2'-O,4'-C-Methylene bridged nucleic acid (2',4'-BNA): synthesis and triplex-forming properties.

Article abstract of DOI:10.1016/S0968-0896(00)00325-4

For development of ideal antisense and antigene molecules, various chemical modifications of oligonucleotides have been studied. However, despite their importance, there is only limited information available on the triplex-forming ability of the conformat

Full text of DOI:10.1016/S0968-0896(00)00325-4

Bioorganic & Medicinal Chemistry 9 (2001) 1001±1011
20-O,40-C-Methylene Bridged Nucleic Acid (20,40-BNA):
Synthesis and Triplex-Forming Properties1
Satoshi Obika, Takeshi Uneda, Tomomi Sugimoto, Daishu Nanbu, Takashi Minami,
Takefumi Doi and Takeshi Imanishi*
Graduate School of Pharmaceutical Sciences, Osaka University, 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan
Received 3 October 2000; accepted 21 November 2000
AbstractÐFor development of ideal antisense and antigene molecules, various chemical modi®cations of oligonucleotides have been
studied. However, despite their importance, there is only limited information available on the triplex-forming ability of the con-
formationally restricted or locked oligonucleotides. We report herein that 20-O,40-C-methylene bridged nucleic acid (20,40-BNA)
modi®cation of triplex-forming oligonucleotide (TFO) signi®cantly enhances the binding anity towards target dsDNA. On Tm
measurements, the triplex with the 20,40-BNA oligonucleotides were found to be stabilized with ÁTm/modi®cation of +4.3 to
+5 ꢀC at pH 6.6 compared to the triplexes with the unmodi®ed oligonucleotide. By means of gel-retardation assay, the binding
constant of the 20,40-BNA oligonucleotide at pH 7.0 was at least 300-fold higher than that of the natural oligonucleotide. In addi-
tion, the 20,40-BNA oligonucleotide clearly showed the inhibition of the NF-kB transcription factor (p50)-target dsDNA binding by
forming a stable triplex at pH 7.0. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
conformation in a suitable form would be quite e€ective
for triplex-formation, as well as binding with com-
plementary ssDNA or ssRNA.7 Recently, we have
achieved the ®rst synthesis and development of a novel
class of modi®cations of nucleic acids, bridged nucleic
acid (BNA), such as 20-O,40-C-methylene BNA (20,40-
BNA; Fig. 1)1,8À15 and 30-O,40-C-methylene BNA (30,40-
BNA).15À18 The 20,40-BNA oligonucleotides were found
to have a potent hybridizing ability towards com-
plementary nucleic acids, especially towards com-
plementary RNA.1,9,19 Furthermore, it was also found
that only one 20,40-BNA modi®cation of TFO promoted
the marked triplex stabilization in a highly sequence-
selective manner.13 Now, we describe the triplex-form-
ing ability of the fully or partially 20,40-BNA modi®ed
homopyrimidine TFO and inhibition of NF-kB (p50)±
DNA binding by the triplex formation.
Homopyrimidine oligodeoxyribonucleotides can bind to
homopurine sequences in dsDNA via Hoogsteen
hydrogen bonding to form triplex DNA.2 In recent
years, interest in the possibility of regulation of gene
expression by triplex-forming oligonucleotides (TFOs)
has grown because of their widely usable application as
molecular biological tools and as a direct way to treat
serious diseases such as cancer and genetic disorders. In
addition, the structure of these triplexes has also been
attracting widespread attention. Since the binding an-
ity of the natural pyrimidine oligonucleotides towards
dsDNA is relatively low under physiological conditions,
extensive e€orts have been directed towards developing
novel DNA and RNA analogues for the practical use of
the antigene technologies and for clari®cation of the
triplex structure.3 However, only a few modi®ed
TFOs3À5 with sucient binding anity under physio-
logical conditions have been reported, and the overall
conformations of triplexes and the structural require-
ments for ideal TFOs are not well understood.6 It
occurred to us that restriction of the oligonucleotide
Results
Synthesis of 20,40-BNA monomers and their oligonucleotide
derivatives
As shown in Scheme 1, 20,40-BNA monomers, 20-O,40-C-
methyleneuridine (1U) and its 5-methyl congener 1T,
were synthesized e€ectively from 1,2-O-cyclohexylidene-
*Corresponding author. Tel.: +81-6-6879-8299; fax: +81-6-6879-
8204; e-mail: imanishi@phs.osaka-u.ac.jp
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00325-4

1002
S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
obtained by dimethoxytritylation and subsequent phos-
phitylation of 1T and 1U, respectively. According to
Xu's procedure,26 the amidites 9T and 9U were trans-
formed into 4-triazolo derivatives 10 and 11, which were
easily converted into the cytidine derivatives by treat-
ment with concd ammonia after oligonucleotide synth-
esis. Oligonucleotide analogues 12±20 containing 20-
O,40-C-methyleneribonucleosides were e€ectively pre-
pared by standard phosphoramidite protocol on a DNA
synthesizer. Crude oligonucleotides were puri®ed by
RP-HPLC and the structure of the prepared oligonu-
cleotides was con®rmed by MALDI-TOF MS analysis.
Figure 1. Structure of 20,40-BNA monomers, 20-O,40-C-methylene-
ribonucleosides (1).
4-hydroxymethyluridine (2).20,21 Treatment of 2 with
p-toluenesulfonyl chloride in pyridine gave the mono-
tosylated compound 3, which was in turn deprotected
under acid conditions to give the triol 4. Without the
protection of the 30-hydroxy group, the oxetane com-
pound was obtained by the ring-closure reaction under
alkaline conditions.16 Therefore, the triol 4 was initially
treated with benzaldehyde in the presence of zinc chlo-
ride to give 20,30-O-benzylidene derivative 5 as an endo-
isomer,22 and then 5 was reduced with sodium cyano-
borohydride and titanium tetrachloride to a€ord the
desired 30-O-benzyl derivative 6. Ring-closure reaction
of 6 was smoothly achieved by using sodium hexa-
methyldisilazide in tetrahydrofuran, and the reductive
debenzylation of 7 gave 20-O,40-C-methyleneuridine
(1U). Furthermore, 1U was e€ectively converted to its
Triplex formation of 20,40-BNA oligonucleotides
At ®rst, the triplex-forming property of the 14-mer 20,40-
BNA oligonucleotides 12±20 towards the 30bp target
duplex was studied by an analysis of the UV melting
curve, as shown in Table 1. The target 30bp DNA is an
upstream region of a mouse inducible NO synthase
(miNOS) gene,27 containing a part of the NF-kB bind-
ing sequence (Fig. 2). Absorbance at 260nm was mea-
sured versus temperature. The triplex was formed by
mixing both strands of the duplex (2 mM) with 2 mM
TFO in a 10mM sodium phosphate bu€er (pH 6.6 or
7.2) containing 100 mM NaCl. The natural TFO 21 was
not able to bind to the target duplex at both pH 6.6 and
7.2. The TFO 22 in which the cytosine base is replaced
by 5-methylcytosine that is well known to enhance the
23
5-methyl congener 1T via 5-bromination and sub-
sequent Pd-catalyzed cross-coupling reaction with
trimethylaluminum.25 The phosphoramidites 9T and 9U,
the suitable building block for DNA synthesis, were
Scheme 1. (a) TsCl, pyridine, 110 ꢀC, 69%; (b) CF3CO2H±H2O (98:2), 20 ꢀC, 94%; (c) PhCHO, ZnCl2, 20 ꢀC, 80%; (d) NaBH3CN, TiCl4, MeCN,
20 ꢀC, 75%; (e) sodium bis(trimethylsilyl)amide, THF, 20 ꢀC, 61%; (f) H2 (1 atm), 10%Pd-C, MeOH, 20 ꢀC, quant; (g) NBS, NaN3, 1,2-dimethoxy-
ethane, 20 ꢀC, 84%; (h) (i) (NH4)2SO4, 1,1,1,3,3,3-hexamethyldisilazane, re¯ux; (ii) PdCl2, Ph3P, Me3Al, THF, re¯ux; (iii) NH4Cl, MeOH±H2O,
re¯ux, 57%; (i) DMTrCl, DMAP, pyridine, 20 ꢀC, 95% (8T), 94% (8U); (j) 2-cyanoethyl N,N,N0,N0-tetraisopropylphosphoramidite, diisopropyl-
ammonium tetrazolide, MeCN-THF, 20 ꢀC, 99% (9T), 81% (9U); (k) 1,2,4-triazole, POCl3, Et3N, MeCN, 0 ꢀC, 91% (10), 89% (11).

