Carba-LNA-5MeC/A/G/T Modified oligos show nucleobase-specific modulation of 3′-exonuclease activity, thermodynamic stability, RNA selectivity, and RNase H elicitation: Synthesis and biochemistry

Article abstract of DOI:10.1021/jo200073q

Using the intramolecular 5-exo-5-hexenyl radical as a key cyclization step, we previously reported an unambiguous synthesis of carba-LNA thymine (cLNA-T), which we subsequently incorporated in antisense oligonucleotides (AON) and investigated their biochemical properties [J. Am. Chem. Soc.2007, 129 (26), 8362-8379]. These cLNA-T incorporated oligos showed specific RNA affinity of +3.5-5 °C/modification for AON:RNA heteroduplexes, which is comparable to what is found for those of LNAs (Locked Nucleic Acids). These modified oligos however showed significantly enhanced nuclease stability (ca. 100 times more) in the blood serum compared to those of the LNA modified counterparts without compromising any RNase H recruitment capability. We herein report the synthesis of 5-methylcytosine-1-yl (^MeC), 9-adeninyl (A), and 9-guaninyl (G) derivatives of cLNA and their oligonucleotides and report their biochemical properties as potential RNA-directed inhibitors. In a series of isosequential carba-LNA modified AONs, we herein show that all the cLNA modified AONs are found to be RNA-selective, but the magnitude of RNA-selectivity of 7′-R-Me-cLNA-G (cLNA-G) (ΔT_m = 2.9 °C/modification) and intractable isomeric mixtures of 7′-(S/R)-Me-cLNA-T (cLNA-T, ΔT_m = 2.2 °C/modification) was found to be better than diastereomeric mixtures of 7′-(S/R)-Me-cLNA-^MeC with trace of cENA-^MeC (cLNA-^MeC, ΔT_m = 1.8 °C/modification) and 7′-R-Me-cLNA-A (cLNA-A, ΔT_m = 0.9 °C/modification). cLNA-^MeC modified AONs however exhibited the best nuclease stability, which is 4-, 7-, and 20-fold better, respectively, than cLNA-T, cLNA-A, and cLNA-G modified counterparts, which in turn was more than 100 times stable than that of the native. When the modification sites are appropriately chosen in the AONs, the cLNA-A, -G, and -^MeC modified sites in the AON:RNA hybrids can be easily recognized by RNase H, and the RNA strand of the hybrid is degraded in a specific manner, which is important for the design of oligos for therapeutic purposes. The cLNA-^MeC modified AON/RNA, however, has been found to be degraded 4 times faster than cLNA-A and G modified counterparts. By appropriately choosing the carba-LNA modification sites in AON strands, the digestion of AON:RNA can be either totally repressed or be limited to cleavage at specific sites or at a single site only (similar to that of catalytic RNAzyme or DNAzyme). Considering all physico- and biochemical aspects of cLNA modified oligos, the work suggests that the cLNA modified antisense oligos have the potential of being a promising therapeutic candidate due to their (i) higher nucleobase-specific RNA affinity and RNA selectivity, (ii) greatly improved nuclease stability, and (iii) efficient RNase H recruitment capability, which can induce target RNA cleavage in a very specific manner at multiple or at a single site, in a designed manner.

Full text of DOI:10.1021/jo200073q

ARTICLE
pubs.acs.org/joc
Carba-LNA-5MeC/A/G/T Modified Oligos Show Nucleobase-Specific
Modulation of 30-Exonuclease Activity, Thermodynamic Stability, RNA
Selectivity, and RNase H Elicitation: Synthesis and Biochemistry
RamShankar Upadhayaya,†,§ Sachin Gangadhar Deshpande,†,§ Qing Li,‡,§ Ramakant Asaram Kardile,†
Aftab Yusuf Sayyed,† Eknath Kamalakar Kshirsagar,† Rahul Vilas Salunke,† Shailesh Satish Dixit,†
Chuanzheng Zhou,‡ Andrꢀas F€oldesi,‡ and Jyoti Chattopadhyaya*,‡
†Institute of Molecular Medicine, International Biotech Park, Genesis Campus, Phase II Opp Infosys, Tal Mulshi, Hinjewadi, Pune 411 057, India
‡Program of Bioorganic Chemistry, Institute of Cell and Molecular Biology, Biomedical Centre, Uppsala University, SE-75123 Uppsala, Sweden
S Supporting Information
b
ABSTRACT:
Using the intramolecular 5-exo-5-hexenyl radical as a key cyclization step, we previously reported an unambiguous synthesis of carba-
LNA thymine (cLNA-T), which we subsequently incorporated in antisense oligonucleotides (AON) and investigated their
biochemical properties [J. Am. Chem. Soc. 2007, 129 (26), 8362ꢀ8379]. These cLNA-T incorporated oligos showed specific RNA
affinity of þ3.5ꢀ5 °C/modification for AON:RNA heteroduplexes, which is comparable to what is found for those of LNAs
(Locked Nucleic Acids). These modified oligos however showed significantly enhanced nuclease stability (ca. 100 times more) in
the blood serum compared to those of the LNA modified counterparts without compromising any RNase H recruitment capability.
We herein report the synthesis of 5-methylcytosine-1-yl (MeC), 9-adeninyl (A), and 9-guaninyl (G) derivatives of cLNA and their
oligonucleotides and report their biochemical properties as potential RNA-directed inhibitors. In a series of isosequential carba-LNA
modified AONs, we herein show that all the cLNA modified AONs are found to be RNA-selective, but the magnitude of RNA-
selectivity of 70-R-Me-cLNA-G (cLNA-G) (ΔTm = 2.9 °C/modification) and intractable isomeric mixtures of 70-(S/R)-Me-cLNA-
T (cLNA-T, ΔTm = 2.2 °C/modification) was found to be better than diastereomeric mixtures of 70-(S/R)-Me-cLNA-MeC with
trace of cENA-MeC (cLNA-MeC, ΔTm = 1.8 °C/modification) and 70-R-Me-cLNA-A (cLNA-A, ΔTm = 0.9 °C/modification).
cLNA-MeC modified AONs however exhibited the best nuclease stability, which is 4-, 7-, and 20-fold better, respectively, than cLNA-
T, cLNA-A, and cLNA-G modified counterparts, which in turn was more than 100 times stable than that of the native. When the
modification sites are appropriately chosen in the AONs, the cLNA-A, -G, and -MeC modified sites in the AON:RNA hybrids can be
easily recognized by RNase H, and the RNA strand of the hybrid is degraded in a specific manner, which is important for the design
of oligos for therapeutic purposes. The cLNA-MeC modified AON/RNA, however, has been found to be degraded 4 times faster
than cLNA-A and G modified counterparts. By appropriately choosing the carba-LNA modification sites in AON strands, the
digestion of AON:RNA can be either totally repressed or be limited to cleavage at specific sites or at a single site only (similar to that of
catalytic RNAzyme or DNAzyme). Considering all physico- and biochemical aspects of cLNA modified oligos, the work suggests that
the cLNA modified antisense oligos have the potential of being a promising therapeutic candidate due to their (i) higher nucleobase-
specific RNA affinity and RNA selectivity, (ii) greatly improved nuclease stability, and (iii) efficient RNase H recruitment capability,
which can induce target RNA cleavage in a very specific manner at multiple or at a single site, in a designed manner.
’ INTRODUCTION
therapeutics are being addressed by different groups including
the problem of target affinity, nuclease stability, delivery, and
ADME-Tox (Absorption, Distribution, Metabolism, Excretion,
A short oligonucleotide sequence in a single-stranded anti-
sense form (AON) or in a double-stranded form (siRNA) can
modulate the gene expression by targeting against the cellular
RNA, which is being exploited as potential therapeutics in several
laboratories.1 Several problems in the design of RNA-directed
Received: January 12, 2011
Published: April 18, 2011
r
2011 American Chemical Society
4408
dx.doi.org/10.1021/jo200073q J. Org. Chem. 2011, 76, 4408–4431
|

The Journal of Organic Chemistry
ARTICLE
Figure 1. Structures of some representatives of 10,20- or 20,40-conformationally locked nucleosides.
and Toxicity). Chemical modifications of oligonucleotides have
been, however, proved to be an effective strategy to counter some
of these problems.2 Among these, the oligonucleotides modified
with LNA (Locked Nucleic Acid, also known as BNA)3 building
blocks have shown several desirable properties, especially their
unprecedented affinity toward complementary RNA. The LNA
monomer nucleoside blocks have a 20,40-bridge [40-CH2-O-20]
locking the sugar conformation in a perfect North-type (N-type)
conformation [pseudorotation angles, P ≈ 17°, and the pucker-
ing amplitude, vmax ≈ 56°].4 The enhanced target binding
property of the North-conformationally constrained bicyclic
sugar units in LNA has been attributed to the improved stacking
between the nearest neighbors by promoting preorganization of
ssLNA owing to restriction of the concerted local backbone
motions in the LNA-modified oligos, thereby reducing the
entropic penalty for formation of duplex with RNA.5 These
striking conformational features have paved the way for LNA-
modified oligos for an extensive application in RNA targeting
therapeutics and in biotechnologies.6
The intrinsic feature of the North-type sugar constraint in
LNA has led to the synthesis of a number of closely related
analogues, in which the 20,40-bridge has been altered7 to 20,
40-BNACOC 13,7d 20,40-BNANC 14,7e 60-substituted LNAs 2ꢀ5,7b
or a new type of 10,20-bridged constraint, such as in the 10,20-
oxetane 88 or 10,20-azetidine 99 analogue (Figure 1). Such
modifications show similar or moderately depressed Tm proper-
ties when compared to LNA, but the nuclease resistance or
RNase H recruitment properties (for example, ENA 10,10 PrNA
12,11 and aza-ENA 1112) (Figure 1) have turned out to be
relatively more favorable than those exhibited by the LNA-
containing AONs.
Replacement of the 20-oxygen of LNA with the 20-CHR- group
leads to carbocyclic locked nucleosides (cLNA and cENA,
Figure 2), and the first cENA and its analogues (18, 19, and
23ꢀ25, Figure 2) were synthesized by Nielsen0s group.13 These
cENA analogues had been incorporated into AONs with a
natural phosphodiester backbone, which led to increased thermal
stability (Tm) by 2.5ꢀ4.5 °C/modification with the complemen-
tary RNA.13 We have recently reported the synthesis14a of 70-Me-
carba-LNA-T (cLNA-T) 27 and 80-Me-carba ENA-T (cENA-T)
22 nucleosides, respectively, involving 5-hexenyl or the 6-hepte-
nyl radical cyclization to a distant double bond tethered at C40
and the radical center at C20 of the ribofuranose ring of
thymidine. We have demonstrated that cLNA-T enhances the
Tm of the modified AON:RNA heteroduplexes by 3.5ꢀ5 and
1.5 °C/modification for cLNA-T and cENA-T, respectively. It
was also noteworthy that they remarkably enhanced the lifetimes
of hydrolysis of these carbocyclic nucleoside-modified AONs in
the human blood serum, which may potentially produce the
highly desired pharmacokinetic properties because of their
unique stability and consequently a net reduction of the required
dosage.
This unique quality as well as their efficient use as the AON in
the RNase H-promoted cleavage of the target RNA make our
cLNA and cENA modifications excellent candidates as potential
antisense therapeutic agents.
Later, we continued to fine-tune the electrostatic effects by
modulating the chemical character of the 20,40-C-ethylene bridge
(B = 1-thyminyl) by the introduction of a variety of substitutions
(H, Me, OH, NH2) at the C60-positions and the C70 (or C80)
positions (28ꢀ31) in cLNA and (20ꢀ22) cENA.14aꢀg However,
this fine-tuning does not result in major structure variations in
4409
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 2. Structures of some known carbocyclic LNA and carbocyclic ENA nucleosides.
Figure 3. Structures of 20,40-constrained cLNA-T/MeC/A/G nucleotides.
terms of recognition of the cDNA and RNA strand, although
some of them lead to quite improved nuclease resistance and
RNase H recruitment properties relative to cLNA-T 27 and
cENA-T 22.15
capability would give us complete insight into how cLNAs in
general could be used as potential antisense or siRNA therapeutic
agents. In this paper, we report our results toward this goal by
highlighting the unique nucleobase-specific properties among
the isosequential cLNA-MeC/T/A/G modified counterparts.
Recently, Seth and co-workers14h have synthesized a few
cLNA analogues (methylene-cLNA 32, see Figure 2), which
have 20,40-bridged carba skeletons on the pentofuranose ring
identical to ours, as in 26ꢀ31.14aꢀf This was achieved by the
intramolecular free-radical cyclization of the C20 radical to the
C40-tethered ꢀCꢁCH as an extension of our original free-radical
cyclization work in which we showed intramolecular addition of
’ RESULT AND DISCUSSION
1.0. Synthesis of cLNA-A, -G, and -MeC and Their Phos-
phoramidite Derivatives. 1.1. Synthesis of cLNA-A, -G, and -MeC
Nucleosides by the Radical Cyclization Method. The independent
syntheses of cLNA-A (41a/b), -G (42a/b), and -MeC (43a/b/c)
nucleosides starts from a common starting intermediate 3614a
(Scheme 1) which has a C4 tethered propenyl side chain for
5-hexenyl-type free radical cyclization. The compound 36 was
subjected to acetolysis using a mixture of acetic anhydride, acetic
acid, and triflic acid to give the corresponding diacetate 37
quantitatively as an R/β anomeric mixture. After workup, the
crude diacetate 37 was coupled with persilylated protected bases,
N6-benzoyladenine,16 N2-acetyl-O6-diphenylcarbomoylguanine,17
and N4-benzoyl-5-methylcytosine,18 in the presence of Lewis acid
TMSOTf to give 38a (86%), 38b (60%), and 38c (80%) in good
14aꢀc,e,g
the C20 radical to the C40-tethered ꢀCHdCH2
or ꢀCHdNOR.14d,f It was also found14h that oligos with carba-
LNA with an exocyclic methylene group at C70 have Tm values
comparable to LNA when hybridized with complementary RNA.
Given the easy accessibility and unique antisense properties of
cLNA-T, it was clear that synthesis of cLNA-A/G/MeC deriva-
tives (33ꢀ35 in Figure 3) and their oligos should give us a nice
opportunity to compare their biophysical and biochemical
properties with those of the cLNA-T modified counterparts.
We argued that such studies as the relative RNA affinity/
selectivity, blood serum stability, and RNase H elicitation
4410
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Scheme 1a
a Reagents and Conditions: (i) Ac2O, AcOH, TfOH, 0 °C to rt, 30 min; (ii) persilylated N6-benzoyladenine, TMSOTf, CH3CN, 50 °C overnight for 38a
or (ii) persilylated protected guanine, TMSOTf, toluene, 80 °C, 1.5 h for 38b or (ii) persilylated N4-benzoyl-5-Me-cytosine, TMSOTf, CH3CN, rt,
overnight for 38c; (iii) 2 M NaOH, THF/H2O, rt, 2 h for 39a or (iii) 2 M NaOH, THF/H2O, 0 °C, 2 h for 39b or (iii) 2 M NaOH, THF/H2O, rt, 2 h for
39c; (iv) 1-methylimidazole, phenyl chlorothionoformate, DCM, rt, 5 h for 40a or (iv) 1-methylimidazole, phenyl chlorothionoformate, DCM, rt,
overnight for 40b or (iv) 1-methylimidazole, phenyl chlorothionoformate, DCM, 1.5 h for 40c; (v) Bu3SnH, toluene, AIBN, reflux, 1.5 h; 41a 38%, 41b
4%; 41c 4% or (v) Bu3SnH, toluene, AIBN, 90 °C, 1.5 h; 48% for 42a, 4% for 42b, and 4% for 42c or (v) Bu3SnH, toluene, AIBN, reflux, 2.5 h, 60% for
43a/b/c.
yields. The separated products 38a, 38b, and 38c have been
confirmed to be pure β isomers by 1D-NOE experiments
(Supporting Information, Figures SII.16/17, SIII.12/13/14,
SIV.16/17).
be a mixture of the second isomer (70S) of the 5-exo cyclization
product 41b and 60-endo cyclization product 41c,14b which were
separated by preparative HPLC. The isolated yields were 38, 4,
and 3.9% for the major isomer 41a (70R), the minor isomer 41b
(70S), and the 6-endo product 41c, respectively. Similarly, radical-
mediated cyclization of guanosine 40b gave two 5-exo cyclization
products 42a (70R, 48%) and 42b (70S, 4%) and one 60-endo
cyclization product 42c (4%). Cyclization of 5-methylcytosine
derivative 40c generated 5-exo cyclization products 43a and 43b
and 60-endo cyclization product 43c as a mixture (overall yield
60%) in a ratio of 79:12:9 by NMR. For unequivocal character-
ization of the reaction products, this mixture was separated by
HPLC to give analytical samples of 43a, 43b, and 43c (see
Supporting Information, SI General methods).
The 20-O-acetyl groups in compounds 38a, 38b, and 38c were
removed using 2 M aqueous NaOH solution in THF/H2O
mixture, yielding nucleosides 39a, 39b, and 39c in 97%, 88%,
and 95% yields, respectively. Notably, during this deacetylation
step, no side reactions, viz., deprotections at nucleobases, were
observed. The key phenoxythiocarbonyl ester derivative at the 20-
O-position was synthesized by reaction of the 20-hydroxy of 39a,
39b, and 39c with phenyl chlorothionoformate and 1-methyli-
midazole in anhydrous DCM, to form 40a, 40b, and 40c in
satisfactory yields (Scheme 1).
