&α-L-ribo-Configured Locked Nucleic Acid (&α-L-LNA): Synthesis and Properties

Article abstract of DOI:10.1021/ja0168763

The syntheses of monomeric nucleosides and 3'-O-phosphoramidite building blocks en route to α-L-ribo-configured locked nucleic acids (α-L-LNA), composed entirely of α-L-LNA monomers (α-L-ribo configuration) or of a mixture of α-L-LNA and DNA monomers (β-D

Full text of DOI:10.1021/ja0168763

Published on Web 02/14/2002
r-L-ribo-Configured Locked Nucleic Acid (r-L-LNA):
Synthesis and Properties
Mads D. Sørensen,† Lisbet Kværnø,†,‡ Torsten Bryld,§ Anders E. Håkansson,†
Birgit Verbeure,| Gilles Gaubert,§ Piet Herdewijn,| and Jesper Wengel*,§
Contribution from the Department of Chemistry, UniVersity of Copenhagen,
UniVersitetsparken 5, DK-2100 Copenhagen, Denmark, Nucleic Acid Center,
Department of Chemistry, UniVersity of Southern Denmark, CampusVej 55,
DK-5230 Odense M, Denmark, and Laboratory of Medicinal Chemistry,
Rega Institute for Medical Research, Katholieke UniVersiteit LeuVen,
Minderbroedersstraat 10, B-3000 LeuVen, Belgium
Received August 17, 2001
Abstract: The syntheses of monomeric nucleosides and 3′-O-phosphoramidite building blocks en route to
R-L-ribo-configured locked nucleic acids (R-L-LNA), composed entirely of R-L-LNA monomers (R-L-ribo
configuration) or of a mixture of R-L-LNA and DNA monomers (â-D-ribo configuration), are described and
the R-L-LNA oligomers are studied. Bicyclic 5-methylcytosin-1-yl and adenin-9-yl nucleoside derivatives
have been prepared and the phosphoramidite approach has been used for the automated oligomerization
leading to R-L-LNA oligomers. Binding studies revealed very efficient recognition of single-stranded DNA
and RNA target oligonucleotide strands. Thus, stereoirregular R-L-LNA 11-mers containing a mixture of
R-L-LNA monomers and DNA monomers (“mix-mer R-L-LNA”) were shown to display ∆Tm values of +1 to
+3 °C per modification toward DNA and +4 to +5 °C toward RNA when compared with the corresponding
unmodified DNA·DNA and DNA·RNA reference duplexes. The corresponding ∆Tm values per modification
for the stereoregular fully modified R-L-LNA were determined to be +4 °C (against DNA) and +5 °C (against
RNA). 11-Mer R-L-LNAs (mix-mer R- L-LNA or fully modified R- L-LNA) were shown in vitro to be significantly
stabilized toward 3′-exonucleolytic degradation. A duplex formed between RNA and either mix-mer R-L-
LNA or fully modified R-L-LNA induced in vitro Escherichia coli RNase H-mediated cleavage, albeit very
slow, of the RNA targets at high enzyme concentrations.
R-LNA (R-D-ribo isomer)4e and their thermal stability toward
Introduction
complementary nucleic acid targets have been reported. These
The prospect of preparing therapeutically active analogues
of natural nucleic acids has stimulated much interest in the
synthesis of effective antisense oligonucleotides (AONs) during
the past decade. Some of the requirements for an AON include
efficient automated synthesis, good aqueous solubility, resistance
toward enzymatic degradation, and high binding affinity toward
nucleic acid targets while maintaining the fidelity of recogni-
tion.1 The unprecedented thermal stability of duplexes involving
LNA (locked nucleic acid, â-D-ribo isomer, Figure 1)2,3 has
inspired us to investigate the properties of the stereoisomers of
LNA. Recently, synthesis of the stereoisomeric analogues termed
xylo-LNA (â-D-xylo isomer),3,4a-c R-L-xylo-LNA (R-L-xylo
isomer),3,4c,5 R-L-LNA (R-L-ribo isomer, Figure 1),3,4a-d,5,6 and
(2) (a) Singh, S. K.; Nielsen, P.; Koshkin, A. A.; Wengel, J. Chem. Commun.
1998, 455. (b) Koshkin, A. A.; Singh, S. K.; Nielsen, P.; Rajwanshi, V.
K.; Kumar, R.; Meldgaard, M.; Olsen, C. E.; Wengel, J. Tetrahedron 1998,
54, 3607. (c) Obika, S.; Nanbu, D.; Hari, Y.; Andoh, J.; Morio, K.; Doi,
T.; Imanishi, T. Tetrahedron Lett. 1998, 39, 5401. (d) Koshkin, A. A.;
Nielsen, P.; Meldgaard, M.; Rajwanshi, V. K.; Singh, S. K.; Wengel, J. J.
Am. Chem. Soc. 1998, 120, 13252. (e) Singh, S. K.; Wengel, J. Chem.
Commun. 1998, 1247.
(3) We have defined LNA as an oligonucleotide containing one or more 2′-
O,4′-C-methylene-â-D-ribofuranosyl nucleotide monomer(s). The natural
â-D-ribo configuration is assigned to LNA (and LNA monomers) as the
positioning of the N1, O2′, O3′, and C5′ atoms are equivalent to those
found in RNA. Analogously, xylo-LNA, R-L-xylo-LNA, and R-L-LNA have
been defined as oligonucleotides containing one or more 2′-O,4′-C-
methylene-â-D-xylofuranosyl nucleotide monomer(s), 2′-O,4′-C-methylene-
R-L-xylofuranosyl nucleotide monomer(s), and 2′-O,4′-C-methylene-R-L-
ribofuranosyl nucleotide monomer(s), respectively. The term “mix-mer R-L-
LNA” is used herein for an R-L-LNA containing a mixture of R-L-LNA
monomers and DNA monomers, and the term “mix-mer LNA” is used for
an LNA containing a mixture of LNA monomers and DNA monomers.
(4) (a) Rajwanshi, V. K.; Håkansson, A. E.; Dahl, B. M.; Wengel, J. Chem.
Commun. 1999, 1395. (b) Rajwanshi, V. K.; Håkansson, A. E.; Kumar,
R.; Wengel, J. Chem. Commun. 1999, 2073. (c) Rajwanshi, V. K.;
Håkansson, A. E.; Sørensen, M. D.; Pitsch, S.; Singh, S. K.; Kumar, R.;
Nielsen, P.; Wengel, J. Angew. Chem., Int. Ed. 2000, 39, 1656. (d)
Håkansson, A. E.; Wengel, J. Bioorg. Med. Chem. Lett. 2001, 11, 935. (e)
Nielsen, P.; Dalskov, J. K. Chem. Commun. 2000, 1179.
† University of Copenhagen.
‡ Present address: Department of Chemistry, The Technical University
of Denmark, Building 201, DK-2800 Lyngby, Denmark.
§ University of Southern Denmark.
| Katholieke Universiteit Leuven.
Funded by the Danish National Research Foundation for Studies on
Nucleic Acid Chemical Biology.
(1) (a) De Mesmaeker, A.; Ha¨ner, R.; Martin, P.; Moser, H. E. Acc. Chem.
Res. 1995, 28, 366. (b) Herdewijn, P. Liebigs Ann. 1996, 1337. (c) Freier,
S. M.; Altmann, K.-H. Nucleic Acids Res. 1997, 25, 4429. (d) Wengel, J.
Acc. Chem. Res. 1999, 32, 301. (e) Herdewijn, P. Biochim. Biophys. Acta
1999, 1489, 167. (f) Uhlmann, E. Curr. Opin. Drug DiscoVery DeV. 2000,
3, 203.
(5) Håkansson, A. E.; Koshkin, A. A.; Sørensen, M. D.; Wengel, J. J. Org.
Chem. 2000, 65, 5161.
(6) Wengel, J.; Petersen, M.; Nielsen, K. E.; Jensen, G. A.; Håkansson, A. E.;
Kumar, R.; Sørensen, M. D.; Rajwanshi, V. K.; Bryld, T.; Jacobsen, J. P.
Nucleosides Nucleotides Nucleic Acids 2001, 20, 389.
9
2164 VOL. 124, NO. 10, 2002 J. AM. CHEM. SOC.
10.1021/ja0168763 CCC: $22.00 © 2002 American Chemical Society

R-L-ribo-Configured Locked Nucleic Acid
A R T I C L E S
Scheme 1. Synthesis of the R-L-LNA Thymine4a,5 and
5-Methylcytosine Phosphoramidite Derivatives 6 and 9 Suitable for
Incorporation of the R-L-LNA Thymine (rLTL) and 5-Methylcytosine
(
rLMeCL) Monomers into Oligonucleotidesa
Figure 1. Structures of the nucleotide monomers of LNA and R-L-LNA
(left) and sketches of their locked N-type7 (C3′-endo/3E) furanose conforma-
tions (right).
preceding studies revealed very efficient recognition of single-
stranded DNA and RNA by both LNA and R-L-LNA. Due to
the fact that the furanose rings of both LNA and R-L-LNA
monomers are part of a dioxabicyclo[2.2.1]heptane skeleton,
they are both efficiently locked in an N-type (C3′-endo/3E)
conformation (Figure 1).7 So far, the properties of R-L-LNA
containing the thymine and the adenine R-L-LNA monomers
have been studied, i.e., homothymine oligomers4a-c and 9-mer
R-L-LNAs containing a mixture of R-L-LNA monomers (three
thymine4c or three adenine R-L-LNA monomers4d) and DNA
monomers (“mix-mer R-L-LNAs”).3,4c,4d On the basis of these
preliminary studies, the surprising indication is that the helical
thermostabilities of R-L-LNA·DNA and R-L-LNA·RNA du-
plexes approach those of the corresponding LNA·DNA and
LNA·RNA duplexes.2,4a-4d
One mode of action of AONs involves recruitment of the
enzyme RNase H for the degradation of the targeted mRNA,
but the exact recognition elements of this substrate-specific
endogenous enzyme are not known at present time.8 When
hybridized to complementary RNA, LNA has been recognized
as an RNA mimic that drives the duplex to adopt an A-type
conformation.2d,9 This feature is disadvantageous in terms of
RNase H recognition because this enzyme recognizes DNA·
RNA duplexes intermediary between the DNA B-type and the
RNA A-type conformations.10 NMR spectroscopic studies of
R-L-LNA·RNA duplexes and molecular dynamics (MD) simula-
tions of fully modified R-L-LNA·RNA duplexes have shown
the overall duplex geometry to be very similar to the corre-
sponding unmodified DNA·RNA hybrid,6,11 pointing to R-L-
LNA being most adequately described as a DNA mimic.
To allow a comprehensive evaluaton of the properties of R-L-
LNA and its potential as an AON, we report herein (a) the
a Reagents, conditions and yields: (i-v) See refs 4a and 5; (i) MsCl,
pyridine, rt (92%); (ii) 6 M NaOH, EtOH, H2O, reflux (58%); (iii) H2/
(20% Pd(OH)2/C)/EtOH, rt (98%); (iv) DMTCl, pyridine, rt (98%); (v)
NC(CH2)2OP(Cl)N(i-Pr)2, EtN(i-Pr)2, CH2Cl2, rt (51%; optimized relative
to ref 4a); (vi) DNA synthesizer; (vii) (a) TMSCl, CH2Cl2, pyridine, rt; (b)
Tf2O, rt; (c) saturated aqueous NH3 in MeOH, rt (80%); (viii) (a) TMSCl,
pyridine, rt; (b) N-benzoyl-1H-tetrazole, rt (80%); (ix) NC(CH2)2OP(Cl)N(i-
Pr)2, EtN(i-Pr)2, CH2Cl2, rt (45%). DMTCl ) 4,4′-dimethoxytrityl chloride.
syntheses of the bicyclic 5-methylcytosin-1-yl and adenin-9-yl
R-L-LNA nucleosides and 3′-O-phosphoramidite building blocks
thereof, (b) binding studies of 11-mer fully modified and 11-
mer mix-mer R-L-LNA toward complementary DNA and RNA
targets, (c) the behavior of R-L-LNA toward 3′-exonucleolytic
degradation in vitro, and (d) the results of Escherichia coli
RNase H-mediated cleavage of the RNA strand of R-L-LNA·
RNA duplexes in vitro.
Results
Synthesis of r-L-LNA Pyrimidine Nucleosides. Conversion
of the known 4′-C-hydroxymethyl nucleoside 1 obtained from
1,2:5,6-di-O-isopropylidine-R-D-glucofuranose as earlier de-
scribed,12 into R-L-LNA pyrimidine phosphoramidite building
blocks was performed as depicted in Scheme 1. Transformation
of nucleoside 1 into the bicyclic thymine R-L-LNA nucleoside
4 and the corresponding phosphoramidite monomer 6 has been
described earlier.4a,5 The key step was a cascade reaction of
tri-O-mesyl nucleoside 2 involving C2′ epimerization, hydrolysis
of the resulting anhydronucleoside intermediate, cyclization, and
removal of the mesyl group at C5′ to give the bicyclic nucleoside
3 in 58% yield.5 The yield of this cascade reaction was later
improved to 89% by use of the corresponding tri-O-tosyl
derivative of nucleoside 2 as reactant instead (unpublished
results).
(7) Despite the apparent differences in the furanose conformations between
an LNA monomer and an R-L-LNA monomer (Figure 1), also the furanose
conformation of an R-L-LNA monomer is of the N-type (C3′-endo, 3E)
because of its L-configuration. For further information about the conforma-
tions of nucleotides, see Eur. J. Biochem. 1983, 131, 9 (Abbreviations and
Symbols for the Description of Conformations of Polynucleotide Chains,
IUPAC-IUB Joint Commission on Biochemical Nomenclature).
(8) (a) Crooke, S. T. Biochim. Biophys. Acta 1999, 1489, 31. (b) Zamaratski,
E.; Pradeepkumar, P. I.; Chattopadhyaya, J. J. Biochem. Biophys. Methods
2001, 48, 189.
(9) (a) Petersen, M.; Nielsen, C. B.; Nielsen, K. E.; Jensen, G. A.; Bondens-
gaard, K.; Singh, S. K.; Rajwanshi, V. K.; Koshkin, A. A.; Dahl, B. M.;
Wengel, J.; Jacobsen, J. P. J. Mol. Recognit. 2000, 13, 44. (b) Bondensgaard,
K.; Petersen, M.; Singh, S. K.; Rajwanshi, V. K.; Kumar, R.; Wengel, J.;
Jacobsen, J. P. Chem. Eur. J. 2000, 6, 2687. (c) Nielsen, K. E.; Singh, S.
K.; Wengel, J.; Jacobsen, J. P. Bioconjugate Chem. 2000, 11, 228. (d)
Nielsen, C. B.; Singh, S. K.; Wengel, J.; Jacobsen, J. P. J. Biomol. Struct.
Dyn. 1999, 17, 175.
Synthesis of the 5-methylcytosine R-L-LNA amidite 9 was
achieved starting from the thymine nucleoside derivative 55
(Scheme 1). Often, the frequently used triazolation strategy13
for conversion of thymine into 5-methylcytosine nucleosides
(10) Fedoroff, O. Y.; Salazar, M.; Reid, B. R. J. Mol. Biol. 1993, 233, 509.
(11) Petersen, M.; Håkansson, A. E.; Wengel, J.; Jacobsen, J. P. J. Am. Chem.
Soc. 2001, 123, 7431.
(12) Rajwanshi, V. K.; Kumar, R.; Hansen, M. K.; Wengel, J. J. Chem. Soc.,
Perkin Trans. 1 1999, 1407.
(13) Lin, T.-S.; Gao, Y.-S.; Mancini, W. R. J. Med. Chem. 1983, 53, 1294.
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2165

