Chemical synthesis of diastereomeric diadenosine boranophosphates (ApbA) from 2′-O-(2-cyanoethoxymethyl)adenosine by the boranophosphotriester method

Article abstract of DOI:10.1016/j.bmc.2008.09.040

We have synthesized diastereomerically pure diadenosine 3′,5′-boranophosphates (Ap^bA) by using the boranophosphotriester method from ribonucleosides protected with the 2′-hydroxy protecting group 2-cyanoethoxymethyl (CEM). Melting curves of the triple-helical complex of the dimer Ap^bA and 2poly(U) at high ionic strength revealed that presumptive (Sp)-Ap^bA had a much higher affinity and presumptive (Rp)-Ap^bA a much lower affinity for poly(U) than the natural dimer ApA did. In contrast, the affinities of these dimers for poly(dT) were similar. Both the (Rp)- and the (Sp)-boranophosphate diastereomers showed much higher resistance to digestion by snake venom phosphodiesterase and nuclease P1 than ApA did. They have potential for use as synthons to be incorporated into boranophosphate oligonucleotides. In particular, because oligonucleotides containing Sp boranophosphate nucleotides are expected to bind more strongly and specifically to RNA than natural oligoribonucleotides do, they may find application in the isolation and detection of functional RNA in basic research and diagnostics.

Full text of DOI:10.1016/j.bmc.2008.09.040

Bioorganic & Medicinal Chemistry 16 (2008) 9154–9160
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Chemical synthesis of diastereomeric diadenosine boranophosphates
(ApbA) from 20-O-(2-cyanoethoxymethyl)adenosine by the
boranophosphotriester method
b
a
a
Yukiko Enya a, Seigo Nagata a, , Yutaka Masutomi , Hidetoshi Kitagawa , Kazuchika Takagaki ,
*
Natsuhisa Oka c, Takeshi Wada c, Tadaaki Ohgi a, Junichi Yano b
a Discovery Research Laboratories, Nippon Shinyaku Co., Ltd, 3-14-1 Sakura, Tsukuba City, Ibaraki 305-0003, Japan
b Discovery Research Laboratories, Nippon Shinyaku Co., Ltd, 14 Nishinosho-Monguchi-Cho, Kisshoin, Minami-ku, Kyoto 601-8550, Japan
c Department of Medical Genome Sciences, Graduate School of Frontier Sciences, The University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa, Chiba 277-8562, Japan
a r t i c l e i n f o
a b s t r a c t
We have synthesized diastereomerically pure diadenosine 30,50-boranophosphates (ApbA) by using the
boranophosphotriester method from ribonucleosides protected with the 20-hydroxy protecting group
2-cyanoethoxymethyl (CEM). Melting curves of the triple-helical complex of the dimer ApbA and
2poly(U) at high ionic strength revealed that presumptive (Sp)-ApbA had a much higher affinity and pre-
sumptive (Rp)-ApbA a much lower affinity for poly(U) than the natural dimer ApA did. In contrast, the
affinities of these dimers for poly(dT) were similar. Both the (Rp)- and the (Sp)-boranophosphate diaste-
reomers showed much higher resistance to digestion by snake venom phosphodiesterase and nuclease P1
than ApA did. They have potential for use as synthons to be incorporated into boranophosphate oligonu-
cleotides. In particular, because oligonucleotides containing Sp boranophosphate nucleotides are
expected to bind more strongly and specifically to RNA than natural oligoribonucleotides do, they may
find application in the isolation and detection of functional RNA in basic research and diagnostics.
Ó 2008 Elsevier Ltd. All rights reserved.
Article history:
Received 1 August 2008
Revised 8 September 2008
Accepted 9 September 2008
Available online 18 September 2008
Keywords:
Boranophosphate
Diadenosine
Poly(U)
2-Cyanoethoxymethyl
Nuclease resistance
Tm value
1. Introduction
ports of the chemical synthesis of oligoribonucleoside boranophos-
phates (BH3–ORNs).14,16,17 In addition, because of undesirable
Manytypesofmodifiedoligonucleotideshavebeenreportedover
the past couple of decades.1–5 Among them, oligonucleoside borano-
phosphates, inwhich onenonbridgingoxygenin the phosphate moi-
ety is replaced by a borane group (BH3–), have attracted much
attention because of their high chemical stability, high nuclease
resistance, and potential for use in boron neutron-capture therapy.6
Furthermore, Hall et al.7 have reported that boranophosphate mod-
ification is useful for increasing the activity of siRNA. Since oligode-
oxyribonucleoside boranophosphates (BH3-ODNs) were first
synthesized in 1990 by Sood et al.,5 both enzymatic and chemical
methodsfor synthesizing BH3-ODNs havebeen reported.8–15 The en-
zymes which have been used for enzymatic synthesis of BH3-ODNs,
such as T7 RNA polymerase, recognize only the Rp diastereomer of
reductions that occur at the N-acyl protecting groups of the base
moiety during the boronation step,18,19 most of the chemical meth-
ods reported for the synthesis of BH3-ODNs and BH3-ORNs are lim-
ited to thymidine or uridine derivatives. Although we have
reported a boranophosphotriester method that can be used to syn-
thesize BH3-ODNs containing all four nucleobases, A, C, G, and T,12
there are still no reports of the chemical synthesis of BH3-ORNs with
nucleobases besidesU. In addition, thecouplingefficiencyof thebor-
anophosphotriester method is lower than that of the phosphorami-
dite method, and this disadvantage would be even more marked
when the boranophosphotriester method is applied to the synthesis
of RNA analogues from ribonucleosides protected with the bulky
conventional t-butyldimethylsilyl (20-O-TBDMS) protecting group.
We have developed an achiral 20-hydroxy protecting group, 2-
cyanoethoxymethyl (CEM), for RNA synthesis.20 Using this protect-
ing group, we have succeeded in efficiently synthesizing RNA olig-
omers up to 110 nt.21 With the CEM method, RNA synthesis can be
carried out much more simply and easily than by the conventional
TBDMS method. Largely because the CEM protecting group shows
less steric hindrance than the TBDMS protecting group, RNA syn-
thesis by the CEM method is comparable in efficiency with DNA
nucleoside 50-(
a-P-borano)triphosphates to incorporate Sp-borano-
phosphate internucleotide linkages, so that Rp linkages cannot be
incorporated into an oligonucleotide by the enzymatic methods.
