at pH 6.0 since this duplex
is pH-sensitive. These data clearly demonstrate the ability of
oligonucleotides with multiple insertions of 5 in the middle of
the sequence separated by three bases to discriminate well
between dsDNA and ssDNA.
The sensitivity to mismatches was studied for parallel
triplexes and duplexes with bulged insertion of 5 in the middle
and at the 5′-end of the sequence (Table 2). In the case of
triplexes, the sensitivity to mismatch was dependent on the site
of insertion of the intercalator. The smallest value of ∆Tm
11.5 °C between matched and mismatched triplexes was
detected when adenine was replaced by guanine in the purine
)
(19) (a) Demidov, V. V.; Frank-Kamenetskii, M. D. Trends Biochem. Sci. 2004,
29, 62-71. (b) Bonnet, G.; Tyagi, S.; Libchaber, A.; Kramer, F. R. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 6171-6176. (c) Roberts, R. W.; Crothers,
D. M. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 9397-9401.
(20) Malakhov, A. D.; Skorobogatyi, M. V.; Prokhorenko, I. A.; Gontarev, S.
V.; Kozhich, D. T.; Stetsenko, D. A.; Stepanova, I. A.; Shenkarev, Z. O.;
Berlin, Y. A.; Korshun, V. A. Eur. J. Org. Chem. 2004, 1298-1307.
(18) (a) Escude, C.; Mohammadi, S.; Sun, J.-S.; Nguyen, C.-H.; Bisagni, E.;
Liquier, J.; Taillandier, E.; Garestier, T.; Helene, C. Chem. Biol. 1996, 3,
57-65.
9
14854 J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005

Hoogsteen-Type Triplexes and Duplexes
A R T I C L E S
Figure 2. Fluorescence spectra of single-strands, antiparallel and parallel duplexes, and parallel triplexes. Measurement conditions: 1 µM of each strand
in a buffer at 10 °C (20 mM sodium cacodylate, 100 mM NaCl, 10 mM MgCl2, pH 6.0), excitation 373 nm (excitation slit 4.0 nm), emission 380-600 nm
(emission slit 2.5 nm for A, B, E, and 0.0 nm for C and D). ON6 and ON12 were used as references in spectra recorded under different conditions.
same oligo in a matched triplex (Figure 2B, ON12/D1). This
means that the pyrene moieties were not in the sufficient
proximity needed for excimer formation upon binding to dsDNA
in the environment of the triplex helix. Similarly, the excimer
band disappeared when ON12 formed a Hoogsteen-paired
dsDNA with ON15 (Figure 2B). On the contrary, very high
monomeric fluorescence intensity and increased excimer fluo-
rescence were observed for the antiparallel duplex (ON12/
ON16) when compared with fluorescence intensities of the
single-stranded ON12 (Figure 2C). This indicates that the two
pyrenyls in the same strand were still in close contact with each
other after formation of the Watson-Crick dsDNA although
this seems not the case in the Hoogsteen-type dsDNA. In this
way, the different properties of TINA toward Watson-Crick-
and Hoogsteen-type helixes were reflected by both hybridization
and fluorescence properties. Moreover, fluorescence data for
(R)-1-O-[4-(1-pyrenylethynyl)phenylmethyl]glycerol (5) as bulged
next-nearest neighbors in a pyrimidine-rich strand can be
summarized as follows: single-strand ON12, medium monomer
fluorescence at 400 and 421 nm and excimer band at 500 nm;
9
J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005 14855

A R T I C L E S
Filichev and Pedersen
parallel triplex ON12/D1, low monomer fluorescence and no
excimer band; parallel duplex ON12/ON15, medium monomer
fluorescence and no excimer band; antiparallel duplex ON12/
ON16, high monomer fluorescence and excimer band.