S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
1003
the triplex-forming ability accurately at lower pH by Tm
measurements. Thus, DNase I footprinting was carried
out on the 521 bp duplex containing the target 30bp
sequence (Fig. 3). At both pH 5.5 and 6.0, DNA clea-
vage by DNase I was perfectly inhibited in a sequence-
dependent manner by addition of the natural TFOs 21,
22 or the partially 20,40-BNA modi®ed TFOs 12±18.
However, it was not suciently inhibited by addition of
the fully 20,40-BNA modi®ed TFO 20. At pH 6.6, the
target sequence was also protected from DNase I clea-
vage by addition of 22 or 12±18, while the natural TFO
21 and the fully 20,40-BNA modi®ed TFO 20 was not
able to inhibit the cleavage. Furthermore, the partially
20,40-BNA modi®ed TFOs 13, 15 and 17 formed a tri-
plex towards the target site and protected from DNase I
digestion even at pH 7.2. These results from DNase I
footprinting experiments coincide with those of Tm
measurements.
Figure 2. DNA sequence of the target 30bp duplex used in this study.
The bold line indicates the site of triplex formation. The boxed
sequence is the NF-kB binding site.
triplex stability under neutral conditions,28 was able to
form the triplex at pH 6.6. On the other hand, the par-
tially 20,40-BNA modi®ed TFOs 12±18 were shown to
have signi®cantly enhanced triplex-forming ability. In
particular, the 20,40-BNA modi®ed TFOs 13, 15 and 17
containing a 5-methylcytosine moiety, showed remark-
able stabilizing e€ect of the triplex with ÁTm/modi®ca-
tion ranging from +4.3 to +5 ꢀC compared to 22 at pH
6.6. Unexpectedly, the fully 20,40-BNA modi®ed (except
for the 30-end) TFOs 19 and 20 exhibited no triplex-
forming ability at both pH 6.6 and 7.2.
Next, the binding anity of the partially 20,40-BNA
modi®ed TFO 17 towards the target duplex was also
elucidated by using gel-retardation assay (Fig. 4). Both
strands of the target duplex were 50-end labeled and
incubated with the indicated concentration of 17 or 22
Because of the overlapping of the transition curves of
duplex and triplex dissociation, it was hard to evaluate
Figure 3. DNase I footprinting assay demonstrating the binding of the 20,40-BNA modi®ed (12±18, 20) and natural (21, 22) TFOs to the oligopurine
target site AGAGAGAGGGAAAA. A 521 bp 32P-labeled fragment (0.2 nM) was treated with indicated oligonucleotides (2 mM) at 18 ꢀC for 48 h in
a 50mM sodium acetate bu€er (pH 5.5) or in a 50mM sodium phosphate bu€er (pH 6.0, 6.6 or 7.2) containing 10mM magnesium chloride, and
then digested by DNase I. Lane C (control) lacking an oligonucleotide is also shown. Lane M contains markers of the indicated sizes in bases.

1004
S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
in 50mM HEPES bu€er (pH 7.0) containing 20mM
MgCl2 and then loaded onto a non-denaturing gel.
Upon addition of 17 to a 2Â10À10 M solution of the 30
bp target duplex, triplex formation was detected with a
midpoint near 3Â10À8 M at pH 7.0(Fig. 4a). The addi-
tion of 17 at 1Â10À7 M ®nal concentration led to the
complete shift of the target duplex (DS) to a triplex
band (TS), and no other shifted bands were detected in
these conditions. On the other hand, addition of TFO
22, a 5-methylcytosine analogue of 21, showed no tri-
plex-formation even at 3Â10À6 M (Fig. 4b). The data in
Fig. 4a have been quanti®ed and are plotted against the
concentration of the triplex-forming oligonucleotides
(Fig. 4c), where the dissociation constant (Kd) of 17 was
estimated to be 2.5Â10À8 M at pH 7.0,29 which was at
least 300-fold lower than that of 22. Considering the
relatively high Gꢁ C base pair contents of the target
sequence and binding conditions (without a polyamine
such as spermine), the 20,40-BNA oligonucleotides
appears to be one of the most potent triplex-forming
oligonucleotides via Hoogsteen base pairing.
Table 1. Tm values (ꢀC) of the triplexes consisting of the target 30-bp
duplex and TFOs 10±20a
Oligonucleotidesb
pH
6.6
7.2
50-d(TCTCTCTCCCTTT)-30
50-d(TmCTmCTmCTmCmCmCTTT)-30
5-0d(TCTCTCTCCCTTTT)-30
50-d(TmCTmCTmCTmCmCmCTTTT)-30
50-d(TCTCTCTCCCTTTT)-30
50-d(TmCTmCTmCTmCmCmCTTTT)-30
50-d(TCTCTCTCCCTTTT)-30
50-d(TmCTmCTmCTmCmCmCTTTT)-30
50-d(TCTCTCTCCCTTTT)-30
50-d(TCTCTCTCCCTTTT)-30
(21)
(22)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
Ðc (64)
28 (63)
39 (64)
53 (64)
41 (64)
54 (64)
41 (64)
55 (64)
48 (63)
Ðc (64)
Ðc (64)
Ðc (64)
Ðc (63)
23 (63)
36 (65)
Ðc (64)
37 (64)
Ðc (63)
33 (64)
28 (63)
Ðc (63)
Ðc (63)
50-d(TmCTmCTmCTmCmCmCTTT)-30
aThe melting temperatures (Tm values) were obtained as the maxima of the
®rst derivatives of the melting curves (A260 versus temperature) using 2 mM
concentration of the triplex in a bu€er containing 10mM sodium phos-
phate (pH 6.6 or 7.2) and 100 mM sodium chloride. The values in the par-
entheses are Tm of the target duplex.
bT, thymidine, C, 20-deoxycytidine, mC, 20-deoxy-5-methylcytidine.
cObvious thermal hypochromicity was not observed.
Figure 4. Gel-retardation experiments showing the binding of the
TFOs to the 30-bp target duplex. The indicated concentration (M) of
the TFOs 17 (panel a) and 22 (panel b) was incubated with a 32P-
labeled target fragment (ꢂ0.2 nM), 50-d(GAACAAACAGAGA-
GAGGGAAAATCCCCCTA)/30 -d(CTTGTTTGTCTCTCTCCCTT-
TTAGGGGGAT), at 4 ꢀC for 48 h in 50mM HEPES bu€er (pH 7.0)
containing 20mM magnesium chloride. TS and DS indicate mobility
of the triplex and duplex, respectively. Panel c shows the titration
curve of TFO 17. The ratio of triplex band (TS) relative to total DNA
(TS+DS) is plotted against the concentration of TFO 17.
