Synthesis, physicochemical and biochemical studies of 1′,2′- oxetane constrained adenosine and guanosine modified oligonucleotides, and their comparison with those of the corresponding cytidine and thymidine analogues

Article abstract of DOI:10.1021/ja048417i

We have earlier reported the synthesis and antisense properties of the conformationally constrained oxetane-C and -T containing oligonucleotides, which have shown effective down-regulation of the proto-oncogene c-myb mRNA in the K562 human leukemia cells.

Full text of DOI:10.1021/ja048417i

Published on Web 08/27/2004
Synthesis, Physicochemical and Biochemical Studies of
1′,2′-Oxetane Constrained Adenosine and Guanosine Modified
Oligonucleotides, and Their Comparison with Those of the
Corresponding Cytidine and Thymidine Analogues
Pushpangadan I. Pradeepkumar,† Pradeep Cheruku,† Oleksandr Plashkevych,†
Parag Acharya,† Suresh Gohil,‡ and Jyoti Chattopadhyaya*,†
Contribution from the Department of Bioorganic Chemistry, Box 581, Biomedical Center,
UniVersity of Uppsala, S-75123, Uppsala, Sweden, and Department of Chemistry,
Swedish UniVersity of Agricultural Sciences, Box 7015, Uppsala, Sweden
Received March 19, 2004; E-mail: jyoti@boc.uu.se
Abstract: We have earlier reported the synthesis and antisense properties of the conformationally
constrained oxetane-C and -T containing oligonucleotides, which have shown effective down-regulation of
the proto-oncogene c-myb mRNA in the K562 human leukemia cells. Here we report on the straightforward
syntheses of the oxetane-A and oxetane-G nucleosides as well as their incorporations into antisense
oligonucleotides (AONs), and compare their structural and antisense properties with those of the T and C
modified AONs (including the thermostability and RNase H recruitment capability of the AON/RNA hybrid
duplex by Michaelis-Menten kinetic analyses, their resistance in the human serum, as well as in the
presence of exo and endonucleases).
Introduction
Antisense oligonucleotides (AONs), based on elegant prin-
ciple of Watson-Crick base pairing with the complementary
mRNA, have been shown to be capable of down-regulation of
the genes of interest.1 AONs are being developed as a gene
knock-off tool in functional genomics as well as therapeutic
agent1, since their first inception in 1978.2 In the clinical
perspective, the antisense technology utilizing AONs, which
elicit the RNase H cleavage of target RNA for the down-
regulation of the disease causing genes (Figure 1), offers
advantages in terms of efficiency and dosage compared to those
AONs, which offer gene silencing by the steric blockage of the
ribosomal read-through.3 The RNase H cleavage efficiency of
the AON/RNA hybrid duplex can be best exploited for potential
therapeutic usage when the AONs show the adequate nuclease
resistance in vivo, sequence specificity, cell deliverability,
nontoxicity, and favorable pharmacokinetics.1,4 Chemical modi-
fications are warranted to achieve those goals, however, in most
of the cases the nature of modifications hampers one or more
of the above-mentioned requirements.4 Since the widely studied
phosphorothioate AONs are marred with their low sequence
specificity and higher (1-3 orders of magnitude) binding affinity
Figure 1. Catalytic RNase H promoted cleavage of the target mRNA
through the formation of antisense Oligonucleotide/RNA Hybrid Duplex.
The kinetic scheme of the RNase H hydrolysis is shown in the bottom part
of the cartoon, where D is AON (antisense Oligo); R is the target RNA;
K
K
d1 is the equilibrium constant of dissociation of the heteroduplex DR and
d2 is equilibrium constant of dissociation of the substrate-enzyme complex
DRE.
for various cellular proteins, especially the heparin binding
proteins, which accounts for many of their reported nonantisense
effects,5 the search for alternative chemical routes of nucleoside
modification without changing the phosphodiester backbone has
been intensified in the past decade.
† University of Uppsala, http://www.boc.uu.se.
‡ Swedish University of Agricultural Sciences.
(1) (a) Scherer, L. J.; Rossi, J. J. Nat. Biotech. 2003, 21, 1457. (b) Dias, N.;
Stein, C. A. Mol. Cancer. Ther. 2002, 1, 347. (c) Opalinska, J. B.; Gewirtz,
A. M. Nat. ReV. Drug. DiscoV. 2002, 1, 5303. (d) Agrawal, S.; Kandimalla,
E. R. Mol. Med. Today 2000, 6, 72.
(2) Zamecnik, P.; Stephenson, M. L. Proc. Natl. Acad. Sci. U.S.A. 1978, 75,
280.
(3) Zamaratski, E.; Pradeepkumar, P. I.; Chattopadhyaya, J. J. Biochem.
Biophys. Methods 2001, 48, 189.
(5) (a) Stein, C. A. J. Clin. InVest. 2001, 108, 641. (b) Lebedeva, I.; Stein C.
A. Annu. ReV. Pharmacol. 2001, 41, 403. (c) Levin, A. A. Biochim. Biophys.
Acta 1999, 1489, 69.
(4) Kurreck, J. Eur. J. Biochem. 2003, 270, 1628.
9
11484
J. AM. CHEM. SOC. 2004, 126, 11484-11499
10.1021/ja048417i CCC: $27.50 © 2004 American Chemical Society

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
In recent times, AONs incorporated with sugar modified
nucleosides have been widely pursued as a valuable alterna-
tive.6,7,17 Among these, the AONs modified with North-
conformationally (-1° < P < 34°)8 constrained nucleotides
have attracted much attention because of their ability to drive
the AON strand of the AON/RNA hybrid duplex to the RNA/
RNA type, thereby increasing the target affinity.6,7 They can
be broadly classified into three groups. The first group comprises
modified AONs, a gap size of 8-10 nucleotides was necessary
to recruit the effective RNase H cleavage capability of the target
RNA.19,20 It is likely that use of such a lengthy unmodified
phosphodiester oligonucleotide gaps could be harmful in view
of the endonuclease susceptibility of the AONs.21
We have developed novel 1′,2′-oxetane locked nucleosides,
[1-(1′,3′-O-anhydro-â-D-psicofuranosyl) nucleosides] which have
a unique fixed North-East sugar conformation,22-24 and belong
to the third class of the above-mentioned conformationally
constrained nucleosides. Even though the Tm of the AON/RNA
hybrid drops by ∼5-6 °C/oxetane-thymidine (T) modification,
the mixmer AON/RNA hybrids incorporated with T units were
found to be substrates for RNase H promoted cleavage as good
as the native hybrid.22-24 However, the incorporation of the
oxetane-cytidine (C) moiety into the AONs imparts only ∼3
°C loss in Tm per oxetane-modification.24 The loss of the
thermodynamic stability in case of oxetane-T and -C was fully
or partly regained by the introduction of the nontoxic25,26 DPPZ
(dipyridophenazine) group25 at the 3′ end, and it gave additional
stability against exonucleases similar to that of the phospho-
rothioate AONs.22c,24 Another interesting observation was that
only 4 deoxynucleotide gaps in the AON strand were needed
to achieve the RNase H cleavage of the RNA in their hybrids
duplexes.22,24 Michaelis-Menten kinetics of the RNase H
cleavage showed that Vmax and Km increase with increase in
the number (one to three) of T/C modifications in the AONs,
indicating higher catalytic activity and lower enzyme binding
affinity of the oxetane-modified AON/RNA hybrids.23,24 The
endonuclease cleavage of these molecules was reduced signifi-
cantly and it was proportional to the number of the oxetane
modified nucleotides per AON molecule: single modification
gave 2-fold protection to the cleavage and double and triple
modification gave 4-fold protection compared to that of the
native phosphodiester oligonucleotide.22c,24
of AONs incorporated with various 2′-O-alkyl moieties [-CH3,9
-CH2-CH2-O-CH3,9 -CH2-CH2-CH2-NH2
etc.] or
7,10
electron withdrawing functional groups such as the F11 in the
ribo configuration. The monomer units of 2-O-alkyl modified
compounds exist in solution in the North-South pseudorotational
equilibrium (with very little preference for the North-type
sugars12), however, upon incorporation into the AON and
subsequent hybridization with the target RNA, they impart 3′-
endo conformation to their sugar moiety and to the neighboring
nucleotides in a varying extent.13 The second group of AONs
encompasses nucleic acids containing pyranose derivatives such
as the hexitol nucleic acids14 and their AON/RNA hybrid
duplexes mimic RNA/RNA type helix.6,14 The third group of
AONs consists of those containing conformationally constrained
bicyclic or tricyclic monomer nucleotide units.15 Because of their
conformational preorganization at the monomer level itself, the
complete or sometimes partial modification of the AON strand
with North-conformationally constrained units16 drive the AON/
RNA duplex to the rigid RNA/RNA type duplex17 which results
in the loss of the RNase H cleavage efficiency.3 However,
RNase H eliciting capability can partly or fully be regained by
adopting various mixmer and gapmer strategies utilizing these
modifications.18,19 In some cases such as in the â-D-LNA
(6) Herdewijn, P. Biochim. Biophys. Acta 1999, 1489, 167.
(7) Manoharan, M. Biochim. Biophys. Acta 1999, 1489, 117.
(8) de Leeuw, H. P. M.; Haasnoot, C. A. G.; Altona, C. Isr. J. Chem. 1980,
20, 108.
The antisense oligo constructs with the oxetane-C and 3′-
DPPZ were found to be nontoxic in K562 human leukemia cells
and have been successfully employed to down-regulate the
proto-oncogene c-myb in a very efficient manner.26 The QRT-
PCR and Western blotting (the “gold standard” in antisense
(9) Martin, P. HelV. Chim. Acta 1995, 78, 486.
(10) Griffey, R. H.; Monia, B. P.; Cummins, L. L.; Frier, S. M.; Greig, M. J.;
Guinosso, C. J.; Lesnik, E.; Manalili, S. M.; Mohan, V.; Owens, S.; Ross,
B. R.; Sasmor, H.; Wancewicz, E.; Wieler, K.; Wheeler, P. D.; Cook, P.
D. J. Med. Chem. 1996, 39, 5100.
(11) Kawasaki, A. M.; Casper, M. D.; Frier, S. M.; Lesnik, E. A.; Zounes, M.
C.; Cummins, L. L.; Gonzalez, C.; Cook, P. D. J. Med. Chem. 1993, 36, 831.
(12) Kawai, G.; Yamamoto, Y.; Amimura, T.; Masegi, T.; Sekine, M.; Hata,
T.; Iimori, T.; Watanabe, T.; Miyazawa, T.; Yokoyama, S. Biochemistry
1992, 31, 1040.
(13) Nishizaki, T.; Iwai, S.; Nakamura, H. Biochemistry 1997, 36, 2577.
(14) Hendrix, C.; Rosenmeyer, H.; Bouvere, B. D.; Aerschot, A. V.; Seela, F.;
Herdewijn, P. Chem. Eur. J. 1997, 3, 1513.
(18) Lima, W. F.; Crooke, S. T. Biochemistry 1997, 36, 390.
(19) (a) Kurreck, J.; Wyszko, E.; Gillen, C.; Erdmann, V. A. Nucleic Acids
Res. 2002, 30, 1911. (b) Braasch, A. D.; Liu, Y.; Corey, D. R. Nucleic
Acids Res. 2002, 30, 5160.
(15) (a) Kvaerno, L.; Wengel, J. Chem. Commun. 2001, 16, 1419. (b) Leumann,
C. Bioorg. Med. Chem. 2002, 10, 841. (c) Meldgaard, M.; Wengel, J. Chem.
Soc., Perkin Trans. 1 2000, 3539.
(20) Frieden, M.; Christensen, S. M.; Mikkelsen, N. D.; Rosenbohm, C.; Thrue,
C. A.; Westergaard, M.; Hansen, H. F.; Orum, H.; Koch, T. Nucleic Acids
Res. 2003, 31, 6325.
(21) (a) Sands, H.; Gorey-Feret, L. J.; Ho, S. P.; Bao, Y.; Cocuzza, A. J.;
Chidester, D.; Hobbs, F. W. Mol. Pharmacol. 1995; 47, 636. (b) Miyao,
T.; Takakura, Y.; Akiyama, T.; Yoneda, F.; Sezaki, H.; Hashida, M.
Antisense Res DeV. 1995; 5, 115. (c) Uhlmann, E.; Peyman, A.; Ryte, A.;
Schmidt, A.; Buddecke, E. Methods Enzymol. 2000, 313, 268. (d) Peyman,
A.; Helsberg, M.; Kretzschmar, G.; Mag, M.; Ryte, A.; Uhlmann, E.
AntiViral Res. 1997, 33, 135.
(22) (a) Pradeepkumar, P. I.; Zamaratski. E.; Foldesi A.; Chattopadhyaya J.
Tetrahedron Lett. 2000, 41, 8601. (b) Pradeepkumar, P. I.; Zamaratski E.;
Foldesi A.; Chattopadhyaya J. J. Chem. Soc., Perkin Trans. 2 2001, 402.
(c) Pradeepkumar, P. I.; Chattopadhyaya J. J. Chem. Soc., Perkin Trans. 2
2001, 2074. (d) Boon, E. M.; Barton, J. K.; Pradeepkumar, P. I.; Isaksson,
J.; Petit, C.; Chattopadhyaya, J. Angew. Chem., Int. Ed. 2002, 41, 3402.
(23) Amirkhanov, N. V.; Pradeepkumar, P. I.; Chattopadhyaya, J. J. Chem. Soc.,
Perkin Trans. 2 2002, 976.
(24) Pradeepkumar, P. I.; Amirkhanov, N. V.; Chattopadhyaya, J. Org. Biomol.
Chem. 2003, 1, 81.
(25) (a) Ossipov, D.; Zamaratski, E.; Chattopadhyaya, J. HelV. Chim. Acta 1999,
82, 2186. (b) Zamaratski, E.; Ossipov, D.; Pradeepkumar, P. I.; Amirkhanov,
N. V.; Chattopadhyaya, J. Tetrahedron 2001, 57, 593.
(26) Opalinska,J. B.; Kalota, A.; Rodriquez, L.; Henningson, H.; Gifford, L.
