Oxepane nucleic acids: Synthesis, characterization, and properties of oligonucleotides bearing a seven-membered carbohydrate ring

Article abstract of DOI:10.1021/ja071336c

The synthesis and properties of oxepane nucleic acids (ONAs) are described. ONAs are sugar - phosphate oligomers in which the pentofuranose ring of DNA and RNA is replaced with a seven-membered (oxepane) sugar ring. The oxepane nucleoside monomers were pr

Full text of DOI:10.1021/ja071336c

Published on Web 06/09/2007
Oxepane Nucleic Acids: Synthesis, Characterization, and
Properties of Oligonucleotides Bearing a Seven-Membered
Carbohydrate Ring
David Sabatino and Masad J. Damha*
Contribution from the Department of Chemistry, McGill UniVersity, 801 Sherbrooke Street West,
Montreal, Quebec H3A 2K6, Canada
Received February 25, 2007; E-mail: masad.damha@mcgill.ca
Abstract: The synthesis and properties of oxepane nucleic acids (ONAs) are described. ONAs are sugar-
phosphate oligomers in which the pentofuranose ring of DNA and RNA is replaced with a seven-membered
(oxepane) sugar ring. The oxepane nucleoside monomers were prepared from the ring expansion reaction
of a cyclopropanated glycal, 1, and their conversion into phosphoramidite derivatives allowed efficient
assembly of ONAs on a solid support. ONAs (oT15 and oA15) were found to be much more resistant toward
nuclease degradation than natural DNA (dT15 and dA15) in fetal bovine serum (FBS) after 24 h of incubation
at 37 °C. ONAs also display several attributes in common with the naturally occurring DNA. For example,
oT15 exhibited cross-pairing with complementary RNA to give a duplex (oT15/rA15) whose conformation
evaluated by CD spectroscopy very closely matched that of the natural DNA/RNA hybrid (dT15/rA15).
Furthermore, oT15 was found to elicit Escherichia coli RNase H-mediated degradation of the rA15 strand.
When we compared the rates of RNase H-mediated degradation induced by 5- (furanose, dT15), 6- (2′-
enopyranose, pT18), and 7-membered (oxepane, oT15) ring oligonucleotides at a temperature that ensures
maximum duplex population (10 °C), the following trend was observed: dT15 . oT15 > pT18. The wider
implications of these results are discussed in the context of our current understanding of the catalytic
mechanism of the enzyme. The homopolymer oT15 also paired with its oxepane complement, oA15, to form
a duplex structure that was different [as assessed by circular dichroic (CD) spectroscopy] and of lower
thermal stability relative to the native dT15/dA15 hybrid. Hence, ONAs are useful tools for biological studies
and provide new insights into the structure and function of natural and alternative genetic systems.
Introduction
host duplex, leading to very efficient RNase H-mediated
cleavage of the complementary RNA. These studies indicated
Oligonucleotide (ON) analogues are important for purposes
of structural studies as well as potential therapeutic agents for
gene silencing applications associated with antisense-,1 antigene-,2
RNAi-,3 and aptamer4 based modalities. In the optimal case for
antisense ONs gene silencing occurs catalytically in concert with
a constitutive ribonuclease H (RNase H) enzyme capable of
recognizing and understanding the mechanisms of substrate
recognition by RNase H provides guidance to the design of
future ONs for therapeutic use.5
that enhanced flexibility of the ON backbone enables better
alignment of the ON/RNA hybrid into the active site of RNase
H, thereby conferring accelerated enzyme-catalyzed degradation
of RNA.7,8 To further test this hypothesis, we turned our
attention to ON backbones that lack the furanose ring altogether
but that would be flexible enough to adjust to the required
duplex trajectory for RNase H recognition and cleavage.
Herein we describe the synthesis of nucleosides and oligo-
nucleotides containing unsaturated 6-membered and saturated
7-membered ring structures, termed 2′-enopyranose and oxepane
nucleic acids (ONA), respectively.9 The ONAs (e.g., oT15) were
compared with native DNA (e.g., dT15) and unsaturated
6-membered (e.g., pT18) ring structures in terms of their ability
to self-associate (e.g., oT15/oA15), cross-pair with native struc-
tures (e.g., oT15/dA15 and oT15/rA15), and elicit RNase H-
We recently synthesized chimeric ONs composed of 2′-deoxy-
2′-fluoro-â-D-arabinose or 2′-â-D-deoxyribose sugars with in-
terspersed acyclic (e.g., 2′,3′-seconucleotide or butyl) residues.6
These compounds combined both preorganization (fluoroara-
binose moiety) and flexibility (acyclic linker moiety) within the
(1) Minshull, J.; Hunt, T. Nucleic Acids Res. 1986, 14, 6433-6451.
(2) Le Doan, T.; Perrouault, L.; Praseuth, D.; Habhoub, N.; Decout, J. L.;
Thoung, N. T.; Lhomme, J.; Helene, C. Nucleic Acids Res. 1987, 15, 7749-
7760.
(3) Sontheimer, E. J.; Carthew, R. W. Science 2004, 305, 1409-1410.
(4) Shoji, A.; Kuwahara, M.; Ozaki, H.; Sawai, H. J. Am. Chem. Soc. 2007,
129 (5), 1456-1464.
(5) Itaya, M.; Kondo, K. Nucleic Acids Res. 1991, 19, 4443-4449.
(6) Mangos, M. M.; Min, K.-L.; Viazovkina, E.; Galarneau, A.; Elzagheid,
M. I.; Parniak, M. A.; Damha, M. J. J. Am. Chem. Soc. 2003, 125 (3),
654-661.
(7) (a) Mangos, M. M.; Damha, M. J. Curr. Top. Med. Chem. 2002, 2 (10),
1147-1171. (b) Yazbeck, D. R.; Min, K.-L.; Damha, M. J. Nucleic Acids
Res. 2002, 30 (14), 3015-3025.
(8) Verbeure, B.; Lescrinier, E.; Wang, J.; Herdewijn, P. Nucleic Acids Res.
2001, 29 (24), 4941-4947.
(9) We had previously attempted to synthesize unsaturated heptose (i.e.,
oxepine) nucleic acids; however, isolation of the oligonucleotides was
prevented by degradation of the oxepine-modified oligonucleotide during
alkaline cleavage of the oligonucelotide from the succinyl linker solid
support (D. Sabatino and M. J. Damha, unpublished results).
9
10.1021/ja071336c CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 8259-8270
8259