Figure 5. Gel-retardation experiments showing the inhibition of the
NF-kB (p50) binding toward the 30-bp target duplex by addition of
TFO 17. The indicated concentration of 17 was incubated with a 32P-
labeled target fragment (ꢂ1 nM), 50-d(GAACAAACAGAGAGAGG-
GAAAATCCCCCTA)/30 - d(CTTGTTTGTCTCTCTCCCTTTTAGG-
GGGAT), and p50(1.2 gel shift units) at 4 ꢀC for 48 h in 9 mM
HEPES bu€er (pH 7.0) containing 45 mM KCl, 0.2 mM EDTA, 9 mM
MgCl2, 2 mM DTT, 9% glycerol, 0.05% NP40, 0.05 mg/mL poly (dI-
dC)ꢁ poly (dI-dC).

S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
1005
Inhibition of NF-ꢀB (p50) binding to target sequence by
triplex formation
prevented for the most part, and the complete inhibition
was observed at 1Â10À5 M.
The 30bp target duplex used in the above experiments
contains an NF-kB binding sequence, and the NF-kB
binding site overlaps the TFO binding site in part (Fig.
2). To elucidate the inhibition of NF-kB-DNA associa-
tion by the 20,40-BNA modi®ed TFO 17, gel retardation
experiments were examined in the presence of a subunit
Discussion
The ideal sugar conformation of TFOs has been dis-
cussed by some groups to date.4,31,32 Gryaznov and his
co-workers reported that the N30±P50 phosphoramidate
oligonucleotides signi®cantly enhance triplex forma-
tion.4 By means of X-ray crystallography of the duplex
consisting of the self-complementary phosphoramidate
oligonucleotide, the phosphoramidate oligonucleotides
were found to prefer existing in A-form (N-form sugar
puckering).33 Recently, Asensio also reported that the
20-OMe oligonucleotide was able to form a stable triplex
with dsDNA where the 20-OMe ribonucleoside deriva-
tives exist in N-conformation predominantly.31
Although these data strongly suggested that the N-form
sugar puckering is suitable for triplex formation, the
investigation by using the oligonucleotides with dis-
of NF-kB, p50, which contains a DNA binding motif
30
and can bind to DNA as a p50dimer.
The labeled
target duplex (1 nM) and p50were incubated with the
indicated concentration of 17 at pH 7.0. In the absence
of 17, the shifted band coming from the p50-target
duplex complex was observed (Fig. 5). No change was
detected in the gel-shift pattern by the addition of 17 at
1Â10À8 M ®nal concentration, while 1Â10À7 M of 17 led
to the formation of a triplex. Further addition of 17
showed inhibition of the p50binding to the target duplex
along with the formation of a triplex. At 1Â10À6 M con-
centration, the formation of a p50±duplex complex was
Figure 6. Comparison of the CD spectra of the fully 20,40-BNA modi®ed oligonucleotide 20 (solid line), partially modi®ed oligonucleotide 17 (dotted
line) and natural oligonucleotide 22 (dashed line) in 10mM sodium phosphate bu€er (pH 7.2) containing 100mM NaCl.

1006
S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
tinctly locked sugar conformation was of great impor-
tance to clarify the relationship between the triplex sta-
bility and the sugar conformation in TFO. The triplex
formation of the conformationally restricted oligo-
nucleotides, in which the sugar conformation was
restricted in S-form, being in opposite conformation to
our oligonucleotides, was reported by Leumann et al.34-
the transcriptional activation mediated by NF-kB in
living cells.
Conclusion
The partailly 20,40-BNA modi®ed TFOs form a very
stable triplex with the target dsDNA in a sequence-spe-
ci®c manner. These hybridization characteristics are
dependent on pH; however, the partially modi®ed TFOs
having 5-methylcytosine moieties show triplex±forma-
tion even under neutral conditions. Furthermore, the
20,40-BNA modi®ed TFO e€ectively prevented the NF-
kB (p50)-dsDNA binding by the triplex±formation at
pH 7.0. On the other hand, the fully modi®ed TFOs lost
their binding anity towards the target dsDNA.
Although further evaluation of the triplex-forming
ability of the fully modi®ed TFO is needed, these results
reveal that the preorganization of oligonucleotides is
quite e€ective for stable triplex±formation and that the
20,40-BNA oligonucleotide is a promising candidate for
a novel and practical antigene molecule, as well as an
antisense molecule.9,19,38
36
Their oligonucleotide analogues with S-form sugar
puckering, `bicyclo-DNA' and `tricyclo-DNA', can
e€ectively bind to complementary RNA;34À36 however,
the stabilization e€ect of triplex formation by these oli-
gonucleotide analogues was moderate (the maximum
ÁTm/modi®cation=+0.9 ꢀC).35 On the contrary, our
data represented here clearly reveal that the 20,40-BNA
oligonucleotides with N-form sugar puckering have sig-
ni®cantly high binding anity towards the target
duplex. On the DNA melting experiments, the partailly
20,40-BNA modi®ed TFOs showed remarkable increase
in Tm value (the maximum ÁTm/modi®cation=+5 ꢀC),
and on the gel retardation experiments, the association
constant of 17 in triplex formation was more than 300-
fold higher comparing to that of natural TFO 22. From
the overall viewpoint, it is suggested that the N-form
sugar puckering is suitable for the pyrimidine motif tri-
plex formation.
Experimental
Unexpectedly, lack of triplex formation by the fully
20,40-BNA modi®ed oligonucleotides 19 and 20 was
observed (Table 1 and Fig. 3). To evaluate these unex-
pected results in detail, we recorded the CD spectra of
natural oligonucleotide 22, partially 20,40-BNA modi®ed
oligonucleotide 17 and fully modi®ed oligonucleotide 20
at several temperatures (Fig. 6). At 20 ꢀC, the 20,40-BNA
oligonucleotides 17 and 20 showed a strong positive
band at around 280nm, compared to the natural oligo-
nucleotide 22, indicating that the modi®ed oligonucleo-
General considerations
All melting points were measured on a Yanagimoto
micro melting point apparatus and are uncorrected. 1H,
13C and 31P NMR spectra were recorded on a Varian
VXR-200 (1H, 200 MHz; 31P, 86.4 MHz), JEOL EX-270
(1H, 270MHz; 13C, 67.5 MHz) or JEOL GX-500 (1H,
500 MHz) spectrometer. IR spectra were recorded on a
JASCO FT/IR-200 spectrometer. Mass spectra were
measured on JEOL JMS-D300 or JMS-600 mass spec-
trometer. Optical rotations were recorded on a JASCO
DIP-370instrument. For column chromatography,
Merck Kieselgel 60 (70±200 mesh) or Fuji Silysia BW-
127ZH (100±200 mesh) was used.
tides have
a relatively ordered structure by the
conformational restriction. On raising the temperature
to 80 ꢀC, the partially modi®ed oligonucleotides exhib-
ited a remarkable decrease in intensity at around
280nm. However, the fully modi®ed oligonucleotide 20
preserved the strong positive band at around 280nm
even at 80 ꢀC. These results suggest that the fully mod-
i®ed oligonucleotide does not cause the conformational
change not only in the sugar puckering but also in the
phosphodiester backbone. Although other evaluations,
such as NMR experiments, are of course needed, we
think the lack of triplex formation by the fully 20,40-
BNA modi®ed oligonucleotide 22 would appear to
come from its too rigid structure to form a triplex for
the present.