For the comparative study of the Sp and Rp diastereomers, therefore,
chemicalsyntheticmethodsarerequired. However, there are fewre-
* Corresponding author. Tel.: +81 29 850 6242; fax: +81 29 850 6217.
E-mail address: s.nagata@po.nippon-shinyaku.co.jp (S. Nagata).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.09.040

Y. Enya et al. / Bioorg. Med. Chem. 16 (2008) 9154–9160
9155
synthesis. Even though the boranophosphotriester method has a
poor coupling yield when used with conventional 20-hydroxy pro-
tecting groups, it can be expected to give a high yield of coupled
product when used with the CEM protecting group.
phosphotriester method. Compound 1, the boranophosphorylating
reagent triethylammonium bis(2-cyanoethyl) boranophosphate 2,
and the condensing reagent 1,3-dimethyl-2-(3-nitro-1,2,4-triazol-
1-yl)-2-pyrrolidin-1-yl-1,3,2-diazaphospholidinium
phosphate (MNTP), were synthesized as previously de-
scribed.21,25,26 Compound
was allowed to react with the
hexafluoro-
In the present paper, we describe the synthesis of diadenosine
30,50-boranophosphate (ApbA) from CEM-protected ribonucleosides
by the boranophosphotriester method. The reason we chose aden-
osine is that the interaction of ApA analogues with poly(U) has of-
ten served as a model system for oligonucleotides with the same
modifications because of the strong base-stacking between the
adenine bases in ApA analogues.22–24 We separated the two ApbA
diastereomers by reverse-phase column chromatography and car-
ried out a comparative study of their physicochemical properties.
We found that the configuration of BH3–ORNs at the phosphorus
center affected their nuclease resistance, the strength of their
base-stacking, and the thermal stability of their complexes with
poly(U). As far as we know, this is the first report of the chemical
synthesis of a diribonucleoside boranophosphate by using conven-
tional nucleobase protection with an acyl group which is suscepti-
ble to reduction by boronation reagents.
1
boranophosphorylating reagent 2 in the presence of MNTP and
2,6-lutidine. The resulting fully protected intermediate 3 was then
treated with Et3N in CH2Cl2 to remove one of the two 2-cyanoethyl
groups of 3 to give the corresponding boranophosphate monomer
4. The monomer 4 was coupled with 20-O-CEM-30-O-TBDMS-N6-
acetyladenosine 5, which was prepared from compound 1, result-
ing in fully protected diastereomeric dimers ApbA 6a and 6b in
57% yield (Scheme 1). Because the boranophosphate group in 6a
and 6b, which bears a 2-cyanoethyl protecting group, is not partic-
ularly stable during workup and purification, the isolated yield of
this reaction is only moderate.
We next set out to remove the protecting groups of the dimers
6a and 6b. The dimethoxytrityl cation (DMTr+) is known to react
with borane groups, resulting in the cleavage of the internucleotide
linkages.9,19 Therefore we used Et3SiH as a DMTr+ scavenger for the
deprotection of the DMTr group, as previously reported.12 Finally,
the other protecting groups of the dimers 6a and 6b were removed
by the usual procedure for 20-O-CEM-protected RNA oligomers to
yield the diastereomeric diadenosine boranophosphates 7a and
7b (Scheme 2). When the diastereomers 7a and 7b were subjected
to reverse-phase column chromatography, 7a eluted earlier than
2. Results and discussion
2.1. Synthesis of diadenosine 30,50-boranophosphate (ApbA)
We synthesized the boranophosphorylated monomer 4 from 50-
O-dimethoxytrityl-20-O-CEM-N6-acetyladenosine 1 by the borano-
Scheme 1. Synthesis of diadenosine 30,50-boranophosphates (6a, 6b). Reagents and conditions: (i) 2 (1.2 equiv), MNTP (2.4 equiv), and 2,6-lutidine (10 equiv) in CH3CN, rt,
1 h; (ii) Et3N (10 equiv) in CH2Cl2, rt, 1 h; (iii) MNTP (3.0 equiv), and 2,6-lutidine (10 equiv) in CH3CN, rt, 1 h.
Scheme 2. Deprotection of diadenosine 30,50-boranophosphates (7a, 7b). Reagents and conditions: (i) 3% dichloroacetic acid in Et3SiH/CH2Cl2 (1:1, v/v), rt, 1 min; (ii) Et3N/
CH3CN (1:1, v/v), rt, 1 h; (iii) concd NH3 aq./EtOH (3:1, v/v), rt, 1 h; (iv) 1 M TBAF in THF containing 1% nitromethane, rt, overnight.

9156
Y. Enya et al. / Bioorg. Med. Chem. 16 (2008) 9154–9160
Figure 1. HPLC analysis of diadenosine 30,50-boranophosphate. (a) 7a, (b) 7b;
m; Imtakt);
buffer A, 5% CH3CN, 50 mM TEAA, pH 7.0; buffer B, 90% CH3CN, 50 mM TEAA, pH
7.0; gradient of 0–20% buffer B in 20 min; flow rate, 0.5 mL/min; 40 °C. UV detection
was at 260 nm.
Scheme 3. Absolute structures of boranophosphate and phosphorothioate diaste-
reomers and their designations according to Prelog’s rule. The substituents at the
phosphorus atom are numbered from 1 (highest priority) to 4 (lowest priority) in
decreasing order of the atomic number of the atoms attached to the stereocenter.
Note that boranophosphate and phosphorothioate diastereomers with the same
absolute configuration have opposite Prelog designations.
Cadenza CD-C18 HPLC reverse-phase column (3 mm ꢀ 75 mm,
3 l
7b (Fig. 1). The structures of 7a and 7b were confirmed by their 1H,
13C, and 31P NMR spectra, reverse-phase HPLC analysis, and elec-
trospray ionization (ESI) mass spectrometry. With a view to deter-
mining the configuration at the chiral phosphorus center in 7a and
7b, we carried out NMR measurements and nuclease-resistance
tests as described below.
verse-phase HPLC, is more resistant than the Sp isomer.16,27,29 In
addition, Shaw and coworkers10 have found that an oligothymidine
boranophosphate dodecamer synthesized as a diastereomeric mix-
ture is much more resistant to SVP than are their natural and
phosphorothioate counterparts.
