Synthesis of 6-amino-2-vinylpurine derivatives for cross-linking and evaluation of the reactivity

Article abstract of DOI:10.1016/j.bmcl.2012.08.122

Oligodeoxynucleotides (ODNs) have been widely used for inhibiting the gene expression in antisense or antigene methods, and the interstrand cross-linking (ICL) forming ODNs have been expected to ensure the inhibition by these methods. Previously, we reported a highly efficient and selective ICL reaction toward cytosine using the 2-amino-6-vinylpurine derivative under acidic conditions. In this Letter, we report the synthesis of ODN containing 6-amino-2-vinylpurine derivatives and evaluation of the cross-linking reactivity.

Full text of DOI:10.1016/j.bmcl.2012.08.122

Bioorganic & Medicinal Chemistry Letters 22 (2012) 6957–6961
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis of 6-amino-2-vinylpurine derivatives for cross-linking and evaluation
of the reactivity
⇑
Shuhei Kusano, Tomoya Sakuraba, Shinya Hagihara, Fumi Nagatsugi
Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai-shi, Miyagi 980-8577, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 April 2012
Revised 25 July 2012
Accepted 30 August 2012
Available online 21 September 2012
Oligodeoxynucleotides (ODNs) have been widely used for inhibiting the gene expression in antisense or
antigene methods, and the interstrand cross-linking (ICL) forming ODNs have been expected to ensure
the inhibition by these methods. Previously, we reported a highly efficient and selective ICL reaction
toward cytosine using the 2-amino-6-vinylpurine derivative under acidic conditions. In this Letter, we
report the synthesis of ODN containing 6-amino-2-vinylpurine derivatives and evaluation of the
cross-linking reactivity.
Keywords:
Cross-linking
Antisense
Ó 2012 Elsevier Ltd. All rights reserved.
Purine derivatives
miRNA
Oligonucleotides
Synthetic oligonucleotides (ONs) are attractive tools in biology,
and they have been developed to control the gene expression both
in vitro and in vivo. These applications are based on the ability of
the ONs to form highly sequence-specific complexes with nucleic
acid targets of interest. The antisense ONs inhibit the translation
of mRNA by induction of the RNase H endonuclease activity to
cleave the RNA–DNA heteroduplex.1 Alternatively, ONs hybridize
with the target RNA to act through steric hindrance (RNase H inde-
pendent mechanism).2 A variety of chemically modified ONs has
been designed and synthesized to overcome the drawbacks of a
natural ON. Some ONs modified in either the carbohydrate part
or the phosphate backbone do not elicit RNase H activity. Recently,
such ONs have been applied to regulation of the gene expression
by a different strategy from a conventional method for inhibiting
the translation.3 20-O-Methyl RNA ONs and phosphorodiamidate
morpholino ONs (PMO) are employed to block splicing signals,
resulting in induced exon skipping.4 In addition, 20-modified
ONs5, PMO6 and a locked nucleic acid (LNA)7 have been used to
study the micro RNA (miRNA), which endogenously expresses
small regulatory non-coding RNAs that play key roles in diverse
cellular processes.8 Interstrand cross-linking reactions (ICL) are ex-
pected to improve the efficiency for regulation of the expression by
irreversibly binding to the target RNA through steric blocking.
There are numerous of reports on ICL forming ODNs using trigger-
ing signals such as photo irradiation9 or chemical reactions.10
Recently, photo-cross-linked reactions with cytosine in RNA have
demonstrated the ability to promote RNA editing and to possibly
manipulate the RNA function.11
We have demonstrated that oligonucleotides bearing 2-amino-
6-vinylpurine (2-AVP:1) achieve a highly selective cross-linking to
the cytosine base in DNA under acidic conditions.12 The high selec-
tivity of 1 might be due to the proximity effect by the formation of
a complex between 1 and the cytosine base with two hydrogen
bondings under acidic conditions. Based on these results, we have
designed 6-amino-2-vinylpurine (6-AVP) derivatives (2), expected
to form a complex with the thymine base under neutral conditions
(Fig. 1A). It is known that the 2-aminopurine-deoxythymine (dT)
pair decreases the duplex stability compared to the deoxyadenine
(dA)-dT pair13 (Fig. 1B), and we expect that 6-AVP (2) could form a
more stable duplex than 2-AVP (1).
