Synthesis and characterization of oligodeoxynucleotides containing a novel tetraazabenzo[cd]azulene:naphthyridine base pair

Article abstract of DOI:10.1016/j.bmc.2010.11.023

The synthesis and thermal stability of oligodeoxynucleotides (ODNs) containing 4-amino-2,3,5,6-tetraazabenzo[cd]azulen-7-one nucleosides 5 (BaO ^N) with the aim of developing new base pairing motif is described. The tricyclic nucleoside 5 was pr

Full text of DOI:10.1016/j.bmc.2010.11.023

Bioorganic & Medicinal Chemistry 19 (2011) 352–358
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and characterization of oligodeoxynucleotides containing a novel
tetraazabenzo[cd]azulene:naphthyridine base pair
Yasuyuki Hirama a, , Noriaki Minakawa b, , Akira Matsuda a,
⇑
a Faculty of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
b Graduate School of Pharmaceutical Sciences, The University of Tokushima, Shomachi 1-78-1, Tokushima 770-8505, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and thermal stability of oligodeoxynucleotides (ODNs) containing 4-amino-2,3,5,6-tetra-
azabenzo[cd]azulen-7-one nucleosides 5 (BaON) with the aim of developing new base pairing motif is
described. The tricyclic nucleoside 5 was prepared starting with the 7-deaza-7-iodopurine derivative 1
via a palladium catalyzed cross-coupling reaction with methyl acrylate, followed by an intramolecular
cyclization. The resulting nucleoside was incorporated into ODNs, and the base pairing property of the
BaON:NaNO (2-amino-7-hydroxy-1,8-naphthyridine nucleoside) pair in the duplex was evaluated by a
thermal denaturation study. The melting temperature (Tm) of the duplex containing the BaON:NaNO pair
showed a higher value than that of the duplexes containing the adenine:thymine (A:T) and the
guanine:cytosine (G:C) pairs, however it was lower than that of the ImON:NaNO (ImON = 7-amino-
imidazo[50,40:4,5]pyrido[2,3-d]pyrimidin-4(5H)-one nucleoside) pair. A temperature-dependent 1H NMR
study revealed that the H-bonding ability of BaON was lower than that of ImON, which would explain
why the BaON:NaNO pair was less thermally stable than the ImON:NaNO pair.
Received 28 October 2010
Revised 5 November 2010
Accepted 6 November 2010
Available online 11 November 2010
Keywords:
Tetraazabenzo[cd]azulene
Hydrogen bond
Nucleobase
Nucleoside
Oligodeoxynucleotide
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
As the first generation of such base pairing motifs, we have designed
complementary pairs consisting of the imidazopyridopyrimidine
In a right-handed double helix of DNA, nucleobases form
Watson–Crick base pairs of adenine:thymine (A:T) and guan-
ine:cytosine (G:C) via two and three hydrogen bonds (H-bonds),
respectively. These selective base pairs play a critical role in not only
conserving and transmitting genetic information but also maintain-
ing the stability of the helical structure. In general, three factors are
thought to contribute to its stability: first, the interaction via
H-bonds between the complementary bases; second, the aromatic
(Im) bases such as ImON:ImNO (Fig. 1a).6,7 Although this base pair
was demonstrated to stabilize the helical structure by means of four
H-bonds and increased stacking interaction due to the expanded
aromatic surface (positive effects), distortion of the helical structure
arising from extension of the intrastrand C10–C10 distance (a nega-
tive effect) was also suggested. Since the ImON:ImNO pair was
thought to not satisfy the shape complementarity, new C-nucleoside
derivatives possessing 1,8-naphthyridines (Na), which are comple-
mentary to the Im bases, were designed (Fig. 1b). The resulting base
pairs, such as ImON:NaNO, markedly stabilized the duplexes ther-
mally, independent of the mode of incorporation of the pairs into
the oligodeoxynucleotides (ODNs) (ca. +8 °C per pair relative to
the A:T pair).7–9 Furthermore, we have succeeded in demonstrating
a selective recognition of the Im:Na pairs by DNA polymerases.10,11
Therefore, the Im:Na base pair appears to be promising as an alter-
native and highly thermally stable base pair beyond the Watson–
Crick base pairs. However, the question has been raised as to
whether the Im:Na pair has the best geometry in the double helix
from the standpoint of shape complementarity. Thus, the Im bases
can be considered as ring-expanded analogs of purine toward the
minor groove direction, while the Na bases are ring-expanded ana-
logs of pyrimidine toward the major groove direction, which may
cause a distortion of the helical structure arising from a shift of the
intrastrand C10–C10 position. In order to avoid this possible
P–P interaction between stacking base pairs (stacking interaction);
and third, the size of the base pairs (shape complementarity) since
Watson–Crick base pairs consist of a combination of purine and
pyrimidine bases.
Thus far, a number of noncanonical base paring motifs have been
designed based on the aforementioned factors in order to apply
them to bioorganic chemistry, biotechnology, and medicinal chem-
istry.1–5 Although the end goal of each base pairing motif is different,
stable and specific interactions between the bases are indispensable
for the development of a new base paring motif. In order to stabilize
and regulate DNA structure, we have been working on a project to
develop new base paring motifs consisting of four H-bonds (Fig. 1).
⇑
Corresponding author. Tel./fax: +81 88 633 7288.
E-mail address: minakawa@ph.tokushima-u.ac.jp (N. Minakawa).