To investigate the effect of modifications at the phosphorus
center on nuclease resistance, we compared the resistance of five
dimers, ApA (9), ApbA (faster-eluting; 7a), ApbA (slower-eluting;
7b), ApsA (Rp; 8a), and ApsA (Sp; 8b) to SVP and nuclease P1 (Figs.
2 and 3).30 The boranophosphate dimers 7a and 7b and the phosp-
horothioate dimers 8a and 8b were significantly more resistant to
both enzymes than was the natural dimer ApA. In particular, 7b
and 8b were more resistant to SVP than were the other dimers.
From this diastereospecific resistance to SVP, we infer that 7b cor-
responds to the Rp configuration, a result that is consistent with
the NMR measurements and that provides independent support
2.2. NMR measurements
The borane groups of the two diastereomers of boron-modified
dimers are in different positions relative to the base-stacking re-
gion.27 In the Sp isomer, the borane group is directed toward the
base-stacking region and is close to the H30 proton of the 50-residue
(pseudoaxial position), while in the Rp isomer, the borane group is
directed away from the base-stacking region and is therefore far
from the H30 proton (pseudoequatorial position). Therefore we car-
ried out a 1D-NOE difference experiment to estimate the distance
between the BH3 protons and the H30 proton of the 50-residue of
each dimer. Irradiation of the H30 proton of the 50-residue of 7a re-
sulted in a small increase in the intensity of the BH3 proton reso-
nances (data not shown). Similar experiments with 7b resulted
in no signal change. These results provide strong evidence that
7a is the Sp isomer and 7b is the Rp isomer. However, the observed
NOE in the 7a 1D-NOE difference experiments was not large en-
ough to allow unambiguous assignment of the absolute configura-
tion of the diastereomers.
2.3. Nuclease resistance of ApA, ApbA, and ApsA
It is important to investigate the nuclease resistance of the di-
mers not only from a biological but also from a chemical point of
view, because the results can be used to determine the configura-
tion at the phosphorus center. An enzymatic method has been re-
ported for the determination of the configuration at the
phosphorus center of phosphorothioates based on the observation
of different rates of hydrolysis of the Rp and Sp diastereomers cat-
alyzed by snake venom phosphodiesterase (SVP).28 According to
Prelog’s rule, boranophosphates and phosphorothioates with the
same absolute configurations have opposite Sp/Rp designations
(see Scheme 3). The boranophosphate diastereomer that is resis-
tant to SVP is proposed to be the Rp configuration.16,18,27 Borano-
phosphate dimers have higher resistance to SVP than natural
dimers do, and the Rp isomer, which elutes more slowly on re-
Figure 2. Digestion of dimers with snake venom phosphodiesterase. Residual
dimer was measured by reverse-phase HPLC after treatment of ApA, ApbA (7a),
ApbA (7b), ApsA (Rp; 8a), or ApsA (Sp; 8b) with SVP.

Y. Enya et al. / Bioorg. Med. Chem. 16 (2008) 9154–9160
9157
tween ApbA and poly(U). The mixing curves for ApbA 7a or 7b and
poly(U) at 260 nm in the presence of 1.0 M NaCl (not shown) re-
vealed a break at ApbA:poly(U) ꢁ 1:2, indicating the formation of
an ApbA/2poly(U) triple-helical complex. We then measured the
Tm values of the dimer/2poly(U) triple-helical complexes to com-
pare the hybridization properties of ApA, ApbA (7a), and ApbA
(7b) in the presence of 1.0 M NaCl (Table 1) (the Tm values were
too low to be measured at low ionic strength). The poly(A)/
2poly(U) triple-helical complex is denatured directly to single
strands on heating at high ionic strength, as evidenced by the sim-
ple shape of the melting curve.33 The melting curve for each dimer/
2poly(U) triple-helical complex (Fig. 5a) also indicates a single-
step transition. The Tm value of the 7a/2poly(U) complex (16 °C)
was substantially higher than that of the natural dimer ApA/
2poly(U) complex (9.1 °C). In contrast, the Tm value of the 7b/
2poly(U) complex (4.0 °C) was substantially lower than that of
ApA/2poly(U) complex. Compound 7a is proposed to be the Sp bor-
anophosphate dimer on the basis of the NMR measurements and
hydrolysis by SVP. These results indicate that the Sp boranophos-
phate adenosine dimer hybridizes more strongly to poly(U) than
the natural dimer does. Meanwhile, the Tm values of the dimer/
2poly(dT) complexes were below 0 °C and similar to each other
(Fig. 5b and Table 1). Therefore, the increase in the strength of
hybridization between the Sp boranophosphate adenosine dimer
and polymer is specific for the RNA polymer. These findings point
the way to the development of a functional nucleic acid to recog-
nize RNA with higher specificity than natural oligomers do.
Figure 3. Digestion of dimers with nuclease P1. Residual dimer was measured by
reverse-phase HPLC after treatment of ApA, ApbA (7a), ApbA (7b), ApsA (Rp; 8a), or
ApsA (Sp; 8b), with nuclease P1.
for the assignment of configuration. Modification of the phospho-
rus center by boronation was as effective as thiolation in providing
resistance to hydrolysis by SVP. Compounds 7b and 8a were
equally resistant to nuclease P1. In the context of Prelog’s rule, it
is worth noting that the rank order of resistance to nuclease P1 ob-
served for the boranophosphate diastereomers, 7a (presumptive
Sp) < 7b (presumptive Rp), was the opposite of that observed for
the phosphorothioate diastereomers, 8b (Sp) < 8a (Rp). Further
investigation of these differences between the two types of modi-
fication is underway.