During the synthesis of 2a, it turned out that the vinyl group
with 2a exhibited a very low reactivity with a nucleophile. Thus,
the carboxyl methyl substitution at the vinyl group was added to
increase the reactivity, because introducing an electron-withdraw-
ing substituent on the vinyl group increases its reactivity.14 In this
Letter, we report the synthesis and evaluation of the cross-linking
reactivity with the 6-AVP derivatives (2).
The synthesis of the oligodeoxynucleotide (ODN) incorporating
6-AVP derivatives (2a) is summarized in Scheme 1. The 6-amino-2-
idodo-purine derivative (3) was synthesized according to
a
previously reported procedure.15 The palladium-catalyzed cross-
coupling reaction of 3 with n-Bu3SnCHCH2 produced the vinyl
derivative which was, without purification, reacted with sodium
methane thiol to afford 4 (73% in 2 steps; Scheme 1). It was found
that the addition of the thiol to the vinyl group proceeded very
⇑
Corresponding author. Tel./fax: +81 22 217 5633.
E-mail address: nagatugi@tagen.tohoku.ac.jp (F. Nagatsugi).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.08.122

6958
S. Kusano et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6957–6961
(A)
(B)
H
DNA
DNA
DNA
H
H2N
N
N
N
N
O
N
O
N
O
N+H
N
O
N
H
N
N
DNA
N
N
DNA
DNA
N
H
O
H
R
1: 2-amino-6-vinylpurine (2-AVP)
2: 6-amino-2-vinylpurine (6-AVP)
2a: R=H
2b: R=COOMe
O
N
O
H
H
N
N
N
O
N
O
N
N
N
N
H
DNA
N
DNA
N
N
DNA
H
H
N
N
DNA
H
2-AP
dT
A
T
Stable complex
Unstable complex
Figure 1. Design of the novel cross-linking agents.
NH2
NH2
N
NH2
N
(a)
(b)
(a)
N
N
N
N
N
N
N
3
N
SMe
N
I
N
SOct
COOMe
TBSO
TBSO
TBSO
O
O
O
3
5
4
10
NHPac
OTBS
OTBS
OTBS
NH2
N
N
N
NHPac
NHPac
N
N
N
N
(c)
N
N
N
N
N
N
N
(c)
(b)
SMe
DMTrO
N
N
SMe
SOct
COOMe
O
COOMe
HO
DMTrO
CCGCGT
TCGCCG
O
O
6
O
P
11
12
O
P
iPr2N
OCH2CH2CN
NH2
OH
NH2
N
iPr2N
OCH2CH2CN
N
N
N
N
(d)
5'
(e)
Scheme 2. Reagents and conditions: (a) (1) n-Bu3Sn(COOMe)C@CH2 (9), Pd(PPh3)4,
CuI, 60 °C, DMF 71%, (2) 1-octanethiol, iPrOH, 49%; (b) (1) (PhOCH2CO)2O, Pyridine,
quant, (2) n-Bu4NF, THF, (3) DMTrCl, Py, 22% in two steps, (4) (i-Pr)2NP(Cl)OC2H4CN,
DIPEA, CH2Cl2 0 °C, 48%; (c) (1) synthesis using an automated DNA synthesizer, (2)
28% aqueous NH3, (3)10% CH3COOH.