A R T I C L E S
Sørensen et al.
of pyrimidine R-L-LNA nucleosides cannot be applied in the
synthesis of the corresponding purine nucleosides because of
the inability of the latter to form C2′-anhydronucleoside
intermediates. Alternatively, the known tetra-O-acyl-L-threo-
pentofuranose 10 (Scheme 2)12,20 was used as a starting material
in the synthesis of the adenine phosphoramidite building block
23.4d Coupling between furanose 10 and 6-N-benzoyladenine
with tin(IV) chloride as the Lewis acid afforded the desired N9-
regioisomeric nucleoside 11 in 52% yield as the only regio-
and stereoisomer isolated, which can be explained by the
presence of the participating 2-O-acetyl group of 10 and by the
fact that the coupling reaction was performed under thermo-
dynamic control. A similar Vorbru¨ggen-type21 coupling reaction
with trimethylsilyltriflate as Lewis acid and N,O-bis(trimeth-
ylsilyl)acetamide as silylating agent provided nucleoside 11 in
only 41% yield (data not shown). By comparison of the 13C
NMR chemical shift values obtained for derivatives 20 and 21
(vide infra) with the published values for N7- and N9-
regioisomeric â-D-ribofuranosyladenines,22 it was possible to
assign the synthesized nucleoside derivatives 11-23 as the N9-
regioisomers. Epimerization at C2′ via intermediates 12 and 13
was performed by a reaction sequence in which the 2′-O-acetyl
group of nucleoside 11 was first deacylated chemoselectively
with half-saturated methanolic ammonia23 to give the intermedi-
ate 12. Subsequently, the reaction of 12 with trifluoromethane-
sulfonic anhydride in a mixture of pyridine and dichloromethane
afforded intermediate 13, which was directly treated with
potassium acetate and 18-crown-6 to afford the desired C2′-
epimeric nucleoside 14 in 84% yield (from 11). Complete
deacylation of nucleoside 14 was accomplished with a mixture
of saturated methanolic ammonia and saturated aqueous am-
monia to give an intermediate, which directly was subjected to
a transient protection protocol16 (O-trimethylsilylation with
chlorotrimethylsilane in pyridine, 6-N-benzoylation with benzoyl
chloride, and desilylation) giving the 6-N-benzoyl derivative 15
in 91% yield from 14. Nucleoside 15 was mesylated selectively
by reaction with 2.2 equiv of mesyl chloride in pyridine at 0
°C, furnishing derivative 16 in 79% yield without any observed
mesylation of the secondary hydroxy group. Cyclization by
intramolecular nucleophilic attack of the 2′-hydroxy group with
concomitant demesylation and debenzoylation was performed
in a mixture of aqueous sodium hydroxide and 1,4-dioxane
(reflux, 72 h) in only 13% yield. Instead, nucleoside 16 was
treated with sodium hydride in THF to give the bicyclic
nucleoside 17, followed by substitution of the remaining
mesyloxy group with acetate by reaction with potassium acetate
and 18-crown-6 in dioxane at reflux to afford nucleoside 18 in
87% yield (from 16). Chemoselective O-deacetylation of 18 in
half-saturated methanolic ammonia afforded nucleoside 19 in
83% yield, and debenzylation of this intermediate with am-
monium formate and Pd/C gave the desired 6-N-benzoyl R-L-
LNA adenine nucleoside 20 in 44% yield in addition to the
fully deprotected nucleoside 21 in 24% yield. The R-L-ribo
configuration of nucleoside 20 was confirmed by 1H NOE
experiments [mutual NOE effects between H2/H8 of the adenine
Figure 2. Nucleobase tautomerization of 4-N-benzoyl-5-methyl-2′-deoxy-
cytidine18 and derivative 8.
gives only moderate overall yields. An alternative method
described by Reese and co-workers14 involving trimethylsily-
lation of the hydroxy groups of nucleoside 4 followed by
activation of the nucleobase with trifluoroacetic anhydride and
subsequent treatment with 4-nitrophenol failed in our hands.
Instead, the conversion was achieved by a slightly different
procedure described by Herdewijn and Van Aerschot.15 In a
transient protection strategy,16 the 3′-hydroxy group of nucleo-
side 5 was initially trimethylsilylated in a mixture of pyridine
and dichloromethane. The nucleobase was subsequently acti-
vated by dropwise addition of 3 equiv of trifluoromethane-
sulfonic anhydride, furnishing the 5-methylcytosin-1-yl nucle-
oside 7 in 81% yield after treatment with saturated methanolic
ammonia. Nucleoside 7 was transiently protected by trimeth-
ylsilylation followed by benzoylation of the nucleobase with
N-benzoyl-1H-tetrazole17 and subsequent desilylation, furnishing
the protected nucleoside 8 in 80% yield. Eventually, standard
phosphitylation with 2-cyanoethyl (N,N-diisopropyl)phosphora-
midochloridite and (N,N-diisopropyl)ethylamine in dichlo-
romethane afforded the desired 4-N-benzoyl-5-methylcytosine
phosphoramidite building block 9 in 45% yield. Because of the
rather efficient synthesis of the thymine nucleoside 5, we chose
the strategy described above for the synthesis of 5-methylcy-
tosine derivative 8 instead of the alternative route involving
coupling between an appropriately protected furanose, e.g.,
derivative 10 or 26 (vide infra), and silylated 5-methylcytosine
(or cytosine). Generally, the 13C NMR chemical shift values of
the nucleobase change only slightly upon 4-N-acylation of
cytosine nucleosides. However, this is not always the case upon
4-N-acylation of 5-methylcytosine nucleosides.18 Thus, the
observed significant downfield shift of the signal for C5 and
upfield shifts of the signals for C2, C4, and C6 of nucleoside 8
compared with the values for the nucleoside 7 with an
unprotected 5-methylcytosine moiety can be explained by a
tautomeric change of the nucleobase (Figure 2). Previously, the
nucleobase of 4-N-benzoyl-5-methyl-2′-deoxycytidine has been
shown to adopt an imino tautomeric form, and the similarity
between the 13C NMR shift values reported for this nucleo-
side18,19 and the corresponding values for the nucleobase of
nucleoside 8 also indicate the imino tautomeric form to be
predominant in this case.
Synthesis of r-L-LNA Purine Nucleosides. In general, a
reaction sequence similar to the one used above for the synthesis
(14) Miah, A.; Reese, C. B.; Song, Q. Nucleosides Nucleotides 1997, 16, 53.
(15) Herdewijn, P.; Van Aerschot, A. Nucleosides Nucleotides 1989, 8, 933.
(16) Ti, G. S.; Gaffney, B. L.; Jones, R. A. J. Am. Chem. Soc. 1982, 104, 1316.
(17) Indications of dibenzoylation of the nucleobase when benzoyl chloride was
used as the reagent led to the use of N-benzoyl-1H-tetrazole instead. Bhat,
B.; Sanghvi, Y. S. Tetrahedron Lett. 1997, 38, 8811.
(20) The synthesis of phosphoramidite 23 from furanose 10 (Scheme 2) has
been described in a preliminary report; see ref 4d.
(18) Busson, R.; Kerremans, L.; Van Aerschot, A.; Peeters, M.; Blaton, N.;
Herdewijn, P. Nucleosides Nucleotides 1999, 18, 1079.
(21) (a) Vorbru¨ggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber. 1981, 114,
1234. (b) Vorbru¨ggen, H.; Ho¨fle, G. Chem. Ber. 1981, 114, 1256.
(22) Chenon, M.; Pugmire, R. J.; Grant, D. M.; Panzica, R. P.; Townsend, L.
B. J. Am. Chem. Soc. 1975, 97, 4627.
(19) 13C NMR shift values for the heterocyclic base carbons of 4-N-benzoyl-
5-methyl-2′-deoxycytidine (see ref 18): [(CD3)2SO] δ 176.9 (CO), 160.2
(C-4), 148.7 (C-2), 139.6 (C-6), 109.9 (C-5), 13.3 (CH3). No significant
change was observed when CDCl3 was used as solvent.
(23) Neilson, T.; Werstiuk, E. S. Can J. Chem. 1971, 49, 493.
9
2166 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

R-L-ribo-Configured Locked Nucleic Acid
A R T I C L E S
Scheme 2. Synthesis of the R-L-LNA Adenine Phosphoramidite Monomer 23 Suitable for Incorporation of the R-L-LNA Adenine Monomer
(
rLAL) into Oligonucleotidesa
a Reagents, conditions and yields: (i) 6-N-benzoyladenine, SnCl4, CH3CN, rt (52%); (ii) half-saturated NH3 in MeOH, 0 °C; (iii) Tf2O, pyridine, CH2Cl2,
-30 to 0 °C; (iv) KOAc, 18-crown-6, toluene, 80 °C (84%, 3 steps); (v) saturated NH3 in MeOH/32% aq NH3 (4/1), rt; (vi) (a) TMSCl, pyridine, rt; (b)
BzCl, pyridine, rt; (c) 32% aq NH3/H2O/MeOH (6/3/4), rt (91%, 2 steps); (vii) MsCl, pyridine, 0 °C (79%); (viii) NaH, THF, rt; (ix) KOAc, 18-crown-6,
1,4-dioxane, 100 °C (87%, 2 steps); (x) half-saturated NH3 in MeOH, 0 °C (83%); (xi) HCOONH4, 10% Pd/C, MeOH, reflux (20, 44%, 21, 24%); (xii)
DMTCl, pyridine, rt (82%); (xiii) NC(CH2)2OP(Cl)N(i-Pr)2, EtN(i-Pr)2, CH2Cl2, rt (63%); (xiv) DNA synthesizer.
Scheme 3. Alternative Synthesis of the Bicyclic Nucleoside 16a
moiety and the protons of the endocyclic methylene group (2%
and 8%/2% when H2/H8 and the endocyclic methylene group,
respectively, were irradiated) and between H1′ and H2′/H3′ (5%/
3% and 11% when H1 and H2′/H3′, respectively, were irradi-
ated)]. To prepare for automated synthesis of R-L-LNA
containing the adenine R-L-LNA monomer rLAL, the primary
hydroxy group of nucleoside diol 20 was first DMT-protected
selectively by reaction with DMTCl in pyridine to give
derivative 22 in 82% yield, and second, phosphitylated under
standard conditions to furnish the desired phosphoramidite
building block 23 in 63% yield (Scheme 2).
In an alternative strategy, the synthesis of intermediate 16
was accomplished with furanose 2424 as starting material
(Scheme 3). Reaction with excess methanesulfonyl chloride in
anhydrous pyridine gave in 79% yield furanose 25, which was
subsequently converted into the 1,2-di-O-acetylated furanose
26 in 63% yield by acetal hydrolysis with 80% trifluoroacetic
acid, followed by acetylation with acetic anhydride in pyridine.
Coupling between furanose 26 and 6-N-benzoyladenine with
tin(IV)chloride as Lewis acid afforded stereoselectively the
desired N9-regioisomeric nucleoside 27 in 57% yield. By
analogy with the synthesis of nucleoside 11, no other nucleoside
products were isolated. Treatment of nucleoside 27 with half-
saturated methanolic ammonia gave chemoselectively the O-
deacylated intermediate 28 in 79% yield. To prepare for
epimerization at C2′, the intermediate 28 was treated with
a Reagents, conditions and yields: (i) MsCl, pyridine, rt (79%); (ii) (a)
80% TFA, rt; (b) Ac2O, pyridine, rt (63%); (iii) 6-N-benzoyladenine, SnCl4,
CH3CN, rt (57%); (iv) half-saturated NH3 in MeOH, rt (79%); (v) (a) Tf2O,
pyridine, CH2Cl2, -30 °C to 0 °C; (b) KOAc, 18-crown-6, toluene, CH2Cl2,
50 °C (70%, 2 steps); (vi) half-saturated NH3 in MeOH, rt (85%).
trifluoromethanesulfonic anhydride in a mixture of pyridine and
dichloromethane and subsequently with potassium acetate and
18-crown-6 in a mixture of toluene and dicloromethane to give
the desired C2′-epimeric nucleoside 29 in 70% combined yield.
In contrast to the unsuccessful O-deacylation of the nucleoside
analogue 14, selective O-deacylation of 29 was achieved with
saturated methanolic ammonia at room temperature affording
the 6-N-benzoyl derivative 16 in 85% yield.25 The two different
routes for synthesis of the adenine monomer differ with respect
to both number of steps and overall yields. Thus, the route from
(24) Youssefyeh, R. D.; Verheyden, J. P. H.; Moffat, J. G. J. Org. Chem. 1979,
44, 1301.
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2167