Tel.: +81 11 706 3228; fax: +81 11 706 4980.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.11.023

Y. Hirama et al. / Bioorg. Med. Chem. 19 (2011) 352–358
353
Figure 1. Base pairing motifs of: (a) ImON:ImNO, (b) ImON:NaNO, (c) newly
designed BaON:NaNO, and (d) A:T. To clarify the intrastrand C10–C10 position, the
C10 positions of each base pair were marked with dot.
distortion, we envisioned a new base paring motif in which both
bases are expanded toward the major groove direction. As one of
the candidates, we designed a nucleoside derivative possessing a
4-amino-2,3,5,6-tetraazabenzo[cd]azulen-7-one (BaON),7 which is
expected to form a base pair with NaNO with four H-bonds
(Fig. 1c). As can be seen by comparison with a natural base pair such
as A:T (Fig. 1d), the proposed base pair, BaON:NaNO, is expected to
have a similar intrastrand C10–C10 position.
In this work, we describe the synthesis of the nucleoside unit 5
(BaON) and the thermal stabilities of the duplex containing the
newly designed base pairing motif, BaON:NaNO.
Scheme 1. Reagents and conditions: (a) methyl acrylate, Et3N, (PPh3)2PdCl2, DMF,
70 °C, 62%; (b) 0.1 M NaOMe in MeOH, 70 °C, 60%; (c) TMSCl, pyridine then
t-butylphenoxyacetyl chloride, 93%; (d) DMTrCl, pyridine, 88%; (e) 2-cyanoetyl
N,N-diisopropylchlorophosphoramidite, iPr2NEt, CH2Cl2, 78%; (f) TBSCl, imidazole,
DMF, 69%.
2. Results and discussion
2.1. Chemistry
We recently reported the synthesis of the 4,7-amino-2,3,5,
6-tetraazabenzo[cd]azulene nucleoside 4 from 1 via 2, and its
spontaneous hydrolysis to the 4-amino-2,3,5,6-tetraazabenzo[cd]
azulen-7-one nucleoside (BaON) 5, the desired compound in this
study.12 To prepare 5 more conveniently, we decided to synthesize
it via 3 (Scheme 1). Thus, the palladium catalyzed cross-coupling
reaction of the 7-iodo derivative 1 with methyl acrylate was first
investigated. The reaction was carried out in DMF at 70 °C in the
presence of a Pd catalyst [Pd(PPh3)4, (dba)3Pd2ꢀCHCl3, (PhCN)2PdCl2,
Pd(OAc)2, or (PPh3)2PdCl2] with or without CuI. The best result was
obtained when (PPh3)2PdCl2 was used without CuI to give 3 in 62%
yield. The resulting 3 was subsequently treated with NaOMe in
MeOH to give 5 in 60% yield, the spectral data of which were identi-
cal with those reported in our previous paper.12
Thus far, only studies directed at the synthesis of nucleoside
derivatives possessing the 2,3,5,6-tetraazabenzo[cd]azulene skele-
ton as a nucleobase have been reported.13,14 However, no reports
on ODNs containing such tricyclic artificial nucleobases have ap-
peared. Since 5 has a lactam moiety, the stability of 5 under the
conditions of ODN synthesis was examined prior to ODN synthesis.
In the general protocol for ODN synthesis, the elongating ODN on a
controlled pore glass (CPG) support is exposed to acidic and oxida-
tive conditions, the former to remove the DMTr group and the lat-
ter to oxidize the phosphorus atom. The resulting ODN is then
treated with basic conditions to remove all the protecting groups
and detach the ODN from the CPG support. Thus, 5 was treated
under the above conditions to check its stability, and it was shown
to be stable against both 80% aqueous AcOH and 0.03 M I2 in THF/
pyridine/H2O (8:1:1) (data not shown).15 On the other hand, 5 was
found to be sensitive under the basic conditions used in the usual
ODN synthesis. Therefore, 5 was treated with concentrated NH4OH
at 55 °C, and its stability was analyzed by HPLC. As a result, decom-
position of 5 was observed and its half-life (t1/2) was estimated to
be 40 min (Fig. 2a). Even when the reaction temperature was de-
creased to room temperature, the estimated t1/2 (5 h) was still
insufficient because treatment for approximately 12 h under these
conditions is required for ODN synthesis (Fig. 2b). Treatment with
0.05 M K2CO3 in MeOH is also applicable to ODN synthesis if more
labile protecting groups are used for the protection of exocyclic
amino groups of adenine, guanine, and cytosine moieties (termed
ultramild phosphoramidites). To our delight, 5 was sufficiently sta-
ble under these conditions (Fig. 2c).
In order to incorporate 5 into ODNs, the exocyclic amino group
of 5 must be protected with an appropriate protecting group,
which can be removed with 0.05 M K2CO3 in MeOH. Sinha et al.
reported using a t-butylphenoxyacetyl (tBPA) group for the protec-
tion of the exocyclic amino groups, which was removed under the
aforementioned conditions.16 Hence, we planned to introduce this
protecting group on 5. Compound 5 was first treated with trimeth-
ylsilyl chloride (TMSCl) in pyridine for selective protection of the 30
and 50-hydroxyl groups. Treatment with tBPA-Cl followed by

354
Y. Hirama et al. / Bioorg. Med. Chem. 19 (2011) 352–358
were dissolved in acetonitrile to be loaded on to the DNA/RNA syn-
thesizer. However, an acetonitrile solution of 8 was found to be too
viscous, necessitating use of dichloromethane as solvent. The fully
protected ODNs (1 lmol scale) linked to the CPG support were
prepared by the usual phosphoramidite method. Then, the result-
ing ODNs were treated with 0.05 M K2CO3 in MeOH for 12 h at
room temperature, followed by C-18 column chromatography.