20,30-O-Cyclohexylidene-40-( p-toluenesulfonyloxymethyl)-
uridine (3). Under N2 atmosphere, p-toluenesulfonyl
chloride (137 mg, 0.721 mmol) was added to a stirred
solution of 220,21 (150mg, 0.424 mmol) in pyridine
(3 mL) at 110 ꢀC. After having been stirred for 3 h at the
same temperature, the reaction mixture was partitioned
with saturated NaHCO3 and the mixture was extracted
with CHCl3. Usual work up and puri®cation by SiO2
column chromatography (CHCl3:MeOH, 15:1) a€orded
3 (149 mg, 0.293 mmol, 69%) as a white powder, mp
19
D
Although the overlapping of the TFO binding site and
the NF-kB consensus sequence is only 6 bp, the 20,40-
BNA modi®ed TFO 17 clearly showed the inhibition of
p50binding to the target duplex at 1 Â10À6 M con-
centration (Fig. 5). Natural TFO was reported to inhibit
the NF-kB speci®c transcriptional activation in living
cells by triplex-formation in spite of its low anity
towards the target duplex (Kd>1Â10À5 M).37 Con-
sidering the high stability of the triplex consisting of the
target duplex and 17, it is thought that the 20,40-BNA
oligonucleotide 17 would be a promising inhibitor for
104±106 ꢀC (benzene/hexane), ‰ꢀŠ À18.5 (c 0.98, ace-
tone). IR n (KBr): 3326, 2929, 2850, 1628, 1577, 1544,
1437, 1311, 1244 cmÀ1. H NMR (acetone-d6): d 1.45±
1
1.67 (10H, m), 2.45 (3H, s), 3.71 (2H, ABq, J=12 Hz),
4.20(2H, ABq, J=11 Hz), 4.92 (1H, d, J=6 Hz), 5.05,
5.06 (1H, dd, J=4, 6 Hz), 5.60(1H, d, J=7 Hz), 5.75
(1H, d, J=4 Hz), 7.48 (2H, d, J=8 Hz), 7.77 (1H, d,
J=8 Hz), 7.81 (2H, d, J=8 Hz), 10.10 (1H, s). 13C
NMR (acetone-d6): d 21.5, 24.1, 24.5, 25.5, 34.8, 36.9,
63.5, 69.7, 82.5, 84.7, 87.8, 92.9, 102.9, 115.4, 128.8,
130.8, 133.9, 142.7, 145.9, 151.3, 163.5. MS (EI): m/z

S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
1007
508 (M+, 81.4), 465 (100). Anal. calcd for
C23H28N2O9Sꢁ 1/3 H2O: C, 53.69; H, 5.61; N, 5.44; S,
6.22. Found: C, 53.99; H, 5.48; N, 5.42; S, 6.10.
4.56 (2H, m), 4.69 (1H, d, J=11 Hz), 5.52 (1H, d,
J=6 Hz), 5.67 (1H, d, J=8 Hz), 7.24±7.29 (7H, m), 7.48
(1H, d, J=8 Hz), 7.70(2H, d, J=9 Hz), 9.91 (1H, s).
13C NMR (CDCl3): d 21.6, 63.2, 69.2, 73.6, 74.6, 78.1,
86.6, 92.9, 102.5, 127.9, 128.2, 128.3, 128.6, 129.9, 132.3,
136.9, 142.4, 145.2, 150.7, 163.8. MS (FAB): m/z 519
(M++H). Anal. calcd for C24H26N2O9S: C, 55.59; H,
5.05; N, 5.40; S, 6.18. Found: C, 55.41; H, 5.02; N, 5.32;
S, 6.15.
40-( p-Toluenesulfonyloxymethyl)uridine (4). Compound
3 (107 mg, 0.21 mmol) was treated with TFA/H2O (98:2,
ꢀ
1 mL) for 10min at 20 C. The reaction mixture was
rapidly evaporated in vacuo and the residue was coeva-
porated several times with EtOH. Puri®cation by SiO2
column chromatography (CHCl3/MeOH, 10:1) a€orded
4 (85.0mg, 0.20mmol, 94%) as a white powder, mp
30-O-Benzyl-20-O,40-C-methyleneuridine (7). Under N2
atmosphere, sodium bis(trimethylsilyl)amide (1 M in
THF, 6.38 mL, 6.38 mmol) was added to a solution of 6
(1.00 g, 1.93 mmol) in THF (10 mL) at 20 ꢀC. After
having been stirred for 3 h at the same temperature, the
reaction mixture was partitioned with saturated
NaHCO3 and the mixture was extracted with CHCl3.
Usual work-up, puri®cation by SiO2 column chromato-
graphy (CHCl3/MeOH, 10:1) and recrystallization from
20
119±120 ꢀC, ‰ꢀŠ À24.4 (c 0.50, acetone). IR n (KBr):
D
3227, 3060, 2932, 2837, 1709, 1508, 1464, 1252 cmÀ1. 1H
NMR (acetone-d6): d 2.31 (3H, s), 2.84 (3H, s), 3.71
(2H, s), 4.13, 4.20(2H, ABq, J=11 Hz), 4.28, 4.31 (1H,
dd, J=9, 6 Hz), 4.36 (1H, d, J=6 Hz), 5.54 (1H, d,
J=8 Hz), 5.75 (1H, d, J=7 Hz), 7.32 (2H, d, J=8 Hz),
7.67 (2H, d, J=8 Hz), 7.70(1H, d, J=8 Hz), 10.14 (1H,
s). 13C NMR (acetone-d6): d 21.5, 63.7, 70.8, 72.7, 74.6,
86.8, 88.8, 103.1, 128.8, 130.7, 133.9, 141.7, 145.8, 151.8,
163.9. MS (EI): m/z 256 (M+ÀTsOH, 9.8), 172 (100).
Anal. calcd for C17H20N2O9Sꢁ 4/5 H2O: C, 46.11; H,
4.92; N, 6.33; S, 7.49. Found: C, 46.19; H, 4.89; N, 6.11;
S, 7.20.
MeOH a€orded 7 (411 mg, 1.1 mmol, 61%) as colorless
23
D
crystals, mp 217±219 ꢀC (MeOH), ‰ꢀŠ +108.4 (c 0.3,
MeOH). IR n (KBr): 3059, 2951, 1688, 1459, 1271,
1053 cmÀ1. 1H NMR (DMSO-d6) d: 3.75, 3.85 (2H, AB,
J=8 Hz), 3.77 (2H, d, J=5 Hz), 3.92 (1H, s), 4.44 (1H,
s), 4.60(2H, s), 5.39 (1H, t, J=5 Hz), 5.48 (1H, s), 7.31
(5H, m), 7.72 (1H, d, J=8 Hz), 11.37 (1H, s). 13C NMR
(DMSO-d6): d 56.0, 71.1, 71.6, 75.8, 76.5, 86.5, 88.3,
100.9, 127.4, 127.6, 128.2, 137.9, 139.0, 150.0, 163.3. MS
(EI): m/z 346 (M+, 1.1), 91 (100). Anal. calcd for
C17H18N2O6: C, 58.96; H, 5.24; N, 8.09. Found: C,
58.67; H, 5.23; N, 8.05.
20,30 -O-Benzylidene-40 -( p-toluenesulfonyloxymethyl)uri-
dine (5). Under N2 atmosphere, a mixture of 4 (400 mg,
0.93 mmol), ZnCl2 (670mg, 5.0mmol) and benzalde-
hyde (2.4 mL) was stirred at room temperature for 5 h.
The reaction mixture was partitioned with saturated
NaHCO3 and the mixture was extracted with CHCl3.