K.; Lu, P.; Jen, K.-Y.; Paradeepkumar, P. I.; Barman, J.; Kim, T. K.; Swider,
C.; Chattopadhyaya, J.; Gewirtz, A. M. Proc. Natl. Acad. Sci. U.S.A. 2004,
under review.
(16) (a) Tarkoy, M.; Bolli, M.; Schweizer, B.; Leumann, C. HelV. Chim. Acta
1993, 76, 481. (b) Altman, K.-H.; Kesselring, R.; Francotte, E.; Rihs, G.
Tetrahedron Lett. 1994, 35, 2331.(c) Siddiqui, M. A.; Ford, H.; George,
C.; Marquez, V. E. Nucleosides Nucleotides 1996, 15, 235 (d) Marquez,
V. E.; Siddiqui, M. A.; Ezzitouni, A.; Russ, P.; Wang, J.; Wanger, W. R.;
Matteucci, D. M. J. Med. Chem. 1996, 39, 3739. (e) Obika S.; Nanbu D.,
Hari Y., Morio, K.; In, Y.; Ishida, T.; Imanishi, T. Tetrahedron Lett. 1997,
38, 8735. (f) Koshkin, A. A.; Singh, S. K.; Nielson, P., Rajwanshi, V. K.;
Kumar R.; Meldgaard, M.; Wengel, J. Tetrahedron 1998, 54, 3607. (g)
Wang, G.; Giradet, J.-L.; Gunic, E. Tetrahedron 1999, 55, 7707. (h).Wengel,
J. Acc. Chem. Res. 1999, 32, 301. (i) Sekine M, Kurasawa O.; Shohda, K.;
Seio K.; Wada.; T. J. Org. Chem. 2000, 12, 3571. (j) Morita, K.; Hasegawa,
C.; Kaneko, M.; Tsutsumi, S.; Sone, J.; Ishikawa, T.; Imanishi, T.; Koizumi,
M. Bioorg. Med. Chem. Lett. 2002, 12, 73. (k) Imanishi, T.; Obika, S.
Chem. Commun. 2002, 16, 1653. (l) Wang, G. Tetrahedron Lett. 1999, 40,
6343. (m) Sugimoto, I.; Shuto, S.; Mori, S.; Shigeta, S.; Matsuda, A.
Tetrahedron Lett. 1999, 9, 385. (n) Steffens, R.; Leumann, C. HelV. Chim.
Acta 1997, 80, 2426. (o) Nielsen, P.; Petersen, M.; Jacobsen, J. P J. Chem.
Soc., Perkin Trans. 1 2000, 3706. (p) Christiensen, N. K.; Andersen, A.
K. L.; Schultz, T. R.; Nielsen, P. Org. Biomol. Chem. 2003, 1, 3738. (q)
Wengel, J.; Peterson, M. Trends Biotechnol. 2003, 21, 74. (r) Kvaerno, L.;
Wengel, J. J. Org. Chem. 2001, 66, 12, 5498.
(17) Acharya, P.; Cheruku, P.; Chatterjee, S.; Acharya, S.; Chattopadhyaya, J.
J. Am. Chem. Soc. 2004, 126, 2862.
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11485

A R T I C L E S
Pradeepkumar et al.
Figure 2. Sequences and Tms of various heteroduplexes containing oxetane-modified AONs and their target RNA. The Tm and the differences in Tm (∆Tm)
with respect to that of the native phosphodiester AON (1) for 20mer AONs are also shown for comparison.
efficacy5a) have shown that rationally designed oxetane-C
modified AONs were highly efficient both in diminishing the
c-myb mRNA (85% has been reduced) and the c-myb protein
(70% of its expression was found to be halted) of the targeted
gene. On the basis of the amount of AON uptake after delivery,
determined by slot blot, it was apparent that the oxetane
modified AONs are 5-6 times more efficient antisense agents
than those of the corresponding isosequential phosphorothioate
analogue.26
The above-mentioned successful antisense design combining
the oxetane-C units and 3′-DPPZ group prompted us to
investigate the properties of the oxetane-purine modified AONs.
Here, we report the synthesis of the oxetane-A [9-(1′,3′-O-
anhydro-â-D-psicofuranosyl)adenine] and the oxetane-G [9-(1′,3′-
O-anhydro-â-D-psicofuranosyl)guanine] units (Figure 2) as well
as the structure, target affinity, RNase H cleavage, exo and
endonuclease tolerance of AONs (2-9) incorporating these
units. Michaelis-Menten kinetics of E. coli RNase H1 promoted
cleavage has been carried out for the oxetane-A and -G modified
AONs (2 and 3)/RNA (13) hybrids, and the kinetic parameters
are compared with those of the oxetane-C modified (4) and
native AON (1)/RNA (13) hybrids.
exhibited melting temperatures very similar to that of the native
AON (1)/RNA (13) hybrid duplex (Figure 2). However, the
introduction of the triple C modifications as in AON (4)/RNA
(13) causes 4 °C drop in Tm per oxetane-modification while
the introduction of the oxetane-T modifications into the AON
(10)/RNA (13) hybrid makes it even more unstable showing a
Tm drop of 5 °C per oxetane-T modification (Figure 2). This
clearly shows that the oxetane-pyrimidine units have consider-
able destabilizing effect on the thermostability of the AON/
RNA hybrids compared to those of their purine counterparts.
The mixed sequences of the oxetane-A and -G moieties, such
as AONs (5-7), have also shown similar target binding affinities
close to that of the native AON (1) (Figure 2). It should be
noted that Tm enhancement effect of the 3′-DPPZ group which
we have observed earlier in the studies of 9mer and 15mer AON/
RNA hybrids22c,24 has been considerably reduced in the case of
20mer AON/RNA hybrids studied here.
To shed more light on the differential binding affinities of
the oxetane modified AONs to the target RNA (13), we have
elucidated thermodynamic parameters (Table 1), based on the
concentration dependent Tm analysis of the native AON (1)/
RNA (13) duplex and those of the triple oxetane-purine (2 and
3) or pyrimidine-modified AONs (4 and 10). It has emerged
that the thermodynamic parameters of the triple oxetane-G
modified AON (3)/RNA (13) duplex are very close to that of
the native AON (1)/RNA (13) duplex. However, ∆H° and ∆S°
of the triple oxetane-A modified AON (2)/RNA (13) duplex
are slightly lower (5.5 kJ/mol drop in ∆G°) than those of the
Results and Discussions
(A) Binding Affinity of the Oxetane-A and -G Modified
AONs to the Target RNA and the Thermodynamic Proper-
ties of the Modified AON/RNA Hybrids. When tested for
target affinity, the triple oxetane-A modified AON (2)/RNA (13)
and the triple oxetane-G modified AON (3)/RNA (13) have
9
11486 J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
Table 1. Thermodynamic Parameters of 20Mer AON (1-4 and
oxetane-G (3) modified AONs offered no resistance toward the
cleavage by the enzyme (Figure 3, Figure S2 in Supporting
Information). The half-life of the AON (7) with the three A
and three G units was found to be about 2 h, which is about
half of that observed for the triple oxetane-C modified AON
(4) studied under similar conditions. These results reveal the
high susceptibility of the oxetane-purine modified AONs toward
the DNase 1 promoted cleavage.
-1
10)/RNA (13) Hybrids Determined from Tm vs ln(CT /4) Plots
−∆G°
Tm
−∆H°
−∆S°
(298 K) ∆∆G°
AONs
(°C)
(kJ/mol)
(J/K‚mol)
−T∆S° (kJ/mol) (kJ/mol)
native 20mer (1) 74 740.4 ( 22.3 2016 ( 64 600.9 139.5
20-3A (2)
20-3G (3)
20-3C (4)
20-3T (10)
73 714.2 ( 22.5 1946 ( 65 580.2 134.0 -5.5
74 747.7 ( 13.9 2034 ( 40 606.4 141.3 +1.8
61 661.5 ( 13.5 1860 ( 40 554.5 107.0 -32.5
59 646.6 ( 20.5 1853 ( 62 544.5 102.3 -37.2
To date, only very limited information is available for the
structural and cleavage characteristics of DNase 1 with single
stranded AONs. Although the DNase 1 cleavage has limited
sequence specificity,27 the crystal structure of the DNase 1 with
various short oligonucleotide duplexes,28 has revealed that in
addition to the minor groove width and DNA flexibility, the
local sequence preference is crucial for the proper alignment
of the phosphodiester bond for the cleavage reaction. It has
emerged that three nucleotides toward the 5′- and 3′-end of the
cleavage site have significant influence on the cleavage proper-
ties. On the basis of the cleavage characteristics of various DNA
duplexes, Herrera and Chaires29 found that among those 6
nucleotides around the cleavage site, the nucleotide sequences
at position 3 (denoted as -3) and 2 (denoted as -2) toward
the 5′-end (Figure 3) and nucleotide at position 2 (denoted as
+2) toward the 3′-end of the cleavage site are very crucial in
determining the major cleavage sites and cleavage rates of DNA
duplexes by DNase 1.29 This is because the NH2 of G at position
-3 cause steric clash with arginine 41 of the enzyme and
hamper the cleavage rate.27,29 Tyrosine-76 has shown favorable
interactions with T/C moieties at the position -2. At the position
+2 toward 3′-end presence of T is highly disfavored.28,29
Our results can be partly explained based on the local
sequence requirements described above. In the native AON (1),
the cleavage sites have been restricted to the 5′-end of the oligo
and the major cleavage site was found to be at G9, which is in
accordance with the aforementioned general requirements
(Figure 3, Figure S2 in Supporting Information). In all of the
oxetane-modified AONs, the major and minor cleavage sites
were observed at one or two nucleotides away from the modified
sites. This clearly shows that the oxetane ring in the nucleotides
interferes with the catalytic core of the enzyme and prevents
the cleavage. Also, the modified residues might alter the
flexibility and the geometrical requirements of phosphate
required for the cleavage. In the triple A modified AON (2) all
the cleavage sites of the native AON (1) are retained because
there is no oxetane modification present at those cleavage sites.
This explains why the extent of cleavage in the triple A modified
AON (2) is very similar to that of the native AON (1). In the
triple G modified AON (3) the major cleavage site has been
shifted to the T8 instead of G9, which is probably because of
the presence of the oxetane-G at G11 position (Figure 3, Figure
S2 in Supporting Information). But the oxetane-G at +3 did
not affect the cleavage rate. Probably, the enzyme might not
have encountered the perturbations caused by the additional
oxetane-ring in the sugar, which is being located far away from
the nucleobase. However, in the case of pyrimidine-oxetane
units, the oxetane ring is close to the pyrimidine ring, which
might interfere with the enzyme. This is evident in the triple
native duplex and the triple oxetane-G modified hybrid duplexes.
The triple oxetane-C modified AON (4)/RNA (13) as well as
the triple oxetane-T modified AON (10)/RNA (13) duplex have
shown large enthalpic destabilization and slight entropic stabi-
lization, which resulted in the net loss of 32.5 kJ/mol in ∆G°
for the triple C and 37.2 kJ/mol net loss for the triple T modified
hybrids compared to that of the native duplex. Apparently, the
large drop in enthalpy of the oxetane-modified pyrimidine AON/
RNA duplexes contributes to the destabilization of the duplex
which reflects in the weakening of the stacking and hydrogen
bonding interactions. But it is not clear which force is
dominantly disturbed during the duplex formation. The nucleic
acid mediated charge transport measured by chronocoulometry22d
has revealed that the stacking/dynamics perturbations owing to
the incorporation of the oxetane-T unit in a DNA/RNA duplex
are similar to those of the native hybrid duplex,22d while the
complete charge attenuation was observed for the oxetane-T
incorporated DNA/DNA homoduplex. However, this better
accommodative nature of North-East constrained oxetane-T unit
in the DNA/RNA duplexes is not reflected in its thermostability
which is somewhat poorer (Tm drop of 6 °C/modification) than
that of the native duplex. Two probable reasons for the
weakening of thermostability could be (i) perturbation of the
intrastrand stacking because of the North-type sugar among the
neighboring South-type sugars along the DNA strand, or (ii) a
possible steric clash between the oxetane ring and the hydrogen-
bonding center of the aglycon, which is more pronounced in
the pyrimidines than in purines (for example the distance be-
tween C1′ to H-bonding center N3H is 4.9 Å in oxetane-T while
the distance between C1′ to H-bonding center N1 in oxetane-A
is 5.4 Å). Clearly, more studies are needed, some of which are
under way in our laboratory, to elucidate the reasons dictating
the stability of the oxetane-modified AON/RNA hybrids.
(B) CD Spectra of the Oxetane-Modified AON/RNA
Hybrids. Since circular dichroism (CD) spectroscopy is an
effective tool to measure the global helical conformation of
nucleic acids, we have recorded CD spectra of the oxetane-
modified AON/RNA hybrids (Figure S1 in Supporting Informa-
tion). Similar to the results obtained in our previous studies22
with the oxetane-T modified 15mer AON/RNA hybrids, it was
found that the introduction of the oxetane-A and oxetane-G units
into the 20mer AONs (2 and 3) has not altered the global
conformation of the duplexes compared to that of the native
AON (1)/RNA (13) duplex (Figure S1 in Supporting Informa-
tion). From these results, it is evident that CD has failed to detect
the local conformational changes brought by the oxetane
modification(s), which has in fact been sensed by RNase H as
it can be seen from its cleavage pattern (vide infra).
(C) Endonuclease, Exonuclease, and Human Serum Sta-
bility of Oxetane-Modified AONs. When treated with the
endonuclease (DNase 1), the triple oxetane-A (2) and triple
(27) Fox, K. R. Methods Mol. Biol.; Fox, K. R., Ed.; Humana Press Inc.: Totowa,
NJ, 1997; Vol. 90, p 1.
(28) Suck, D. J. Mol. Recognit. 1994, 7, 65.
(29) Herrera, J. E.; Chaires, J. B. J. Mol. Biol. 1994, 236, 405.
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11487

A R T I C L E S
Pradeepkumar et al.