The ODNs were detritylated with 80% aqueous AcOH, and each
ODN obtained showed a single peak on reverse-phase HPLC. Each
ODN was completely hydrolyzed to the corresponding nucleosides
by a mixture of nuclease P1, snake venom phosphodiesterase, and
alkaline phosphatase, and the nucleoside composition was ana-
lyzed by HPLC. In the case of ODN I and ODN III (X = BaON), the
peak corresponding to the tricyclic nucleoside 5, confirmed by
co-elution with an authentic sample, was clearly observed, and
the nucleoside composition calculated from the areas of the peaks
supported the structure of each (Supplementary data). On the
other hand, no peak corresponding to 5 was observed when the
resulting ODN V0 (which was not desired ODN V and termed as
ODN V0) was hydrolyzed. The matrix-assisted laser desorption/ion-
ization time-of-flight mass (MALDI-TOF/MS) spectrum of ODN V0
suggested that three tBPA groups still remained on its structure
(data not shown). In the case of ODNs I and III, the tBPA groups
were removed by treatment with 0.05 M K2CO3 in MeOH at room
temperature. However, the tBPA groups on three consecutive BaON
residues like ODN V0 were resistant under the same conditions.
Therefore, the ODN linked to the solid support was treated with
0.05 M K2CO3 in MeOH at 55 °C for 8 h, and the resulting ODN V
was hydrolyzed and analyzed by HPLC. The peak corresponding
to 5 was observed, and the nucleoside composition calculated from
the areas of the peaks supported the structure of ODN V.
Figure 2. HPLC profiles of treatment of 5 with various basic conditions.
aqueous workup provided 6 in 93% yield. Since it is necessary to
check whether the tBPA group can be removed without decompo-
sition, 6 was treated with 0.05 M K2CO3 in MeOH, and the tBPA
group was removed smoothly without decomposition of the lac-
tam moiety (see Supplementary data). Protection of the 50-hydro-
xyl group with a dimethoxytrityl (DMTr) group and subsequent
treatment of the resulting 7 with N,N-diisopropylchlorophosph-
oramidite in the presence of Hünig’s base afforded the phospho-
ramidite unit 8 (Scheme 1).
2.3. Thermal stability of the DNA duplexes containing
BaON:NaNO pairs
The thermal stability of the duplexes formed by ODN I:ODN II
(containing one X:Y pair), ODN III:ODN IV (containing three
nonconsecutive X:Y pairs), and ODN V:ODN VI (containing three
consecutive X:Y pairs) was measured by thermal denaturation in
a buffer of 10 mM sodium cacodylate (pH 7.0) containing 1 mM
NaCl. Each profile of the thermal denaturation showed a single
transition corresponding to a helix-to-coil transition to give a melt-
ing temperature (Tm). For comparison, Tm values of DNA duplexes
containing natural A:T and G:C pair(s) and noncanonical ImON:
NaNO pair(s) along with those of BaON:NaNO pair(s) were
2.2. Synthesis of oligodeoxynucleotides
We next prepared ODNs on a DNA/RNA synthesizer. In one ser-
ies in our studies, we used three sets of complementary 17mers
(ODN I:ODN II, ODN III:ODN IV, and ODN V:ODN VI) (Table 1) to
investigate the thermal stability of the duplexes containing new
base pairing motifs.8 Thus, the BaONs were incorporated in the X
position of ODNs I, III, and V. In general, the phosphoramidite units
measured under the same conditions. The resulting Tms and
DTms
were calculated based on the Tm of the duplex (X:Y = A:T, common
to ODN I:ODN II, ODN III:ODN IV, and ODN V:ODN VI) and are
listed in Table 1. When one A:T pair of ODN I:ODN II was replaced
with a G:C pair, the Tm value was increased to 48.5 °C from 47.2 °C
Table 1
Sequences of ODNs and hybridization dataa
(D
Tm = 1.3 °C), which arises from a two to three increase in
a
b
H-bonds. The substitution of X:Y into the ImON:NaNO pair remark-
Duplex
X
Y
Tm
DTm
(°C)
(°C)
ably increased the Tm value to 55.3 °C (
D
Tm = 8.1 °C) as reported
A
G
T
C
47.2
48.5
55.3
52.0
—
ODN I
ODN II
50-GCACCGAAXAAACCACG-30
30-CGTGGCTTYTTTGGTGC-50
+1.3
+8.1
+4.8
ImON
BaON
NaNO
NaNO
Table 2
Measurement of stacking abilities of each nucleobase
a
b
G
C
55.1
79.8
65.2
+7.9
+32.6
+18.0
Duplex
Z
Tm (°C)
D
Tm (°C)
D
Tm/Zb (°C)
ODN III
ODN IV
50-GCXCGAAXAAACCXCG-30
30-CGYGGCTTYTTTGGYGC-50
ImON
BaON
NaNO
NaNO
None
T
40.8
49.1
54.9
57.4
61.1
–
–
+4.2
+8.3
+14.1
+16.6
+20.3
G
C
55.2
77.7
65.7
+5.0
+30.5
+18.5
A
+7.1
+8.3
+10.2
Im-ON
Ba-ON
ODN V
ODN VI
50-GCACCGAXXXAACCACG-30
30-CGTGGCTYYYTTGGTGC-50
ImON
BaON
NaNO
NaNO
a
a
Experimental conditions are described in the Section 3. The data presented are
average of triplicates.
Experimental conditions are described in Section 3. The data presented are
averages of triplicates.
b
b
The
DTm values were obtained by subtracting data for the Tm possessing
The
DTm values were obtained by subtracting data for the Tm possessing
X:Y = A:T from that for each duplex.
Z = none from that for each duplex.

Y. Hirama et al. / Bioorg. Med. Chem. 19 (2011) 352–358
355
previously.8 This elevation was thought to be due to the effects of:
(1) an increase of H-bonds from two to four; (2) an increase of
stacking ability of the expanded aromatic surface; and (3) a suit-
able distance between the base pair at the C10 (purine:pyrimidine
type). As explained in the Introduction, we designed a BaON:NaNO
pair to enhance the duplex stability by adopting a better geometry
of the intrastrand C10–C10 position than an ImON:NaNO pair.
