Synthesis of TMG-capped RNA-DNA chimeric oligonucleotides

Article abstract of DOI:10.1016/S0040-4039(02)02828-9

This paper deals with the synthesis of m₃^2,2,7G^5′pppAmpUmpApd(CpTpTpApCpCpT), a 5′-TMG-capped RNA-DNA chimeric oligonucleotide, which is expected to be conveyed to the nucleus by snurportin 1, a nuclear-transport protein. This 5′-TMG-capped RNA-DNA chimeric molecule was synthesized by the enzymatic condensation of m₃^2,2,7G^5′pppAmpUmpA with A^5′ppd(CpTpTpApCpCpT) in the presence of RNA ligase. An inherent serious side reaction was disclosed in the 5′-adenylation of d(CpTpTpApCpCpT) on the solid support, but the use of an active ester intermediate as the pA donor gave an improved result.

Full text of DOI:10.1016/S0040-4039(02)02828-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 1703–1707
Synthesis of TMG-capped RNA–DNA chimeric oligonucleotides
Mitsuo Sekine,* Masatoshi Ushioda, Takeshi Wada and Kohji Seio
Department of Life Science, Tokyo Institute of Technology, Nagatsuta, Midoriku, Yokohama 226-8501, Japan
Received 7 August 2002; revised 8 December 2002; accepted 13 December 2002
Abstract—This paper deals with the synthesis of m23,2,7G5%pppAmpUmpApd(CpTpTpApCpCpT), a 5%-TMG-capped RNA–DNA
chimeric oligonucleotide, which is expected to be conveyed to the nucleus by snurportin 1, a nuclear-transport protein. This
5%-TMG-capped RNA–DNA chimeric molecule was synthesized by the enzymatic condensation of m23,2,7G5%pppAmpUmpA with
A5%ppd(CpTpTpApCpCpT) in the presence of RNA ligase. An inherent serious side reaction was disclosed in the 5%-adenylation
of d(CpTpTpApCpCpT) on the solid support, but the use of an active ester intermediate as the pA donor gave an improved result.
© 2003 Published by Elsevier Science Ltd.
The mechanism of nuclear transport of U1snRNA1
having a unique 2,2,7-trimethylguanosine (TMG)-cap
structure at the 5%-end was clarified by cooperative
studies of Lu¨hrmann and us in 1998.2 The mechanism
disclosed involves participation of snurportin 1, a spe-
cific protein that can bind to the TMG-cap structure of
U1snRNA. This TMG-binding protein can also bind to
importin b which is well known to carry nucleoproteins
from the cytoplasm to the nucleus by triparticle forma-
tion of nucleoprotein–importin a–improtin b.3 Thus, it
turned out that the nuclear transport of U1snRNA
occurs by formation of a triparticle of U1snRNA–snur-
portin 1–importin b in a manner very similar to that of
nuclear proteins.4 Namely, snurportin 1 plays an
important role in transporting U1snRNA as the substi-
tute of importin b. Based on this finding, we came up
with a new idea that, if antigene oligonucleotides can be
capped by the TMG-cap structure at the 5%-terminal
site, they can be conveyed to the nucleus with the help
of snurportin 1 and importin b so that they can bind
selectively to DNA duplexes. This process creates a new
promising tool in antigene strategy.5
A5%ppd(CpTpTpApCpCpT) (23), with m32,2,7G5%ppp-
AmpUmpA (24), i.e. 5%-terminal fragment of
a
U1snRNA, since we recently established a facile solid-
phase synthesis of the latter TMG-capped RNA trimer
block without extensive workup.6 There are few papers
regarding the synthesis of RNA–DNA chimeric
oligonucleotides by use of RNA ligase.7 McLaughlin8
reported the details of RNA ligation of RNA oligomers
with 5%-adenylated deoxynucleoside 3%,5%-bisphosphate.
Without 5%-adenylation, in general RNA ligase cannot
recognize sufficiently the substrate so that the coupling
yield is extremely low.8 Therefore, we studied a more
straightforward method for the synthesis of 5%-adenyl-
In this paper, we report the synthesis of a TMG-
capped
RNA–DNA
chimeric
oligonucleotide,
m23,2,7G5%pppAmpUmpApd(CpTpTpApCpCpT) (26), as
a possible molecule having the above-designed function
in antigene strategy.
