Convenient RNA synthesis using a phosphoramidite possessing a biotinylated photocleavable group

Article abstract of DOI:10.1021/ol101489v

A convenient RNA synthesis using a biotin-streptavidin interaction and a photocleavable protecting group is described. The biotinylated photocleavable group was introduced at the 2′-position of the uridine derivative. Using the phosphoramidite 12, we attempted the synthesis of a 21mer RNA, which is pure enough to show potent RNAi activity compared with a conventionally prepared and HPLC-purified 21mer RNA with the same sequence.

Full text of DOI:10.1021/ol101489v

ORGANIC
LETTERS
2010
Vol. 12, No. 17
3836-3839
Convenient RNA Synthesis Using a
Phosphoramidite Possessing a
Biotinylated Photocleavable Group
Kota Tomaya,† Mayumi Takahashi,† Noriaki Minakawa,*,‡ and Akira Matsuda†
Faculty of Pharmaceutical Sciences, Hokkaido UniVersity, Kita-12, Nishi-6, Kita-ku,
Sapporo 060-0812, Japan, and Graduate School of Pharmaceutical Sciences,
The UniVersity of Tokushima, Shomachi 1-78-1, Tokushima 770-8505, Japan
minakawa@ph.tokushima-u.ac.jp
Received July 1, 2010
ABSTRACT
A convenient RNA synthesis using a biotin-streptavidin interaction and a photocleavable protecting group is described. The biotinylated
photocleavable group was introduced at the 2′-position of the uridine derivative. Using the phosphoramidite 12, we attempted the synthesis
of a 21mer RNA, which is pure enough to show potent RNAi activity compared with a conventionally prepared and HPLC-purified 21mer RNA
with the same sequence.
After the discovery of RNA interference (RNAi),1 much
effort has been dedicated to the application of short interfer-
ing RNAs (siRNAs), not only as biological tools but also as
therapeutic agents.2 In addition, investigation has also
focused on the discovery of potentially new functions of
RNAs, especially noncoding RNAs such as microRNAs
(miRNAs).3 This trend translates into a high demand for
development of a more rapid and efficient synthetic protocol
for solid-phase RNA synthesis. However, because of the
existence of the 2′-hydroxyl group in its structure, the RNA
synthesis is still comparatively difficult versus the optimized
solid-phase DNA synthesis. Thus, selective protection of the
2′-hydroxyl group is first required. This protecting group
must be stable and less bulky for a coupling reaction with a
phosphoramidite unit throughout the solid-phase synthesis;
yet it must be readily removable under mild conditions. In
addition, after removal of the protecting groups at the 2′-
hydroxyl position, the resulting free 2′-hydroxyl group may
cause cleavage and/or migration of the RNA strand, espe-
cially under basic conditions. Therefore, a more rapid method
of purification after removal of the protecting group at the
2′-position is also required.
For the protecting group at the 2′-position, a tert-
butyldimethylsilyl (TBDMS) group is a popular choice, the
phosphoramidite units of which are widely used for RNA
synthesis. However, since this is not a robust method,
extensive effort is underway to develop a more suitable
protecting group at the 2′-position, such as triisopropylsily-
loxymethyl (TOM) and bis(2-acetoxyethyloxy)methyl (ACE)
protecting groups.4-9 Ohgi et al. have recently developed
one of the most successful examples with a 2-cyanoet-
† Hokkaido University.
‡ The University of Tokushima.
(1) Fire, A.; Xu, S.; Montgomery, M. K.; Kostas, S. A.; Driver, S. E.;
Mello, C. C. Nature 1998, 391, 806–811.
(2) Bumcrot, D.; Manoharan, M.; Koteliansky, V.; Sah, D. W. Nat.
Chem. Biol. 2006, 2, 711–719.
(4) Schwartz, M. E.; Breaker, R. R.; Asteriadis, G. T.; deBear, J. S.;
Gough, G. R. Bioorg. Med. Chem. Lett. 1992, 2, 1019–1024.
(3) Bartel, D. P. Cell 2004, 116, 281–297.
10.1021/ol101489v  2010 American Chemical Society
Published on Web 08/09/2010