Figure 3. Cleavage pattern of the DNase 1 degradation of AON (1)-(4) and (7) by the PAGE analysis. The % of AON left after 4 h of incubation: 0%
of AON (1), 0% of AON (2), 0% of AON (3), 50% of AON (4), and 30% of AON (7). The vertical arrows show the DNA cleavage sites and the relative
length of an arrow shows the relative extent of cleavage at that site.
oxetane-C modified AON (4), where only minor cleavage was
observed at the G9 along with the G11 while the major cleavage
was observed at the C3 and it is very close to the 5′-end.
However, the rate of cleavage was low at those sites probably
owing to the perturbations of the AON-enzyme interaction by
the oxetane-C, which is situated just three nucleotide away from
the cleavage site. The AON (7) incorporated with 6 oxetane-
purines has shown a half-life of the cleavage the cleavage to
be about 2 h. These results lead us to suggest that to get effective
endonuclease resistance with the oxetane-purine units, extensive
(probably every alternate nucleotide) modifications of the AON
strand are needed. This, however, may slow the RNase H
recruiting capability (vide infra). Another viable approach can
be the use of a mixture of the oxetane-purines and a minimum
amount of the oxetane-pyrimidines.
To investigate the tolerance of the oxetane-modifications
toward an exonuclease (snake venom phosphodiesterase,
SVPDE), we have synthesized an AON (12) with 3 consecutive
oxetane-C, -A, and -G units at the 3′-end. Incubation with the
SVPDE showed that the presence of the oxetane-modified units
at the 3′-end offers resistance toward cleavage: 36% of AON
(12) was left after 24h incubation with the enzyme, while the
native PO-AON (1) was completely degraded in less than 2 h
of incubation (Figure S3 in Supporting Information). The same
AON showed half-life of around 9 h in human blood serum
where major degrading enzymes are exonucleases (Figure S4
in Supporting Information). Complete exonuclease protection
was achieved in case of the 3′-DPPZ conjugated AONs (8 and
9, (Figure S3 in Supporting Information), which were found to
have slightly better stability than the native PS-AON in human
serum (Figure S4 in Supporting Information).
(D) RNase H1 Cleavage Pattern of the Complementary
RNA and the Extent (%) of RNA Cleavage in Hetero-
duplexes Containing Oxetane-A, -G, -C Units and Their
DPPZ Conjugates. All of the oxetane-A and -G modified AON
(2-7)/RNA (13) hybrids and their DPPZ conjugates (8 and 9)
were found to be substrates for the Escherichia coli RNase H1
with varying cleavage potential. In the RNase H1 cleavage
pattern of all the oxetane-A and -G modified AON/RNA hybrids,
except for the AON (7)/RNA (13), a region of 5 nucleotides in
the RNA strand toward the 3′-end from the site opposite to the
oxetane modification, was found to be resistant toward RNase
H promoted cleavage (Figures 4 and 5). This is presumably
owing to the local steric and structural alterations brought about
by the modification, thereby preventing the flexibility and
accessibility required for the RNase H cleavage.3 This 5-nucle-
otide footprint is identical to what was found earlier for the
oxetane-T and -C modified AON/RNA hybrids.22-24 The
cleavage of the AON (7)/RNA (13) at A11 site was surprising
because this cleavage site is located within the anticipated
footprint region of the oxetane-A modification (Figure 4).
Our earlier studies of the oxetane-T and -C modified AON/
RNA hybrids revealed that the RNase H cleavage potentials of
these modified hybrids depend on the concentration of the RNA
as a substrate.22b,23,24 This has prompted us to test all our
modified AONs at two different substrate (RNA) concentrations.
We have deduced the extent of cleavage at high (0.8 µM) and
low (0.01 µM) RNA concentrations (Figure 6) under saturation
9
11488 J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
Figure 4. (A) Autoradiograms of 20% denaturing PAGE, showing the cleavage kinetics of 5′-32P-labeled target RNA (13) by E. coli RNase H1 in the native
AON (1)/RNA (13) and the oxetane-A, -G, and -C modified AON (2-4 and 7)/RNA (13) hybrid duplexes. Lanes 1 to 5 represent the aliquots of the digest
taken after 5, 10, 15, 30, and 60 min, respectively. Conditions of cleavage reaction: RNA (0.8 µM) and AONs (4 µM) in buffer, containing 20 mM Tris-HCl
(pH 8.0), 20 mM KCl, 10 mM MgCl2 and 0.1 mM DTT at 21 °C; 0.8 U of RNase H. Total reaction volume was 30 µL. The % of RNA left intact after 1
h of incubation: 28% for AON (1), 36% for AON (2), 50% for AON (3), 26% for AON (4) and AON for 32% (7). (B) RNase H1 cleavage pattern of hybrid
duplexes. The vertical arrows show the RNA cleavage sites and relative length of an arrow shows the relative extent of cleavage at that site.
Figure 5. (A) Autoradiograms of 20% denaturing PAGE, showing the cleavage kinetics of 5′-32P-labeled target RNA (13) by E. coli RNase H1 in the native
AON (1)/RNA (13) and the oxetane-A, -G, and -C modified AON (5, 6, 8, and 9)/RNA (13) hybrid duplexes. Lanes 1 to 5 represent the aliquots of the digest
taken after 5, 10, 15, 30, and 60 min respectively. Conditions of the cleavage reaction: RNA (0.8 µM) and AONs (4 µM) in buffer, containing 20 mM
Tris-HCl (pH 8.0), 20 mM KCl, 10 mM MgCl2 and 0.1 mM DTT at 21 °C; 0.8 U of RNase H. Total reaction volume was 30 µL. The % of RNA left after
1h of incubation: 34% for AON (1), 65% for AON (5), 64% for AON (6), 37% for AON (8) and 64% for AON (9). (B) RNase H cleavage pattern of hybrid
duplexes. The vertical arrows show the RNA cleavage sites and relative length of an arrow shows the relative extent of cleavage at that site.
conditions (the enzyme is completely saturated by the substrate).
At the low RNA concentration, the triple oxetane-A (2) and
triple oxetane-G (3) modified AON/RNA hybrid duplexes and
their respective DPPZ counterparts (8) and (9) have shown the
extent of the cleavage slightly lower than that of the native AON
(1)/RNA (13) hybrid duplex (Figure 6). The mixed A and G
incorporated AON (5 and 6)/RNA (13) hybrids have also shown
similar cleavage characteristics. Exception was found for the
AON (7)/RNA (13), incorporated with 6 oxetane-purine
moieties, where the extent of cleavage was extremely low (17
( 3%) compared to that of the native AON (1)/RNA (13) hybrid
duplex. It should be noted that under these conditions the triple
oxetane-C modified AON (4)/RNA (13) hybrid duplex has
shown a comparable or slightly better extent of cleavage (78 (
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11489

A R T I C L E S
Pradeepkumar et al.
Figure 8. Bar graphs showing the relative (with respect to the native AON)
Vmax, 1/Km and Vmax/Km values for the RNase H promoted RNA hydrolysis
in AON/RNA hybrid duplexes (Figure 2) formed by the native 20mer AON
(1), oxetane-A, -G, and -C modified AONs (2-4).
Figure 6. Bar graphs showing the extent (%) of the E. coli RNase H1
promoted target RNA (13) hydrolysis in AON/RNA hybrid duplexes after
1 h incubation with the enzyme for native 20mer AON (1) and the oxetane-
A, -G, and -C modified AONs (2-9) at high (0.8 µM) and low (0.01 µM)
substrate (RNA) concentration. Conditions of cleavage reactions: RNA (0.01
or 0.8 µM) and AONs (4 µM) in buffer, containing 20 mM Tris-HCl (pH
8.0), 20 mM KCl, 10 mM MgCl2 and 0.1 mM DTT at 21 °C; 0.08 U of
RNase H for reactions under low RNA concentration and 0.8 unit of enzyme
for reactions under high RNA concentration. Total reaction volume was
30 µL.
oxetane-G modified AON (3), its 3′-DPPZ containing counter-
part (9) and the AONs (5 and 6) containing both the oxetane-A
and oxetane-G units have shown relatively low cleavage by the
RNase H (Figure 6). Surprisingly, the AON (7)/RNA (13)
hybrid, which had shown very low cleavage at low RNA
concentration, showed RNase H cleavage close to that of the
native hybrid duplex. This underscores the differential cleavage
characteristics of the oxetane-purine modified AON/RNA
hybrids at different substrate concentrations.
(E) Michaelis-Menten Kinetics of the Oxetane-A and -G
Modified AON/RNA Hybrids and Their Comparison with
the Oxetane-C and Native AON/RNA Hybrid Duplexes. To
obtain a deeper understanding of the E. coli RNase H1
recognition and cleavage of the oxetane-A and -G modified
AON (3 and 4)/RNA (13) hybrids, a detailed Michaelis-Menten
kinetics has been carried out and the kinetic parameters were
subsequently compared with those of the native (1) and the triple
oxetane-C modified AON (4)/RNA (13) hybrids (Table 2,
Figures 7 and 8).
It is evident from the kinetic data that the triple oxetane-A
modified AON (2)/RNA (13) duplex has a maximum velocity,
V
max, slightly lower than that of the native (1) and the triple
oxetane-C modified AON (4)/RNA (13) hybrids. However, its
enzyme binding affinity, 1/Km, has found to be very close to
the native yielding an effective enzyme activity, kcat/Km, close
to that of the native. The triple oxetane-G modified AON (3)/
RNA (13) has shown a lower Vmax and Km value compared to
those of the native, triple oxetane-A and the triple oxetane-C
modified AON/RNA hybrids (Table 2, Figures 7 and 8). Thus,
the kcat/Km value of oxetane-G modified AON (3)/RNA (13) is
half of that of the native AON (1)/RNA (13) duplex. Noteworthy
the fact that the Vmax value appeared to be higher for the triple
oxetane-C modified AON/RNA duplex than for the native
(Table 2, Figures 7 and 8). This can be attributed the lower
binding affinity (1/Km) and lower Tm of the triple C modified
AON/RNA duplex similar to the observations made in our
previous studies on various oxetane-C and -T modified 15mer
AON/RNA hybrids.23,24 The reason for the lower Vmax and high
1/Km of the triple oxetane-G modified AON/RNA duplex is not
Figure 7. Initial velocity of the hydrolysis (Vo) of the 20mer target RNA
(13) in the AON (1-4)/RNA hybrids (see Figure 2) by E. coli RNase H as
a function of the RNA concentration (logarithmic scale). Conditions of the
cleavage reaction: AONs (4 µM) and RNA (13) in buffer containing 20
mM Tris-HCl (pH 8.0), 20 mM KCl, 10 mM MgCl2 and 0.1 mM DTT at
21 °C, 0.08 U of RNase H. Total reaction volume was 30 µL.
1%) than that of the native hybrid duplex (75 ( 1%). These
data clearly show that the triple oxetane-A and triple oxetane-G
modified AON/RNA hybrids are capable of recruiting RNase
H in an efficient manner (52-60% of RNA was cleaved) at
low substrate concentrations, and it is important in view of
extremely low concentrations of target mRNAs in many infected
cells.30
At high substrate concentration (0.8 µM RNA), the hybrid
duplexes of the triple oxetane-A modified AON (2) and its DPPZ
counterpart (8) showed the extent of cleavage comparable to
that of the native (1) and triple C (4) modified AON/RNA
hybrids (Figure 6) However, the hybrid duplexes of the triple
(30) Baer, M. R.; Augustinos, P.; Kinniburgh, A. Blood 1992, 79, 1319.
9
11490 J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
Table 2. Kinetic Characteristics of the RNA Cleavage by E. coli RNase H1 of the AON (1-4)/RNA (13) Hybrid Duplexes
V
max, 10-3
Km, 10-3
µM
kcat
,
(Vmax/Km),
min-1
(kcat/Km),
µM-1 min-1
relative
(kcat/Km)
AONs
Tm (°C)
µM min-1
min-1
native 20mer (1)
74
7.3 ( 0.3
44.4 (6.3
38.3 ( 6.4
25.6 ( 4.1
70.2 ( 13.8
24.2 ( 1
0.164
0.15
545.1
501.3
246.1
383.7
1
3A-oxetane-modified AON
20-3A (2)
73
74
61
5.8 ( 0.2
1.9 ( 0.1
8.1 ( 0.5
19.2 ( 0.6
6.3 ( 0.35
26.9 ( 1.5
0.92
0.45
0.70
3G-oxetane-modified AON
20-3G (3)
0.074
0.11
3C-oxetane-modified AON
20-3C (4)
a Vmax ) kcatE0 (E0 ) 0.08 U/30 µL )2.66 × 10-3 U/µL ) 3.016 × 10-4 µM), specific activity ) 420 000 U/mg or 2.38 × 10-6 mg/U ) 1.133 86 ×
10-13 mol/unit, MW ) 21 000 g/mol.