Scheme 1. Reagents and conditions: (a) TMSCl (2 equiv.),
pyridine, rt, 20 min; (b) DMTrCl (1.5 equiv.), DMAP (cat.),
To synthesize the TMG-capped RNA–DNA chimeric
oligonucleotide, we used the RNA ligase-mediated con-
densation of a 5%-adenylated heptadeoxynucleotide,
pyridine, rt,
4 h, 62%; (c) cyclohexylammonium S,S-
diphenylphosphoroditioate (1.2 equiv.), 1H-tetrazole (4
equiv.), isodurenedisulfonyl dichloride (2 equiv.), pyridine, rt,
30 min (68%); (d) 5 M pyridinium phosphinate in pyridine–
tiethylamine (2:1, v/v), rt, 1 h (93%).
* Corresponding author. Tel.: +81-45-924-5706; fax: +81-45-924-5772;
e-mail: msekine@bio.titech.ac.jp
0040-4039/03/$ - see front matter © 2003 Published by Elsevier Science Ltd.
PII: S0040-4039(02)02828-9

1704
M. Sekine et al. / Tetrahedron Letters 44 (2003) 1703–1707
ated deoxyoligonucleotides as effective donor molecules
for the ligation. Horn and Urdea reported the 5%-phos-
phorylation of oligodeoxynucleotides synthesized on
polymer supports by using a 5%-terminal phosphorylating
reagent having a DMTr group,9 by which the coupling
yield of the 5%-phosphorylation can be estimated by UV
colorimetric assay of the DMTr cation. In the case of the
5%-adenylation, it is convenient to see if the coupling is
successfully carried out by using such an expeditous
method. Therefore, we synthesized a DMTr-containing
5%-adenylating reagent, PhSpDMTrAmM (4: triethylammo-
nium salt), from 2%,3%-O-methoxymethyleneadenosine
(1), as shown in Scheme 1. Reaction of 1 with DMTrCl
gave the N-tritylated product 2 in 62% yield. Reaction
Scheme 2. Reagents and conditions: (a) DMTrO(CH2)2-
SO2(CH2)2O-P(NiPr2)(OCH2CH2CN) (20 equiv.) 1H-tetrazole
(80 equiv.), CH3CN, rt, 5 min; (b) 0.12 MDMAP, pyridine–
Ac2O (9:1, v/v, rt, 2 min); (c) I2, pyridine–THF–H2O (10:10:1,
v/v/v), rt, 15 s×3; (d) 2.5% dichloroacetic acid, CH2Cl2, rt, 5
s×3; (e) 10% DBU, pyridine–bis(trimethylsilyl)acetamide (1:1,
v/v), rt, 10 min; (f) 0.1 M of 4 (20 equiv.) in pyridine, I2 (60
equiv.), rt, 5 min; (g) trifluoroacetic acid, CH2Cl2; (h) NH3–
EtOH (4:1, v/v), rt, 1 h.
of
2
with cyclohexylammonium S,S%-diphenyl
phosphorodithioate10 (PSS) in the presence of
isodurenedisulfonyl dichloride10 (DDS) gave the 5%-O-
phosphorylated species in 68% yield. The selective
dephenylthiolation of 3 with the phosphinate reagent10,11
gave the O,S-phosphodiester derivative in quantitative
yield. Before its application to the oligomer-level synthe-
sis, we checked the reaction of 4 with a 5%-phosphorylated
species 6 derived from a T-loaded aminopropyl-CPG12
resin.
The reaction of 4 with this resin in the presence of I2 in
pyridine at room temperature for 5 min followed by the
successive workup with 1% TFA in CH2Cl2 for 1 min and
NH3–EtOH (4:1, v/v) at room temperature for 1 h gave
the desired product (8: A5%ppT) in 48% yield (HPLC). To
our surprise, a symmetric pyrophosphate derivative (9:
T5%ppT) was also obtained in 12% yield. A considerable
amount (40%) of the unreacted species (10: pT) was
recovered, as shown in Scheme 2.