3. Conclusions
We have synthesized diadenosine 30,50-boranophosphate
(ApbA) from CEM-protected ribonucleosides by the boranophos-
photriester method and isolated the pure ApbA diastereomers. As
far as we know, this is the first report of the chemical synthesis
of a diribonucleoside boranophosphate by using conventional
nucleobase protection with an acyl group which is susceptible to
reduction by boronation reagents. It is inferred from the analysis
of 1D-NOE difference spectra and nuclease resistance to SVP of
the dimers that 7a is the Sp isomer. The Tm values of the dimer/
2poly(U) complexes revealed that 7a had a much higher and 7b
a much lower affinity for poly(U) than did the natural dimer,
ApA. In contrast, the Tm values of the dimer/2poly(dT) complexes
were all similar. These results indicate that the diastereopure bor-
anophosphate RNA, which is inferred to be the Sp isomer, hybrid-
izes better to RNA than natural RNA does. The diastereomeric
boranophosphate dimers 6a and 6b synthesized by this method
can be subjected to specific removal of the 30-protecting group,
TBDMS, and transformed by the phosphitylating reagent bis(N,N-
diisopropylamino)cyanoethylphosphite into the fully protected di-
mer block CEM-amidite suitable for use in the solid-phase synthe-
sis of oligoribonucleotides having 30,50-boranophosphate
internucleotide linkages. Therefore these dimers have the potential
to be synthons that can be incorporated into oligonucleotides.
Although it is known that dimers may differ from oligomers in
their hybridization properties and that longer oligomers interact-
ing with mixed-base RNA tend to form double-helical rather than
triple-helical complexes, studies of dimers are sometimes carried
out to provide an initial insight into the hybridization properties
of oligomers.22–24 Accordingly, we propose to extend the chemistry
described in this paper to oligomers containing 30,50-boranophos-
phate linkages and to assess their interaction with complementary
RNA. Meanwhile, our preliminary studies on the physicochemical
properties of the diastereopure oligoribonucleotide boranophos-
phate dimers suggest that BH3-ORNs synthesized with the Sp dia-
2.4. CD spectra of ApA and ApbA
We recorded circular dichroism (CD) spectra of natural and bor-
anophosphate dimers (7a and 7b) at 0 °C (Fig. 4). All dimers
showed typical CD spectra with a positive band at about 270 nm
and a negative band at 252 nm. Compound 7a showed greater
CD intensity than the natural dimer ApA or 7b did. These results
suggest that all dimers have a right-handed stacked conformation
and that the two purine bases show stronger stacking interactions
in 7a than in the other dimers.
2.5. Thermal stability of ApA/poly(U) and ApbA/poly(U)
complexes
The adenosine dimer ApA forms an ApA/2poly(U) triple-helical
complex in the presence of 1 M NaCl.31,32 We performed mixing
experiments to investigate the stoichiometry of the interaction be-
Figure 4. CD spectra of 7a (blue), 7b (red), and ApA (black) at 0 °C. The total base
concentration was 0.1 mM. The spectra were taken in 10 mM sodium phosphate,
pH 7.4, containing 1.0 M NaCl.
stereomers can be developed as
a promising new class of

9158
Y. Enya et al. / Bioorg. Med. Chem. 16 (2008) 9154–9160
Table 1
from Wako Pure Chemical Industries, Ltd (Osaka, Japan) and tetra-
butylammonium fluoride (TBAF) was from Nacalai Tesque (Kyoto,
Japan).
Tm values of dimer/2poly(U) complexes and dimer/2poly(dT) complexes
dimer
Tm (°C) with poly(U)
Tm (°C) with poly(dT)
ApA
9.1
16.0
4.0
ꢂ1.9
ꢂ2.8
ꢂ0.5
ApbA 7a
ApbA 7b
4.3. Chemical synthesis
4.3.1. Triethylammonium 50-O-dimethoxytrityl-20-O-CEM-N6-
acetyladenosine cyanoethyl boranophosphate (4)
functional RNA that will form stable complexes with complemen-
tary RNA.
50-O-Dimethoxytrityl-20-O-CEM-N6-acetyladenosine 1 (500 mg,
0.72 mmol) and triethylammonium bis(2-cyanoethyl) boranophos-
phate 2 (262 mg, 0.86 mmol) were dried by repeated coevapora-
tion with dry toluene followed by dry pyridine and finally
dissolved in dry acetonitrile (7.2 mL). To the solution were succes-
sively added 2,6-lutidine (0.84 mL, 7.2 mmol) and MNTP (769 mg,
1.73 mmol). After having been stirred at rt for 1 h, the mixture was
diluted with CH2Cl2 (10 mL). The solution was washed with satu-
rated NaHCO3 aq. (3 ꢀ 10 mL) and the aqueous layer was back-ex-
tracted with CH2Cl2 (10 mL). The organic layer and washings were
combined, dried over Na2SO4, filtered, and concentrated to dryness
under reduced pressure. The residue was chromatographed on sil-
ica gel (10 g) with a gradient of 10–50% acetone in acetonitrile with
0.5% Et3N as the eluent. Fractions containing 50-O-dimethoxytrityl-
O-CEM-N6-acetyladenosine cyanoethyl boranophosphate (3),
unseparated by-products, and reagents, were combined and con-
centrated to dryness under reduced pressure. To a solution of 3
in CH2Cl2 (1 mL) was added Et3N (1 mL, 7.2 mmol). After having
been stirred at rt for 1 h, the mixture was concentrated to dryness
under reduced pressure. The residue was chromatographed on sil-
ica gel (10 g) with a gradient of 0–50% acetone in acetonitrile with
0.5% Et3N as the eluent. Fractions containing triethylammonium 50-
O-dimethoxytrityl-20-O-CEM-N6-acetyladenosine cyanoethyl bor-
anophosphate (4) were combined and concentrated to dryness un-
der reduced pressure. Excess Et3N was removed by repeated
coevaporation with toluene to yield 4 (358 mg; 54% from 1) as a
colorless foam. 1H NMR (300 MHz, CDCl3): d 0–1 (bq, 3H); 1.27
(t, 9H, J = 7.3 Hz); 2.43–2.71 (m, 4H); 2.59 (s, 3H); 3.01 (q, 6H,
J = 7.3 Hz); 3.43–4.04 (m, 6H); 3.79 (s, 6H); 4.47–4.57 (m, 1H);
4.79 (dd, 1H, J = 7.2, 10.2 Hz); 4.98–5.07 (m, 3H); 6.26–6.29 (m,
1H); 6.78–6.85 (m, 4H); 7.20–7.45 (m, 9H); 8.20 (d, 1H,
J = 5.9 Hz); 8.61 (s, 1H); 8.68 (bs, 1H); 13C NMR (300 MHz, CDCl3):
d 8.6, 14.2, 18.7, 18.8, 20.0, 20.1, 25.6, 29.3, 29.7, 45.7, 55.3, 58.6,
62.9, 63.0, 63.2, 71.6, 86.9, 87.1, 94.9, 95.0, 113.2, 117.7, 117.8,
117.9, 127.1, 127.1, 127.9, 128.4, 130.3, 135.4, 135.4, 135.5,
141.3, 144.3, 149.2, 151.1, 151.2, 152.4, 152.5, 158.6, 158.6,
170.2; 31P NMR (500 MHz, CDCl3): d 95.0–99.5 (m). HRMS (ESIꢂ)
calcd for C40H45BN7O10P [MꢂH]ꢂ, 824.29748; found, 824.29600.