Usual work up and puri®cation by SiO2 column chro-
matography (CHCl3/MeOH, 40:1) a€orded 5 (380mg,
20-O,40-C-Methyleneuridine (1U). Under H2 atmosphere,
10% Pd-C (25 mg) was added to a stirred solution of 7
(25 mg, 0.072 mmol) in MeOH (2.5 mL) at 20 ꢀC. After
having been stirred for 15 h, the reaction mixture was
®ltered, and the ®ltrate was concentrated. Puri®cation
by SiO2 column chromatography (CHCl3/MeOH, 10:1
0.74 mmol, 80%) as a white powder, mp 99±102 ꢀC
23
D
(CH2Cl2/hexane). ‰ꢀŠ À26.7 (c 1.0, CHCl3). IR n
(KBr): 3059, 1691, 1460, 1362, 1269, 1218, 1177 cmÀ1
.
1H NMR (CDCl3) d: 2.41 (3H, s), 3.25 (1H, br), 3.79
(2H, m), 4.19 (2H, s), 5.09 (1H, d, J=7 Hz), 5.28 (1H,
dd, J=4, 7 Hz), 5.60(1H, d, J=4 Hz), 5.73 (1H, d,
J=8 Hz), 5.94 (1H, s), 7.24 (1H, d, J=8 Hz), 7.38 (2H,
d, J=9 Hz), 7.42 (5H, br), 7.69 (2H, d, J=9 Hz), 9.11
(1H, br). 13C NMR (CDCl3): d 21.6, 63.5, 68.3, 77.2,
82.8, 84.2, 87.7, 94.9, 102.6, 107.5, 126.5, 127.9, 128.5,
129.7, 132.2, 135.0, 143.0, 145.0, 150.4, 163.5. MS (EI):
m/z 516 (M+, 28.7), 91(100). Anal. calcd for
C24H24N2O9Sꢁ 1/3 H2O: C, 55.17; H, 4.76; N, 5.36; S,
6.14. Found: C, 55.19;H, 4.66; N, 5.29; S, 5.98.
to 5:1) a€orded 1U (18.3 mg, quant.) as colorless crys-
23
D
tals, mp 239±243 ꢀC (MeOH). ‰ꢀŠ +92.2 (c 0.3,
MeOH). IR n (KBr): 3331, 3091, 3059, 2961, 1689,
1463, 1272, 1049 cmÀ1. 1H NMR (CD3OD) d: 3.76, 3.96
(2H, AB, J=8 Hz), 3.90(2H, s), 4.04 (1H, s), 4.28 (1H,
s), 5.55 (1H, s), 5.69 (1H, d, J=8 Hz), 7.88 (1H, d,
J=8 Hz). 13C NMR (DMSO-d6): d 56.0, 68.7, 71.1,
78.9, 86.4, 89.0, 100.9, 139.2, 150.1, 163.4. MS (EI): m/z
256 (M+, 27.2), 69 (100). Anal. calcd for C10H12N2O6:
C, 46.88; H, 4.72; N,10.93. Found: C, 46.74; H, 4.70; N,
10.84.
30-O-Benzyl-40-( p-toluenesulfonyloxymethyl)uridine (6).
Under N2 atmosphere, NaBH3CN (92 mg, 1.5 mmol)
was added to a stirred solution of 5 (150mg, 0.29 mmol)
in MeCN (3 mL) at room temperature, then TiCl4
(0.16 mL, 1.5 mmol) was added dropwise to the mixture
at 0 ꢀC. After having been stirred for 15 h, the reaction
mixture was partitioned with saturated NaHCO3 and
the mixture was extracted with CHCl3. Usual work up
and puri®cation by SiO2 column chromatography
(CHCl3/MeOH, 25:1) a€orded 6 (112 mg, 0.22 mmol,
5-Bromo-20-O,40-C-methyleneuridine (1BrU). A solution
of sodium azide (183 mg, 2.82 mmol) in H2O (0.55 mL)
and N-bromosuccinimide (136 mg, 0.764 mmol) were
added to a stirred solution of 20-O,40-C-methyleneuri-
dine (150mg, 0.585 mmol) in 1,2-dimethoxyethane
(88 mL) at 20 ꢀC. After having been stirred for 27 h at
the same temperature, the reaction mixture was con-
centrated under reduced pressure and the residue was
puri®ed by SiO2 column chromatography (CHCl3/
75%) as colorless crystals, mp 195±197 ꢀC (AcOEt-hex-
23
D
ane). ‰ꢀŠ À14.6 (c 1.0, CHCl3). IR n (KBr): 3033, 2885,
MeOH, 9:1 to 3:1) to a€ord 1BrU (165 mg, 84%) as col-
23
D
2820, 1726, 1470, 1361, 1274, 1175, 1119 cmÀ1
.
1H
orless crystals, mp 255±256.5 ꢀC (MeOH), ‰ꢀŠ +24.0ꢀ
NMR (CDCl3) d: 2.40(3H, s), 3.59±3.77 (3H, m),
4.10,4.24 (2H, AB, J=11 Hz), 4.32 (1H, d, J=6 Hz),
(c 0.30, MeOH). IR nmax (KBr): 3557, 3330, 3178, 3059,
2958, 2831, 1697, 1615, 1441, 1271, 1126, 1059 cmÀ1. 1H

1008
S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
NMR (CD3OD): d 3.73, 3.94 (2H, AB, J=8 Hz), 3.90
(2H, s), 4.06 (1H, s), 4.29 (1H, s), 5.53 (1H, s), 8.21 (1H,
s). 13C NMR (DMSO): d 55.3, 68.2, 70.9, 78.6, 86.5,
89.0, 95.3, 138.8, 149.3, 159.3. Anal. calcd for
C10H11N2O6Br: C, 35.84; H, 3.31; N, 8.36; Br, 23.84.
Found: C, 35.72; H, 3.29; N, 8.32; Br, 23.85.
30-O-[2-Cyanoethoxy(diisopropylamino)phosphino]-50-O-
(4,40 -dimethoxytrityl)-5-methyl-20 -O,40-C-methyleneuri-
dine (9T). Under N2 atmosphere, 2-cyanoethyl-
N,N,N0,N0-tetraisopropylphosphorodiamidite (142 mg,
0.472 mmol) was added to a solution of 8T (221 mg,
0.39 mmol) and diisopropylammonium tetrazolide
(50mg, 0.271 mmol) in MeCN-THF (5 mL, 3:1) at
20 ꢀC. The stirring was continued for 3 h and the reac-
tion mixture was concentrated under reduced pressure.
The residue was puri®ed by SiO2 column chromato-
graphy (AcOEt/n-hexane/Et3N, 66:33:1) to a€ord 9T
(295 mg, 99%) as a white solid. 31P NMR (CDCl3): d
149.1, 149.2.