The formation of T5%ppT (9) is unique since we have never
observed such symmetric pyrophosphate derivatives in
Figure 1. The mechanism of the side reaction observed in the
solid-phase synthesis of A5%ppT (8).
the liquid phase synthesis in our long experience. This
DMTr 5%
result implies that the once-formed
A
ppT-poly-
mM
mer (7: X in Fig. 1) reacted with an activated methaphos-
phate species (11) of 4 to give a triphosphate derivative
(Y) which in turn allowed attack of a neighbouring pT
residue (Z) to decompose to give (pT)2-polymer (W) and
(pDMTrAmM)2. This might be because some of the pT
residues were brought into closer proximity.
To elucidate this hypothesis, an independent reaction of
N4,3%-O-dibenzoyl-5%-deoxycytidylic acid (12: pdBzCBz)
with 2 equiv. of PhS-pDMTAmM 4 was carried out in the
presence of 6 equiv. of I2 in pyridine–pyridine-d5 (3:1,
v/v).13 Actually, the 31P NMR spectrum obtained after
10 min showed that the desired product 13 at −7.5 ppm
was formed as a minor product, as shown in Figure 2.
Instead, more significant new broad signals X and Y at
around −9 and −18 ppm were observed in the ratio of
2:1. However, it was found that addition of water to the
mixture resulted in convergence of the signals to the
region of typical pyrophosphate derivatives at −7.5
ppm.14 From these results, it is likely that the initial
resonance signals observed at around −18 ppm are due
to the tri-substituted triphosphates14 14 and 15 as
depicted by asterisks in Figure 2, which were formed by
Figure 2. The 13P NMR spectra of the reaction of 4 with 12
in the presence of I2. Panel A: the spectrum after a mixture of
2 and 12 were treated with I2 for 10 min. Panel B: the spectrum
obtained after the mixture was quenched with water for 30 min.

M. Sekine et al. / Tetrahedron Letters 44 (2003) 1703–1707
1705
the reaction of the once-formed initial product 13 with
the methaphosphate derivative 11.
product pd(DMTrAmM) (18) of the active ester, as shown
in Figure 4(B2).
Now, it was concluded that this over-activation of
When the 6-(trifluoromethyl)-1-hydroxybenzotriazole
was replaced with 6-(trifluoromethyl)-4-nitrobenzotri-
azole, the coupling required longer periods of time as
shown in Figure 4(A2). Kadokura et al. recently
reported a pyrophosphate bond formation by use of
such active esters derived from phosphorimidazolidate
derivatives.15 Actually, they found that the use of the
active ester from phosphorimidazolidate derivatives
gave better results than that from S-phenyl phospho-
rothioate derivatives, although the reason is still
unclear. Thus, we also synthesized the active ester 17b
by the same route.
DMTr 5%
A ppT-polymer might be due to the extreme
mM
reactivity of the generated methaphosphate ester
DMT
mM
derivative 11 of pA
, as shown in Figure 2. There-
fore, to control the strong reactivity of the methaphos-
phate intermediate 11, the phosphorothioate derivative
PhSpDMTrAmM 4 was converted to a 6-(trifluoro-
methyl)-4-nitrobenzotriazolyl-1-oxy ester derivative 17b
by treatment with 3 equiv. of 6-(trifluoromethyl)-1-
hydroxyl-4-nitrobenzotriazole 16b in the presence of 3
equiv. of I2 in pyridine–pyridine-d5 (3:1, v/v).
The conversion was completed within 10 min, as evi-
denced by 31P NMR (Fig. 3) which showed that the
starting material at 2.8 ppm disappeared and a new
signal at 1.2 ppm was observed as a single peak.
Addition of this active ester 17b to the triethylammo-
nium salt of 12 resulted in slow formation of the
desired pyrophosphate derivative 13 which appeared at
−8.5 ppm. The pyrophosphate derivative was quantita-
tively formed after 10 min, as shown in Figure 4(B1).
Addition of water after the reaction gave a simple
mixture of the desired product 13 and the hydrolysed
Treatment of 2 with 7 equiv. of diphenyl phosphonate
in pyridine at room temperature for 2 h followed by
addition of triethylamine–H2O (1:1, v/v) gave the 5%-
phosphonate derivative 19 in 68% yield, as shown in
Scheme 3. The reaction of 19 with 1.2 equiv. of
trimethylsilylimidazole in the presence of 4 equiv. of
triethylamine in CH3CN–CCl4 at room temperature for
20 min followed by addition of methanol gave the
desired phosphorimidazolidate derivative 20 in quanti-
tative yield, as evidenced by Figure 5(A).