N
N
N
SMe
5'
CCGCGT
3'
3'
TCGCCG
CCGCGT
TCGCCG
7
8
Scheme 1. Reagents and conditions: (a) (1) n-Bu3SnCH@CH2, Pd(CH3CN)2Cl2,
CH3CN, 60 °C, (2) CH3SNa, CH3CN, H2O 73% for two steps; (b) (1) PhOCH2COCl,
1-HOBT, CH3CN, Pyridine, 93%, (2) n-Bu4NF, THF, 82%; (c) (1) DMTrCl, Py, 92%, (2)
(i-Pr)2NP(Cl)OC2H4CN, DIPEA, CH2Cl2 0 °C, 29%; (d) (1) synthesis using an auto-
mated DNA synthesizer, (2) 28% aqueous NH3, (3)10% CH3COOH; (e) (1) 2 equiv
MMPP, (2) aqueous NaOH 3 days.
Attempts to protect the 6-amino group of the nucleoside deriv-
ative (10) using phenoxyacetyl chloride resulted in producing a
complex mixture. The protection of the amino group of 10 was per-
formed with phenoxyacetic anhydride to give the desired product
in good yields. After deprotection of the TBS group, the conven-
tional method produced the phosphoramidite precursor (11).
The methyl acrylate-bearing ODN (12) was synthesized on an
automated synthesizer using 11 and cleaved from the support with
28% ammonia, which had eliminated the sulfide group.
The purified ODN was (12) treated with 10% CH3COOH to give
the deprotected ODN. The structures of all the ODNs were deter-
mined by MALDI-TOF measurements.
The cross-linking reactions were conducted under acidic and
neutral conditions between the functionalized ODNs (8 and 12)
and target DNAs (13a) (N = dG, dA, dC, dT) or RNAs (13b) (N = G,
A, C, U) labeled with fluorescein at the 50 end and analyzed by
gel electrophoresis with 20% denaturing gel (Fig. 2A).
The formation of the slower mobility band indicates the cross-
linked adduct, and the yields of the adduct can be obtained by
quantification of both the lower and higher mobility bands. The
ODN (8) did not give the adduct at all, as anticipated from the
reactivity of the vinyl derivative (2a) with the thiol. In contrast,
the ODN (12) incorporating the methyl acrylate nucleoside pro-
duced adducts with the target DNA and RNA. Figure 2 illustrates
the reactivity of the ODN (12) toward the different target site
(N = A, G, C, T or U) in DNA or RNA. Under acidic conditions, the
slowly (3 days), and this suggested that the reactivity of the vinyl
group with 2a is very low compared to 1, which reacted with the
thiol group within 2 h. The 6-amino group of the nucleoside deriv-
ative (4) was protected with phenoxyacetyl, and subsequent
deprotection of the TBS group with TBAF afforded the diol product
(5). Two subsequent steps following the standard method were
carried out to prepare the phosphoramidite precursor (6) that is
the building block for incorporation into ODN by automated DNA
synthesis. The sulfide-protected ODN (7) was obtained in good
yield by subjecting 6 to an automated DNA synthesizer after puri-
fication with RP-HPLC. The ODN (7) was then converted to ODN (8)
by oxidation with magnesium monoperphthalate, following elimi-
nation of the sulfoxide group under alkaline conditions.
The synthesis of ODN (2b) incorporating a carboxyl-substituted
vinyl derivative is summarized in Scheme 2. The vinylstannane re-
agent (9) was prepared by regioselective hydrostannation of the
terminal alkyne.16 The cross-coupling reaction between 3 and 9
was performed with Pd(PPh3)4 and CuI in DMF to produce the vinyl
derivative in a good yield. The vinyl derivative was dissolved in
2-propanol (0.01 M) and the solution was slowly added to excess
1-octanethiol to produce the octanethio derivative (10).