5-Methyl-20-O,40-C-methyleneuridine (1T). A mixture of
1BrU (100 mg, 0.298 mmol), ammonium sulfate (1.5 mg,
0.011 mmol)
and
1,1,1,3,3,3-hexamethyldisilazane
(10mL) was re¯uxed under N 2 atmosphere for 24 h. The
reaction mixture was concentrated under reduced
pressure. To the residue were added palladium chloride
(5 mg,
0.029 mmol),
triphenylphosphine
(16 mg,
0.061 mmol) and THF (12 mL), and the mixture was
re¯uxed under N2 atmosphere for 10min. Then, tri-
methylaluminium (2 M in hexane, 0.45 mL, 0.90 mmol)
was added dropwise and the mixture was re¯uxed for
67 h. The solvent was removed under reduced pressure,
water was added to the residue, and the mixture was
extracted with CHCl3. The extract was dried and con-
centrated under reduced pressure. To the residue was
added a solution of ammonium chloride (95 mg,
1.78 mmol) in water (1.5 mL) and MeOH (30mL) and
the mixture was re¯uxed for 5 h. After removing the
solvent, the residue was puri®ed by SiO2 column chro-
matography (H2O/MeOH, 1:0to 4:1) to a€ord 1T
1-[30-O-[2-Cyanoethoxy(diisopropylamino)phosphino]-50-
O-(4,40-dimethoxytrityl)-20-O,40-C-methylene-ꢁ-D-ribo-
furanosyl]-5-methyl-4-(1,2,4-triazol-1-yl)pyrimidinone
(10). To a stirred suspension of 1,2,4-triazole (2.14 g,
31.0mmol) in MeCN (38 mL) was added slowly POCl 3
(1.00 g, 6.54 mmol) followed by Et3N (3.34 g,
ꢀ
33.0mmol) at 0 C. After having been stirred for 30min
at 0 ꢀC, a solution of 9T (400 mg, 0.519 mmol) in MeCN
(7.7 mL) was added over a period of 15 min. The stirring
was continued for 100 min at the same temperature. The
reaction was quenched by addition of saturated
NaHCO3 solution, and then the mixture was extracted
with CH2Cl2. Usual work up and puri®cation by SiO2
column chromatography (AcOEt/n-hexane, 3:2 to 1:0)
a€orded 10 (391 mg, 91%) as a white solid. 31P NMR
(CDCl3): d 149.0, 149.6.
(46 mg, 57%) as colorless crystals, mp 204±205 ꢀC
23
D
(MeOH), ‰ꢀŠ +58.1ꢀ (c 1.01, EtOH). IR nmax (KBr):
3323, 3163, 3027, 2889, 2826, 1689, 1471, 1276,
1057 cmÀ1. 1H NMR (CD3OD): d 1.89 (3H, q, J=1 Hz),
3.74, 3.95 (2H, AB, J=8 Hz), 3.90(2H, s), 4.07 (1H, s),
4.26 (1H, s), 5.53 (1H, s), 7.74 (1H, d, J=1 Hz). 13C
50 -O-(4,40 -Dimethoxytrityl)-20 -O,40 -C-methyleneuridine
(8U). Under N2 atmosphere, 4,40-dimethoxytrityl chlo-
ride (292 mg, 0.86 mmol) and 4-dimethylaminopyridine
(8.8 mg, 0.072 mmol) were added to a stirred solution of
1U (184 mg, 0.72 mmol) in pyridine (1.7 mL) at 20 ꢀC.
After having been stirred for 5 h at the same tempera-
ture, the reaction mixture was partitioned with satu-
rated Na2CO3 solution, and extracted with AcOEt.
Usual work up and puri®cation by SiO2 column chro-
NMR (CD3OD): d 12.6, 57.6, 70.3, 72.4, 80.8, 88.3,
+
90.4, 110.7, 136.8, 151.8, 166.5. MS (EI): m/z 270(M
ꢁ
,
47.7), 69 (100). Anal. calcd for C11H14N2O6 3/4H2O: C,
46.42; H, 5.38; H, 9.61. Found: C, 46.56; H, 5.51; N, 9.87.
50-O-(4-40-Dimethoxytrityl)-5-methyl-2 -O,40-C-methylene-
0
uridine (8T). Under N2 atmosphere, 4,40-dimethoxytrityl
chloride (477 mg, 1.41 mmol) and 4-dimethylaminopyr-
idine (14 mg, 0.114 mmol) were added to a stirred solu-
tion of 1T (318 mg, 1.18 mmol) in pyridine (2.6 mL) at
20 ꢀC. After having been stirred for 4 h at the same
temperature, the reaction mixture was partitioned with
saturated Na2CO3 solution, and extracted with AcOEt.
Usual work up and puri®cation by SiO2 column chro-
matography (AcOEt n-hexane:Et3N, 74:25:1 to AcOEt/
matography (CHCl3/MeOH, 40:1) a€orded 8U (438 mg,
94%) as a white solid, mp 117±120 ꢀC (CHCl3), ‰ꢀŠ
23
D
+17.2ꢀ (c 1.0, CHCl3). IR nmax (KBr): 3393, 3101, 2885,
1
1689, 1464, 1272, 1047 cmÀ1. H NMR (CDCl3): d 2.59
(1H, br), 3.49, 3.57 (2H, AB, J=11 Hz), 3.78 (7H, m),
3.87 (1H, d, J=7 Hz), 4.26 (1H, s), 4.47 (1H, s), 5.60
(1H, d, J=9 Hz), 5.63 (1H, s), 5.84 (4H, d, J=9 Hz),
7.22±7.45 (9H, m), 7.93 (1H, d, J=9 Hz). 13C NMR
(CDCl3): d 50.1, 53.2, 64.8, 66.9, 72.2, 74.1, 81.6, 82.1,
83.1, 96.8, 108.4, 122.0, 122.9, 125.1, 130.1, 134.4, 139.7,
145.3, 153.6, 159.3. MS (EI): m/z 558 (M+, 5.2), 69
MeOH/Et3N, 94:5:1) a€orded 8T (640mg, 95%) as a
26
D
white solid, mp 115±116 ꢀC, ‰ꢀŠ +4.16ꢀ (c 1.01,
CHCl3). IR nmax (KBr): 4401, 4069, 3156, 3048, 2952,
2835, 2552, 2249, 2050, 1710, 1606, 1508, 1463, 1252,
ꢁ
(100). Anal. calcd for C31H30N2O8 1/4H2O: C, 66.12; H,
5.46; N, 4.97. Found: C, 65.91; H, 5.50; N, 4.91.
1179, 1054 cmÀ1
.
1H NMR (CDCl3): d 1.70(3H, d,
J=1 Hz), 3.46, 3.55 (2H, AB, J=11 Hz), 3.80(6H, s),
3.81, 3.87 (2H, AB, J=8 Hz), 4.27 (1H, s), 4.42 (1H, s),
5.63 (1H, s), 6.85 (4H, dd, J=1, 9 Hz), 7.21±7.47 (9H,
m), 7.64 (1H, d, J=1 Hz), 8.32 (1H, br). 13C NMR
(CDCl3): d 12.6, 55.2, 58.2, 70.3, 71.7, 79.3, 86.7, 87.0,
88.1, 110.4, 113.3, 127.1, 128.0, 128.0, 130.0, 134.6,
135.3, 144.4, 149.9, 158.7, 164.1. MS (FAB): m/z 573
(M +H). Anal. calcd for C32H32N2O8 3/4H2O: C,
65.57; H, 5.76; N, 4.78. Found: C, 65.46; H, 5.74; N,
4.63.
30-O-[2-Cyanoethoxy(diisopropylamino)phosphino]-50-O-
(4,40-dimethoxytrityl)-20-O,40-C-methyleneuridine (9U).
Under N2 atmosphere, 2-cyanoethyl-N,N,N0,N0-tetra-
isopropylphosphorodiamidite (114 mg, 0.37 mmol) was
added to a solution of 8U (173 mg, 0.31 mmol) and di-
isopropylammonium tetrazolide (40mg, 0.23 mmol) in
MeCN-THF (4 mL, 3:1) at 20 ꢀC. The stirring was con-
tinued for 1.5 h and the reaction mixture was con-
centrated under reduced pressure. The residue was
+
ꢁ

S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
1009
puri®ed by SiO2 column chromatography (AcOEt/n-
hexane/Et3N, 75:25:1) to a€ord 9U (181 mg, 81%) as a
white solid. 31P NMR (CDCl3): d 149.9.
from 5 to 95 ꢀC at a rate of 0.2 ꢀC/min, and the absor-
bance at 260nm was recorded every 2.5 min. Samples
were dissolved at 2 mM strand concentration in 10mM
sodium phosphate bu€er (pH 6.6 or 7.2) containing
100 mM NaCl. The mixture was incubated at 85 ꢀC for
5 min and then cooled back slowly to 4 ꢀC before the Tm
mesurements. The melting temperature (Tm) was eval-
uated as the temperature of half-dissociation of the
formed triplexes determined by the ®rst derivative of the
melting curve.