A R T I C L E S
Sørensen et al.
Table 1. Melting Temperatures (Tm Values) toward Matched and
Singly Mismatched Complementary DNAa
Tm values/°C
T1
MM1
MM2
MM3
MM4
MM5
MM6
DNA-1
DNA-2
r-L-LNA-1
r-L-LNA-2
LNA-1
36
42
39
65
54
69
17
21
20
40
30
45
10
16
11
36
27
38
16
22
15
40
28
42
21
26
18
46
37
51
27
32
33
57
45
61
25
28
22
48
39
50
LNA-2
a Tm values for matched sequences are shown in boldface type. Tm values
were measured as the maximum of the first derivative of the melting curve
(A260 vs temperature) recorded in medium salt hybridization buffer (10 mM
sodium phosphate, 100 mM sodium chloride, and 0.1 mM EDTA, pH 7.0)
with 1.5 µM concentrations of the two strands under the assumption of
identical extinction coefficients of modified and unmodified nucleotides.
Figure 3. Various oligonucleotides (R-L-LNAs, LNAs, and unmodified
controls and targets) used in this study. R-L-LNA monomers are denoted
RLTL, rLMeCL, and rLAL; LNA monomers are denoted TL, MeCL, and AL;
the natural DNA/RNA monomers and 2′-OMe monomers are denoted A,
C, MeC, G, T and U; d(sequence) denotes a DNA strand, r(sequence) an
RNA strand, 2′-OMe(sequence) a 2′-O-methyl-oligoribonucleotide strand,
rL(sequence)L a fully modified R-L-LNA strand, and (sequence)L a fully
modified LNA strand. Mismatched bases in target sequences MM1-MM8
are underlined.
Figure 4. Tm curves for (4) DNA-2·T1, (]) r-L-LNA-2·T1, and (0) LNA-
2·T1 duplexes. See the Experimental Section and Table 1 for conditions.
shown in Table 1. For comparison, the properties of the
corresponding DNAs (DNA-1 and DNA-2, containing cytosine
and 5-methylcytosine monomers, respectively) and LNAs
(LNA-1 and LNA-2) are also shown.
the starting material 24 via furanose 10 to nucleoside 16
(Scheme 2) proceeded in overall 22% yield (nine steps) whereas
the alternative route (Scheme 3) from 24 via 26 to 16 proceeded
in an overall yield of only 14% (five steps).
Compared with the reference DNA-1, the mix-mer r-L-
LNA-1 increases the Tm value by 3 °C toward the fully matched
complementary DNA target T1. This increase is surprisingly
small compared with the results obtained for 9-mer R-L-LNAs
containing three R-L-LNA monomers (either three thymine or
three adenine R-L-LNA monomers), for which an increase in
the Tm value of 8 °C toward the DNA targets was observed.4c,4d
In contrast, the LNA mix-mer LNA-1 induces an increase in
the Tm value of 18 °C compared with DNA-1. However, the
Tm values for the two fully modified 11-mers r-L-LNA-2 and
LNA-2 when paired with T1 are comparable and both signifi-
cantly increased (by 23 and 27 °C, respectively, compared with
the value for the DNA reference DNA-2). The decreases in
thermal stabilities obtained for r-L-LNA-1 and r-L-LNA-2
toward singly mismatched DNA targets (MM1-MM6) are
comparable with those obtained for DNA-1, DNA-2, LNA-1,
and LNA-2 (Table 1), although the pairing selectivities of R-L-
LNA and LNA in general appear slightly improved compared
with that of DNA, thereby confirming the results obtained earlier
in other sequence contexts.2b,4c,4d,26 The normalized melting
curves for the duplexes DNA-2·T1, r-L-LNA-2·T1, and LNA-
2·T1 are depicted in Figure 4.
We decided to evaluate R-L-LNA containing thymine,
5-methylcytosine, and adenine monomers (rLTL , rLMeCL, and
rLAL, respectively). Having access to two pyrimidine R-L-LNA
monomers and one purine R-L-LNA monomer, we envisioned
this would allow us to perform the necessary experiments.
Preliminary synthetic procedures for preparation of guanine R-L-
LNA nucleosides are included in the Supporting Information.
Thermal Denaturation Studies. To test the properties of
mix-mer R-L-LNA and fully modified R-L-LNA, we synthe-
sized r-L-LNA-1, composed of alternating R-L-LNA (three
adenine and two thymidine monomers) and DNA monomers,
and the fully modified r-L-LNA-2 (see Figure 3 for the
composition of oligonucleotides). The phosphoramidite approach
on an automated DNA synthesizer was used for synthesis of
the oligomers (see the Experimental Section for details).
The data illustrating the hybridization properties of r-L-
LNA-1 and r-L-LNA-2 toward matched and singly mismatched
DNA complements (T1 and MM1-MM6, respectively) are
(25) We have, in a preliminary way, attempted yet another alternative way of
introducing the bicyclic skeleton of an R-L-LNA adenine nucleoside (data
not reported). Thus, the 2′-O-mesyl-5′-hydroxy-4′-C-hydroxymethyl deriva-
tive of nucleoside 12 was refluxed for 94 h with 10 equiv of sodium
hydroxide in a 1/1 mixture of water and dioxane. Disappointingly, no ring
closure by intramolecular nucleophilic attack from the 5′-hydroxy group
on the 2′-position was observed. Attempted ring closure of the correspond-
ing 2′-O-trifluoromethanesulfonyl analogue with approximately 6 equiv of
sodium hydride in DMF (at room temperature up to 140 °C) for several
hours was likewise unsuccessful. In contrast, Robins et al. have reported
successful ring closure of the structurally similar nucleoside 9-(3-azido-
3-deoxy-2-O-mesyl-â-D-xylofuranosyl)adenine, affording the corresponding
2′,5′-anhydronucleoside derivative: Robins, M. J.; Hawrelak, S. D.; Kanai,
T.; Siefert, J.-M.; Mengel, R. J. Org. Chem. 1979, 44, 1317.
In Table 2, the hybridization properties of the mix-mer r-L-
LNA-1 and the fully modified r-L-LNA-2 toward the matched
RNA complement T2 and two singly mismatched RNA
complements (MM7 and MM8) are shown and compared with
those of the corresponding DNA (DNA-1 and DNA-2) and LNA
(LNA-1 and LNA-2) oligomers.
(26) Ørum, H.; Jacobsen, M. H.; Koch, T.; Vuust, J.; Borre, M. B. Clin. Chem.
1999, 45, 1898.
9
2168 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

R-L-ribo-Configured Locked Nucleic Acid
A R T I C L E S
Table 2. Melting Temperatures (Tm Values) toward Matched and
Singly Mismatched Complementary RNAa
Tm values/°C
T2
MM7
MM8
DNA-1
DNA-2
r-L-LNA-1
r-L-LNA-2
LNA-1
30
35
49
75
63
77
<10
<10
25
50
39
15
18
32
56
48
61
LNA-2
54
Figure 6. CD spectra of (0) DNA-1·T1, (]) DNA-1·T2, (4) r-L-LNA-
1·T1, and (×) r-L-LNA-1·T2 duplexes. See Table 1 for hybridization
conditions.
a Tm values for matched sequences are shown in boldface type. See
caption below Table 1 for conditions.
Figure 5. Tm curves for (4) DNA-2·T2, (]) r-L-LNA-2·T2, and (0) LNA-
2·T2 duplexes. See the Experimental Section and Table 1 for conditions.
Figure 7. CD spectra of (0) DNA-2·T1, (]) DNA-2·T2, (4) r-L-LNA-
2·T1, and (×) r-L-LNA-2·T2 duplexes. See Table 1 for hybridization
conditions.
The Tm values for r-L-LNA-1 and r-L-LNA-2 toward T2
are increased by 19 and 40 °C, respectively, compared with
the references (DNA-1 and DNA-2, respectively). The corre-
sponding increases for LNA-1 and LNA-2 are 33 and 42 °C,
respectively. The decreases in thermal stabilities obtained for
r-L-LNA-1 and r-L-LNA-2 toward the two singly mismatched
RNA strands MM7 and MM8 are comparable with those
obtained for the corresponding DNAs and LNAs, indicating that
the hybridization of R-L-LNA with an RNA complement occurs
with satisfactory selectivity. The normalized melting curves for
the duplexes DNA-2·T1, r-L-LNA-2·T1, and LNA-2·T1 are
depicted in Figure 5. Taken together, the results obtained clearly
demonstrate antiparallel Watson-Crick-type hybridization for
R-L-LNA.
The results from the hybridization experiments (Tables 1 and
2) not only confirm the data reported earlier but also reveal
important new information. The preferential binding of mix-
mer R-L-LNA toward RNA has been demonstrated earlier,4c,4d
although the results obtained herein [compare r-L-LNA-1·T1
(∆Tm +3 °C) with r-L-LNA-1·T2 (∆Tm +19 °C)] are even
more pronounced. It is clear from earlier reported data2b,4c,4d,26
as well as from the data reported herein that mix-mer LNA
displays superior hybridization properties toward especially
DNA complements. Importantly, it is shown herein that the
binding affinity of fully modified R-L-LNA is comparable to
that of fully modified LNA toward both complementary DNA
and RNA. A possible reason for the different hybridization
behavior of R-L-LNA and LNA as mix-mers can be extracted
from results obtained by NMR structural analyses and molecular
modeling.6,9,11 Thus, LNA monomers in mix-mer LNAs have
been shown strongly to induce the furanose conformations of
neighboring unmodified DNA monomers to shift toward an
N-type. This leads to preorganization of single-stranded mix-
mer LNA and overall A-type conformations of a mix-mer LNA
strand in duplexes with both DNA and RNA complements.9
This conformational tuning toward the duplex-stabilizing A-type
conformation1 explain the very high binding affinities of mix-
mer LNAs as also seen in this study (compare data for DNA-1,
DNA-2, LNA-1, and LNA-2). A similar effect is not seen in
mix-mer R-L-LNAs, as the furanose rings of DNA monomers
neighboring R-L-LNA monomers retain their S-type conforma-
tions.6,11 The reason for the relatively much more efficient
hybridization of mix-mer R-L-LNA with RNA (∆Tm values per
modification of R-L-LNA·RNA duplexes compared with DNA·
RNA reference duplexes of ca. +5 °C reported earlier4,11 and
ca. +4 °C obtained herein) than with DNA is presently not clear.
However, it has been established that R-L-LNA·RNA duplexes
are able to adopt overall conformations ranging from an A-form
to a form intermediate between the standard A- and B-forms,11
and differences in single-strand and/or duplex hydration patterns
are likely of importance for the preferential RNA-binding of
R-L-LNA, as reported earlier for other high-affinity oligonucle-
otide analogues.27
CD Spectral Analysis. CD (circular dichroism) spectral
analysis allows an estimate of the overall duplex conformation
and thus of the general influence of introducing modifications
into oligonucleotides. In Figure 6, CD curves of duplexes
involving the mix-mer r-L-LNA-1 are shown compared with
those of the unmodified duplexes, whereas Figure 7 depicts the
corresponding curves of duplexes involving the fully modified
r-L-LNA-2 and its relevant reference duplexes. Compared with
the reference duplexes, a general trend is toward a less intense
negative CD band around 240-250 nm and a more intense band
around 260-280 nm, the latter being shifted toward shorter
wavelengths for the modified duplexes. However, for both r-L-
LNA-1 and r-L-LNA-2 hybridized to RNA, similarities with
the reference DNA·RNA duplexes are demonstrated, indicating
the overall duplex conformations as being intermediary between
the A- and B-forms. R-L-LNA·DNA duplexes more clearly
deviate from the reference DNA·DNA duplexes since the use
(27) (a) Egli, M.; Portmann, S.; Usman, N. Biochemistry 1996, 35, 8489. (b)
Adamiak, D. A.; Rypniewski, W. R.; Milecki, J.; Adamiak, R. W. Nucleic
Acids Res. 2001, 29, 4144.
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2169