Contrary to our expectation, the Tm value of ODN I:ODN II contain-
ing a BaON:NaNO pair was lower than that of the ImON:NaNO pair
(52.0 °C vs. 55.3 °C), although the BaON:NaNO pair was more stable
ability of the nucleobases. When ImON was introduced into the Z
position, the Tm value was increased to 57.4 °C (DTm = 8.3 °C/base),
which was higher than those of the natural nucleobases.6 Contrary
to our speculation, the duplex possessing BaON in the Z position
added +10.2 °C/base of Tm, which was the highest among the
nucleobases examined. Accordingly, it was revealed that the lower
thermal stability of a BaON:NaNO pair compared with that of an
ImON:NaNO pair was not due to their stacking ability.
As an alternative explanation, we considered the difference of
H-bonding ability between BaON and ImON. In our previous paper,6
we showed that ImON formed an anti-parallel homodimer via four
H-bonds in a nucleoside level.19 Since BaON is also expected to
form an anti-parallel homodimer via four H-bonds, we compared
its ability to form the homodimer to that of ImON by a tempera-
ture-dependent 1H NMR study. The results of ImON are shown in
Figure 3a.6 Thus, the amide proton was observed at around
12.6 ppm as a broad singlet, and the amino protons were observed
at around 7.0 ppm as a single coalesced signal at 60 °C. These re-
sults showed that ImON prefers to exist as a monomer at that tem-
perature. On the other hand, the amide proton signal was shifted
downfield with decreasing temperature. In addition, the amino
protons were split into two sets of broad singlets. These spectral
changes indicated formation of the expected anti-parallel homodi-
mer, which had already been observed at 0 °C and more clearly at
than the A:T (DTm = 4.8 °C) and G:C (DTm = 3.5 °C) pairs. The same
tendency was observed in the results of ODN III:ODN IV and ODN
V:ODN VI. Hence, the newly designed base pairing motif BaON:NaNO
stabilized the duplex via four H-bonds irrespective of the mode of
incorporation,17 although its stabilizing effect was smaller than
that of the ImON:NaNO pair. In order to understand why the BaON:
NaNO pair is less thermally stable than the ImON:NaNO pair in the
DNA duplex, two experiments were carried out, as discussed in the
next section.
2.4. Considerations as to why the BaON:NaNO pair is less
thermally stable than the ImON:NaNO pair
As one of the possible explanations of the above results, the
BaON may have a weaker stacking ability than the ImON in the
DNA duplex. To examine this possibility, the stacking ability of
BaON was estimated. According to the method reported by Guckian
et al.,18 a series of self-complementary ODN VII, where Z was added
at the 50-end of each paired duplex (dangling end), were prepared,
and the Tm values were determined in a buffer of 0.01 M sodium
phosphate (pH 7.0) containing 1 M NaCl (Table 2). As a comparison,
the Tm data for non-dangling and natural bases (T and A) at the
dangling end are listed (Table 2). The two unpaired Ts and As
added +8.3 °C (+4.2 °C/base) and +14.1 °C (+7.1 °C/base) to the
Tms, respectively, relative to the control duplex (Z = none). These
increments of Tm values can be estimated as their own stacking
ꢁ60 °C. Based on these observations, a temperature-dependent 1
H
NMR study of BaON was carried out. First, the hydroxyl groups of 5
were silylated with tert-butyldimethylsilyl chloride to give 9
(Scheme 1). 1H NMR measurements of 9 were conducted at vari-
able temperatures in CDCl3 at 25 mM concentrations to observe
the anti-parallel homodimer formation. As can be seen in Figure
3b, the amide proton was observed at around 9.0 ppm as a broad
singlet, and the amino protons were observed at around 5.0 ppm
as a single coalesced signal at 60 °C. These results indicated the
existence of BaON as a monomer, as was the case for ImON. In addi-
tion, the amide and amino proton signals were shifted downfield
with decreasing temperature. Finally, the amino protons were split
Figure 3. Partial 1H NMR spectra obtained in the course of carrying out a temperature-dependent measurements of: (a) ImON and (b) BaON at 25 mM in CDCl3.

356
Y. Hirama et al. / Bioorg. Med. Chem. 19 (2011) 352–358
into two sets of broad singlets at around 8.5 and 5.0 ppm at ꢁ60 °C.
From these results, it was confirmed that BaON also formed an anti-
parallel homodimer in CDCl3 via four H-bonds. However, clear
differences between ImON and BaON were observed in the spectra
at 30, 0, and ꢁ30 °C. For example, the homodimer formation of
ImON has already been observed at 0 °C as described above. In
contrast, BaON existed preferentially as a monomer at the same
temperature, because the amino protons were still observed at
around 6.0 ppm as a single coalesced signal. Since the ImON ring
system has fourteen electrons and satisfies Hückel’s rule, this ring
system would be planer. In contrast, the BaON ring system has thir-
teen electrons, which does not satisfy Hückel’s rule. Therefore, the
BaON ring system would probably not be planer. These results and
considerations indicated that the H-bonding ability of BaON would
be weaker than that of ImON and would support the lesser thermal
stability of a BaON:NaNO pair than an ImON:NaNO pair.
In conclusion, we have designed a new base pairing motif
consisting of a BaON:NaNO pair to thermally stabilize the DNA
duplex. The corresponding BaON nucleoside was prepared starting
from the 7-deaza-7-iodopurine derivative 1 via a palladium cata-
lyzed cross-coupling reaction with methyl acrylate, followed by
intramolecular cyclization. After the incorporation of BaON into the
ODNs, the thermal stability of the resulting duplex containing the
BaON:NaNO pair(s) was evaluated. The BaON:NaNO pair thermally
stabilized the duplexes better than the A:T and G:C pairs, however
thestabilizingeffect was less thanthat ofthe ImON:NaNO pair, which
we previously developed.8 A temperature-dependent 1H NMR study
revealed that the H-bonding ability of BaON was weaker than that of
ImON, which would explain why the BaON:NaNO pair was less ther-
mally stablethan theImON:NaNO pair. In order toclarify whetherthe
BaON and NaNO units form base pairs with four H-bonds in the du-
plex and whether the resulting BaON:NaNO pair adopts the intra-
strand C10–C10 position similar to natural A:T pair, an NMR study is
required, which is now under investigation.