4. Experimental
4.1. General methods
UV spectra, mixing curves, and melting curves were recorded
with a Hitachi U-2810 double-beam spectrophotometer. CD spec-
tra were taken with a Jasco J-720 spectropolarimeter fitted with
a Jasco temperature controller. 1H, 13C, and 31P NMR spectra were
recorded on a Varian VNMRS600, Bruker DRX 500 or Bruker DPX
300 spectrometer with 3-trimethylsilylpropionate-2,2,3,3-d4 so-
dium salt or H3PO4 as the standard. High resolution mass spectra
(HRMS) were recorded on a JEOL JMS-T100LC mass spectrometer.
Analytical high performance liquid chromatography (HPLC) was
performed on a Shimadzu HPLC system equipped with a Cadenza
CD-C18 (75 ꢀ 3 mm, Imtakt) or Mightysil RP-18 GP (150 ꢀ
4.6 mm, Kanto Kagaku) column. Thin-layer chromatography (TLC)
was performed on Merck silica gel 60 F254 precoated plates and
column chromatography on Wako silica gel C-200 (Wako) or
LiChroprep RP-18 GP (Merck). DowexÒ 50WX8-200 hydrogen-
form ion-exchange resin (Sigma–Aldrich) was used after conver-
sion to the pyridinium or sodium form by washing with pyri-
dine/water (1:1, v/v) or 1.0 M sodium hydroxide solution,
respectively. We used published molar extinction coefficients for
ApA (e e
= 13,600 Mꢂ1 cmꢂ1 at 258 nm)31 and poly(U) ( = 9200 Mꢂ1
cmꢂ1 at 260 nm).34 The amount of the final products 7a and 7b was
determined by measuring the absorbance at 258 nm and convert-
ing to micromoles by using the published extinction coefficient for
ApA.
4.2. Materials
Sodium polyuridylate (poly(U)) and nuclease P1 were pur-
chased from Yamasa Corp. (Chiba, Japan) and snake venom phos-
phodiesterase from Worthington Biochemical Corp. (Lakewood,
NJ). Reagents and solvents were from commercial suppliers and
were used without further purification. Anhydrous solvents were
Figure 5. Melting curves of 7a (blue), 7b (red), or ApA (black) complexed with (a) poly(U) or (b) poly(dT). The spectra were taken in 10 mM sodium phosphate, pH 7.4,
containing 1.0 M NaCl. Each dimer and poly(U) or poly(dT) were mixed at a ratio of 1:2 to give a total base concentration of 0.1 mM.

Y. Enya et al. / Bioorg. Med. Chem. 16 (2008) 9154–9160
9159
4.3.2. 20-O-CEM-30-O-TBDMS-N6-acetyladenosine (5)
50-O-Dimethoxytrityl-20-O-CEM-N6-acetyladenosine
117.7, 117.7, 122.2, 122.4, 124.0, 127.1, 127.2, 128.0, 128.1,
130.1, 130.2, 135.2, 135.2, 142.0, 142.2, 144.2, 149.5, 150.9,
151.2, 151.2, 152.5, 152.5, 158.7, 158.7, 170.4, 170.5, 170.5; 31P
NMR (500 MHz, CDCl3): d 120.3–121.2 (m). HRMS (ESIꢂ) calcd for
C62H77BN13O15PSi [MꢂH]ꢂ, 1312.51783; found, 1312.51717.
1
(2.0 g,
2.87 mmol) was dried by repeated coevaporation with dry toluene
followed by dry pyridine and finally dissolved in dry dimethyl-
formamide (30 mL). To the solution were successively added imid-
azole (0.43 g, 6.31 mmol) and tert-butyldimethylsilyl chloride
(0.87 mg, 5.74 mmol). After having been stirred at rt for 20 h, the
mixture was diluted with AcOEt (50 mL). The solution was then
washed successively with water (2 ꢀ 50 mL) and brine (50 mL).
The organic layers were combined and dried over Na2SO4, filtered,
and concentrated to dryness under reduced pressure. The residue
was chromatographed on silica gel (40 g) with a mixture of AcOEt,
hexane, and acetone as the eluent. Fractions were combined and
concentrated to dryness under reduced pressure to yield 50-O-
dimethoxytrityl-20-O-CEM-30-O-TBDMS-N6-acetyladenosine (1.72 g,
74%) as a colorless foam. The resulting intermediate (1.72 g,
2.12 mmol) was dissolved in CH2Cl2 (60 mL). To the solution were
successively added dichloroacetic acid (1.75 mL, 21.2 mmol). After
having been stirred at rt for 15 min, the mixture was diluted with
CH2Cl2. The solution was then washed with saturated NaHCO3 aq.
(2 ꢀ 50 mL) and brine (50 mL). The organic layers were dried over
Na2SO4, filtered, and concentrated to dryness under reduced pres-
sure. The residue was chromatographed on silica gel with a mix-
ture of CH2Cl2 and MeOH as the eluent. Fractions were combined
and concentrated to dryness under reduced pressure to yield 20-
O-CEM-30-O-TBDMS-N6-acetyladenosine 5 (730 mg, 68%) as a col-
orless foam. 1H NMR (300 MHz, CDCl3): d 0.13 (s, 6H); 0.95 (s,
9H); 2.22–2.27 (m, 2H); 2.65 (s, 3H); 3.27–3.46 (m, 2H); 3.69–
3.81 (m, 1H); 3.97 (d, 1H, J = 12.3 Hz); 4.23 (s, 1H); 4.57 (d, 1H,
J = 4.6 Hz); 4.59 (s, 2H); 5.02 (dd, 1H, J = 4.6, 7.7 Hz); 5.92 (d, 1H,
J = 12.3 Hz); 6.02 (d, 1H, J = 7.7 Hz); 8.14 (s, 1H); 8.67 (s, 1H);
8.83 (s, 1H); 13C NMR (300 MHz, CDCl3): d ꢂ4.7, ꢂ4.6, 18.2, 18.6,
25.8, 25.9, 62.9, 63.1, 72.2, 78.8, 89.5, 95.0, 117.2, 123.2, 143.3,
150.0, 152.0, 170.6. HRMS (ESIꢂ) calcd for C22H34N6O6Si [MꢂH]ꢂ,
505.22254; found, 505.22072.