A R T I C L E S
Sørensen et al.
Figure 8. Time course of snake venom phosphodiesterase digestion of R-L-LNA and LNA. See the Experimental Section for details.
of fully modified R-L-LNA seems to induce a shift of the overall
conformation of R-L-LNA·DNA duplexes toward an intermedi-
ate form approaching that of RNA·DNA duplexes. In summary,
the CD spectra shown in Figures 6 and 7 indicate that R-L-
LNA oligomers are able to structurally mimic not only DNA
but also to some extent RNA.
3′-Exonucleolytic Stability of r-L-LNA in Vitro. To
evaluate the stability of R-L-LNA toward 3′-exonucleolytic
degradation, we used snake venom phosphodiesterase (SVPDE)
in a simple in vitro assay (see the Experimental Section for
Figure 9. RNase H-mediated hydrolysis of chimeric T3 upon hybridization
with either DNA-3, r-L-LNA-3, or LNA-3. Lane 1, partial alkaline
hydrolysis (0.2 M Na2CO3, 80 °C, 5 min) of T3; lanes 2 and 3, control
tubes with the DNA-3·T3 duplex without RNase H (incubation times 10
min and 21 h); lanes 4 and 5, RNase H activity toward the DNA-3·T3
duplex (incubation times 10 min and 21 h); lanes 6-10, RNase H activity
toward the r-L-LNA-3·T3 duplex (incubation times 10 min and 1, 4, 8,
and 21 h); lanes 11 and 12, RNase H activity toward the LNA-3·T3 duplex
(incubation times 10 min and 21 h).
details).28,29 During SVPDE digestion of unmodified oligo-
nucleotides, the absorbance (hyperchromicity) at 260 nm
increases due to the conversion of the oligonucleotide into
monomeric constituents.30 Figure 8 shows the results for the
mix-mers r-L-LNA-1 and LNA-1, the fully modified r-L-
LNA-2 and LNA-2, and the corresponding unmodified DNA
controls DNA-1 and DNA-2. A rapid increase in absorbance
for the unmodified DNA-1 (and DNA-2) follows immediately
after addition of SVPDE and declines after less than 5 min.
This indicates the expected fast 3′-exonucleolytic degradation
of the unmodified DNA.28,29 It can be seen that both the mix-
mers r-L-LNA-1 and LNA-1 as well as fully modified r-L-
LNA-2 and LNA-2 are significantly stabilized toward degra-
dation by SVPDE since no (or only limited) increase in
absorbance is observed after 1 h (the time scale depicted is
limited to 30 min for clarity). These qualitative experiments
indicate that mix-mer R-L-LNA, mix-mer LNA,31 fully modified
R-L-LNA, and fully modified LNA2a exhibit significant stabi-
lization toward 3′-exonucleolytic degradation.
H,32 the method uses a double chimeric approach with six of
the desired modified monomers centered in a mixed DNA
sequence. To avoid strand cleavage outside the modified region,
the RNA complement has 2′-O-methylribonucleotide monomers
positioned opposite to the unmodified DNA monomers since
2′-O-methyl-RNA/DNA hybrids are not substrates of RNase
H.33 E. coli RNase H was used because of its easy availability
and the fact that its cleavage properties are not very different
from those of the mammalian enzyme.34
Initial evaluation of the ability of R-L-LNA to activate RNase
H as compared to LNA by the above-mentioned approach is
shown in Figure 9. Lane 1 shows partial alkaline hydrolysis of
the chimeric RNA target, while lanes 2 and 3 are negative
controls of the double chimeric DNA/RNA duplex (DNA-3·
T3; see Figure 3) after 10 min and 21 h, respectively, without
any addition of enzyme. As positive control, the quick degrada-
tion of the same DNA/RNA duplex 10 min and 21 h after
addition of RNase H, respectively, is shown in lanes 4 and 5.
Lanes 6-10 depict cleavage of the r-L-LNA-3·T3 duplex after
10 min and 1, 4, 8, and 21 h, respectively. As can be seen,
considerable cleavage of the RNA strand is only seen after
prolonged exposure to the reaction conditions. In contrast, no
cleavage at all can be detected for the LNA-3·T3 duplex, after
10 min or 21 h, as shown in lanes 11 and 12.
In Vitro Cleavage of r-L-LNA/RNA and LNA/RNA
Hybrids by E. coli RNase H. Recently, a successful method
for evaluating the ability of an oligonucleotide analogue to
activate RNase H-induced cleavage of its complementary RNA
strand was reported that only requires incorporation of the
adenine monomer in the oligonucleotide analogue.32 As poly-
(dA)/poly(rU) duplexes are poor substrates for E. coli RNase
(28) Rosemeyer, H.; Seela, F. HelV. Chim. Acta 1991, 74, 748.
(29) Svendsen, M. L.; Wengel, J.; Dahl, O.; Kirpekar, F.; Roepstorff, P.
Tetrahedron 1993, 49, 11341.
(30) Newman, P. C.; Nwosu, V. U.; Williams D. M.; Cosstick, R.; Seela, F.;
Connolly, B. A. Biochemistry 1990, 29, 9891.
(31) Wahlestedt, C.; Slami, P.; Good, L.; Kela, J.; Johnsson, T.; Ho¨kfeldt, T.;
Broberger, C.; Porreca, F.; Lai, J.; Ren, K.; Ossipov, M.; Koshkin, A.;
Jakobsen, N.; Skouv, J.; Oerum, H.; Jacobsen, M. H.; Wengel, J. Proc.
Natl. Acad. Sci. U.S.A. 2000, 97, 5633.
(32) Wang, J.; Verbeure, B.; Luyten, I.; Lescrinier, E.; Froeyen, M.; Hendrix,
C.; Rosemeyer, H.; Seela, F.; Van Aerschot, A.; Herdewijn, P. J. Am. Chem.
Soc. 2000, 122, 8595.
(33) Inoue, H.; Hayase, Y.; Iwai, S.; Ohtsuka, E. FEBS Lett. 1987, 215, 327.
(34) Monia, B. P.; Lesnik, E. A.; Gonzalez, C.; Lima, W. F.; McGee, D.;
Guinosso, C. J.; Kawasasaki, A. M.; Cook, P. D.; Freier, S. M. J. Biol.
Chem. 1993, 268, 14514.
9
2170 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

R-L-ribo-Configured Locked Nucleic Acid
A R T I C L E S
are used as monomers in oligonucleotides because of the risk
of conversion of the 5-methylcytosine moiety into a thymine
moiety during the standard basic deprotection conditions used
after automated oligonucleotide synthesis.18 Because of the
difference of only one mass unit between a 5-methylcytosine
and a thymine base, MALDI-MS analysis does not reveal
whether such a transformation has taken place for the oligomers
DNA-2, r-L-LNA-2, and LNA-2 studied herein. However, the
obtained Tm values, especially as these follow the expected
Watson-Crick base-pairing rules, strongly support the identity
and stability of the 5-methylcytosine bases under the conditions
used.35
Figure 10. RNase H-mediated hydrolysis of T2 upon hybridization with
either DNA-1, r-L-LNA-1, LNA-1, or LNA-2. Lane 1, partial alkaline
hydrolysis (0.2 M Na2CO3, 80 °C, 7 min) of RNA target; lanes 2 and 3,
control tubes with the DNA-1·T2 duplex without RNase H (incubation times
10 min and 9.5 h); lanes 4 and 5, RNase H activity toward the DNA-1·T2
duplex (incubation times 10 min and 9.5 h); lanes 6-9, RNase H activity
toward the r-L-LNA-1·T2 duplex (incubation times 10 min and 1, 4, and
9.5 h); lanes 10-13, RNase H activity toward the LNA-1·T2 duplex
(incubation times 10 min and 1, 4, and 9.5 h); lanes 14 and 15, RNase H
activity toward the LNA-2·T2 duplex (incubation times 10 min and 9.5 h).
The binding studies have demonstrated Watson-Crick hy-
bridization of both mix-mer R-L-LNA and fully modified R-L-
LNA toward both complementary DNA and RNA. High-affinity
recognition comparable to that of LNA is possible toward
complementary DNA only for fully modified R-L-LNA, whereas
both mix-mer R-L-LNA and fully modified R-L-LNA display
high binding affinity toward RNA (∆Tm values per modification
ca. +4 °C). We have shown earlier4d that R-L-LNA·R-L-LNA
base pairing is very strong, as is LNA·LNA base pairing.2d
Therefore, careful design of sequences or, alternatively, the use
of mix-mer sequences is necessary for in vivo applications. It
is thus a very important feature that it is possible to combine
DNA monomers and R-L-LNA monomers in mix-mer R-L-LNA
oligomers without compromising the high-affinity recognition
of RNA. This fact, in combination with the relatively low
affinity of mix-mer R-L-LNAs toward DNA, suggests (mix-
mer) R-L-LNA to be potentially very attractive for antisense
applications. Other important points that have been convincingly
demonstrated herein are satisfactory discrimination between
matched and mismatched complementary sequences and en-
hanced stability against 3′-exonucleolytic degradation.
Successful activation of RNase H by LNA mix-mers and gap-
mers has previously been reported,31 for which reason the mix-
mer r-L-LNA-1 was compared with the analogous mix-mer
LNA-1. As shown in Figure 10, cleavage of an 11-mer RNA
strand when hybridized to either DNA-1, r-L-LNA-1, LNA-
1, or LNA-2 was studied. Lane 1 shows partial alkaline
hydrolysis of the RNA strand. Lanes 2 and 3 show DNA/RNA
(DNA-1·T2) control tubes without enzyme addition after 10 min
and 9.5 h, respectively. Lanes 4 and 5 depict the rapid
degradation of the unmodified DNA/RNA duplex after 10 min
and 9.5 h, respectively. Lanes 6-9 show the progress in the
cleavage of the r-L-LNA-1·T2 duplex after 10 min and 1, 4,
and 9.5 h, respectively, while no degradation could be detected
for the analogous mix-mer LNA-1 (lanes 10-13) within the
same time intervals. The latter result contrasts with the previ-
ously reported RNA cleavage obtained with mix-mer LNAs of
a different sequence.31 A similar negative result was obtained
for the fully modified LNA-2 when hybridized to T2 after 10
min and 9.5 h, as shown in lanes 14 and 15, respectively. The
different mobility observed in lanes 14 and 15 can be attributed
to migration of the RNA strand as complexes with LNA, despite
the denaturing conditions of the gel. This emphasizes the indeed
very strong hybridization between LNA and RNA. Identical
cleavage patterns at different RNase H concentrations were
obtained for DNA-1 and the corresponding DNA with 5-me-
thylcytosine nucleobases (DNA-2) (data not shown). Therefore
any unfavorable interactions between oligonucleotides contain-
ing 5-methylcytosine nucleobases and the enzyme (data not
shown) can be excluded, a result that supports the general
applicability of 5-methylcytosine nucleobases for RNase H
studies.
An AON has basically two possible modes of action, which
both involve hybridization to the RNA target. One is simply
steric blocking of the mRNA; the other is recognition of the
RNA·AON duplex as a substrate for the enzyme RNase H,
which then cleaves the RNA strand of an RNA/DNA duplex.
RNase H has been reported to bind in the minor groove of
substrate RNA·DNA heteroduplexes adopting a duplex form
intermediate between the A- and B-form, with a minor-groove
width also intermediate between that of the A- and B-forms.36
The furanose conformations in the RNA strand are of the N-type,
whereas hybridization of a DNA strand to the RNA strand
causes the furanose conformations of the DNA strand to change
from the typical S-type (C2′-endo) into E-type conformations
(O4′-endo range).36 Thus, the activation of RNase H proposedly
requires AONs with furanose rings able to adopt E-type (O4′-
(35) The potential conversion of 5-methylcytosine LNA monomers into thymine
LNA monomers during the deprotection of LNA oligomers has been studied
by analytical ion-exchange chromatography at pH 11, allowing a clear
distinction between oligomers of identical sequence but with, e.g., one
5-methylcytosine monomer being converted into a thymine monomer
(Daniel S. Pedersen, Anders M. Sørensen and Troels Koch personal
communication; manuscript in preparation). The stability of 5-methylcy-
tosine LNA monomers was confirmed (under deprotection conditions
similar to those used herein), which further supports the assigned composi-
tions of DNA-2, r-L-LNA-2, and LNA-2.
(36) (a) Salazar, M.; Fedoroff, O. Y.; Miller, M.; Ribeiro, N. S.; Reid, B. R.
Biochemistry 1993, 32, 4207. (b) Gyi, J. I.; Conn, G. L.; Lane, A. N.;
Brown, T. Biochemistry 1996, 35, 2538. (c) Nakamura, H.; Oda, Y.; Iwai,
S.; Inoue, H.; Ohtsuka, E.; Kanaya, S.; Kimura, S.; Katsuda, C.; Katayanagi,
K.; Morikawa, K.; Miyashiro, H.; Ikehara, M. Proc. Natl. Acad. Sci. U.S.A.
1991, 88, 11535. (d) Daniher, A. T.; Xie, J.; Mathur, S.; Bashkin, J. K.
Bioorg. Med. Chem. 1997, 5, 1037.
Discussion and Conclusion
With the procedures introduced earlier4a,4c,4d,5 and herein,
feasible synthetic routes toward the pyrimidine and purine R-L-
LNA nucleoside derivatives have been developed. Oligomer-
ization of the corresponding phosphoramidite derivatives on a
DNA synthesizer at extended couplings times (∼10 min) allows
the synthesis of mix-meric or fully modified R-L-LNA. The
tautomeric characteristics of 5-methylcytosine nucleosides
described above (Figure 2) are a point of concern when these
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2171

