Methylcytosine-selective fluorescence quenching by osmium complexation

Article abstract of DOI:10.1016/j.bmc.2006.12.023

We report on the control of the emission from a fluorophore fixed on DNA using the methylcytosine-selective addition of an osmium-bipyridine complex. We have synthesized DNA modified by a microenvironment-sensitive fluorophore, 2-dimethylamino-6-acyl-naph

Full text of DOI:10.1016/j.bmc.2006.12.023

Bioorganic & Medicinal Chemistry 15 (2007) 1615–1621
Methylcytosine-selective fluorescence quenching by
osmium complexation
Kazuo Tanaka,a Kazuki Tainakaa and Akimitsu Okamotoa,b,
*
aFrontier Research System, RIKEN (The Institute of Physical and Chemical Research), Wako, Saitama 351-0198, Japan
bDepartment of Synthetic Chemistry and Biological Chemistry, Faculty of Engineering, Kyoto University, Kyoto 615-8510, Japan
Received 7 November 2006; revised 12 December 2006; accepted 13 December 2006
Available online 15 December 2006
Abstract—We report on the control of the emission from a fluorophore fixed on DNA using the methylcytosine-selective addition of
an osmium-bipyridine complex. We have synthesized DNA modified by a microenvironment-sensitive fluorophore, 2-dimethylami-
no-6-acyl-naphthalene. The emission from the fluorophore tethered to a probe DNA was effectively quenched by a methylcytosine
glycol-osmium-bipyridine triad, which was located in the immediate neighborhood of the fluorophore. The discrimination of the
cytosine methylation status at a methylation hot spot in the p53 gene was also executed using a well-designed fluorescent DNA
probe.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
Here, we report on the control of the fluorescence emis-
sion from a fluorophore fixed on DNA using the meth-
ylcytosine-selective addition of an osmium–bipyridine
complex. The DNA modified with a microenviron-
ment-sensitive fluorophore was synthesized, and the
fluorescence was effectively quenched by the addition
of the osmium–bipyridine complex to methylcytosine,
which is located in the immediate neighborhood of the
fluorophore.
Cytosine methylation plays a crucial role in the regula-
tion of chromatin stability, gene regulation, parental
imprinting, and X chromosome inactivation in fe-
males.1–4 In addition, erroneous DNA methylation
may contribute to the etiopathogenesis of tumorigenesis
and aging.5,6 Therefore, a simple diagnosis of cytosine
methylation status is highly desirable for further under-
standing of the role of 5-methylcytosine (M) in human
genes.
2. Results and discussion
There have been very few reports on cytosine methyla-
tion analysis based on chemical concepts; for example,
cytosine hydrolysis with sodium bisulfite.7–9 To con-
struct an effective method for a high-throughput detec-
tion of methylation in the future, the establishment of
optical analysis systems, particularly one utilizing fluo-
rescence emission, is essential, but there remains the dif-
ficulty in developing these, because there is only a slight
difference in the photophysical properties of cytosine
and methylcytosine.
2.1. Synthesis of a fluorescence-labeled nucleoside, D
We designed a DNA probe with a microenvironment-
sensitive fluorophore, 6-dimethylamino-2-acylnaphtha-
lene (DAN). This has a unique fluorescence property:
the emission intensity and wavelength vary sharply with
the degree of protonation and solvent polarity, and with
any microstructural changes.10–12 We incorporated this
fluorophore into DNA. The purine base modified with
DAN (D) forms a wobble base pair with pyrimidines,
and in addition, the acyl group of the DAN is expected
to form an extra hydrogen bond with the amino group
of the base-pairing cytosine (Fig. 1).13,14 The synthesis
of D-containing DNA is outlined in Scheme 1. The
DAN derivative 2, which was prepared from 6-dimeth-
ylamino-2-naphthaldehyde (1) in four steps, was cou-
Keywords: DNA methylation; Fluorescent probes; FRET; Nucleobases;
Oxidation.
*
Corresponding author. Tel.: +81 48 467 9238; fax: +81 48 467
9205; e-mail: aki-okamoto@riken.jp
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.12.023

1616
K. Tanaka et al. / Bioorg. Med. Chem. 15 (2007) 1615–1621
(X = M) opposite a D base (Tm = 33.2 and 31.9 ꢁC,
respectively, Table 2), although the addition of a hydro-
phobic fluorophore decreased the duplex stability, com-
pared with the stability of the duplex with a C/A
mismatched base pair. ODN(C)/cODN(D) was more
stable than ODN(M)/cODN(D), implying the formation
of an extra hydrogen bond between cytosine and D as
shown in Figure 1 and/or the M/D base pair destabiliza-
tion caused by steric hindrance between a 5-methyl
group of methylcytosine and the DAN chromophore
of D. These fluorescent duplexes showed similar fluores-
cence intensities at 515 nm (UF = 0.075 and 0.061,
respectively, excited at 390 nm, Table 3). At this stage,
the fluorescence of D was not much affected by the pres-
ence/absence of the C5 methyl group of a cytosine ring.
To more effectively distinguish 5-methylcytosine from
cytosine, we modified ODN(X) by the addition of an
osmium–bipyridine complex. The addition of osmium
tetroxide and bipyridine to pyrimidine bases is known
to provide a pyrimidine glycol-osmium-bipyridine tri-
ad.16–18 We have previously established the sequence-se-
lective complexation of 5-methylcytosine in DNA using
potassium osmate and bipyridine.19 This treatment dras-
tically changed the fluorescence intensity of the D-con-
taining duplex with a high methylcytosine selectivity
(Fig. 2a). ODN(X) (5 lM) was added to a solution of
5 mM potassium osmate, 100 mM potassium hexacy-
anoferrate(III), 100 mM bipyridine, and 1 mM EDTA
in 100 mM Tris–HCl buffer (pH 7.7) and 10% acetoni-
trile. This mixture was then incubated at 37 ꢁC for 3 h.
