Synthesis of 2',5'-oligoadenylate analogs containing an adenine acyclonucleoside and their ability to activate human RNase L.

Article abstract of DOI:10.1016/j.bmcl.2004.06.071

This paper described synthesis of 2',5'-oligoadenylate (2-5A) analogs containing the purine acyclonucleoside, 9-[(2'S,3'R)-2',3',4'-trihydroxybutyl]adenine (2). The ability of the analogs to activate recombinant human RNase L was evaluated using 5'-32P-r(

Full text of DOI:10.1016/j.bmcl.2004.06.071

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4431–4434
Synthesis of 20,50-oligoadenylate analogs containing an adenine
acyclonucleoside and their ability to activate human RNase L
Yoshihito Ueno, Shinji Ishihara, Yasutomo Ito and Yukio Kitade*
Department of Biomolecular Science, Faculty of Engineering, Gifu University,
Yanagido, Gifu 501-1193, Japan
Received 10 May 2004; accepted 17 June 2004
Abstract—This paper described synthesis of 20,50-oligoadenylate (2–5A) analogs containing the purine acyclonucleoside, 9-
[(20S,30R)-20,30,40-trihydroxybutyl]adenine (2). The ability of the analogs to activate recombinant human RNase L was evaluated
using 50-32P-r(C11U2C7)-30 as a substrate. The EC50 value (the concentration of the 2–5A required to cleave half of the RNA) of
the parent 2–5A tetramer 13 was 1.0nM, whereas those of the analog 14 incorporating 2 at the second position from the 50-end
and the analog 15 incorporating 2 at the third position from the 50-end were 9.0 and 1.7nM, respectively. The analogs 14 and
15 were only 9- and 1.7-fold less potent than the parent 2–5A 13 itself, in RNase L activation ability. Furthermore, the oligodeoxy-
nucleotide containing 2 was more resistant to nucleolytic hydrolysis by snake venom phosphodiesterase (a 30-exonuclease) than the
unmodified oligodeoxynucleotide. Thus, incorporation of an acyclonucleoside into 2–5A may be useful for developing an antiviral
agent based on the 2–5A system.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Interferons mediate diverse and important cellular re-
sponses such as the induction of an antiviral state, pro-
liferation, and various immunological processes.1,2 One
extensively characterized mechanism of interferon ac-
tion is the 2–5A system.1–3 Oligoadenylates that linked
20–50 (2–5A) are formed from ATP by 2–5A synthetase
in response to interferon and double-stranded RNA.
These 2–5As bind to and allosterically activate a latent
endoribonuclease (RNase L). The activated RNase L
cleaves viral and cellular RNAs on the 30-sides of UpNp
sequences.4 The RNA degradation results in inhibition
of protein synthesis and thereby inhibition of viral rep-
lication. Another member in the 2–5A system is the
20–50 phosphodiesterase that rapidly degrades the 2–
5As.
chains mimicking the sugar portion of naturally occur-
ring nucleosides have also been synthesized to evaluate
their antiviral activities.6–8 It is expected that incorpora-
tion of acyclonucleosides into 2–5As instead of adeno-
sine (1) (Fig. 1) would enhance their stability against
nucleolytic degradation by nucleases found inside cells.
However, the biological property of a 2–5A analog
incorporating an acyclonucleoside has not been re-
ported.9 In this paper, we wish to report the synthesis
of the 2–5A analogs containing the purine acyclonucleo-
side, 9-[(20S,30R)-20,30,40-trihydroxybutyl]adenine (2),
and their ability to activate recombinant human RNase
L. The resistance of the oligodeoxynucleotide containing
2 to nucleolytic hydrolysis by snake venom phosphodi-
esterase was also examined.
During the past two decades, various 2–5A analogs have
been synthesized to study the biological function of 2–
5A and its structure–activity relationshipand to explore
the potential for deriving an antiviral agent based on the
2–5A system.5 On the other hand, various acyclonucleo-
sides in which the carbohydrate moieties are acyclic
NH2
NH2
N
N
N
N
N
N
N
HO
N
O
HOH2C
H
H
HO OH
HO OH
Keywords: Adenine; Acyclonucleoside; 2–5A; RNA; RNase L.
*
1
2
Corresponding author. Tel./fax: +81-58-293-2640; e-mail: kitade@
biomol.gifu-u.ac.jp
Figure 1. Structures of nucleosides.
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.071