The radical cyclization reaction was carried out under reflux in
degassed (N2) anhydrous toluene using Bu3SnH with radical
initiator AIBN, which were added slowly in a dropwise manner to
ensure that the radical generated has an adequate lifetime to
capture the double bond before it is quenched by hydrogen. The
radical cyclization reaction for adenosine 40a resulted in two
spots on the TLC. The higher-Rf spot was major compound 41a
which was characterized as one of the isomers (70R) formed by
the 5-exo cyclization reaction, and the lower-Rf spot was found to
It is known that cyclization of 5-hexenyl-1-radical is a kineti-
cally controlled process and generally prefers the exo cyclization
product (exo/endo > 98/2).19 In the present study, treatment of
the radical precursor 40a, 40b, or 40c with Bu3SnH and AIBN in
refluxing toluene gave C20 carbon radical intermediates TS I, TS
II, and TS III, respectively (Figure 4), followed by intramolecular
addition of the C20 radical to the tethered CdC, giving 5-exo
cyclization products (∼92%) plus ∼8% of 6-endo cyclization
products. We previously reported that radical cyclization of the
4411
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
3
thymine counterpart also furnished 5-exo and 6-endo products in
a ratio of 92:8.14b Hence, it seems no matter what nucleobase is at
the anomeric center, cyclization of 5-hexenyl-1-radical in these
nucleoside systems generates a slightly increased 6-endo product
compared with normal cases (8% versus 2%).
COSY spectra (Figure 5), and second, by the observable JHC
HMBC correlation between H10 and C70, H20 and C60, and 2JHC
HMBC correlation between H70 and C20 (Figure 6). These
protonꢀproton connectivities in the COSY and proton to
carbon connectivities in HMBC proved that the oxa-bicyclo-
[2.2.1]heptane ring systems in cLNA-A, -G, and -MeC nucleo-
sides have indeed been formed during the radical cyclization.
The configuration of C10 and C70 centers in cLNA nucleosides
41a/b, 42a/b, and 43a/b has been determined by 1D NOE
experiments (see Supporting Information Figure SII 65ꢀ66,
86ꢀ87, SIII 48ꢀ51, 68ꢀ69, SIV 72ꢀ73, 91ꢀ92). For 41a,
irradiation of H30 led to 3% of NOE enhancement for H8,
suggesting that the 9-adeninyl moiety is in β configuration and in
an anti conformation across the glycoside bond. The fact that the
NOE enhancement of 2.9% for H10 upon irradiation on 70-CH3
has been observed indicates that the methyl group on C70 is in
1.2. Structural Evidence from NMR for cLNA-A, -G, and -MeC
Nucleosides As Well As 6-endo Cyclization Products. cLNA
nucleosides 41a/b, 42a/b, and 43a/b and 6-endo cyclization
products carbocyclic ENA nucleosides 41c, 42c and 43c have
been unambiguously characterized using 1D and 2D NMR
13
including 1H, 13C, DEPT, as well as NOE, COSY, 1Hꢀ C
HMQC, and long-range 1Hꢀ C correlation (HMBC)
13
experiments.
The formation of the C20ꢀC70 bond in cLNA nucleosides
41a/b, 42a/b, and 43a/b has been confirmed unequivocally:
first, by the presence of correlation between H20 and H70 in
close proximity of H10 (dH1 ,7 Me ≈ 2.1 Å) in 41a, thereby
confirming the R-configuration at the C70 center (Figure 7). On
the other hand, in the minor product 41b, irradiation of H10 led
0
0
to 5.5% NOE enhancement for H70 (dH1 ,H7 ≈ 2.1 Å) but none
0
0
for 70-Me (dH1 ,7 Me ≈ 3.8 Å). Hence, the stereochemistry at C70
has been assigned to S configuration in compound 41b. In a
similar way, the C70 of the major cyclization products 42a and
43a has been assigned to R-configuration, and the minor
products 42b and 43b to C70-S configuration (Figure 7).
For the 6-endo cyclization product 41c, the ring closure
by formation of the C20ꢀC80 bond has been unequivocally
0
0
Figure 4. 5-Hexenyl-1-radical cyclization through competing 5-exo
versus 6-endo pathway.
3
confirmed by the following NMR observations: (1) JHH
Figure 5. COSY spectra of compounds' major 5-exo products 41a (panel A), 42a (panel B), and 43a (panel C). The correlations between H70 and H20
indicate that the C20ꢀC70 bonds have been formed in these compounds.
4412
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 6. HMBC spectra of compounds 41a (panel A), 42a (panel B), and 43a (panel C). The correlations between H10 and C70 (indicated by red
dashed lines), H30 and C70 (indicated by red dashed lines), H20 and C60 (indicated by turquoise dashed lines), and H70 with C20 (indicated by black
dashed lines) indicate that the C20ꢀC70 bonds have been formed in these compounds.
correlation between H80 and H20 in the COSY spectrum
addition of benzoyl chloride. Subsequently, treatment with
aqueous ammonia furnished the desired the 30,50-debenzylated
cLNA-ABz (45a). The overall yield of this conversion from 41a to
45a in three steps was 35% (Scheme 2A). 30,50-Debenzylated
cLNA-MeCBz (47a/b/c) as a mixture was also obtained in
moderate yield starting from 43a/b/c by a similar strategy
(Scheme 2C). It should be noted that we have, however,
separated and characterized each component of the mixture of
43a/b/c (79:12:9 by NMR) in a small scale (see the Experi-
mental Section). This separation, however, did not work in
our hands at a larger scale required for the preparation of
the corresponding phosphoramidites (required 4ꢀ5 steps,
Scheme 2). Hence, we proceeded with the mixture of 43a/b/c
to make the phosphoramidites 51a/b/c as described in Scheme 2
(43a/b/c f47a/b/c f 50a/b/c f 51a/b/c).
For the transformation of guanine derivative 42a to phosphor-
amidites, it was critically important to selectively remove the
bulky O6-diphenylcarbamoyl group first. The basic conditions
such as aqueous NaOH, methanolic NH3,21 or fluoride ion22
only led to a mixture of desired and undesired products. Another
strategy for this purpose is full deprotection followed by selective
reprotection of N2, but it was laborious and time-consuming. We
have also tried to treat 42a with 90% trifluoroacetic acid;23
however, only 25ꢀ30% of the desired compound 42a0 was
3
(Figure 8A); (2) JHC HMBC correlations between H10 and
C80 (Figure 8C); (3) DEPT (Supporting Information, Figure
SII.98) and HMQC (Figure 8B) spectra showed that both C70
and C80 are secondary carbons, each of them having two protons
attached. Similar NMR experiments have also been carried out
to prove that the oxa-bicyclo[3.2.1]octane ring system has
been formed for compounds 42c (Figure 8DꢀF) and 43c
(Figure 8GꢀJ).
1.3. Synthesis of Phosphoramidites of cLNA-A, -G, and -MeC.
To transform the cLNA-ABz nucleoside to the corresponding
phosphoramidites, the major isomer 41a was subjected to 30,50-
debenzylation, using the Pd(OH)2/ammonium formate
procedure.14a This reaction led to desired debenzylated com-
pound 45a which was however contaminated with N-debenzoy-
lated product 44a. To avoid formation of any N-alkylated
product later on, we decided to deprotect the N6-benzoyl group
of 41a before 30,50-debenzylation (Scheme 2A). Thus, treatment
of 41a with methanolic ammonia at rt for 16 h furnished N6-
debenzoylated compound 41a0, which was then subjected to
classical 20% Pd(OH)2/C, ammonium formate procedure, giv-
ing debenzylated nucleoside 44a in moderate yield. Following
Jones0 transient protection method,20 the nucleoside 44a was
treated with TMS-Cl in dry pyridine and followed by in situ
4413
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 7. NOE contacts in compounds 41a/b, 42a/b, and 43a/b. In 41a, 42a, and 43a, the observation of NOE enhancement between H10 and 70-Me
and complete absence of NOE between H10 and H70 suggest the C70-R configuration. The reverse observation in compounds 41b, 42b, and 43b
indicates the C70-S configuration for them. The β orientation and anti conformation across the glycoside bond of the base moieties for all of these
compounds can be confirmed by the strong NOE between H30 and H6 or H8 of nucleobases.
obtained, and the rest was a mixture of polar deglycosylated
products. However, a treatment with glacial acetic acid under
moderately warm conditions, followed by RT stirring, was found
successful, giving the desired O6-deprotected nucleoside 42a0 in
60ꢀ62% yields. Debenzylation of 42a0 using the classical Pd-
(OH)2/ammonium formate method furnished only N2-deace-
tylated product, and 30,50-dibenzyl groups were completely
intact. Fortunately, use of formic acid as a hydride source instead
of ammonium formate was found to be helpful.24 Thus, 42a0 was
debenzylated using 20% Pd(OH)2/C and formic acid in dry
methanol, yielding the desired diol 45b in 60% yield
(Scheme 2B).
The 50-primary hydroxyl group in 45a, 45b, and 47a/b/c
was selectively protected by DMTr according to a standard
procedure,14a yielding 48a, 48b, and 50a/b/c, respectively.
Phosphitylation of 48a, 48b, and 50a/b/c was first tried using
4414
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 8. COSY, HMQC, and HMBC spectra of compounds 41c (panels AꢀC), 42c (panels DꢀF), and 43c (panels GꢀJ). The 3JHH correlation
between H80 and H20 in COSY spectra (panels A, D, and G, indicated by pink dashed lines) and 3JHC HMBC correlations between H10 and C80 (panels
C, F, and J, indicated by green dashed lines) and H800 and C10 (panel H, indicated by red dashed lines) suggest that the oxa-bicyclo[3.2.1]octane ring
system has been formed for compounds 41c, 42c, and 43c, which can be further confirmed by HMQC (panels B, E, and I) spectra, showing both C70 and
C80 have two protons (correlations are indicated by pink dashed lines for C80 with H80 and H800, indicated by orange dashed lines for C70 with H70 and
H700) attached.
4415
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Scheme 2a
a Reagents and conditions: (i) methanolic ammonia, rt, 16 h; (ii) 20% Pd(OH)2/C, ammonium formate, methanol, reflux, 16 h; (iii) a. TMSCl,
dry pyridine; b. BzCl, c. aq NH3; (iv) acetic acid, 55 °C; (v) 20% Pd(OH)2/C, formic acid, methanol, reflux, 2.5 h; (vi) Bz2O, pyridine, rt, overnight;
(vii) DMTrCl, pyridine, rt, overnight; (viii) 2-cyanoethyl-N,N-diisopropylphosphoramidochloridite, DIPEA, dry DCM, rt, 2 h.
standard conditions,14a 2 equiv of 2-cyanoethyl N,N-diisopro-
pylphosphoramidochloridite in the presence of DIPEA with
anhydrous and radical free THF as solvent, rt, overnight under
the inert N2 environment. However, this methodology led to a
consistent polar impurity in invariable amount along with desired
phosphoramidite products 49a, 49b, and 51a/b/c. These polar
impurities show mass equal to M þ 16 and were inseparable
by column chromatography and various means of precipitations
and washings. The polar contamination for the MeC derivative
51a/b/c was found more pronounced. 31P NMR of a crude
reaction mixture of 51a/b/c showed six peaks at δ 150.20,
150.04, 149.46, 149.22, 148.94, and 148.80 and additional signals
between δ 7.0 and 8.4. The upfield shift in 31P NMR spectra and
the mass analysis suggest the polar impurities are oxidized
phosphate P(V) derivatives.25,26
Clearly, to circumvent the oxidation problem in the phosphi-
tylation reaction, use of (oxygen-free) nonpolar solvents and the
reduction of reaction time as well as lesser exposure of reaction
4416
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Table 1. Tm Values of Duplexes Formed by Modified AONs with Complementary RNA and DNAa
T
m of
ΔTm/mod of
AON/RNAb
Tm of
ΔTm/mod of
AON/DNAc
d
oligo name
sequence of AONs
AON/RNA
AON/DNA
ΔTm
AON 1
AON 2
AON 3
AON 4
AON 5
AON 6
AON 7
AON 8
50 TCC CGC CTG TGA CAT GCA TT
50 TCC CGC CTG TGA CAT GCA TT
50 TCC CGC CTG TGA CAT GCA TT
50 TCC CGC CTG TGA CAT GCA TT
50 TCC CGC CTG TGA CAT GCA TT
50 TCC CGC CTG TGA CAT GCA TT
50 TCC CGC CTG TGA CAT GCA TT
50 TCC CGC CTG TGA CAT GCA TT
74.0 °C
77.5 °C
80.0 °C
82.0 °C
83.0 °C
78.0 °C
79.0 °C
87.5 °C
0 °C
71.0 °C
74.7 °C
71.2 °C
76.7 °C
76.4 °C
72.4 °C
72.8 °C
76.3 °C
0 °C
þ3.0 °C
þ2.8 °C
þ8.8 °C
þ5.3 °C
þ6.6 °C
þ5.6 °C
þ6.2 °C
þ11.2 °C
þ1.2 °C
þ2.0 °C
þ2.7 °C
þ3.0 °C
þ1.3 °C
þ1.7 °C
þ4.5 °C
þ1.2 °C
þ0.1 °C
þ1.9 °C
þ1.8 °C
þ0.5 °C
þ0.6 °C
þ1.8 °C
a Tm values measured at the maximum of the first derivative of the melting curve (A260 nm vs temperature) in medium salt buffer (60 mM tris-HCl at pH
7.5, 60 mM KCl, 0.8 mM MgCl2) with temperature 20ꢀ90 °C using 1 μM concentrations of two complementary strands. The values of Tm given are
averages of three independent measurements (the error of the three consecutive measurements is within (0.3 °C). A = cLNA-A; G = cLNA-G; C =
cLNA-MeC; T = cLNA-T. b ΔTm of AON/RNA/mod was obtained by comparing the Tm of AON/RNA with that of the native AON 1/RNA. c ΔTm of
AON/DNA/mod was obtained by comparing the Tm of AON/DNA with that of the native AON 1/DNA. d RNA selectivity: ΔTm = (Tm of AON/
RNA) ꢀ (Tm of AON/DNA). See footnote under ref 31 for why the AONꢀRNA duplex is more stable than the AONꢀDNA counterpart.
mixture to the environmental conditions were inevitable condi-
tions to follow. Thus, modified phosphitylation reactions were
carried out in dry DCM (dried over P2O5), using 2 equiv of
2-cyanoethyl N,N-diisopropylphosphoramidochloridite in the
presence of DIPEA for invariably 2ꢀ2.5 h of reaction time.
Phosphoramidites 49a, 49b, and 51a/b/c were obtained in 81%,
77%, and 75% yield, respectively. However, under this condition,
MeC-phosphoramidites 51a/b/c were found still contaminated
by 3ꢀ5% of P(V) impurity 52, which was hard to be removed.
Given the nature of the P(V) impurity that is relatively inert
compared to active P(III) phosphoramidite, 3ꢀ5% of this
impurity has not proven to be harmful for the oligomer synthesis
in purity and yield. Hence, we did not invest further efforts for
purification of 51a/b/c.
2.0. Thermal Denaturation Studies. The modified phos-
phoramidites 49a and 49b and a mixture of 51a/b/c were
incorporated into a 20mer AON through the solid phase synth-
esis protocol27 on an automated DNA/RNA synthesizer. The
sequences of AONs and Tm values of duplexes formed by these
AONs with complementary RNA and DNA are listed in Table 1.
2.1. cLNA-T, A, G, and MeC Modifications in AON Strands Lead
to Different Tm Increase for AON/RNA Hybrids. Just like cLNA-T,
cLNA-A, -G, and -MeC modifications in the same AON sequence
also led to Tm increase for the AON:RNA duplex, but the ΔTm
varied significantly with different base types. Triple cLNA-A,
triple cLNA-G, triple cLNA-MeC, and triple cLNA-T modification
in the AON sequence resulted in a 3.5, 6, 8, and 9 °C increase,
respectively. Hence, their RNA affinities decrease in the following
rank: cLNA-T (þ3.0 °C/modification) > cLNA-MeC (þ2.7 °C/
modification) > cLNA-G (þ2.0 °C/modification) > cLNA-A
(þ1.2 °C/modification). It seems that pyrimidine-type cLNAs
show higher RNA affinity than purine-type counterparts, which is
in contrast to 10,20-oxetane modification.8 Therefore, introduction
of pyrimidine-type cLNA modification in the oligos is preferable
for the design of the antisense strand in the antisense approach.
This presumably suggests that the relative strength of cLNA-
modified-AON:RNA base pairing versus oxetane-modified-AON:
RNA base pairing is dictated by the choice of modification site as
well as by the type of chemical modification employed because the
relative stability of the duplex is dictated by the combined strength
of stacking and base pairing ability in each strand which are
different depending upon the sequence context.
For AONs containing both cLNA-A and cLNA-G modifica-
tions such as AON 6 and 7, Tm rise also followed 1 °C/cLNA-A
modification and 2 °C/cLNA-G modification. Interestingly, the
Tm of AON 8/RNA was found to be 13.5 °C higher than the
native counterpart, which is much higher than expectation given
that AON 8 contains three cLNA-A and three cLNA-G mod-
ifications. This observation indicates that cLNA-A and cLNA-G
modification can lead to a much more pronounced Tm rise in the
AON sequence with multiple modifications.
2.2. cLNA-T, A, -G, and -MeC Modified AONs Are RNA-
Selective, But the Magnitudes of RNA Selectivity Are Different.