1-[30-O-[2-Cyanoethoxy(diisopropylamino)phosphino]-50-
O-(4,40-dimethoxytrityl)-20-O,40-C-methylene-ꢁ-D-ribo-
furanosyl]-4-(1,2,4-triazol-1-yl)pyrimidinone (11). To a
stirred suspension of 1,2,4-triazole (1.99 g, 28.8 mmol) in
MeCN (45 mL) was added slowly POCl3 (1.00 g,
6.54 mmol) followed by Et3N (3.34 g, 33.0mmol) at
ꢀ
0 ꢀC. After having been stirred for 40min at 0 C, a
solution of 9U (632 mg, 0.847 mmol) in MeCN (3.6 mL)
was added over a period of 15 min. The stirring was
continued for 4 h at the same temperature and for 2.5 h
at 20 ꢀC. The reaction was quenched by addition of
saturated NaHCO3 solution, and then the mixture was
extracted with CHCl3. Usual work up a€orded 11
(605 mg, 89%) as a white solid. 31P NMR (CDCl3): d
150.0.
Gel retardation experiments
Target oligonucleotides were labeled at their 50-end with
T4 polynucleotide kinase and [32P]ATP. Samples were
prepared by mixing ꢂ0.2 nM labeled strands and the
appropriate concentration of the third strand in a
50mM HEPES bu€er (pH 7.0) containing 20mM
MgCl2. The mixture was incubated at 95 ꢀC for 5 min
and then cooled back slowly to room temperature
before storage at 4 ꢀC for 48 h. Gel electrophoresis was
run on a 10% polyacrylamide/bisacrylamide (40:1)
nondenaturing gel in a 100 mM HEPES bu€er (pH 7.0)
containing 40mM MgCl 2 with 3 W power at 4 ꢀC. The
gel was dried and visualized by autoradiography.
Synthesis of the 20,40-BNA modi®ed oligonucleotides
Synthesis of the modi®ed oligonucleotides was per-
formed with solid-phase phosphoramidite methodology
on the 0.2 mmol scale using a Pharmacia LKB Gene
AssemblerTM Plus instrument. Reagent solutions were
purchased from Amersham Pharmacia Biotech. The
assembly of modi®ed nucleotides was performed
according to the standard synthesis cycles, with the
exception of a prolonged coupling time (4 min). All
syntheses were performed in the trityl on mode, ending
with 50-tritylated oligonucleotides. After synthesis, the
solid support was suspended in concentrated NH3 solu-
tion (ca. 1 mL) and left for 3 h at 70 ꢀC. The crude pro-
duct was detritylated and puri®ed by NENSORBTM
PREP cartridge (NEN Life Science Products), and then
puri®ed by RP-HPLC on a CHEMCOBOND 5-ODS-H
(4.6Â250mm) column (Chemco). The MALDI-TOF-
MS data of the isolated oligonucleotides were obtained
by using a Voyager DE instrument (PerSeptive Biosys-
tems).
Preparation of 521 bp DNA fragment for DNase I
footprinting experiments
A 36 bp DNA fragment containing target sequence was
synthesized and inserted into Hind III/Bam HI site of
pcDNA3. The 32P-labeled 521 bp DNA fragment used
in DNase I footprinting experiments was prepared by
digestion of the plasmid construct with Nde I and Xba I
followed by 30-end labeling with [32P]dCTP using Kle-
now fragment of DNA polymerase. The labeled DNA
fragment was puri®ed by Sephadex G-50.
DNase I footprinting experiments
The 32P-labeled DNA fragment (ꢂ0.2 nM) was incu-
bated with 2.2 mM third strand for 48 h at 18 ꢀC in a
50mM sodium acetate bu€er (pH 5.5) or in a 50mM
sodium phosphate bu€er (pH 6.0, 6.6 or 7.2) containing
10mM MgCl 2. The mixture was digested with DNase I
for 1 min, and the reaction was terminated by the addi-
tion of 0.2 M NaCl, 20 mM EDTA and 0.05 mg/mL gly-
cogen. The digested DNA fragment was precipitated
with ethanol and resuspended in 98% formamide con-
taining bromophenol blue and xylene cyanol and dena-
tured by heating at 90 ꢀC for 2 min. The sample was
quickly cooled on ice and then analyzed on a 6% poly-
acrylamide/bisacrylamide (19:1) sequence gel in the
presence of 7 M urea. The gel was dried and visualized
by autoradiography.
MALDI-TOF-MS data [M+H]+ for the modi®ed oligo-
nucleotides 12±20
Oligonucleotides
Calculated
4246.7
4330.8
4274.7
4358.9
4274.7
4358.9
4303.7
4471.7
4555.04556.3
Found
4247.9
4332.6
4276.6
4358.0
4275.9
4359.7
4305.9
4473.9
12
13
14
15
16
17
18
19
20
Circular dichroismspectroscopy
DNA melting experiments
Circular dichroism (CD) measurements were carried out
on a JASCO J-720W spectropolarimeter equipped with
a temperature controller PTC-348W. Samples were dis-
solved at 2 mM strand concentration in 10mM sodium
phosphate bu€er (pH 7.2) containing 100 mM NaCl.
DNA melting experiments were carried out on a Beck-
man DU650spectrophotometer equipped with Tm ana-
lysis accessory using quartz cuvettes of 1 cm optical path
length. The temperature of the cell holder was increased

1010
S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
The concentration used to calculate the CD spectra was
that of the nucleotide unit.
the A form. See: Soyfer, V. N.; Potaman, V. N. Triple-
Helical Nucleic Acids; Springer-Verlag: New York, 1996
pp 100±150.
7. Recent reviews for conformationally restricted or locked oli-
gonucleotides. See: Herdewijn, P. Liebigs Ann. Chem. 1996, 1337.
Kool, E. T. Chem. Rev. 1997, 97, 1473. Wengel, J. Acc. Chem.
Res. 1999, 32, 30 1.
8. Obika, S.; Nanbu, D.; Hari, Y.; Morio, K.; In, Y.; Ishida,
T.; Imanishi, T. Tetrahedron Lett. 1997, 38, 8735.
9. Obika, S.; Nanbu, D.; Hari, Y.; Andoh, J.; Morio, K.; Doi,
T.; Imanishi, T. Tetrahedron Lett. 1998, 39, 5401.
10. Obika, S.; Andoh, J.; Sugimoto, T.; Miyashita, K.; Ima-
nishi, T. Tetrahedron Lett. 1999, 40, 6465.
11. Obika, S.; Hari, Y.; Morio, K.; Imanishi, T. Tetrahedron
Lett. 2000, 41, 215.
12. Obika, S.; Hari, Y.; Morio, K.; Imanishi, T. Tetrahedron
Lett. 2000, 41, 221.
13. Obika, S.; Hari, Y.; Sugimoto, T.; Sekiguchi, M.; Ima-
nishi, T. Tetrahedron Lett. 2000, 41, 8923.