A R T I C L E S
Sørensen et al.
over the course of 10 min. The mixture was allowed to warm to room
temperature, and after 18 h analytical TLC (CH2Cl2/MeOH/pyridine,
89.8/10.0/0.2 v/v/v) indicated complete conversion to a single product
with very low mobility. The mixture was poured into a solution of
saturated NH3 in MeOH (30 mL) and the mixture was kept in a sealed
flask at room temperature overnight. Evaporation to dryness under
reduced pressure followed by silica gel column chromatography with
CH2Cl2/MeOH/pyridine (0-5% MeOH and 0.5% pyridine by volume)
as eluent yielded nucleoside 7 (241 mg, 81%) as an off-white solid
material. Rf 0.16 (CH2Cl2/MeOH/pyridine, 89.8/10.0/0.2 v/v/v); FAB-
endo), or perhaps S-type (C2′-endo), conformations. The locked
furanose conformations of R-L-LNA might therefore explain
the only limited ability of R-L-LNA·RNA duplexes to act as
substrates for RNase H (high enzyme concentration and
extended reaction time) despite the DNA-mimicking nature of
R-L-LNA in R-L-LNA·RNA duplexes. In addition, the changed
positioning of atoms and/or groups compared with natural DNA·
RNA duplexes in the minor groove might contribute to the lack
of efficient RNase H activation despite the additional 2′-C,4′-
C-oxymethylene linker being conveniently accommodated on
the brim of the major groove.11
Duplexes between RNA and arabinonucleic acids, an RNA
stereoisomer, have been reported to be effecient substrates of
RNase H.37 It is noteworthy, however, that both stereoregular
fully modified R-L-LNA and stereoirregular mix-mer R-L-LNA
are able to induce RNase H cleavage at all when hybridized to
RNA. Thus, whereas the backbone, e.g., the configurations at
C3′ and C4′, is identical in RNA and arabinonucleic acids, it
is, if the base is used as a point of reference, inverted at both
C3′ and C4′ in R-L-LNA. The results obtained with R-L-LNA
therefore not only substantiate the lack of stereoselectivity for
Watson-Crick hybridization as demonstrated earlier, e.g., by
the arabinonucleic acids37 or the R-oligodeoxynucleotides,38 but
also question the (stereo)selectivity of nucleic acid-protein
interactions.
1
MS m/z 572 [M + H]+; H NMR (CDCl3) δ 7.62-7.22 (m, 10H),
6.86 (d, 4H, J ) 7.9 Hz), 5.98 (s, 1H), 4.77 (s, 1H), 4.48 (s, 1H),
4.25-4.00 (m, 2H), 3.80-3.45 (m, 8H), 1.90 (s, 3H); 13C NMR (CDCl3)
δ 165.1, 158.3, 156.4, 144.5, 137.6, 135.4, 135.2, 129.8, 127.8, 127.6,
126.6, 113.0, 101.3, 89.5, 87.9, 85.9, 79.0, 73.5, 73.0, 60.4, 54.9, 13.2.
(1S,3R,4S,7R)-3-(4-N-Benzoyl-5-methylcytosin-1-yl)-1-[[(4,4′-
dimethoxytrityl)oxy]methyl]-7-hydroxy-2,5-dioxabicyclo[2.2.1]-
heptane (8). To a solution of nucleoside 7 (90 mg, 0.16 mmol) in
anhydrous pyridine (5 mL) was added chlorotrimethylsilane (0.04 mL,
0.31 mmol), and the mixture was stirred for 1.5 h at room temperature.
N-Benzoyl-1H-tetrazole17 (42 mg, 0.24 mmol) was added and stirring
was continued for 12 h at room temperature. H2O (10 mL) was added,
and after being stirred for an additional 5 h at room temperature, the
mixture was diluted with CH2Cl2 (100 mL), washed with a saturated
aqueous solution of NaHCO3 (3 × 100 mL), and dried (Na2SO4). The
solvent was removed under reduced pressure and the residue was
purified by silica gel column chromatography (CH2Cl2/MeOH/pyridine,
0-4% MeOH and 0.5% pyridine by volume) to give nucleoside 8 as
a white solid material (86 mg, 80%). Rf 0.61 (CH2Cl2/MeOH/pyridine,
Experimental Section
1
89.8/10.0/0.2 v/v/v); FAB-MS m/z 676 [M + H]+; H NMR (CDCl3)
General. Reactions were conducted under an atmosphere of nitrogen
when anhydrous solvents were used. All reagents were obtained from
commercial suppliers and were used without further purification.
Petroleum ether of the distillation range 60-80 °C was used. The silica
gel (0.040-0.063 mm) used for column chromatography was purchased
from Merck. After organic phases were dried with Na2SO4, filtration
was performed. After column chromatography, fractions containing
product were pooled, evaporated under reduced pressure, and dried
overnight under high vacuum to give the product unless otherwise
specified. 1H NMR spectra were recorded at 300 or 400 MHz, 13C NMR
spectra at 75.5 or 100.6 MHz, and 31P NMR spectra at 121.5 MHz.
Chemical shifts are reported in parts per million (ppm) relative to either
tetramethylsilane or the deuterated solvent as the internal standard for
1H and 13C and relative to 85% H3PO4 as the external standard for 31P.
Assignments of NMR spectra, when given, are based on 2D spectra
and follow the standard carbohydrate/nucleoside nomenclature (the
carbon atom of the 4′-C-substituent is numbered C-5′′ even though the
systematic compound names of the bicyclic structures are given
according to the von Baeyer nomenclature). The assignments of
methylene carbon and hydrogen atoms may, in general, be interchanged.
Fast-atom bombardment mass spectra (FAB-MS) were recorded in
positive ion mode.
δ 8.32 (d, 2H, J ) 7.3 Hz, Bz), 7.74 (s, 1H, 6-H), 7.55-7.22 (m, 12H,
DMT, Bz), 6.86 (d, 4H, J ) 8.5 Hz, DMT), 5.99 (s, 1H, 1′-H), 4.59 (s,
1H, 2′-H), 4.42 (s, 1H, 3′-H), 4.13 (d, 1H, J ) 9.1 Hz, 5′-Ha), 4.07 (d,
1H, J ) 9.1 Hz, 5′-Hb), 3.79 (s, 6H, OCH3), 3.59 (d, 1H, J ) 10.8 Hz,
5′′-Ha), 3.52 (d, 1H, J ) 10.8 Hz, 5”-Hb), 2.19 (s, 3H, CH3); 13C NMR
(CDCl3) δ 179.0 (CdO), 159.7 (C-4), 158.4 (DMT), 147.8 (C-2), 144.2,
135.2, 132.3, 129.9, 129.7, 128.0, 127.8, 126.8 (DMT, Bz), 136.9 (C-
6), 113.1 (DMT), 110.4 (C-5), 89.7, 86.2 (C-4′, DMT), 87.7 (C-1′),
78.7 (C-2′), 74.1 (C-3′), 72.7 (C-5′), 59.8 (C-5′′), 55.1 (DMT), 13.8
(CH3). The 1H NMR data obtained are in agreement with the 1H NMR
data obtained for the corresponding enantiomeric derivative (unpub-
lished results, Poul Nielsen, Department of Chemistry, University of
Southern Denmark).
(1S,3R,4S,7R)-7-[2-Cyanoethoxy(diisopropylamino)phosphinoxy]-
1-[[(4,4′-dimethoxytrityl)oxy]methyl]-3-(4-N-benzoyl-5-methylcytosin-
1-yl)-2,5-dioxabicyclo[2.2.1]heptene (9). To a stirred solution of
nucleoside 8 (255 mg, 0.38 mmol) in anhydrous CH2Cl2 (8 mL) at 0
°C was added (N,N-diisopropyl)ethylamine (0.20 mL, 1.13 mmol). After
dropwise addition of 2-cyanoethyl [(N,N-diisopropyl)phosphoramido]-
chloridite (0.13 mL, 0.57 mmol), the mixture was allowed to warm to
room temperature and stirring was continued for 12 h. MeOH (0.1 mL)
was added, and the mixture was diluted with CH2Cl2 (30 mL), washed
with a saturated aqueous solution of NaHCO3 (3 × 30 mL), and dried
(Na2SO4). The organic phase was evaporated to dryness under reduced
pressure, and the residue was purified by silica gel column chroma-
tography with EtOAc/n-hexane/NEt3 (49.5/49.5/1.0 v/v/v) as eluent to
give amidite 9 as a white foam (148 mg, 45%). Rf 0.88 (CH2Cl2/MeOH/
pyridine, 89.8/10.0/0.2 v/v/v); 31P NMR [(CD3)2SO] δ 150.5, 150.4.
(1S,3R,4S,7R)-1-[[(4,4′-Dimethoxytrityl)oxy]methyl]-7-hydroxy-
3-(5-methylcytosin-1-yl)-2,5-dioxabicyclo[2.2.1]heptane (7). To a
stirred solution of nucleoside 55 (300 mg, 0.52 mmol) in a mixture of
anhydrous CH2Cl2 and anhydrous pyridine (20 mL, 3/2 v/v) was added
chlorotrimethylsilane (0.13 mL, 1.04 mmol). After being stirred for 1
h at room temperature, the mixture was cooled to 0 °C and trifluo-
romethanesulfonic anhydride (0.26 mL, 1.56 mmol) was added dropwise
9-[2-O-Acetyl-5-O-benzoyl-4-C-[(benzoyloxy)methyl]-3-O-benzyl-
r-L-threo-pentofuranosyl]-6-N-benzoyladenine (11). Furanose 1012
(2.05 g, 3.65 mmol) was dissolved in anhydrous acetonitrile (30 mL).
6-N-Benzoyladenine (1.86 g, 7.78 mmol) and SnCl4 (1.30 mL, 11.11
mmol) were added, and the mixture was stirred at room temperature
for 4 h. The reaction was quenched by addition of a saturated aqueous
solution of NaHCO3 until the evolution of CO2 had ceased, whereupon
(37) (a) Damha, M. J.; Wilds, C. J.; Noronha, A.; Brukner, I.; Borkow, G.;
Arion, D.; Parniak, M. A. J. Am. Chem. Soc. 1998, 120, 12976. (b) Wilds,
C. J.; Damha, M. J. Nucleic Acids Res. 2000, 28, 3625. (c) Noronha, A.
M.; Wilds, C. J.; Lok, C. N.; Viazovkina, K.; Arion, D.; Parniak, M. A.;
Damha, M. J. Biochemistry 2000, 39, 7050.
(38) (a) Gagnor, C.; Bertrand, J.-R.; Thenet, S.; Lemaitre, M.; Morvan, F.;
Rayner, B.; Malvy, C.; Lebleu, B.; Imbach, J.-L.; Paoletti, C. Nucleic Acids
Res. 1987, 15, 10419. (b) Thuong, N. T.; Asseline, U.; Roig, V.; Takasugi,
M.; Hele´ne`, C. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 5129.
9
2172 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