All the Tm values of cLNA modified AON 2ꢀ8/DNA are higher
than that of native AON 1/DNA, suggesting that cLNA mod-
ifications in the AON strand may also result in increased affinity
toward cDNA compared to the native type, while these results
were found to be different from the Tm results obtained in our
lab,14a in which we have shown that cLNA-T (isomeric mixture
of 70-R/S-Me) modified oligos lead to a small decrease in
thermodynamic stability with cDNA compared to the native
type. This is most probably due to the different sequences
(15mer14a versus 20mer in this paper) and modification numbers
and sites employed here, which caused a positive effect on the
stability of cLNA-T modified AON/DNA duplexes versus the
native AON/DNA duplex. However, we found for all of the
modified AON strands that the Tm of AON/DNA is lower
than the Tm of AON/RNA, indicating cLNA modified AONs
are RNA-selective (see footnote under ref 32 for why the
AONꢀRNA duplex is more stable than the AONꢀDNA coun-
terpart). The magnitude of RNA-selectivity, ΔTm (ΔTm = [Tm of
AON/RNA] ꢀ [Tm of AON/DNA]), was found to be 2.8, 8.8,
5.3, and 6.6 °C, respectively, for triple cLNA-A modified AON 2,
triple cLNA-G modified AON 3, triple cLNA-C modified AON
4, and triple cLNA-T modified AON 5. Hence, it is likely that
RNA selectivity of each type of cLNA decreases in the following
rank: cLNA-G (ΔTm = 2.9 °C/modification) > cLNA-T (ΔTm =
2.2 °C/modification) > cLNA-C (ΔTm = 1.8 °C/modification)
> cLNA-A (ΔTm = 0.9 °C/modification). For AONs 6ꢀ8, which
contain both cLNA-A and cLNA-G, each cLNA-A and cLNA-G
modification seems to lead to a magnitude of RNA selectivity
very similar to that found for AON 2 and AON 3.
3.0. Nuclease Stability Studies of cLNA Modified AONs.
Chemically modified AONs are required especially to have
4417
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 9. Denaturing PAGE analysis of the SVPDE degradation of AONs with cLNA modifications. The AON % left after 24 h incubation for each
AON is shown below the corresponding band. Digestion conditions: AON 3 μM (50-end 32P labeled with specific activity 80 000 cpm) in 100 mM Tris-
HCl (pH 8.0) and 15 mM MgCl2, 21 °C, total reaction volume 30 μL, SVPDE concentration (6.7 ng/μL).
resistance toward nuclease degradation for their in vivo applications,
whereas the predominant nuclease activity in vivo comes from 30-
exonuclease.28 Therefore, the newly synthesized AONs 2ꢀ8 were
treated with 30-exonuclease (phosphodiesterase I from Crotalus
adamanteus venom [SVPDE] used in this study) for testing their
nuclease stability. The selected AONs were labeled at the 50-end
with 32P and then incubated with SVPDE [SVPDE 6.7 ng/μM,
100 mM Tris-HCl (pH 8.0), 15 mM MgCl2, total volume 30 μL]
at 21 °C. Aliquots were taken out at appropriate time intervals and
analyzed by 20% denaturing PAGE.
Thus, total percentages of integrated various bands of AONs
mentioned above were plotted against time points to give the
SVPDE digestion curve for AONs 1ꢀ8 in Figure 10, and
pseudofirst-order reaction rates were obtained by fitting the
curves to single-exponential decay functions. A comparison of
digestion rates of AON 2ꢀ8 gave the following results and
implications.
3.1. Relative 30-Exonucleolytic Stabilities of cLNA-A, -G, -MeC,
and -T Modified AONs. Through comparison of pseudofirst-
order reaction rates, we found that the stabilities of AONs 2ꢀ8
toward SVPDE incubation decreased in the following order:
cLNA-MeC modified AON 4 (k = 0.0002 minꢀ1) > cLNA-T
modified AON 5 (k = 0.0008 minꢀ1) > cLNA-G and A modified
AON 7 (k = 0.0009 ( 0.0001 minꢀ1) > cLNA-G and A modified
AON 6 (k = 0.0013 ( 0.0001 minꢀ1) > cLNA-G and A modified
AON 8 (k = 0.0013 ( 0.0001 minꢀ1) > cLNA-A modified AON
2 (k = 0.0014 ( 0.0001 minꢀ1) > cLNA-G modified AON 3 (k =
0.0040 minꢀ1). This result suggests all the cLNA modified AONs
2ꢀ8 (despite various positions or base modifications) are much
more stable than the native counterpart, which degraded com-
pletely within several minutes.
From the gel pictures (Figure 9), we found the native AON 1
was completely degraded in several minutes under the present
conditions, while the other cLNA modified AONs 2ꢀ8 showed
considerably improved 30-exonuclease resistance to a variable
extent (most of them stand until 24 h). Because the first
modification site from the 30-end of AON 2ꢀ8 is located at
different positions within the oligo (see Table 1), the band
corresponding to 19mer of AON 2, 17mer of AON 3, 18mer and
19mer of AON 4, 16mer plus 17 and 18mer of AON 5, 19mer of
AON 6, 17mer of AON 7, and 19mer of AON 8 could be
observed on the PAGE pictures (Figure 9).
Since SVPDE has no phosphatase activity, the 50-32P labels
should be intact during SVPDE incubation. Hence, the
relative band intensities quantified after autoradiography can
reflect their real relative quantity for bands in the same lane.
3.2. Effect of Various Base Moieties of cLNAs on 30-Exonu-
cleolytic Stability. We found cLNA-MeC modified AON 4 was
the most stable among all cLNA modified AONs. It is 4 times
more stable than cLNA-T modified AON 5, 7 times more stable
4418
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 10. SVPDE digestion curves of native AON 1 and modified AONs 2ꢀ8. The pseudofirst-order rates shown here were obtained by fitting the
curve to the single-exponential decay function. Digestion conditions: AON 3 μM DNA concentration (50-end 32P labeled with specific activity 80 000
cpm), 100 mM Tris-HCl (pH 8.0), 15 mM MgCl2, [SVPDE] = 6.7 ng/μL, total reaction volume was 30 μL, 21 °C.
Table 2. Comparison of MichaelisꢀMenten Kinetic Parameters of Cleavage of Oligonucleotides Modified at the 30-End with
Different Native Deoxyribo- and Ribonucleotides by SVPDEa
oligo name
sequence
KM (μM)
kcat (minꢀ1
)
kcat/KM (μMꢀ1 minꢀ1
)
relative kcat/KM
3
AON 9
(Tp)13TpsTpT
(Tp)13TpsTpC
(Tp)13TpsTpA
(Tp)13TpsTpG
(Tp)13TpsTprU
(Tp)13TpsTprC
(Tp)13TpsTprA
(Tp)13TpsTprG
2.88 ( 0.65
2.63 ( 0.59
2.63 ( 0.62
0.66 ( 0.19
1.80 ( 0.32
3.19 ( 0.53
3.97 ( 0.91
3.47 ( 0.81
36.55 ( 2.1
22.76 ( 1.4
11.8 ( 2.5
14.0 ( 3.6
11.4 ( 2.0
15.8 ( 7.9
8.3 ( 1.2
5.4 ( 1.5
6.0 ( 3.7
5.6 ( 1.9
1
AON 10
AON 11
AON 12
AON 13
AON 14
AON 15
AON 16
1.19
0.97
1.34
0.70
0.46
0.51
0.47
26.2 ( 1.0
13.1 ( 0.7
16.3 ( 0.7
18.58 ( 0.76
19.4 ( 1.1
17.7 ( 1.1
a The digestion curves are shown in Supporting Information Figure SI.2ꢀ9.
than cLNA-A modified AON 2, and even 20 times more stable
than cLNA-G modified AON 3. Although given that cLNA-T
contains 70-R/S-Me product and cLNA-MeC contains 70-R/S-Me
(46a/b) and endo-product (46c) as mixtures, the results of their
incorporation into oligonucleotides clearly show their usefulness
as RNA-directed agents.31 Especially for cLNA-MeC, only a small
amount (≈10%) of 70-S-Me-cLNA-MeC and the endoproduct
(46c) exist (≈3ꢀ5%) in the mixtures, which is not expected
to significantly influence the results of nuclease stability com-
pared to the obvious difference of nuclease resistance between
cLNA-MeC and cLNA-T/A/G (see Figure 10). Similarly, the Tm
difference found is well within the error limit, given the poor
accuracy of the Tm measurement ((0.5 °C among three con-
secutive measurements). It is noteworthy that while the bio-
chemical properties of oligos containing cLNA-MeC and -T
represent the total property contributed by each isomeric
component, the properties of cLNA-A/G modified oligos re-
present the properties contributed by 70-R-Me-cLNA-A and -G.
(k = 0.0040 minꢀ1). These results suggest the cLNAs with
pyrimidine moieties seemed to give higher nuclease resistance
than those of cLNAs with the purine moieties.
3.3. Effect of Various Nucleobase Moieties on 30-Exonucleo-
lytic Stability for Natural Nucleotides. To figure out if the
digestion rate of natural nucleotides by SVPDE is also nucleobase
dependent, MichaelisꢀMenten kinetics analysis of AONs 9ꢀ16
has been performed. AONs 9ꢀ16 are 15mer oligonucleotides
(with general formula [50-(Tp)13TpsTpN-30, where N = A/G/T/
C/rA/rG/rU/rC], see Table 2), and the only difference between
them is that they have different 30-end nucleotides (N). For all of
them, the second internucleotidyl phosphate linkage from the 30-
end has been modified to phosphorothioate (ps), which is known
to be fairly resistant toward SVPDE.29 Hence, the enzyme
only recognizes and cleaves the first nucleotide from the 30-end
at the initial stage of the reaction (see Supporting Information
Figure SI.1).
The MichaelisꢀMenten kinetic parameters for digestion of
each oligonucleotide by SVPDE are listed in Table 2. For AONs
9ꢀ12 which have deoxyribonucleotides (N) in the 30-end,
their digestion efficiencies (kcat/KM) are in the same magnitude,
In addition, cLNA-T modified AON 5 (k = 0.0008 minꢀ1
)
was also found to be more stable than cLNA-A modified AON 2
(k = 0.0014 ( 0.0001 minꢀ1) and cLNA-G modified AON 3
4419
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
around 101 μMꢀ1 Minꢀ1. (Tp)13TpsTpC (AON 10) and
modification site in the 20-deoxyoligo-AON strand to adopt an
RNA-type conformation, which has been evidenced by the
change in RNase H cleavage pattern of the corresponding
AON:RNA hybrid compared to the native counterpart.14aꢀf By
appropriately choosing the cLNA modification sites in the 20-
deoxy-AON strand, the digestion centers of 20-deoxy-AON:RNA
by RNase H can be totally repressed in certain regions or can be
engineered to cleave at a specific site (as in catalytic RNA) due to
conformational transformation of those site(s) to the RNA:RNA
type due to cLNA modifications.
3
(Tp)13TpsTpG (AON 12) can be cleaved slightly faster than
(Tp)13TpsTpT (AON 9) and (Tp)13TpsTpA (AON 11). How-
ever, SVPDE digestion of 30-end ribonucleotides in AONs
13ꢀ16 was found much slower than digestion of 30-deoxyribo-
nucleotides in AONs 9ꢀ12, which is consistent with a previous
report that deoxyribonucleotides are hydrolyzed faster than the
ribonucleotides by SVPDE.30 It seems that both increased KM
and decreased kcat contribute to the decreased digestion effi-
ciency (kcat/KM) for 30-ribonucleotides in AONs 13ꢀ16 com-
pared to digestion of 30-deoxyribonucleotides in AONs 9ꢀ12.
The kcat/KM of digestion of AONs 13ꢀ16 varied from 5.4 to
If the modification sites are chosen appropriately as in AON 2,
AON 4, AON 6, and AON 7, the complementary RNA strand of
the modified AON:RNA hybrids can be recognized and cleaved
by RNase H reflecting the site-specific presence of cLNA
modifications. On the contrary, RNase H degradation of the
RNA strand in the AON 3/RNA, AON 5/RNA, and AON 8/
RNA is totally repressed because in them the modification
stretch covers the whole possible cleavage sites (Figure 12).
Hence, we can effectively protect a complementary RNA from
RNase H digestion (e.g., for steric blocking using the antisense
approach) by engineering an appropriately modified 20-deoxy-
AON in the 20-deoxy-AON/RNA heteroduplex, as duplexes
AON 3/RNA, AON 5/RNA, and AON 8/RNA (Figure 11).
Alternatively, we can, at will, engineer multiple RNA cleavage
site or a single RNA cleavage site at a specific point within the
RNA strand in the 20-deoxy-AON/RNA heteroduplex by judi-
ciously choosing the position of cLNA modification in the
complementary AON strand, which is evidenced by the single
site cleavage at the 30f 50 phosphodiester bond of A13ꢀG14 in
the AON 4/RNA heteroduplex (Figure 11).31 This cLNA
modified oligonucleotide-induced engineering of a specific clea-
vage point in the complementary RNA strand with the help of
RNase H (which is omnipresent in the mammalian cells) is
reminiscent of the characteristic RNA cleavage at a single
cleavage site found in RNA catalysis because of the change
brought about by the change of the internal folding motif within
RNA with or without a cofactor as in, e.g., hairpin ribozyme, or
those induced by a second RNA strand (e.g., hammerhead) or in
DNAzyme in which the RNA folding motif is induced by a
complexing DNA strand.
8.3 μMꢀ1 minꢀ1, suggesting SVPDE does not differentiate the
3
different nucleobases among the 30-end ribonucleotides.
Thus, SVPDE digests of four different native deoxyribonucleo-
tides (AONs 9ꢀ12) with similar efficiency (well within the error
limit of MichaelisꢀMenten kinetic parameters) as those found
for four different native ribonucleotides (AONs 13ꢀ16) show
that there is no native nucleobase preference for digestion by
SVPDE (Table 2). Hence, the much different reactivity of cLNA-T, -
A, -G, and -MeC toward SVPDE, vis-ꢁa-vis native counterparts, as
described in Section 3.2, is a property of cLNAs.
4.0. Human Blood Serum Studies of Carba-LNAs Modified
AONs. The newly synthesized AONs 2ꢀ8 with modifications of
cLNA-A, -G, -MeC, and -T at different positions were also assayed
for stability in human blood serum. Thus, the AONs (50-end 32P
labeled) were incubated with human blood serum (male, type
AB) for up to 48 h at 21 °C, and aliquots were taken out at regular
time intervals and then analyzed by 20% denaturing PAGE. The
gel pictures obtained by autoradiography are shown in the
Supporting Information Figure SI.10. Because of the presence
of alkaline phosphatase in blood serum which gradually removes
the 50-end 32P label, the determination of the accurate degrada-
tion rate for each AON by quantifying the gel picture is not
possible. Therefore, through visual comparison of the gel pic-
tures, we found nearly all modified 20mer AONs 2ꢀ8 can sustain
in blood serum for more than 48 h, while the native AON 1 can
only stand until 12 h under identical conditions. Among all the
studied AONs 1ꢀ8, cLNA-MeC modified AON 4 was found to
be the most stable oligonucleotide upon blood serum treatment,
which was well in agreement with that observed upon treatment
of 30-exonuclease. The relative stabilities in blood serum for
all modified AONs in the present study are as follows: cLNA-MeC
modified AON 4 > cLNA-T modified AON 5 ≈ cLNA-G and A
modified AON 8 > cLNA-G modified AON 3 ≈ cLNA-G and
A modified AON 6 ≈ cLNA-G and A modified AON 7 > cLNA-A
modified AON 2 > native AON 1. It therefore appeared that the
order of relative stabilities of AONs in human blood serum was
similar with those observed upon treatment of 30-exonuclease
SVPDE.
Additionally, certain specific cLNA modification in AON, as
in the AON 4/RNA heteroduplex, made the corresponding 20-
deoxy-AON:RNA hybrid an almost 2-fold better substrate to
RNase H (Figure 13). So what we show here is that simple
designer construction of cLNA modified 20-deoxyoligo, when
complexed to complementary RNA, can in fact perform a dual
job with the help of RNase H ((a) specific cleavage of the
complementary RNA strand at a specific site) and can also cause
a (b) considerable enhancement of the RNA cleavage rate
compared to that of the native.
5.0. RNase H Digestion Study of cLNA Modified AONs.
The RNase H recruitment study of cLNA-A, -G, -MeC, and -T
modified AONs 2ꢀ8 have been carried out and compared with
the native AON 1. From the gel picture (Figure 11), we found for
the native AON 1/RNA hydrid that the Escherichia coli RNase
H1 cleaves the central part of the RNA strand (between A6 and
G14) with a preference on A13. Incorporation of cLNA mod-
ifications in the AON strands leads to a stretch of a 5ꢀ6 base
pairs region in the RNA strand which is completely resistant to
RNase H1 mediated cleavage (Figure 12). This is due to the
reason that cLNA-T, cLNA-A, G, and MeC modifications in AON
can modulate the conformation 5ꢀ6 base pairs region next to the
So, we present evidence that shows that these cLNA modified
oligos can become an important tool for RNA engineering which
potentially can complement ribozyme function.
For AON 1/RNA, AON 2/RNA, AON 4/RNA, AON 6/
RNA, and AON 7/RNA, total percentages of intact RNA were
plotted against time points to give the RNase H digestion curve
for every AON:RNA hybrid, and pseudofirst-order reaction rates
were obtained by fitting the curves to single-exponential decay
functions. Through comparison of digestion rates of each AON:
RNA hybrid (Figure 13), we have found that the cleavage rates of
AON 2, 6, or 7/RNA hybrids are only around half that of the
native AON:RNA hybrid, while the cLNA-MeC modified AON
4420
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 11. Autoradiograms of 20% denaturing PAGE pictures, showing the cleavage kinetics of 50-32P-labeled target RNA by E. coli RNase H1 in the
AON/RNA hybrids. The intact RNA% left after 1 h incubation for each AON/RNA is shown below the corresponding band. Conditions of cleavage
reactions: RNA (0.1 μM) and AONs (2 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 H1, total volume 30 μL.