4432
Y. Ueno et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4431–4434
9-[(20S,30R)-40-Hydroxy-20,30-isopropylidenedioxybu-
tyl] adenine (4), which was prepared by the reported
method,10,11 was treated with TMSCl in pyridine and
then reacted with benzoyl chloride (BzCl) to produce
the N6-Bz derivative 5 (Scheme 1). Treatment of 5 with
80% CH3CO2H afforded the triol 6 in 44% yield from 4.
The primary hydroxyl group of 6 was protected with a
4,40-dimethoxytrityl (DMTr) groupto give 7 in 94%
yield. O-DMTr derivative 7 was treated with tert-but-
yldimethylsilyl chloride (TBDMSCl) to afford 20-O-
TBDMS and 30-O-TBDMS derivatives, 812 and 9,13 in
7% and 35% yields, respectively. 30-O-TBDMS deriva-
tive 9 was phosphitylated by a standard procedure to
give the corresponding phosphoramidite 1014 in 65%
yield.
The 2–5A analogs 14 and 15 were synthesized using the
phosphoramidites 10, 11, and 1215 with a DNA/RNA
synthesizer (Fig. 2). The analog 14 has 2 at the second
position from the 50-end, whereas the analog 15 has 2
at the third position from the 50-end. To examine the
resistance of the oligodeoxynucleotide containing 2 to
nucleolytic hydrolysis by snake venom phosphodiest-
erase, 50-d(A2G2A32GAG2A3GA)-30 (16)18 was also
synthesized. The fully protected 2–5As (1lmol each)
linked to the solid support were treated with concen-
trated NH4OH/EtOH (3:1, v/v) at 55ꢁC for 12h and
then 1M TBAF/THF at room temperature for 12h.
The released 2–5As were purified by reversed phase
HPLC to give deprotected 2–5As 14 and 15 in 1.8 and
2.4 OD260 units, respectively. These 2–5As were ana-
NHR
N
N
Ad
AdBz
H
AdBz
H
N
HO
N
O
HOH2C
H
ROH2C
H
DMTrOH2C
H
ref. 11
b
d
H
R1O OR2
O
O
O
O
HO OH
8: R1 = H, R2 = TBDMS
9: R1 = TBDMS, R2 = H
10: R1 = TBDMS, R2 =
4: R = H
5: R = Bz
6: R = H
7: R = DMTr
3
a
c
e
-P Ni-Pr2
OCH2CH2CN
NHBz
N
N
N
O
S
N
DMTrO
O
DMTrO
O
P
Ni-Pr2
O
OCH2CH2CN
12
TBDMSO
O
P
Ni-Pr2
OCH2CH2CN
11
Scheme 1. (a) (1) TMSCl, pyridine, rt; (2) BzCl, pyridine, rt; (b) 80% CH3CO2H, 60ꢁC, 44% from 4; (c) DMTrCl, pyridine, rt, 94%; (d) TBDMSCl,
imidazole, DMF, rt, 7% (8), 35% (9); (e) 2-cyanoethyl N,N-diisopropylchlorophosphoramidite, N,N-diisopropylethylamine, THF, rt, 65%.
NH2
NH2
N
NH2
N
N
N
N
N
N
N
N
O
O
P
O–
O
N
HO
P
O–
O
O
N
N
HO
O
HO P
O
O
NH2
N
NH2
N
O
O–
NH2
N
N
N
N
N
N
N
HO
O
HO
O
HO
O
N
N
O
P
O–
O
O
N
O P
O–
O
O
P
O–
O
NH2
NH2
N
O
NH2
N
H
H
N
N
N
N
N
N
N
HO
O
HO
O
HO
O
N
N
O P
O
N
O P
O–
O
O
O
O P
O–
O
O–
NH2
N
NH2
N
NH2
H
H
N
N
N
N
N
N
N
HO
O
HO
O
P
HO
O
N
N
N
O
P
O–
O
O
O
O
P
O–
O
O
O
O
O–
13
14
15
HO OH
HO OH
HO OH
Figure 2. Structures of 2–5As.