The fluorescence spectra of the resulting ODNs were
measured after desalting and hybridization with
cODN(D). The fluorescence of the duplex ODN(MOs)/
cODN(D), which was produced via osmium addition
to ODN(M), was significantly quenched (UF = 0.018,
Table 3) compared to the fluorescence emission of
ODN(C)/cODN(D) and ODN(M)/cODN(D) (Figs. 2b
and c). The decay in fluorescence was strongly influ-
enced by the osmium complex formation. A new rapid
component (s1 = 43 ps) was observed only for ODN
(MOs)/cODN(D) (Table 4), suggesting that photo-in-
duced electron transfer from the excited D base to the
osmium complex occurred, and that this contributed
to the quenching process.
Figure 1. Our proposed base-pairing model of a fluorescence-labeled
purine base, D, and cytosine.
Scheme 1. Synthesis of a fluorescent nucleoside, D. Reagents and
conditions: (a) O-(tert-butyldimethylsilyl)propargyl alcohol, n-butyl
lithium, THF, ꢀ78 ꢁC, rt, 30 min, 96%; (b) H2, Pd–C, methanol, rt,
9 h, 65%; (c) tetrapropylammonium perruthenate, 4-methylmorpholine
N-oxide, dichloromethane, rt, 2 h, 87%; (d) tetrabutylammonium
fluoride, THF, rt, 6 h, 95%; (e) 30O,50O-bis(t-butyldimethylsilyl)-20-
deoxyguanosine, triphenylphosphine, diethyl azodicarboxylate, diox-
ane, 70 ꢁC, 6 h, 73%; (f) tetrabutylammonium fluoride, THF, rt, 2 h,
73%; (g) dimethylformamide dimethylacetal, DMF, rt, 6 h, 94%; (h)
4,40-dimethoxytrityl chloride, pyridine, 50 ꢁC, 2 h, 87%; and (i)
(iPr2N)2PO(CH2)2CN, 1H-tetrazole, acetonitrile, rt, 2 h.
Accordingly, the distance between the D base and the
osmium complex is very important for efficient fluores-
cence quenching. We measured the fluorescence intensi-
ty of several duplexes that had different distances
between D and the osmium complex attached to 5-meth-
ylcytosine (Fig. 3) or thymine (Fig. 4). Little significant
fluorescence quenching was observed for all of the
duplexes that contained osmium complexes on one or
two of the bases next to the opposite D base. The results
clearly show that fluorescence quenching occurred only
on osmium complexation at the methylcytosine located
opposite to the D base. Although many 5-methylcyto-
sines and thymines that can form complexes with osmi-
um may be included in a DNA strand, the fluorescence
of D was very sensitive only to complexation of 5-meth-
ylcytosine opposite D. The selective quenching suggests
that the methylation status of cytosine at the site of
pled with silyl-protected 20-deoxyguanosine15 to give 3.
The deprotection of 3 provided a nucleoside 4, which
was subsequently converted into the phosphoramidite
5 for use in a DNA autosynthesizer. The synthesized
oligodeoxyribonucleotides (ODN) are summerized in
Table 1.
2.2. Quenching of D fluorescence by osmium complexation
The ODN containing a D base, 50-d(TTTTTTCDC
TTTTTT)-30 (cODN(D)), formed a stable duplex with
ODNs, 50-d(AAAAAAGXGAAAAAA)-30 (ODN(X)),
containing cytosine (X = C) or 5-methylcytosine

K. Tanaka et al. / Bioorg. Med. Chem. 15 (2007) 1615–1621
Table 1. ODN sequences used in this study and their MS values
1617
Sequences
Calcd
Founda
ODN(C)
ODN(M)
ODN(MOs
ODN(T)
50-AAA AAA GCG AAA AAA-30
50-AAA AAA GMG AAA AAA-30
50-AAA AAA GMOsG AAA AAA-30
50-AAA AAA GTG AAA AAA-30
50-AAA AAA GTOsG AAA AAA-30
50-TTT TTT CDC TTT TTT-30
—
—
4657.13b
5051.54c
—
4657.14
5051.36
—
)
ODN(TOs
)
5052.53c
4734.24b
—
5052.72
4734.51
—
cODN(D)
cODN(G)
50-TTT TTT CGC TTT TTT-30
ODN(M50-1)
ODN(M50-2)
ODN(M30-1)
ODN(M30-2)
ODN(MOs50-1)
ODN(MOs50-2)
ODN(MOs30-1)
ODN(MOs30-2)
ODN(T50-1)
ODN(T50-2)
ODN(T30-1)
ODN(T30-2)
ODN(TOs50-1)
ODN(TOs50-2)
ODN(TOs30-1)
ODN(TOs30-2)
cODN(D50-1)
cODN(D50-2)
cODN(D30-1)
cODN(D30-2)
p53(C)
50-AAA AAM GCG AAA AAA-30
50-AAA AMA GCG AAA AAA-30
50-AAA AAA GCG MAA AAA-30
50-AAA AAA GCG AMA AAA-30
50-AAA AAMOs GCG AAA AAA-30
50-AAA AMOsA GCG AAA AAA-30
50-AAA AAA GCG MOsAA AAA-30
50-AAA AAA GCG AMOsA AAA-30
50-AAA AAT GCG AAA AAA-30
50-AAA ATA GCG AAA AAA-30
50-AAA AAA GCG TAA AAA-30
50-AAA AAA GCG ATA AAA-30
50-AAA AATOs GCG AAA AAA-30
50-AAA ATOsA GCG AAA AAA-30
50-AAA AAA GCG TOsAA AAA-30
50-AAA AAA GCG ATOsA AAA-30
50-TTT TTT CDC ATT TTT-30
4633.11b
4633.11b
4633.11b
4633.11b
5027.52c
5027.52c
5027.52c
5027.52c
—
4633.76
4633.81
4633.13
4633.86
5028.01
5028.51
5027.66
5027.71
—
—
—
—
—
—
—
5028.89c
5028.89c
5028.89c
5028.89c
4743.25b
4743.25b
4743.25b
4743.25b
—
5028.87
5029.77
5028.65
5027.31
4743.99
4743.15
4743.01
4744.31
—
50-TTT TTT CDC TAT TTT-30
50-TTT TTA CDC TTT TTT-30
50-TTT TAT CDC TTT TTT-30
50-GCTATCTGAGCAGCGCTCATGGTGGGGG
CAGCGCCTCACAACCTCCGTCATGTGCTGTGA-30
50-GCTATCTGAGCAGCGCTCATGGTGGGGGCAGM
GCCTCACAACCTCCGTCATGTGCTGTGA-30
50-GTG AGG CDC TGC CCC-30
p53(M)
—
—
cODN(p53)
4808.27b
4809.62
a Mass spectra were taken by MALDI-TOF MS with 20,30,40-trihydroxyacetophenone as a matrix.