insertions of (R)-1-O-(1-pyrenylmethyl)glycerol using the same
sequences.9a,d
When two pyrenyl intercalators 5 were separated by one base-
pair, an excimer band observed for the ssON24 (Figure 2E)
did not disappear upon formation of the antiparallel duplex
(ON24/ON25 and ON24/ON27, Figure 2E). This observation
is opposite to the above observations for parallel triplex and
parallel duplex with TINA and is also contrary to the previously
obtained results for INA.9b The presence of the excimer band
upon formation of the antiparallel duplex in both homopyrimi-
dine and mixed pyrimidine/purine strands (ON12/ON16 and
ON24/ ON25, respectively) with bulged 5 as next-nearest
neighbors indicates that two pyrenyl residues were positioned
very closely and communicated with each other and were not
fully embedded into stacking interactions with neighboring
Watson-Crick base-pairs. This can also explain the decrease
of the antiparallel duplex stability upon incorporation of 5 as a
bulge.
The ability of structure 5 to affect the stability of the parallel
triple helix upon its incorporation into the Watson-Crick duplex
part of the triplex is presented in Table 3. The triplex was
stabilized in all cases when 5 was inserted as a bulge in the
pyrimidine strand of the duplex (ON1, ON6, and ON9 toward
D5) when compared to the unmodified triplex (ON1/D1). Two
transitions (Tm ) 36.5 and 55.5 °C) were detected for the triplex
ON1/D1 in the thermal denaturation experiment at 373 nm [λmax
of 5] which corresponded to triplex and duplex meltings,
respectively. Detection of the meltings by the 373 nm absor-
bance indicated that the intercalator was involved in both the
duplex and triplex formation. The insertion of 5 as bulges in
both Watson-Crick and Hoogsteen pyrimidine strands opposite
to each other (ON6/D5) did not change the melting of the triplex
when compared with a triplex with an intercalator only in the
duplex part of the triplex (ON1/D5). When two pyrene moieties
5, one in each of the pyrimidine strands, were placed as bulges
separated by three base-pairs (ON9/D5), the triplex melting was
very close to the duplex melting temperature, which was also
observed above for the double incorporation of 5 into TFO.
Decreased triplex and parallel duplex stabilities compared to
the unmodified complexes were observed when compound 5
was inserted in the purine strand as a bulge (ON1 toward D6
and toward ON21).
We then checked whether an excimer bond could be formed
for duplexes and triplexes if two or three dyes were placed
opposite to each other, in each of their complementary strand.
No excimer band was observed in either parallel duplex ON21/
ON6 (Figure 2D) or antiparallel duplexes ON20/ON21 (Figure
2D) and ON23/ON26 (Figure 2E). This result correlates with
the work of A. Malakhov et al.20 when 4-[4-(1-pyrenylethynyl)-
phenyl]-1,3-butanediol was positioned opposite one another in
the complementary strands of antiparallel dsDNA with mixed
sequences. Only for the triplex with three 4-(1-pyrenylethynyl)-
phenyl moieties placed opposite to each other in all three strands
was a weak excimer band at 500 nm detected (ON6/D7, Figure
2D). The conclusion is that the communication of the 4-(1-
pyrenylethynyl)phenyl moieties positioned in different strands
of the parallel and antiparallel duplexes and parallel triplex is
impeded, which makes zipping of intercalators together with
excimer formation unlikely, contrary to what was found for INA.
Thus, zipping of two pyrene moieties of INA situated opposite
to each other in a duplex have been observed in an NMR
structure,9f and this duplex structure led to formation of an
excimer band at 480 nm in a steady-state fluorescence spectra
upon excitation at 343 nm (unpublished data).
We also studied the hybridization affinity of TINA possessing
5 toward mixed purine/pyrimidine sequences of ssDNA and
ssRNA in Watson-Crick-type duplexes (Table 4) using the
same sequence and conditions as have been described for
INA.9a,d Considerable destabilization of TINA/DNA (∆Tm in
the range -8.0 to -15.5 °C) and TINA/RNA (∆Tm ) -10.0
°C) was observed for 5 as a bulge in the middle of the sequence
when compared with the wild-type duplexes. The insertion of
the second intercalator 5 as a next-nearest neighbor into DNA
(ON24) led to further destabilization of the duplex (ON24/ON25
and ON24/ON27). The incorporation of 5 opposite to each other
into two complementary mixed purine-pyrimidine strands, as
the complex ON23/ON26, resulted in a Tm value of 36.0 °C
which was at the same level of magnitude as TINA/DNA
duplexes (ON23/ON25 and ON22/ON26). However, when INA
was inserted in the same positions in INA/INA duplexes, they
were less stable (Tm ) 43.6 °C) than INA/DNA (Tm ) 51.5
°C).9d
The differences in fluorescence spectra and hybridization
properties of the two different pyrene intercalating nucleic acids
INA and TINA in Watson-Crick-type duplexes clearly illustrate
the consequence of adding an extra 1-phenylethynyl moiety to
the aromatic part of (R)-1-O-(1-pyrenylmethyl)glycerol (INA).