4/RNA hybrid exhibits 2 times higher cleavage rate compared to
the native AON:RNA hybrid. Given the fact that AON 2/RNA,
AON 4/RNA, AON 6/RNA, and AON 7/RNA showed a very
similar cleavage pattern by RNase H (Figures 11 and 12) but with
much different cleavage rates (Figure 13), we can arrive at the
conclusion that cLNA-A, -G, -T, and -MeC nucleoside modifica-
tion in the same AON sequence can lead to a much different
effect on RNase H digestion efficiency.
modification) and cLNA-T (þ3 °C/modification) mod-
ifications showed higher RNA affinity than purine type
cLNA-G (þ2 °C/modification) and cLNA-A (þ1 °C/
modification) modifications. For an explanatory note of
why AON/RNA duplexes are generally more stable than
the AON/DNA counterpart, see ref 32.
(2) All the cLNAs-A, G, and MeC modified 20mer AONs were
found to be RNA selective. One cLNAs-G modification
(ΔTm = 2.9 °C/modification) leads to the greatest RNA
selectivity, and one cLNA-A modification leads to the
least RNA selectivity (ΔTm = 0.9 °C/modification), with
’ CONCLUSIONS AND IMPLICATIONS
cLNA-T (ΔTm = 2.2 °C/modification) and cLNA-Me
(ΔTm = 1.8 °C/modification) in between.
C
In the present study, cLNA-A, -G, and cLNA-MeC nucleosides
have been synthesized through an intramolecular radical cycliza-
tion. Biological properties of AONs containing 70R-Me-cLNA-A,
70R-Me-cLNA-G, and (exo/endo products) cLNA-MeC modifica-
tions have been evaluated and compared with 70R/S-Me-cLNA-
(3) The nucleolytic stability of cLNA was found to vary
greatly with different nucleobase moieties. cLNA-MeC
modified AON was found to be most stable toward
SVPDE, followed by cLNA-T and cLNA-A modified
counterparts. cLNA-G modified AON was found to be
most labile. On the contrary, studies using native DNA
and RNA models showed that SVPDE exhibited very
similar activity toward four deoxyribonucleotides as well
T
14a modified counterpart as well as with that of the native AON.
The major conclusions are as follows:
(1) All the cLNAs-A, -G, and -MeC modified 20mer AONs
showed improved RNA affinity compared to the native
counterpart, but pyrimidine-type cLNA-MeC (þ2.5 °C/
4421
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Figure 12. RNase H1 cleavage pattern of hybrid duplexes. Modifications in the 20-deoxy-AON strand in the heteroduplex are shown by an
underscore þ italic þ bold. The boxes show the conformational transmission effect to RNA-type conformation in the AON strand owing to North-
locked cLNA modification(s) in the proximity, and hence the resulting RNA:RNA pseudoduplex is uncleavable by RNase as evidenced by PAGE
footprints in Figure 11. The vertical “red” arrows show the RNA cleavage sites, and the relative length of an arrow shows the relative extent of cleavage
at that site. The square boxes show the stretch of the modification, which is resistant to RNase H1 cleavage. Duplexes AON 3/RNA, AON 5/RNA,
and AON 8/RNA can not be degraded by RNase H because of their overlapping resistance region owing to cLNA modification stretch over the whole
possible cleavage sites. On the other hand, heteroduplex AON 4/RNA has been designed such that we have engineered a single cleavage site in the
target RNA strand.
4422
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Commun. 2011, 2, 206] and allele selective inhibition of
Mutant Huntingtin expression [Biochemistry 2010, 49,
10166ꢀ10178] involving cell culture in vitro and
in vivo experiments. These studies clearly showed that
cLNA is better than any other modification including
LNA in term of IC50, selective inhibition of HTT, and
human blood serum stability. Apart from these studies,
Seth et al. [J. Am. Chem. Soc. 2010, 132, 14942ꢀ14950]
also have studied the potential of cLNA versus LNA
and methylene-cLNA in cell culture (brain endothelial
cells using electroporation) and animal experiments
(subchronic dosing schedule in mice). They found that
cLNA was slightly less potent compared to LNA and
methylene-cLNA modified AONs under employed experi-
mental conditions. It is also clear that it is not always
straightforward to understand why particularly modified
AONs are more effective than others, especially with such
a wide variety of targets and experimental conditions. A
number of factors can complicate an oligonucleotide’s
ability to reach its target RNA and inhibit gene expression.
Simple explanations might include poor transfection or
release from endosomes after cell entry and cellular uptake
(see footnote under ref 33 for exact AONs or siRNA
construct).
Figure 13. Bar plots show the cleavage rates of RNase H1 mediated
degradation of RNA in native and modified AON:RNA hybrid duplexes.
Duplexes AON 3/RNA, AON 5/RNA, and AON 8/RNA can not be
degraded by RNase H, and hence they are not shown here. On the other
hand, the cleavage rate of the AON 4/RNA heteroduplex (with a single
cleavage site) is ca. 2-fold better than that of the native AON 1/RNA
counterpart.
as four ribonucleotides. Hence, the discrimination of
various cLNA modified nucleotides in AON by SVPDE is
a unique property of the cLNAs.
(8) We also present experimental evidence that shows that
these cLNA modified oligos can become an important
tool for RNA engineering which potentially can comple-
ment a ribozyme action.
(4) Just like cLNA-T, cLNA-A, G, and MeC, modifications in
AON change the RNase H cleavage pattern for the AON/
RNA hybrid. By appropriately choosing the modification
sites in AON strands, the digestion of AON/RNA can be
totally repressed or limited to cleavage at a specific site.
(5) When the cleavage patterns are the same, cLNA-MeC
modified AON/RNA can be degraded by RNase H nearly
four times faster than cLNA-A and c-LNA-G modified
counterparts and two times faster than the native
counterpart.
(9) Taken together, cLNA-MeC modified antisense oligonu-
cleotides exhibit higher RNA affinity, RNA selectivity,
greatly improved nuclease stability, and efficient RNase H
recruitment capability compared to cLNA-A, -G, and -T
modified counterparts. Hence, in a mixed sequence the
modification with cLNA-MeC nucleotide units could be
more preferentially recommended than other cLNAs in a
promising therapeutic antisense or siRNA for RNA
targeting and inhibition.
(6) It has been shown that cLNA-T, cLNA-A, G, and MeC
modifications in AON can modulate the neighboring
nucleotides (up to 5ꢀ6 nucleotides) to adopt an RNA-
type conformation, which has been evidenced by the
change in the RNase H cleavage pattern of the corre-
sponding AON/RNA hybrid compared to the native
counterpart. By appropriately choosing the cLNA mod-
ification sites in the 20-deoxy-AON strand, the digestion
centers of 20-deoxy-AON/RNA by RNase H can be
totally repressed in certain regions or can be engineered
to cleave at a specific site (as in catalytic RNA). Addi-
tionally, certain specific cLNA modification in AON made
the corresponding 20-deoxy-AON/RNA hybrid almost a
2-fold better substrate to RNase H than that of the native.
So what we show here is that simple designer construction
of the cLNA modified oligo, when complexed to com-
plementary RNA, can in fact perform a dual job with the
help of RNase H: (a) it can cause specific cleavage of
complementary RNA at a specific site, and (b) it also can
promote a considerable enhancement of the cleavage rate
compared to that of the native.
’ EXPERIMENTAL SECTION
9-[4-C-Allyl-3,5-di-O-benzyl-2-O-acetyl-β-D-ribofuranosyl]-
N6-benzoyladenine (38a). Compound 36 (25.0 g, 60.93 mmol) was
dissolved with acetic anhydride (65 mL, 686 mmol) and acetic acid
(347 mL) and cooled to 0 °C, and then triflic acid (0.5 mL) was added to
this mixture. After 30 min of stirring at 0 °C, the reaction was quenched
with cold saturated NaHCO3 solution and extracted with DCM. The
organic layer was dried and evaporated. The obtained residue was
coevaporated with dry CH3CN (3 ꢂ 650 mL) and dissolved in the
same (650 mL). N6-Benzoyladenine (16 g, 66 mmol) and N,
O-bis(trimethylsilyl)acetamide (33.3 mL, 144 mmol) were added to this
solution and refluxed for 45 min. The solution was cooled to 0 °C, and
TMSOTf (15 mL, 82 mmol) was added dropwise and then allowed to
warm to 50 °C and kept constant at this temperature overnight. The
reaction was quenched with saturated NH4Cl solution and was extracted
with EtOAc (3 ꢂ 450 mL). The organic layer was dried and evaporated,
and the obtained residue was subjected to column chromatography over
silica gel eluting with 70% EtOAc in hexane (v/v) to give 38a as a white
foam (30 g, 86%): Rf 0.60 (5% methanol in DCM, v/v). 1H NMR (600
MHz, CDCl3): δ 8.98 (1H, s, NH), 8.77 (1H, s, H2), 8.33 (1H, s, H8),
8.02(2H, d, J = 7.8 Hz, ArꢀH), 7.60 (1H, t, J = 7.8 Hz, ArꢀH), 7.52 (2H,
t, J = 7.8 Hz, ArꢀH), 7.2ꢀ7.4 (10H, m, ArꢀH), 6.40 (1H, d, J = 4.8 Hz,
H10), 5.92 (1H, t, J = 4.8 Hz, H20), 5.8ꢀ5.95 (1H, m, H70), 5.10 (1H, d,
(7) Recently, carba-LNA (cLNA) incorporated antisense
oligonucleotides (AON) and siRNA have been studied
against various biological targets to evaluate the true
potential of cLNA in various disease models like the
single cell replication model for HIV-1 [Med. Chem.
4423
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
J = 4.8 Hz, H80), 5.08 (1H, m, H80), 4.63 (1H, d, J = 11.4 Hz, CH2Ph),
4.58 (1H, d, J = 5.4 Hz, H30), 4.53 (2H, t, J = 10.8 Hz, CH2Ph), 4.44 (1H,
d, J = 12 Hz, CH2Ph), 3.64 (1H, d, J = 10.2 Hz, H500), 3.42 (1H, d, J = 10.2
Hz, H50), 2.72 (1H, dd, J = 6 and 15 Hz H600), 2.41 (1H, dd, J = 8.4 and
14.4 Hz, H60), 2.07 (3H, s, CH3). 13C NMR (150.9 MHz, CDCl3): δ
169.9 (CH3CdO), 164.5 (PhCdO), 152.7 (C2), 151.8 (C4), 149.4
(C6), 141.8 (C8), 137.5 (ArꢀC), 137.4 (ArꢀC), 137.2 (ArꢀC), 133.8
(ArꢀC), 132.7 (C70), 129.9 (ArꢀC), 128.9 (ArꢀC), 128.7 (ArꢀC),
128.6 (ArꢀC), 128.53 (ArꢀC), 128.49 (ArꢀC), 128.1 (ArꢀC), 128
(ArꢀC), 127.9 (ArꢀC), 127.8 (ArꢀC), 127.7 (ArꢀC), 123.2 (C5),
118.6 (C80), 87.9 (C40), 85.9 (C10), 78.3 (C30), 75.7 (C20), 74.4
(CH2Ph), 73.6 (CH2Ph), 72.5 (C50), 37.2 (C60), 20.6 (CH3). MAL-
DI-TOF m/z [MþH]þ calcd for C36H36N5O6 634.267, found 634.264.
9-[4-C-Allyl-3,5-di-O-benzyl-β-D-ribofuranosyl]-N6-ben-
zoyladenine (39a). Compound 38a (30 g, 47.4 mmol) was dissolved
in THF (120 mL) and cooled to 0 °C, to which a cold solution of NaOH
(7.5 g, 189 mmol) in water (90 mL) was added dropwise over 15 min at
the same temperature, and the mixture was allowed to warm to rt and
stirred for another 2 h. The reaction mixture was quenched with water
and extracted with EtOAc (300 mL). The organic layer was dried over
Na2SO4, concentrated under reduced pressure, and dried under vacuum
132.5 (C70), 129.6 (ArꢀC), 129.5 (ArꢀC), 129.3 (ArꢀC), 128.8
(ArꢀC), 128.7 (ArꢀC), 128.6 (ArꢀC), 128.5 (ArꢀC), 128.2
(ArꢀC), 128.14 (ArꢀC), 128.09 (ArꢀC), 128 (ArꢀC), 127.9
(ArꢀC), 127.8 (ArꢀC), 127.7 (ArꢀC), 126.8 (ArꢀC), 123.1 (C5),
121.6 (ArꢀC), 118.8 (C80), 88.2 (C40), 85.0 (C10), 83.6 (C20), 78.2
(C30), 74.8 (CH2Ph), 73.7 (CH2Ph), 73.0 (C50), 37.3 (C60). MALDI-
TOF m/z [M þ H]þ calcd for C41H38N5O6S 728.254, found 728.249.
(1R,3R,4R,5R,7S)-3-(N6-Benzoyladenin-9-yl)-7-benzyloxy-
1-benzyloxymethyl-5-methyl-2-oxa-bicyclo[2.2.1]heptane
(41a), (1R,3R,4R,5S,7S)-3-(N6-Benzoyladenin-9-yl)-7-benzyloxy-
1-benzyloxymethyl-5-methyl-2-oxa-bicyclo[2.2.1]heptane (41b),
and (1R,5R,7R,8S)-7-(N6-Benzoyladenin-9-yl)-8-benzyloxy-5-ben-
zyloxmethyl-6-oxa-bicyclo[3.2.1]octane (41c). Compound 40a
(10.0 g, 13.7 mmol) was dissolved in dry toluene (400 mL) and bubbled
with N2 gas for 30 min. Bu3SnH (3.7 mL, 13.7 mmol) was dissolved in
40 mL of toluene, and half of this solution was added dropwise to
refluxing solution over 30 min. AIBN (2.3 g, 13.7 mmol) was dissolved in
40 mL of dry toluene and added to the above solution dropwise, and
simultaneously the remaining solution of Bu3SnH was added over 60ꢀ
70 min. After 60 min of reflux, the solution was cooled, CCl4 (10 mL)
added, and the mixture stirred for 20 min. A solution of iodine in DCM
was added to the above solution until a faint coloration persisted, and
then solvent was evaporated. The organic layer was dried, evaporated,
and chromatographed over silica gel (10ꢀ32% EtOAc in hexane, v/v) to
give pure 41a as an off-white foam (3.0 g, 38%). Rf 0.4 (60% EtOAc in
hexane, v/v). The remaining mixture was subjected to preparative HPLC
separation to yield compound 41b as white foam solid (0.30 g, 3.95%)
(Rf 0.35, 60% EtOAc in hexane, v/v) and white solid 41c (0.29 g, 3.89%)
(Rf 0.35, 60% EtOAc in hexane, v/v).
1
to give crude 39a as an off-white foam (27.0 g, 97%). H NMR (600
MHz, CDCl3): δ 9.14 (1H, s, NH), 8.70 (1H, s, H2), 8.19 (1H, s, H8),
7.99 (2H, d, J = 7.8 Hz, ArꢀH), 7.58 (1H, t, J = 7.2 Hz, ArꢀH), 7.49(2H,
t, J = 7.2 Hz, ArꢀH), 7.2ꢀ7.4 (10H, m, ArꢀH), 6.10 (1H, d, J = 5.4 Hz,
H10), 5.82ꢀ5.91 (1H, m, H70), 5.08ꢀ5.16 (2H, m, H80 and H800), 4.89
(1H, m, J = 4.8 Hz, H20), 4.76 (1H, d, J = 11.4 Hz, CH2Ph), 4.72 (1H, d,
J = 11.4 Hz, CH2Ph), 4.53 (1H, d, J = 12 Hz, CH2Ph), 4.47 (1H, d,
J = 11.4 Hz, CH2Ph), 4.38 (1H, d, J = 5.4 Hz, H30), 3.78 (1H, d, J = 4.8
Hz, OH), 3.59 (1H, d, J = 10.2 Hz, H500), 3.46 (1H, d, J = 9.6 Hz, H50),
2.77 (1H, dd, J = 6 and 14.4 Hz, H600), 2.46 (1H, dd, J = 8.4 and 14.4 Hz,
H60). 13C NMR (150.9 MHz, CDCl3): δ 164.5 (PhCdO), 152.5 (C2),
151.6 (C4), 149.4 (C6), 141.8 (C8), 137.4 (ArꢀC), 137.1 (ArꢀC),
133.76 (ArꢀC), 132.86 (C70), 132.7 (ArꢀC), 128.8 (ArꢀC), 128.7
(ArꢀC), 128.6 (ArꢀC), 128.5 (ArꢀC), 128.3 (ArꢀC), 128.1 (ArꢀC),
128.0 (ArꢀC), 127.8 (ArꢀC), 127.7 (ArꢀC), 123.1 (C5), 118.6 (C80),
89.0 (C10), 87.8 (C40), 80.2 (C30), 75.5 (C20), 74.8 (CH2Ph), 73.7
(CH2Ph), 73.1 (C50), 37.2 (C60). MALDI-TOF m/z [M þ H]þ calcd
for C34H34N5O5 592.256, found 592.253.