b These values were calculated by [MꢀH]ꢀ.
c These values were calculated by [MꢀOH]ꢀ.
Table 2. Melting temperatures of D-containing ODNs
Table 3. Fluorescence quantum yields of DAN in ODNs
Tm/ꢁCa
a
UF
Duplexes
ODN(C)/cODN(G)
ODN(C)/cODN(A)
ODN(C)/cODN(D)
ODN(M)/cODN(D)
ODN(MOs)/cODN(D)
48.6
35.1
33.2
31.9
35.5
cODN(D)
0.068
0.068
0.075
0.061
0.035
0.018
ODN(T)/cODN(D)
ODN(C)/cODN(D)
ODN(M)/cODN(D)
ODN(TOs)/cODN(D)
ODN(MOs)/cODN(D)
a 2.5 lM duplexes in 50 mM sodium phosphate and 0.1 M sodium
chloride (pH 7.0). Detection wavelength was at 260 nm.
a cODN(D) hybridized with ODNs was measured (2.5 lM in 50 mM
sodium phosphate, 0.1 M sodium chloride, pH 7.0, 25 ꢁC).
kex = 390 nm, kem = 515 nm.
interest could be assessed fluorometrically without any
sequence dependence using osmium complexation.
Therefore, this phenomenon would be applicable to a
fluorescent epigenotyping method.
desalting through a gel filter, the recovered DNA was
hybridized with D-containing DNA probe,
a
cODN(p53), in a 96-well microliter plate, and its fluores-
cence intensity at 510 nm was evaluated. Compared to
the control sample, a small difference in the fluorescence
intensity was observed for the nonmethylated sample,
whereas the fluorescence of the methylated sample was
markedly decreased (Fig. 5). Using this fluorescent sys-
tem, cytosine methylation at the sequence of interest
can be effectively analyzed.
2.3. Detection of methylcytosine in p53 sequence
We prepared
a
60-mer DNA strand, 50-d(. . .
GGCAGXGCCTC. . .)-30, which contained a methyla-
tion hotspot (X = M or C) at codon 175 in exon 5 of
the p53 gene (p53(X)).20 We treated this DNA with the
osmium oxidation reagents described above. After

1618
K. Tanaka et al. / Bioorg. Med. Chem. 15 (2007) 1615–1621
10
5
I
/a.u.
0
600
400
500
λ
/nm
Figure 4. Fluorescence spectrum of the duplexes with different
distances between T and D, ODN(T50-1)/cODN(D50-1) (blue line),
ODN(T50-2)/cODN(D50-2) (red line), ODN(T30-1)/cODN(D30-1) (pink
line), ODN(T30-2)/cODN(D30-2) (light blue line), ODN(TOs50-1)/
cODN(D50-1) (green line), ODN(TOs50-2)/cODN(D50-2) (light green
line),ODN(TOs30-1)/cODN(D30-1) (purple line),ODN(TOs30-2)/cODN
(D30-2) (light purple line). Reaction solutions containing 2.5 lM
duplexes and 0.1 M sodium chloride in 50 mM sodium phosphate (pH
7.0) were excited at 390 nm at 25 ꢁC.
Figure 2. Fluorescence quenching by 5-methylcytosine-selective osmi-
um complexation: (a) a new way to distinguish methylcytosine from
cytosine using a fluorophore-tethered modified base; (b) fluorescence
spectra of 2.5 lM cODN(D) (light blue line) hybridized with 2.5 lM
ODN(C) (green line), ODN(M) (orange line), and ODN(MOs
)
(magenta line) in 50 mM sodium phosphate (pH 7.0) and 0.1 M
sodium chloride at 25 ꢁC. Excitation was at 390 nm; and (c) fluores-
cence image of ODN(C)/cODN(D) (left) and ODN(MOs)/cODN(D)
(right). The image was taken using a transilluminator (366 nm).