hoxymethyl (CEM) protecting group and have succeeded in
preparing a 110mer precursor miRNA using this method.10,11
Apart from these efforts, there have been few reports of novel
technology for rapid purification, which would avoid cleav-
age and/or migration of the resulting RNA strand caused by
the tedious purification procedure. To this end, Olejnik et
al. developed a rapid purification of DNA using a non-
nucleosidic photocleavable biotin phosphoramidite.12 In their
approach, the phosphoramidite is incorporated at the 5′-end
of the DNA strand, so that the resulting 5′-end biotin-labeled
DNA can be selectively isolated from the undesired shorter
DNA fragments by incubation with immobilized streptavidin.
After irradiation with UV light, the photocleavable biotin
moiety at the 5′-end is cleaved to afford the desired DNA
strand. This approach is ingenious for a rapid purification;
however, the resulting DNA strand is inevitably 5′-phos-
phorylated. Fang and Bergstrom reported an RNA synthesis
using a similar strategy based on biotin-avidin interaction.13
In their method, the biotin group is introduced at the 5′-end
of the resulting oligoRNA via a silyl acetal linkage, neces-
sitating the ACE-protected phosphoramidite units,6 and thus
requires unusual reaction conditions in solid-phase RNA
synthesis.
temperature (t1/2 estimated to be 12 h), whereas under 0.05
M K2CO3 in MeOH, it was converted into a 4-methoxypy-
rimidine nucleoside (t1/2 estimated to be 1.3 h). Thus, it was
concluded that incorporation of the functional group at the
uracil moiety was not suitable for our approach (data not
shown).
We next planed to introduce the photocleavable group at
the 2′-position via an acetal linkage, which can be introduced
selectively into the desired position and will not migrate to
the 3′-position during further chemical conversion, and which
works as a protecting group. The synthesis of the phos-
phoramidite unit possessing a biotinylated photocleavable
group is shown in Schemes 1 and 2. Starting with 5-methyl-
Scheme 1
In view of the apparent limitations of existing methods,
we report herein a new approach toward a convenient RNA
synthesis. Our approach consists of the synthesis of a
phosphoramidite possessing a biotinylated photocleavable
group at the 2′-position, its incorporation into the 5′-end of
the RNA strand under the usual reaction conditions in solid-
phase RNA synthesis, and selective isolation by means of
the biotin-streptavidin interaction, followed by photocleav-
age to afford the desired RNA strand pure enough for RNAi
experiments.
In order to simplify our proposed approach, the type and
introduction of a biotinylated photocleavable group in the
phosphoramidite unit initially had to be considered. We first
planned to introduce the photocleavable group at the nucleo-
base moiety, according to the report by Ho¨bartner and
Silverman.14 The uridine derivative 1 was prepared, as shown
in Figure 1. Prior to conversion into the corresponding
2-nitrobenzoic acid (2), we prepared compound 3 in four
steps.15 The bromo group of 3 was then converted to an azido
group to afford 4. Methylthiomethylation of 4 was achieved
by treatment with acetic anhydride in a mixture of dimethyl
sulfoxide and acetic acid to afford 5,16 which was introduced
into the 2′-position of the uridine derivative 6 (Scheme 2).
Thus, a mixture of 5, 6, and N-iodosuccinimide (NIS) in THF
was treated with trifluoromethanesulfonic acid (TfOH) at
-40 °C, and the 1,1,3,3-tetraisopropyldisiloxane-1,3-diyl
(TIPDS) group of the crude product was removed by
ammonium fluoride to give 7 in 81% yield in two steps. The
5′-hydroxyl group was then protected by the dimethoxytrityl
(DMTr) group, and the azido group of the resulting 8 was
reduced by Ph3P in the presence of H2O under reflux to give
9. In order to introduce a biotin unit, compound 9 was treated
with 1017 in the presence of O-(benzotriazol-1-yl)-N,N,N′,N′-
tetramethyluronium hexafluorophosphate (HBTU), 1-hy-
(5) Pitsch, S.; Weiss, P. A.; Jenny, L.; Stutz, A.; Wu, X. HelV. Chem.
Acta 2001, 84, 3773–3795.
(6) Scaringe, S. A.; Wincott, F. E.; Caruthers, M. H. J. Am. Chem. Soc.
1998, 120, 11820–11821.
(7) Semenyuk, A.; Foldesi, A.; Johansson, T.; Estmer-Nilsson, C.;
Blomgren, P.; Bra¨nnvall, M.; Kirsebom, L. A.; Kwiatkowski, M. J. Am.
Chem. Soc. 2006, 128, 12356–12357.
(8) Lackey, J. G.; Sabatino, D.; Damha, M. Org. Lett. 2007, 9, 789–
792.
(9) Gieœlak, J.; Grajkowski, A.; Kauffman, J. S.; Duff, R. J.; Beaucage,
S. L. J. Org. Chem. 2008, 73, 2774–2783.
(10) Ohgi, T.; Masutomi, Y.; Ishiyama, K.; Kitagawa, H.; Shiba, Y.;
Yano, J. Org. Lett. 2005, 7, 3477–3480.
Figure 1. 2′-O-Methyluridine possessing a photocleavable group
(11) Shiba, Y.; Masuda, H.; Watanabe, N.; Ego, T.; Takagaki, K.;
Ishiyama, K.; Ohgi, T.; Yano, J. Nucleic Acids Res. 2007, 35, 3287–3296.
(12) Olejnik, J.; Krzymanska-Olejnik, E.; Rothschild, K. J. Nucleic Acids
Res. 1996, 24, 361–366.
in its nucleobase moiety.
(13) Fang, S.; Bergstrom, D. E. Tetrahedron Lett. 2004, 45, 7987–7990.
(14) Ho¨bartner, C.; Silverman, S. S. Angew. Chem., Int. Ed. 2005, 44,
7305–7309.
phosphoramidite unit, the stability of 1 against conditions
used in solid-phase RNA synthesis had to be addressed. As
a result, 1 was converted into 2′-O-methylcytidine when it
was treated with concentrated NH4OH/EtOH (3:1) at room
(15) Saran, D.; Burke, D. H. Bioconjugate Chem. 2007, 18, 275–279.
(16) Pojer, P. M.; Angyal, S. J. Tetrahedron Lett. 1976, 17, 3067–3068.
(17) Zhong, M.; Strobel, S. A. Org. Lett. 2006, 8, 55–58.
Org. Lett., Vol. 12, No. 17, 2010
3837