Scheme 1 a
a Reagents and conditions: (i) 0.4 M methanolic HCl, r.t., 12 h; (ii) MsCl, pyridine, 4 °C, 12 h; (iii) 90% TFA/water, r.t., 20 min (iv) acetic anhydride,
pyridine, r.t., 3 h; (v) 30% HBr/AcOH, 0 °C to r.t., 12 h; (vi) persilylated N6-benzoyl-adenine, CH3CN, SnCl4, 70 °C, 4 h, r.t., 12 h for 18a; (vii) persilylated
N2-acetyl-O6-diphenylcarbamoylguanine, CH3CN-DCE, SnCl4, 70 °C, 5 h, r.t., 12 h followed by 90% TFA/water, r.t., 20 min for 18b; (viii) methanolic NH3,
r.t., 48 h; (ix) TIPDS-Cl2, pyridine, 4 °C, 30 min, r.t., 2 h; (x) NaHMDS, THF, 4 °C, 2 h; (xi) PAC-Cl, pyridine, 4 °C, 2 h., r.t., 2 h for 21a; (xii)
N,N-dimethylformamide dimethylacetal, MeOH, r.t., 12 h for 21b; (xiii) 1M TBAF, THF, r.t., 4 min; (xiv) DMTrCl, pyridine, r.t., 12 h; (xv)
(2-cyanoethoxy)bis(N,N-diisopropylamino)phosphine, N,N-diisopropylammonium tetrazolide, DCM, r.t., 2 h. Abbreviations: Tol ) 4-methylbenzoyl, r.t. )
room temperature, Ms ) methanesulfonyl, TFA ) trifluoroacetic acid, Ac ) acetyl, DCE ) dichloroethane, TIPDS ) 1,1′,3,3′-tetraisopropyldisiloxane-
1,3-diyl, NaHMDS ) sodium bis(trimethylsilyl) amide, PAC ) phenoxyacetyl, TBAF ) tetrabutylammonium fluoride, THF ) tetrahydrofuran, DMTr )
4,4′-dimethoxytrityl, DCM ) dichloromethane.
clear. Clearly, more modified sequences have to be tested before
drawing any general conclusions regarding the RNase H
recruitment by the oxetane-G and -A modified AON/RNA
hybrids.
benzoyladenine and N2-acetyl-O6-diphenylcarbamoylguanine in
the presence of SnCl4 as Lewis acid catalyst. This afforded the
â-nucleosides 18a; and yielded 18b (together with the short
treatment with TFA to remove the DPC: 25%). The yield of
two step transformation from 16 to 18 was 42% compared to
95% reported.31 Along with 18a we got very little amount of R
anomers with a lot of nonnucleosidic impurities. This has
reduced the overall yield. For 18b we got inseparable mixtures
of N7 R and â nucleosides along with inseparable mixtures of
N9 (both R and â) and N7 (both R and â) psiconucleosides
without the DPC protection. Moreover, there also were non-
nucleosidic impurities. Unfortunately, due to the use of repeated
columns for purification we could not quantify the yields of all
those side products. The â configurations of the products were
determined by NOE measurements. NOE observed between H-8
and H-3′, which are separated by 1.9 Å, in 18a was 2%. In the
case of 18b NOE between H-8 and H-3′, was found to be 2.7%.
Such kind of NOE cannot be observed in the R anomers owing
to the 3.9 Å distance between H-8 and H-3′. The N9-
connectivity of the base in 18b was evidenced by the chemical
shift of C5 of the base at 122.0 ppm in the 13C NMR spectra,
(F) Synthesis of the Oxetane-A and -G Building Blocks.
The oxetane-A and oxetane-G phosphoramidites, 23a and 23b,
required for incorporation into AONs (2-9 and 12) were
synthesized using a strategy shown in Scheme 1. The synthesis
started with the conversion of 6-O-(4-toluoyl)-1,2:3,4-diisopro-
pylene-â-D-psicofuranose (14),22b into 16, a possible key
intermediate in the nucleobase coupling, using a reported
procedure.31 It is noteworthy that some yields obtained in our
hands were significantly lower in certain steps compared to that
reported earlier.31 Thus, the transformation from 14 to 15
(Scheme 1) yielded in only 60% compared to the 91%
reported.31 Since the attempts to couple the silylated purine
nucleobase derivatives with 16 were unsuccessful, we have
decided to transform 16 into the bromosugar 17 followed by
immediate coupling of the crude 17 with the silylated N6-
(31) Roivainen, J.; Vepsalainen, J.; Azhayev, A.; Mikhailopulo, I. A. Tetrahedron
Lett. 2002, 43, 6553.
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11491

A R T I C L E S
Pradeepkumar et al.
3
Figure 9. Calculated and experimental values of the JH,H vicinal proton coupling constants (Hz) (see Table S1 in Supporting Information). Theoretical
dependences (color lined in frames (A)-(C)) of the 3JH,H were calculated by varying the phase angle of pseudorotation (P) from 0° to 360° using generalized
Karplus-type equation (P1 ) 13.24, P2 ) -0.91, P3 ) 0, P4 ) 0.53, P5 ) -2.41, P6 ) 15.5°) developed by Altona et al.36 at fixed puckering amplitudes
of 30° (blue dashed line), 35° (red dot-dashed line), 40° (black line), and 45° (black dotted line). The experimental coupling constants for the oxetane-
1
modified nucleosides (1a)-(1d) were obtained through 600 MHz H NMR spectra (DMSO-d6 + CD3OD) at 298 K. The experimental coupling constants
of the North-type sugar (2a),38 (3a),39 (3b),16e and (4a),16b and the South-type sugar (5a)40 and (6a)42 compounds are presented from available NMR and
X-ray data. Note that the numbering of the sugar atoms used in this figure differs from the nomenclature defined numbering (Scheme 1) employed throughout
the paper.
which is a characteristic for the N9-isomer.32 Methanolysis of
18a-b followed by simultaneous protection of 4′ and 6′
hydroxyl groups using 1,3-dichloro-1,1′,3,3′-tetraisopropyl-
disiloxane (TIPDSCl2) furnished the precursors, 19a (72%) and
19b (63%), required for the oxetane ring formation. The
cyclization was achieved by the treatment of 19a-b with
NaHMDS in THF at 4 °C, giving 20a (81%) and 20b (83%).
The exocyclic amino group of the 20a was protected by
phenoxyacetyl group and that of 20b by dimethylformamidine
group to furnish 21a (60%) and 21b (91%). The removal of
the TIPDS group from 21a-b followed by one-pot dimethoxy-
tritylation afforded 22a (80%) and 22b (83%). Compounds
22a-b were transformed to their corresponding phosphor-
amidites 23a (60%) and 23b (75%), the building blocks required
for the solid-phase synthesis of the modified AONs (2-9 and 12).
(G) Conformational Analysis of the Oxetane-Fused Sugar
Moiety. Conformational Analysis Based on Experimental
3JH,H Coupling Data. The conformation adopted by the pentose
in the bicyclic (1′,3′-O-anhydro-â-D)-furanosyl oxetane-modified
nucleosides (compounds 1a-1d in Figure 9, frame D), is not
possible to estimate directly from the vicinal proton-proton
coupling constants due to the lack of an important endocyclic
3JH-2′,H-3′ (3JH-1′,H-2′ in Figure 9) coupling. But a fair amount
of information about this conformation can be extracted from
3
3
the experimental JH-3′,H-4′ and JH-4′,H-5′ (referred to as
3JH-2′,H-3′ and 3JH-3′,H-4′ in Figure 9) coupling constants. Using
generalized Karplus-type equation developed by Altona et
al.,36,37 a set of empirical rules relating 3JH,H′ coupling constants
of pentose can be used in analysis of the vicinal couplings. It is
3
well established that a high or low value of the JH-1′,H-2′ in
3
(32) Garner, P.; Ramakanth, S. J. Org. Chem. 1988, 53, 1294.
conjunction with a low or a high value of JH-3′,H-4′, respec-
9
11492 J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
3
Table 3. Proton Chemical Shifts (δ Scale, ppm) and JH,H Coupling Constants (Hz) for the Oxetane-C (1a),a -T (1b),a -A (1c), and -G (1d)
(1H NMR spectra were recorded in DMSO-d6 with methanol-d4 at 600 MHz using DMSO-d6 (2.6 ppm) peak as internal standard)
1
a The H NMR data for the oxetane-C (1a) in D2O and the oxetane-T (1b) in methanol-d4 have been reported in refs 24 and 22b, respectively.
tively, are considered to be strong indicators of the South- or
the North-type conformation (see the 3JH-1′,H-2′ and 3JH-3′,H-4′
dependencies on the pseudorotational phase angle P in frames
A and B in Figure 9) as it is shown for the typical North-type
conformation for compounds 2a (3JH-1′,H-2′ ) 0.0 Hz, P )
3JH-3′,H-4′ (frame C, Figure 9) shows the combination of these
coupling constants for the oxetane-modified nucleosides to be
typical for P ) 20-60° range which unambiguously points to
the North-conformation as opposed to South-type conformation
3
for compound 6a (Figure 9, frames C and D). The JH-2′,H-3′
3
16.5°),38 3a (3JH-1′,H-2′ ) 0.0 Hz, P ) 17°),39 3b (3JH-1′,H-2′
)
)
and JH-3′,H-4′ (see numbering in Figure 9) of the oxetane-C,
3
0.0 Hz, P ) 17.4°),16e 4a (3JH-1′,H-2′ ) 0.0 Hz, JH-1′,H-2′′
-T, -A, and -G nucleosides have been found to be temperature-
6.8 Hz, P ) 343°)16b and the typical South-type conformation
independent, thus the H NMR data suggest that the sugar
1
for compounds 5a (3JH-1′,H-2′ ) 7.5 Hz, JH-2′,H-3′ ) 0.0 Hz,
moiety of the oxetane-modified nucleosides are locked in the
North-type conformation.16e-f,16h,43,44
3
P ) 136°)40 and 6a (3JH-2′,H-3′ ) 3.0 Hz, 3JH-2′′,H-3′ ) 5.4 Hz,
3JH-3′,H-4′ ) 0.0 Hz, P ) 190°)42 (molecular structures and atom
numbering are shown in Figure 9, frame D).
Ab Initio Calculations. In an alternative analysis, the ab initio
geometry optimizations for each of the oxetane modified
nucleosides 1a-1d (Figure 9, frame D) were performed to
obtain the geometrical parameters. It was followed by the 0.5
ns MD simulations in the explicit aqueous medium to probe a
conformational hyperspace available for these systems. The ab
initio calculations utilizing 6-31G* Hartree-Fock geometry
optimization by Gaussian 98 program45 resulted in the H3′-
C3′-C4′-H4′ and H4′-C4′-C5′-H5′ torsions to be ∼43.4°
and ∼163.5°, respectively and these values varied only by (0.4°
for all the oxetane-A, -G, -C, and -T nucleosides 1a-1d (Table
S1 in Supporting Information). The positions of the heavy atoms
of sugar moieties of the respective oxetane-modified nucleosides
vary slightly between the nucleosides resulting in less than 0.01
Å RMS difference as it can clear be seen in Figure 10, frame
A. The theoretical vicinal coupling constants estimated using
these H3′-C3′-C4′-H4′ and H4′-C4′-C5′-H5′ torsions and
the generalized Karplus equation36 agree well with the experi-
mental 3JH,H constants (max RMSd ) 0.46 Hz, see Table S1 in
The JH-3′,H-4′ coupling (3JH-4′,H-5′ in the nomenclature of
3
the oxetane-modified nucleosides) of >∼8 Hz, along with
3JH-1′,H-2′ of <∼1.5 Hz, would indicate that the pentose-sugar
is locked in the North-type conformation (Figure 9, frame B)
and, in contradistinction, a 3JH-3′,H-4′ < ∼1.0 Hz and 3JH-1′,H-2′
of >∼8 Hz would finger-point a constrained South-type
conformation (the exact limits depend on the electronegativity
of the substituents and the substitution pattern)36,37,41 Thus, a
3JH-3′,H-4′ ≈ 0 Hz for the compound 6a alone indicates the
South-type structure (Figure 9, frame D) which agrees with its
available X-ray structure.42 The 3JH-4′,H-5′ values of the oxetane-
modified compounds 1a-1d (Figure 9, frame D) are ∼8.6 Hz
(Table 3) which leads to conclusion that the sugar moieties of
the oxetane-C, -T, -A, and -G nucleosides are locked in the
3
North-type conformation. The dependence JH-2′,H-3′ versus
(33) Nielsen, P. E. Curr. Opin. Mol. Ther. 2000, 2, 282.
(34) Gryaznov, S. M. Biochim. Biophys. Acta 1999, 1489, 131.
(35) Summerton, J.; Weller, D. Antisense Nucleic Acid Drug DeV. 1997, 7, 187.
(36) Haasnoot, C. A. G.; DeLeeuw, F. A. A. M.; Altona, C. Tetrahedron 1980,
36, 2783, and references therein.
(43) Ravn, J.; Nielsen, P. J. Chem. Soc., Perkin Trans. 1 2001, 985
(44) Tarko¨y, M.; Bolli, M.; Schweizer, B.; Leumann, C. HelV. Chim. Acta 1993,
76, 481.
(37) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc. 1972, 94, 8205, and
references therein.
(45) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, Jr., J. A.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck,
A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.;
Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.;
Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.;
Head-Gordon, M.; Replogle E. S.; Pople, J. A. Gaussian 98, Revision A.6;
Gaussian, Inc.: Pittsburgh, PA, 1998.
(38) Wang, G.; Gunic, E.; Girardet, J.-L.; Stoisavljevic, V. Bioorg. Med. Chem.
Lett. 1999, 9, 1147.
(39) Nielsen, C. B.; Singh, S. K.; Wengel, J.; Jacobsen, J. P. J. Biomol. Struct.
Dyn. 1999, 17, 175.
(40) Obika S.; Morio, K.-i.; Hari, Y.; Imanishi, T. Chem. Commun. 1999, 2423.
(41) Thibaudeau, C.; Chattopadhyaya, J. Stereoelectronic Effects in Nucleosides
and Nucleotides and their Structural Implications; Department of Bioor-
ganic Chemistry, Uppsala University Press (jyoti@boc.uu.se): Sweden,
1999 (ISBN 91-506-1351-0).
(42) Altman, K.-H.; Imwinkelried, R.; Kesselring, R.; E.; Rihs, G. Tetrahedron
Lett. 1994, 35, 7625.
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11493

A R T I C L E S
Pradeepkumar et al.
B2-E2) with the average RMSd being max 0.10 ( 0.03 Å
along the trajectories and the major movements are observed
for the aglycon part of the oxetane- modified nucleosides (Figure
10, frames from B1, B2 to E1, E2).
(H) Lessons for Future Design of AONs of Therapeutic
Potentials. To show bona fide antisense action the AONs should
fulfill broad spectrum of requirements. These include sequence
specificity, high exo and endonuclease resistance, RNase H
recruitment capability, deliverability, low toxicity, and favorable
pharmacokinetics. Achieving all of these goals in a balanced
manner by chemical modification of AONs has indeed proven
to be a challenging task. The off-target effects and nonantisense
properties associated with phosphorothioate AONs5 pose serious
concerns over its therapeutic utilities. Other backbone modifica-
tions such as PNA,33 N3′-P5′ phosphoramidates34 and mor-
pholino AONs35 were less specific and unable to activate RNase
H degradation of the target RNA. Hence, any potential
modifications without altering the phosphodiester backbone of
AONs haVe been opted as a Viable alternatiVe. Among these,
the modifications of the sugar moieties of the nucleosides have
been intensively pursued in recent years. However, the loss of
RNase H action by extensive modifications of carbohydrate
residues such as 2′-O-CH3,9 2′-O-CH2-CH2-O-CH3,9 2′-O-
CH2-CH2-CH2-NH2,7,10 LNA16f etc. in an oligochain limits
the antisense potential of these constructs. The “gapmer”
design,7 utilizing such modifications at the ends of the oligo
chain and keeping a long stretches of 6-8 unmodified residues,
was helpful to partly or fully regain the RNase H recruitment.