1H, J = 5.7 and 8.3 Hz), 6.36 (br s, 2H, exchangeable with D2O), 6.30
(d, 1H, J = 15.5 Hz), 5.82 (br s, 2H, exchangeable with D2O), 5.22 (d,
1H, J = 3.5 Hz, exchangeable with D2O), 5.02 (t, 1H, J = 5.8 Hz,
exchangeable with D2O), 4.31 (m, 1H), 3.77 (m, 1H), 3.68 (s, 3H),
3.55 and 3,49 (each m, each 1H), 2.38 (ddd, 1H, J = 8.3, 5.5, and
13.2 Hz), 2.09 (ddd, 1H, J = 5.7, 2.9, and 13.2 Hz); 13C NMR (DMSO-
d6) d 167.1, 160.0, 158.5, 154.3, 133.1, 120.8, 114.0, 112.1, 92.4,
87.2, 82.3, 71.0, 62.0, 51.1, 39.5; Anal. Calcd for C15H19N5O5: C,
51.57; H, 5.48; N, 20.05. Found: C, 51.71; H, 5.42; N, 20.02.
3.3. 4-Amino-2-(2-deoxy-b-D-erythro-pentofuranosyl)-2,6-dihy
dro-7H-2,3,5,6-tetraazabenzo[cd]azulen-7-one (5)12
A solution of 3 (255 mg, 0.73 mmol) in 0.1 M NaOMe in MeOH
(73 mL) was heated at 70 °C for 12 h. The reaction mixture was
cooled to 0 °C, and the resulting precipitate was corrected and
washed with MeOH to give 5 (89 mg, 38%) as yellow solid. The
filtrate was removed in vacuo, and the residue was purified by
an aminosilica gel column, eluted with MeOH in CHCl3 (5–25%),
to give additional 5 (50 mg, 22%). The physical data of 5 were iden-
tical with those of reported previously: UV kmax (H2O) 392 nm
(
e
= 1170), 301 nm (
HCl) 311 nm ( = 6820), 279 nm (
kmax (0.5 M NaOH) 243 nm (
= 22,100); 1H NMR (DMSO-d6) d
e
= 8830), 242 nm (
e
= 33,300); kmax (0.5 M
e
e
= 10,600), 243 nm (
e
= 33,000);
e
10.08 (d, 1H, J = 1.2 Hz, exchangeable with D2O), 7.34 (s, 1H),
6.93 (d, 1H, J = 12.0 Hz), 6.29 (dd, 1H, J = 5.9 and 7.9 Hz), 6.25 (br
s, 2H, exchangeable with D2O), 5.58 (d, 1H, J = 11.6 Hz), 5.26 (d,
1H, J = 3.6 Hz, exchangeable with D2O), 5.02 (t, 1H, J = 5.8 Hz,
exchangeable with D2O), 4.30 (m, 1H), 3.77 (m, 1H), 3.49 (m,
2H), 2.49 (m, 1H), 2.12 (ddd, 1H, J = 6.1, 2.3, and 12.8 Hz).
3.4. 4-(4-tert-Butylphenoxyacetylamino)-2-(2-deoxy-b-D-ery
thro-pentofuranosyl)-2,6-dihydro-7H-2,3,5,6-tetraazabenzo
[cd]azulen-7-one (6)
3. Experimental section
3.1. General methods
To a suspension of 5 (400 mg, 1.26 mmol) in pyridine (25 mL)
was added chlorotrimethylsilane (0.97 mL, 7.6 mmol), and the
reaction mixture was stirred at room temperature. After being stir-
red for 1 h, tert-butylphenoxyacetyl chloride (0.39 mL, 1.9 mmol)
was added to the mixture, and the whole was stirred for additional
1 h. The reaction was quenched by addition of H2O, and the solvent
was removed in vacuo. The residue was purified by a silica gel col-
umn, eluted with MeOH in CHCl3 (2–5%), to give 6 (595 mg, 93%) as
a yellow solid. An analytical sample was crystallized from MeOH to
give pale yellow crystals: mp 153 °C (decomp.); FAB-LRMS m/z 508
[(M+H)+]; 1H NMR (DMSO-d6) d 10.74 (br s, 1H, exchangeable with
D2O), 10.31 (br s, 1H, exchangeable with D2O), 7.66 (s, 1H), 7.28 (d,
2H, J = 8.6 Hz), 7.04 (d, 1H, J = 12.2 Hz), 6.83 (d, 2H, J = 8.6 Hz), 6.40
(dd, 1H, J = 6.1 and 7.9 Hz), 5.68 (d, 1H, J = 12.2 Hz), 5.31 (d, 1H,
J = 4.3 Hz, exchangeable with D2O), 5.08 (s, 2H), 4.92 (t, 1H,
J = 4.3 Hz, exchangeable with D2O), 4.34 (m, 1H), 3.82 (m, 1H),
3.52 (m, 2H), 2.44 (ddd, 1H, J = 7.9, 5.5, and 13.4 Hz), 2.20 (ddd,
1H, J = 6.1, 3.0, and 13.4 Hz), 1.23 (s, 9H); 13C NMR (CDCl3) d
166.8, 156.0, 154.3, 153.2, 153.0, 143.5, 134.2, 126.4, 122.9,
121.0, 114.3, 113.2, 102.2, 87.7, 83.2, 71.1, 67.9, 62.0, 39.9, 34.1,
31.6; Anal. Calcd for C26H29N5O6ꢀ0.5H2O: C, 60.46; H, 5.85; N,
13.56. Found: C, 60.20; H, 5.74; N, 13.40.
Physical data were measured as follows: melting points are
uncorrected. 1H and 13C NMR spectra were recorded at 270 or
500 MHz and 67.5 or 125 MHz instruments in CDCl3 or DMSO-d6
as the solvent with tetramethylsilane as an internal standard.