S. Kusano et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6957–6961
6959
5’
12 : CCGCGT X TCGCCG3’ (X: 6-AVP)
13a: 3’d (GGCGCA N AGCGGC)-FAM5’
13b: 3’r (GGCGCA N AGCGGC)-FAM5’
(A)
(b) pH7.0
(a) pH5.0
DNAtarget
RNAtarget
DNAtarget
RNAtarget
adduct
adduct
ss
N
ss
N
A
G
C
T
A
G
C
U
A
G
C
T
A
G
C
U
(B)
DNA target ((12+13a (N=dG, dA, dC))
(b) pH 7.0
(a) pH 5.0
N=
dG
40
40
30
N=
dA
dG
30
20
10
dC
dA
20
10
dC
10
20
10
20
Reaction time (h)
Reaction time (h)
RNA target ((12+13b (N=G, A, C))
(a) pH 5.0
(b) pH 7.0
N=
40
30
40
G
C
N=
A
30
20
10
20
10
A
C
G
10
Reaction time (h)
20
10
20
Reaction time (h)
Figure 2. Analysis of cross-linking reactions by gel electrophoresis. (A) Gel electrophoresis of the cross-linking reaction after 24 h (B) Comparison of the reaction yields
calculated from the gel electrophoresis analysis of the cross-linking using 12 and target DNA (13a) (N = dG, dA, dC) or RNA (13b) (N = G, A, C). The reaction was performed
with 10 lM ODN (12) and 5 lM target ODN (13) in 0.1 M NaCl, 50 mM MES, pH 5.0 (a) or pH 7.0 (b), at 30 °C.
reactivity decreased in the order of A ꢀ G > C ꢁ T to DNA and
changed to G > C > A ꢁ T, U under neutral conditions. Contrary to
our design, ODN (12) produced the adducts to thymine or uracil
A > G > C > U to RNA. On the other hand, the order of reactivity

6960
S. Kusano et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6957–6961
12:
5’d(CCGCGT
X
N
TCGCCG)3’ (X: 6-AVP)
13a: 3’d(GGCGCA
AGCGGC)-FAM5’
(A)
N=dA(pH=5.0)
N=dG(pH=7.0)
N=dC(pH=7.0)
adduct
adduct
adduct
13
12
13
12
13
12
48 h
6 h
48 h
6 h
48 h
6 h
2 h
0 h
2 h
0 h
2 h
0 h
10
(min.)
10
(min.)
(min.)
(B)
(C)
0
5
10
15
0
5
10
15
(min.)
0
5
10
15
(min.)
(min.)
12a: calcd. (M-H)- 3994.6
12a: calcd. (M-H)- 3994.6
b : calcd. (M-H )- 3980.6
12a: calcd. (M-H)- 3994.6
b : calcd. (M-H )- 3980.6
b : calcd. (M-H )- 3980.6
13(N=dG): calcd. (M-H)- 4562.1
13(N=dA): calcd. (M-H)- 4546.1
13(N=dC): calcd. (M-H)- 4522.1
4545.088
4523.945
8549.992
8505.499
4565.765
3981.901
3979.635
3978.588
8527.915
Figure 3. Analysis of the cross-linking reactions with 12 and target DNA (13a) (N = dG, dA, dC) by HPLC. The reaction was performed with 50 lM ODN (12) and 25 lM target
ODN (13) in 0.1 M NaCl, 50 mM MES, pH 5.0 or pH 7.0 at 30 °C. (A) Time course of cross-linking reactions analyzed by reversed-phase HPLC. (B) HPLC charts of purified
adducts. (C) MALDI-TOF Mass spectra of adducts.
in low yields. The time course of the cross-linking to the target
DNA and RNA of ODN (12) is shown in Figure 2B. The cross-linked
adducts were formed in the highest yields by the reaction with
adenine in DNA and RNA under acidic conditions. On the other
hand, under neutral conditions, the ODN (12) produced adducts
with guanine in DNA and RNA in the highest yields. It is interesting
that the change in the pH conditions causes a drastic change in the
base selectivity. The thermal stability of the duplex was estimated
by measuring the melting temperature (Tm) of the duplex formed
between the stable ODN (7) or ODN (8) and DNA targets under
neutral conditions (Supplementary data). The melting profile of
ODN (7) or (8) suggested that the stable duplex would be formed
with the target DNAs under the reaction conditions. The most
favorable Tm values were found when the 6-AVP derivative is
placed opposite to dT, however, the corresponding duplexes are
less stable than the A:T duplex by 4 °C. A comparison of the Tm
values between the 2-AVP derivative and 6-AVP derivative indi-
cates that the destabilizing effect of the 6-AVP derivatives is lower
than that of the 2-AVP derivative. These results have suggested
that 6-AVP (2) could form a more stable duplex than the 2-ami-
no-6-vinylpurine (2-AVP) (1). However, the reactivity of 2-AVP
(1) with the target base is higher than that of 6-AVP (2).12 The
differences in the thermal stability could not explain the reaction
efficiency and base selectivity of the cross-linking reactions.