Table 4. Fluorescence analysis for D-containing ODNsa
s1/ns (a1) s2/ns (a2) s3/ns (a3) v2
ODN(M)/cODN(D)
—
1.156 (60) 3.352 (40) 0.96
ODN(MOs)/cODN(D) 0.043 (50) 1.448 (29) 3.814 (21) 1.06
a cODN(D) hybridized with ODN(M) or ODN(MOs) was measured
(2.5 lM in 50 mM sodium phosphate, 0.1 M sodium chloride, pH
7.0, 25 ꢁC). kex = 337 nm, kem = 510 nm.
Figure 5. Distinction between cytosine and methylcytosine at a p53
methylation hotspot using the fluorescence quenching of cODN(p53)
(50 mM sodium phosphate, 0.1 M sodium chloride, pH 7.0, 25 ꢁC).
Excitation was at 380 10 nm, and the fluorescence intensity was
measured through 510 5 nm band-pass filter. The data points
represent the averages of five experimental results, and the 10% error
bars are retained from the individual sets.
10
5
3. Conclusion
In conclusion, we have synthesized a DAN-labeled
nucleoside, D, whose fluorescence is very sensitive to
osmium complexation at the 5-methylcytosine site oppo-
site D. This fluorescence quenching enables us to distin-
guish 5-methylcytosines from cytosines. Thus, the
combination of a DAN-labeled DNA probe and osmi-
um complexation would be useful as a fluorescent epi-
genotyping system.
I
/a.u.
0
400
500
600
λ/nm
Figure 3. Fluorescence spectrum of the duplexes with different
distances between M and D, ODN(M50-1)/cODN(D50-1) (blue line),
ODN(M50-2)/cODN(D50-2) (red line), ODN(M30-1)/cODN(D30-1)
(pink 0 line), ODN(M30-2)/cODN(D30-2) (light blue line), ODN
(MOs5 -1)/cODN(D50-1) (green line), ODN(MOs50-2)/cODN(D50-2)
(light green line),ODN(MOs30-1)/cODN(D30-1) (purple line),ODN
(MOs30-2)/cODN(D30-2) (light purple line). Reaction solutions con-
taining 2.5 lM duplexes and 0.1 M sodium chloride in 50 mM sodium
phosphate (pH 7.0) were excited at 390 nm at 25 ꢁC.
4. Experimental
1H NMR spectra were measured with Varian Mercury
400 (400 MHz) spectrometer. 13C NMR spectra were
measured with JEOL JNM a-500 (125 MHz) spectrom-
eter. Coupling constants (J value) are reported in Hertz.

K. Tanaka et al. / Bioorg. Med. Chem. 15 (2007) 1615–1621
1619
The chemical shifts are expressed in ppm downfield from
tetramethylsilane, using residual chloroform (d = 7.24 in
1H NMR, d = 77.0 in 13C NMR) as an internal stan-
dard. FAB mass spectra were recorded on JEOL JMS
DX-300 spectrometer or JEOL JMS SX-102A spectrom-
eter. Masses of oligodeoxyribonucleotides were deter-
To a solution of the alcohol (189 mg, 0.51 mmol) in
dichloromethane (4 mL) was added flame-dried molecu-
lar sieves (4A, 1/16, 100 mg), 4-methylmorpholine N-ox-
ide (90 mg, 0.77 mmol), and tetrapropylammonium
perruthenate (18 mg, 0.05 mmol) at 0 ꢁC, and the mix-
ture was stirred at room temperature for 2 h. After the
reaction was completed, the mixture was diluted with
diethyl ether (80 mL) and Florisil (150–250 lm, 60–100
mesh, 100 mg) was added to the solution and stirred at
room temperature for 15 min. The mixture was filtered
through Celite and washed with diethyl ether. The fil-
trate and washings were combined and evaporated un-
der reduced pressure. The residue was extracted with
Na2SO3 aq and ethyl acetate. The organic layer was
washed with brine, dried over Na2SO4, filtered, and
evaporated under reduced pressure. The crude product
was purified by silica gel column chromatography (hex-
ane/ethyl acetate = 5:1) to yield the ketone (162 mg,
mined with
a
MALDI-TOF mass spectroscopy
(PerSeptive Biosystems Voyager Elite, acceleration volt-
age 21 kV, negative mode) with 20,30,40-trihydroxyace-
tophenone as matrix, using dT8 ([MꢀH]ꢀ 2370.61) and
dT17 ([MꢀH]ꢀ 5108.37) as an internal standard. HPLC
was performed on a Cosmosil 5C-18AR or CHEMCO-
BOND 5-ODS-H column (4.6· 150 mm) with a Gilson
Chromatography Model 305 using a UV detector Model
118 at 254 nm.