But such designs will probably be less effective in invoking
the antisense action in vivo considering the endonuclease
vulnerability observed for some of these gapmers.21 A recent
report20 showed the utility of R-L-LNA to partly circumvent
this problem. A state-of-the-art design would consist of a
minimum number of modifications and a maximum antisense
action using a minimum amount of antisense reagent, which
can only be exploited by employing the catalytic cleavage
property of RNase H. Our studies with the oxetane-C modified
mixmer AONs have proven to be in the right direction toward
achieving these goals.26 We have demonstrated recently in a
cellular system the effective down-regulation of the target RNA
using the oxetane-C modified AONs conjugated with the
nontoxic DPPZ group.26 The fully phosphodiester AONs of such
design were in fact ∼5 to 6 times more effective in target
silencing than the corresponding phosphorothioate AONs.26 One
inherent problem with the oxetane-C modified AONs was the
loss of thermodynamic stability associated with the modification,
even though it has partly or fully been regained by the DPPZ
conjugation to short oligomer sequences. The loss of the target
affinity can be bypassed by modifications of AONs by the
oxetane-A and oxetane-G units, as reported here. The oxetane-
modified AONs, unlike many other conformationally con-
strained nucleic acids, require only 4 nucleotide gaps to achieve
efficient target cleavage by RNase H. Since the flexibility of
AON/RNA hybrids is perhaps decisive in dictating their RNase
H cleavage rates,3 it is particularly important that three to four
oxetane modifications do not hamper the hybrid flexibility
required for the RNase H cleavage, as it is shown in our studies.
At low RNA concentration (10 nM) both the oxetane-A and -G
modified AONs and their DPPZ conjugates were found to be
good substrates for RNase H with extent of cleavage comparable
Figure 10. Molecular structures of the oxetane-modified nucleosides shown
from the top (A1-E1) and rotated clockwise by 90° around the y-axis (A2-
E2). Projections A2-E2 show that the oxetane-fused sugars are indeed
locked in the North-type constrained conformation. Frames (A1, A2) show
superimposed ab initio optimized geometries of the oxetane-C (1a),
oxetane-T (1b), oxetane-A (1c), and oxetane-G (1d). Frames (B1, B2)-
(E1, E2) contain 10 superimposed molecular structures collected every 10
ps of the last 0.4-0.5 ns interval of the MD simulations of the oxetane-C
(B1, B2), oxetane-T (C1, C2), oxetane-A (D1, D2), and oxetane-G (E1,
E2), respectively.
Supporting Information). Figure 11 shows that a combination
3
3
of the experimentally determined JH-3′,H-4′ and JH-4′,H-5′
constants and ab initio obtained values of the respective H3′-
C3′-C4′-H4′ and H4′-C4′-C5′-H5′ torsions correlates well
with the empirical 3JH,H vs torsion dependence predicted by the
Karplus equation, thereby indicating the reliability of the
theoretically obtained geometry parameters. The conformation
analysis based on the values of the sugar torsion from the ab
initio geometries collected in Table 4 for all the oxetane-
modified nucleosides reveals that the furanose ring is locked in
the North-East-type conformation with the phase angle of 40-
43° and the puckering amplitude 35-37°.
Molecular dynamics simulations (details are in the Experi-
mental section below) clearly demonstrated that the oxetane ring
is firmly restricting the dynamics of the oxetane-fused furanose
ring within the typical North-East-type (16° < P < 56°, 23° <
φm < 41°) conformation range (Table 4) with the average values
30.6° < P < 36.6° and 31.7° < φm < 32.7° somewhat lower
than the corresponding values from the ab initio geometries
(39.8° < P < 42.8°, 35.1° < φm < 36.6°) (Table 4). Ten
snapshots of the last 100 ps of the MD trajectory (harvested at
every 10 ps) are superimposed from the respective MD
simulations of the oxetane-modified nucleosides (Figure 10,
frames from B1, B2 to E1, E2). This shows that that the
oxetane-fused furanose ring is indeed rigid and locked in to
North-East-type conformation (shown in Figure 10, projections
9
11494 J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
3
Figure 11. Empirical standard and generalized Karplus dependencies compared to the experimental JH,H constants (Hz) (see Table S1 in Supporting
Information) for the H-3′, H-4′ and H-4′, H-5′ proton-proton interactions in the oxetane-modified nucleosides plotted against the values of the H3′-C3′-
C4′-H4′ and H4′-C4′-C5′-H5′ torsions from the 6-31G* ab initio calculated molecular structures of the oxetane-C (1a), -T (1b), -A (1c), and -G (1d)
nucleosides.
Table 4. Conformational Parameters of the Sugar Moiety of the Oxetane-C (1A), -T (1b), -A (1c), and -G (1d) Nucleosides Obtained (a)
from the ab Initio 6-31G* Hartree-Fock Optimized by Gaussian9845 Molecular Geometries and (b) from Unrestrained 0.5 ns MD Simulation
by Amber 746 Averaged in 0.1-0.5 ns Time Interval of the Respective Simulations (shown in parentheses in italics)
sugar conformational parameters
oxetane-C (1a)
oxetane-T (1b)
oxetane-A (1c)
oxetane-G (1d)
ν0: C5′-O5′-C2′-C3′
ν1: O5′-C2′-C3′-C4′
ν2: C2′-C3′-C4′-C5′
ν3: C3′-C4′-C5′-O5′
ν4: C4′-C5′-O5′-C2′
phase angle P
-15.20° (-9.7 ( 8.5°)
-8.57° (-10.7 ( 5.9°)
26.34° (25.0 ( 6.7°)
-35.76° (-32.0 ( 9.5°)
32.66° (26.7 ( 10.5°)
42.79° (35.3 ( 18.2°)
36.55° (32.3 ( 8.7°)
-14.30° (-10.5 ( 9.1°)
-8.89° (-10.2 ( 6.2°)
26.03° (24.9 ( 7.5°)
-34.88° (-32.4 ( 10.7°)
31.52° (27.5 ( 11.6°)
41.87° (36.6 ( 20.7°)
35.60° (32.9 ( 9.5°)
-12.91° (-7.1 ( 7.8°)
-9.96° (-13.6 ( 5.9°)
26.45° (27.0 ( 5.9°)
-34.50° (-32.8 ( 7.9°)
30.37° (25.3 ( 8.9°)
39.80° (30.6 ( 15.3°)
35.07° (32.7 ( 6.8°)
-14.72° (-9.3 ( 9.2°)
-8.82° (-10.6 ( 6.5°)
26.49° (24.5 ( 7.8°)
-35.55° (-31.4 ( 10.9°)
32.11° (25.8 ( 11.7°)
42.06° (35.7 ( 20.9°)
36.27° (31.7 ( 9.7°)
puckering amplitude
to that of the native AON/RNA hybrids. The DPPZ group offers
excellent 3′-exonuclease protection and augmented stabilities
in the Human serum. However, the vulnerability of the
oxetane-A and G modified AONs toward endonucleases might
be a drawback. Since we have not checked their tolerance toward
cellular endonucleases other than DNase 1, it is not possible to
comment at this time on their potential nuclease resistance in
the cellular systems, although oxetane-C modified AONs were
found to be adequately stable in K562 human leukemia cells.26
It is also worth to mention that, there are reports existing in the
literature which show the importance of the modifications at
the pyrimidine sites in an AON rather than at purines to achieve
endonuclease resistance.21c,d Further biological tests are in
progress to examine the potential target down-regulation by the
oxetane-A and -G modified AONs in cellular systems.
1H NMR and theoretical ab initio and MD simulations. The
combined experimental and theoretical studies have demon-
strated that the oxetane-fused furanose ring is indeed locked in
the typical North-East-type conformation with the pseudo-
rotational phase angle (P) and puckering amplitude (φm) for the
ab initio optimized geometries (6-31G* HF) varying from 39.8°
< P < 42.8°, 35.1° < φm < 36.6° for all four oxetane-modified
nucleosides. The last 100 picoseconds (ps) of 0.5 nanosecond
(ns) MD simulation starting from the respective ab initio
geometries have shown for the oxetane-modified sugars acces-
sible conformational range of P and φm to be 16° < P < 56°,
23° < φm < 41°.
(3) Oxetane-A and -G modified AONs showed their target
affinity with complementary RNA, and the thermodynamic
parameters to be identical or very close to the native AON/
RNA duplex. However, the oxetane-C and oxetane-T modified
duplexes showed large enthalpic destabilization and slight
entropic stabilization compared to that of the native hybrid
duplex. The large drop in enthalpy amounts to 32 to 37 kJ/mol
drop in the free energy.
Conclusions
(1) Oxetane-A and -G phosphoramidites were synthesized and
incorporated into 20mer AONs and targeted to 20mer RNA,
and their antisense properties have been evaluated in vitro.
(2) The molecular structures of the oxetane-C, T, A, and G
monomer units have been studied by means of the high-field
(4) The global helical structure of all the oxetane modified
AON/RNA hybrids, as revealed by the CD spectra, was very
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11495

A R T I C L E S
Pradeepkumar et al.
with pyridine three times and dissolved in 105 mL of the same solvent.
The mixture was cooled in an ice bath and methanesulfonyl chloride
(3.2 mL, 41.3 mmol) was added dropwise to the mixture and the stirring
was continued for 15 min at the same temperature. The reaction was
kept at 4 °C for 12 h, poured into cold saturated NaHCO3 solution and
extracted with CH2Cl2. The organic phase was washed with brine, dried
over MgSO4, filtered, evaporated, and coevaporated with toluene and
treated with 60 mL of 90% CF3COOH in water for 25 min at r.t. The
reaction mixture was evaporated, coevaporated three times with toluene
and three times with dry pyridine. The residue was dissolved in 75
mL of dry pyridine, acetic anhydride (8.6 mL, 90 mmol) was added
and stirred at room temperature for 3 h. The reaction mixture was
poured into saturated NaHCO3 solution and extracted with CH2Cl2. The
organic phase was washed with brine, dried over MgSO4, filtered,
evaporated, and coevaporated with toluene. Column chromatography
afforded 16 (5.3 g, 11.3 mmol, 75% after three steps). Rf ) 0.6 (MeOH/
similar to the native AON/RNA duplex showing that the CD
failed to detect the local conformational perturbations brought
about by the North-East conformationally constrained oxetane
modifications.
(5) All of the oxetane-A and oxetane-G modified AON/RNA
hybrid duplexes were found to be good substrates for the E.
coli RNase H1. In the oxetane-A and-G modified AON/RNA
hybrids, except for one case, a region of 5 nucleotides in the
RNA strand in the 3′-end direction from the site opposite to
the oxetane modification, was found to be insensitive toward
RNase H cleavage presumably owing to the local structural
perturbations brought about by the conformationally constrained
modifications.
(6) Michaelis-Menten kinetic analysis showed that triple
oxetane-A modified AON/RNA duplex has Vmax slightly lower
than those of the native and the triple oxetane-C modified AON
/RNA hybrids. However, the Km is very close to the native
yielding a kcat/Km close to it. The triple G modified AON/RNA
duplex showed low Km and low Vmax compared to the native
counterpart resulting in a kcat/Km value being half of that of the
native AON/RNA hybrid.
(7) The conjugation of nontoxic DPPZ group at the 3′-end
of the oxetane-A and -G modified AONs increases the Tm of
the respective AON/RNA duplexes only by 1-2 °C in com-
parison with their unconjugated counterparts. However, the
DPPZ conjugation has helped the AONs to achieve substantial
stability against exonuclease (t1/2 more than 24 h) and nucleases
in human serum (t1/2 more than 9 h).
1
CH2Cl2 97:3 v/v). H NMR (270 MHz, CDCl3): 7.95 (d, d, J ) 8.2
Hz, 2H, 4-toluoyl), 7.25 (d, 2H, 4-toluoyl), 5.65 (dd, JH-3,
) 4.6
H-4
Hz, JH-4, H-5 ) 7.9 Hz, 1H, H-4), 5.44 (d, 1H, H-3), 4.58 (dd, Jgem
11.8 Hz, JH-5, H-6 ) 3.3 Hz, 1H, H-6), 4.51-4.45 (m, 1H, H-5), 4.38
(Abq, Jgem ) 11.5 Hz, 2H, H-1, H-1′), 4.35 (dd, JH-5, ) 3.3 Hz,
)
H-6
1H, H-6′), 3.29 (s, 3H, OCH3), 3.05 (s, 3H, mesyl), 2.42 (s, 3H,
4-toluoyl), 2.16 (s, 3H, O-acetyl), 2.01 (s, 3H, O-acetyl). 13C NMR
(67.9 MHz, CDCl3): 169.3, 169 (CdO, acetyl), 166 (CdO, 4-toluoyl),
144.0, 129.6, 129.1, 126.7 (4-toluoyl), 106.2 (C-2), 79.0 (C-5), 73.6
(C-3), 71.0 (C-4), 63.5 (C-6), 61.3 (C-1), 48.9 (OCH3), 38.0 (CH3,
mesyl), 21.6 (CH3, 4-toluoyl), 20.4, 20.3 (CH3, O-acetyl). FAB-
HRMS: [MH]+ 475.1263; calcd 475.1276.
9-[1′-O-Methanesulfonyl-3′,4′-O-acetyl-6′-O-(4-toluoyl)-â-D-psico-
furanosyl]-N6-benzoyladenine (18a). The sugar 16 (10 g, 21 mmol)
was dissolved in 70 mL of dichloromethane (DCM) and treated with
HBr/acetic acid solution (70 mL) for 12 h. The mixture was poured
into ice cold water and extracted three times with DCM. The combined
organic phase was washed once with saturated NaHCO3 solution
followed by saturated NaCl solution. Finally, the organic phase was
dried over MgSO4, filtered, evaporated, and coevaporated twice with
dry toluene. The crude bromosugar was dissolved in dry dichloroethane-
CH3CN (6:1; 140 mL) and added to silylated N6-benzoyladenine.