Addition of 6-(trifluoromethyl)-1-hydroxy-4-nitro-benzo-
triazole to this product gave an equilibrated mixture of
the phosphorimidazolidate (−8.5 ppm) and the active
ester (0.8 ppm), as shown in Figure 5(B). This mixture
was allowed to react with pT-polymer at room temper-
ature for 2 h. Consequently, A5%ppT was formed in 64%
yield, while the formation of T5%ppT was greatly sup-
pressed to only 1% and unreacted pT was recovered in
33% yield. Further attempts to improve the coupling
yield failed. The difficulty in completing the reaction is
an inherent serious problem. It should be noted that the
direct use of 20 without addition of 6-(trifluoromethyl)-
Figure 3. The 13P NMR spectra of the active esters 17a,b
obtained by the reaction of 4 with 16a,b.
Scheme 3. Reagents and conditions: (a) diphenylphosphonate
(7 equiv.), pyridine, rt, 2 h; (b) triethylamine–H2O (1:1, v/v),
rt, 30 min, 68% from 2; (c) trimethylsilylimidazole (1.2
equiv.), triethylamine (4 equiv.), CH3CN–CCl4 (1:1, v/v), rt,
20 min; (d) MeOH, rt, 5 min, quant.
Figure 5. The 13P NMR spectra of the mixture of 20 and 17b
obtained by the reaction of 20 with 16b in pyridine at room
temperature for 15 min. Panel A: before addition of 16b.
Panel B: after addition of 16b.
Figure 4. The 13P NMR profiles of the reaction of 12 with
17a or 17b.

1706
M. Sekine et al. / Tetrahedron Letters 44 (2003) 1703–1707
1-hydroxy-4-nitro-benzotriazole gave lower yields of
the product.
Application of this method to the synthesis of
A5%ppd(CpTpTpApCpCpT) (21) was carried out by use
of 6-N,2%,3%-O-tribenzoyladenosine 5%-phosphorimida-
zolidate (22) which was similarly synthesized from the
corresponding 5%-phosphonate derivative (23) and acti-
vated by addition of 6-(trifluoromethyl)-1-hydroxy-4-
nitrobenzotriazole, as shown in Scheme 4. Thus, the
desired 5%-adenylated product 21 was isolated in 16%
yield. Since, usually, the isolated yields of unmodified
oligodeoxynucleotides are ca. 30%, this isolated yield is
reasonable in consideration of the efficiency of ca. 65%
in the pyrophosphate bond formation.
Figure 6. Anion exchange HPLC profile of the crude 24
obtained by ligation of a mixture of 26 with 25 or 21. Panel
A: without 5%-adenylation; 50 mM Tris–HCl (pH 8), 10 mM
MgCl2, 10 mM DTT, 10 mM ATP. Panel B: with 5%-adenyla-
tion; 50 mM Tris–HCl (pH 8), 10 mM MgCl2, 10 mM DTT.
Finally, we used this 5%-adenylated oligodeoxynucle-
otide derivative to obtain m23,2,7G5%pppAmpUmpApd-
(CpTpTpApCpCpT) (24). To demonstrate the
efficiency of the adenylation on ligase reaction, we
studied both reactions of A5%ppd(CpTpTpApCpCpT)
(21)
and
pd(CpTpTpApCpCpT)
(25)
with
m32,2,7G5%pppAmpUmpA (26)6 in the presence of RNA
ligase under the conditions described in Figure 6. It was
clearly shown that the adenylated species gave better
results than the unmodified oligodeoxynucleotide. After
48 h, the product was formed in 58% yield (HPLC)
while the product was formed in only 10% in the case
of the unmodified oligodeoxynucleotide. Separation of
the desired product by HPLC followed by workup gave
m23,2,7G5%pppAmpUmpApdd(CpTpTpApCpCpT) (24) in
an isolated yield of 19%.