Scheme 2
droxybenzotriazole (HOBt), and Et3N to give 11. Finally,
between 11 and 13 min. The solution containing the full-
length and failure sequences and other impurities (50% of
the 0.2 µmol synthesis) was then incubated with magnetic
beads having streptavidin (Dynabeads MyOne Streptavidin
C1) in a buffer for 30 min at room temperature (step 4).
Since the full-length RNA strand has a biotin functional
group, only the desired RNA will be immobilized to the
beads through the high-affinity biotin-streptavidin interac-
compound 11 was converted into the corresponding phos-
phoramidite 12. Since this conversion was slow and the
starting material was not consumed, 4-dimethylaminopyridine
(DMAP) was added as a catalyst to promote the reaction.
With the desired phosphoramidite 12 in hand, we next
examined the solid-phase RNA synthesis. The sequence
prepared in this study was a 21mer RNA with two consecu-
tive 2′-deoxythymidine (t) units at its 3′-end (RNA1), which
can directly be used for gene silencing experiments by RNAi
if it does not require further purification. The schematic
representation of our approach is shown in Figure 2. First, a
20mer sequence was prepared on a DNA/RNA synthesizer
using combination of 2′-deoxythymidine and 2′-O-TBDMS
phosphoramidite units following the standard procedure (step
1). After deprotection of the DMTr group at the 5′-end, 12
(represented as U possessing an asterisk in Figure 2) was
coupled with the resulting 5′-hydroxyl group to give the
21mer sequence on the CPG support. In the usual synthesis,
oligoDNA or RNA is synthesized as DMTr-On at its 5′-end
in order to ease its purification, whereas in our approach,
the DMTr group at the 5′-end was removed (step 2). After
completion of the solid-phase synthesis, the CPG support
was treated with concentrated NH4OH/EtOH (3:1) for 16 h
at room temperature to remove the protecting groups and
detach the RNA strand from the CPG support. The CPG
support was filtered off, and the filtrate was successively
treated with Et3N·3HF18 for 14 h at room temperature to
remove the 2′-O-TBDMS groups (step 3). In this stage, an
aliquot of the reaction mixture was analyzed by HPLC. As
can be seen in Figure 3a, a full-length sequence was observed
at around 17 min, which was observed as two peaks arising
from the existence of diastereomers at the photocleavable
moiety, and undesired shorter sequences were observed
Figure 2. Schematic representation of our approach.
3838
Org. Lett., Vol. 12, No. 17, 2010