4.3.4. Triethylammonium diadenosine 30,50-boranophosphate
(ApbA) (7a and 7b)
The diastereomeric mixture of 6a and 6b (69.8 mg, 53.1 lmol)
was added to a solution of 3% dichloroacetic acid (DCA) in Et3-
SiH/CH2Cl2 (1:1 v/v, 5.3 mL), and the mixture was stirred at rt for
1 min. The mixture was then diluted with CH2Cl2 and washed with
saturated NaHCO3 aq. (2 ꢀ 10 mL), and the aqueous layer was
back-extracted with CH2Cl2 (1 ꢀ 10 mL). The organic layer and
washings were combined, dried over Na2SO4, filtered, and concen-
trated to dryness under reduced pressure. To a solution of the res-
idue in acetonitrile (1 mL) was added Et3N (1 mL). After having
been stirred at rt for 1 h, the mixture was concentrated to dryness
under reduced pressure. The residue was then dissolved in NH3
aq./EtOH (2 mL) and stirred at rt for 1 h, after which the reaction
mixture was concentrated to dryness under reduced pressure. A
solution of 1.0 M TBAF in tetrahydrofuran (THF) solution contain-
ing 1% nitromethane (2 mL) was added to the residue, and the solu-
tion was stirred at rt overnight. The reaction mixture was
subjected to reverse-phase column chromatography (LiChroprep
RP-18 GP (Merck)) with a linear gradient of 0–20% acetonitrile in
50 mM triethylammonium acetate (TEAA) buffer, pH 7.0. The flow
rate was 10 mL/min and the effluent was monitored at 260 nm.
Fractions containing triethylammonium diadenosine 30,50-borano-
phosphate were concentrated to dryness under reduced pressure.
This purification procedure afforded diastereopure 7a (394
A258 units, equivalent to 14.5
lmol) and 7b (347 A258 units, equiv-
alent to 12.8 mol) in 51% yield over 4 steps from 6a + 6b). NMR of
l
7a (Na salt): 1H NMR (D2O) d 0–1 (bq, 3H); 3.85–3.90 (m, 1H);
3.99–4.03 (m, 1H); 4.14–4.18 (m, 1H); 4.35–4.40 (m, 3H); 4.41–
4.44 (m, 1H); 4.47–4.50 (m, 1H); 4.64–4.70 (m, 2H); 5.89–5.90
(m, 1H); 5.91–5.92 (m, 1H); 7.90–7.92 (m, 1H); 8.10–8.12 (m,
1H); 8.19–8.21 (m, 1H); 8.23–8.24 (m, 1H); 13C NMR (D2O) d
63.4, 64.1, 71.9, 73.5, 74.3, 75.9, 76.6, 77.6, 85.4, 86.4, 88.8, 90.8,
91.2, 92.7, 141.5, 142.4, 143.4, 154.7, 155.5, 157.6; 31P NMR
4.3.3. 50-O-Dimethoxytrityl-20-O-CEM-N6-acetyladenosine-30-yl
30-O-TBDMS-N6-acetyladenosine-50-yl cyanoethyl
boranophosphate (6a and 6b)
Compounds 5 (85.8 mg, 92.5 lmol) and 4 (46.9 mg, 92.5 lmol)
were dried by repeated coevaporation with dry toluene followed
by dry pyridine and finally dissolved in dry acetonitrile
(1.00 mL). To the solution were successively added 2,6-lutidine
(0.11 mL, 0.925 mmol) and MNTP (123.5 mg, 0.278 mmol). After
having been stirred at rt for 1 h, the mixture was diluted with
CH2Cl2 (10 mL). The solution was washed with saturated NaHCO3
aq. (3 ꢀ 10 mL), and the aqueous layer was back-extracted with
CH2Cl2 (10 mL). The organic layer and washings were combined,
dried over Na2SO4, filtered, and concentrated to dryness under re-
duced pressure. The residue was chromatographed on silica gel
(5 g) with a gradient of 0–50% acetone in acetonitrile containing
0.5% pyridine as the eluent. Fractions containing 50-O-dimethoxy-
trityl-20-O-CEM-N6-acetyladenosine-30-yl 30-O-TBDMS-N6-acetyl-
adenosine-50-yl cyanoethyl boranophosphate were combined to
yield a diastereomeric mixture of 6a and 6b. 1H NMR (300 MHz,
CDCl3): d 0–1 (bq, 3H); 0.15 (s, 6H); 0.94 (s, 9H); 2.25–2.33 (m,
2H); 2.48–2.70 (m, 10H); 3.24–3.77 (m, 6H); 3.78 (s, 6H); 4.02–
4.52 (m, 6H); 4.60–4.87 (m, 6H); 5.16–5.35 (m, 2H); 6.17–6.23
(m, 2H); 6.79–6.83 (m, 4H); 7.22–7.31 (m, 7H); 7.38–7.41 (m,
2H); 8.16–8.25 (m, 2H); 8.56–8.57 (m, 1H); 8.68 (s, 1H); 8.81(bs,
1H); 8.87 (bs, 1H); 13C NMR (300 MHz, CDCl3): d ꢂ4.8, ꢂ4.8,
ꢂ4.6, 8.7, 14.1, 18.1, 18.1, 18.5, 18.5, 18.7, 18.8, 18.8, 19.6, 19.7,
22.7, 24.3, 25.6, 25.7, 25.7, 25.8, 26.5, 26.6, 29.3, 29.6, 29.7, 31.5,
31.5, 31.8, 46.3, 46.6, 46.6, 47.3, 47.5, 53.5, 55.3, 55.3, 61.5, 62.3,
62.7, 63.1, 63.2, 63.3, 70.5, 70.7, 76.3, 76.8, 78.7, 78.8, 82.6, 83.3,
86.0, 87.2, 87.7, 87.8, 89.4, 95.4, 95.4, 113.3, 116.3, 117.5, 117.5,
(D2O)
d
97.88 (d, JP,B = 413 Hz). HRMS (ESIꢂ) calcd for
C20H28BN10O9P [MꢂH]ꢂ, 593.17877; found, 593.17938. NMR of
7b (Na salt): 1H NMR (D2O) d 0–1 (bq, 3H); 3.81–3.89 (m, 1H);
4.16–4.32 (m, 1H); 4.34–4.38 (m, 2H); 4.47–4.50 (m, 1H); 4.55–
4.59 (m, 1H); 4.69–4.72 (m, 1H); 4.80–4.86 (m, 1H); 5.85–5.88