R-L-ribo-Configured Locked Nucleic Acid
A R T I C L E S
the mixture was filtered through a layer of Celite 545. The layer of
Celite 545 was washed with CH2Cl2 (2 × 50 mL) and the combined
filtrates were washed successively with a saturated aqueous solution
of NaHCO3 (3 × 150 mL) and H2O (2 × 150 mL), dried (Na2SO4),
and evaporated to dryness under reduced pressure. The residue was
purified by silica gel column chromatography with EtOAc/petroleum
ether (40-60% EtOAc by volume) as eluent to give nucleoside 11 as
a white solid material (1.40 g, 52%). Rf 0.40 (MeOH/CH2Cl2, 2/23 v/v);
FAB-MS m/z 742 [M + H]+; 1H NMR (CDCl3) δ 9.60 (br s, 1H, 6-H),
8.73 (s, 1H, 8-H), 8.51 (s, 1H, 2-H), 8.02-7.98 (m, 4H), 7.89-7.86
(m, 2H), 7.58-7.16 (m, 14H), 6.58 (d, 1H, J ) 2.6 Hz, 1′-H), 5.88 (t,
1H, J ) 2.2 Hz, 2′-H), 4.93 (d, 1H, J ) 12.0 Hz, Bn), 4.79 (d, 1H, J
) 12.0 Hz, Bn), 4.70-4.61 (m, 4H), 4.34 (d, 1H, J ) 1.8 Hz, 3′-H),
2.17 (s, 3H, Ac); 13C NMR (CDCl3) δ 169.3, 165.5, 165.3, 152.3, 151.3,
149.4, 135.5, 133.3, 133.1, 132.9, 132.3, 129.3, 129.3, 129.1, 128.7,
128.3, 128.2, 128.1, 128.1, 128.0, 128.0, 127.6, 122.8 (Bn, Bz, Ac,
C-4, C-5, C-6, C-8), 141.1 (C-2), 87.2, 87.0 (C-1′, C-4′), 80.8 (C-3′),
80.1 (C-2′), 72.2 (Bn), 62.7, 62.3 (C-5′, C-5′′), 20.5 (Ac).
residue was purified by silica gel column chromatography with MeOH/
CH2Cl2 (4-6% MeOH by volume) as eluent, yielding nucleoside 15
(656 mg, 91%) as a white foam, which was used in the next step without
further purification. Rf 0.47 (MeOH/CH2Cl2, 1/9); FAB-MS m/z 492
1
[M + H]+; H NMR [(CD3)2SO] δ 11.18 (br s, 1H, 6-NH), 8.75 (s,
1H, 2-H), 8.66 (s, 1H, 8-H), 8.07-7.95 (m, 2H), 7.67-7.29 (m, 8H),
6.48 (d, 1H, J ) 4.8 Hz, 1′-H), 5.92 (br s, 1H, 2′-OH), 5.35 (br s, 1H),
5.09 (br s, 1H), 4.81 (d, 1H, J ) 11.6 Hz), 4.67 (m, 1H, 2′-H), 4.62 (d,
1H, J ) 11.6 Hz) 4.40 (d, 1H, J ) 5.0 Hz, 3′-H), 3.70 (d, 1H, J )
11.6 Hz), 3.63 (d, 1H, J ) 11.5 Hz), 3.54 (s, 2H); 13C NMR [(CD3)2-
SO] δ 164.7, 151.5, 150.1, 138.4, 133.5, 132.8, 132.7, 129.3, 128.6,
128.5, 128.3, 127.6, 127.5, 125.1 (Bn, Bz, C-4, C-5, C-6), 152.3 (C-
2), 144.4 (C-8), 110.4, 87.8 (C-4′), 83.7 (C-1′), 79.4 (C-3′), 72.5 (Bn),
69.9 (C-2′), 63.3, 61.1 (C-5′, C-5′′).
9-[3-O-Benzyl-5-O-(methanesulfonyl)-4-C-[[(methanesulfonyl)-
oxy]methyl]-r-L-erythro-pentofuranosyl]-6-N-benzoyladenine (16).
Nucleoside 15 (1.75 g, 3.57 mmol) was coevaporated with anhydrous
pyridine (100 mL) and dissolved in anhydrous pyridine (100 mL). The
mixture was cooled to 0 °C and methanesulfonyl chloride (0.61 mL,
7.9 mmol) was added dropwise. After the mixture was stirred for 2 h,
H2O (10 mL) was added and the mixture was evaporated to dryness
under reduced pressure. The residue was dissolved in CH2Cl2 (75 mL)
and washed with a saturated aqueous solution of NaHCO3 (3 × 75
mL) and dried (Na2SO4). The organic phase was evaporated to dryness
under reduced pressure and the residue was purified by silica gel column
chromatography with CH2Cl2/MeOH (2-4% MeOH by volume) as
eluent, affording nucleoside 16 (1.83 g, 79%) as a white solid material.
9-[2-O-Acetyl-5-O-benzoyl-4-C-[(benzoyloxy)methyl]-3-O-benzyl-
r-L-erythro-pentofuranosyl]-6-N-benzoyladenine (14). To a solution
of nucleoside 11 (3.63 g, 4.90 mmol) in MeOH (75 mL) was added a
saturated solution of NH3 in MeOH (75 mL). The solution was stirred
at 0 °C for 2 h and evaporated to dryness under reduced pressure. The
residue was coevaporated with anhydrous toluene (2 × 40 mL) and
dissolved in a mixture of anhydrous CH2Cl2 (100 mL) and anhydrous
pyridine (20 mL). After the mixture was cooled to -30 °C, trifluo-
romethanesulfonic anhydride (1.7 mL, 10.3 mmol) was added dropwise.
The reaction mixture was allowed to warm to room temperature, and
after 1.5 h additional trifluoromethanesulfonic anhydride (0.5 mL, 3.03
mmol) was added. After being stirred for an additional 30 min, the
mixture was diluted with CH2Cl2 (130 mL), washed with a saturated
aqueous solution of NaHCO3 (3 × 200 mL) and dried (Na2SO4). After
evaporation of the solvents, the residue was coevaporated with
anhydrous toluene (2 × 75 mL) and dissolved in anhydrous toluene
(150 mL). KOAc (2.45 g, 25.0 mmol) and 18-crown-6 (2.56 g, 9.70
mmol) were added, and the mixture was stirred for 30 min at room
temperature and then heated to 80 °C for 30 min. The mixture was
cooled to room temperature and evaporated to dryness under reduced
pressure. The residue was purified by silica gel column chromatography
with MeOH/CH2Cl2 (0-2% MeOH by volume) as eluent, affording
nucleoside 14 (3.05 g, 84%) as a white solid material. Rf 0.44 (MeOH/
1
Rf 0.63 (MeOH/CH2Cl2, 1/9 v/v); FAB-MS m/z 648 [M + H]+; H
NMR (CDCl3) δ 9.10 (s, 1H, 6-NH), 8.52 (s, 1H, 8-H), 8.48 (s, 1H,
2-H), 7.96-7.95 (m, 2H), 7,62-7.31 (m, 8H), 6.46 (d, 1H, J ) 3.7
Hz, 1′-H), 5.42 (br s, 1H, 2′-OH), 5.04 (d, 1H, J ) 11.7 Hz, 3′-H),
4.79 (d, 1H, J ) 11.7 Hz, Bn), 4.68 (br s, 1H, 2′-H), 4.58 (d, 1H, J )
11.7 Hz, Bn), 4.45-4.26 (m, 4H, 5′-H, 5′′-H), 3.03 (s, 3H, Ms), 3.01
(s, 3H, Ms); 13C NMR (CDCl3) δ 174.3, 164.8, 151.6, 148.5, 136.3,
133.4, 132.9, 128.9, 128.8, 128.7, 128.5, 127.9, 121.8 (Bn, Bz, C-4,
C-5, C-6), 152.1 (C-2), 143.5 (C-8), 84.8 (C-1′), 82.4 (C-4′), 78.8, 68.3
(C-5′, C-5′′), 73.6 (Bn), 70.3 (C-2′), 69.0 (C-3′), 37.7 (Ms), 37.6 (Ms).
(1S,3R,4S,7R)-1-(Acetoxymethyl)-3-(6-N-benzoyladenin-9-yl)-7-
benzyloxy-2,5-dioxabicyclo[2.2.1]heptane (18). Nucleoside 16 (1.31
g, 2.02 mmol) was coevaporated with anhydrous toluene (50 mL) and
dissolved in anhydrous THF (50 mL). The mixture was cooled to 0 °C
and NaH [60% suspension in mineral oil (by weight), 175 mg, 4.38
mmol] was added. After being stirred at 0 °C for 6.5 h, the reaction
mixture was evaporated to dryness under reduced pressure and the
residue was dissolved in CH2Cl2 (100 mL). Washing was performed
with a saturated aqueous solution of NaHCO3 (3 × 100 mL) and the
organic phase was dried (Na2SO4) and evaporated to dryness under
reduced pressure. The residue was dissolved in 1,4-dioxane (50 mL),
and KOAc (1.02 g, 10.4 mmol) and 18-crown-6 (1.08 g, 4.09 mmol)
were added. The mixture was heated under reflux for 19 h and
subsequently evaporated to dryness under reduced pressure. The residue
was purified by silica gel column chromatography with CH2Cl2/MeOH
(0.5% MeOH by volume) as eluent, yielding nucleoside 18 (911 mg,
87%) as a white solid material, which was used in the next step without
further purification. Rf 0.60 (MeOH/CH2Cl2, 1/9 v/v); FAB-MS m/z
1
CH2Cl2, 1/19 v/v); FAB-MS m/z 742 [M + H]+; H NMR (CDCl3) δ
9.00 (br s, 1H, 6-NH), 8.76 (s, 1H, 8-H), 8.59 (s, 1H, 2-H), 8.05-7.98
(m, 6H), 7.64-7.41 (m, 10H), 7.28-7.23 (m, 4H), 6.77 (d, 1H, J )
4.6 Hz, 1′-H), 5.86 (m, 1H, H-2′), 5.23 (d, 1H, J ) 12.3 Hz), 4.74-
4.66 (m, 3H), 4.56-4.50 (m, 3H), 2.04 (s, 3H, Ac); 13C NMR (CDCl3)
δ 168.8, 165.8, 142.7, 136.0, 133.5, 133.3, 132.7, 129.6, 129.6, 128.8,
128.6, 128.5, 128.4, 128.4, 128.1, 127.8 (Bn, Bz, Ac, C-2, C-4, C-5,
C-6, C-8), 83.8 (C-4′), 82.2 (C-1′), 78.4 (C-3′), 74.3 (Bn), 70.8 (C-2′),
64.7, 63.4 (C-5′, C-5′′), 20.5 (Ac).
9-[3-O-Benzyl-4-C-(hydroxymethyl)-r-L-erythro-pentofuranosyl]-
6-N-benzoyladenine (15). Nucleoside 14 (1.09 g, 1.47 mmol) was
dissolved in a saturated solution of NH3 in MeOH (40 mL). After the
mixture was stirred for 30 min at room temperature, an aqueous solution
of NH3 (10 mL, 32% by weight) was added. After 22 h, the mixture
was evaporated to dryness under reduced pressure and the residue, after
coevaporation with anhydrous pyridine (3 × 20 mL), was dissolved in
anhydrous pyridine (20 mL). Chlorotrimethylsilane (2.8 mL, 22 mmol)
was added and stirring was continued for 2.5 h at room temperature.
The reaction mixture was cooled to 0 °C, BzCl (0.9 mL, 7.8 mmol)
was added, and stirring was continued for 16.5 h at room temperature.
After the mixture was cooled to 0 °C, water (40 mL) was added and
the mixture was evaporated to dryness under reduced pressure. The
residue was dissolved in H2O/MeOH/saturated methanolic ammonia
(65 mL, 3/4/6 v/v/v), and the mixture was stirred for 6.5 h at room
temperature and evaporated to dryness under reduced pressure. The
1
516 [M + H]+; H NMR (CDCl3) δ 9.37 (br s, 1H, 6-NH), 8.72 (s,
1H, 8-H), 8.51 (s, 1H, 2-H), 8.04-8.02 (m, 2H), 7.60-7.33 (m, 8H),
6.46 (s, 1H, 1′-H), 4.76 (d, 1H, J ) 11.9 Hz, Bn), 4.66 (d, 1H, J )
11.7 Hz, Bn), 4.65 (s, 1H, 2′-H), 4.50 (d, 1H, J ) 12.6 Hz, 5′′-H),
4.38 (d, 1H, J ) 12.6 Hz, 5′′-H), 4.33 (s, 1H, 3′-H), 4.30 (d, 1H, J )
8.4 Hz, 5′-H), 4.11 (d, 1H, J ) 8.2 Hz, 5′-H), 2.06 (s, 3H, Ac); 13C
NMR (CDCl3) δ 170.1 (Ac), 164.3, 152.4, 151.9, 149.5, 141.2, 136.4,
133.5, 132.5, 128.6, 128.5, 128.2, 127.8, 127.6, 122.8, (C-2, C-4, C-5,
C-6, C-8, Bz, Bn), 87.0 (C-4′), 84.5 (C-1′), 79.7 (C-3′), 77.1 (C-2′),
73.5 (C-5′), 72.3 (Bn), 59.7 (C-5′′), 20.5 (Ac).
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2173