[Silylation of N6-benzoyladenine: Nucleobase (7.2 g, 30 mmol) was
suspended in hexamethyldisilazane (100 mL) and trimethylchlorosilane
(6 mL) was added. The reaction mixture was stirred at 120 °C under
argon atmosphere for 12 h. The volatile material was evaporated,
coevaporated with dry toluene and the residue was dried for 20 min
using an oil pump]. The mixture was cooled in an ice bath and SnCl4
(5.25 mL) was added. After stirring in an ice bath for 10 min, the
reaction mixture was heated at 70 °C for 3h and continued stirring
overnight at r.t. The reaction mixture was poured into saturated NaHCO3
solution, passed through diatomaceous earth and extracted with CH2-
Cl2. The combined organic phase was washed with brine, dried over
MgSO4, filtered and evaporated. The residue on chromatography
furnished 18a (6.1 g, 8.8 mmol, 42%). Rf ) 0.5 (MeOH/CH2Cl2 96:4
(8) Modification of AONs with the oxetane-A and -G units
failed to offer resistance toward DNase 1 endonuclease degrada-
tion in comparison with the oxetane-C modified AONs. Thus,
DNase 1 appeared to have varying tolerance toward the oxetane-
purine and oxetane-pyrimidine nucleotides.
(9) This study unravels the pros and cons of the AONs
modified with the oxetane-purine and oxetane-pyrimidine
nucleosides and provides valuable information regarding the
optimal design of AONs having completely natural phospho-
diester backbone for the therapeutic applications.
Experimental Section
Comparative 13C chemical shifts in ppm (δ scale) of the oxetane
nucleosides 1a (C), 1b (T), 1c (A), and 1d (G) are shown in Table S2
in Supporting Information.
2-O-Methyl-3,4-O-isopropylidine-6-O-(4-toluoyl)-D-psicofur-
anose (15). The sugar 14 (9.5 g, 25 mmol) was treated with 0.4 M
methanolic HCl (125 mL) overnight. The reaction mixture was cooled,
subsequently neutralized with triethylamine, and evaporated. The crude
mixture was partitioned between CH2Cl2 and water. The aqueous layer
was washed three times with CH2Cl2 and the combined organic layer
was dried over MgSO4, filtered, and evaporated. Column chromatog-
raphy of the residue afforded 15 (5.2 g, 15 mmol, 60%). Rf ) 0.5
(CH2Cl2/MeOH 97:3 v/v). 1H NMR (270 MHz, CDCl3): 7.96 (d, 2H,
1
v/v). H NMR (270 MHz, CDCl3): 8.85 (br s, 1H, NH), 8.69 (s, 1H,
H-2), 8.20 (s, 1H, H-8), 7.99 (d, J ) 7.92 Hz, 2H, 4-toluoyl), 7.67-
7.52 (m, 5H, benzoyl), 7.01(d, 2H, 4-toluoyl,) 6.53 (d, JH-3′,
)
H-4′
5.32 Hz, 1H, H-3′), 5.61 (dd, JH-4′, H-5′ ) 3.96 Hz, 1H, H-4′), 4.97-
4.79 (m, J H-1′, H-1′′ ) 11.5, 4H, H-5′, H-1′, H-1′′, H-6′), 4.37 (dd, Jgem
J ) 8.29, 4-toluoyl), 7.24 (d, 2H, 4-toluoyl), 4.82 (dd, 1H, JH-3, H-4
)
6.06 Hz, JH-4, ) 1.2 Hz, H-4), 4.66 (d, H-3), 4.55-4.50 (m, 1H,
H-5
) 12.6 Hz, JH-5′,
) 2.97 Hz, H-6′′), 2.84 (s, 3H, CH3, mesyl),
H-6′
Jgem ) 7.2 Hz, JH-5, H-6 ) 1 Hz, H-6), 4.42-4.30 (m, 2H, H-5, H-6′),
3.84 (Abq, Jgem ) 9.2 Hz, 2H, H-1, H-1′), 3.29 (s, 3H, OMe), 2.4 (3H,
4-toluoyl), 1.53 (s, 3H, O-acetyl), 1.35 (s, 3H, O-acetyl). 13C NMR
(67.9 MHz, CDCl3): 166 (CdO, 4-toluoyl), 143.7, 129.5, 128.9, 126.7
(4-toluoyl), 112.9 (C-Me2), 110.1 (C-2), 85.3 (C-3), 84.0 (C-6), 82.0
(C-4), 64.6 (C-5), 58.3 (C-1), 48.4(OCH3), 26.0, 24.7 (CH3, isopropyl),
21.4 (CH3, 4-toluoyl). FAB-HRMS: [MH]+ 353.1624; calcd 353.1602.
1-O-Methanesulfonyl-2-O-methyl-3,4-O-acetyl-6-O-(4-toluoyl)-D-
psicofuranose (16). Compound 15 (5.2 g, 15 mmol) was coevaporated
2.28 (s, 6H, CH3, 4-toluoyl, CH3, O-acetyl), 2.16 (s, 3H, CH3, O-acetyl).
13C NMR (67.9 MHz, CDCl3): 169.3, 168.5 (CdO, acetyl), 165.5
(CdO, 4-toluoyl), 164 (CdO, benzoyl), 152.6 (C-6), 150.1 (C-4), 149.4
(C-2), 144.5 (4-toluoyl), 140.7 (C-8), 133.5, 132.7 (benzoyl) 129.7,
129.2, 128.9, 127.6, 125.3 (4-toluoyl, benzoyl), 123.0 (C-5), 94.9 (C-
2′), 82.4 (C-5′), 74.8 (C-3′), 71.7 (C-4′), 68.0 (C-1′), 62.8 (C-6′), 37.8
(CH3, mesyl), 21.4 (CH3, 4-toluoyl), 20.3 (CH3, O-acetyl). FAB-
HRMS: [MH]+ 682.1767; calcd 682.1819.
9
11496 J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
9-[1′-O-Methanesulfonyl-4′,6′-O-(tetraisopropyldisiloxane-1,3-
diyl)-â-D-psicofuranosyl]adenine (19a). Compound 18a (6.1 g, 9
mmol) was treated with methanolic NH3 solution for 48 h at room
temperature. The mixture was evaporated, triturated with dichloro-
methane, coevaporated three times with pyridine and dissolved in 90
mL of the same solvent. The mixture was cooled in an ice bath and
1,3-dichloro-1,1′,3,3′-tetraisopropyldisiloxane (2.88 mL, 9 mmol) was
added dropwise. The mixture was stirred at 0 °C for 30 min and at r.t.
for 2 h. It was poured into saturated NaHCO3 solution and extracted
with CH2Cl2. The organic phase was dried, evaporated and coevaporated
with toluene. Column chromatography of the residue afforded com-
pound 19a (4.1 g, 6.47 mmol, 72%). Rf ) 0.6 (CH2Cl2/MeOH 94:6
v/v). 1H NMR (270 MHz, CDCl3): 8.34 (s, 1H, H-2), 8.21(s, 1H, H-8),
5.86 (br s, 2H, NH2), 5.28 (d, Jgem ) 11.8 Hz, 1H, H-1′), 4.94 (d, 1H,
9-[1′-3′-O-Anhydro-6′-O-(4,4′-dimethoxytrityl)-â-D-psicofur-
anosyl]-N6-(phenoxyacetyl)-adenine (22a). Compound 21a (1.7 g, 2.7
mmol) was dissolved in THF (27 mL) and treated with 1.0 M tetra-
n-butylammonium fluoride in THF (2.7 mL) for 4 min at room
temperature. The mixture was evaporated and coevaporated three times
with pyridine. It was then dissolved in pyridine (22 mL) and
4,4′-dimethoxytrityl chloride (2 g, 5.94 mmol) was added, and the
mixture was stirred at room-temperature overnight. Saturated NaHCO3
solution was added and extracted with CH2Cl2. The organic phase was
washed with brine, dried over MgSO4, filtered, evaporated, and
coevaporated with toluene. The residue on column chromatography
afforded 22a (1.54 g, 2.1 mmol, 80% in two steps). Rf ) 0.4 (ethyl
1
acetate/cyclohexane 93:7 v/v). H NMR (270 MHz, CDCl3): 8.80 (s,
1H, H-2), 8.08 (s, 1H, H-8), 7.40-7.01 (m, 14H, DMTr & PAC), 6.79-
6.75 (m, 4H, DMTr), 5.79 (d, JH-3′, H-4′ ) 4.33 Hz, 1H, H-3′), 5.65 (d,
Jgem ) 7.92 Hz, 1H, H-1′), 4.978 (d, 1H, H-1′′), 4.87 (s, 1H, CH2,
JH-3′,
) 4.4, H-3′), 4.9 (d, 1H, H-1′′), 4.41 (dd, JH-3′,
) 8.54 Hz, 1H, H-4′), 4.32 (dt, 1H, JH-5′,
) 4.3
H-4′
H-4′
Hz, JH-5′,
) 2.47,
H-6′
H-4′
H-5′), 4.25-4.03 (ddd, Jgem ) 13.2 Hz, JH-5′, H-6′ ) 2.72 Hz, 2H, H-
6′, H-6′′), 1.07-0.85 (m, 28 H, Si-CH(CH3)2 and CH3 from 4-iPr).
13C NMR (67.9 MHz, CDCl3): 155.5 (C-6), 152.7 (C-2), 148.2 (C-4),
139.4 (C-8), 120.6 (C-5), 94.9 (C-2′), 82.5 (C-5′),75.3 (C-3′), 69.6 (C-
4′), 68.8 (C-1′), 60.3 (C-6′), 37.5 (CH3, mesyl), 17.1, 17.0, 16.9, 16.8
(Si-CH(CH3)2), 13.3, 12.9, 12.5, 12.4 (Si-CH(CH3)2). FAB-HRMS:
[MH]+ 618.2427; calcd 618.2449.
PAC), 4.52 (dd, JH-4′,
H-5′), 3.76 (s, OCH3, DMTr); 3.58 (dd, Jgem ) 10.76 Hz, JH-5′, H-6′
2.6 Hz, H-6′), 2.98 (dd, JH-5′,
) 8.3 Hz, 1H, H-4′), 4.47-4.43 (m, 1H,
H-5′
)
) 2.7 Hz, 1H, H-6′′). 13C NMR
H-6′′
(67.9 MHz, CDCl3): 166.5 (CdO), 158.5, 157 (DMTr), 152.9 (C-2),
151.2 (C-4), 148.4 (C-6), 144.5 (DMTr), 140.3 (C-8), 135.5, 130, 129.9,
129.8, 128.9, 128.1, 127.9, 127.7, 126.8, 125.2, 122.8, 122.4 (DMTr
& PAC), 114.9 (C-5), 113.1 (DMTr), 88.6 (C-3′), 86.3 (C-2′), 84.0
(C-5′), 79.7 (C-1′), 71.5 (C-4′), 68.1 (CH2, PAC), 62.5 (C-6′), 55.1
(OCH3, DMTr). FAB-HRMS: [MH]+ 716.2784; calcd 716.2720.
9-{1′,3′-O-Anhydro-4′-O-[2-cyanoethoxy(diisopropylamino)-phos-
phino]-6′-O-(4,4′-di-methoxytrityl)-â-D-psicofuranosyl}-N6-(phe-
noxyacetyl)-adenine (23a). To a stirred solution of 22a (763 mg, 1.07
mmol) in 7 mL of CH2Cl2, 2-cyanoethoxy-bis(N,N-diisopropylamino)
phosphine (0.62 mL, 1.6 mmol) was added followed by N,N-
diisopropylammonium tetrazolide (92 mg, 0.54 mmol) and left for
stirring overnight. The reaction mixture was diluted with of ethyl
acetate, poured into saturated NaHCO3 solution and extracted. The
organic layer was washed with saturated brine solution, dried over
MgSO4, filtered, and evaporated. The residue on chromatography (40-
60% ethyl acetate, cyclohexane + 2% Et3N) furnished 23a (640 mg,
0.7 mmol, 65%). Rf ) 0.6 (ethyl acetate/cyclohexane 93:7 v/v). The
compound was dissolved in CH2Cl2 (2 mL) and precipitated from
hexane at -40 °C. 31P NMR (109.4 MHz, CDCl3): 151.39; 150.40.
FAB-HRMS: [MH]+ 916.3730; calcd 916.3799.
9-[1′-O-Methanesulfonyl-3′,4′-O-acetyl-6′-O-(4-toluoyl)-â-D-psico-
furanosyl]-N2-acetylguanine (18b). The sugar 16 (4.7 g, 10 mmol)
was brominated using the same procedure as used for 18a. After being
kept on an oil pump for 30 min, the crude bromosugar was dissolved
in dry (CH2)2Cl2 (60 mL) and added to silylated N2-acetyl-O6-
diphenylcarbamyguanine. [The silylation was carried out by heating
N2-acetyl-O6-diphenyl carbamyguanine (5.8 g, 15 mmol) with N,O-
bis(trimethylsilyl)acetamide (30 mL) in (CH2)2Cl2 (75 mL) at 80 °C
for 30 min. The mixture was evaporated, coevaporated with dry toluene
and dried on an oil pump for 20 min.] The mixture was cooled in an
ice bath and SnCl4 (2.5 mL) was added. After being stirred in an ice
bath for 20 min. the reaction mixture was heated at 70 °C for 3 h and
continued stirring at r.t. overnight. The reaction mixture was poured
into saturated NaHCO3 solution and passed through diatomaceous earth.