4.1. Synthesis
1
4.1.1. Compound 2. To a solution of O-(tert-butyldim-
ethylsilyl)propargyl alcohol (2.31 g, 13.5 mmol) in
THF (40 mL) was added 1.6 M n-BuLi (8.48 ml,
13.5 mmol) at ꢀ78 ꢁC, and stirred at 0 ꢁC for
30 min. 6-Dimethylamino-2-naphthaldehyde (1.50 g,
7.54 mmol) in THF (15 mL) was added to the mixture
at ꢀ78 ꢁC and then stirred at room temperature for
30 min. After the reaction was quenched with NH4Cl
aq at 0 ꢁC, the mixture was evaporated under reduced
pressure. The reaction mixture was extracted with
NH4Cl aq and ethyl acetate. The organic layer was
washed with brine, dried over Na2SO4, filtered, and
evaporated under reduced pressure. The crude product
was purified by silica gel column chromatography (hex-
ane/ethyl acetate = 6:1) to yield the adduct (2.88 g,
0.44 mmol, 87%) as a yellow solid: H NMR (CDCl3)
d 8.35 (d, 1H, J = 1.5 Hz), 7.95 (dd, 1H, J = 1.8,
8.8 Hz), 7.79 (d, 1H, J = 9.2 Hz), 7.63 (d, 1H,
J = 8.8 Hz), 7.15 (dd, 1H, J = 2.7, 9.2 Hz), 6.87 (d, 1H,
J = 2.4 Hz), 3.75 (t, 2H, J = 6.0 Hz), 3.14 (t, 2H,
J = 7.1 Hz), 3.09 (s, 6H), 2.05–1.98 (m, 2H), 0.93 (s,
9H), 0.08 (s, 6H): 13C NMR (CDCl3) d 99.6, 150.1,
137.5, 130.6, 129.7, 126.1, 125.0, 124.5, 116.2, 105.2,
62.4, 40.3, 34.5, 27.8, 25.9, 18.3, ꢀ5.3: MS (FAB) m/z
371 [M+]; HRMS (FAB) calculated for C22H33O2NSi
[M+] 371.2281; found: 371.2283.
To a solution of the ketone (2.87 g, 7.72 mmol) in tet-
rahydrofuran (20 mL) was added a 1.0 M solution of
tetra-n-butylammonium fluoride in tetrahydrofuran
(9.24 mL, 9.24 mmol), and the mixture was stirred for
6 h at room temperature. The reaction mixture was
concentrated and purified by column chromatography
on silica gel, eluting with hexane/ethyl acetate (1:1) to
give 2 (1.90 g, 95 %) as a yellow solid:1H NMR
(CDCl3) d 8.35 (d, 1H, J = 1.6 Hz), 7.93 (dd, 1H,
J = 1.8, 8.8 Hz), 7.78 (d, 1H, J = 9.0 Hz), 7.62 (d, 1H,
J = 9.0 Hz), 7.16 (dd, 1H, J = 2.6, 9.2 Hz), 6.86 (d,
1H, J = 2.6 Hz), 3.77 (t, 2H, J = 5.9 Hz), 3.21 (t, 2H,
J = 7.0 Hz), 3.10 (s, 6H), 2.15 (br, 1H), 2.09–2.02 (m,
2H): 13C NMR (CDCl3) d 200.1, 150.1, 137.6, 130.6,
130.2, 129.9, 126.1, 124.9, 124.4, 116.1, 105.1, 62.3,
40.2, 35.0, 27.3: MS (FAB) m/z 257 [M+]; HRMS
(FAB) calculated for C16H19O2N [M+] 257.1416;
found: 257.1417.
1
7.49 mmol, 96%) as a yellow solid: H NMR (CDCl3)
d 7.82 (s, 1H), 7.70 (d, 1H, J = 9.2 Hz), 7.66 (d, 1H,
J = 8.6 Hz), 7.52 (dd, 1H, J = 1.8, 8.6 Hz), 7.17 (dd,
1H, J = 2.6, 9.2 Hz), 6.92 (s, 1H), 5.59 (s, 1H), 4.44 (d,
2H, J = 1.8 Hz), 3.05 (s, 6H), 2.36 (s, 1H), 0.92 (s,
9H), 0.13 (s, 6H): 13C NMR (CDCl3) d 148.8, 134.7,
134.0, 128.9, 126.7, 126.3, 125.2, 125.0, 116.6, 106.4,
85.1, 84.8, 64.7, 51.8, 40.8, 25.8, 18.2, ꢀ5.2: MS
(FAB), m/z 369[M+]; HRMS (FAB) calculated for
C22H31O2NSi [M+] 369.2131; found: 369.2124.
A mixture of the adduct (2.88 g, 7.49 mmol) and 10%
Pd/C (2.0 g) in methanol (100 mL) was stirred under
hydrogen atmosphere for 9 h at room temperature.
The mixture was filtered through Celite and then
washed with methanol. The filtrate and washings were
combined and evaporated under reduced pressure. The
crude product was purified by silica gel chromatogra-
phy (hexane/ethyl acetate = 5–3:1) to give the alcohol
4.1.2. Compound 3. Triphenylphosphine (262 mg,
1 mmol), 30O,50O-bis(tert-butyldimethylsilyl)-20-deoxy-
guanosine (496 mg, 1 mmol), and 2 (257 mg, 1 mmol)
were placed in a dried flask under vacuum for 15 min.
After the flask was refilled with argon, dry dioxane
(5 mL) was added via syringe and stirred at 70 ꢁC.
Diethylazodicarboxylate (460 lL, 1 mmol) was added
to the resulting suspension slowly via syringe. After
the solution became homogeneous, the reaction mixture
was stirred for an additional 6 h. The reaction mixture
was concentrated and purified by column chromatogra-
phy on silica gel, eluting with 1 : 2 hexane–ethyl acetate
1
(1.74 g, 4.66 mmol, 65%) as a yellow solid: H NMR
(CDCl3)
d
7.71–7.64 (m, 3H), 7.39 (d, 1H,
J = 9.5 Hz), 7.19 (d, 1H, J = 7.7 Hz), 6.98 (br s, 1H),
4.81 (t, 1H, J = 6.0 Hz), 3.68 (dt, 2H, J = 5.9,
1.6 Hz), 3.05 (s, 6H), 1.95–1.89 (m, 2H), 1.70–1.58
(m, 4H), 0.91 (m, 9H), 0.07 (m, 6H): 13C NMR
(CDCl3) d 148.6, 138.5, 134.4, 128.7, 126.6, 126.5,
124.5, 124.2, 116.6, 106.6, 74.4, 63.4, 40.9, 36.2, 29.2,
25.9, 18.3, ꢀ5.4: MS (FAB) m/z 373 [M+]; HRMS
(FAB) calculated for C22H35O2NSi [M+] 373.2437;
found: 373.2439.