(m, 1H); 5.96–5.99 (m, 1H); 7.97–8.01 (m, 1H); 8.10–8.13 (m,
1H); 8.20–8.23 (m, 1H); 8.32 (s, 1H); 13C NMR (D2O) d 63.5, 64.4,
65.6, 72.4, 73.5, 76.1, 76.5, 77.2, 85.8, 86.7, 88.7, 90.6, 91.2, 92.1,
141.9, 142.8, 151.1, 155.4, 155.5, 158.0; 31P NMR (D2O) d 97.38
(d, JP,B = 411 Hz). HRMS (ESIꢂ) calcd for C20H28BN10O9P [MꢂH]ꢂ,
593.17877; found, 593.17813. The two diastereoisomers were des-
ignated 7a and 7b according to their order of elution on reverse-
phase HPLC, and they showed no evidence of diastereomeric
cross-contamination on analytical reverse-phase HPLC.
4.4. Preparation of diadenosine (ApA)
ApA was synthesized on a 15-lmol scale by the phosphorami-
dite method modified for CEM chemistry on an ÄKTA Oligopilot
10 (GE Healthcare Bio-Science Corp., NJ). Commercially available
controlled-pore glass (CPG) derivatized with N-benzoyl-20-O-acet-
ylriboadenosine was used as the solid support. 20-O-CEM-phospho-
ramidite adenosine was synthesized as previously described.15
Cleavage from the resin and deprotection of the phosphate and
base moiety were carried out by treatment with concentrated
ammonia in EtOH at 35 °C for 15 h. The other protecting groups

9160
Y. Enya et al. / Bioorg. Med. Chem. 16 (2008) 9154–9160
were removed by 1 M TBAF in THF and the ApA was purified as de-
scribed above.
and 1 mM spermidine; final concentrations) were added to each
dimer (0.1 A260 units) to give a total volume of 70 L and the reac-
l
tion mixture was incubated at 37 °C. Samples of the reaction mix-
ture were taken at 0.5, 3, 24, and 48 h and analyzed directly by
reverse-phase HPLC.
4.5. Mixing curves
Poly(U) and dimer were mixed in various proportions to main-
tain a constant total base concentration of 0.1 mM. Each nucleotide
mixture contained, in a total volume of 4 mL, nucleotide dimer
and/or polymer, 1.0 M NaCl, and 10 mM sodium phosphate, pH
7.4. The mixture was scanned at 0 °C from 320 to 205 nm, and from
each spectrum A260 was determined and plotted against the pro-
portion of poly(U).
4.10.2. Digestion by nuclease P1
Each dimer (0.1 A260 units) was incubated with nuclease P1
(2 U) at 37 °C. Samples of the reaction mixture were taken at 0.5,
3, 24, and 48 h and analyzed directly by reverse-phase HPLC.
Acknowledgments
4.6. Melting curves
This research was supported in part by grants from the New En-
ergy and Industrial Technology Development Organization (NEDO)
of Japan for its Functional RNA Project. We thank Dr. G.E. Smyth,
Discovery Research Laboratories, Nippon Shinyaku Co. Ltd, for
helpful discussions and suggestions during the preparation of the
manuscript. Funding to pay the Open Access publication charges
for this article was provided by Nippon Shinyaku Co. Ltd.
Melting curves (absorbance–temperature profiles) were re-
corded in the presence of 10 mM sodium phosphate, pH 7.4, con-
taining 1.0 M NaCl. Each dimer and poly(U) were mixed at a ratio
of 1:2 to give a total base concentration of 0.1 mM. The absorbance
was monitored at 260 nm as the temperature was increased from
0 °C to 30 °C at a rate of 0.5 °C/min. Tm values were determined as
the temperature corresponding to the peak of the first-derivative
plot of the melting profile.
References and notes
1. Kurreck, J. Eur. J. Biochem. 2003, 270, 1628–1644.
4.7. CD spectroscopy
2. Morita, K.; Hasegawa, C.; Kaneko, M.; Tsutsumi, S.; Sone, J.; Ishikawa, T.;
Imanishi, T.; Koizumi, M. Bioorg. Med. Chem. Lett. 2002, 12, 73–76.
3. Kawasaki, A. M.; Casper, M. D.; Freier, S. M.; Lesnik, E. A.; Zounes, M. C.;
Cummins, L. L.; Gonzalez, C.; Cook, P. D. J. Med. Chem. 1993, 36, 831–841.
4. Teplova, M.; Wallace, S. T.; Tereshko, V.; Minasov, G.; Symons, A. M.; Cook, P.
D.; Manoharan, M.; Egli, M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 14240–14245.
5. Sood, A.; Shaw, B. R.; Spielvogel, B. F. J. Am. Chem. Soc. 1999, 112, 9000–9001.
6. Li, P.; Sergueeva, Z. A.; Dobrikov, M.; Shaw, B. R. Chem. Rev. 2007, 107, 4746–
4796.
The spectropolarimeter was calibrated with (+)-10-camphor-
sulfonic acid. Spectra were recorded in a 1-cm-path-length quartz
cell at a scan speed of 50 nm/min, and the recorded spectra were
digitized at 0.1-nm resolution and stored on a floppy disk. The
bandwidth was 1 nm and the response time was 1.0 s. At least se-
ven spectra were accumulated for each determination.