A R T I C L E S
Sørensen et al.
(1R,3R,4S,7R)-3-(6-N-Benzoyladenin-9-yl)-7-benzyloxy-1-(hy-
droxymethyl)-2,5-dioxabicyclo[2.2.1]heptane (19). To a solution of
nucleoside 18 (1.13 g, 2.19 mmol) in MeOH (25 mL) was added a
saturated solution of NH3 in MeOH (25 mL). The reaction mixture
was stirred at 0 °C for 17 h and evaporated to dryness. The residue
was purified by silica gel column chromatography with CH2Cl2/MeOH
(1-5% MeOH by volume) as eluent, affording nucleoside 19 (862 mg,
83%) as a white solid material. Rf 0.22 (MeOH/CH2Cl2, 2/23 v/v); FAB-
(1S,3R,4S,7R)-7-[2-Cyanoethoxy(diisopropylamino)phosphinoxy]-
1-[[(4,4′-dimethoxytrityl)oxy]methyl]-3-(6-N-benzoyladenin-9-yl)-
2,5-dioxabicyclo[2.2.1]heptane (23). Nucleoside 22 (145 mg, 0.21
mmol) was coevaporated with anhydrous pyridine (2 × 3 mL) and
dissolved in anhydrous CH2Cl2 (5 mL). N,N-Diisopropylethylamine
(0.15 mL) was added and the solution was cooled to 0 °C, whereupon
2-cyanoethyl (N,N-diisopropyl)phosphoramidochloridite (0.10 mL, 0.52
mmol) was added. The mixture was stirred in the dark for 14.5 h at
room temperature, CH2Cl2 (45 mL) was added, and the mixture was
washed with a saturated aqueous solution of NaHCO3 (3 × 35 mL).
The organic phase was dried (Na2SO4) and evaporated to dryness under
reduced pressure. The residue was purified by silica gel column
chromatography with n-hexane/EtOAc/triethylamine (5/4/1 v/v/v) as
eluent, affording amidite 23 (118 mg, 63%) as a white foam. Rf 0.70,
0.77 (MeOH/CH2Cl2, 2/23 v/v); 31P NMR δ (CH3CN) 150.2, 150.0.
3-O-Benzyl-5-O-(methanesulfonyl)-4-C-[[(methanesulfonyl)oxy]-
methyl]-1,2-O-isopropylidene-â-L-threo-pentofuranose (25). Com-
pound 2424 (2.00 g, 6.40 mmol) was coevaporated with anhydrous
pyridine (20 mL) and dissolved in anhydrous pyridine (15 mL). The
solution was stirred and cooled to 0 °C, whereupon methanesulfonyl
chloride (1.25 mL, 16.1 mmol) was added dropwise. The mixture was
allowed to warm to room temperature, and after it was stirred for 2 h,
additional methanesulfonyl chloride (1.00 mL, 12.9 mmol) was added.
After another 2 h, ice-cold H2O (40 mL) was added, and extraction
was performed with diethyl ether (2 × 150 mL). The combined organic
phase was washed with brine (100 mL), dried (Na2SO4), and evaporated
to dryness under reduced pressure. The residue was purified by silica
gel column chromatography with MeOH/CH2Cl2 (1/99 v/v) as eluent,
giving furanose 25 (2.00 g, 66%) as a white semisolid material, which
was used in the next step without further purification. FAB-MS m/z
467 [M + H]+; 1H NMR δ (CDCl3) 7.37 (s, 5H), 5.80 (d, 1H, J ) 3.3
Hz), 4.89 (d, 1H, J ) 11.7 Hz), 4.79 (d, 1H, J ) 11.7 Hz), 4.67 (t, 1H,
J ) 4.4 Hz), 4.59 (d, 1H, J ) 11.8 Hz), 4.44 (d, 1H, J ) 12.1 Hz),
4.34 (d, 1H, J ) 11.0 Hz), 4.21-4.13 (m, 2H), 3.10 (s, 3H), 2.98 (s,
3H), 2.16 (s, 3H), 1.68 (s, 3H), 1.34 (s, 3H); 13C NMR δ (CDCl3)
136.5, 128.4, 128.2, 128.2, 128.1, 127.9, 113.8, 104.3, 83.0, 78.2, 77.6,
72.6, 69.3, 68.5, 37.8, 37.2, 26.0, 25.4.
1
MS m/z 474 [M + H]+; H NMR (CDCl3) δ 9.20 (br s, 1H, 6-NH),
8.74 (s, 1H, 8-H), 8.51 (s, 1H, 2-H), 8.04-8.01 (m, 2H), 7.60-7.27
(m, 8H), 6.47 (s, 1H, 1′-H), 4.79 (d, 1H, J ) 11.9 Hz), 4.69 (d, 1H, J
) 11.9 Hz), 4.62 (s, 1H, 2′-H), 4.37 (s, 1H, 3′-H), 4.23 (d, 1H, J ) 8.1
Hz), 4.04 (d, 1H, J ) 8.4 Hz), 3.99-3.89 (m, 3H, 5′′-H, 5′′-OH); 13
C
NMR (CDCl3) δ 164.8, 151.6, 149.3, 141.3, 141.3, 136.7, 133.4, 132.5,
128.5, 128.4, 128.0, 127.8, 127.5, 122.2 (C-4, C-5, C-6, C-8, Bz, Bn),
152.4 (C-2), 89.9 (C-4′), 84.5 (C-1′), 79.3 (C-3′), 77.0 (C-2′), 73.3 (C-
5′), 72.3 (Bn), 57.8 (C-5′′).
(1R,3R,4S,7R)-3-(6-N-Benzoyladenin-9-yl)-7-hydroxy-1-(hydroxy-
methyl)-2,5-dioxabicyclo[2.2.1]heptane (20) and (1R,3R,4S,7R)-3-
(Adenin-9-yl)-7-hydroxy-1-(hydroxymethyl)-2,5-dioxabicyclo[2.2.1]-
heptane (21). To a solution of nucleoside 19 (503 mg, 1.06 mmol) in
MeOH (40 mL) was added a catalytical amount of Pd/C (10% by
weight) and HCOONH4 (201 mg, 3.52 mmol). The mixture was heated
under reflux for 12 h and evaporated to dryness under reduced pressure.
The residue was purified by silica gel column chromatography with
CH2Cl2/MeOH (3-8% MeOH by volume) as eluent, yielding nucleo-
sides 20 (177 mg, 44%) and 21 (72 mg, 24%), both as white solid
materials. Data for 20: Rf 0.23 (MeOH/CH2Cl2, 13/87 v/v); FAB-MS
1
m/z 384 [M + H]+; H NMR [(CD3)SO] δ 11.10 (br s, 1H, 6-NH),
8.72 (s, 1H, 2-H), 8.65 (s, 1H, 8-H), 8.04 (d, 2H, J ) 8.2 Hz, Bz),
7.65-7.51 (m, 3H, Bz), 6.51 (s, 1H, 1′-H), 5.94 (d, 1H, J ) 4.4 Hz,
3′-OH), 4.93 (t, 1H, J ) 5.6 Hz, 5′′-OH), 4.44-4.42 (m, 2H, 2′-H,
3′-H), 4.09 (d, 1H, J ) 8.2 Hz, 5′-H), 4.02 (d, 1H, J ) 8.2 Hz, 5′-H),
3.73 (d, 2H, J ) 5.5 Hz, 5′′-H); 13C NMR [(CD3)SO] δ 152.2, 151.6,
150.3, 133.4, 132.5, 128.5, 125.2 (Bz, C-2, C-4, C-5, C-6), 142.5 (C-
8), 90.7 (C-4′), 84.4 (C-1′), 79.4 (C-2′), 72.9 (C-3′), 72.6 (C-5′), 57.5
(C-5′′). Data for 21: Rf 0.12 (MeOH/CH2Cl2, 13/87 v/v); FAB-MS
1
m/z 280 [M + H]+; H NMR [(CD3)SO] δ 8.34 (s, 1H, 2-H), 8.11 (s,
1,2-Di-O-acetyl-3-O-benzyl-5-O-(methanesulfonyl)-4-C-[[(meth-
anesulfonyl)oxy]methyl]-L-threo-pentofuranose (26). Compound 25
(5.00 g, 10.7 mmol) was dissolved in 80% TFA (40 mL). After being
stirred for 4 h at room temperature, the mixture was evaporated to
dryness under reduced pressure. The residue was coevaporated with
anhydrous pyridine (50 mL) and dissolved in anhydrous pyridine (50
mL), whereupon Ac2O (4.0 mL, 42.0 mmol) was added. The mixture
was stirred for 3 h at room temperature, a saturated aqueous solution
of NaHCO3 (150 mL) was added, and extraction was performed with
CH2Cl2 (3 × 200 mL). The combined organic phase was dried (Na2-
SO4) and evaporated to dryness under reduced pressure. The residue
was purified by silica gel column chromatography with CH2Cl2 as
eluent, affording furanose 26 (3.50 g, 63%, 4/5 anomeric mixture) as
a white solid material. FAB-MS m/z 511 [M + H]+; 13C NMR δ
(CDCl3) 169.5, 169.5, 169.1, 136.6, 136.5, 128.7, 128.6, 128.4, 128.3,
128.1, 99.5, 92.3, 86.5, 82.8, 81.5, 81.2, 79.8, 76.3, 73.7, 72.8, 68.8,
68.6, 67.9, 66.7, 37.9, 37.8, 37.7, 37.5, 21.0, 20.9, 20.7, 20.4.
1H, 8-H), 6.32 (s, 1H, 1′-H), 5.90 (br s, 1H, 3′-OH), 4.90 (br s, 1H,
5′′-OH), 4.36 (s, 1H, 3′-H), 4.32 (s, 1H, 2′-H), 4.03 (d, 1H, J ) 8.1
Hz, 5′-H), 3.99 (d, 1H, J ) 8.3 Hz, 5′-H), 3.69 (s, 2H, 5′′-H); 13C
NMR [(CD3)SO] δ 156.1 (C-6), 152.8 (C-2), 149.6 (C-4), 139.1 (C-
8), 128.7, 118.5 (C-5), 90.6 (C-4′), 84.0 (C-1′), 79.6 (C-2′), 73.1 (C-
3′), 72.7 (C-5′), 57.7 (C-5′′).
(1S,3R,4S,7R)-1-[[(4,4′-Dimethoxytrityl)oxy]methyl]-7-hydroxy-
3-(6-N-benzoyladenin-9-yl)-2,5-dioxabicyclo[2.2.1]heptane (22). Com-
pound 20 was coevaporated with anhydrous pyridine (30 mL, 12 mL)
and dissolved in anhydrous pyridine (6 mL) at room temperature,
whereupon 4,4′-dimethoxytrityl chloride (170 mg, 0.53 mmol) was
added. After being stirred for 11 h, the mixture was poured into
petroleum ether/CH2Cl2/H2O (30 mL, 1/1/1 v/v/v), and the organic phase
was separated, washed with a saturated aqueous solution of NaHCO3
(3 × 15 mL), and evaporated to dryness under reduced pressure. The
residue was coevaporated with anhydrous toluene (6 mL) and purified
by silica gel column chromatography with CH2Cl2/MeOH/pyridine (0-
2.5% MeOH and 0.5% pyridine by volume) as eluent, affording
nucleoside 22 (147 mg, 82%) as a white solid material. Rf 0.46 (MeOH/
9-[2-O-Acetyl-3-O-benzyl-5-O-(methanesulfonyl)-4-C-[[(methane-
sulfonyl)oxy]methyl]-r-L-threo-pentofuranosyl]-6-N-benzoylade-
nine (27). Furanose 26 (6.20 g, 12.1 mmol) was coevaporated with
anhydrous CH3CN (50 mL) and dissolved in anhydrous CH3CN (125
mL) under stirring at room temperature. 6-N-Benzoyladenine (7.20 g,
30.1 mmol) and then SnCl4 (5.0 mL, 42.7 mmol) were added. Shortly
thereafter the mixture turned clear, and after it was stirred for 1.5 h,
additional SnCl4 (3.2 mL, 27.3 mmol) was added. After 4.5 h, a
saturated aqueous solution of NaHCO3 was added slowly until CO2
evolution ceased. The mixture was filtered through Celite 545 three
times [after each filtration the Celite 545 was washed MeOH/CH2Cl2
(500 mL, 8/2 v/v)]. The combined filtrate was evaporated to dryness
1
CH2Cl2, 1/9 v/v); FAB-MS m/z 686 [M + H]+; H NMR (CDCl3) δ
9.35 (br s, 1H, 6-NH), 8.72 (s, 1H, 2-H), 8.52 (s, 1H, 8-H), 8.01-7.99
(m, 2H), 7.56-7.14 (m, 12H), 6.82-6.79 (m, 4H), 6.52 (s, 1H, 1′-H),
4.58-4.56 (m, 2H), 4.21-4.19 (m, 2H), 4.14 (d, 1H, J ) 8.6 Hz, 5′-
H), 3.75 (s, 6H, OCH3), 3.52 (s, 2H, 5′′-H); 13C NMR (CDCl3) δ 164.7,
158.4, 149.2, 144.2, 135.2, 135.1, 133.4, 132.6, 129.8, 128.9, 128.6,
128.0, 127.8, 127.8, 127.8, 126.8, 125.1, 122.4, 113.1 (DMT, C-4, C-5,
C-6, Bz), 152.4 (C-2), 141.4 (C-8), 89.2, 86.3 (DMT, C-4′), 84.6 (C-
1′), 79.3, 74.6 (C-2′, C-3′), 73.2 (C-5′), 60.0 (C-5′′), 55.1 (DMT).
9
2174 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