A R T I C L E S
Sabatino and Damha
pairing than DNA, but it does not cross-pair with native DNA
or RNA strands. The reason for this tighter base pairing was
postulated to be a consequence of the higher rigidity of a
pyranose compared to a furanose ring, resulting in a preorga-
nization of the single strand’s backbone conformation favoring
a compact quasi-linear duplex.14,15 Since then, more flexible
unsaturated six-membered ring nucleic acids (e.g., cyclohexene
NA) have been developed to promote binding to the natural
systems in hybrid conformations that mimic the native du-
plexes.16 Alternatively, contracting the carbohydrate moiety to
a four-carbon skeleton, as in the R-L-threofuranosyl sugar,
maintains a furanose half-chair sugar pucker conformation
similar to native systems, allowing R-L-threofuranosyl nucleic
acids, TNAs, to pair with its complement as well as to single-
stranded ssDNA and ssRNA.17,18 Here we report the synthesis
and characterization of ONs bearing a 6- or 7-membered
carbohydrate ring in place of the normal 5-membered ribofura-
nose ring.9 We surmised that these relatively unconstrained ONs
would exhibit RNase H induction properties as well as provide
some insight toward elucidating the structural factors that
provide the “optimal” ON/RNA substrate of the enzyme.7,8
Conformational Analysis. Molecular modeling of nucleoside
conformation is commonly used to study carbohydrate geometry
in duplex structures.19 To gain insight into the conformation of
oxepane and 2′-enopyranose oligomers, we conducted compara-
tive computational analyses of the thymine nucleosides, oT and
pT, along with the native nucleosides dT and rU (Figure 1).
Unsaturated 6-membered rings have been shown to be most
stable in their twisted half-chair conformations with barriers to
interconversion of 8-12 kcal/mol, oscillating predominantly via
the boat (bent) conformation.12,20 The nucleoside pT favors a
°H5′ half-chair conformation such that the dihedral angle about
the C4′-C5′ bond in the 2′-enopyranose (δ ) 85°) falls within
the range of ribose in A-RNA-type helices (C3′-C4′, δ ) 77°-
97°) (Figure 1).21 The pseudorotational equilibrium for oxepane
carbohydrates is more complex and involves many low-energy
transitions between the boat (B), twist boat (TB), chair (C), and
the favored twist chair (TC) forms.11,22 The most common
transition from C to TC geometry proceeds with relatively low
interconversion energy barriers of 0.5-2 kcal/mol, whereas
some of the less populated TC to B or B to TB transitions have
interconversion energy barriers of 5-10 kcal/mol.11 An ab initio
geometry optimization was performed on oT nucleoside, 6, in
vacuo, with the 3-21 G basis set using HyperChem version 7.0.
The Polak-Ribiere conjugate gradient algorithm was used, with
a root-mean-square gradient termination cutoff of 0.05 kcal Å-1
mol-1. The calculations provided compact energy-minimized
structures where the heptose carbohydrate was found to be either
Figure 1. Structures and conformations of nucleosides.
mediated cleavage of their target RNA (rA15). We show that
the trend relating flexibility and RNase H activity persists at a
temperature ensuring a maximum duplex population (10 °C)
when we compare the RNase H activity of 5- (furanose, dT15),
6- (2′-enopyranose, pT18), and 7-membered (oxepane, oT15) ring
oligonucleotides with the flexibility of the constituent sugar rings
(i.e., furanose10 (5) > oxepane11 (7) > 2′-enopyranose12 (6)
(Figure 1)]. In addition, oT15 and oA15 were found to be com-
pletely stable in fetal bovine serum after 24 h of incubation at
37 °C. These findings are significant in that they contribute to
our knowledge about what types of chemical modifications are
tolerated in an oligonucleotide strand for the development of
RNase H-active, nuclease-stable ONs and their antisense ap-
plications.
Results and Discussion
To date, the largest expansion of the sugar moiety in DNA
is that reported for six-membered ring hexopyranosyl nucleic
acids and their assembly into (4′ f 6′) linked oligo(2′,3′-
dideoxy-â-D-glucopyranosyl)nucleotides (i.e., “homo-DNA”).13,14
This architecture exhibits much stronger Watson-Crick base
(15) Lescrinier, E.; Froeyen, M.; Herdewijn, P. Nucleic Acids Res. 2003, 31
(12), 2975-2989.
(16) Wang, J.; Verbeure, B.; Luyten, I.; Lescrinire, E.; Froeyen, M.; Hendrix,
C.; Rosemeyer, H.; Seela, F.; Van Aershot, A.; Herdewijn, P. J. Am. Chem.
Soc. 2000, 122, 8595-8602.
(17) Schoning, K.-U.; Scholz, P.; Guntha, S.; Wu, X.; Krishnamurthy, R.;
Eschenmoser, A. Science 2000, 290 (5495), 1347-1351.
(18) Wilds, C. J.; Wawrzak, Z.; Krishnamurthy, R.; Eschenmoser, A.; Egli, M.
J. Am. Chem. Soc. 2002, 124, 13716-13721.
(10) Gabb, H. A.; Harvey, S. C. J. Am. Chem. Soc. 1993, 115, 4218-4227.
(11) Espinosa, A.; Gallo, M. A.; Entrena, A.; Gomez, J. A. J. Mol. Struct. 1994,
323, 247-256.
(12) Choo, J.; Lee, S.-N.; Lee, K.-H. Bull. Kor. Chem. Soc. 1996, 17 (1), 7-11.
(13) (a) Augustyns, K.; Vandendriessche, F.; Van Aerschot, A.; Busson, R.;
Urbanke, C.; Herdewijn, P. Nucl. Acids Res. 1992, 20, 4711-4716. (b)
Augustyns, K.; Van Aerschot, A.; Urbanke, C.; Herdewijn, P. Bull. Soc.
Chim. Belg. 1992, 101, 119-130. (c) Hunzinker, von J.; Roth, H-J.;
Bohringer, M.; Giger, A.; Diederichsen, U.; Gobel, M.; Krishnan, R.; Jaun,
B.; Leumann, C.; Eschenmoser, A. HelV. Chim. Acta. 1993, 76, 259-352.
(14) Eschenmoser, A. Science 1999, 284, 2118-2124.
(19) Noy, A.; Perez, A.; Marquez, M.; Luque, J.; Orozco, M. J. Am. Chem.
Soc. 2005, 127, 4910-4920.
(20) (a) Ferrier, R. J.; Sankey, G. H. J. Chem. Soc. C 1966, 2345-2349. (b)
Laane, J.; Choo, J. J. Am. Chem. Soc. 1994, 116, 3889-3891.
(21) Felder, E.; Gattlen, R.; Ossola, F.; Baschang, G. Nucleosides Nucleotides
1992, 11 (9), 1667-1671.
(22) DeMatteo, M. P.; Snyder, N. L.; Morton, M.; Baldisseri, D. M.; Hadad, C.
M.; Peczuh, M. W. J. Org. Chem. 2005, 70, 24-38.
9
8260 J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007

Oxepane Nucleic Acids
A R T I C L E S
Scheme 1. Synthesis of Oxepane Nucleosides and Phosphoramiditesa
a Conditions and reagents: (i) persilylated base, TMSOTf, MeCN, reflux, 12-24 h, Thy 40% and AdeNBz 45%; (ii) 1 M TBAF/THF, 0 °C, 1 h, Thy 90%
and AdeNBz 61%; (iii) 1 atm H2, Pd/C, MeOH, 22 °C, 4 h, Thy 70% and AdeNBz 99%; (iv) MMTCl, pyridine, 22 °C, 3 h, Thy 66% and AdeNBz 50%; (v)
ClP(OCEt)N(i-Pr)2, EtN(i-Pr)2, THF, 22 °C, 2 h, Thy 80% and AdeNBz 90%.
Table 1. Selectivity of the Glycosylation Reaction of 1
in a chair or slightly twisted chair conformation (O′6′TC4′5′), with
C4′ and C5′ lying below and the ring oxygen and C6′ above
the plane of the molecule defined by C1′-C2′-C3′ (Figure 1).
Comparison of the oT conformation with the C2′-endo23 sugar
pucker of dT (Figure 1) revealed an expanded oxepane
carbohydrate structure in which the in-ring diameter of oT is
extended relative to the pentofuranose ring by 0.8 Å (i.e., for
oT, C1′-C5′ ) 3.11 Å, and for dT, C1′-C4′ ) 2.35 Å), while
the O-O distance in oT was found to be reduced by almost
0.8 Å (i.e., for oT, C7′OH-C5′OH ) 3.57 Å, and for dT,
C5′OH-C3′OH ) 4.26 Å). Therefore, from this perspective,
the oxepane modification provides a more compact oligonucle-
otide structure that can, in principle, hybridize to DNA and,
even more favorably, to the more compact C3′-endo23 geometry
of RNA (i.e., for oT, C7′OH-C5′OH ) 3.35 Å, and for rU,
side, 2, along with the diene byproduct, 3. The product
distribution for the reaction was found to be dependent on the
C5′OH-C3′OH ) 3.80 Å) (Figure 1). However, two factors
may limit the base-pairing potential of ONA with complemen-
tary DNA and RNA: (a) the increased ring size of the oxepane
ring, as observed with pyranose-based ONs,14,24 and (b) the
relative orientation of the glycosidic bond (C1′-N1) with respect
to the C5′-OH bond (Figure 1).25
Synthesis and Characterization of Nucleosides and Oli-
gonucleotides. The oxepane nucleosides, oT and oA, were
prepared by a procedure similar to the ring expansion reaction
of cyclopropanated glycal sugars reported by Hoberg26 (Scheme
1). We have expanded the scope of this reaction to nucleoside
synthesis by introducing silylated thymine and N6-benzoylad-
enine as nucleophiles. Thus, Vorbru¨ggen-type27 glycosylation
reactions with sugar 1 afforded the unsaturated oxepine nucleo-
potency of the nucleophile (Table 1). Coupling with N6-
benzoyladenine proceeded more rapidly and more efficiently
than with thymine (0.5 day vs 1 day reflux) at the expense of
a poorer diastereoselectivity (2:1 â:R for N6-benzoyladenine and
10:1 â:R for thymine). The â- and R-anomeric configurations
of the nucleosides were unequivocally established by NOESY
experiments (see Supporting Information Figure S16). Following
desilylation and reduction of the double bond, the oxepane 5′-
O-phosphoramidite derivatives 10 and 11 were prepared and
purified by standard methods (Scheme 1). While the dimethox-
ytrityl protecting group is more commonly used in ON synthesis,
we chose the 5′-MMT group as it is more stable and, in our
hands provides higher yield of nucleosides.
The synthesis of the unsaturated pyranose nucleoside, pT,
and its corresponding phosphoramidite was accomplished with
a slightly modified procedure,28 starting from commercially
available tri-O-acetyl-D-glucal (Scheme 2). The reaction of
silylated thymine with tri-O-acetyl-D-glucal in the presence of
Lewis acid catalyst TMSOTf proceeded to completion by
refluxing in MeCN for 3 h.
(23) (a) Pauling, L.; Corey, R. B. Proc. Natl. Acad. Sci. U.S.A. 1953, 39 (2),
84-97. (b) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen,
A. G.; Taylor, R. J. Chem. Soc., Perkin Trans. 2 1987, S1-S19. (c) Gelbin,
A.; Schneider, B.; Clowney, L.; Hsieh, S.-H.; Olson, W. K.; Berman, H.
M. J. Am. Chem. Soc. 1996, 118, 519-529.
(24) Kerremans, L.; Schepers, G.; Rozenski, J.; Busson, R.; Van Aerschot, A.;
Herdewijn, P. Org. Lett. 2001, 3 (26), 4129-4132.
(25) Premraj, B. J.; Yathindra, S. R. N. Biophys. Chem. 2002, 95, 253-272.
(26) Hoberg, J. O. J. Org. Chem. 1997, 62 (19), 6615-6618.
(27) Skalmowski, B. B.; Krolikiewicz, K.; Vorbruggen, H. Tetrahedron Lett.
1995, 36 (43), 7845-7848.
(28) Pedersen, H.; Perdersen, E. K. Heterocylces 1992, 34 (2), 265-272.
9
J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007 8261