1
to give 3 (537 mg, 73%) as a yellow solid: H NMR
(CDCl3) d 8.35 (d, 1H, J = 1.6 Hz), 7.93 (dd, 1H,

1620
K. Tanaka et al. / Bioorg. Med. Chem. 15 (2007) 1615–1621
J = 1.8, 8.8 Hz), 7.90 (s, 1H), 7.78 (d, 1H, J = 9.2 Hz),
7.62 (d, 1H, J = 8.8 Hz), 7.16 (dd, 1H, J = 2.6, 9.2 Hz),
6.87 (d, 1H, J = 2.4 Hz), 6.31 (t, 1H, J = 6.6 Hz), 4.80
(br, 2H), 4.64 (t, 2H, J = 6.0 Hz), 4.58 (dt, 1H, J = 3.5,
5.7 Hz), 3.99–3.96 (m, 1H), 3.81 (dd, 1H, J = 4.2,
11.2 Hz), 3.75 (dd, 1H, J = 3.3, 11.0 Hz), 3.31 (t, 2H,
J = 7.0 Hz), 3.10 (s, 6H), 2.56 (ddd, 1H, J = 5.9, 7.14,
13.0 Hz), 2.38–2.31 (m, 3H), 0.91 (d, 18H, J = 1.8 Hz),
0.10 (s, 6H), 0.08 (d, 6H, J = 1.1 Hz): 13C NMR(CDCl3)
d 198.9, 161.2, 159.3, 153.5, 150.1, 137.5, 137.3, 130.6,
130.4, 129.8, 126.0, 125.0, 124.4, 116.1, 115.7, 105.2,
87.5, 83.5, 71.9, 66.2, 62.7, 62.0, 40.8, 40.3, 34.5, 25.9,
25.7, 23.7, 18.3, 17.9, 14.3, ꢀ4.8, ꢀ4.9, ꢀ5.5, ꢀ5.6:
MS (FAB) m/z 734 [M+]; HRMS (FAB) calculated for
C38H58O5N6Si2 [M+] 734.4007; found: 734.4017.
To a solution of the 2-NH2-protected nucleoside (84 mg,
0.15 mmol) in pyridine (3 mL) was added 4,40-dimeth-
oxytrityl chloride (56 mg, 0.17 mmol), and the mixture
was stirred for 2 h at 50 ꢁC. The reaction mixture was
concentrated and purified by column chromatography
on silica gel, eluting with chloroform–methanol (30:1)
to give the tritylated nucleoside (112 mg, 87%) as a yel-
1
low solid: H NMR (DMSO-d6) d 8.57 (s, 1H), 8.44 (s,
1H), 8.16 (s, 1H), 7.87–7.81 (m, 2H), 7.65 (d, 1H,
J = 8.8 Hz), 7.32–7.15 (m, 10H), 6.92 (d, 1H,
J = 2.2 Hz), 6.79–6.74 (m, 4H), 6.37 (t, 1H, J = 6.4 Hz),
5.37 (d, 1H, J = 4.4 Hz), 4.59 (t, 2H, J = 7.0 Hz), 4.44–
4.42 (m, 1H), 3.90–3.95 (m, 1H), 3.69 (s, 6H), 3.26 (t,
2H, J = 7.1 Hz), 3.15 (dd, 1H, J = 4.8, 10.6 Hz), 3.12–
3.11 (m, 1H), 3.09 (s, 3H), 3.04 (s, 6H), 3.00 (s, 3H),
2.87–2.73 (m, 1H), 2.37–2.30 (m, 1H), 2.18 (quintet,
2H, J = 6.6 Hz): 13C NMR (DMSO-d6) d 198.5, 160.0,
158.14, 158.10, 158.06, 153.5, 150.3, 145.0, 139.8, 137.4,
135.71, 135.67, 130.7, 130.0, 129.9, 129.8, 129.7, 127.8,
126.7, 126.0, 124.7, 124.0, 117.2, 116.5, 113.2, 104.9,
86.0, 85.6, 83.0, 79.3, 70.8, 65.5, 64.4, 55.08, 55.07,
40.3, 40.0, 34.7, 34.1, 23.8: MS (FAB) m/z 864
[(M+H)+]; HRMS (FAB) calculated for C50H54O7N7
[(M+H)+] 864.4085; found: 864.4083.
4.1.3. Compound 4. To a solution of 3 (1.66 g, 2.26
mmol) in tetrahydrofuran (20 mL) was added a 1.0 M
solution of tetra-n-butylammonium fluoride in tetrahy-
drofuran (6.77 mL), and the mixture was stirred for
2 h at room temperature. The reaction mixture was con-
centrated and purified by column chromatography on
silica gel, eluting with 30:1 chloroform/methanol to give
1
4 (830 mg, 73%) as a yellow solid: H NMR (CDCl3) d
8.34 (d, 1H, J = 1.5 Hz), 7.92 (dd, 1H, J = 1.8, 8.6 Hz),
7.77 (d, 1H, J = 9.2 Hz), 7.62–7.60 (m, 2H), 7.14 (dd,
1H, J = 2.6, 9.2 Hz), 6.85 (d, 1H, J = 2.6 Hz), 6.21 (dd,
1H, J = 5.7, 9.5 Hz), 4.92 (br, 2H), 4.74 (d, 1H,
J = 5.1 Hz), 4.62 (t, 2H, J = 6.0 Hz), 4.20–4.19 (m,
1H), 3.95 (dd, 1H, J = 1.6, 12.6 Hz), 3.75 (dd, 1H,
J = 1.8, 12.8 Hz), 3.29 (t, 2H, J = 7.1 Hz), 3.09 (s, 6H),
2.99–2.96 (m, 1H), 2.33 (quintet, 2H, J = 6.4 Hz),
2.27–2.22 (m, 1H): 13C NMR (CDCl3) d 199.0,161.7,
158.7,152.2,150.2,138.8,137.6, 130.7,130.3, 130.0,126.1,
125.0,124.4,116.8,116.2,105.2,89.2,87.3,73.1,66.4,63.4,40.3,
34.5,25.3,23.7: MS(FAB) m/z 506 [M+]; HRMS(FAB)
calculated for C26H30O5N6 [M+] 506.2278; found:
506.2285.