In order to evaluate the quality of the oligoRNA prepared
in our approach, the RNAi activity of RNA1 was compared
with the 21mer RNA having the same sequence using a
conventional method and HPLC purification. The sequence
of RNA1 is complementary to firefly luciferase mRNA
(pGL3). Thus, RNA1 and its complementary strand (5′-
GUGCCAGAGUCCUUCGAUAtt-3′) prepared separately
were annealed to form siRNA (duplex), and its RNAi activity
was measured by transfection into HeLa cells constitutively
expressing the luciferase gene at 5, 25, and 50 nM concen-
trations. As a result, the siRNA consisting of RNA1 showed
almost equal RNAi activity compared with the siRNA
prepared by the usual method (the data are given in the
Supporting Information).
In conclusion, we have developed a rapid and efficient
RNA synthesis using a phosphoramidite unit possessing a
biotinylated photocleavable group. The biotinylated photo-
cleavable group was introduced at the 2′-position of uridine.
Since this functional group was bonded via an acetal linkage,
no migration to the 3′-hydroxyl group occurred and it was
stable toward conditions used in solid-phase RNA synthesis.
After incorporation of the phosphoramidite 12 at the 5′-
terminal of the RNA strand, followed by appropriate
procedures, the desired full-length RNA was selectively
“caught” by magnetic beads possessing streptavidin and then
“released” by photoirradiation to give the desired RNA. Since
the resulting RNA is pure enough to show potent RNAi
activity compared with a conventionally prepared and HPLC-
purified RNA, the “catch and release” approach presented
in this work would be applicable to high-throughput screen-
ing to determine the best sequence of siRNAs. Application
of this approach for the synthesis of longer RNA sequences,
such as pre-miRNA, is now in progress.
Figure 3. HPLC profiles of (a) reaction mixture after step 3 (see
Figure 2); (b) supernatant after step 4; (c) resulting RNA after step
5.
tion. Figure 3b shows an HPLC chart of the supernatant after
the procedure of step 4.19 As expected, the peaks corre-
sponding to the full-length completely disappeared, whereas
the undesired shorter RNAs were observed. As the last step
of our approach, a 0.2 M phosphate buffer of the magnetic
beads loaded with the full-length RNA was irradiated with
a high-pressure Hg lamp (400 W) using a Pyrex filter (>300
nm) for 30 min (step 5). After filtration of the magnetic
beads, the filtrate was analyzed by HPLC. Accordingly, a
single peak was observed at 13 min (Figure 3c), which was
identical with that of RNA1 prepared by the usual protocol
(i.e., purification by C-18 silica gel, followed by HPLC).
Additionally, the structure of RNA1 prepared by this method
was confirmed by matrix-assisted laser desorption/ionization
time-of-flight mass spectrometry (MALDI-TOF/MS).20
Acknowledgment. This work was supported in part by a
Grant-in-Aid for Scientific Research from the Japan Society
for Promotion of Science. We would like to thank Ms. M.
Kiuchi (Center for Instrumental Analysis, Hokkaido Uni-
versity) for elemental analysis. We also would like to thank
Ms. S. Oka (Center for Instrumental Analysis, Hokkaido
University) for measurement of mass spectra.
(18) Westman, E.; Stro¨mberg, R. Nucleic Acids Res. 1994, 22, 2430–
2431.
Supporting Information Available: Experimental pro-
cedure for new compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
(19) The peaks appearing at about 3 min came from the washing buffer
solution. This was confirmed by injection of this solution alone.
(20) RNA1 calculated mass C201H250N77O144P20 6664.9 (M - H),
observed mass 6666.0.
OL101489V
Org. Lett., Vol. 12, No. 17, 2010
3839