41a: 1H NMR (400 MHz, CDCl3): δ 9.06 (1H, s, NH-Bz), 8.77 (1H,
s, H2), 8.47 (1H, s, H8), 8.03 (2H, d, J = 7.72 Hz, ArꢀH), 7.20ꢀ7.60
(1H, m, ArꢀH), 7.49ꢀ7.51 (2H, m, ArꢀH), 7.26ꢀ7.35 (8H, m,
ArꢀH), 7.10ꢀ7.20 (2H, m, ArꢀH), 6.24 (1H, s, H10), 4.66 (1H, d, J
= 12.48 Hz, CH2Ph), 4.59 (1H, d, J = 12.36 Hz, CH2Ph), 4.40 (2H, q, J =
14.4 Hz, CH2Ph), 4.20 (1H, s, H30), 3.75 (1H, d, J = 11.0 Hz, H50), 3.66
(1H, d, J = 11.0 Hz, H50), 2.77 (1H, d, J = 3.4 Hz, H20), 2.63ꢀ2.76 (1H,
m, H70), 2.10 (1H, dd, J = 8.6 and 19.6 Hz, H600), 1.35 (3H, d, J = 7.24
Hz, 70-CH3), 1.26 (1H, dd, J = 4.76 and 12.72 Hz, H60). 13C NMR
(125.8 MHz, CDCl3): δ 164.6 (PhCdO), 152.4 (C2), 150.7 (C4),
149.2 (C6), 141.8 (C8), 137.9 (ArꢀC), 137.5 (ArꢀC), 133.7 (ArꢀC),
132.7 (ArꢀC), 128.8 (ArꢀC), 128.5 (ArꢀC), 128.3 (ArꢀC), 127.9
(ArꢀC), 127.88 (ArꢀC), 127.8 (ArꢀC), 127.7 (ArꢀC), 127.6
(ArꢀC), 127.4 (ArꢀC), 123.8 (C5), 88.8 (C40), 83.4 (C10), 80.3
(C30), 73.5 (CH2Ph), 72.1 (CH2Ph), 67.4 (C50), 48.2 (C20), 37.7 (C60),
28.7 (C70), 15.7(CH3). MALDI-TOF m/z [M þ H]þ calcd for
C34H34N5O4 576.261, found 576.259. 41b: 1H NMR (500 MHz,
CDCl3): δ 9.23 (1H, s, NH-Bz), 8.79 (1H, s, H2), 8.51 (1H, s, H8),
8.04 (2H, d, J = 7 Hz, ArꢀH), 7.58ꢀ7.62 (1H, m, ArꢀH), 7.50ꢀ7.55
(2H, m, ArꢀH), 7.13ꢀ7.40 (10H, m, ArꢀH), 5.89 (1H, s, H10), 4.69
(1H, d, J = 12.5 Hz, CH2Ph), 4.61 (1H, d, J = 12.5 Hz, CHPh), 4.47 (1H,
d, J = 11.5 Hz, CHPh), 4.40 (1H, d, J = 11.5 Hz, CHPh), 4.12 (1H, s,
H30), 3.78 (2H, q, J = 11.0 Hz, H50/500), 2.70 (1H, s, H20), 2.25ꢀ2.33
(1H, m, H70), 2.06 (1H, dd, J = 9 and 12 Hz, H600), 1.64 (1H, dd, J = 5.5
and 12.5 Hz, H60), 1.29 (3H, d, J = 7 Hz, 70-CH3). 13C NMR (125.8
MHz, CDCl3): δ 164.9 (CdO), 152.0 (C2), 150.8 (C6), 149.3 (C4),
142.8 (C8), 137.7 (ArꢀC), 137.6 (ArꢀC), 132.7 (ArꢀC), 129.4
(ArꢀC), 129.0 (ArꢀC), 128.6 (ArꢀC), 128.4 (ArꢀC), 128.4
(ArꢀC), 128.0 (ArꢀC), 127.9 (ArꢀC), 127.8 (ArꢀC), 127.7
(ArꢀC), 127.4 (ArꢀC), 123.7 (C5), 88.5 (C40), 88.1 (C10), 79.8
(C30), 73.5 (CH2Ph), 72.3 (CH2Ph), 67.0 (C50), 48.1 (C20), 36.6 (C60),
33.7 (C70), 20.2 (70-CH3). MALDI-TOF m/z [M þ H]þ calcd for
C34H34N5O4 576.261, found 576.264. 41c: 1H NMR (500 MHz,
CDCl3): δ 9.47 (1H, s, NH), 8.85 (1H, s, H8), 8.77 (1H, s, H2),
7.50ꢀ8.07 (5H, m, ArꢀH), 7.23ꢀ7.36 (10H, m, ArꢀH), 6.32 (1H, s,
9-[4-C-Allyl-3,5-di-O-benzyl-2-O-phenoxythiocarbonyl-β-
D-ribofuranosyl]-N6-benzoyladenine (40a). Compound 39a
(27.0 g, 45.76 mmol) was dissolved in 400 mL of dry DCM and was
cooled to 0 °C. To this precooled solution was added 1-methyl
imidazole (14.4 mL, 183 mmol), and the solution was stirred for 20
min at the same temperature. This was followed by addition of phenyl
chlorothionoformate (19 mL, 137 mmol) dropwise, and the reaction
was stirred for 5 h at room temperature. The reaction was quenched with
saturated NaHCO3 solution and extracted with DCM (3 ꢂ 350 mL).
The organic layer was washed with saturated citric acid solution several
times and dried over Na2SO4 and concentrated, and the crude com-
pound was chromatographed over silica gel (1% methanol in DCM, v/v)
to give 40a as off-white foam solid (30 g, 90%). Rf 0.60 (3% methanol in
1
DCM, v/v). H NMR (500 MHz, CDCl3): δ 9.0 (1H, bs, NH), 8.76
(1H, s, H2), 8.33 (1H, s, H8), 8.01 (2H, d, J = 7.75 Hz, ArꢀH),
7.57ꢀ7.62 (1H, m, ArꢀH), 7.49ꢀ7.55 (2H, m, ArꢀH), 7.23ꢀ7.40
(13H, m, ArꢀH), 6.95 (2H, d, J = 8 Hz, ArꢀH), 6.58 (1H, d, J = 5.5 Hz,
H10), 6.46 (1H, t, J = 5.5 Hz, H20), 5.80ꢀ5.88 (1H, m, H70), 5.08ꢀ5.16
(2H, m, H80 and H800), 4.75ꢀ482 (2H, m, H30and CH2Ph), 4.63 (1H, d,
J = 11. Hz, CH2Ph), 4.59 (1H, d, J = 11 Hz, CH2Ph), 4.51 (1H, d, J = 9
Hz, CH2Ph), 3.68 (1H, d, J = 10 Hz, H500), 3.50 (1H, d, J = 10 Hz, H50),
2.74 (1H, dd, J = 6.5 and 14.5 Hz, H600), 2.46 (1H, dd, J = 8.0 and 14.5
Hz, H60). 13C NMR (125.8 MHz, CDCl3): δ 194.1 (ꢀCSOPh), 164.6
(PhCdO), 153.3 (ArꢀC), 152.7 (C2), 151.9 (C4), 149.4 (C6), 141.8
(C8), 137.34 (ArꢀC), 137.2 (ArꢀC), 133.6 (ArꢀC), 132.7 (ArꢀC),
4424
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
H10), 4.38ꢀ4.68 (4H, m, 2 ꢂ CH2Bn), 4.39 (1H, d, J = 5.0 Hz, H30),
3.64 (1H, d, J = 10.5 Hz, H500), 3.52 (1H, d, J = 10.5 Hz, H50), 2.76 (1H,
s, H20), 2.02 (1H, m, H800), 1.72ꢀ1.90 (4H, m, H600, H70, H700, H80),
1.40 (1H, dd, H60). 13C NMR (125.8 MHz, CDCl3): δ 164.9 (CdO),
152.2 (C2), 150.8 (C4), 149.3 (C6), 142.0 (C8), 137.7 (ArꢀC), 137.6
(ArꢀC), 133.1 (ArꢀC), 132.7 (ArꢀC), 128.8 (ArꢀC), 128.6 (ArꢀC),
128.4 (ArꢀC), 128.1 (ArꢀC), 127.9 (ArꢀC), 127.8 (ArꢀC), 127.7
(ArꢀC), 127.4 (ArꢀC), 123.6 (C5), 86.9 (C10), 85.1 (C40), 73.7 (C30),
73.3 (CH2Ph), 71.9 (CH2Ph), 70.2 (C50), 42.9 (C20), 26.7 (C60), 21.0
(C80), 17.8 (C70). MALDI-TOF m/z [M þ H]þ calcd for C34H34N5O4
576.261, found 576.263.
(1R,3R,4R,5R,7S)-3-(Adenin-9-yl)-7-hydroxyl-1-(hydroxy-
methyl)-5-methyl-2-oxa-bicyclo[2.2.1]heptane(44a).Compound
41a (9 g, 15.65 mmol) was dissolved in methanolic ammonia (400 mL) and
stirred for 16 h at room temperature. The reaction mixture was concentrated
under vacuum to give 41a0 as a crude gummy solid. Rf = 0.19 (100% EtOAc).
To a solution of 41a0 (8 g, 16.9 mmol) in methanol (500 mL) were added
20% Pd(OH)2/C (8 g) and ammonium formate (64.2 g, 1019 mmol), and
the solution was heated at 60 °C for 16 h in a water bath. The reaction was
allowed to reach to rt, and the suspension was filtered over Celite. Pd(OH)2/C
was washed with (5 ꢂ 400 mL) methanol. The organic solvent was
evaporated, and then the dried residue was washed with EtOAc (150 mL)
to give 44a as white powder (3.2 g, 64%). Rf 0.11 (10% methanol in DCM,
v/v). 1H NMR (400 MHz, DMSO-d6):δ8.24(1H, s, H8), 8.10(1H, s, H2),
7.23 (2H, s, NH2, D2O exchangeble), 6.00 (1H, s, H10), 5.27 (1H, d, J = 3.6
Hz, 30-OH, D2O exchangeble), 4.93 (1H, t, J = 5.7 Hz, 50-OH, D2O
exchangeble), 4.16 (1H, s, H30), 3.50ꢀ3.70 (2H, m, H50 and H500),
2.50ꢀ2.73 (1H, m, H70), 2.38 (1H, d, J = 3.12 Hz, H20), 1.90 (1H, t, J =
12.2 Hz, H60), 1.20 (3H, d, J = 7 Hz, CH3), 0.98 (1H, dd, J = 5 and 12.5 Hz,
H60). 13C NMR (125.8 MHz, CDCl3): δ 155.8 (C6), 152.4 (C2), 148.2
(C4), 138.2 (C8), 119.2 (C5), 89.8 (C40), 81.8 (C10), 71.9 (C30), 58.8
(C50), 50.1 (C20), 36.7 (C60), 28.1 (C70), 15.5 (CH3). MALDI-TOF m/z
[M þ H]þ calcd for C13H18N5O3 292.141, found 292.139.
dried and evaporated, and the crude compound was chromatographed
over silica gel (1% pyridine/methanol in DCM, v/v) to give 48a (0.55 g,
80%) as yellow foam. Rf 0.40 (10% methanol in DCM, v/v). 1H NMR
(400 MHz, CDCl3): δ 9.02 (1H, s, NH-Bz), 8.79 (1H, s, H2), 8.41 (1H,
s, H8), 8.02 (2H, d, J = 8.72 Hz, ArꢀH), 7.60ꢀ7.70 (4H, m, ArꢀH), 7.44
(2H, d, J = 7.6 Hz, ArꢀH), 7.30ꢀ7.35 (6H, m, ArꢀH), 6.84 (4H, d, J =
8.72 Hz, ArꢀH), 6.23 (1H, s, H10), 4.38 (1H, s, H30), 3.78 (6H, s, Ar-
OMe), 3.56 (1H, d, J = 10.7 Hz, H500), 3.36 (1H, d, J = 10.8 Hz, H50),
2.65ꢀ2.90 (2H, m, H20 and H70), 2.12 (1H, d, J = 10.8 Hz, H600), 1.36
(3H, d, J = 6.9 Hz, CH3), 1.26 (1H, dd, J = 5.0 and 12.9 Hz, H60).
13C NMR (100.6 MHz, CDCl3): δ 164.7 (COPh), 158.5 (ArꢀC),
152.5 (C2), 150.6 (C4), 149.0 (C6), 144.4 (ArꢀC), 141.2 (C8), 135.6
(ArꢀC), 133.7 (ArꢀC), 132.6 (ArꢀC), 129.0 (ArꢀC), 128.9 (ArꢀC),
128.8 (ArꢀC), 128.2 (ArꢀC), 128.0 (ArꢀC), 127.9 (ArꢀC), 127.8
(ArꢀC), 127.0 (ArꢀC), 123.4 (C5), 113.2 (ArꢀC), 88.8 (C40), 86.5
(DMTr-C), 83.2 (C10), 74.5 (C30), 61.5 (C50), 55.2 (DMTr-OMe),
50.4 (C20), 37.3 (C60), 28.6 (C70), 15.6 (CH3). MALDI-TOF m/z [M þ
H]þ calcd for C41H40N5O6 698.298, found 698.301.
(1R,3R,4R,5R,7S)-3-(N6-Benzoyladenin-9-yl)-7-(2-cyanoethoxy-
diisopropylamino-phosphinoxy)-1-(4,40-dimethoxytrityloxy)-
methyl-5-methyl-2-oxa-bicyclo[2.2.1]heptane (49a). Compound
48a (0.950 g, 1.36 mmol) was dissolved in 45 mL of dry DCM, and DIPEA
(1.18 mL, 6.80 mmol) was added at 0 °C followed by 2-cyanoethyl N,N-
diisopropylphosphoramidochloridite (0.605 mL, 2.72 mmol). The reaction
was stirred for 24 h at room temperature. Reaction was quenched by the
addition of 0.1 mL of methanol and aqueous NaHCO3 under cooling
conditions and extracted twice (25 mL) with DCM. The organic phase was
dried and evaporated. The obtained crude compound was purified by
precipitation in n-hexane, to give 49a (1.0 g, 81%) as a mixture of two
isomers. 31P NMR (161.97 MHz, CDCl3): 149.38, 149.19. LC-MS m/z
[M þ H]þ calcd for C50H57N7O7P 898.40, found 898.30.
9-[4-C-Allyl-3,5-di-O-benzyl-2-O-acetyl-β-D-ribofuranosyl]-
N2-Acetyl-O6-diphenylcarbamoyl-guanine (38b). N,O-Bis-
(trimethylsilyl)acetamide (BSA) (32.0 mL, 131.6 mmol) was added to a
stirred suspension of O6-diphenylcarbamoyl-N2-acetylguanine (13.6 g,
36.2 mmol) in dried 1,2-dichloroethane (164 mL). Stirring was con-
tinued under N2 at 80 °C for 30 min. The clear solution was evaporated
under reduced pressure. The residue was redissolved in dry toluene
(100 mL). Trimethylsilyl triflate (TMSOTf) (7.74 mL, 42.8 mmol) and a
solution of 37 (15.0 g, 32.9 mmol) in dried toluene (65 mL) were added.
The solution was stirred at 80 °C for 1 h and was allowed to cool to rt.
The reaction was quenched with saturated NH4Cl solution, and volatiles
were evaporated. The residual was redissolved in EtOAc, and the solution
was washed with saturated aqueous NaHCO3 solution. The organic layer
was dried over anhydrous Na2SO4 and filtered. The filtrate was con-
centrated to dryness. The crude product obtained was purified by silica
gel column chromatography (15ꢀ20% ethyl acetate in n-hexane, v/v) to
give 38b(15.4 g, 60%). 1H NMR(400MHz, CDCl3): δ8.24(1H, s, H8),
7.89 (1H, s, NH), 7.40ꢀ7.25 (20H, m, ArꢀH), 6.21 (1H, d, H10, J = 5.0
Hz), 5.77ꢀ5.83 (2H, m, H70, H20), 5.07 (2H, d, J = 12 .4 Hz, H80, H800,
4.40ꢀ4.59 (5H, m, 2 ꢂ CH2Ph, H30), 3.60 (1H, d, J = 10.3 Hz, H50),
3.39(1H, d, J = 10.3 Hz, H500), 2.65ꢀ2.71 (1H, dd, J = 6.3, 14.7 Hz, H60),
2.53 (3H, s, 2-COCH3), 2.32ꢀ2.39 (1H, m, H600), 2.05 (3H, s,
20-COCH3). 13C NMR (100.6 MHz, CDCl3): δ 170.8 (NHCdO),
169.7 (20-CdO), 156.0, 154.5 (C4), 152.1, 150.3 (CONPh2), 142.5
(C8), 141.7, 137.4, 137.2, 132.6 (C70), 129.1, 128.5, 127.9, 127.8, 127.5,
126.4, 125.9, 120.9 (C5), 118.6(C80), 87.8 (C40), 85.9 (C10), 78.3 (C30),
75.5 (C20), 74.3 (OCH2Ph), 73.4 (OCH2Ph), 72.3 (C50), 37.1 (C60),
25.1 (NHCOCH3), 20.6 (OCOCH3). MALDI-TOF m/z [M þ H]þ
calcd for C44H42N6O8 783.315, found 783.312.
(1R,3R,4R,5R,7S)-3-(N6-Benzoyladenin-9-yl)-7-hydroxyl-1-
hydroxymethyl-5-methyl-2-oxa-bicyclo[2.2.1]heptane (45a).