14. Torigoe, H.; Hari, Y.; Sekiguchi, M.; Obika, S.; Imanishi,
T. J. Biol. Chem., in press.
Inhibition of NF-ꢀB (p50) binding
Target oligonucleotides were labeled at their 50-end with
T4 polynucleotide kinase and [32P]ATP. Samples were
prepared by mixing ꢂ1 nM labeled strands in a 9 mM
HEPES bu€er (pH 7.0) containing 9 mM MgCl2,
45 mM KCl and 0.2 mM EDTA. The mixture was incu-
bated at 95 ꢀC for 5 min and then cooled back slowly to
room temperature before addition of 2 mM DTT, 9%
glycerol, 0.05% NP40, 0.05 mg/mL poly (dI-dC), 1.2 gel
shift units of p50(Promega) and the appropriate con-
centration of 15. After incubation at 18 ꢀC for 48 h, gel
electrophoresis was run on a 10% polyacrylamide/
bisacrylamide (40:1) nondenaturing gel in a 10 mM
HEPES bu€er (pH 7.0) with 3 W power at 4 ꢀC. The gel
was dried and visualized by autoradiography.
15. Obika, S. Yakuga Kuzasshi 2000, 120, 147.
16. Obika, S.; Morio, K.; Nanbu, D.; Imanishi, T. Chem.
Commun. 1997, 1643.
17. Obika, S.; Morio, K.; Hari, Y.; Imanishi, T. Bioorg. Med.
Chem. Lett. 1999, 9, 515.
Acknowledgements
18. Obika, S.; Morio, K.; Hari, Y.; Imanishi, T. Chem. Com-
mun. 1999, 2423.
A part of this work was supported by a Grant-in Aid
for Scienti®c Research (B), Nos 11470469 and
12557201, from Japan Society for the Promotion of
Science. We are also grateful to Takeda Science Foun-
dation for ®nancial support.
19. Singh, S. K.; Nielsen, P.; Koshkin, A. A.; Wengel, J.
Chem. Commun. 1998, 455. Koshkin, A. A.; Singh, S. K.;
Nielsen, P.; Rajwanshi, V. K.; Kumar, R.; Meldgaard, M.;
Olsen, C. E.; Wengel, J. Tetrahedron 1998, 54, 3607.
20. Cook, S. L.; Secrist, J. A. J. Am. Chem. Soc. 1979, 101,
1554.
21. Jones, G. H.; Taniguchi, M.; Tegg, D.; Mo€att, J. G. J.
Org. Chem. 1979, 44, 1309.
References and Notes
22. The stereochemistry of 5 was con®rmed by the observa-
tion of NOE between the benzyliden hydrogen and 20- or 30-
hydrogen. It is thought that the endo-isomer was selectively
obtained thermodynamically. See: Li, C.; Vasella, A. Helv.
Chim. Acta 1993, 76, 211.
23. Recently, Wengel et al. reported the synthesis of 1 from 4-
C-hydroxymethylribofuranose derivative.19,24 Independently,
we also achieved the e€ective synthesis of 1 in a similar
manner.15
1. A part of this work appeared in a preliminary report: Ima-
nishi, T.; Obika, S. J. Synth. Org. Chem., Jpn 1999, 57, 969.
2. Moser, H. E.; Dervan, P. B. Science 1987, 238, 645. Le
Doan, T.; Perrouault, L.; Praseuth, D.; Habhoub, N.; Decout,
J. L.; Thuong, N. T.; Lhomme, J.; Helene, C. Nucleic Acids
Res. 1987, 15, 7749. Lyamichev, V. I.; Mirkin, S. M.; Frank-
Kamenetskii, M. D.; Cantor, C. R. Nucleic Acids Res. 1988,
16, 2165. Fox, K. R. Curr. Med. Chem. 2000, 7, 17.
24. Koshkin, A. A.; Rajwanshi, V. K.; Wengel, J. Tetrahedron
Lett. 1998, 39, 4381.
25. Kumar, R.; Wiebe, L. I.; Knaus, E. E. Can. J. Chem.
1994, 72, 2005. Hirota, K.; Kitade, Y.; Kanbe, Y.; Isobe, Y.;
Maki, Y. Synthesis 1993, 213.
26. Xu, Y.-Z.; Zheng, Q.; Swann, P. F. J. Org. Chem. 1992,
57, 3839.
27. Xie, Q.; Nathan, C. J. Leukoc. Biol. 1994, 56, 576.
28. Lee, J. S.; Woodsworth, M. L.; Latimer, L. J.; Morgan,
A. R. Nucleic Acids Res. 1984, 12, 6603.
29. The association constants were calculated according to the
literature. See: Akiyama, T.; Hogan, M. E. J. Biol. Chem.
1996, 271, 29126.
30. Baeuerle, P. A. Biochim. Biophys. Acta 1991, 1072, 63.
31. Asensio, J. L.; Carr, R.; Brown, T.; Lane, A. N. J. Am.
Chem. Soc. 1999, 121, 11063.
32. Hendrix, C.; Rosemeyer, H.; Verheggen, I.; Seela, F.; Van
Aershot, A.; Herdewijn, P. Chem. Eur. J. 1997, 3, 110.
33. Tereshko, V.; Gryaznov, S.; Egli, M. J. Am. Chem. Soc.
1998, 120, 269.
34. Bolli, M.; Leumann, C. Angew. Chem., Int. Ed. Engl. 1995,
34, 694.
3. Thuong, N. T.; Helene, C. Angew. Chem., Int. Ed. Engl.
1993, 32, 666. Luyten, I.; Herdewijn, P. Eur. J. Med. Chem.
1998, 33, 515. Gowers, D. M.; Fox, K. R. Nucleic Acids Res.
1999, 27, 1569.
4. N30±P50 Phosphoramidate oligonucleotides are known to
form a stable triplex with a homopurineꢁ homopyrimidine tract
in dsDNA. See: Gryaznov, S. M.; Chen, J.-K. J. Am. Chem.
Soc. 1994, 116, 3143. Escude, C.; Giovannangeli, 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.
Gryaznov, S. M.; Winter, H. Nucleic Acids Res. 1998, 26,
4160.
5. Only C-rich homopyrimidine sequences of peptide nucleic
acids (PNA) can bind to dsDNA to form PNAꢁ DNAꢁ DNA
triplex. See: Wittung, P.; Nielsen, P.; Norden, B. Biochemistry
1997, 36, 7973. Uhlmann, E.; Peyman, A.; Breipohl, G.; Will,
D. W. Angew. Chem. Int. Ed. 1998, 37, 2796. However, except
for the C-rich sequences, the strand invasion (PNAꢁ DNAꢁ PNA
triplex formation) was favored. See: Nielsen, P. E.; Egholm,
M.; Berg, R. H.; Buchardt, O. Science 1991, 254, 1497.
6. Although the conformation of triplexes was believed to be
A-form-like structures (N-form sugar puckering) from the
X-ray ®ber analysis of poly(dT)ꢁ poly(dA)ꢁ poly(dT), recent
studies have shown that some triplex structures di€er from
35. Ste€ens, R.; Leumann, C. J. Am. Chem. Soc. 1999, 121,
3249.

S. Obika et al. / Bioorg. Med. Chem. 9 (2001) 1001±1011
1011
36. Tarkoy, M.; Leumann, C. Angew. Chem., Int. Ed. Engl. 38. Wahlestedt, C.; Salmi, P.; Good, L.; Kela, J.; Johnsson,
T.; Hokfelt, T.; Broberger, C.; Porreca, F.; Lai, J.; Ren, K.;
È
È
1993, 32, 1432. Tarkoy, M.; Bolli, M.; Leumann, C. Helv.
È
Chim. Acta 1994, 77, 716.
37. Kochetkova, M.; Shannon, M. F. J. Biol. Chem. 1996,
271, 14438.
Ossipov, M.; Koshkin, A.; Jakobsen, N.; Skouv, J.; Oerum,
H.; Jacobsen, M. H.; Wengel, J. Proc. Natl. Acad. Sci. U.S.A.
2000, 97, 5633.