A R T I C L E S
Sabatino and Damha
Scheme 2. Synthesis of 2′-Enopyranosyl Nucleoside O-Phosphoramiditesa
a Reagents and conditions: (i) bis(trimethylsilyl)thymine, TMSOTf, MeCN, reflux, 3 h (75%); (ii) 0.1 M NaOMe/MeOH 1 h, (93%); (iii) MMTCl,
pyridine, 22 °C, 3 h (78%); (iv) ClP(OCEt)N(i-Pr)2, EtN(i-Pr)2, THF, 22 °C, 1 h (80%).
Table 2. Comparison of UV Thermal Melt and Percent
Hyperchromicity Values for Pairing and Cross-Pairing of
The reaction after purification yielded 75% 13 in a separable
1.1:1 ratio in R:â diastereoselectivity at the anomeric C1′-
2′-Enopyranose (pT) and Oxepane (oT and oA) Oligonucleotides
position. The structures of 13a and 13b were confirmed by 1H
and 13C NMR, and the chemical shift data obtained were in
complete agreement with the literature data.29 Nucleoside 13a
was subsequently treated with 0.1 M NaOMe/MeOH to yield
the completely deprotected nucleoside 14 in a °H5′ half-chair
conformation as identified by the characteristic coupling
constants, that is, J1′2′ ∼ 1 Hz and J4′5′ ) 6.8 Hz (see
Experimental Section).29 The nucleoside was monomethoxytri-
tylated at the primary C6′ hydroxyl and then treated with Cl-
P(OCEt)N(i-Pr)2 to yield the phosphoramidite 16, suitable for
solid-phase oligonucleotide synthesis.
and Control DNA (dT and dA) and RNA (rU and rA) Sequencesa
sequences
Tm
(°
C)
% H
1. oT15 + oA15
2. oT15 + dA15
3. oT15 + rA15
4. dT15 + oA15
5. rU15 + oA15
6. pT18 + dA18
7. pT18 + rA18
8. dT15 + dA15
9. dT15 + rA15
10. rU15 + rA15
11. rU15 + dA15
12b
<5
13b
<5
12b
34b
25b
37
4.5
8.7
11.3
3.3
3.4
11.3
15.3
24.8
25.4
28.7
20.7
32
25
16
The homopolymeric oT15, oA15, and pT18 strands were
assembled on a commercially available 500 Å long-chain
alkylamino Unylinker CPG support.30 With a limited supply of
oT and oA monomers in hand, phosphoramidites 10 and 11
were used as 0.05 M solutions (in anhydrous CH2Cl2 for 10
and in anhydrous MeCN for the more polar nucleoside 11).
Oligonucleotide syntheses were conducted on a 0.5 µmol scale
with extended amidite coupling times (30 min) and 5-ethyl-
thiotetrazole, ETT, as the activator (0.25 M in MeCN). The
detritylation step was extended to 2.5 min in order to ensure
complete removal of the monomethoxytrityl (MMT) group. All
other sequences were synthesized by conventional procedures.31
The desired oligonucleotides (i.e., oT15, oA15, and pT18)
constituted 65-70% of the crude material isolated after depro-
tection (see Supporting Information Figures S1 and S2),
indicating that under the conditions used, the monomers coupled
with 98-99% efficiency. Recoveries of oT15 and oA15 from
the solid support were ca. 25 OD units (25-40% yield).
Following purification by HPLC and/or denaturing PAGE
conditions, the oligomers were desalted by size-exclusion
a Duplex concentration of 3.04 µM in sodium phosphate buffer: 140
mM KCl, 1 mM MgCl2, and 5 mM Na2HPO4 adjusted to pH ) 7. b Early
and/or broad transition curve.
chromatography (Sephadex G-25). The structure of the oligo-
nucleotides was confirmed by MALDI-TOF mass spectrometry
(see Supporting Information Table S1).32
Hybridization Properties of Oxepane-Modified Oligo-
nucleotides. 1. oT15/oA15 Duplex. Hybridization of oligonucleo-
tide strands was first assessed by UV-monitored thermal
denaturation experiments; phosphate buffers typically contained
1 mM Mg2+ and 5 mM Na+ (see Experimental Section) and
were maintained at pH ) 7 (Table 2 lists the sequences that
were investigated). As anticipated from previous studies,33,40
the control duplexes dT15/dA15, rU15/rA15, dT15/rA15, and rU15/
dA15 all exhibited a single-phase transition with 20-30%
hyperchromicity (Table 2). By contrast, binding experiments
on oT15/oA15 showed a weak but clearly detectable early
transition curve (Tm ) 12 °C) (Figure 2B). In addition, changes
in hyperchromicity (% H), were quite small, suggesting the
formation of a weaker complex relative to the native duplex
(29) (a) Herscovici, J.; Montserret, R.; Antonakis, K. Carbohydr. Res. 1988,
176, 219-229. (b) Doboszewski, B.; Blaton, N.; Herdewijn, P. J. Org.
Chem. 1995, 60, 7909-7919.
(30) Guzaev, A. P.; Manoharan, M. J. Am. Chem. Soc. 2003, 125, 2380-2381.
(31) (a) Caruthers, M. H.; Barone, A. D.; Beaucage, S. L.; Dodds, D. R.; Fisher,
E. F.; McBride, L. J.; Mateucci, M.; Stabinsky, Z.; Tang, Y.-J. Methods
Enzymol. 1987, 154, 287-313. (b) Damha, M. J.; Ogilvie, K. K. Methods
Mol. Biol. 1993, 20, 81-114.
(32) Distler, A. M.; Allison, J. Anal. Chem. 2001, 73, 5000-5003.
(33) (a) Kim, K.; Jhon, M. S. Biochim. Biophys. Acta 1979, 565, 131-147. (b)
Aymaan, J.; Coll, M.; Frederick, C. A.; Wang, A. H.-J.; Rich, A. Nucleic
Acids Res. 1989, 17, 3229-3245. (c) Hall, K. B.; McLaughin, L. W.
Biochemistry 1991, 30, 10606-10613. (d) Ratmeyer, L.; Vinayak, R.;
Zhong, Y. Y.; Zon, G.; Wilson, W. D. Biochemistry 1994, 33, 5298-5304.
9
8262 J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007

Oxepane Nucleic Acids
A R T I C L E S
Figure 2. Comparison of Tm plots for (A) dT15/dA15 and (B) oT15/oA15 in addition to their single-stranded oligonucleotides. Duplex concentration was 3.04
µM in sodium phosphate buffer: 140 mM KCl, 1 mM MgCl2, and 5 mM Na2HPO4 adjusted to pH ) 7.
Figure 3. UV mixing curves at 5 °C for (A) oT15/oA15 and (B) dT15/dA15. Concentration of each oligonucleotide stock solution was 5 µM, in duplex
binding buffer: 10 mM Na2HPO4 and 100 mM NaCl, pH ) 7.2.
Figure 4. Temperature-dependent CD curves and representative plots for the change in molar ellipticity with temperature at λmax 248 nm or λmin 265 nm
for oT15/oA15. (Inset)Percent change in molar ellipticity as a function of temperature at 265 nm and hybrid thermal denaturation curve. Concentration of
duplex was 3.04 µM in sodium phosphate buffer: 140 mM KCl, 1 mM MgCl2, and 5 mM Na2HPO4, pH ) 7.
dA15/dT15 (Tm ) 37 °C) (Figure 2A; Table 2). Experiments with
the individual strands alone showed, in the case of dA15 and
oA15, a slight increase in absorbance with temperature (Figure
2), consistent with a good stacking propensity observed for
purine strands. This was not the case for oT15, however, which
displayed a decrease in absorbance with temperature (Figure
2B). Since the hyperchromicity change in the UV melting curve
of oT15/oA15 was small, more experiments were needed to
evaluate whether the putative oT15/oA15 duplex was formed.
Data from the UV mixing (Job Plot)34 studies were more
conclusive, strongly suggesting that oT15 and oA15 base-pair
into a helical complex with 1:1 stoichiometry at 5 °C (Figure
3A). The native dT15/dA15 duplex also showed, as expected,
UV mixing data consistent with a 1:1 complex (Figure 3B).
As another test for oT15/oA15 complexation, we performed
CD-monitored melting experiments (Figure 4). CD is particu-
larly well suited for monitoring the melting of duplexes,35 as
single-phase transition from helix to coil state is often associated
with the largest change in amplitude in the Cotton effect (molar
ellipticity) at a given wavelength.36 Samples were prepared as
described above and the CD spectra were normalized against a
blank solution containing the phosphate buffer. A parallel
(35) Chan, S. S.; Breslauer, K. K.; Hogan, M. E.; Kessler, D. J.; Austin, R. H.;
Ojemann, J.; Passner, J. M.; Wiles, N. C. Biochemistry 1990, 29, 6161-
6171.
(34) Pilch, D. S.; Levenson, C.; Shafer, R. H. Proc. Natl. Acad. Sci. U.S.A.
1990, 87, 1942-1946.
(36) Gray, D. M.; Hung, S.-H.; John, K. H. Methods Enzymol. 1995, 246, 19-
34.
9
J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007 8263