It was extracted with CH2Cl2 and the combined organic phase was
washed with brine, dried over MgSO4, filtered, and evaporated. The
residue on chromatography afforded DPC protected nucleoside, which
was then treated with 90% triflouroacetic acid in water for 30 min to
remove the DPC group. The reaction mixture was evaporated, co-
evaporated with toluene and a quick column furnished 18b (1.6 g, 2.5
mmol, 25%). Rf ) 0.5 (CH2Cl2/MeOH 90:10 v/v). 1H NMR (270 MHz,
CDCl3): 12.01 (s, 1H, NH), 9.59 (s, 1H, NH), 7.98 (s, 1H, H-8), 7.67
(d, J ) 8.16 Hz, 2H, 4-toluoyl), 7.20 (d, 4-toluoyl), 6.56 (d, J ) 4.82
Hz, 1H, H-3′), 5.52 (dd, J H-3′, H-4′ ) 7.42 Hz, J H-4′, H-5′ ) 2.72 Hz,
1H, H-4′), 4.86 (d, Jgem ) 11.38 Hz 1H, H-1′), 4.7-4.64 (m, 3H, H-5′,
H-1′, H-6′), 4.53 (dd, Jgem ) 12.8 Hz, JH-6′, H-5′ ) 4.33 Hz, 1H, H-6′),
9-[1′-3′-O-Anhydro-4′,6′-O-(tetraisopropyldisiloxane-1,3-diyl)-â-
D-psicofuranosyl]adenine (20a). Compound 19a (3.19 g, 5.2 mmol,
1M solution in THF) was dissolved in THF (100 mL) and cooled in
an ice bath. Sodium bis(trimethylsilyamide) (10 mL, 10 mmol, 1 M
solution in THF) was added in dropwise and the stirring was continued
for 2 h. The reaction was quenched by adding saturated NaHCO3 and
extracted with excess of ethyl acetate. The combined organic phase
was dried, filtered, and evaporated. The residue on column chroma-
tography yielded 20a (2.2 g, 4.2 mmol, 81%). Rf ) 0.5 (ethyl acetate/
cyclohexane 93:7 v/v). 1H NMR (270 MHz, CDCl3): 8.28 (s, 1H, H-2),
7.79 (s, 1H, H-8), 5.86 (br s, 2H, NH2), 5.74 (d, JH-3′,
) 3.9 Hz,
H-4′
1H, H-3′), 5.58 (d, Jgem ) 7.8 Hz, 1H, H-1′), 5.01 (d, 1H, H-1′′), 4.83
(dd, JH-4′, H-5′ ) 8.5 Hz, H-4′), 4.49 (dt, JH-5′,H-6′ ) 2.5 Hz, 1H,H-5′),
4.18 (ddd, JH-5′,
) 3 Hz, Jgem)13.5, 2H, H-6′,H-6′′), 1.12-1.03
H-6′′
(m, 28 H, Si-CH(CH3)2 and CH3 from 4-iPr).13C NMR (67.9 MHz,
CDCl3): 155.7 (C-6), 153.4 (C-2), 149.6 (C-4), 137.2 (C-8), 119.4 (C-
5), 88.9 (C-3′), 87.8 (C-2′); 81.4 (C-5′), 79.5 (C-1′), 70.6 (C-4′), 59.8
(C-6′), 17.21, 17.15, 17.03, 16.9 (Si-CH(CH3)2), 13.3, 12.9, 12.4 (Si-
CH(CH3)2). FAB-HRMS: [MH]+ 522.2540; calcd 522.2568. A small
amount of 20a was deprotected to get 9-(1′,3′-O-anhydro-â-D-psicofur-
anosyl)adenine (1c) and data is shown in Table 3 and Table S2 in the
Supporting Information. FAB-HRMS: [MH]+ 280.1040; calcd 280.1046.
9-[1′-3′-O-Anhydro-4′,6′-O-(tetraisopropyldisiloxane-1,3-diyl)-â-
D-psicofuranosyl]-N6-(phenoxy-acetyl)-adenine (21a). Compound 20a
(1.8 g, 3.5 mmol) was dissolved in pyridine (25 mL) and phenoxyacetyl
chloride (0.64 mL, 4.6 mmol) was added dropwise to the reaction
mixture. Stirring was continued for 3 h at r.t. The mixture was poured
into sat. NaHCO3 solution extracted with CH2Cl2. Combined organic
phase evaporated and coevaporated with toluene. Column chromatog-
raphy of the residue furnished compound 21a (1.4 g, 2.1 mmol, 60%).
Rf ) 0.6 (CH2Cl2/MeOH 94:6 v/v). 1H NMR (270 MHz, CDCl3): 9.5
(br s, 1H, NH), 8.74 (s, 1H, H-2), 8.02 (s, 1H, H-8), 7.37-7.01 (m,
5H, PAC), 5.76 (d, J ) 3.9 Hz, 1H, H-3′), 5.59 (d, 1H, Jgem ) 7.8 Hz,
H-1′), 5.03 (d, 2H, Jgem ) 8.0 Hz, H-1′′), 4.92-4.82 (m, JH-4′, H-5′
3.9 Hz, JH-4′, H-3′ ) 8.7 Hz, 3 H, H-4′, CH2 PAC), 4.50 (dt, JH-5′, H-6′
) 2.0 Hz, JH-4′, H-5′, ) 8.7 Hz, 1H, H-5′), 4.25-4.12 (ddd, Jgem
13.9 Hz, JH-5′, ) 2.4 Hz, JH-5′, ) 2.7 Hz, 2H, H-6′, H-6′′),
)
)
H-6′
H-6′′
1.13-1.05 (m, 28 H, Si-CH(CH3)2 and CH3 from 4-iPr). 13C NMR
(67.9 MHz, CDCl3): 166.5 (CdO), 156.9 (C-6), 152.9 (C-2), 151.3
(C-4), 148.4, 140.1 (C-8), 129.7, 122.6, 122.3, 114.8 (C-5), 88.7 (C-
3′), 87.9 (C-2′); 81.7 (C-5′), 79.3 (C-1′), 70.5 (C-4′), 68.0 (CH2, PAC),
59.7 (C-6′), 17.2, 17.0, 16.9, 16.8 (Si-CH(CH3)2), 13.3, 12.9, 12.4 (Si-
CH(CH3)2). FAB-HRMS: [MH]+ 656.2949; calcd 656.2936.
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11497

A R T I C L E S
Pradeepkumar et al.
2.87 (s, 3H, CH3, mesyl), 2.38 (s, 3H, CH3, 4-toluoyl), 2.27(s, 3H,
CH3, O-acetyl), 2.2 (s, 3H, CH3, O-acetyl), 2.05 (s, 3H, CH3, N-acetyl).
13C NMR(67.9 MHz, CDCl3): 172.0(CdO, C-6), 169.1 (CdO, N-2
acetyl), 168.8 (CdO, 4-toluoyl), 165.8, 155.2 (C-6), 147.0 (C-2), 146.8
(C-4), 144.4 (4-toluoyl), 136.8 (C-8), 129.2, 129.0, 125.6 (4-toluoyl),
122.0 (C-5), 94.4 (C-2′); 80.0 (C-5′), 73.3 (C-3′), 69.7 (C-4′), 67.0
(C-1′), 61.6 (C-6′), 37.7 (CH3, mesyl), 24.1 (CH3, O-acetyl), 21.4 (CH3,
4-toluoyl), 20.2 (CH3, O-acetyl) 20.1 (CH3, N-acetyl). FAB-HRMS:
[MH]+ 636.1667; calcd 636.1564.
(270 MHz, CDCl3): 8.42 (s, 1H, NdCH), 7.63 (s, 1H, H-8), 7.42-
7.17 (m, 9H, DMTr), 6.79-6.75 (m, 4H, DMTr), 5.71 (d, JH-3′, H-4′
)
3.6 Hz, 1H, H-3′), 5.54 (d, Jgem ) 7.7 Hz, 1H, H-1′), 4.94 (d, 1H,
H-1′′), 4.45-4.37 (m, 2H, H-4′, H-5′), 3.77 (s, OCH3, DMTr); 3.58-
3.45 (ddd, Jgem ) 11 Hz, JH-5′, H-6′ ) 2.6 Hz, JH-5′, H-6′′ ) 2.7 Hz, 2H,
H-6′, H-6′′), 3.05, 3.01 (2s, 6H, -N(CH3)2). 13C NMR (67.9 MHz,
CDCl3): 158.4 (DMTr), 157.9 (NdCH), 157.5, 156.7 (C-6), 149.8 (C-
4), 144.4, 135.59, 135.54 (DMTr), 135.1 (C-8), 129.9, 128.1, 127.7,
126.7 (DMTr), 120.7 (C-5), 113.0 (DMTr), 88.6 (C-3′), 86.2 (C-2′),
83.1 (C-5′), 79.9 (C-1′), 71.6 (C-4′), 62.4 (C-6′), 55.1 (OCH3, DMTr),
41.3, 35.1 (N(CH3)2). FAB-HRMS: [MH]+ 653.2749; calcd 653.2724.
9-{1′,3′-O-Anhydro-4′-O-[2-cyanoethoxy(diisopropylamino)-phos-
phino]-6′-O-(4,4′-di-methoxytrityl)-â-D-psicofuranosyl}-N2-(N,N-
dimethylaminomethylene)-guanine (23b). Compound 22b (392 mg,
0.6 mmol) was converted to the corresponding phosphoramidite using
the conditions described for 23a. Workup and chromatography (80-
100% CH2Cl2, cyclohexane + 2% Et3N) furnished 23b (341 mg, 0.44
mmol, 73%). Rf ) 0.6 (CH2Cl2/MeOH 94:6 v/v). The compound was
dissolved in CH2Cl2 (2 mL) and precipitated from hexane at -40 °C.
31P NMR (109.4 MHz, CDCl3): 152.12; 149.89. FAB-HRMS: [MH]+
858.3817; calcd 853.3802.
Synthesis, Deprotection and Purification of Oligonucleotides. All
oligonucleotides were synthesized using an automated DNA/RNA
synthesizer by Applied Biosystems, model 392. For modified AONs
containing oxetane-A, -G, -C, and -T units, fast deprotecting phosphor-
amidites (nucleobases were protected using the following groups: Ac
for C, iPr-PAC for G, and PAC for A) were used. The protected AONs
were deprotected at room temperature by the treatment of aqueous NH3
for 24 h. For the 3′-DPPZ conjugated AONs, a CPG support containing
DPPZ was used.25 All AONs and the target RNA were purified by
20% polyacrylamide/7M urea) PAGE, extracted with 0.3 M NaOAc,
desalted with C18-reverse phase catridges and their purity (greater than
95%) was confirmed by PAGE. MALDI-MS analysis: AON (2) [MH]-
) 6128.37; calcd ) 6127.01; AON (3) [MH]- ) 6127.18; calcd )
6127.01; AON (4) [MH]- ) 6127.86; calcd ) 6127.01; AON (5)
[MH]- ) 6129.72; calcd ) 6127.01; AON (6) [MH]- ) 6127.44; calcd
) 6127.01; AON (7) [MH]- ) 6211.51; calcd ) 6211.04; AON (8)
[MH]- ) 6735.68; calcd ) 6736.48; AON (9) [MH]- ) 6737.88; calcd
) 6736.48; AON (12) [MH]- ) 6178.81; calcd ) 6178.02.
UV Melting Experiments. Determination of the Tm of the AON/
RNA hybrids was carried out in the following buffer: 57 mM Tris-
HCl (pH 7.5), 57 mM KCl, 1 mM MgCl2. Absorbance was monitored
at 260 nm in the temperature range from 20 °C to 90 °C using UV
spectrophotometer equipped with Peltier temperature programmer with
the heating rate of 1 °C per minute. Prior to measurements, the samples
(1 µΜ of AON and 1 µM RNA mixture) were preannealed by heating
to 90 °C for 5 min followed by slow cooling to 4 °C and 30 min
equilibration at this temperature.
Thermodynamic Calculations from the UV Experiments. The
thermodynamic parameters characterizing the helix-to-coil transition
for the DNA/RNA hybrids were obtained from Tm measurements over
the concentration range from 0.6 to 16 µM (total strand concentration).
Values of 1/Tm were plotted versus ln(CT/4) and ∆H° and ∆S°
parameters were calculated from slope and intercept of fitted line: 1/Tm
) (R/∆H°)ln(CT/S) + ∆S°/∆H°, where S reflects the sequence
symmetry of the self (S ) 1) or nonself-complementary strands (S ) 4).
CD Experiments. CD spectra were recorded from 300 to 200 nm
in 0.2 cm path length cuvettes. Spectra were obtained with a AON/
RNA duplex concentration of 5 µM in 57 mM Tris-HCl (pH 7.5), 57
mM KCl, 1 mM MgCl2. All the spectra were measured at 5 °C and
each spectrum is an average of 5 experiments from which CD value of
the buffer was subtracted.
9-[1′-O-Methanesulfonyl-4′,6′-O-(tetraisopropyldisiloxane-1,3-
diyl)-â-D-psicofuranosyl]guanine (19b). Compound 18b (2.2 g, 3.5
mmol) was deprotected and reated with 1,3-dichloro-1,1′,3,3′-tetraiso-
propyldisiloxane (1.1 mL, 3.5 mmol) using the same procedure
described for compound 19a. After workup and chromatography
afforded 19b (1.4 g, 2.2 mmol, 63% yield). Rf ) 0.6 (CH2Cl2/MeOH
90:10 v/v). 1H NMR (270 MHz, CD3OD): 8.03 (s, 1H, H-8), 5.29 (d,
Jgem ) 11.75 Hz, 1H, H-1′), 5.04 (d, JH-3′, H-4′ ) 3.34 Hz, 1H, H-3′),
4.79 (d, 1H, Jgem ) 11.75 Hz, H-1′′), 4.45-4.34 (m, 3H, H-4′, H-5′,
H-6′), 4.13 (dd, Jgem ) 15.96, JH-5′,
) 2.60 Hz, 1H, H-6′′), 3.03
H-6′
(s, 1H, mesyl), 1.2-1.02 (m, 28 H, Si-CH(CH3)2 and CH3 from 4-iPr).
13C NMR (67.9 MHz, CD3OD): 159.7 (C-6), 155.2 (C-2), 151.5 (C-
4), 138.3 (C-8), 119.2 (C-5), 96.6 (C-2′), 83.6 (C-5′), 76.3 (C-3′), 70.9
(C-1′), 70.7 (C-4′), 61.1 (C-6′), 37.5 (OMs), 18.2, 18.1, 17.7, 17.7 (Si-
CH(CH3)2), 14.9, 14.5, 14.2, 14.1 (Si-CH(CH3)2). FAB-HRMS:
[MH]+ 634.2432; calcd 634.2401.