NH2
N
N
N
12a: R1=Me
12b: R1=H
N
O
COOR1
O
12
O
NH2
N
R1=H,
R1=H,
R1=H,
R2=13a (N=dA): cal: 8527.7
N
R2=13a (N=dG): cal: 8543.7
R2=13a (N=dC): cal: 8503.7
N
N
R2
COOR1
O
O
R1=CH3, R2=13a (N=dA): cal: 8541.7
R1=CH3, R2=13a (N=dG): cal: 8557.7
R1=CH3, R2=13a (N=dC): cal: 8517.7
O
14
Figure 4. Suggested structure of adducts from molecular weight.
The cross-linking reactions between ODN (12) and the target
DNA bearing G, A or C were also analyzed by HPLC, and the chro-
matograms are summarized in Figure 3.
The mixing of 12a and 13a in the buffer produced three or four
peaks, which might be attributed to the duplex or some complexes,
that were observed in the HPLC chart. After 48 h, these peaks

S. Kusano et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6957–6961
6961
essentially decreased and converged into a main peak with the
Supplementary data
retention time of around 10 min in each reaction. Each new peak
was isolated and its MALDI-TOF mass spectra measured. The ob-
served molecular weight of the cross-linked adducts was in accord
with the general structure (14) containing the target strand (13a)
and reactive ODN (12b) with a carboxylic acid (R1=H) (Fig. 4).
The unreacted ODNs (12) after 48 h in each reaction displayed a
molecular weight corresponding to 12 with a carboxylic acid and
methyl ester (R1=CH3). The 6-AVP with carboxylic acid (12b) was
isolated by HPLC and the cross-linking reactions with target DNA
(13) were examined, but no cross-linked products were observed.
These results suggested that the ODN (12a) with a methyl ester re-
acted with target DNA and that the cross-linked products might be
hydrolyzed under the reaction conditions.
Each purified cross-linked product was cleaved by a hydroxyl
radical for confirmation that the cross-linking reactions of 12 took
place at the opposing base in the target strand. However, the cross-
linked products were unstable under the hydroxyl radical cleavage
conditions and decomposed into each strand. To obtain more in-
sight into the cross-linking reaction, the purified cross-linked prod-
uct was digested by enzymatic hydrolysis, but the cross-linked
product was not observed in the hydrolysates. Thus, it is difficult
to precisely determine each structure.
In conclusion, we have synthesized the 6-AVP derivatives, ex-
pected to cross-link with thymine, and evaluated the cross-linking
reactivity. The ODN containing the non-substituted 6-AVP deriva-
tive (2a) can form a stable duplex with the target DNA under the
stated reaction conditions but did not produce any adducts to
the target DNA or RNA, due to the low reactivity of the vinyl group.
On the other hand, introduction of a carboxyl group onto the vinyl
group of the 6-AVP derivative (2b) increased the cross-linking
reactivity to produce the adducts except for the thymine target
base. These results have provided useful information for the design
of new cross-linking agents.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/
j.bmcl.2012.08.122.
References and notes
1. Prakash, T. P. Chem. Biodivers. 2011, 8, 1616.
2. (a) Bennett, C. F.; Swayze, E. E. Annu. Rev. Pharmacol. Toxicol. 2010, 50, 259; (b)
Aboul-Fadl, T. Curr. Med. Chem. 2005, 12, 2193.
3. Lennox, K. A.; Behlke, M. A. Gene Ther. 2011, 18, 1111.
4. (a) Yokota, T.; Takeda, S.; Lu, Q. L.; Partridge, T. A.; Nakamura, A.; Hoffman, E. P.