A R T I C L E S
Sabatino and Damha
Figure 6. UV mixing curves at 5 °C for (A) oT15/rA15 and (B) dT/rA15.
Concentration of each oligonucleotide stock solution was 5 µM in duplex
binding buffer: 10 mM Na2HPO4 and 100 mM NaCl, pH ) 7.2.
Figure 5. CD spectral signatures of the single-stranded and duplex
oligonucleotides for (A) oT15/oA15 and dT15/dA15 duplexes and (B) oT15,
oA15, dT15, and dA15 single strands. Duplex concentration was 3.04 µM
and single strand concentration was 1.52 µM in 140 mM KCl, 1 mM MgCl2,
and 5 mM Na2HPO4, pH ) 7.
agreement with the Tm values estimated from the temperature-
dependent CD experiments (Figure 7 and Supporting Informa-
tion Figure S7). The corresponding native hybrids dT15/rA15 and
dA15/rU15 had melting temperatures of 32 and 16 °C, respec-
tively (Table 2).33,40
No association (Tm < 5 °C) was detected between the oxepane
oligomers and their complementary ssDNA strands (UV and
CD thermal denaturation experiments). Consistent with this
notion, the CD spectrum of oT15/dA15 exhibited only a modest
change in amplitude of the molar ellipticity with respect to the
single strands, and the CD-monitored melting curve of the oT15
+ dA15 mixture produced only a broad early transition of less
than 5 °C (see Supporting Information Figure S8).
3. oT15/rA15 and dT15/rA15 Form an A-like Helix. The CD
profiles of oT15/rA15 and the native dT15/rA15 systems are
characteristic of A-like helices,37 both displaying a strong
positive band at 265 nm, a strong negative band near 248 nm,
and a crossover point at around 254 nm. These similarities can
be ascribed to the conformation of the underlying rA15 strand,
which greatly influences the CD spectrum (Figure 8A). Given
that the CD trace of the oT15 single strand alone (Figure 8B) is
so different from that of rA15 and the oT15/rA15 duplex, we
conclude that oT15 readily adjusts to the dominant conformation
of the purine RNA strand (rA15). Compared to dT15 and rU15,
the oxepane oT15 single strand exhibited a unique spectral
signature with small negative bands and no crossover points
(Figure 8B). However, pairing between oT15 and rA15 provided
a 1:1 complex at 5 °C (Figure 6) whose helical conformation
very closely resembled that of the dT15/rA15 hybrid (Figure 8A).
Escherichia coli RNase H-Mediated Degradation of oT/
rA and pT/rA Hybrids. The RNase H family comprises a class
of enzymes that recognize and cleave the RNA strand of RNA-
DNA heteroduplexes having topologies that are intermediate
between the pure A- or B-form helices adopted by dsRNA and
dsDNA, respectively.41 Among the expanding list of sugar-
modified ONs that elicit RNase H-mediated cleavage of RNA
experiment was conducted with the native dT15/dA15 duplex (see
Supporting Information Figure S4) and the individual component
strands (Figure 5B). The data obtained provided additional
evidence for the interaction between oT15 and oA15. Specifically,
the CD spectra of the mixed oligomers oT15 and oA15 showed
clear differences from those of the single strands (Figure 5),
and a plot of the change in molar ellipticity as a function of
temperature at 265 nm for oT15/oA15 produced a sigmoidal
hyperchromic transition from which a Tm value of ca. 12 °C
was determined (Figure 4). This was in complete agreement
with the Tm value obtained from UV melting measurements.
Of note, oT15/oA15 exhibited a strikingly different CD profile
from dT15/dA15 (Figure 5A). The dT15/dA15 hybrid exhibited
the expected CD profile for a B-form helix, including a negative
peak at 248 nm and a positive one at 278 nm.37 By contrast,
the oT15/oA15 system displayed a nearly opposite signature that
was characterized by negative minima at 265 and 220 nm and
a negative maximum at 248 nm, suggesting the presence of a
different helical form (Figure 5A). While the trace observed
may be suggestive of a left-handed helix,38 NMR and/or X-ray
crystallographic studies will be required to conclusively establish
the oT15/oA15 duplex structure.39
2. Oxepane Nucleic Acid Cross-Pairs with RNA but Not
Single-Stranded DNA. The single strands oT15 and oA15 were
found to associate with their respective complementary RNA
strands in 1:1 ratios to form oT15/rA15 (Tm ) 13 °C) and oA15/
rU15 (Tm ) 12 °C) hybrids, respectively (Figure 6; Table 2 and
Supporting Information Figure S3). These values were in
(37) Steely, H. T.; Gray, D. M.; Ratliff, R. L. Nucleic Acids Res. 1986, 14,
10071-10090.
(38) (a) Wang, A. H.-J.; Quigley, G. J.; Kolpak, F. J.; Crawford, J. L.; van
Boom, J. H.; van der Marel, G.; Rich, A. Nature 1979, 282, 680-686. (b)
Nejedly, K.; Klysik, J.; Palecek, E. FEBS Lett. 1989, 243 (2), 313-317.
(c) Kagawa, T. F.; Howell, M. L.; Tseng, K.; Ho, P. S. Nucleic Acids Res.
1993, 21 (25), 5978-5986.
(39) Tomasz, M.; Barton, J. K.; Magliozzo, C. C.; Tucker, D.; Lafer, E. M.;
Stollar, B. D. Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 2874-2878.
(40) Martin, F. H.; Tinoco, I. Nucleic Acids Res. 1980, 8 (10), 2295-2299.
9
8264 J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007

Oxepane Nucleic Acids
A R T I C L E S
Figure 7. Temperature-dependent CD curves and representative plots for the change in molar ellipticity with temperature at λmin 248 nm or λmax 265 nm
for oT15/rA15. (Inset) Percent change in molar ellipticity as a function of temperature at λmax 265 nm and hybrid thermal denaturation curve. Concentration
of duplex was 3.04 µM in sodium phosphate buffer: 140 mM KCl, 1 mM MgCl2, and 5 mM Na2HPO4 pH ) 7.
strand is critical as well.45 A rigid ON seems to be unable to
distort to allow its RNA complement to assume the conformation
necessary for hydrolysis.46,48
The study of other carbohydrate systems, including non-
furanose sugars, allows us to extend our understanding of the
impact of backbone flexibility on RNase H activity.6,7a,8,16 The
insertion of a double bond in cyclohexane nucleic acids (i.e.,
cyclohexene NAs) is necessary for cross-pairing with RNA and
activation of the RNase H enzyme.8,16,47 To test whether the
same effect operates in a pyranose system, we prepared a variant
of the homo-DNA structure, the 2′-enopyranosyl modification
(pT18) previously reported by Felder et al.21 The oligomer we
prepared (pT18) was capable of cross-pairing with the native
rA18 complement strand (Table 1, Tm ∼ 25 °C) and displayed
a global helical conformation that resembled that of the native
dT18/rA18 substrate (see CD spectra, Supporting Information
Figure S11). The oxepane and 2′-enopyranose backbones thus
provide ring structures that are rendered more flexible than
homo-DNA13,14 and [3.3.0]bc-ANA45 and conceivably are more
likely to engage in interactions (i.e., H-bonding and base
Figure 8. CD spectral signatures of single-stranded and duplex oligo-
nucleotides for (A) oT15/rA15 and dT15/rA15 duplexes and (B) oT15, rA15,
stacking) with the opposing RNA strand of the formed ON/
RNA heteroduplex. This is demonstrated in Figure 9, which
shows that both oT15/rA15 and pT18/rA18 support detectable
cleavage by E. coli RNase H. Significantly less hydrolysis occurs
for oT15/rA15 when compared with the native substrate, which
can be rationalized, at least in part, by the lower affinity of
oT15 for the RNA target (Tm ∼ 13 °C). This property acts to
present a lower effective concentration of substrate duplex to
the enzyme, thereby diminishing the overall rate of catalysis.
As we were interested in the relative rates of cleavage induced
by the different ONs, this information was extracted from the
assay conducted at 10 °C, at a temperature that ensures the
highest hybrid population for these systems and under which
enzyme activity is retarded just enough to enable a comparison
and dT15 single strands. Duplex concentration was 3.04 µM and single strand
concentration was 1.52 µM in 140 mM KCl, 1 mM MgCl2, and 5 mM
Na2HPO4, pH ) 7.
are arabino- (ANA) and 2′-deoxy-2′-fluoro-â-D-arabinonucleic
acids (2′-F-ANA),42 cyclohexene nucleic acids (CeNA),8,16 and
R-L-ribo-configured locked nucleic acids (R-L-LNA).43 Chemical
changes of the furanose or alterations in the orientation of the
furanose to the base can dramatically diminish RNase H
activation.25,44 The discovery that 2′-arabino-configured locked
bicyclonucleotide [3.3.0]bc-ANA satisfies the conformational
requirements for RNase H recognition and cleavage without
activating this enzyme suggests that the flexibility of the ON
(41) (a) Uchiyama, Y.; Miura, Y.; Inoue, H.; Ohtsuka, E.; Ueno, Y.; Ikehara,
M.; Iwai, S. J. Mol. Biol. 1994, 243, 782-791. (b) Nishizaki, T.; Iwai, S.;
Ohtsuka, E.; Nakamura, H. Biochemistry 1997, 36, 2577-2585.
(42) (a) Damha, M. J.; Wilds, C. J.; Noronha, A.; Brukner, I.; Borkow, G.;
Arion, D.; Parniak, M. A. J. Am. Chem. Soc. 1998, 120, 12976-12977.
(b) Noronha, A. M.; Wilds, C. J.; Lok, C. N.; Viazovkina, K.; Arion, D.;
Parniak, M. A.; Damha, M. J. Biochemistry 2000, 39, 7050-7062.
(43) Sorensen, M. D.; Kvaerno, L.; Bryld, T.; Hakansson, A. E.; Verbeure, B.;
Gaubert, G.; Herdewijn, P.; Wengel, J. J. Am. Chem. Soc. 2002, 124, 2164-
2176.
(45) Nielsen, P.; Pfundheller, H. M.; Wengel, J. Chem. Commun. 1997, 825-
826.
(46) (a) Oda, Y.; Iwai, S.; Ohtsuka, E.; Ishikawa, M.; Ikehara, M.; Nakamura,
H. Nucleic Acids Res. 1993, 21, 4690-4695. (b) Li, J.; Wartell, R. M.
Biochemistry 1998, 37, 5154-5161.
(47) Maurinsh, Y.; Rosemeyer, H.; Esnouf, R.; Medvedovici, A.; Wang, J.;
Ceulemans, G.; Lescrinier, E.; Hendrix, C.; Busson, R.; Sandra, P.; Seela,
F.; Van Aerschot, A.; Herdewijn, P. Chem.sEur. J. 1999, 5 (7), 2139-
2150.
(44) Baker, B. F.; Monia, B. P. Biochim. Biophys. Acta 1999, 1489, 3-18.
9
J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007 8265