Chemical shifts are reported in parts per million (d), and signals
are expressed as s (singlet), d (doublet), t (triplet), q (quartet), m
(multiplet), or br (broad). All exchangeable protons were detected
by addition of D2O. TLC was done on Merck Kieselgel F254 pre-
coated plates. Silica gel used for column chromatography was
YMC Gel 60A (70–230 mesh). Iatrobeads used for column chroma-
tography was 6RS-8090 (Mitsubishi Chemical Medience Co.).
Aminosilica gel was Chromatorex NH-DM1020 (Fuji Silysia Chem-
ical Ltd).
3.2. 2,4-Diamino-5-[(E)-1-(methoxycarbonyl)-2-ethenyl]-7-(2-
deoxy-b-D-erythro-pentofuranosyl)pyrrolo[2,3-d]pyrimidine (3)
To a solution of 1 (391 mg, 1.0 mmol) in DMF (10 mL) including
Et3N (0.28 mL, 2.0 mmol) and (PPh3)2PdCl2 (70 mg, 0.1 mmol) was
added methyl acrylate (1.8 mL, 20 mmol), and the reaction mixture
was heated at 70 °C for 5 h. The solvent was removed in vacuo, and
the residue was purified by an aminosilica gel column, eluted with
MeOH in CHCl3 (3%), to give 3 (215 mg, 62%) as a yellow solid. An
analytical sample was crystallized from MeOH to give pale yellow
crystals: mp 195–196 °C; FAB-LRMS m/z 350 [(M+H)+]; 1H NMR
(DMSO-d6) d 7.82 (d, 1H, J = 15.5 Hz), 7.71 (s, 1H, H-6), 6.36 (dd,
3.5. 4-(4-tert-Butylphenoxyacetylamino)-2-(2-deoxy-5-O-4,
40-dimethoxytrityl-b-
D-erythro-pentofuranosyl)-2,6-dihydro-
7H-2,3,5,6-tetraazabenzo[cd]azulen-7-one (7)
To a solution of 6 (600 mg, 1.2 mmol) in dry pyridine (6 mL)
was added DMTrCl (1.8 mmol, 610 mg), and the reaction mixture

Y. Hirama et al. / Bioorg. Med. Chem. 19 (2011) 352–358
357
was stirred for 1 h at room temperature. The reaction was
quenched by addition of ice. The mixture was concentrated in va-
cuo. The residue was diluted with AcOEt, which was washed with
H2O and saturated aqueous NaHCO3, followed by brine. The organ-
ic layer was dried (Na2SO4) and concentrated in vacuo. The residue
was purified by a silica gel column, eluted with CHCl3/acetone
(19:1–9:1), to give 7 (840 mg, 88%) as a yellow solid: FAB-LRMS
m/z 810 [(M+H)+]; FAB-HRMS calcd for C47H47N5O8 809.3425,
found 810.3509 [(M+H)+]; 1H NMR (CDCl3) d 9.25 (br s, 1H,
exchangeable with D2O), 8.95 (br s, 1H, exchangeable with D2O),
7.41 (d, 2H, J = 7.3 Hz), 7.31-7.20 (m, 10H), 6.88 (d, 2H,
J = 8.6 Hz), 6.68 (d, 4H, J = 9.2 Hz), 6.68 (t, 1H, J = 6.4 Hz), 6.45 (d,
1H, J = 11.6 Hz), 5.72 (dd, 1H, J = 11.6 and 1.8 Hz), 4.77 (m, 1H),
4.63 (br s, 2H), 4.19 (m, 1H), 3.76 (s, 6H), 3.51 (br s, 1H, exchange-
able with D2O), 3.45 (dd, 1H, J = 3.7 and 10.4 Hz), 3.40 (dd, 1H,
J = 3.7 and 10.4 Hz), 2.59 (m, 2H), 1.28 (s, 9H); 13C NMR (CDCl3) d
166.8, 158.7, 155.1, 153.6, 153.0, 152.8, 147.4, 145.0, 144.7,
139.6, 135.8, 135.7, 133.8, 130.3, 129.3, 128.4, 128.0, 128.0,
127.9, 127.2, 127.1, 126.6, 122.7, 120.7, 114.5, 113.5, 113.3,
103.3, 86.8, 86.3, 83.9, 77.4, 72.4, 68.1, 63.9, 55.4, 55.4, 41.3,
34.3, 31.6.
3.8. Stability of 5 under basic conditions (experiments shown in
Fig. 2)
Compound of 5 (ca. 1 mg) was dissolved in concentrated
NH4OH (2 mL) in a vial tube, and the solution was heated at
55 °C (or kept at room temperature). An aliquot of the reaction
mixture (25 mL) at the appropriate time was diluted with 0.1 M
TEAA buffer (pH 7.0) (450 mL), and this solution was analyzed by
HPLC, using a J’sphere ODN M80 column (4.6 ꢂ 250 mm, YMC)
with a linear gradient of acetonitrile (from 9.5% to 41% over
20 min) in 0.1 M TEAA buffer (pH 7.0). In the case of basic condi-
tions shown in Figure 3c, an aliquot of the reaction mixture
(25 mL) was diluted with 0.1 M TEAA buffer (pH 7.0)(450 mL)
and 0.05 M aqueous AcOH (25 mL), and this solution was analyzed
by HPLC under the same conditions.