Arch. Neurol. 2009, 66, 32; (b) Aartsma-Rus, A. RNA Biol. 2010, 7, 453.
5. (a) Esau, C.; Kang, X. L.; Peralta, E.; Hanson, E.; Marcusson, E. G.; Ravichandran,
L. V.; Sun, Y. Q.; Koo, S.; Perera, R. J.; Jain, R.; Dean, N. M.; Freier, S. M.; Bennett,
C. F.; Lollo, B.; Griffey, R. J. Biol. Chem. 2004, 279, 52361; (b) Krutzfeldt, J.;
Rajewsky, N.; Braich, R.; Rajeev, K. G.; Tuschl, T.; Manoharan, M.; Stoffel, M.
Nature 2005, 438, 685; (c) Davis, S.; Lollo, B.; Freier, S.; Esau, C. Nucleic Acids Res.
2006, 34, 2294; (d) Hutvagner, G.; Simard, M. J.; Mello, C. C.; Zamore, P. D. Plos
Biol. 2004, 2, 465.
6. Flynt, A. S.; Li, N.; Thatcher, E. J.; Solnica-Krezel, L.; Patton, J. G. Nat. Genet. 2007,
39, 259.
7. (a) Chan, J. A.; Krichevsky, A. M.; Kosik, K. S. Cancer Res. 2005, 65, 6029; (b)
Orom, U. A.; Kauppinen, S.; Lund, A. H. Gene Ther. 2006, 372, 137.
8. (a) Berezikov, E. Nat. Rev. Genet. 2011, 12, 846; (b) McDermott, A. M.;
Heneghan, H. M.; Miller, N.; Kerin, M. J. Pharm. Res. 2011, 28, 3016.
9. (a) Xu, Y.; Ito, K.; Suzuki, Y.; Komiyama, M. J. Am. Chem. Soc. 2010, 132, 631; (b)
Higuchi, M.; Kobori, A.; Yamayoshi, A.; Murakami, A. Bioorg. Med. Chem. 2009,
17, 475; (c) Qiu, Z.; Lu, L.; Jian, X.; He, C. J. Am. Chem. Soc. 2008, 130, 14398; (d)
Peng, X. H.; Hong, I. S.; Li, H.; Seidman, M. M.; Greenberg, M. M. J. Am. Chem. Soc.
2008, 130, 10299; (e) Yoshimura, Y.; Fujimoto, K. Org. Lett. 2008, 10, 3237.
10. Op de Beeck, M.; Madder, A. J. Am. Chem. Soc. 2011, 133, 796; (b) Stevens, K.;
Madder, A. Nucleic Acids Res. 2009, 37, 1555; (c) Weinert, E. E.; Dondi, R.;
Colloredo-Melz, S.; Frankenfield, K. N.; Mitchell, C. H.; Freccero, M.; Rokita, S. E.
J. Am. Chem. Soc. 2006, 128, 11940.
11. Fujimoto, K.; Konishi-Hiratsuka, K.; Sakamoto, T.; Yoshimura, Y. Chem.
Commun. 2010, 46, 7545.
12. Kawasaki, T.; Nagatsugi, F.; Ali, M. M.; Maeda, M.; Sugiyama, K.; Hori, K.;
Sasaki, S. J. Org. Chem. 2005, 70, 14.
13. McLaughlin, L. W.; Leong, T.; Benseler, F.; Piel, N. Nucleic Acids Res. 1988, 16,
5631.
14. Nagatsugi, F.; Tokuda, N.; Maeda, M.; Sasaki, S. Bioorg. Med. Chem. Lett. 2001,
11, 2577.
Acknowledgments
15. Nair, V.; Purdy, D. F. Tetrahedron 1991, 47, 365.
16. Darwish, A.; Lang, A.; Kim, T.; Chong, J. M. Org. Lett. 2008, 10, 861.
This work was supported by a Grant-in-Aid for Scientific Re-
search (B) and (S) from Japan Society for the Promotion of Science
(JSPS), CREST from Japan Science and Technology Agency.