A R T I C L E S
Sabatino and Damha
Figure 9. E. coli RNase H-mediated degradation of (A) 2′-OMe rU18/rA18, pT18/rA18, dT15/rA15, and oT15/rA15 duplexes at 10 °C, for 0, 20, and 40 min and
1, 2, and 4 h; (B) 2′-OMe rU18/rA18, pT18/rA18, dT15/rA15, and oT15/rA15 duplexes at 20 °C for 0, 10, 20, and 40 min and 1 and 2 h; (C) 2′-OMe rU18/rA18,
pT18/rA18, dT15/rA15, and oT15/rA15 duplexes at 37 °C for 0, 20, and 40 min and 1 and 2 h. The plots represent the hydrolysis profile of rA15 or rA18 as a
function of temperature at 10, 20, and 37 °C for dT15, pT18, and oT15 complementary antisense strands.
of the extent of RNase H-mediated cleavage (Figure 9A). When
we compared the relative rates of RNase H-mediated degrada-
tion induced by 5- (furanose, dT15), 6- (2′-enopyranose, pT18),
and 7-membered (oxepane, oT15) ring oligonucleotides at 10
°C, the following trends were observed: dT15 . oT15 > pT18,
which parallels the decreasing conformational flexibility of the
constituent sugar rings, that is, furanose10 (5) . oxepane11 (7)
> 2′-enopyranose12 (6). These results support the notion that
the plasticity of the DNA/RNA hybrid is essential for efficient
RNase H catalysis, in which an enzyme-induced altered trajec-
tory of the bound substrate facilitates optimal interaction with
RNase H’s catalytic site.6,7a,48
Resistance of oT15 and oA15 toward Nuclease Degradation.
Since resistance to exo- and endonucleases is imperative for in
vivo applications of oligonucleotide-based therapeutics such as
AONs, siRNAs, and aptamers,49 we next determined the extent
of nuclease resistance of oT15 and oA15 against 10% fetal bovine
serum (FBS) at 37 °C. Under these conditions, the native dT15
and dA15 strands were degraded within 2 and 8 h, respectively,
whereas the oxepane-modified oligomers (oT15 and oA15) were
completely resistant to cleavage under the same conditions with
no noticeable degradation observed after 1 day (Supporting
Information).
Conclusions
Replacement of the furanose carbohydrate with oxepane
(oT15) confers resistance to nucleases while retaining the ability
to direct RNase H-mediated cleavage of a target RNA. The oT15
and pT18 oligomers of this study represent two new cases of
chemically modified ONs capable of activating RNase H when
bound to RNA, the others being, in chronological order, ANA,42b
2′-F-ANA,42a CeNA,8,16 and R-L-LNA.43 Although the extent
to which cleavage occurs is lower than that observed for the
(49) Kandimalla, E. R.; Manning, A.; Zhao, Q.; Shaw, D. R.; Byrn, R. A.;
(48) Nielsen, J. T.; Stein, P. C.; Petersen, M. Nucleic Acids Res. 2003, 31, 5858-
Sasisekharan, V.; Agrawal, S. Nucleic Acids Res. 1997, 25(2), 370-
5867.
378.
9
8266 J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007

Oxepane Nucleic Acids
A R T I C L E S
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5,7-di-tert-butylsilanediyl)-â-ox-
epinyl]thymine (2a). The product 2a was collected as its pure â-anomer
as a white foam (650 mg, 35%): Rf (hexanes/EtOAc 2:1) 0.22; H
wild-type hybrid, it is significant that oT15 and pT18 are able to
affect cleavage within the RNA component at all, in light of
the dramatically different structure of their sugar moiety. This
highlights the importance of sugar conformer flexibility along
the ON strand, which likely acts in concert with the global
helical architecture of the duplex to govern interactions between
RNase H and its substrate. We are undertaking further studies
with other ONA and 2′-enopyranose analogues obtained by
functionalizing the double bond of nucleosides 2 and 13a. These
endeavors should afford further insight into the substrate
requirements of the enzyme and provide a new repertoire of
potent antisense and siRNA ONs.
1
NMR (CDCl3, 400 MHz) δ 8.56 (br s, NH), 7.19 (1H, d, J ) 1 Hz,
H6), 5.91 (1H, ddd, J ) 2.4, 2.4, 12 Hz, H4′), 5.74 (1H, dd, J ) 6, 10
Hz, H1′), 5.68 (1H, m, H3′), 4.59 (1H, dd, J ) 2.4, 8.8 Hz, H5′), 4.06
(1H, dd, J ) 8.4, 10.5 Hz, H7′), 3.85 (1H, dd, J ) 8.4, 10.5 Hz, H7′′),
3.64 (1H, m, H6′), 2.59 (1H, m, H2′), 2.39 (1H, m, H2′′), 1.94 (3H, d,
J ) 1 Hz, H7), 1.05 (9H s, t-Bu Me), 0.99 (9H s, t-Bu Me); 13C NMR
1
(CDCl3, 100 MHz, H-decoupled at 400 MHz) δ 163.4 (CO), 149.8
(CO), 140 (C4′), 135.3, 122 (C3′), 111.36, 83.81 (C1′), 78.5 (C6′), 77
(C5′), 66.8 (C7′), 36.44 (C2′), 27.82 (t-Bu Me), 27.40 (t-Bu Me), 23.02
(t-Bu), 20.42 (t-Bu), 13.04; ESI-MS calcd for C20H32O5N2Si 408.6,
found 408.8.
Experimental Section
Characterization of (1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5,7-di-tert-
butylsilanediyl)-â-oxepinyl]-N6-benzoyladenine (2b). The product 2b
was collected as its pure â-anomer in yields of 710 mg (30%) as a
General. Thin-layer chromatography was performed with Merck
Kieselgel 60 F-254 aluminum-back analytical silica gel sheets (0.2 mm
thickness, EM Science, Gibbstown, NJ) and visualized with a solution
of 10% H2SO4 in MeOH and/or UV light. 31P NMR spectra were
acquired on a Varian Gemini 200 MHz spectrometer at 80 MHz with
1H decoupled and H3PO4 as external standard. Phosphoramidite samples
were further characterized by ESI-MS on a Finnigan LCQ DUO mass
spectrometer in methanol or acetone. UV spectra were measured at
260 nm on a Cary 300 dual-beam spectrophotometer.
1
white foam: Rf (EtOAc/hexanes 2:1) 0.24; H NMR (acetone-d6, 500
MHz) δ 10.0 (1NH, s, H6), 8.68 (1H, s, H8), 8.47 (1H, s, H2), 8-7.2
(5H, m, ar), 6.02 (1H, d, J ) 9 Hz H1′), 5.90 (1H, dd, J ) 8, 12 Hz,
H3′), 5.76 (1H, dd, J ) 8.5, 11 Hz, H4′), 4.74 (1H, d, J ) 9 Hz, H5′),
4.04 (1H, ddd, J ) 14, 10, 6 Hz, H7′), 3.91 (1H, dd, J ) 9.5, 4.5 Hz,
H6′), 3.86 (1H, d, J ) 18.5 Hz, H7′′), 3.40 (1H, d, J ) 14 Hz, H2′),