In conclusion, we were able to synthesize the TMG-
capped RNA–DNA chimeric molecule. It should be
noted that the 5%-adenylation on the polymer support
should be improved. Now work is under way in this
direction.
Acknowledgements
This work was supported by a Grant-in-Aid (14208074)
for Scientific Research from the Ministry of Education,
Science, Sports, and Culture, Japan.
References
1. Massenet, S.; Mougin, A.; Branlant, C. In Modification
and Editing of RNA; Grosjean, H.; Benne, R., Eds.
Posttranscriptional modifications in the U small nuclear
RNAs. ASM Press: Washington, DC, 1988; pp. 201–227.
2. Huber, J.; Cronshagen, U.; Kadokura, M.; Marshallsay,
C.; Wada, T.; Sekine, M.; Lu¨hrmann, R. EMBO J. 1998,
17, 4114–4126 and references cited therein.
Scheme 4. Reagents and conditions: (a) chain elongation; (b)
DMTrO(CH2)2SO2(CH2)2O-P(NiPr2)(OCH2-CH2CN)
(20
3. Will, C. L.; Lu¨hrmann, R. Curr. Opin. Cell Biol. 2001, 13,
290–301.
equiv.), 1H-tetrazole (80 equiv.), CH3CN, rt, 5 min; (c) 0.12 M
DMAP, pyridine–Ac2O (1:9, v/v), rt, 2 min; (d) I2, pyridine–
THF–H2O (10:10:1, v/v/v), rt, 15 s×3; (e) 2.5% dichloroacetic
acid, CH2Cl2, rt, 5 s×3; (f) 10% DBU, pyridine–bis(trimethyl-
silyl)acetamide (1:1, v/v), rt, 10 min; (h) an equilibrilized (40
equiv., 0.1 M) of 22 and the active ester obtained by pre-acti-
vation of 22 by addition of 6-(trifluoromethyl)-1-hydroxy-4-
nitrobenzotriazole (120 equiv.) at rt for 10 min, pyridine, rt, 2
h; (i) NH3–EtOH (4:1, v/v), 55°C, 4 h.
4. Izaurralde, E.; Adams, S. RNA 1998, 4, 351–364.
5. Pharmaceutical Aspects of Oligonucleotides; Couvreur, P.;
Malvy, C., Eds.; Taylor & Francis: Philadelphia, 2000.
6. (a) Kadokura, M.; Wada, T.; Seio, K.; Moriguchi, T.;
Huber, J.; Lu¨hrmann, R.; Sekine, M. Tetrahedron Lett.
2001, 42, 8853–8856; (b) Sekine, M.; Kadokura, M.;
Satoh, T.; Seio, K.; Wada, T.; Fischer, U.; Sumper, V.;
Lu¨hrmann, R. J. Org. Chem. 1996, 61, 4412–4422.

M. Sekine et al. / Tetrahedron Letters 44 (2003) 1703–1707
1707
7. England, T. E.; Bruce, A. G.; Uhlenbeck, O. C. Methods
Enzymol. 1980, 65 (Nucleic Acids, Pt. I), 65–74.
8. Hoffmann, P. U.; McLaughlin, L. W. Nucleic Acids Res.
1987, 15, 5289–5303.
Jpn. 1981, 54, 3815–3816; (b) Sekine, M.; Hata, T. Curr.
Org. Chem. 1998, 3, 25–66.
12. Groeger, G.; Seliger, H. Nucleosides Nucleotides 1988, 7,
773–778.
13. Usman, N.; Pon, R. T.; Ogilvie, K. K. Tetrahedron Lett.
1985, 26, 4567–4570.
9. Horn, T.; Urdea, M. S. Tetrahedron Lett. 1986, 27,
4705–4708.
14. Ludwig, J.; Eckstein, F. J. Org. Chem. 1989, 54, 631–635.
15. Kadokura, M.; Wada, T.; Urashima, C.; Sekine, M.
Tetrahedron Lett. 1997, 38, 8359–8362.
10. Sekine, M.; Matsuzaki, J.; Hata, T. Tetrahedron Lett.
1981, 22, 3209–3212.
11. (a) Sekine, M.; Hamaoki, K.; Hata, T. Bull. Chem. Soc.