3.9. Synthesis of ODNs I, III, V and VII containing BaON
ODNs were synthesized on
a DNA synthesizer (Applied
Biosystem Model 3400) using BaON phosphoramidite unit 8, and
commercially available deoxyribonucoside phosphoramidite units
(phenoxyacetyl protected dA, acetyl protected dC, and isopropyl-
phenoxyacetyl protected dG) at 1 lmol scale following the standard
3.6. 4-(4-tert-Butylphenoxyacetylamino)-2-{2-deoxy-3-O-
[(N,N-diisopropylamino)-2-cyanoethoxyphosphino]-5-O-4,40-
dimethoxytrityl-b-D-erythro-pentofuranosyl}-2,6-dihydro-7H-
procedure. For the incorporation of BaON into the ODNs, a 0.12 M
solution in dry CH2Cl2, and coupling time of 10 min was used. After
completion of synthesis, the CPG support was then treated with
0.05 M K2CO3 in MeOH for 12 h at room temperature. In the case
of ODN V (containing three consecutive BaON), the CPG support
was treated with 0.05 M K2CO3 in MeOH at 55 °C for 8 h. After filtra-
tion of theCPG support, the released ODN protectedby a DMTr group
at the 50-end was chromatographed on a C-18 silica gel column
(1 ꢂ 13 cm) with a linear gradient of acetonitrile (from 0 to 50%) in
0.1 M TEAA buffer (pH 7.0). The fractions were concentrated, and
the residue was treated with aqueous 80% AcOH at room tempera-
ture for 10 min, then the solution was concentrated, and the residue
was coevaporated with H2O. The residue was dissolved in H2O and
the solution was washed with AcOEt, then the H2O layer was con-
centrated to give a deprotected ODN. The ODN was further purified
2,3,5,6-tetraazabenzo[cd]azulen-7-one (8)
To a solution of 7 (700 mg, 0.86 mmol) in dry CH2Cl2 (9 mL)
were added N,N-diisopropylethylamine (600 mL, 3.5 mmol) and
2-cyanoethyl N,N-diisopropylchlorophosphoramidite (580 mL,
2.6 mmol), and the reaction mixture was stirred for 30 min at room
temperature. The reaction was quenched by addition of ice. The
reaction mixture was diluted with AcOEt, which was washed with
H2O (twice) and saturated brine. The organic layer was dried
(Na2SO4) and concentrated in vacuo. The residue was purified by
a silica gel column, eluted with hexane/AcOEt (3:2–1:1), to give
8 (680 mg, 78%) as a yellow solid: FAB-LRMS m/z 1010 [(M+H)+];
FAB-HRMS calcd for C56H64N7O9P 1009.4503, found 1010.4598
[(M+H)+]; 31P NMR (CDCl3) d 149.3 and 149.1.
by reverse-phase HPLC, using
a J’sphere ODN M80 column
(4.6 ꢂ 150 mm, YMC) with linear gradient of acetonitrile (from 5
to 20% over 20 min) in 0.1 M TEAA buffer (pH 7.0) to give a highly
purified ODNs.
3.7. 4-Amino-2-(2-deoxy-3,5-di-O-tert-butyldimethylsilyl-b-D-
erythro-pentofuranosyl)-2,6-dihydro-7H-2,3,5,6-tetraazabenzo
[cd]azulene-7-one (9)
3.10. Hyperchromicity, extinction coefficients and nucleoside
composition of the ODNs
To a solution of 5 (93 mg, 0.29 mmol) in DMF (3 mL) containing
imidazole (120 mg, 1.8 mmol) was added tert-butyldimethylsilyl
chloride (160 mg, 0.88 mmol), and the reaction mixture was stirred
for 3 h at 50 °C. The reaction was quenched by addition of MeOH
and the solvent was removed in vacuo. The residue was dissolved
in a mixture of CHCl3 and MeOH (9:1), and the organic solution
was washed with 1 N HCl, saturated aqueous NaHCO3, and satu-
rated brine. The organic layer was dried (Na2SO4) and concentrated
in vacuo. The residue was purified by a silica gel column, eluted
with hexane/AcOEt (3:1–2:1), to give 9 (110 mg, 69%) as a yellow
solid. An analytical sample was crystallized from MeOH to give
yellow crystals: mp 147–148 °C; EI-LRMS m/z 545 (M+); 1H NMR
(CDCl3) d 9.81 (br s, 1H, exchangeable with D2O), 7.11 (s, 1H),
6.81 (d, 1H, J = 11.5 Hz), 6.43 (m, 1H), 5.88 (br s, 2H, exchangeable
with D2O), 5.75 (d, 1H, J = 11.5 Hz), 4.54 (m, 1H), 3.93 (m 1H), 3.82
(dd, 1H, J = 11.2 and 3.6 Hz), 3.75 (dd, 1H, J = 11.2 and 3.0 Hz), 2.33
(m, 2H), 0.94 and 0.92 (each s, each 9H), 0.10 (s, 12H); 13C NMR
(CDCl3) d 169.0, 162.4, 155.0, 153.8, 134.9, 120.3, 119.9, 114.3,
99.6, 87.7, 83.5, 72.2, 63.2, 41.6, 26.3, 26.1, 18.8, 18.4, ꢁ4.3, ꢁ4.5,
ꢁ5.0, ꢁ5.1; Anal. Calcd for C26H43N5O4Si2: C, 57.21; H, 7.94; N,
12.83. Found: C, 57.15; H, 7.82; N, 12.53.
Each ODN (0.5 OD units at 260 nm) was incubated with
snake venom phosphodiesterase (6
lL, 1 mg/0.5 mL), nuclease P1
(12 L, 1 mg/mL), and alkaline phosphatase (10
l
l
L, 0.1 U/mL) in
a buffer containing 100 mM Tris–HCl (pH 7.7) and 2 mM MgCl2
(total 528 mL) at 37 °C for 8 h. Hyperchromicity of each ODN was
determined by comparing UV absorbencies at 260 nm of the solu-
tions before and after hydrolysis. The extinction coefficients (at
260 nm) of each ODN were determined using the following equa-
tion: eODN = the sum of enucleoside/hyperchromicity. The extinction
coefficients (at 260 nm) of nucleosides used for calculations were
as follows: dA, 15,400; dG, 11,700; dT, 8800; dC, 7300; BaON,
13,900. After the reaction mixture was heated in boiling water
for 5 min, the enzymes were removed from the reaction mixture
by filtration with MicropureÒ-EZ device (Millipore), and the filtrate
was concentrated. Nucleoside composition was determined by
analysis of the residue with reverse-phase HPLC, using a J’sphere
ODN 80 column (4.6 ꢂ 250 mm, YMC) with a linear gradient of
acetonitrile (from 5% to 80% over 30 min) in 0.1 M TEAA buffer
(pH 7.0). The results were shown in the Supplementary data
(Fig. S2 and Table S1).