By this work we have succeeded in placing pyrene appropriately
in the Hoogsteen-type triplex. The ability of intercalators to
stabilize parallel triplex structures with only little influence on
the stability of dsDNA is known. Thus, addition of 2-(2-
naphthyl)quinolin-4-amine and analogues thereof6e,f lead to
considerable stabilization of triplex DNA [∆Tm ) 35.6 °C for
2-(2-naphthyl)quinolin-4-amine] with only a little increased
hybridization affinity of duplex DNA (∆Tm ) 5.5 °C). Y. Aubert
et al.8e have in a similar work reported the synthesis and
hybridization properties of oligodeoxynucleotides with perylene
coupled either directly or via a propyl linker to the anomeric
position of a 2′-deoxyribose residue. One of the advantages of
TINA with polycyclic moieties over monomeric triplex-specific
intercalators is that TINA can be inserted several times into
desired positions of the sequence instead of using excess of the
The fluorescence properties of complexes with the 4-(1-
pyrenylethynyl)phenyl moiety in the Watson-Crick dsDNA as
a duplex alone and as a part of the triplex are shown in Figure
2C-E. The monomer fluorescence was considerably increased
when 5 was inserted into the purine strand (ssON21) compared
with the insertion into the pyrimidine strand (ssON6, Figure
2D). Slightly decreased fluorescence intensity was seen upon
assembling of the triplexes and duplexes with unmodified DNA
and ssON21 (data not shown). It was a surprising finding that
the strong sensitivity of the monomer fluorescence of 4-(1-
pyrenylethynyl)phenyl moieties in homopyrimidine sequences
upon the formation of antiparallel duplexes completely disap-
peared for mixed sequences (ON23/ON25, ON23/ON27, Figure
2E). This differs also from previous results reported for bulged
9
14856 J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005

Hoogsteen-Type Triplexes and Duplexes
A R T I C L E S
intercalator in the solution. Moreover, high parallel triplex and
duplex stabilization together with destabilization of antiparallel
duplexes as described here for TINA have never been observed
hitherto for other intercalating systems covalently attached to
the oligodeoxynucleotides. In this context, TINA, when incor-
porated as multiple bulge insertions into oligodeoxynucleotides,
is a unique molecule with the ability to discriminate dsDNA
over ssDNA. This feature is clearly seen for ON14 when their
triplex and antiparallel duplex stabilities are compared at pH
6.0 and 7.2 (Table 1). This opens up the possibility of reducing
the number of false positives coming from duplex formation
when parallel triplex formation is to be detected. This could
for example be the case for fluorescence in situ hybridization
(FISH) on genomes under nondenaturing conditions21 and for
the purification of plasmid DNA using triple-helix affinity
chromatography22 or triple-helix affinity precipitation23 which
can be performed at pH 6.0 or 5.0. This type of discrimination
of parallel triplex formation over duplex formation cannot be
achieved with triplex forming oligos such as PNA, LNA, or
N3′fP5′ phosphoramidates which are also known to stabilize
antiparallel duplexes.