R-L-ribo-Configured Locked Nucleic Acid
A R T I C L E S
under reduced pressure and redissolved in CH2Cl2 (300 mL), washed
successively with saturated aqueous solutions of NaHCO3 (2 × 100
mL) and brine (2 × 100 mL), dried (Na2SO4), and evaporated to dryness
under reduced pressure. The residue was purified by silica gel column
chromatography with MeOH/CH2Cl2 (1.5/98.5 v/v) as eluent, yielding
nucleoside 27 (4.80 g, 57%) as a white solid material. FAB-MS m/z
690 [M + H]+; 1H NMR δ (CDCl3) 9.11 (s, 1H), 8.78 (s, 1H), 8.38 (s,
1H), 8.03 (d, 2H, J ) 7.5 Hz), 7.61-7.27 (m, 8H), 6.45 (d, 1H, J )
2.7 Hz), 5.94 (t, 1H, J ) 2.4 Hz), 4.81-4.65 (m, 3H), 4.43-4.28 (m,
4H), 3.00 (s, 3H), 2.96 (s, 3H), 2.16 (s, 3H); 13C NMR δ (CDCl3)
169.5, 164.6, 152.9, 151.6, 149.7, 141.2, 135.7, 133.4, 132.8, 128.8,
128.8, 128.7, 128.6, 128.5, 128.4, 128.0, 127.8, 123.0, 87.3, 86.0, 81.3,
79.4, 73.3, 67.4, 65.5, 37.7, 37.6, 20.7.
solid material. Analytical data were identical with those reported above
for 16.
Synthesis, Deprotection, and Purification of Oligonucleotides. All
LNA and R-L-LNA oligomers were prepared by the phosphoramidite
approach as described earlier2a,4 on a Biosearch 8750 DNA synthesizer,
generally on CPG solid supports. The stepwise coupling efficiencies
for R-L-LNA phosphoramidite 6 (with 10 min coupling time) and
unmodified deoxynucleoside phosphoramidites (with 2 min standard
coupling time) were generally >99% with 1H-tetrazole as activator.
The stepwise coupling yield for synthesis of the 11-mix-mer r-L-
LNA-1 was 81-96% for amidite 23 (with 10 min coupling time). For
synthesis of the fully modified r-L-LNA-2, a universal CPG support
(BioGenex) was applied. The stepwise coupling yields obtained were
85-89% for amidite 23, 90-92% for amidite 9, and >99% for amidite
6 (with 10 min coupling time). For synthesis of the 16-mer r-L-LNA-
3, the stepwise coupling yields obtained were 98-99% for amidite 23
(with 10 min coupling time). The different stepwise coupling yields
obtained may be explained by slight variations in purity between the
phosphoramidite samples. After standard deprotection and cleavage
from the solid support by 32% aqueous ammonia (containing 1.5%
LiCl in the case of the universal support) r-L-LNA-1, r-L-LNA-2,
and r-L-LNA-3 were purified by DMT-ON reversed-phase chroma-
tography on disposable purification cartridges which includes detrity-
lation. The composition of the R-L-LNAs were confirmed by MALDI-
MS analysis and the purity (>90%) by capillary gel electrophoresis.
MALDI-MS analysis: r-L-LNA-1 [M - H]- 3405.8, calcd 3408.2;
r-L-LNA-2 [M - H]- 3636.2, calcd 3632.4; r-L-LNA-3 [M - H]-
5025.4, calcd 5026.2.
Thermal Denaturation Studies. The thermal denaturation studies
were carried out in a medium salt buffer solution (10 mM sodium
phosphate, 100 mM sodium chloride, and 0.1 mM EDTA, pH 7.0).
Concentrations of 1.5 mM of the two complementary strands were used,
with the assumption of identical extinction coefficients for modified
and unmodified oligonucleotides. The absorbance was monitored at
260 nm while the temperature was raised at a rate of 1 °C/min. The
melting temperatures (Tm values) of the duplexes were determined as
the maximum of the first derivatives of the melting curves obtained.
CD Experiments. The CD spectra were recorded on a Jasco J710
spectropolarimeter at room temperature in 0.5 cm cuvettes with the
same oligonucleotide concentrations and hybridization buffer solution
as those used for the thermal denaturation studies.
RNase H Experiments. The ribonucleotides T2 and T3 were
radiolabeled (32P) at the 5′-end with T4 polynucleotide kinase (Gibco-
BRL) and [γ-32P]adenosine 5′-triphosphate (ATP) (4500 Ci/mmol, ICN)
by standard procedures39 and purified on NAP-5 columns (Pharmacia).
Mixtures of the radiolabeled T3 (2.5 pmol) and either the complemen-
tary deoxyribonucleotide DNA-3 (5 pmol), r-L-LNA-3 (5 pmol), or
LNA-3 (5 pmol), as well as a mixture of the radiolabeled T2 (2.5 pmol)
and either the complementary deoxyribonucleotide DNA-1 (5 pmol),
r-L-LNA-1 (5 pmol), LNA-1 (5 pmol), or LNA-2 (5 pmol), were
incubated at 37 °C for 15 min in a total volume of 25 µL (pH 7.5)
containing 10 mM Tris-HCl, 25 mM KCl, and 0.5 mM MgCl2 to allow
the duplex to be formed. Cleavage reactions were started by the addition
of 10 units of RNase H (E. coli ribonuclease H, Life Technologies) to
the mixtures, while control tubes received no enzyme. Aliquots were
taken at appropriate time intervals, mixed with an equal volume of
stop mix (50 mM EDTA, 0.1% xylene cyanol FF, and 0.1% bromophe-
nol blue in 90% formamide), and chilled in a dry ice/acetone bath.
Samples were analyzed by denaturing 20% polyacrylamide gel elec-
trophoresis (PAGE) containing urea (50%) with TBE buffer39 at 1000
V for 1.25 h, followed by visualization by phosphorimaging (Packard
Cyclone, OptiQuant software).
9-(3-O-Benzyl-5-O-(methanesulfonyl)-4-C-[[(methanesulfonyl)-
oxy]methyl]-r-L-threo-pentofuranosyl)-6-N-benzoyladenine (28). Nu-
cleoside 27 (2.10 g, 3.00 mmol) was dissolved in a mixture of MeOH
(35 mL) and saturated methanolic ammnonia (15 mL). The mixture
was stirred at room temperature for 1.5 h and then evaporated to dryness
under reduced pressure. The residue was coevaporated with 96% EtOH
(50 mL) and purified by silica gel column chromatography with MeOH/
CH2Cl2 (1/40 v/v) as eluent, affording nucleoside 28 (1.56 g, 79%) as
a white solid material. FAB-MS m/z 648 [M + H]+; 1H NMR δ (CDCl3)
9.07 (s, 1H), 8.46 (s, 1H), 8.30 (s, 1H), 7.96 (d, 2H, J ) 6.9 Hz),
7.60-7.25 (m, 9H), 6.12 (d, 1H, J ) 2.4 Hz), 5.30 (d, 1H, J ) 2.4
Hz), 5.01 (dd, 1H, J ) 11.4, 2.4 Hz), 4.79 (dd, 1H, J ) 11.4, 2.7 Hz),
4.63 (dd, 1H, J ) 10.5, 2.7 Hz) 4.46-4.24 (m, 4H), 3.03 (s, 3H), 2.89
(s, 3H); 13C NMR δ (CDCl3) 165.0, 152.4, 151.2, 148.5, 142.0, 136.9,
133.3, 133.0, 129.0, 128.6, 128.5, 128.3, 127.8, 121.7, 87.4, 82.8, 81.8,
78.6, 73.2, 68.5, 68.3, 37.7, 37.4.
9-(2-O-Acetyl-3-O-benzyl-5-O-(methanesulfonyl)-4-C-[[(methane-
sulfonyl)oxy]methyl]-r-L-erythro-pentofuranosyl)-6-N-benzoylade-
nine (29). Nucleoside 28 (1.40 g, 2.16 mmol) was dissolved in a mixture
of anhydrous CH2Cl2 (50 mL) and anhydrous pyridine (10 mL). The
stirred solution was cooled to -30 °C and trifluoromethanesulfonic
anhydride (0.80 mL, 4.80 mmol) was added. After 1.5 h, the reaction
mixture was allowed to warm to 0 °C, additional trifluoromethane-
sulfonic anhydride (0.20 mL, 1.20 mmol) was added, and stirring at 0
°C was continued for 2.5 h. A saturated aqueous solution of NaHCO3
(10 mL) was added, followed by addition of CH2Cl2 (150 mL). The
organic phase was separated, washed with a saturated aqueous solution
of NaHCO3 (3 × 70 mL), dried (Na2SO4), and evaporated to dryness
under reduced pressure. The residue was dissolved in a mixture of
anhydrous toluene (100 mL) and anhydrous CH2Cl2 (20 mL), and KOAc
(1.20 g, 12.2 mmol) and 18-crown-6 (1.20 g, 4.50 mmol) were added
at room temperature under stirring. The temperature was raised to 50
°C and stirring was continued for 16 h. After the mixture was cooled
to room temperature, CH2Cl2 (200 mL) was added, and the resulting
mixture was washed with a saturated aqueous solution of NaHCO3 (3
× 90 mL), dried (Na2SO4), and evaporated to dryness under reduced
pressure. The residue was purified by silica gel column chromatography
with MeOH/CH2Cl2 (1-2% MeOH by volume) as eluent, affording
nucleoside 29 (1.05 g, 70%) as a white solid material. FAB-MS m/z
690 [M + H]+; 1H NMR δ (CDCl3) 9.17 (s, 1H), 8.74 (s, 1H), 8.57 (s,
1H), 8.04 (d, 2H, J ) 7.2 Hz), 7.61-7.26 (m, 9H), 6.74 (d, 1H, J )
3.9 Hz), 5.74 (s, 1H), 4.90-4.64 (m, 4H), 4.58 (d, 1H, J ) 10.8 Hz),
4.40 (d, 1H, J ) 10.8 Hz), 4.28 (d, 1H, J ) 10.8 Hz), 3.04 (s, 3H),
3.02 (s, 3H), 2.03 (s, 3H); 13C NMR δ (CDCl3) 168.6, 164.6, 152.8,
151.7, 149.5, 142.3, 135.9, 132.8, 128.8, 128.8, 128.7, 128.5, 128.4,
127.8, 122.4, 82.4, 82.3, 78.2, 74.6, 70.4, 68.0, 67.9, 38.0, 37.7, 20.5.
Alternative Synthesis of 16. Nucleoside 29 (1.00 g, 1.45 mmol)
was dissolved in a mixture of MeOH (35 mL) and saturated methanolic
ammnonia (15 mL). The mixture was stirred for 3.5 h at room
temperature and then evaporated to dryness under reduced pressure.
The residue was coevaporeted with EtOH (50 mL) and purified by
silica gel column chromatography with CH2Cl2/MeOH (2-4% MeOH
by volume) as eluent, yielding nucleoside 16 (0.80 g, 85%) as a white
Stability toward 3′-Exonucleolytic Degradation. A solution of the
oligonucleotides (∼0.2 OD) in 2 mL of a buffer (0.1 M Tris-HCl, pH
(39) Maniatis, T.; Fritsch, E.; Sambrook, J. Molecular Cloning: A Laboratory
Manual; Cold Spring Harbor Laboratory: Cold Spring Harbor, NY, 1989.
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2175

A R T I C L E S
Sørensen et al.
8.6, 0.1 M NaCl, and 14 mM MgCl2) was digested with 1.2 units of
snake venom phosphodiesterase [30 µL of a solution in 5 mM Tris-
HCl, pH 7.5, and 50% glycerol (by volume)] at 25 °C. The increase in
absorbance at 260 nm during digestion was followed.
Supporting Information Available: Description and experi-
mental details of the individual synthetic steps of the preliminary
conversion of furanose 26 into the guanine R-L-LNA nucleoside
35 and Scheme S1 illustrating this conversion; copies of the
13C NMR spectra of compounds 7, 8, 11, 14, 16, 19, 20-22,
26-32, 34, and 35; copies of selected NOE spectra of compound
20; and copies of the 31P NMR spectra of compounds 9 and 23
(PDF). This material is available free of charge via the Internet
at http:/pubs.acs.org.
Acknowledgment. The Danish Natural Science Research
Council, The Danish Technical Research Council, and Exiqon
A/S are acknowledged for financial support. Ms. Britta M. Dahl
is thanked for oligonucleotide synthesis, Dr. Michael Meldgaard,
Exiqon A/S, for MALDI-MS analysis, and Ms. Karen Jørgensen,
Department of Chemistry, University of Copenhagen, for CD
spectroscopy.
JA0168763
9
2176 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002