358
Y. Hirama et al. / Bioorg. Med. Chem. 19 (2011) 352–358
2. (a) Krueger, A. T.; Kool, E. T. Curr. Opin. Chem. Biol. 2007, 11, 588; (b) Loakes, D.;
3.11. Thermal denaturation
Gallego, J.; Pinheiro, V. B.; Kool, E. T.; Holliger, P. Journal. Am. Chem. Soc. 2009,
131, 14827.
Each sample containing an appropriate duplex (6 lM for the
3. (a) Hirao, I. Curr. Opin. Chem. Biol. 2006, 10, 622; (b) Kimoto, M.; Mitsui, T.;
Yokoyama, S.; Hirao, I. Journal. Am. Chem. Soc. 2010, 132, 4988.
4. (a) Henry, A. A.; Romesberg, F. E. Curr. Opin. Chem. Biol. 2003, 7, 727; (b)
Malyshev, D. A.; Seo, Y. J.; Ordoukhanian, P.; Romesberg, F. E. Journal. Am. Chem.
Soc. 2009, 131, 14620.
experiments in Table 1 and 5 lM for the experiments in Table 2)
in a buffer solution (10 mM sodium cacodylate (pH 7.0) containing
1 mM NaCl for the experiments in Table 1 and 10 mM sodium
cacodylate (pH 7.0) containing 1 M NaCl for the experiments in
Table 2) was heated at 95 °C for 5 min, cooled gradually to an
appropriate temperature, and used for the thermal denaturation
study. Thermal-induced transitions of each mixture of ODNs were
monitored at 260 nm on a Beckman DU 650 spectrophotometer.
The sample temperature was increased 0.5 °C/min.
5. Clever, G. H.; Kaul, C.; Carell, T. Angew. Chem., Int. Ed. 2007, 46, 6226.
6. Minakawa, N.; Kojima, N.; Hikishima, S.; Sasaki, T.; Kiyosue, A.; Atsumi, N.;
Ueno, Y.; Matsuda, A. Journal. Am. Chem. Soc. 2003, 125, 9970.
7. For the sake of simplicity, the aglycons and nucleoside units of
imidazopyridopyrimidine derivatives were referred to as ImON and ImNO,
and those of naphthyridine possessing amino-oxo substituents and
tetraazabensoazulene possessing oxo-amino substituents were referred to as
NaNO and BaON, respectively.
8. Hikishima, S.; Minakawa, N.; Kuramoto, K.; Fujisawa, Y.; Ogawa, M.; Matsuda,
A. Angew. Chem., Int. Ed. 2005, 44, 596.
9. Hikishima, S.; Minakawa, N.; Kuramoto, K.; Ogata, S.; Matsuda, A. ChemBioChem
2006, 7, 1970.
Acknowledgments
This investigation was supported in part by a Grant-in-Aid for
Scientific Research on Priority Areas from the Ministry of Educa-
tion, Science, Sports, and Culture of Japan. We would like to thank
Ms. M. Kiuchi (Center for Instrumental Analysis, Hokkaido Univer-
sity) for elemental analysis. We also would like to thank Ms. S. Oka
(Center for Instrumental Analysis, Hokkaido University) for mea-
surement of Mass spectra.
10. Minakawa, N.; Ogata, S.; Takahashi, M.; Matsuda, A. Journal. Am. Chem. Soc.
2009, 131, 1644.
11. Ogata, S.; Takahashi, M.; Minakawa, N.; Matsuda, A. Nucleic Acids Res. 2009, 37,
5602.
12. Hirama, Y.; Abe, H.; Minakawa, N.; Matsuda, A. Tetrahedron 2010, 66, 8402.
13. Seela, F.; Zulauf, M. Journal. Chem. Soc., Perkin Trans. 1 1998, 3233.
14. Cook, P. D.; Ewing, G.; Jin, Y.; Lambert, J.; Prhavc, M.; Rajappan, V.; Rajwanshi,
V. K.; Sakthivel, K. Int. Pat. WO 2005 021568 A2, 2005.
15. Compound 5 was stable (>95% remaining) under both conditions for 12 h at
room temperature.
16. Sinha, N. D.; Michaud, D. P.; Roy, S. K.; Casale, R. A. Nucleic Acids Res. 1994, 22,
3119.
Supplementary data
17. One can image a stable base pair between BaON and T via three H-bonds.
However, the Tm value of ODN I:ODN II containing a BaON:T pair was 47.0 °C,
which was same as A:T pair. Therefore, the BaON:NaNO would be thermally
stable and specific.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.11.023.
18. Guckian, K. M.; Schweitzer, B. A.; Ren, R. X.-F.; Sheils, C. J.; Paris, P. L.;
Tahmassebi, D. C.; Kool, E. T. Journal. Am. Chem. Soc. 1996, 118, 8182.
19. In this case, ImON means 30,50-O-disilylated derivative.
References and notes
1. (a) Benner, S. A.; Sismour, A. M. Nat. Rev. Genet. 2005, 6, 533; (b) Hoshika, S.;
Chen, F.; Leal, N. A.; Benner, S. A. Angew. Chem., Int. Ed. 2010, 49, 5554.