Using the Sonogoshira-type postsynthetic modification of
oligonucleotides possessing (R)-1-O-(4-iodophenyl)methyl-
glycerol, we screened several twisted intercalating nucleic acids
(TINAs) for their ability to increase the thermal stability of
Hoogsteen-paired duplexes and triplexes. The insertion of (R)-
1-O-[4-(1-pyrenylethynyl)phenylmethyl]glycerol (5) as a bulge
in oligodeoxynucleotides was found to be the most effective
TINA with good discriminating properties between matched and
mismatched sequences. The Watson-Crick-type DNA/DNA
and DNA/RNA duplexes were destabilized upon insertion of
TINA in the middle of the sequence compared with native
duplexes. We believe that TINA is the first intercalating system
covalently attached to oligodeoxynucleotides as a bulge showing
increased affinity toward Hoogsteen-type base-pairing and
decreased affinity toward Watson-Crick-type helixes. The short
synthetic route to phosphoramidite 8 and postsynthetic Sono-
gashira modification of oligonucleotides are competitive ad-
vantages of TINA over other triplex-stabilizing nucleic acids.
From studying double insertions of TINA (5) in one strand, it
could be concluded that placing of three nucleobases between
two bulged (R)-1-O-[4-(1-pyrenylethynyl)phenylmethyl]-
glycerols may be an optimum for high thermal stability of
Hoogsteen DNA helixes. On the other hand, the different
luminescence properties (excimer band formation) upon inser-
tion of 5 as next-nearest bulged neighbors in the pyrimidine
DNA sequence could be used to distinguish between parallel
triplex, parallel dsDNA, and antiparallel dsDNA. Increasing the
thermal stability in the range 12-19 °C for TINA with single
bulged insertion of 5 can be applied to reduce the required length
of the TFO. Moreover, good thermal stability for Hoogsteen-
type duplexes and triplexes could be obtained at pH 7.0 even
in the presence of several cytosines in the sequence (up to 36%
in the present work). The multiple insertions of 5 can be used
to increase the melting temperature of less stable Hoogsteen
duplexes to the level of Watson-Crick duplexes of the same
length under proper conditions (sequence, pH, salt concentration,
etc.). Considering the development of modified nucleic bases
with high affinity for C-G24 and T-A25 inversion sites in
dsDNA along with alternate-stranded triplexes,7e,26 we think such
improvements of triplex formation will expand the applicability
of TINA. The ability to stabilize the triplex upon insertion of 5
into the pyrimidine strands of circular oligodeoxynucleotides
or clamps to target ssDNA and ssRNA is also an obvious
possibility. As a next step, studies are devoted to the influence
of insertion of TINA and INA on the stability of nucleic acid
helixes different from the classical Watson-Crick and Hoogs-
teen complexes. Thus, there is still limited availability of nucleic
acid analogues which can stabilize reverse-Hoogsteen base-
pairing, i-motifs (C-C+ base-pairs), or quadruplexes (G-rich
sequences). We believe that the ability of TINA to stabilize
parallel triplexes and duplexes along with discrimination of
Hoogsteen-type over Watson-Crick-type nucleic acid helixes
can make TINA very useful in the design of DNA-based tools
in bio- and nanotechnology where specific recognition, high
thermal stability, and self-organization or reorganization are
vital.27
Experimental Section
General Methods. NMR spectra were recorded on a Varian Gemini
2000 spectrometer at 300 MHz for 1H and 75 MHz for 13C. The internal
1
standard used in H NMR spectra was TMS (δ: 0.00) for CDCl3; in
13C NMR the standard was CDCl3 (δ: 77.0). Accurate ion mass
determinations were performed using the 4.7 T Ultima Fourier transform
(FT) mass spectrometer (Ion Spec, Irvine, CA). The [M + Na]+ ions
were peak matched using ions derived from the 2,5-dihydroxybenzoic
acid matrix. Thin-layer chromatography (TLC) analyses were carried
out with use of TLC plates 60 F254 purchased from Merck and were
visualized in UV light (254 nm). The silica gel (0.040-0.063 mm)
used for column chromatography was purchased from Merck. Solvents
used for column chromatography were distilled prior to use, while
reagents were used as purchased.