Y. Ueno et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4431–4434
4433
lyzed by matrix-assisted laser desorption/ionization
time-of-flight mass spectrometry (MALDI-TOF/MS),
and the observed molecular weights supported their
structures.18
The ability of 2–5A analogs 14 and 15 to activate RNase
L was estimated by monitoring the cleavage of a syn-
thetic RNA by the activated RNase L. In this study,
50-r(C11U2C7)-30 was used as a substrate. Carroll et al.
reported that initial cleavage of RNA by RNase L
occurs on the 30-side of r(C11U2) to yield r(C11UpUp),
and second cleavage occurs on the 30-side of r(C11U)
to give r(C11Up) with a higher enzyme concentration
or longer incubation time.4 The abilities of 2–5A analogs
14 and 15 to activate the enzyme were represented in
terms of the concentrations of the analogs required to
cleave half of the RNA (EC50). Recombinant human
RNase L was expressed in E. coli and purified according
to the reported procedure.19 The RNA labeled at the 50-
end with 32P was incubated with the enzyme (60nM)
that had been pre-incubated with the 2–5A analogs 14
and 15.20 The reactions were analyzed by polyacryla-
mide gel electrophoresis under denaturing conditions
(Fig. 3a). The densities of radioactive bands on the gel
were determined with a bio-imaging analyzer (Fig. 3b).
The EC50 value of the 2–5A tetramer 13 was 1.0nM,
whereas those of the analogs 14 and 15 were 9.0 and
1.7nM, respectively. Although the abilities of 14 and
15 to activate RNase L were weaker than that of the
parent 2–5A 13, the EC50 values of 14 and 15 were only
9- and 1.7-fold higher than that of 13, respectively.
Figure 4. Polyacrylamide gel electrophoresis of 50-32P-labeled oligode-
oxynucleotides hydrolyzed by snake venom phosphodiesterase: (lanes
1–6) 50-32P-d(A2G2A4GAG2A3GA)-30; (lanes 7–12) 50-32P-d(A2G2A32-
GAG2A3GA)-30. The oligodeoxynucleotides were incubated with
snake venom phosphodiesterase for 0min (lanes 1 and 7), 5min (lanes
2 and 8), 10min (lanes 3 and 9), 20min (lanes 4 and 10), 30min (lanes 5
and 11), and 60min (lanes 6 and 12).
of incubation. In contrast, it turned out that the phos-
phodiester linkage at the 30-side of dG at the 30-side of
the analog 2 was highly resistant to the enzyme.
In conclusion, we have demonstrated the synthesis of
the 2–5A analogs containing the purine acyclonucleo-
side, 9-[(20S,30R)-20,30,40-trihydroxybutyl]adenine (2).
The abilities of the analogs to activate recombinant
human RNase L were evaluated using 50-32P-r(C11U2C7)-
30 as a substrate. It was found that the analog 14 con-
taining 2 at the second position from the 50-end and
the analog 15 containing 2 at the third position
from the 50-end were only 9- and 1.7-fold less potent
than the parent 2–5A tetramer itself, in RNase L activa-
tion ability. Furthermore, the oligodeoxynucleotide con-
taining 2 was more resistant to nucleolytic hydrolysis by
snake venom phosphodiesterase than the unmodified
oligodeoxynucleotide. Thus, incorporation of an acyclo-
nucleoside into 2–5A may be useful for developing a
novel antiviral agent based on the 2–5A system.
We next examined the stability of the oligodeoxynu-
cleotide 16 containing 2 to nucleolytic hydrolysis by
snake venom phosphodiesterase (a 30-exonuclease).
The oligodeoxynucleotide 16 labeled at the 50-end with
32P was incubated with the enzyme, and the reac-
tions were analyzed by polyacrylamide gel electropho-
resis under denaturing conditions (Fig. 4). The control
oligodeoxynucleotide,
was hydrolyzed randomly by the enzyme after 60min
50-d(A2G2A4GAG2A3GA)-30,
Figure 3. (a) Polyacrylamide gel electrophoresis of 50-32P-C11U2C7-30 hydrolyzed by recombinant human RNase L activated with 2–5A 13 (lanes
1–5), 2–5A analog 14 (lanes 6–10), or 2–5A analog 15 (lanes 11–15). Concentrations of 2–5As: 0nM (lanes 1, 6, and 11), 1nM (lanes 2, 7, and 12),
5nM (lanes 3, 8, and 13), 10nM (lanes 4, 9, and 14), and 20nM (lanes 5, 10, and 15). Concentration of the RNA substrate: 100nM. Concentration of
recombinant human RNase L: 60nM. (b) The graph represents percent RNA cleavage determined with a bio-imaging analyzer.