9-[1′-3′-O-Anhydro-4′,6′-O-(tetraisopropyldisiloxane-1,3-diyl)-â-
D-psicofuranosyl]guanine (20b). Compound 19b (1.4 g, 2.2 mmol)
was cyclized using the procedure as described for 20a. After work up
and chromatography of the residue furnished 20b (990 mg, 1.8 mmol,
83%) Rf ) 0.6 (CH2Cl2/MeOH 88:12 v/v). 1H NMR (270 MHz, DMSO-
d6): 7.85 (s, 1H, H-8), 6.32 (s, 2H, NH2), 5.70 (d, JH-3′, H-4′ ) 3.9 Hz,
1H, H-3′), 5.39 (d, Jgem ) 8.3 Hz, 1H, H-1′), 4.95 (d, 1H, H-1′′), 4.58
(dd, JH-4′, H-5′ ) 8.3 Hz 1H, H-4′), 4.44 (dt, JH-5′, H-6′ ) 3.0 Hz, 1H,
H-5′), 4.13 (d, 2H, H-6′, H-6′′), 1.18-1.09 (m, 28 H, Si-CH(CH3)2
and CH3 from 4-iPr). 13C NMR (67.9 MHz, DMSO-D6): 156.8 (C-6),
153.8 (C-2), 151 (C-4), 135 (C-8), 116.9 (C-5), 87.6 (C-3′), 87.4 (C-
2′), 80.7 (C-5′), 79.0 (C-1′), 71.8 (C-4′), 60.7 (C-6′), 17.3, 17.2, 17.1,
16.9 (Si-CH(CH3)2), 12.8, 12.6, 12.3, 12.2 (Si-CH(CH3)2). FAB-
HRMS: [MH]+ 538.2525; calcd 538.2517. A small amount of 20b
was deprotected to get 9-(1′,3′-O-anhydro-â-D-psicofuranosyl)guanine
(1d) and the data are shown in Table 3 and in Table S2 in Supporting
Information. FAB-HRMS: [MH]+ 296.0990; calcd 296.0997.
9-[1′-3′-O-Anhydro-4′,6′-O-bis (tetraisopropyldisiloxane-1,3-diyl)-
â-D-psicofuranosyl]-N2-(N,N-dimethylamino methylene)guanine (21b).
To a stirred solution of compound 20b (733 mg, 1.4 mmol) in dry
MeOH (13 mL), N,N-dimethylformamide dimethylacetal (0.9 mL, 6.8
mmol) was added and allowed to stir at r.t. overnight. The mixture
was evaporated and coevaporated with toluene. Column chromatography
of the residue furnished compound 21b (757 mg, 1.27 mmol, 91%). Rf
1
) 0.5 (CH2Cl2/MeOH 90:10 v/v). H NMR (270 MHz, CDCl3): 8.5
(s, 1H, NdCH), 7.59 (s, 1H, H-8), 5.67 (d, Jgem ) 9 Hz, 1H, H-1′),
5.64 (d, JH-3′,
) 4.3 Hz, 1H, H-3′), 4.93 (d, 1H, H-1′′), 4.44 (dt,
H-4′
JH-4′,
) 8.5 Hz, 1H, H-5′), 4.38 (dd, 1H, H-4′), 4.23-4.10 (ddd,
H-5′
Jgem ) 13.0 Hz, JH-5′
,
) 2.5 Hz, JH-5′,
) 2.4 Hz, 2H, H-6′,
H-6′
H-6′′
H-6′′), 3.19, 3.13 (2s, 6H, N(CH3)2), 1.10-1.04 (m, 28 H, Si-CH(CH3)2
and CH3 from 4-iPr). 13C NMR (67.9 MHz, CDCl3): 157.7 (NdCH),
157.5 (C-2), 156.7 (C-6), 149.8 (C-4), 135.1 (C-8), 120.9 (C-5), 88.9
(C-3′), 88.6 (C-2′); 81.0 (C-5′), 80.2 (C-1′), 71.5 (C-4′), 60.0 (C-6′),
41.3, 35.2 (N(CH3)2), 17.22, 17.13, 17.0, 16.9, (Si-CH(CH3)2), 13.5, 12.9,
12.6 (Si-CH(CH3)2). FAB-HRMS: [MH]+ 593.2951; calcd 593.2939.
9-[1′-3′-O-Anhydro-6′-O-(4,4′-dimethoxytrityl)-â-D-psicofuran-
osyl]-N2-(N,N-(dimethylamino)-methylene)-guanine (22b). From com-
pound 21b (680 mg, 1.2 mmol), the TIPS group was removed and
treated with 4,4′-dimethoxytrityl chloride (780 mg, 2.3 mmol) using
the method described for compound 22a to afford 22b (614 mg, 1
mmol, 83% in two steps). Rf ) 0.5 (CH2Cl2/MeOH 90:10 v/v). 1H NMR
Theoretical Calculations. The structural parameters of the oxetane-
modified nucleosides and conformational hyperspace available for the
compounds have been determined by the ab initio geometry optimiza-
tions followed by the 0.5 ns molecular dynamics simulations. The
9
11498 J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004

1′,2′-Oxetane Constrained Modified Oligonucleotides
A R T I C L E S
geometry optimizations of the oxetane-modified nucleosides have been
carried out by the Gaussian 98 program package45 at the Hartree-
Fock level using 6-31G* basis set. The atomic charges and optimized
geometries of the oxetane-C (1a), -T (1b), -A (1c), and -G (1d) were
then used for the Amber force field parameters employed in the MD
simulations. The protocol of this MD simulations is based on
Cheathan-Kollman’s47 procedure employing modified version of
Amber 1994 force field as it is implemented in Amber 7 program
package.46 The TIP3P water model was used to introduce explicit
solvent molecules in the MD calculations. Periodic boxes containing
1048, 1125, 1084, and 1122 water molecules were created around the
oxetane-C (1a), -T (1b), -A (1c), and -G (1d), respectively, extending
12.0 Å from these molecules in all three dimensions.
tetraacetic acid (EDTA), 0.05% (w/v) bromophenol blue and 0.05%
(w/v) xylene cyanol in 95% formamide, and subjected to 20% 7 M
urea denaturing gel electrophoresis. The kinetic parameters Km and Vmax
were obtained from V0 versus [S0] plots obeying Michaelis-Menton
equation: V0 ) Vmax‚S0/Km + S0. Values of Km and Vmax at this method
were determined directly from V0 versus [S0] plots by using of
correlation Program, where correlation equation was: y ) ax/(b + x)
where a ) Vmax and b ) Km. Since the Vmax ) E0*kcat (E0 ) initial
enzyme concentration), and for all of our RNA concentration dependent
kinetics, the E0 was identical, hence Vmax is proportional to kcat. In other
words, the kcat can be understood by comparing simply the Vmax
.
Exonuclease Degradation Studies. Stability of the AONs toward
3′-exonucleases was tested using snake venom phosphodiesterase from
Crotalus adamanteus. All reactions were performed at 3 µM DNA
concentration (5′-end 32P labeled with specific activity 50 000 cpm) in
56 mM Tris-HCl (pH 7.9) and 4.4 mM MgCl2 at 22 °C. Exonuclease
concentration of 35 ng/µL was used for digestion of oligonucleotides.
Total reaction volume was 14 µL. Aliquots were taken at 1, 2, 4, and
24 h and quenched by addition of the same volume of 50 mM EDTA
in 95% formamide. Reaction progress was monitored by 20% denatur-
ing (7 M urea) PAGE and autoradiography.
Endonuclease Degradation Studies. Stability of AONs toward
endonuclease was tested using DNase 1 from bovine pancreas.
Reactions were carried out at 0.9 µM DNA concentration (5′-end 32P
labeled with specific activity 50 000 cpm) in 100 mM Tris-HCl (pH
7.5) and 10 mM MgCl2 at 37 °C using 20 unit of DNase 1. Total
reaction volume was 22 µL. Aliquots were taken at 1, 2, 4, 8 and 12
h and quenched with the same volume of 50 mM EDTA in 95%
formamide. They were resolved in 20% polyacrylamide denaturing (7
M urea) gel electrophoresis and visualized by autoradiography.
Stability Studies in Human Serum. AONs (6 µL) at 2 µM
concentration (5′-end 32P labeled with specific activity 90 000 cpm)
were incubated in 30 µL of human serum (male AB) at 22 °C (total
reaction volume was 36 µL). Aliquots (3 µL) were taken at 0, 15 and
30 min, 1, 2 and 9 h, and quenched with 6 µL of solution containing
8 M urea and 50 mM EDTA, resolved in 20% polyacrylamide
denaturing (7 M urea) gel electrophoresis and visualized by auto-
radiography.
32P Labeling of Oligonucleotides. The oligoribonucleotide, oligo-
deoxyribonucleotides were 5′-end labeled with 32P using T4 polynucleo-
tide kinase, [γ-32P]ATP and standard procedure.48 Labeled AONs and
RNA were purified by 20% denaturing PAGE and specific activities
were measured using Beckman LS 3801 counter.
Kinetics of RNase H Hydrolysis. (A) Calibration of RNase H
Concentration Based on its Cleavage Activity. The source of RNase
H1 (obtained from Amersham Bioscience) was Escherichia coli
containing clone of RNase H gene. The solutions of 20mer AON (1)/
RNA (13) duplex: [AON] ) 10-6 M, [RNA] ) 10-7 M (the first control
substrate) and 20mer AON (3)/RNA (13) duplex: [AON] ) 10-6 M,
[RNA] ) 10-7 M (the second control substrate) in a buffer, containing
20 mM Tris-HCl (pH 8.0), 20 mM KCl, 10 mM MgCl2, 0.1 mM EDTA
and 0.1 mM dithiothreitol (DTT) at 21 °C in 30 µL of the total reaction
volume have been used as standard substrates to calibrate the amount
of RNase H actually used. The percentage of RNA cleavage was
monitored by gel electrophoreses as a function of time (1-5 min), with
using of 0.08 U of RNase H, to give the initial velocity. Thus, the
initial Velocity of the RNase H cleaVage reaction, 5.4171 × 10-3 µM/
min (for the first standard substrate control) or 2.0869 × 10-3 µM/
min (for the second standard substrate control) under the aboVe
condition corresponds to 0.08 units actiVity of enzyme in 30 µL of the
total reaction mixture. These are based on 9 independent experiments.
Since Michaelis equation suggests that the initial velocity of the reaction
linearly depends on the enzyme concentration, therefore, using the initial
velocity for the first standard substrate control or the second standard
substrate control corresponding to 0.08 units/30 µL concentration of
the RNase H, two correlation coefficients were found by dividing the
observed experimental initial velocity by the standard initial velocity
of 5.4171 × 10-3 or 2.0869 × 10-3 µM/min. Then, the real enzyme
concentration as well as the initial velocity in each experiment was
corrected using average value of the correlation coefficients of both
the standard substrate controls, which were used to calibrate the initial
velocity of the RNase H promoted cleavage reaction for each substrate
presented in this work.
Acknowledgment. We thank Swedish Natural Science Re-
search Council (Vetenskapsrådet), the Swedish Foundation for
Strategic Research (Stiftelsen fo¨r Strategisk Forskning) and
Philip Morris USA Inc. for generous financial support. We also
thank the Swedish National Allocation Committee for the High
Performance Computing at the National Supercomputer Center
(Linko¨ping, Sweden).
Supporting Information Available: (1) General experimental
procedures; (2) Figure S1: CD spectra of AON (1-4, 7, and
10)/RNA (13) duplexes; (3) Figure S2: PAGE analysis of the
DNase 1 degradation of AON (1)-(4), (7), and (11); (4) Figure
S3: PAGE analysis of the snake venom phosphodiesterase
(SVPDE) degradation of AONs (1), (7), (12), (8), (9) and (11);
(5) Figure S4: PAGE analysis of the degradation of AONs (1),
(7), (12), (8), (9) and (11) in human serum; (6) Table S1:
(B) RNA Concentration Dependent Experiments. 32P-Labeled
RNA (0.01 to 0.8 µM, specific activity 50 000 cpm) with AONs (4
µM) were incubated with 0.08 units of RNase H in buffer, containing
20 mM Tris-HCl (pH 8.0), 20 mM KCl, 10 mM MgCl2 and 0.1 mM
DTT at 21 °C. Total reaction volume was 30 µL. Prior to the addition
of the enzyme reaction components were preannealed in the reaction
buffer by heating at 80 °C for 5 min followed by 1.5 h equilibration at
21 °C. After 1-7 min, aliquots (3 µL) were mixed with stop solution
(6 µL), containing 0.05 M disodium salt of the ethylenediamine-
3
Experimental and calculated vicinal JH,H coupling constants;
(7) Table S2: Carbon chemical shifts in ppm (δ scale) of the
oxetane nucleosides 1a (C), 1b (T), 1c (A), and 1d (G); (8)
Figures S4-S16: 13C NMR spectra of compounds 15-22; (9)
Figures S17-S18: 31P NMR spectra of compounds 23a-b; (10)
Figure S19: 13C NMR spectrum of oxetane-A (1c); (11) Figure
S20: 13C NMR spectrum of oxetane-G (1d). This material is
available free of charge via the Internet at http://pubs.acs.org.
(46) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham, T. E., III.; Wang,
J., Ross, W. S.; Simmerling, C. L.; Darden, T. A.; Merz, K. M.; Stanton,
R. V.; Cheng, A. L.; Vincent, J. J.; Crow-ley, M.; Tsui, V.; Gohlke, H.;
Radmer, R. J.; Duan, Y.; Pitera, J.; Massova, I.; Seibel, G. L.; Singh, U.
C.; Weiner, P. K.; Kollman, P. A. 2002, AMBER 7, University of
California, San Francisco.
(47) Cheatham, T. E., III.; Kollman, P. A. J. Am. Chem. Soc. 1997, 119, 4805.
(48) Ausubel, F. M.; Brent, R.; Kingston, R. E.; Moore, D. D.; Seidman, J. G.;
Smith, J. A.; Struhl, K. Current Protocols in Molecular Biology; John Wiley
& Sons: New York, 1995; Vol. 3, 10.3.
JA048417I
9
J. AM. CHEM. SOC. VOL. 126, NO. 37, 2004 11499