2.84 (1H, m, H2′′), 1.07 (9H, s, t-Bu Me), 1.04 (9H, s, t-Bu Me); 13C
(acetone-d6, 125.7 MHz, 1H-decoupled at 500 MHz) δ 152.16, 141.65,
139.3 (C3′), 132.6, 131.4, 128.75, 128.5, 128.4, 127.8, 127.6, 123.2
(C4′), 84.29 (C1′), 77.84 (C6′), 77.31 (C5′), 66.58 (C7′), 35.30 (C2′),
27.11 (t-Bu- Me), 26.78 (t-Bu- Me), 22.46 (t-Bu), 19.86 (t-Bu); ESI-
MS calcd for C27H35N5O4SiNa 544.7, found 544.1.
General Procedure for Desilylation Reaction. The starting material
2 (3.3 mmol) was dried overnight under vacuum. Under N2 at 0 °C,
the starting material was dissolved in 7 mL of THF. A solution of 1 M
TBAF in dry THF (7 mL) was added with stirring over 5 min. The
reaction mixture turned slightly cloudy as the deprotected nucleoside
slowly began to precipitate from the solvent. The reaction was complete
by TLC after 1 h, so the solvent was removed and the resulting viscous
oil was purified by column chromatography (9:1 CH2Cl2/MeOH) to
give a white foam.
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-hydroxy-7-hydroxymethyl)-â-
oxepinyl]thymine (4). Yield 0.854 g (90%); Rf (9:1 CH2Cl2/MeOH)
0.37; 1H NMR (DMSO-d6, 500 MHz) δ 11.30 (1NH, s, H3), 7.57 (1H,
s, H6), 5.74 (1H, t, J ) 9 Hz, H4′), 5.60 (1H, d, J ) 10 Hz, H1′), 5.55
(1H, d, J ) 9 Hz, H3′), 4.05 (1H, d, J ) 7.5 Hz, H5′), 3.60 (1H, d, J)
11 Hz, H7′), 3.46 (1H, d, J ) 11 Hz, H7′′), 3.36 (1H, t, J ) 6.5 Hz,
H6′), 3.14 (d, OH), 2.67 (1H, t, J ) 12 Hz, H2′), 2.32 (1H, dd, J ) 6,
7.5 Hz, H2′′), 1.76 (3H, s, 7), 1.55 (s, OH); 13C NMR (DMSO-d6, 125
MHz, 1H-decoupled at 500 MHz) δ 164.22 (CO), 150.33 (CO), 138.03
(C3′), 136.31 (C6), 83.45 (C1′), 34.89 (C2′), 122.27 (C4′), 68.91 (C5′),
84.80 (C6′), 61.95 (C7′), 23.06 (C7), 108.93 (C5); ESI-MS calcd for
C12H16N2O5 268.7, found 269.
3-Acetyl-1,5-anhydro-2-deoxy-1,2-C-methylene-4,6-O-(di-tert-bu-
tylsilanediyl)-D-glucal (1). Compound 1 was prepared according to
slight modifications of the known literature procedure.26,50 Commercially
available tri-O-acetyl-D-glucal was used as starting material and initially
deacetylated with 0.13 M NaOMe/MeOH for 1 h at room temperature.
After evaporation of the solvent and concentration of the deprotected
glycal as a white crystalline solid, the product was subsequently
protected with di-tert-butylsilyldi(trimethylsilyl)trifluoromethanesulfonate,
t-Bu2Si(OTf)2, in DMF at -40 °C to yield the 4,6-O-(di-tert-
butylsilanediyl)-D-glucal after extraction and column chromatographic
purification. The silylated glycal was subsequently subjected to
cyclopropanation conditions according to Furukawa’s modification51
of the Simmons-Smith reaction with 1 M ZnEt2 in hexanes and CH2I2.
This reaction was followed by acetylation with Ac2O in pyridine to
yield the target compound 1: 1H NMR (1, CDCl3, 400 MHz) δ 5.18
(1H, t, J ) 4.8 Hz), 4.09 (1H, m), 4.09 (1H, dd, J ) 6.4, 10 Hz), 3.69
(1H, t, J ) 6.8 Hz), 3.65 (1H, t, J ) 6.4 Hz), 3.42 (1H, m), 2.13 (3H,
s), 1.58 (1H, m), 1.01 (9H, m), 0.98 (9H, m), 0.71 (2H, m); 13C NMR
1
(1, CDCl3, 100 MHz, H-decoupled at 400 MHz) δ 75.9, 75.1, 73.5,
66.2, 55.3, 27.95, 27.5, 23.23, 21.87, 16.7, 13.19; ESI-MS calcd for
C16H28O5SiNa 342.5, found 343.3.
General Procedure for Glycosylation Reaction (2). The starting
material 1 (4.7 mmol) was dried overnight under high vacuum. Simi-
larly, in a separate flask, the base (thymine or N6-benzoyladenine) (24
mmol) and drying agent (NH4)2SO4 (2.5 mmol) were also dried under
vacuum. Under a nitrogen atmosphere at 22 °C, 85 mL of dry MeCN
was added to the flask containing the base and (NH4)2SO4. Hexameth-
yldisilazane (HMDS, 38 mmol) was added dropwise to the resulting
suspension, and the reaction was refluxed for 3-4 h until the MeCN-
soluble silylated base was formed. The solvent was evaporated, a
solution of 1 in 20 mL of dry MeCN was added and stirred at 22 °C
under N2, TMSOTf (1.6 mmol) was added dropwise, and the reaction
was refluxed at 90 °C. The crude product was extracted with EtOAc
(150 mL) followed by quenching and washing the upper organic layer
with 100 mL each of saturated NaHCO3 and H2O. The upper organic
layer was dried over MgSO4 and evaporated to dryness, and the residue
was purified on a column of silica gel [eluent 4:1 to 2:1 (v/v) hexanes/
EtOAc].
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-hydroxy-7-hydroxymethyl)-â-
oxepinyl]-N6-benzoyladenine (5). Yield 184 mg (61%); Rf [9:1 (v/v)
1
CH2Cl2/MeOH] 0.33; H NMR (MeOH-d4, 500 MHz) δ 8.71 (1H, s,
H8), 8.58 (1H, s, H2), 8.08 (2H, d, ar), 7.65 (1H, m, ar), 7.56 (2H, m,
ar), 6.09 (1H, dd, J ) 9.5, 2.5 Hz, H1′), 5.92 (ddd, 1H, J ) 2.5, 13
Hz, H3′), 5.76 (1H, m, H4′), 4.28 (1H, dd, J ) 2, 9 Hz, H5′), 3.89
(1H, dd, J ) 4.5, 9.5 Hz, H7′), 3.79 (1H, ddd, J ) 2.5, 5, 9 Hz, H6′),
3.68 (1H, dd, J ) 6, 11.5 Hz, H7′′), 3.17 (1H, m, H2′), 3.05 (OH),
2.90 (1H, ddd, J ) 2, 7, 16.5 Hz, H2′′), 1.65 (OH); 13C NMR (MeOH-
d4, 125.7 MHz, 1H-decoupled at 500 MHz) δ 175.23 (CO), 152.0 (C8),
142.2 (C2), 137.6 (C3′), 132.7 (ar), 128.6 (ar), 128.2 (ar), 122.3 (C4′),
84.90 (C1′), 84.29 (C6′), 69.92 (C5′), 62.83 (C7′), 35.23 (C2′); ESI-
MS calcd for C19H19N5O4 402.6, found 404.
(50) Oplinger, J. A.; Paquette, L. A. J. Org. Chem. 1998, 53 (13), 2953-
2959.
General Procedure for Hydrogenation Reaction. The product from
the desilylation reaction (0.565 mmol) and 152 mg of 10% Pd/C catalyst
were dried overnight under vacuum. Dry MeOH (11.5 mL) was added
(51) (a) Furukawa, J.; Kawabata, N.; Nishimura, J. Tetrahedron 1968, 24, 53-
58. (b) Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. ReV. 1993, 93,
1307-1370.
9
J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007 8267