Compound 44a (0.5 g, 1.7 mmol) was coevaporated twice with dry
pyridine (15 mL) and suspended in the same (15 mL) and cooled
solution of the above reaction mixture.TMSCl (1.1 mL, 8.6 mmol) was
added over 45 min and stirred at 0 °C for 30 min. Then benzoyl chloride
(0.3 mL, 2.6 mmol) was added over 15 min, and the solution was stirred
at room temperature for 2 h. Then ammonia solution (1.8 mL, in 9 mL of
water) was added slowly at 0 °C and stirred for 30 min. The reaction
mixture was allowed to warm to room temperature and stirred for
another 16 h. After evaporation of solvent, the residual was chromato-
graphed over silica gel (2.5ꢀ4.5% MeOH in DCM, v/v) to give 45a as
an off-white solid (0.35 g, 52%). Rf 0.4 (10% MeOH in DCM, v/v). 1H
NMR (600 MHz, CDCl3): δ 8.58 (1H, s, H2), 8.51 (1H, s, H8), 7.96
(2H, d, J = 7.2 Hz, ArꢀH), 7.53 (1H, t, J = 7.8 Hz, ArꢀH), 7.40ꢀ7.48
(2H, m, ArꢀH), 6.10 (1H, s, H10), 4.51 (1H, s, H30), 3.74 (1H, d, J = 12.6
Hz, H500), 3.69 (1H, d, J = 12.6 Hz, H50), 2.68ꢀ2.80 (1H, m, H70), 2.53
(1H, d, J = 3 Hz, H20), 1.95 (1H, t, J = 12 Hz, H600), 1.24 (3H, d, J = 7.2
Hz, CH3), 0.98 (1H, dd, J = 4.8 and 12.6 Hz, H60). 13C NMR (150.9
MHz, CDCl3): δ 166.2 (PhCdO), 152.1 (C2), 150.8 (C4), 148.8 (C6),
141.8 (C8), 133.3 (ArꢀC), 132.8 (ArꢀC), 128.7 (ArꢀC), 128 (ArꢀC),
122.9 (C5), 90.8 (C40), 83.0 (C10), 72.5 (C30), 59.5 (C50), 50.4þ(C20),
36.6 (C60), 28.8 (C70), 15.6 (CH3). MALDI-TOF m/z [M þ H] calcd
for C20H22N5O4 396.167, found 396.163.
(1R,3R,4R,5R,7S)-3-(N6-Benzoyladenin-9-yl)-1-(4,40-dimeth-
oxytrityloxy)methyl-5-methyl-7-hydroxyl-2-oxa-bicyclo[2.2.1]-
heptane (48a). Compound 45a (0.5 g, 1.26 mmol) was coevaporated
twice with dry pyridine (100 mL) and suspended in the same (100 mL).
4,40-Dimethoxytrityl chloride (1.27 g, 3.78 mmol) was added and stirred
for 16 h at rt. The reaction mixture was quenched with saturated aqueous
NaHCO3 and extracted three times with DCM. The organic layer was
9-[4-C-Allyl-3,5-di-O-benzyl-β-D-ribofuranosyl]-N2-acetyl-
O6-diphenylcarbamoyl-guanine (39b). A solution of compound
38b (14 g, 17.9 mmol) in THF:water (2:1) (450 mL) was cooled to 0 °C
and 2 N aqueous NaOH (34.4 mL, 4 equivalent) solution was added.
4425
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
The reaction was stirred at 0 °C for 2 h. The reaction mixture was
partitioned between diethyl ether and the aqueous layer. The organic
layer was dried over anhydrous Na2SO4 and filtered. The filtrate was
evaporated to dryness. The crude product was washed several times in
5% acetone/n-hexane mixture to yield white solid 39b (11.6 g, 88%). 1H
NMR (400 MHz, CDCl3): δ 8.15 (1H, s, H8), 8.03 (1H, s, NH),
7.17ꢀ7.42 (18H, m, ArꢀH), 7.06 (2H, m, ArꢀH), 5.95 (1H, d, J = 5.6
Hz, H10), 5.78ꢀ5.81 (1H, m, H70), 5.06ꢀ5.11 (3H, m, CH2Ph and
H80), 4.92 (1H, bs, H20), 4.67 (1H, d, J = 11.4 Hz, CH2Ph), 4.33 (2H, s,
CH2Ph), 4.26 (1H, d, J = 5.6 Hz, H30), 3.41 (2H, s, H50and H500),
2.68ꢀ2.70 (1H, m, H60), 2.51ꢀ2.53 (1H, m, H600), 2.32 (3H, bs,
NHCOCH3). 13C NMR (100.6 MHz, CDCl3): δ 169.4 (CdO), 155.7,
153.8 (C4), 151.1, 150.3 (CONPh2), 142.7 (C8), 141.6, 138.0, 137.1,
133.0 (C70), 129.1, 128.3, 128.2, 127.9, 127.6, 127.5, 121.1 (C5), 118.4
(C80), 90.6 (C10), 89.0 (C40), 80.3 (C30), 76.8 (C20), 74.4 (OCH2Ph),
73.4 (OCH2Ph), 73.1 (C50), 37.2 (C60), 24.7 (NHCOCH3). MALDI-
TOF m/z [M þ H]þ calcd for C42H40N6O7 741.304, found 741.308.
9-[4-C-Allyl-3,5-di-O-benzyl-2-O-phenoxythiocarbonyl-β-
D-ribofuranosyl]-N2-acetyl-O6-diphenylcarbamoylguanine
(40b). To a mixture of compound 39b (10 g, 13.5 mmol) and 1-N-
methylimidazole (2.25 mL, 28.4 mmol) in dry DCM (150 mL) was
added phenyl chlorothionoformate (3.15 mL, 22.9 mmol), and the
solution was stirred overnight at rt. The reaction was poured in
saturated NaHCO3 solution and extracted with DCM (3 ꢂ 250 mL).
Combined organic phases were washed with saturated citric acid
solution (3 ꢂ 50 mL), dried with Na2SO4, and filtered, and the filtrate
was evaporated to dryness. The residual was purified over silica gel
column chromatography (15ꢀ20% ethyl acetate in n-hexane, v/v) to
(1H, d, J = 11.0, H500), 2.59ꢀ2.67 (1H, m, H70), 2.58 (1H, m, H20), 2.55
(3H, s, NHCOCH3), 2.03ꢀ2.11 (1H, m, H600), 1.28 (3H, d, J = 7.2 Hz,
70-CH3), 1.22 (1H, m, H60). 13C NMR (100.6 MHz, CDCl3): δ 171.0
(CdO), 155.9, 153.2 (C4), 151.8, 150.4 (CONPh2), 142.4 (C8), 141.7,
137.8, 137.4, 129.1, 128.5, 128.2, 127.7, 127.5, 127.4, 126.0, 121.5 (C5),
88.7 (C40), 83.4 (C10), 80.1 (C30), 73.4 (OCH2Ph), 72.2 (OCH2Ph),
67.1 (C50), 48.3 (C20), 37.5 (C60), 28.5 (C70), 25.1 (NHCOCH3), 15.5
(70-CH3). MALDI-TOF m/z [M þ H]þ calcd for C42H40N6O6
1
725.309, found 725.310. 42b: H NMR (400 MHz, CDCl3): δ 8.41
(1H, s, H8), 8.17 (1H, s, NH), 7.14ꢀ7.60 (20H, m, 2 ꢂ Bn, DPC), 5.73
(1H, s, H10), 4.64 (2H, q, J = 12.5 Hz, BnCH2), 4.41 (2H, q, J = 12.0 Hz,
BnCH2), 4.05 (1H, s, H30), 3.75 (2H, q, J = 11.0 Hz, H50 and 500), 2.61
(1H, bs, H20), 2.55 (3H, s, Ac), 2.22 (1H, m, H70), 2.02 (1H, dd, J = 8.5
Hz, H600), 1.62 (1H, dd, J = 5.0 and 12.5 Hz, H60), 1.27 (3H, d, J = 7.5
Hz, CH3). 13C NMR (100.6 MHz, CDCl3): δ 171.1 (CdO), 155.8
(C6), 153.5 (C4), 151.8 (C2), 150.5 (OC(O)), 142.4 (C8), 141.8,
137.8, 137.3, 129.2, 128.5, 128.3, 127.9, 127.81, 127.8, 127.6, 127.5,
121.6 (C5), 88.6 (C40), 88.2 (C10), 79.8 (C30), 73.4 (BnCH2), 72.3
(BnCH2), 67.0 (C50), 48.1 (C20), 38.8 (C60), 33.7 (C70), 25.1
(CH3C(O)), 15.6 (CH3). MALDI-TOF m/z [M þ H]þ calcd for
C42H40N6O6 725.309, found 725.313. 42c: 1H NMR (500 MHz,
CDCl3): δ 8.71 (1H, s, H8), 8.26 (1H, s, NH), 7.22ꢀ7.60 (20H, m, 2
ꢂ Bn, DPC), 6.16 (1H, s, H10), 4.36ꢀ4.69 (4H, m, 2 ꢂ BnCH2), 4.32
(1H, d, J = 5.0 Hz, H30), 3.61 (1H, d, J = 11.0 Hz, H500), 3.49 (1H, d,
J = 11.0 Hz, H50), 2.69 (1H, bs, H20), 2.55 (3H, s, Ac), 2.01 (1H, m,
H800), 1.72ꢀ1.90 (4H, m, H600, H70, H700 and H80), 1.39 (1H, dd, H60).
13C NMR (125.8 MHz, CDCl3): δ 171.1 (CdO), 155.8 (C6), 153.7
(C4), 151.8 (C2), 150.5 (OC(O)), 143.0 (C8), 141.7, 137.7, 137.6,
129.2, 128.6, 128.4, 127.9, 127.81, 127.8, 127.4, 121.5 (C5), 87.1 (C10),
85.1 (C40), 73.8 (C30), 73.3 (BnCH2), 72.0 (BnCH2), 70.1 (C50), 42.8
(C20), 26.7 (C60), 25.1 (CH3C(O)), 21.1 (C80), 17.8 (C70). MALDI-
TOF m/z [M þ H]þ calcd for C42H40N6O6 725.309, found 725.310.
(1R,3R,4R,5R,7S)-3-(N2-Acetylguanin-9-yl)-1-benzyloxy-
methyl-7-benzyloxy-5-methyl-2-oxa-bicylco[2.2.1]heptane
(42a0). A solution of 42a (2.75 g, 3.78 mmol) in acetic acid (75 mL) was
warmed at 55 °C with constant stirring for 12 h followed by stirring at rt
overnight. The volatiles were evaporated under reduced pressure to give
brownish oil. The residue was diluted with EtOAc (50 mL). To this
mixture was added water (50 mL), and subsequently sonication led to
white precipitate formation, which was filtered. The precipitate was
washed with EtOAc/hexane (1:1) (150 mL), and residual solid was
dissolved in DCM (50 mL). The organic layer was dried over Na2SO4
and filtered. The filtrate was concentrated to dryness to obtain white
solid 42a0 (1.25 g, 62%). 1H NMR (400 MHz, CDCl3): δ 11.92 (1H, bs,
NH), 8.81 (1H, bs, NH), 8.00 (1H, s, H8), 7.17ꢀ7.32 (10H, m, ArꢀH),
5.90 (1H, s, H10), 4.63 (1H, d, J = 12.2 Hz, OCH2Ph), 4.55 (1H, d, J =
12.2 Hz, OCH2Ph), 4.42 (2H, s, OCH2Ph), 4.19 (1H, s, H30), 3.70 (1H,
d, J = 11.3 Hz, H50), 3.63 (1H, d, H500), 2.67 (1H, m, H70), 2.53 (1H, d, J
= 5.5 Hz, H20), 2.27 (3H, s, NHCOCH3), 2.03ꢀ2.11 (1H, m, H600),
1.23 (4H, d, J = 8.2 Hz, 70-CH3 and H60). 13C NMR (100.6 MHz,
CDCl3): δ 173.0 (CdO), 156.0 (C6), 147.6 (C4), 147.3 (C2), 137.7,
137.43 (C8), 137.35, 128.3, 127.7, 121.1 (C5), 88.5 (C40), 82.9 (C10),
79.8 (C30), 73.3 (OCH2Ph), 71.8 (OCH2Ph), 67.5 (C50), 48.3 (C20),
37.4 (C60), 28.5 (C70), 24.3 (NHCOCH3), 15.3 (70-CH3). MALDI-
TOF m/z [M þ H]þ calcd for C29H31N5O5 530.241, found 530.244.
(1R,3R,4R,5R,7S)-3-(N2-Acetylguanin-9-yl)-7-hydroxyl-1-
hydroxymethyl-5-methyl-2-oxa-bicylco[2.2.1]heptane (45b).
Compound 42a0 (0.775 g, 1.46 mmol) was dissolved in methanol
(75 mL) to which N2 gas was purged for half an hour. To the reaction
mixture was added formic acid (0.575 mL, 14.6 mmol) and 20%
Pd(OH)2/C (0.630 g). The reaction mixture was heated to reflux for
2.5 h. The reaction mixture was filtered over a bed of Celite and washed
with hot methanol several times. The crude product obtained was
purified on a preparative thin layer chromatography plate to obtain white
1
yield 40b as a white foam solid (9.79 g, 82%). H NMR (400 MHz,
CDCl3): δ 8.24 (1H, s, H8), 7.88 (1H, s, NH), 7.25ꢀ7.41 (23H, m,
ArꢀH), 6.92ꢀ6.94 (2H, d, J = 7.8 Hz, ArꢀH), 6.40 (1H, d, J = 5.8 Hz,
H10), 6.28 (1H, t, J = 5.6 Hz, H20), 5.78ꢀ5.84 (1H, m, H70), 5.08 (2H, d,
J = 12.6 Hz, H80, H800), 4.71ꢀ4.76 (2H, m, OCH2Ph andH30), 4.60(2H,
d, J = 12.8 Hz, OCH2Ph), 4.47 (1H, d, J = 11.7 Hz, OCH2Ph), 3.64 (1H,
d, J = 10.2 Hz, H50), 3.47 (1H, d, J = 10.2 Hz, H500), 2.67ꢀ2.73 (1H, dd,
J = 6.1 and 14.5 Hz, H60), 2.56 (3H, bs, NHCOCH3), 2.40ꢀ2.44 (1H, m,
H600). 13C NMR (100.6 MHz, CDCl3): δ 194.0 (CdS), 170.9 (CdO),
156.1, 154.7 (C4), 153.2, 152.2, 150.3 (CONPh2), 142.5 (C8), 141.7,
137.3, 137.1, 132.4 (C70), 129.6, 129.1, 128.5, 128.1, 128.0, 127.9, 126.8,
125.7, 121.5 (C5), 118.8 (C80), 88.1 (C40), 85.0 (C10), 83.5 (C20), 77.9
(C30), 74.8 (OCH2Ph), 73.6 (OCH2Ph), 72.9 (C50), 37.2 (C60), 25.3
(NHCOCH3). MALDI-TOF m/z [M þ H]þ calcd for C49H44N6O8S
877.303, found 877.304.
(1R,3R,4R,5R,7S)-3-(N2-Acetyl-O6-diphenylcarbamoyl-
guanin-9-yl)-1-benzyloxymethyl-7-benzyloxy-5-methyl-2-oxa-
bicylco[2.2.1]heptane (42a), (1R,3R,4R,5S,7S)-3-(N2-Acetyl-
O6-diphenylcarbamoylguanin-9-yl)-1-benzyloxymethyl-7-
benzyloxy-5-methyl-2-oxa-bicylco[2.2.1]heptane (42b), and
(1R,5R,7R,8S)-7-(N2-Acetyl-O6-diphenylcarbamoylguanin-
9-yl)-5-benzyloxymethyl-8-benzyloxy-6-oxa-bicyclo[3.2.1]-
octane (42c). Compound 40b (4.5 g, 5.13 mmol) was dissolved in
anhydrous toluene (200 mL) to which N2 gas was purged for half an
hour. The reaction mixture was heated at 90 °C, and Bn3SnH (1.38 mL
in 10 mL of anhydrous toluene) and AIBN (0.84 g in 10 mL of
anhydrous toluene) were added dropwise over 1.5 h and heated at
90 °C for another 30 min. Solvent was evaporated, and residue was
applied to silica short column chromatography (15ꢀ25% EtOAc in
n-hexane) to obtain compound 42a as white foam (1.8 g, 48%). The
remaining mixture was subjected to preparative HPLC separation to
yield compound 42b as a white foam solid (0.16 g, 4.3%) and white
solid 42c (0.150 g, 4%). Compound 42a: 1H NMR (400 MHz, CDCl3):
δ 8.39 (1H, s, H8), 7.97 (1H, s, NH), 7.14ꢀ7.43 (20H, m, ArꢀH),
6.11 (1H, s, H10), 4.62 (2H, q, J = 12.4 Hz, OCH2Ph), 4.40 (2H,
s, OCH2Ph), 4.14 (1H, s, H30), 3.72 (1H, d, J = 11.0 Hz, H50), 3.64
4426
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
yellow solid 45b (0.3 g, 60%). 1H NMR (400 MHz, DMSO-d6): δ 12.00
(1H, bs, NH), 11.56 (1H, bs, NH), 8.11 (1H, s, H8), 5.89 (1H, s, H10),
5.35 (1H, d, J = 3.3 Hz, 30-OH), 4.88 (1H, t, J = 5.5 Hz, 50-OH), 4.18 (1H,
s, H30), 3.59 (2H, m, H50, H500), 2.67 (1H, m, H70), 2.43 (1H, d, J = 3.2
Hz, H20), 2.17 (3H, s, NHCOCH3), 1.86ꢀ1.96 (1H, m, H60), 1.22 (3H,
d, J = 7.3 Hz, 70-CH3), 0.97ꢀ1.01 (1H, m, H600). 13C NMR (100.6 MHz,
CDCl3): δ 173.1 (CdO), 154.4 (C6), 147.2 (C4), 147.0 (C2), 136.6
(C8), 120.2 (C5), 89.6 (C40), 81.0 (C10), 71.7 (C30), 58.3 (C50), 49.4
(C20), 36.3 (C60), 27.8 (C70), 23.4 (NHCOCH3), 15.3 (70-CH3).
MALDI-TOF m/z [M þ H]þ calcd for C15H19N5O5 350.147, found
350.151.