To a solution of the tritylated nucleoside (70 mg,
81.0 lmol) in acetonitrile (900 lL) were added 2-cyano-
ethyl tetraisopropyldiphosphoramidite (31 lL, 98.9
lmol) and tetrazole (7 mg, 98.9 lmol). The reaction
mixture was stirred at room temperature for 2 h. The
mixture of 5 was filtered and used with no further
purification.
4.1.5. Oligodeoxyribonucleotide synthesis and character-
ization. Oligodeoxyribonucleotides (ODNs) were syn-
thesized by the conventional phosphoramidite method
by using an Applied Biosystems 392 DNA/RNA synthe-
sizer. Synthesized ODNs were purified by reverse phase
HPLC on a 5-ODS-H column (10· 150 mm, elution
with a solvent mixture of 0.1 M triethylammonium ace-
tate (TEAA), pH 7.0, linear gradient over 30 min from
5% to 20% acetonitrile at a flow rate of 3.0 mL/min).
An aliquot of purified ODN solution was fully digested
with calf intestine alkaline phosphatase (50 U/mL),
snake venom phosphodiesterase (0.15 U/mL), and P1
nuclease (50 U/mL) at 37 ꢁC for 3 h. Digested solution
was analyzed by HPLC on Cosmosil 5C-18AR or
CHEMCOBOND 5-ODS-H column (4.6· 150 mm),
elution with a solvent mixture of 0.1 M TEAA, pH
7.0, linear gradient over 20 min from 0% to 20% aceto-
nitrile at a flow rate of 1.0 mL/min. Concentration of
each ODN was determined by comparing a given peak
area with those of 0.1 mM standard solution containing
dA, dC, dG, and dT. Each ODN was characterized by
MALDI-TOF MS.
4.1.4. Compound 5. A solution of 4 (253 mg, 0.5 mmol)
and N,N-dimethylformamide dimethylacetal (1 mL) in
N,N-dimethylformamide (3 mL) was stirred for 6 h at
room temperature. The reaction mixture was concen-
trated to a yellow oil and purified by column chroma-
tography on silica gel, eluting with 10:1 ethyl acetate–
methanol to give the 2-NH2-protected nucleoside
1
(266 mg, 94%) as a yellow solid: H NMR (DMSO-d6)
d 8.60 (s, 1H), 8.44 (d, 1H, J = 1.1 Hz), 8.27 (s, 1H),
7.86 (d, 1H, J = 9.2 Hz), 7.82 (dd, 1H, J = 1.6, 8.6 Hz),
7.65 (d, 1H, J = 8.8 Hz), 7.24 (dd, 1H, J = 2.7, 9.2 Hz),
6.92 (d, 1H, J = 2.4 Hz), 6.34 (dd, 1H, J = 6.2, 7.7 Hz),
5.29 (br, 1H), 5.09 (br, 1H), 4.59 (t, 2H, J = 6.6 Hz),
4.41–4.39 (m, 1H), 3.88–3.86 (m, 1H), 3.61 (dt, 1H,
J = 4.6, 11.7 Hz), 3.52 (ddd, 1H, J = 4.6, 5.5, 10.8 Hz),
3.25 (t, 2H, J = 7.1 Hz), 3.11 (s, 3H), 3.04 (s, 6H), 2.99
(s, 3H), 2.66 (ddd, 1H, J = 5.9, 7.7, 13.4 Hz), 2.25
(ddd, 1H, J = 2.9, 6.2, 13.2 Hz), 2.17 (quintet, 2H,
J = 6.8 Hz): 13C NMR (DMSO-d6) d 198.5, 161.9,
160.0, 158.3, 153.5, 150.3, 140.0, 137.4, 130.7, 130.0,
129.9, 126.0, 124.7, 124.0, 117.1, 116.5, 104.9, 88.0,
83.6, 79.3, 71.0, 65.4, 62.0, 40.5, 40.0, 34.6, 34.0, 23.8:
MS (FAB) m/z 561 [M+]; HRMS (FAB) calculated for
C29H35O5N7 [M+] 561.2700; found: 561.2692.
4.2. Photophysical and photochemical measurements
4.2.1. Melting temperature (Tm) measurements. All Tms
of the ODNs (2.5 lM, final duplex concentration) were
taken in 50 mM sodium phosphate buffer (pH 7.0) con-
taining 100 mM sodium chloride. Absorbance vs tem-
perature profiles were measured at 260 nm using a

K. Tanaka et al. / Bioorg. Med. Chem. 15 (2007) 1615–1621
1621
Shimadzu UV-2550 spectrophotometer equipped with a
Peltier temperature controller using 1 cm path length
cell. The absorbance of the samples was monitored at
260 nm from 5 to 90 ꢁC with a heating rate of 1 ꢁC/
min. From these profiles, first derivatives were analyzed
to determine Tm values.
picosecond fluorescence lifetime measurement system
(C4780, Hamamatsu). A N2 laser (337 nm) was used
as the excitation light source. The fluorescence emission
was collected at 510 nm.