Synthesis of Target Compound 8. (S)-1-(4,4′-Dimethoxytri-
phenylmethyloxy)-3-(4-iodo-benzyloxy)-propan-2-ol. (S)-(+)-2,2-Di-
methyl-1,3-dioxolane-4-methanol (6, 1.17 g, 8.9 mmol) and 4-iodo-
benzylbromide (2.5 g, 8.4 mmol) were refluxed under Dean-Stark
conditions in toluene (80 mL) in the presence of KOH (8.8 g, 154.0
mmol) for 12 h. The reaction mixture was allowed to cool, and H2O
(30 mL) was added. After separation of the phases, the water layer
was washed with toluene (2 × 15 mL). Combined organic layers were
washed with H2O (30 mL) and concentrated in vacuo. The residue was
treated with 80% aq AcOH (25 mL) for 48 h at room temperature.
(24) Huang, C.-Y.; Cushman, C. D.; Miller, P. S. J. Org. Chem. 1993, 58, 5048-
5049. (b) Prevot-Halter, I.; Leumann, C. J. Bioorg. Med. Chem. Lett. 1999,
9, 2657-2660. (c) Obika, S.; Hary, Y.; Sekiquchi, M.; Imanishi, T. Chem.
Eur. J. 2002, 8, 4796-4802. (d) Buchini, S.; Leumann, C. J. Angew. Chem.
2004, 116, 4015-4018; Angew. Chem., Int. Ed. 2004, 43, 3925-3928. (e)
Ranasinghe, R. T.; Rusling, D.; Powers, V. E. C.; Fox, K. R.; Brown, T.
Chem. Commun. 2005, 2555-2557.
(25) (a) Guianvarc’h, D.; Benhida, R.; Fourrey, J. L.; Maurisse, R.; Sun, J. S.
Chem. Commun. 2001, 1814-1815. (b) Guianvarc’h, D.; Fourrey, J. L.;
Maurisse, R.; Sun, J. S.; Benhida, R. Bioorg. Med. Chem. 2003, 11, 2751-
2759.
(26) (a) Ono, A.; Chen, C.-N.; Kan, L.-S. Biochemistry 1991, 30, 9914-9921.
(b) Froehler, B. C.; Terhorst, T.; Shaw, J.-P.; McCurdy, S. N. Biochemistry
1992, 31, 1603-1609. (c) de Bizemont, T.; Duval-Valentin, G.; Sun, J.-
S.; Bisagni, E.; Garestier, T.; Helene, C. Nucleic Acids Res. 1996, 24, 1136-
1143. (d) de Bizemont, T.; Sun, J.-S.; Garestier, T.; Helene, C. Chem. Biol.
1998, 5, 755-762. (e) Ueno, Y.; Mikawa, M.; Hoshika, S.; Matsuda, A.
Bioconjugate Chem. 2001, 12, 635-642.
(27) (a) Wengel, J. Org. Biomol. Chem. 2004, 2, 277-280. (b) Samori, B.;
Zuccheri, C. Angew. Chem. 2005, 117, 1190-1206; Angew. Chem., Int.
Ed. 2005, 44, 1166-1181.
(21) (a) Anguiano, A. J. Clin. Lig. Assay 2000, 23, 33-42. (b) Hausmann, M.;
Winkler, R.; Hildenbrand, G.; Finsterle, J.; Weisel, A.; Rapp, A.; Schmitt,
E.; Janz, S.; Cremer, C. BioTechniqes 2003, 35, 564-577.
(22) (a) Wils, P.; Escriou, V.; Warnery, A.; Lacroix, F.; Lagneaux, D.; Ollivier,
M.; Crouzet, J.; Mayaux, J.-F.; Scherman, D. Gene Ther. 1997, 4, 323-
330. (b) Schuelp, T.; Cooney, C. L. Nucl. Acids Res. 1998, 26, 4524-
4528.
(23) (a) Costioli, M. D.; Fisch, M. D.; Garret-Flaudy, F.; Hilbrig, F.; Freitag,
R. Biotechnol. Bioeng. 2003, 81, 535-545.
9
J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005 14857

A R T I C L E S
Filichev and Pedersen
The solvent was removed in vacuo, and the residue was coevaporated
twice with toluene/EtOH (30 mL, 5:1, v/v). The residue was dried under
diminished pressure to afford (R)-3-(4-iodo-benzyloxy)-propane-1,2-
diol (7, 100%, 2.3 g) as yellowish oil that was used in the next step
without further purification.
washed with the reaction mixture several times by syringes. After every
45 min, the last operation was repeated. After 3-4 h, the reaction
mixture was removed from the support, and columns were washed with
DMF (2 × 0.5 mL) and CH3CN (2 × 1 mL), and dried. In the cases
of ON12-14, ON21, and ON24, CPG-supports were treated one more
time with freshly prepared Sonogashira-coupling reaction mixture.