13.22 (1H, s, NH, D2O exchangeable), 8.30 (2H, d, J = 7.2 Hz, Ar- H),
7.69 (1H, s, H6), 7.48ꢀ7.53 (1H, m, ArꢀH), 7.28ꢀ7.42 (12H, m,
ArꢀH), 6.3 (1H, d, J = 4.0 Hz, H10), 5.74ꢀ5.92 (1H, m, H70), 5.45 (1H,
t, J = 5.4 Hz, H20), 5.07ꢀ5.12 (2H, m, H80), 4.65 (1H, d, J = 11.4 Hz,
CH2ꢀPh), 4.42ꢀ4.61 (3H, m, CH2ꢀPh), 4.38 (1H, d, J = 6.0 Hz, H30),
3.72 (1H, d, J = 10.2 Hz, H500), 3.41 (1H, d, J = 10.2 Hz, H50), 2.68 (1H,
dd, J = 6.0 and 14.4 Hz, H600), 2.35 (1H, dd, J = 8.4 Hz, H60), 2.10 (3H, s,
COCH3), 1.60 (3H, s, Cy-CH3). 13C NMR (150.9 MHz, CDCl3): δ
179.6 (COBz), 170.0 (COCH3), 159.6 (C4), 148.1 (C2), 137.5
(ArꢀC), 137.3 (ArꢀC), 137.2 (ArꢀC), 136.9 (C6), 132.6 (C70),
132.4 (ArꢀC), 127.6ꢀ129.9 (ArꢀC), 118.7 (C80), 112.2 (C5), 87.3
(C40), 86.5 (C10), 77.6 (C30), 75.4 (C20), 74.4 (OBn), 73.7 (OBn), 73.0
(C50), 37.3 (C60), 20.7 (COCH3), 13.1 (Cy-CH3). MALDI-TOF m/z of
[M þ H]þ calcd for C36H38N3O7 624.271, found 624.270.
(1R,3R,4R,5R,7S)-3-(N2-Acetylguanin-9-yl)-1-(4,40-dimeth-
oxytrityloxymethyl)-7-hydroxyl-5-methyl-2-oxa-bicylco[2.2.1]-
heptane (48b). Compound 45b (0.4 g, 1.14 mmol) was coevaporated
twice with dry pyridine (25 mL) and suspended in the same (25 mL).
4,40-Dimethoxytrityl chloride (1.16 g, 3.42 mmol) was added and stirred
overnight at rt. The reaction was quenched with saturated aqueous
NaHCO3 and extracted with DCM (3 ꢂ 50 mL). The combined organic
phase was dried and evaporated, and the residue was chromatographed
over silica gel (1% pyridine/DCM/methanol, v/v) to give 48b (0.6 g,
0.92 mmol, 80%) as yellow foam. Rf 0.3 (10% methanol in DCM, v/v).
1H NMR (400 MHz, CDCl3): δ 11.95 (1H, bs, NH), 9.3 (1H, s, NH),
8.00 (1H, s, H8), 7.12ꢀ7.42 (10H, m, ArꢀH), 6.78 (3H, d, J = 8.8 Hz,
ArꢀH), 5.85 (1H, s, H10), 4.39 (1H, s, H30), 3.73 (6H, s, Trityl-OMe),
3.42 (1H, d, J = 10.6 Hz, H50), 3.35 (1H, d, H500), 2.66ꢀ2.67 (1H, m,
H70), 2.48 (1H, m, H20), 2.21 (3H, s, NHCOCH3), 2.08ꢀ2.11 (1H, m,
H60), 1.24 (1H, m, H600), 1.28 (3H, d, J = 7.2 Hz, 70-CH3). 13C NMR
(100.6 MHz, CDCl3): δ 172.4 (CdO), 158.4, 155.8 (C6), 147.3 (C4),
147.1 (C2), 144.7, 137.8, 137.5 (C8), 135.8, 135.7, 130.0, 128.9, 128.2,
128.1, 127.8, 126.8, 125.3, 121.1 (C5), 113.1, 88.8 (C40), 86.2 (DMTr-
C), 82.6 (C10), 74.3 (C30), 61.8 (C50), 55.1 (OMe), 50.7 (C20), 37.þ3
(C60), 28.3 (C70), 24.3 (NHCOCH3), 15.5 (70-CH3). m/z [M þ H]
calcd for C36H37N5O7 652.27, found 652.20.
1-[4-C-Allyl-3,5-di-O-benzyl-β-D-ribofuranosyl]-N4-benzoyl-
5-methylcytosine (39c). Compound 38c (13 g, 20.8 mmol) was
dissolved in 50 mL of THF and cooled to 0 °C. To this cold solution (2.7 g,
67.5 mmol) NaOH in 15 mL of water was added dropwise over 15 min at
the same temperature, and the mixture was stirred for 2 h at rt. The
reaction mixture was quenched with water and extracted with EtOAc
(200 mL). The organic layer was dried over Na2SO4, concentrated under
reduced pressure, and chromatographed over silica gel eluting with 15%
ethyl acetate in hexane (v/v) to give 39c as a white foam solid (11.5 g,
1
95%). Rf 0.3 (30% ethyl acetate in hexane, v/v). H NMR (600 MHz,
CDCl3): δ 13.25 (1H, s, NH, D2O exchangeable), 8.27 (2H, m, ArꢀH),
7.62 (1H, s, H6), 7.25ꢀ7.58 (13H, m, ArꢀH), 6.01 (1H, d, J = 5.4 Hz,
H10), 5.81ꢀ5.89 (1H, m, H70), 5.09ꢀ5.12 (2H, m, H80), 4.70 (2H, q, J =
11.4 Hz, CH2ꢀPh), 4.53 (2H, q, J = 11.4 Hz, CH2ꢀPh), 4.39 (1H, dd,
J = 6.0 and 12.5 Hz, H20), 4.20 (1H, d, J = 6.0 Hz, H30), 3.60 (1H, d, J =
10.2 Hz, H500), 3.45 (1H, d, J = 10.2 Hz, H50), 2.99 (1H, d, J = 7.2 Hz,
OH, D2O exchangeable), 2.75 (1H, dd, J = 6.0 and 14.4 Hz, H600), 2.34
(1H, dd, J = 8.4 and 15.0 Hz, H60), 1.73 (3H, s, CH3). 13C NMR (150.9
MHz, CDCl3): δ 179.5 (COBz), 159.6 (C4), 148.5 (C2), 137.4 (ArꢀC),
137.3 (ArꢀC), 137.1 (ArꢀC), 137.0 (C6), 132.7 (C70), 132.4 (ArꢀC),
129.9 (ArꢀC), 128.7 (ArꢀC), 128.7 (ArꢀC), 128.6 (ArꢀC), 128.4
(ArꢀC), 128.2 (ArꢀC), 128.1 (ArꢀC), 128.0 (ArꢀC), 127.6 (ArꢀC),
118.7 (C80), 111.9 (C5), 89.4 (C10), 87.2 (C40), 79.6 (C30), 75.4 (C20),
74.8 (OBn), 73.8 (OBn), 73.7 (C50), 37.5 (C60), 13.2 (Cy-CH3).
MALDI-TOF m/z [M þ H]þ calcd for C34H36N3O6 582.261, found
582.263.
9-[4-C-Allyl-3,5-di-O-benzyl-2-O-phenoxythiocarbonyl-β-
D-ribofuranosyl]-N4-benzoyl-5-methyl-cytosine (40c). Com-
pound 39c (10.5 g, 18.07 mmol) was dissolved in 230 mL of dry DCM
and cooled to 0 °C, and to this precooled solution was added
1-methylimidazole (2.85 mL, 36.0 mmol) and stirred for 20 min at
the same temperature. To this mixture was added dropwise phenyl
chlorothionoformate (3.49 mL, 25.3 mmol), and the reaction was stirred
for 1.5 h at rt. The reaction was quenched with saturated solution of
NaHCO3 and extracted with DCM. The organic layer was dried over
Na2SO4, concentrated, and chromatographed over silica gel eluting with
10% ethyl acetate in hexane (v/v) to give 40c as a white foam (11.9 g,
92%). Rf 0.5 (20% ethyl acetate in hexane, v/v). 1H NMR (600 MHz,
CDCl3): δ 13.1 (1H, s, NH, D2O exchangeable), 8.28 (2H, d, J = 7.8 Hz,
ArꢀH), 7.70 (1H, s, H6), 7.27ꢀ7.58 (15H, m, ArꢀH), 7.0 (2H, d, J =
7.8 Hz, ArꢀH), 6.48 (1H, d, J = 6.0 Hz, H10), 6.0 (1H, dd, J = 5.4 Hz,
H20), 5.79ꢀ5.86 (1H, m, H70), 5.08ꢀ5.12 (2H, m, H80), 4.79 (1H, d, J =
10.8 Hz, OBn), 4.50ꢀ4.57 (4H, m, CH2ꢀPh and H30), 3.75 (1H, d, J =
10.2 Hz, H500), 3.50 (1H, d, J = 10.2 Hz, H50), 2.68 (1H, dd, J = 6.0 and
15.0 Hz, H600), 2.37 (1H, dd, J = 7.8 and 14.4 Hz, H60), 1.67 (3H, s, Cy-
CH3). 13C NMR (150.9 MHz, CDCl3): δ 194.5 (CdS), 179.4 (COBz),
159.5 (C4), 153.5 (ArꢀC), 148.0 (C2CO), 137.3 (ArꢀC), 137.2
(ArꢀC), 137.16 (ArꢀC), 137.1 (C6), 132.5 (C70), 132.45 (ArꢀC),
130.0 (ArꢀC), 129.6 (ArꢀC), 129.4 (ArꢀC), 128.7 (ArꢀC), 128.5
(ArꢀC), 128.2 (ArꢀC), 128.15 (ArꢀC), 128.1 (ArꢀC), 127.9
(1R,3R,4R,5R,7S)-3-(N2-Acetylguanin-9-yl)-7-(2-cyanoethoxy-
diisopropylamino-phosphinoxy)-1-(4,40-dimethoxytrityloxy-
methyl)-5-methyl-2-oxa-bicyclo[2.2.1]heptane (49b). Com-
pound 48b (0.4 g, 0.614 mmol) was dissolved in 22 mL of dry DCM,
and DIPEA (0.534 mL, 3.07 mmol) was added at 0 °C followed by
2-cyanoethyl-N,N-diisopropylphosphoramidochloridite (0.273 mL,
1.23 mmol). The reaction was stirred overnight at rt. Methanol
(0.5 mL) was added followed by aqueous saturated NaHCO3 and
extracted twice (25 mL) with dichloromethane. The organic phase was
dried, evaporated, and chromatographed on silica gel (5% Et3N/DCM
v/v) to give 49b (0.4 g, 77%) as a mixture of two isomers. 31P NMR
(161.97 MHz, CDCl3): 149.57, 148.26. LC-MS m/z [M þ H]þ calcd for
C45H55N7O8P 852.38, found 852.305.
1-[4-C-Allyl-3,5-di-O-benzyl-2-O-acetyl-β-D-ribofuranosyl]-
N4-benzoyl-5-methylcytosine (38c). Acetic anhydride (35.3 mL,
343 mmol) and acetic acid (196.2 mL) were added to 36 (13.5 g, 32.9
mmol) and cooled, and triflic acid (0.28 mL) was added to it and stirred.
After 30 min the reaction was quenched with cold saturated NaHCO3
solution and extracted with ethyl acetate. The organic layer was washed
with saturated NaHCO3 until the aqueous layer is neutralized. The crude
mixturewas co-evaporated withdryCH3CN(3 ꢂ 350 mL) anddissolved
in the same. N4-Benzoyl-5-methylcytosine (7.92 g, 34.4 mmol) and N,
O-bis(trimethylsilyl)acetamide (15.3 mL, 85.2 mmol) were added to this
solution and refluxed for 45 min until the suspension becomes a clear
solution. This solution was cooled to 0 °C, and TMSOTf (5.72 mL, 23.2
mmol) was added dropwise and stirred overnight at rt. The organic
solvent was evaporated andquenched with saturated NH4Cl solution and
extracted with ethyl acetate (3 ꢂ 250 mL). The organic layer was dried,
evaporated, and chromatographed over silica gel eluting with 10% ethyl
acetate in hexane (v/v) to give 38c as a white foam solid (15.0 g, 80%). Rf
0.6 (30% ethyl acetate in hexane, v/v).1H NMR (600 MHz, CDCl3): δ
4427
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
(ArꢀC), 126.8 (ArꢀC), 121.7 (ArꢀC), 118.9 (C80), 112.3 (C5), 87.5
(C40), 85.8 (C10), 83.3 (C20), 77.9 (C30), 75.0 (OBn), 73.8 (OBn), 73.7
(C50), 37.5 (C60), 13.1 (Cy-CH3). MALDI-TOF m/z [M þ H]þ calcd
for C41H40N3O7S 718.259, found 718.261.
(ArꢀC), 128.07 (ArꢀC), 127.9 (ArꢀC), 127.89 (ArꢀC), 127.5
(ArꢀC), 127.4 (ArꢀC), 127.3 (ArꢀC), 110.3 (C5), 88.0 (C10), 85.6
(C40), 73.6 (OBn), 72.9 (C30), 71.7 (OBn), 70.7 (C50), 42.7 (C20), 26.8
(C60), 20.7 (C80), 17.7 (C70), 12.9 (Cy-CH3). MALDI-TOF m/z [M þ
H]þ calcd for C34H36N3O5 566.266, found 566.264.
(1R,3R,4R,5R,7S)-3-(N4-Benzoyl-5-methylcytosin-1-yl)-1-
benzyloxy-7-benzyloxymethyl-5-methyl-2-oxa-bicylco[2.2.1]-
heptane (43a), (1R,3R,4R,5S,7S)-3-(N4-Benzoyl-5-methylcy-
tosin-1-yl)-1-benzyloxy-7-benzyloxymethyl-5-methyl-2-oxa-
bicylco[2.2.1]heptane (43b), and (1R,5R7R,8S)-7-(N4-Ben-
zoyl-5-methylcytosin-1-yl)-8-benzyloxy-5-benzyloxmethyl-
6-oxa-bicyclo[3.2.1]octane (43c). Compound 40c (11.0 g, 15.34
mmol) was dissolved in 600 mL of dry toluene and purged with dry
nitrogen for 30 min. Bu3SnH (4.16 mL, 15.34 mmol) was dissolved in
50 mL of dry toluene, and half of this solution was added dropwise to
refluxing solution over 30 min. AIBN (3.0 g, 18.29 mmol) was dissolved
in 50 mL of dry toluene and added to the above solution dropwise, and
simultaneously was added the remaining solution of Bu3SnH over
60ꢀ70 min. After 60 min of reflux, the solution was cooled, CCl4
(100 mL) added, and the mixture stirred for 20 min. A solution of iodine
in DCM was added to the above solution until a faint coloration
persisted, and then solvent was evaporated. The solid obtained was
taken up in EtOAc. The organic layer was dried, evaporated, and
chromatographed over silica gel (10ꢀ12% EtOAc in hexane, v/v) to
give a mixture of 43a/b/c as a white foam (5.2 g, 60%, 1H NMR showed
that 43a:43b:43c = 79:12:9). Rf 0.6 (30% ethyl acetate in hexane, v/v).
Compound 43a: 1H NMR (500 MHz, CDCl3): δ 13.4 (1H, s, NH, D2O
exchangeable), 8.31 (2H, m, ArꢀH), 7.93 (1H, s, H6), 7.24ꢀ7.53 (13H,
m, ArꢀH), 5.81 (1H, s, H10), 4.43ꢀ4.63 (4H, m, 2 ꢂ OBn), 4.03 (1H, s,
H30), 3.82 (1H, d, J = 11.0 Hz, H500), 3.72 (1H, d, J = 10.5 Hz, H50), 2.66
(1H, m, H20), 2.60ꢀ2.66 (1H, m, H70), 2.08 (1H, dd, J = 11.0 and 12.0
Hz, H600), 1.71 (3H, s, Cy-CH3), 1.28 (3H, d, J = 7.5 Hz, 70CH3), 1.18
(1H, dd, J = 4.5 and 12.5 Hz, H60). 13C NMR (125.8 MHz, CDCl3): δ
179.5 (BzCO), 160.2 (C4), 147.5 (C2), 138.1 (C6), 137.8 (ArꢀC),
137.5 (ArꢀC), 137.4 (ArꢀC), 132.3 (ArꢀC), 129.8 (ArꢀC), 128.6
(ArꢀC), 128.4 (ArꢀC), 128.3 (ArꢀC), 128.1 (ArꢀC), 128.0 (ArꢀC),
127.9 (ArꢀC), 127.8 (ArꢀC), 127.5 (ArꢀC), 110.4 (C5), 89.1 (C40),
84.6 (C10), 78.6 (C30), 73.7 (OBn), 71.9 (OBn), 67.6 (C50), 47.6 (C20),
37.7 (C60), 28.8 (C70), 15.5 (70CH3), 13.2 (Cy-CH3). MALDI-TOF
m/z [M þ H]þ calcd for C34H36N3O5 566.266, found 566.269. 43b: 1H
NMR (500 MHz, CDCl3): δ 13.4 (1H, s, NH, D2O exchangeable),
8.30ꢀ8.32 (2H, m, ArꢀH), 7.92 (1H, s, H6), 7.25ꢀ7.53 (13H, m,
ArꢀH), 5.43 (1H, s, H10), 4.57ꢀ4.63 (3H, m, OBn), 4.23ꢀ4.45 (1H, d,
J = 11.5 Hz, OBn), 3.93 (1H, s, H30), 3.86 (1H, d, J = 11.0 Hz, H500), 3.80
(1H, d, J = 11.0 Hz, H50), 2.60ꢀ2.61 (1H, m, H20), 2.16ꢀ2.21 (1H, m,
H70), 1.95 (1H, dd, J = 9.0 and 12.5 Hz, H600), 1.67 (3H, s, Cy-CH3),
1.60 (1H, dd, J = 5.5 and 12.5 Hz, H60), 1.25 (3H, d, J = 7.5 Hz, 70CH3).