Acknowledgments
4.2.2. Osmium oxidation of oligodeoxyribonucleotides.
Osmium oxidation and fluorescence measurement of
ODNs were carried out as follows: The ODN (5 lM)
to be examined was incubated in a solution of 5 mM
potassium osmate, 100 mM potassium hexacyanofer-
rate(III), 100 mM bipyridine, and 1 mM EDTA in
100 mM Tris–HCl buffer (pH 7.7) and 10% acetonitrile
at 37 ꢁC for 3 h. ODN samples were purified by reverse
phase HPLC or through gel filters. Recovered ODN
samples were hybridized with D-containing DNA
probes to obtain 2.5 lM duplex solutions in 50 mM
sodium phosphate buffer (pH 7.0) containing 100 mM
sodium chloride. Fluorescence spectra of the duplexes
were taken at 25 ꢁC using a Shimadzu RF-5300PC spec-
trofluorophotometer and 1 cm path length cell. The
excitation bandwidth was 1.5 nm. The emission band-
width was 1.5 nm.
This research was supported by Industrial Technology
Research Grant Program in 2006 from New Energy
and Industrial Technology Development Organization
(NEDO) of Japan.
References and notes
1. Jones, P. L.; Wolffe, A. P. Semin. Cancer Biol. 1999, 9,
339–347.
2. Tate, P. H.; Bird, A. P. Curr. Opin. Genet. Dev. 1993, 3,
226–231.
3. Colot, V.; Rossignol, J. L. BioEssays 1999, 21, 402–411.
4. Feil, R.; Khosla, S. Trend. Genet. 1999, 15, 431–435.
5. Jones, P. A.; Laird, P. W. Nature Genet. 1999, 21, 163–
167.
6. Robertson, K. D. Nat. Rev. Genet. 2005, 6, 597–610.
7. Hayatsu, H.; Wataya, Y.; Kai, K.; Iida, S. Biochemistry
1970, 9, 2858–2866.
8. Frommer, M.; McDonald, L. E.; Millar, D. S.; Collis, C.
M.; Watt, F.; Grigg, G. W.; Molloy, P. L.; Paul, C. L.
Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 1827–1831.
9. Herman, J. G.; Graff, J. R.; Myo¨ha¨nen, S.; Nelkin, B. D.;
Baylin, S. B. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 9821–
9826.
10. Okamoto, A.; Tainaka, K.; Saito, I. Bioconjug. Chem.
2005, 16, 1105–1111.
11. Cohen, B. E.; McAnaney, T. B.; Park, E. S.; Jan, Y. N.;
Boxer, S. G.; Jan, L. Y. Science 2002, 296, 1700–1703.
12. Kimura, T.; Kawai, K.; Majima, T. Org. Lett. 2005, 7,
5829–5832.
13. Lin, K.-Y.; Matteucci, M. D. J. Am. Chem. Soc. 1998, 120,
8531–8532.
14. Ziegel, R.; Shallop, A.; Jones, R.; Tretyakova, N. Chem.
Res. Toxicol. 2003, 16, 541–550.
15. Rajeev, K. G.; Sanjayan, G. J.; Ganesh, K. N. J. Org.
Chem. 1997, 62, 5169–5173.
16. Subbaraman, L. R.; Subbaraman, J.; Behrman, E. J.
Bioinorg. Chem. 1971, 1, 35–55.
17. Cfhang, C.-H.; Ford, H.; Behrman, E. J. Inorg. Chim.
Acta 1981, 55, 77–80.
18. Dizdaroglu, M.; Holwitt, E.; Hagan, M. P.; Blakely, W. F.
Biochem. J. 1986, 235, 531–536.
4.2.3. Fluorescence quantum yields. The emission band-
width was 1.5 nm. The fluorescence quantum yields
(UF) were determined using 9,10-diphenylanthracene as
a reference with a known UF of 0.95 in ethanol accord-
ing to the reference, J. Phys. Chem. 1976, 80, 969–974.
The area of the emission spectrum was integrated using
the software available in the instrument, and the quan-
tum yield is calculated according to the following
equation,
UFðSÞ=UFðRÞ ¼ ½A =AðRÞꢁ ꢂ ½ðAbsÞðRÞ=ðAbsÞðSÞ
ꢁ
ðSÞ
ꢂ ½n2ðSÞ=n2ðRÞ
ꢁ
ð1Þ
Where, UF(S) and UF(R) are the fluorescence quantum
yield of the sample and the reference, respectively. A(S)
and A(R) are the area under the fluorescence spectra of
the sample and the reference, respectively, (Abs)(S) and
(Abs)(R) are the respective optical densities of the sample
and the reference solution at the wavelength of excita-
tion, and n(S) and n(R) are the values of refractive index
for the respective solvents used for the sample (1.333)
and the reference (1.383).
4.2.4. Fluorescence decay measurements. ODN solutions
were prepared as described in the Tm measurement
experiment. Fluorescence decay was measured by a
two-dimensional photon-counting method with the
19. Okamoto, A.; Tainaka, K.; Kamei, T. Org. Biomol. Chem.
2006, 4, 1638–1640.
20. Tornaletti, S.; Pfeifer, G. P. Oncogene 1995, 10, 1493–
1499.