A R T I C L E S
Sabatino and Damha
to the evacuated flask and the dark suspension was stirred at room
temperature (22 °C). A balloon filled with H2 was attached to the flask
by piercing the septum with a needle. Small aliquots were periodically
withdrawn and evaporated to dryness, and the extent of reaction was
3, 6.5, 15.6 Hz, H2′), 2.16 (1H, ddd, J ) 4.8, 9, 15.6 Hz, H2′′), 1.99
(1H, dd, J ) 4.8, 17.2 Hz, H3′), 1.92 (1H, dd, J ) 2.8, 17.2 Hz, H3′′),
1.82 (1H, ddd, J ) 4, 8.4, 16 Hz, H7′), 1.73 (1H, t, J ) 8.4 Hz, H7′′);
1
13C (CDCl3, 125.7 MHz, H-decoupled at 500 MHz) δ 158.9 (CO),
1
verified by H NMR. After 4 h, the remaining reaction mixture was
152.8 (C8), 144.3, 140.6 (C2), 144.1, 135.2, 132.98, 130.5, 129.5, 129.1,
128.48, 128.13, 128.1, 128.08, 127.4, 127.3, 113.5, 95, 87.2, 86.44
(C1′), 83.71 (C5′), 73.28 (C6′), 65.88 (C4′), 55.44 (OMe), 36.08 (C2′),
33.61 (C7′), 18.37 (C3′); ESI-MS calcd for C39H37N5O5 655.9, found
655.7.
filtered and evaporated to dryness. The crude product was purified by
flash silica gel column chromatography in 9:1 CH2Cl2/MeOH.
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-hydroxy-7-hydroxymethyl)-â-
oxepanyl]thymine (6). The extent of reaction could not be accurately
monitored by TLC, as the Rf values for the starting material and product
were found to be identical in eluent system 9:1 CH2Cl2/MeOH. The
product was obtained in 70% yield (104 mg) as a white foam after
silica gel column chromatography (9:1 CH2Cl2/MeOH): Rf (9:1 CH2Cl2/
MeOH) 0.21; 1H NMR (MeOH-d4, 500 MHz) δ 7.56 (1H, s, H6), 5.78
(1H, d, J ) 9.5 Hz, H1′), 3.77 (1H, s, H5′), 3.70 (1H, d, J ) 11.5 Hz,
H7′), 3.55 (1H, m, H6′), 3.55 (1H, m, H7′′), 1.97 (1H, m, H2′), 1.90
(2H, m, H3′, H3′′), 1.90 (3H, s, H7), 1.88 (2H, m, H4′, H4′′), 1.66
(1H, d, J ) 6 Hz, H2′′); 13C NMR (MeOH-d4, 125 MHz, 1H-decoupled
at 500 MHz) δ 150.9 (CO), 137 (C6), 110 (CO), 95 (C5), 86.64 (C6′),
86.5 (C1′), 70.8 (C5′), 63.37 (C7′), 34.87 (C4′), 33.29 (C3′), 17.9 (C2′)
11.18 (C7); ESI-MS calcd for C12H18N2O5Na 279.3, found 293.1.
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-hydroxy-7-hydroxymethyl)-â-
oxepanyl]-N6-benzoyladenine (7). Compound 7 was isolated in 99%
yield (149 mg) as a white foam after silica gel column chromatography
(9:1 CH2Cl2/MeOH): Rf (9:1 CH2Cl2/MeOH) 0.22; 1H NMR (MeOH-
d4, 500 MHz) δ 9.81 (1NH, s), 8.71 (1H, s, H8), 8.58 (1H, s, H2),
8.08 (2H, ar), 7.65 (1H, ar), 7.56 (2H, ar), 6.06 (1H, dd, J ) 5, 5.5 Hz,
H1′), 3.80 (1H, dd, J ) 4, 8.25 Hz, H6′), 3.76 (1H, dd, J ) 3, 6 Hz,
H4′), 3.58 (1H, dd, J ) 6.35, 12 Hz, H4′′), 3.12 (1H, t, J ) 8 Hz,
H5′), 2.39 (2H, m, H2′, H2′′), 2.04 (2H, m, H3′, H3′′), 1.93(1H, m,
H7′), 1.81 (1H, m, H7′′), 1.67 (OH), 1.42 (OH); 13C NMR (MeOH-d4,
125.7 MHz, 1H-decoupled at 500 MHz) δ 151.9 (C8), 142 (C2), 132.7
(ar), 128.6 (ar), 128.3 (ar), 87.25 (C1′) 70.68 (C5′), 63.54 (C4′), 52.90
(C6′), 35.08 (C2′), 33.76 (C3′), 18.0 (C7′); ESI-MS calcd for
C19H21N5O4Na 406.4, found 406.4.
General Procedure for Tritylation Reaction. Nucleoside 6 or 7
(0.377 mmol) and MMT-Cl (0.44 mmol), were dried overnight under
vacuum. Pyridine (1.5 mL) was added at 22 °C under nitrogen. The
reaction was stirred for 4 h until TLC (9:1 CH2Cl2/MeOH) indicated
completion. The reaction was diluted with EtOAc (60 mL) and washed
with saturated aqueous NaHCO3 (2 × 60 mL). The organic layer was
then dried over MgSO4, concentrated, and purified by silica gel
chromatography (9:1 CH2Cl2/MeOH).
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-hydroxy-7-[4-(methoxyphenyl)-
diphenyl])-â-oxepanyl]thymine (8). Yield 66% (135 mg) as a white
foam; Rf (9:1 CH2Cl2/MeOH) 0.68; 1H NMR (CDCl3, 500 MHz) δ 8.85
(1H, s, NH), 7.34 (4H, ar), 7.2 (1H, s, H6),. 7.19 (8H, ar), 6.76 (2H,
ar), 5.74 (1H, dd, J ) 3.6, 9.8 Hz, H1′), 3.83 (H1, ddd, J ) 3.6, 4.8,
7.6 Hz, H5′), 3.72 (s, OMe), 3.6 (1H, dd, J ) 6, 12.6 Hz, H6′), 3.31
(1H, dd, J ) 5.6, 9.6 Hz, H7), 3.11 (1H, dd, J ) 5.6, 9.6 Hz, H7′′),
1.98 (1H, dd, J ) 3.6, 9 Hz, H3′), 1.78 (1H, m, H3′′), 1.80 (2H, m,
H4′, H4′′), 1.86 (3H, d, J ) 1 Hz, H7), 1.62 (2H, dd, J ) 8.4, 14.4 Hz,
H2′, H2′′); 13C (CDCl3, 125.7 MHz, 1H-decoupled at 500 MHz) δ
163.79 (CO), 158.93 (CO), 149, 144.3,144,1, 135.9 (C6), 135.2, 130.6,
128.5, 128.2, 127.3, 113.5, 110.9, 86.53 (C1′), 87.16 (C5), 83.58 (C6′),
73.41 (C5′), 65.90 (C7′), 55.47 (OMe), 35.81 (C3′), 33.48 (C4′), 18.31
(C2′), 12.86 (C7); ESI-MS calcd for C32H34N2O6Na 565.6, found
565.1.
General Procedure for Phosphitylation Reaction. The tritylated
nucleoside (0.224 mmol) was dried under vacuum overnight prior to
reaction. Dry THF (1.2 mL) was added under N2. To the resulting
solution were added dropwise, over a span of 10 min, EtN(i-Pr)2 (0.89
mmol) and Cl--P(OCEt)N(i-Pr)2 (0.246 mmol). The reaction mixture
was stirred for 2 h at 22 °C, and the reaction progress was monitored
by TLC (hexanes/EtOAc 2:1). The progression of the reaction is also
observable by the formation of a white precipitate, Cl-+NH(Et)(i-Pr)2.
After the reaction reached completion, EtOAc (15 mL) was added and
the mixture was washed twice with saturated aqueous NaHCO3, dried
over MgSO4, and concentrated to a yellowish foam, which was purified
by silica gel chromatography (hexanes/EtOAc 2:1 to 1:2 containing
3% TEA).
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-phosphoramidous-7-[4-(meth-
oxyphenyl)diphenyl])-â-oxepanyl]thymine (10). Yield 131 mg (80%)
as a white foam; Rf (2:1 hexanes/EtOAc) 0.52; 31P NMR (CDCl3, 80.99
1
MHz, H-decoupled at 200 MHz) δ 149.5 and 149.02; ESI-MS calcd
for C41H51N4O7PNa 765.86, found 765.2.
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-phosphoramidous-7-[4-(meth-
oxyphenyl)diphenyl])oxepanyl]-N6-benzoyladenine (11). Yield 135
mg (92%) as white foam; Rf (2:1 hexanes/EtOAc) 0.33; 31P NMR
1
(CDCl3, 80.99 MHz, H-decoupled at 200 MHz) δ 148.4 and 147.9;
ESI-MS calcd for C48H54N7O6PNa 878.9, found 878.3.
1-(2,3-Dideoxy-4,6-di-O-acetyl-D-erythro-hex-2-enopyranosyl)thy-
mine r- and â-Anomers (13). The starting material 12 (7.35 mmol)
was dried separately from thymine (14.7 mmol) and the drying agent
(NH4)2SO4 (1.4 mmol) overnight under high vacuum. Anhydrous MeCN
(56 mL) under N2 at 22 °C was added to the reaction flask containing
thymine and the drying agent, and the mixture was stirred to a white
slurry suspension prior to dropwise addition of HMDS (23.85 mmol)
to initiate the silylation reaction of the base. The reaction was refluxed
to completion at 100 °C for 3-4 h and the solvent was evaporated to
a viscous yellowish oil. A solution of the starting material 12 in 30
mL of anhydrous MeCN was stirred at 22 °C under N2 and transferred
to the silylated thymine. The coupling reaction was initiated after
dropwise addition of TMSOTf (11 mmol) and completed for 3 h under
reflux at 90 °C as was monitored by TLC: Rf (2:1 hexanes/EtOAc) )
0.29 and 0.20 for the â- and R-anomers, respectively. The crude reaction
mixture was diluted with EtOAc (200 mL) and in turn quenched and
washed with NaHCO3 and H2O (150 mL). The organic layer was dried
over MgSO4 and concentrated for chromatographic purification with
eluent gradient of 4:1 hexanes/EtOAc to 2:1 EtOAc/hexanes. The
purified â-anomer, 13a, was collected with a yield of 765 mg (35%)
and the R-anomer, 13b, was collected with a yield of 935 mg (40%).
1-(2,3-Dideoxy-4,6-di-O-acetyl-â-D-erythro-hex-2-enopyranosyl)-
thymine (13a). 1H NMR (CDCl3, 400 MHz) δ 8.16 (s, NH), 6.96 (1H,
s, H5), 6.52 (1H, s, H1′), 6.16 (1H, d, H3′), 5.75 (1H, d, J ) 10 Hz,
H2′), 5.38 (1H, d, J ) 6.5 Hz, H4′), 4.20 (2H, d, J ) 3.5 Hz, H6′,
H6′′), 4.00 (1H, t, J ) 4.5 Hz, H5′), 2.11 (3H, s, OAc), 2.08 (3H, s,
1
OAc), 1.91 (3H, s); 13C NMR (CDCl3 125.7 MHz, H-decoupled at
(1R)-1-[(2,3,4-Trideoxy-(5S,6R)-5-hydroxy-7-[4-(methoxyphenyl)-
diphenyl])-â-oxepanyl]-N6-benzoyladenine (9). Yield 50% (120 mg)
500 MHz) δ 169.3 (CO), 168.8 (CO), 161.7 (CO), 148.8 (CO), 131.2
(C3′), 134.1 (C6), 126.3 (C2′), 110.77 (C5), 77.19 (C1′), 73.77 (C5′),
62.89 (C4′), 61.53 (C6′), 19.83 (OAc), 19.73 (OAc), 11.42 (C7); ESI-
MS calcd for C15H18N2O7 338.9, found 338.3.
1-(2,3-Dideoxy-4,6-di-O-acetyl-r-D-erythro-hex-2-enopyranosyl)-
thymine (13b). 1H NMR (CDCl3, 400 MHz) δ 8.18 (s, NH), 7.24 (1H,
s, H5), 6.40 (1H, s, H1′), 6.30 (1H, d, J ) 10 Hz, H3′), 5.86 (1H, d,
1
as a white foam; Rf (9:1 CH2Cl2/MeOH) 0.65; H (CDCl3, 500 MHz)
δ 9.03 (1H, s, H4), 8.14 (1H, s, H2), 8.72 (1H, s, H8), 7.95 (2H, ar),
7.52 (1H, ar), 7.44 (2H, ar), 7.29 (3H, ar), 7.17 (8H, ar), 6.71 (2H, ar),
5.99 (1H, dd, J ) 2.8, 10 Hz, H1′), 3.90 (1H, t, J ) 8.4 Hz, H6′), 3.79
(1H, dd, J ) 5.6, 12.6 Hz, H5′), 3.71 (OMe), 3.30 (1H, dd, J ) 5.6,
9.6 Hz, H4′), 3.14 (1H, dd, J ) 6, 9.6 Hz, H4′′), 2.27 (1H, ddd, J )
9
8268 J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007