4434
Y. Ueno et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4431–4434
J=5.2Hz, 30-OH), 4.37 (2H, m, 10-H), 4.21 (1H, m, 20-H),
Acknowledgements
3.75–3.69 (7H, m, 30-H and OMe·2), 3.10 (1H, dd,
J=11.3 and 5.0Hz, 40-H), 3.03 (1H, dd, J=11.3 and
6.4Hz, 40-H), 0.57 (9H, s, Sit-Bu), ꢀ0.30 (3H, s, SiMe),
ꢀ0.54 (3H, s, SiMe), the assignments were in agreement
with COSY spectrum; FAB-HRMS calcd for
C43H50N5O6Si (MH+), 760.3530; found, 760.3535.
This research was in part supported by a Grant-in-Aid
for Scientific Research (B) (KAKENHI 15390036) from
Japan Society for the Promotion of Science (JSPS).
13. Physical data of 9: 1H NMR (400MHz, DMSO-d6): d
11.12 (1H, s, 6-NH), 8.70 (1H, s, 8- or 2-H), 8.31 (1H, s, 2-
or 8-H), 8.05–6.86 (18H, m, Bz and DMTr), 5.21 (1H, d,
J=5.2Hz, 20-OH), 4.48 (1H, d, J=10.0Hz, 10-H), 4.09–
4.05 (2H, m, 10-H and 20-H), 3.82 (1H, m, 30-H), 3.72 (6H,
s, OMe·2), 3.19 (1H, dd, J=9.9 and 5.2Hz, 40-H), 3.05
(1H, dd, J=9.9 and 5.4Hz, 40-H), 0.85 (9H, s, Sit-Bu), 0.08
(3H, s, SiMe), ꢀ0.06 (3H, s, SiMe), the assignments were
in agreement with COSY spectrum; FAB-HRMS calcd for
C43H50N5O6Si (MH+), 760.3530; found, 760.3535.
References and notes
1. Lengyel, P. Annu. Rev. Biochem. 1982, 51, 251.
2. Pestka, S.; Langer, J. A.; Zoon, K. C.; Samuel, C. E.
Annu. Rev. Biochem. 1987, 56, 727.
3. Torrence, P. F.; Xiao, W.; Li, G.; Khamnei, S. Curr. Med.
Chem. 1994, 1, 176.
4. Carroll, S. S.; Chen, E.; Viscount, T.; Geib, J.; Sardana,
M. K.; Gehman, J.; Kuo, L. C. J. Biol. Chem. 1996, 271,
4988.
5. Player, M. R.; Torrence, P. F. Pharmacol. Ther. 1998, 78,
55.
6. Lerner, L. M. Carbohydr. Res. 1984, 127, 141.
7. Bessodes, M.; Antonakis, K. Tetrahedron Lett. 1985, 26,
1305.
8. Chu, C. K.; Cutler, S. J. J. Heterocycl. Chem. 1986, 23,
289.
9. Synthesis of the 2–5A analogs consisting of 9-[(20R,40S)-
10,50-dihydroxy-40-hydroxymethyl-30-oxapent-2-yl]adenine
has been reported: Mikhailov, S. N.; Pfleiderer, W.
Tetrahedron Lett. 1985, 26, 2059.
10. Kitade, Y.; Hirota, K.; Maki, Y. Tetrahedron Lett. 1993,
34, 4835.
14. Physical data of 10: 31P NMR (162MHz, CDCl3): d
149.78, 146.81; FAB-HRMS calcd for C52H67Z7O7PSi
(MH+), 960.4609; found, 960.4612.
15. The phophoramidite 11 was synthesized as reported
previously.16 The phophoramidite 1217 was purchased
from GLEN Research.
16. Ueno, Y.; Kato, Y.; Okatani, S.; Ishida, N.; Nakanishi,
M.; Kitade, Y. Bioconjugate Chem. 2003, 14, 690.
17. Horn, T.; Urdea, M. S. Tetrahedron. Lett. 1986, 27, 4705.
18. Data of MALDI-TOF/MS. 2–5A 14: calcd for
C39H51N20O24P4 (MH+), 1307.2; found, 1307.1. 2–5A
15: calcd for C39H51N20O24P4 (MH+), 1307.2; found,
1306.0. The oligodeoxynucleotide 16: calcd for C169H205
N85O89P16 (M+), 5346.5; found, 5346.2.
-
11. Hirota, K.; Monguchi, Y.; Sajiki, H.; Sako, M.; Kitade, Y.
J. Chem. Soc., Perkin. Trans. 1 1998, 941.
12. Physical data of 8: 1H NMR (400MHz, DMSO-d6): d
11.13 (1H, s, 6-NH), 8.70 (1H, s, 8- or 2-H), 8.33 (1H, s, 2-
or 8-H), 8.06–6.86 (18H, m, Bz and DMTr), 5.44 (1H, d,
19. Yoshimura, A.; Nakanishi, M.; Yatome, C.; Kitade, Y. J.
Biochem. 2002, 132, 643.
20. The extinction coefficients of the 2–5As at 260nm were
estimated as 4.16·104.