Afterward, the 5′-DMT-on oligonucleotides were cleaved off from the
solid support (room temperature, 2 h) and deprotected (55 °C, overnight)
using 32% aqueous ammonia. Purification of 5′-O-DMT-on TINAs was
accomplished using a reverse-phase semipreparative HPLC on a Waters
Xterra MS C18 column. The ODNs were DMT deprotected in 100 µL
10% aq acetic acid (30 min), diluted with 32% aqueous ammonia (1
mL), and purified again on HPLC. Corresponding fractions with ODNs
were evaporated and diluted with 1 M aq NaOAc (150 µL), and ODNs
were precipitated from ethanol (550 µL). The modified ODNs were
confirmed by MALDI-TOF analysis on a Voyager Elite biospectrometry
research station from PerSeptive Biosystems. The purity of the final
TFOs was checked by ion-exchange chromatography using a LaChrom
system from Merck Hitachi on a GenPak-Fax column (Waters).
Melting Temperature Measurements. Melting temperature mea-
surements were performed on a Perkin-Elmer UV-vis spectrometer
Lambda 35 fitted with a PTP-6 temperature programmer. The triplexes
were formed by first mixing the two strands of the Watson-Crick
duplex, each at a concentration of 1.0 µM in the corresponding buffer
solution. The solution was heated to 80 °C for 5 min and cooled to
room temperature, and the third (TFO) strand was added and then kept
at 15 °C for 30 min. The duplexes were formed by mixing the two
strands, each at a concentration of 1.0 µM in the corresponding buffer
solution followed by heating to 70 °C for 5 min and then cooling to
room temperature. The melting temperature (Tm, °C) was determined
as the maximum of the first derivative plots of the melting curves
obtained by measuring absorbance at 260 nm against increasing
temperature (1.0 °C per min). A lower speed of increasing the
temperature (0.5 °C per min) resulted in the same curves. Experiments
were also done at 373 nm. All melting temperatures are within the
uncertainty ( 0.5 °C as determined by repetitive experiments.
Fluorescence Measurements. Fluorescence measurements were
performed on a Perkin-Elmer luminescence spectrometer LS-55 fitted
with a Julabo F25 temperature controller. The triplexes and duplexes
were formed in the same way as for Tm measurements except that only
1.0 µM of TFOs were used in all cases. The spectra were recorded at
10 °C in the buffer 20 mM sodium cacodylate, 100 mM NaCl, and 10
mM MgCl2 at pH 6.0.
This oil (2.3 g, 8.4 mmol) was dissolved in anhydrous pyridine (25
mL), and 4,4′-dimethoxytrityl chloride (3.5 g, 10.4 mmol) was added
under nitrogen. After 24 h, MeOH (2 mL) followed by EtOAc (150
mL) were added, and the mixture was extracted with saturated aq
NaHCO3 (40 mL × 2). The water phase was extracted with EtOAc
(20 mL × 2). The combined organic layers were dried (Na2SO4),
filtered, and evaporated under diminished pressure. The residue was
coevaporated twice with toluene/EtOH (25 mL, 1:1, v/v). The residue
was adsorbed on silica gel (3.0 g) from EtOAc (30 mL) and purified
using dry column vacuum chromatography with EtOAc (0-30%, v/v)
in petroleum ether to afford compound (S)-1-(4,4′-dimethoxytriphenyl-
methyloxy)-3-(4-iodo-benzyloxy)-propane-2-ol (70%, 3.6 g) as a yellow
foam. 1H NMR (CDCl3) δ 2.42 (br s, 1H), 3.20 (m, 2H), 3.56 (m, 2H),
3.78 (s, 6H), 3.97 (m, 1H), 4.43 (s, 2H), 6.78 (d, 4H, J ) 8.5 Hz),
7.00 (d, 2H, J ) 8.0 Hz), 7.30-7.45 (m, 9H), 7.63 (d, 2H, J ) 8.0
Hz). 13C NMR (CDCl3) δ 55.2, 62.2, 69.9, 71.6, 72.6, 86.1, 93.1, 129.4,
137.4, 137.7, 113.1, 126.7, 127.8, 128.1, 130.0, 135.9, 144.7, 158.5.