7. Hall, A. H. S.; Wan, J.; Shaughnessy, E. E.; Shaw, B. R.; Alexander, K. A. Nucleic
Acids Res. 2004, 32, 5991–6000.
8. Zhang, J.; Terhorst, T.; Matteucci, M. D. Tetrahedron Lett. 1997, 38, 4957–4960.
9. Higson, A. P.; Sierzchala, A.; Brummel, H.; Zhao, Z.; Caruthers, M. H. Tetrahedron
Lett. 1998, 39, 3899–3902.
4.8. Preparation of NMR samples
10. Sergueev, D. S.; Shaw, B. R. J. Am. Chem. Soc. 1998, 120, 9417–9427.
11. Sergueev, D. S.; Sergueeva, Z. A.; Shaw, B. R. Nucleosides Nucleotides Nucleic
Acids. 2001, 20, 789–795.
12. Wada, T.; Shimizu, M.; Oka, N.; Saigo, K. Tetrahedron Lett. 2002, 43, 4137–4140.
13. Shimizu, M.; Wada, T.; Oka, N.; Saigo, K. J. Org. Chem. 2004, 69, 5261–5268.
14. Wan, J.; Shaw, B. R. Nucleosides Nucleotides Nucleic Acids 2005, 24, 943–946.
15. McCuen, H. B.; Noe, M. S.; Sierzchala, A. B.; Higson, A. P.; Caruthers, M. H. J. Am.
Chem. Soc. 2006, 128, 8138–8139.
16. Chen, Y.-Q.; Qu, F.-C.; Zhang, Y.-B. Tetrahedron Lett. 1995, 36, 745–748.
17. He, K.; Sergueev, D. S.; Sergueeva, Z. A.; Shaw, B. R. Tetrahedron Lett. 1999, 40,
4601–4604.
18. Sergueeva, Z. A.; Sergueev, D. S.; Shaw, B. R. Nucleosides Nucleotides Nucleic
Acids 2000, 19, 275–282.
19. Sergueeva, Z. A.; Sergueev, D. S.; Shaw, B. R. Nucleosides Nucleotides Nucleic
Acids 2001, 20, 941–945.
20. Ohgi, T.; Masutomi, Y.; Ishiyama, K.; Kitagawa, H.; Shiba, Y.; Yano, J. Org. Lett.
2005, 7, 3477–3480.
21. Shiba, Y.; Masuda, H.; Watanabe, N.; Ego, T.; Takagaki, K.; Ishiyama, K.; Ohgi, T.;
Yano, J. Nucleic Acids Res. 2007, 35, 3287–3296.
Compounds 7a and 7b synthesized as described above were
subjected to reverse-phase column chromatography (LiChroprep
RP-18 GP (Merck)) with a linear gradient of 0–20% acetonitrile in
50 mM triethylammonium hydrogen carbonate (TEAB) buffer, pH
7.4. The flow rate was 10 mL/min and the effluent was monitored
at 260 nm. Fractions containing 7a or 7b were concentrated to dry-
ness under reduced pressure, coevaporated twice with H2O to re-
move TEAB buffer, and passed through
a pyridinium-form
DowexÒ 50WX8–200 ion-exchange resin followed by a sodium-
form resin. The dimers were eluted with water at each ion-ex-
change step, and so were obtained in the sodium form. The dimers
were coevaporated twice with 99.96% D2O (Merck) and lyophilized
twice from 99.96% D2O for NMR analysis.
4.9. HPLC
ˇ
22. Pressová, M.; Endová, M.; Tocik, Z.; Liboska, R.; Rosenberg, I. Bioorg. Med. Chem.
Lett. 1998, 8, 1225–1230.
23. Hanuš, J.; Barvík, I.; Ruszová-Chmelová, K.; Štꢀpánek, J.; Turpin, P.-Y.; Bok, J.;
Rosenberg, I.; Petrová-Endová, M. Nucleic Acids Res. 2001, 29, 5182–5194.
24. Letsinger, R. L.; Bach, S. A.; Eadie, J. S. Nucleic Acids Res. 1986, 14, 3487–3499.
25. Shimizu, M.; Saigo, K.; Wada, T. J. Org. Chem. 2006, 71, 4262–4269.
26. Oka, N.; Shimizu, M.; Saigo, K.; Wada, T. Tetrahedron 2006, 62, 3667–3673.
27. Li, H.; Huang, F.; Shaw, B. R. Bioorg. Med. Chem. 1997, 5, 787–795.
28. Burgers, P. M. J.; Eckstein, F. Biochemistry 1979, 18, 592–596.
29. Sergueeva, Z. A.; Sergueev, D. S.; Ribeiro, A. A.; Summers, J. S.; Shaw, B. R. Helv.
Chim. Acta 2000, 83, 1377–1391.
The HPLC buffer stock solution was 1 M TEAA, pH 7.0 (1 M in
triethylamine, titrated to pH 7.0 with acetic acid), and the HPLC
buffers were prepared by mixing the appropriate amounts of 1 M
TEAA, pH 7.0, and acetonitrile with distilled deionized water. The
column was operated at a flow rate of 1 mL/min and the effluent
was monitored at 260 nm.
30. Sierzchala, A.; Okruszek, A.; Stec, W. J. J. Org. Chem. 1996, 61, 6713–6716.
31. Tazawa, I.; Tazawa, S.; Stempel, L. M.; Ts’o, P. O. P. Biochemistry 1970, 9, 3499–
3514.
4.10. Nuclease-digestion experiments
32. Ishiyama, K.; Smyth, G. E.; Ueda, T.; Masutomi, Y.; Ohgi, T.; Yano, J. J. Am. Chem.
Soc. 2004, 126, 7476–7485.
33. Blake, R. D.; Massouliè, J.; Fresco, J. R. J. Mol. Biol. 1967, 30, 291–308.
34. Ts’o, P. O. P.; Rapaport, S. A.; Bollum, F. J. Biochemistry 1966, 5, 4153–4170.
4.10.1. Digestion by snake venom phosphodiesterase
Snake venom phosphodiesterase (0.3 U) and buffer (7
50 mM Tris–HCl, pH 9.3, containing 1 mM MgCl2, 0.1 mM ZnCl2
lL;