Oxepane Nucleic Acids
A R T I C L E S
J ) 10.5 Hz, H2′), 5.24 (1H, d, J ) 3 Hz, H4′), 4.27 (1H, dd, J ) 12
Hz, H6′), 4.16 (1H, dd, J ) 3, 12.25 Hz, H6′′), 4.01 (1H, m, J ) 4.5
Hz, H5′), 2.12 (s, OAc), 2.09 (s, OAc), 1.93 (3H, s, H7); ESI-MS calcd
for C15H18N2O7 338.9, found 338.3.
2.5 min. Oligonucleotides were cleaved from the CPG and base-
protecting groups were removed by overnight incubation at 55 °C in
saturated NH4OH/EtOH (3:1). The deprotected oligonucleotides were
evaporated to dryness, extracted from the CPG with 1 mL of sterile
H2O, and quantitated by UV/vis spectrophotometry. Crude ONA
oligonucleotides were purified by denaturing PAGE or anion-exchange
HPLC with a 30% gradient of 1 M LiClO4. Purified oligonucleotides
were extracted and desalted with Nap 10 Sephadex columns.
1-(2,3-Dideoxy-â-D-erythro-hex-2-enopyranosyl)thymine (14). The
starting material, 13a (2.07 mmol), was vacuum-dried overnight prior
to reaction. Under N2 and at 22 °C, the starting material was dissolved
in 10 mL of anhydrous MeOH followed by dropwise addition of a 0.1
M NaOMe/MeOH (7 mL) solution. The reaction proceeded for 1 h,
and the product was confirmed by TLC [Rf (9:1 CH2Cl2/MeOH) )
0.2]. The crude reaction mixture was concentrated to a white slurry
and purified by column chromatography with eluent system 9:1 CH2-
Cl2/MeOH. The purified product was collected as a white foam, yield
0.485 g (93%); 1H NMR (DMSO-d6, 400 MHz) δ 11.34 (s, NH), 7.16
(1H, s, H6), 6.23 (1H, s, H1′), 6.08 (1H, d, J ) 11 Hz, 3′), 5.64 (1H,
d, J ) 10.4 Hz, H2′), 5.27 (s, OH), 4.71 (s, OH), 4.03 (1H, d, J ) 6.8
Hz, H4′), 3.67 (1H, d, J ) 11 Hz, H6′), 3.46 (1H, m, H6′′), 3.30 (1H,
s, H5′), 1.76 (1H, s, H7); 13C NMR (DMSO-d6, 125.7 MHz, 1H-
decoupled at 500 MHz) δ 163.8 (CO), 150.5 (CO), 137.3 (C3′), 136.5
(C6), 124.9 (C2′), 109.9 (C5), 81.06 (C5′), 77.99 (C1′), 61.24 (C4′),
60 (C6′), 11.99 (C7); ESI-MS calcd for C11H14N2O5 254.3, found
254.2.
Characterization of Oligonucleotides by MALDI-TOF Mass
Spectrometry. All spectra were acquired on a Kratos Kompact-III
instrument operated in the negative linear mode for 200 pmol of
oligonucleotide sample dissolved in 1 µL of H2O, 1 µL of 6-aza-2-
thiothymine/spermine, and 1 µL of 50 mM aqueous L-fucose mix-
ture.
UV Thermal Denaturation Experiments, Tm. The studies were
conducted on a UV-vis Cary 300 dual-beam spectrophotometer for 3
µM duplex in a phosphate buffer that mimics the physiological ionic
strength: 140 mM KCl, 5 mM NaH2PO4, and 1 mM MgCl2, pH ) 7.
The solutions were denatured at 85 °C for 10 min and then cooled to
room temperature over 2.5 h and kept at 5 °C overnight prior to the
experiment. The thermal melt measurements were run from 5 -to 85
°C with temperature gradient increases of 0.5 °C/min and data point
intervals of 0.5/min at a wavelength of 265 nm.
1-{(4S,5R)-4-Hydroxy-6-[4-(methoxyphenyl)diphenyl]-2,3-dideoxy-
â-D-erythro-hex-2-enopyranosyl}thymine (15). The starting material,
14 (1.91 mmol), and MMT--Cl (2.20 mmol) were dried under high
vacuum prior to the start of the reaction. The reaction was initiated at
22 °C under N2 by dropwise addition of anhydrous pyridine (8 mL)
and stirred to complete reaction for 3 h, as was monitored by TLC [Rf
(9:1 CH2Cl2/MeOH) ) 0.68]. The crude reaction mixture was diluted
in 35 mL of EtOAc or Et2O and quenched with NaHCO3 (2 × 20 mL)
prior to being dried over MgSO4 and evaporated to dryness. The crude
reaction mixture was purified by silica gel chromatography, with eluent
gradient ranging from 151 to 9:1 CH2Cl2/MeOH. The product was
collected as a white foam, 0.78 mg (78% yield); 1H NMR (CDCl3, 400
MHz) δ 8.61 (NH), 7.40, 7.27, 6.84, (MMT), 6.94 (1H, d, J ) 2 Hz,
H6), 6.42 (1H, d, J ) 2 Hz, H1′), 6.20 (1H, ddd, J ) 2, 10 Hz, H3′),
5.64 (1H, ddd, J ) 1.8, 3.6, 10 Hz, H2′), 4.30 (1H, dd, J ) 2.4, 8.4
Hz, H4′), 3.82 (OCH3), 3.80 (1H, m, H5′), 3.58 (1H, dd, J ) 4.8, 9.2
Hz, H6′), 3.31 (1H, dd, J ) 6.4, 9.2 Hz, H6′′), 1.90 (1H, d, H5); 13C
CD Experiments. Duplexes were prepared as for UV Tm analysis.
Samples were analyzed on a Jasco Model J-710 or J-810 spectropo-
larimeter with temperature controlled with a circulating bath (VWR
Scientific) in the 310-210 nm wavelength range as the sum of three
replicates which were finally baseline-corrected from the phosphate
buffer used (140 mM KCl, 5 mM NaH2PO4, and 1 mM MgCl2, pH )
7). For the temperature-dependent CD-melt experiments, the samples
were allowed to equilibrate for 5-10 min under N2 at a fixed
temperature prior to the runs.
UV Mixing Curves. The studies were conducted on a UV-vis Cary
300 dual beam spectrophotometer. Equimolar stock solutions of each
oligonucleotide (2.5 nmol) were prepared in 0.5 mL of buffer (10 mM
Na2HPO4 and 100 mM NaCl, pH ) 7.2). At low temperature (5 °C),
100 µL aliquots (0.5 nmol) of T solution were titrated into the stock
solution containing A and allowed to hybridize for 5-10 min prior to
measuring the UV absorbance at 260 nm under constant flow of N2.
Absorbance values were measured at 260 nm for each mol fraction of
T titrated into a fixed concentration of complementary A stock solu-
tion.
1
NMR (CDCl3, 100 MHz, H-decoupled at 400 MHz) δ 163.4 (CO),
158.8 (CO), 150.4, 149.1, 143.6, 136.3 (C2′), 135.89 (C6), 134.7, 130.4,
128.3, 128.2, 127.9, 127.4 (C3′), 125.4, 123.96 (C7′), 111.7(C5), 87.8,
78.64 (C1′), 77.64 (C5′), 66.3 (C4′), 65.5 (C6′), 55.6 (OMe), 13.01;
ESI MS calcd for C31H30N2O6 549.4, found 549.1.
Serum Stability of Oligonucleotides. Pure oligothymidylate and
oligoadenylate (0.7 OD) were evaporated to dryness. To each dried
oligonucleotide sample was added 300 µL of 10% FBS (0.2 µM, in
10% fetal bovine serum diluted with Eagle’s medium) and incubated
at 37 °C for up to 24 h. For each time point, a 50 µL aliquot of the
reaction mixture was removed and frozen on dry ice for 10 min followed
by evaporation in a SpeedVac concentrator. Aliquots were redissolved
in deionized formamide and analyzed by denaturing 24% PAGE,
visualized by Stains-All dye.
1-[(2,3-Dideoxy-(4S,5R)-5-[4-(methoxyphenyl)diphenyl]-4-phos-
phoramidous-â-D-erythro-hex-2-enopyranosyl)thymine (16). The
starting material, 15 (1.48 mmol), was vacuum-dried overnight prior
to the start of the reaction. At 22 °C and under N2, the starting material
was dissolved in anhydrous THF (8 mL) and the reaction was ini-
tiated by dropwise addition of EtN(i-Pr)2 (5.6 mmol) and Cl--
P(OCEt)N(i-Pr)2 (1.56 mmol). The reaction was monitored to comple-
tion by TLC [Rf (2:1 EtOAc/hexanes) ) 0.59]. The crude reaction
mixture was diluted in EtOAc or Et2O (70 mL) and quenched/washed
with saturated NaHCO3 (2 × 30 mL). The upper organic layer was
dried over MgSO4 and evaporated to a yellow-tinged foam. The product
was then purified by column chromatography with eluent gradient of
2:1 hexanes/EtOAc to 2:1 EtOAc/hexanes with 3% TEA. The purified
phosphoramidite diastereomers were collected with a yield of 825 mg
RNase H Assays. The RNA strand (200 pmol) was initially
5′-radiolabeled with T4 polynucleotide kinase and [γ-32P]ATP. A 1.8-
fold excess of 15- or 18-mer oligothymidylate was added to the purified
RNA strand and annealed in 100 µL solutions containing 0.5× reaction
buffer (i.e., 20 pmol in 10 µL of buffer; 100 mM Tris-HCl, pH )
7.5, 100 mM KCl, 50 mM MgCl2, 0.5 mM EDTA, and 0.5 mM DTT).
E. coli RNase H assays were performed at different temperature
conditions (10, 20, and 37 °C) in 10 µL reactions containing 2 pmol
of duplex substrate with 2 µL of reaction buffer (100 mM Tris-HCl,
pH ) 7.5, 100 mM KCl, 50 mM MgCl2, 0.5 mM EDTA, and 0.5 mM
DTT) and 0.5 µL of E. coli RNase H enzyme (concentration 5 units/
µL in storage buffer: 20 mM Tris-HCl, pH ) 7.9, 100 mM KCl, 10
mM MgCl2, 0.1 mM EDTA, 0.1 mM DTT, and 50% glycerol). Each
reaction was quenched at various time points by heating to 90 °C and
1
(80%) as a white foam: 31P NMR (CDCl3, 80 MHz, H-decoupled at
200 MHz) δ 151.8 and 150.1; ESI-MS calcd for C40H47N4O7PNa 749.8,
found 749.2.
Oligonucleotide Synthesis. Synthesis of modified oligonucleotides
was performed on a 500 Å Unylinker CPG on a 0.5-1 µmol scale.
Phosphoramidites were prepared as 0.05-0.1 M solutions (10 in CH2-
Cl2, 11 and 16 in MeCN). The coupling times were 30 min with 0.25
M ETT in MeCN as activator. The detritylation time was extended to
9
J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007 8269

A R T I C L E S
Sabatino and Damha
addition of 10 µL of stop reaction solution (50 mM EDTA in
formamide) prior to analysis by 16% denaturing PAGE.
the 2007 recipient of the Bernard Belleau Award (Canadian
Society for Chemistry).
Acknowledgment. Dedicated to Professor Kelvin Kenneth
Ogilvie on the occasion of his 65th birthday. We thank the
Canadian Institute of Health Research (CIHR), the National
Science & Engineering Council of Canada (NSERC), and
McGill University for funding. We thank J. K. Watts and R.
A. Donga for critical proofreading of the manuscript. M.J.D. is
Supporting Information Available: Figures and tables as
described in the text. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA071336C
9
8270 J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007