HR-MALDI-MS calcd for C31H31IO5Na [M + Na]+ m/z 633.1108;
found m/z 633.1116.
(S)-2-O-[2-Cyanoethoxy(diisopropylamino)phosphino]-1-O-(4,4′-
dimethoxy triphenylmethyl)-3-O-(4-iodo-benzyl)glycerol (8). (S)-1-
(4,4′-Dimethoxytriphenylmethyloxy)-3-(4-iodo-benzyloxy)-propane-2-
ol (2.0 g, 3.3 mmol) was dissolved under nitrogen in anhydrous CH2Cl2
(50 mL). N,N-Diisopropylammonium tetrazolide (0.850 g, 5.0 mmol)
was added followed by dropwise addition of 2-cyanoethyl tetraisopro-
pylphosphordiamidite (1.1 g, 3.7 mmol) under external cooling with
an ice-water bath. After 16 h, analytical TLC showed no more starting
material, and the reaction was quenched with H2O (30 mL). Layers
were separated, and the organic phase was washed with H2O (30 mL).
Combined water layers were washed with CH2Cl2 (25 mL). The organic
phase was dried (Na2SO4) and filtered, silica gel (1.5 g) and pyridine
(0.5 mL) were added, and solvents were removed under reduced
pressure. The residue was purified using silica gel dry column vacuum
chromatography with NEt3(0.5%, v/v)/EtOAc(0-25%,)/petroleum ether.
Combined UV-active fractions were evaporated in vacuo affording the
final compound 8 (1.8 g, 67%) as a foam that was used in ODN
synthesis. 32P NMR (CDCl3) δ 149.8, 149.9 in ratio 1:1. HR-ESI-MS
calcd for C40H46IO6N2PLi [M + Li]+ m/z 817.2449, found m/z 817.2447.
Synthesis and Purification of TINAs. ODNs were synthesized on
an Expedite nucleic acid synthesis system model 8909 from Applied
Biosystems using 4,5-dicyanoimidazole as an activator and an increased
deprotection time (100 s) and coupling time (2 min) for 0.075 M
solution of the phosphoramidite 8 in a 1:1 mixture of dry MeCN/CH2-
Cl2. After the DNA synthesis, the columns with CPG-supports and
DMT-on oligonucleotides possessing 4-iodophenyl moieties were
flushed with argon (2 min) prior to the coupling reaction. Sonogashira-
coupling reagent mixture containing Pd(PPh3)4 or Pd(PPh3)2Cl2 (7.5
mM), an aromatic structure possessing a terminal acetylene (22.5 mM),
and CuI (7.5 mM) in dry DMF/Et3N (3.5/1.5, 500 µL) was prepared in
a 1 mL plastic syringe under dry conditions at room temperature.
Syringes were also flushed with argon prior to use. The syringe with
Sonogashira-coupling reagent mixture was attached to the column with
the CPG, and another empty syringe was connected from another side
of the column. The CPG-support with modified oligonucleotide was
Acknowledgment. The Nucleic Acid Center is funded by The
Danish National Research Foundation for studies on nucleic acid
chemical biology.
Note Added in Proof. TINA oligonucleotides and their
monomeric phosphoramidites have been made commerical
available from University of Southern Denmark after manu-
script submission. See http://www.sdu.dk/Nat/Chem/INFO/
DNAmonomers.pdf.
Supporting Information Available: MALDI-TOF MS and
reverse-phase (DMT-on) and ion-exchange (DMT-off) HPLC
analysis of oligonucleotides synthesized. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA053645D
9
14858 J. AM. CHEM. SOC. VOL. 127, NO. 42, 2005