13C NMR (125.8 MHz, CDCl3): δ 179.4 (BzCO), 160.1 (C4), 147.7
(C2), 137.9 (C6), 137.8 (ArꢀC), 137.3 (ArꢀC), 132.3 (ArꢀC), 129.8
(ArꢀC), 129.0 (ArꢀC), 128.6 (ArꢀC), 128.5 (ArꢀC), 128.4 (ArꢀC),
128.1 (ArꢀC), 128.0 (ArꢀC), 127.9 (ArꢀC), 127.7 (ArꢀC), 127.5
(ArꢀC), 110.2 (C5), 89.1 (C10), 88.9 (C40), 78.2 (C30), 73.7 (OBn),
72.0 (OBn), 67.5 (C50), 47.7 (C20), 38.6 (C60), 33.6 (C70), 20.1
(70CH3), 13.1 (Cy-CH3). MALDI-TOF m/z [M þ H]þ calcd for
C34H36N3O5 566.266, found 566.267. 43c: 1H NMR (600 MHz,
CDCl3): δ 13.45 (1H, s, NH, D2O exchangeable), 8.30 (2H, m, ArꢀH),
8.28 (1H, s, H6), 7.29ꢀ7.52 (13H, m, ArꢀH), 5.87 (1H, s, H10),
4.60ꢀ4.62 (2H, m, OBn), 4.55 (1H, d, J = 11.4 Hz, OBn), 4.45 (1H, d,
J = 12.0 Hz, OBn), 4.17 (1H, d, J = 4.8 Hz, H30), 3.71 (1H, d, J = 10.8 Hz,
H500), 3.58 (1H, d, J = 10.2 Hz, H50), 2.65 (1H, m, H20), 1.91ꢀ1.97 (1H,
m, H800), 1.84ꢀ1.89 (1H, m, H600), 1.77ꢀ1.79 (1H, m, H80), 1.69ꢀ1.75
(2H, m, H70/700), 1.59 (3H, s, Cy-CH3), 1.35ꢀ1.37 (1H, m, H60). 13C
NMR (150.9 MHz, CDCl3): δ 179.5 (BzCO), 160.2 (C4), 147.9 (C2),
138.3 (C6), 137.8 (ArꢀC), 137.5 (ArꢀC), 137.4 (ArꢀC), 132.2
(ArꢀC), 129.8 (ArꢀC), 128.6 (ArꢀC), 128.5 (ArꢀC), 128.1
Isomeric Mixture of (1R,3R,4R,5R,7S)-7-Hydroxyl-1-hydroxy-
methyl-5-methyl-3-(5-methylcytosin-1-yl)-2-oxa-bicylco[2.2.1]-
heptane (46a), (1R,3R,4R,5S,7S)-7-Hydroxyl-1-hydroxymethyl-
5-methyl-3-(5-methylcytosin-1-yl)-2-oxa-bicylco[2.2.1]heptane
(46b), and (1R,5R,7R,8S)-8-Hydroxyl-5-hydroxymethyl-
7-(5-methylcytosin-1-yl)-6-oxa-bicyclo[3.2.1]octane (46c). A
mixture of 43a/b/c (3.5 g, 5.84 mmol) was dissolved in methanolic
ammonia (100 mL) and stirred under a nitrogen atmosphere for 8 h at rt.
The reaction mixture was concentrated under vacuum to get 430 as off-
white foam (2.5 g, 90%), Rf 0.2 (40% EtOAc in hexane, v/v). m/z of [M
þH]þ found 461.90, calculated mass of 461.23. To a solution of 430 (2.0
g, 4.33 mmol) in methanol (200 mL) were added 20% Pd(OH)2/C (4.0
g) and ammonium formate (16.3 g, 258 mmol), and the mixture was
heated at 60 °C for 16 h. After the reaction was finished as seen by TLC,
the suspension was filtered through Celite, and Pd(OH)2/C was washed
with (5 ꢂ 400 mL) methanol. The solvent was evaporated. The crude
compound was chromatographed over silica gel eluting with 10%
methanol in DCM (v/v) to give 46a/b/c as a white foamy solid
(0.833 g, 70%). Rf 0.13 (10% methanol in DCM, v/v). 46a is the major
product, and 46b is the minor product. 1H NMR (400 MHz, DMSO-
d6): δ 8.2 (0.09H, s, H6, Minor), 7.79ꢀ7.80 (1H, m, H6, Major), 7.2
(1H, s, NH, D2O exchangeable), 6.7 (1H, s, NH, D2O exchangeable),
5.55 (1H, s, H10, Major), 5.02ꢀ5.2 (2.3H, m, H10, Minor and OH, D2O
exchangeable), 3.8ꢀ3.98 (1H, m, H30 Major and Minor), 3.52ꢀ3.74
(2H, m, H50, H500), 2.52ꢀ2.59 (1H, m, H70, Major), 2.09ꢀ2.2 (1H, m,
H20 Major and Minor), 1.99 (0.5H, m, H70 Minor), 1.80ꢀ1.90 (4.0H,
m, Cy-CH3 Major and Minor, H600 Major), 1.71ꢀ
1.75 (0.5H, m, H600, Minor), 1.39 (0.3H, s, H60, Minor), 1.13ꢀ1.22
(3.2H, m, 70CH3, Major and Minor), 0.90 (1H, dd, J = 4.5 and 12.2 Hz,
H60 Major). 13C NMR (100.6 MHz, DMSO-d6): δ 164.9 (C2), 154.4,
154.3 (C4, Major), 138.1, 137.9 (C6), 99.1, 98.9 (C5), 89.5 (C40), 87.5
(C10, Minor), 82.4 (C10, Major), 70.0, 69.8 (C30), 58.1, 57.8 (C50), 49.3
(C20), 36.3 (C60, Minor), 36.4 (C60, Major), 33.1 (C70, Minor), 27.8
(C70, Major), 20.1 (70CH3, Minor), 15.1 (70CH3, Major), 13.0 (Cy-CH3
Major and Minor). MALDI-TOF m/z [M þ H]þ calcd for C13H20N3O4
282.146, found 282.150.
Isomeric Mixture of (1R,3R,4R,5R,7S)-3-(N4-Benzoyl-5-
methylcytosin-1-yl)-7-hydroxyl-1-hydroxymethyl-5-methyl-
2-oxa-bicylco[2.2.1]heptane (47a), (1R,3R,4R,5S,7S)-3-(N4-
Benzoyl-5-methylcytosin-1-yl)-7-hydroxyl-1-hydroxymethyl-
5-methyl-2-oxa-bicylco[2.2.1]heptane (47b), and (1R,5R,7R,-
8S)-7-(N4-Benzoyl-5-methylcytosin-1-yl)-8-hydroxyl-5-hydro-
xymethyl-6-oxa-bicyclo[3.2.1]octane (47c). A mixture of com-
pound 46a/b/c (2.5 g, 8.89 mmol) was evaporated twice with dry
pyridine and suspended in the same (15 mL). To this solution benzoic
anhydride (4.0 g, 17.6 mmol) was added and stirred at rt overnight and
diluted with EtOH (32.5 mL), and 2 M NaOH (50 mL) was added. After
stirring for 1 h, AcOH (6.25 mL) was added to the solution, and the
solventwasremoved under reduced pressure. Theresidue was purified by
chromatography over silica gel, 40% ethyl acetate in hexane (v/v) to give
three portions: 47a/47b/47c as a mixture (2.7 g, 77%), 47a (0.600 g,
13%), and 47b (0.080 g, 10%). Rf 0.4 (60% EtOAc in hexane, v/v). 47a:
1H NMR (600 MHz, DMSO-d6) δ 13.3 (1H, s, NH, D2O exchangeable),
8.2 (3H, bs, ArꢀH), 7.58ꢀ7.60 (1H, m, ArꢀH), 7.48ꢀ7.54 (2H, m,
ArꢀH), 5.67 (1H, s, H10), 5.32 (1H, s, OH, D2O exchangeable), 5.18
(1H, t, J = 4.8 Hz, OH, D2O exchangeable), 4.00 (1H, s, H30), 3.67 (1H,
dd, J = 4.8 and 12.6 Hz, H500), 3.60 (1H, dd, J = 4.8 and 12.0 Hz, H50),
2.57ꢀ2.65 (1H, m, H70), 2.31 (1H, d, J = 3.0 Hz, H20), 2.0 (3H, s, Cy-
CH3), 1.91 (1H, t, J = 11.4 Hz, H600), 1.17 (3H, d, J = 7.2 Hz, 70CH3),
4428
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
0.95 (1H, dd, J = 4.2 and 12.6 Hz, H60). 13C NMR (150.9 MHz, DMSO-
d6) δ 139.3 (C6), 132.9 (ArꢀC), 132.3 (ArꢀC), 129.2 (ArꢀC), 128.2
(ArꢀC), 90.7 (C40), 83.9 (C10), 70.4 (C30), 58.2 (C50), 49.5 (C20),
36.6 (C60), 28.2 (C70), 15.4 (70CH3), 13.4 (Cy-CH3). m/z [MþH]þ
calcd for C20H24N3O5 384.16, found 384.00. 47b: 1H NMR (500 MHz,
DMSO-d6) δ 13.3 (1H, s, NH, D2O exchangeable), 8.23 (1H, s, H6),
8.02ꢀ8.31 (2H, m, ArꢀH), 7.54ꢀ7.62 (1H, m, ArꢀH), 7.42ꢀ7.51
(2H, m, ArꢀH), 5.33 (1H, d, J = 3.0 Hz, 30-OH, D2O exchangeable), 5.29
(1H, s, H10), 5.22 (1H, bt, 50-OH, D2O exchangeable), 3.95 (1H, s, H30),
3.70 (2H, s, H50 and H500), 2.28 (1H, s, H20), 1.21 (3H, d, J = 7.0 Hz,
70CH3), 2.06 (1H, m, H70), 2.01 (3H, s, Cy-CH3), 1.76 (1H, dd, J = 9.0
and 12.0 Hz, H600), 1.44 (1H, dd, J = 5.0 and 12.5 Hz, H60). 13C NMR
(125.8 MHz, DMSO-d6) δ 139.6 (C6), 132.3 (ArꢀC), 129.1 (ArꢀC),
128.6 (ArꢀC), 128.2 (ArꢀC), 107.9 (C5), 90.7 (C40), 88.4 (C10), 70.1
(C30), 58.6 (C50), 49.4 (C20), 37.6 (C60), 33.3 (C70), 20.4 (70CH3),
13.3 (Cy-CH3). MALDI-TOF m/z [M þ H]þ calcd for C20H24N3O5
386.172, found 386.171.
extracted with dry DCM (2 ꢂ 15 mL). The organic phase was dried and
washed with hexane:DCM (4:1) 3 times to give 51a/b/c (0.250 g, 75%)
as a mixture of six isomers as a white solid. Rf 0.75 (40% EtOAc/Hexane,
v/v). 31P NMR (161.97 MHz, CDCl3): δ 150.20, 150.04, 149.46,
149.22, 148.94, 148.80. LC-MS m/z [M þ H]þ calcd for C50H59N5O8P
888.41, found 888.50.
Oligonucleotide Synthesis and Purification. All AONs were
synthesized using an automated DNA/RNA synthesizer based on
phosphoramidite chemistry. For native A, G, and C building blocks,
fast deprotecting phosphoramidites (Ac for C, iPr-Pac for G, Pac for A)
were used. Standard DNA synthesis reagents and cycle were used except
that 0.25 M 5-[3,5-bis(trifluoromethyl)phenyl]-1H-tetrazole (activator
42) was used as the activator and Tac2O as the cap A. For incorporation
of modified nucleotides, extended coupling time (10 min compared to
25 s for native nucleotides) was used. The deprotections were carried
out in 33% aqueous NH3 for 10 h at 55 °C. After deprotection, all crude
oligos were purified by denaturing PAGE (20% polyacrylamide with 7 M
urea), extracted with 0.3 M NaOAc, and desalted with a C-18 reverse
phase cartridge to give AONs in >99% purity, and correct masses have
been obtained by MALDI-TOF mass spectroscopy for each of them.
RNA was also synthesized by a solid-supported phosphoramidite
approach based on the 20-O-TEM strategy.
Isomeric Mixture of (1R,3R,4R,5R,7S)-3-(N4-Benzoyl-
5-methylcytosin-1-yl)-1-(4,40-dimethoxytrityloxymethyl)-
7-hydroxyl-5-methyl-2-oxa-bicylco[2.2.1]heptane (50a),
(1R,3R,4R,5S,7S)-3-(N4-Benzoyl-5-methylcytosin-1-yl)-1-(4,40-
dimethoxytrityloxymethyl)-7-hydroxyl-5-methyl-2-oxa-bicylco-
[2.2.1]heptane (50b), and (1R,5R,7R,8S)-7-(N4-Benzoyl-5-
methylcytosin-1-yl)-5-(4,40-dimethoxytrityloxymethyl)-8-
hydroxyl-6-oxa-bicyclo[3.2.1]octane (50c). A mixture of 47a/b/c
(0.190 g, 0.493 mmol) was evaporated twice with dry pyridine and
suspended in the same (10 mL). 4,40-Dimethoxytrityl chloride (0.256 g,
0.757 mmol) was added and stirred overnight at rt. The reaction was
quenchedwith saturated aqueousNaHCO3 and extracted with DCM (3 ꢂ
10 mL). The organic layer was dried and evaporated, and chromatographic
separation over silica gel (1% pyridine/EtOAc in hexane, v/v) furnished
50a/50b/50c (0.180 g, 53%) as yellow foam. Rf 0.60 (40% EtOAc/
Hexane, v/v). 50a is the major product, and 50b is the minor product. 1H
NMR (400 MHz, CDCl3): δ 13.4 (1H, s, NH, D2O exchangeable), 8.30
(2H, d, J = 7.3 Hz, ArꢀH), 8.0 (1H, s, H6), 7.21ꢀ7.54 (14H, m, ArꢀH),
6.86 (4H, dd, J = 3.5 and 5.4 Hz, ArꢀH), 5.7 (1H, s, H10 Major), 5.36
(0.27H, s, H10 Minor), 4.28 (1.2H, s, H30 Major and Minor), 3.79 (7H, s,
OMe, Major and Minor), 3.60ꢀ3.63 (1.4H, m, H50 Major and Minor),
3.37 (0.38H, d, J = 11.1 Hz, H50, Minor), 3.30 (1H, d, J = 11.1 Hz, H50),
2.64 (1H, bs, H70 Major), 2.54 (1H, d, J = 3.0 Hz, H20 Major), 2.18 (0.3H,
m, H70 Minor), 2.46 (0.3H, s, H20 Minor), 1.89ꢀ1.99 (1.4H, m, H600
Major and Minor), 1.76 (3H, s, Cy-CH3), 1.71 (1H, s, Cy-CH3, Minor),
1.47 (0.5H, dd, J = 4.7 and 11.3 Hz, H60, Minor), 1.21ꢀ1.26 (4H, m,
70CH3), 1.11 (1H, d, J = 4.9 and 12.8 Hz, H60 Major). 13C NMR (100.6
MHz, CDCl3): δ 179.1 (COBz), 160.2 (C4), 158.7 (ArꢀC), 144.6 (C2),
137.5, 137.3 (C6), 135.6 (ArꢀC), 135.5 (ArꢀC), 132.3 (ArꢀC), 130.1
(ArꢀC), 129.8 (ArꢀC), 128.1 (ArꢀC), 127.1 (ArꢀC), 113.3 (ArꢀC),
110.6 (C5), 89.7 (C40), 86.7 (DMTr-C), 84.5 (C10), 72.9 (C30), 61.0
(C50), 55.3(OMe), 50.2 (C20), 37.4(C60), 28.5(C70), 15.4(70CH3), 13.5
(Cy-CH3). m/z [M þ H]þ calcd for C41H42N3O7 688.30, found 688.20.
(1R,3R,4R,5R,7S)-3-(N4-Benzoyl-5-methylcytosin-1-yl)-7-
(2-cyanoethoxy-diisopropylamino-phosphinoxy)-1-(4,40-di-
methoxytrityloxymethyl)-5-methyl-2-oxa-bicylco[2.2.1]heptane
(51a), (1R,3R,4R,5S,7S)-3-(N4-Benzoyl-5-methylcytosin-1-yl)-7-
(2-cyanoethoxy-diisopropylamino-phosphinoxy)-1-(4,40-di-
methoxytrityloxymethyl)-5-methyl-2-oxa-bicylco[2.2.1]heptane
(51b), and (1R,5R,7R,8S)-7-(N4-Benzoyl-5-methylcytosin-1-yl)-
8-(2-cyanoethoxy-diisopropylamino-phosphinoxy)-5-(4,40-
dimethoxytrityloxymethyl)-6-oxa-bicyclo[3.2.1]octane (51c).
Compound 50a/b/c (0.250 g, 0.364 mmol) was dissolved in 10 mL of
dry DCM, and DIPEA (0.32 mL, 1.802 mmol) was added at 0 °C
followed by 2-cynoethyl-N,N-diisopropylphosphoramidochloridite
(0.16 mL, 0.728 mmol). The reaction was stirred at rt for 2 h. Methanol
(0.3 mL) was added followed by aqueous saturated NaHCO3 and
UV Melting Experiments. Determination of the Tm of the AON/
RNA hybrids or AON/DNA duplex was carried out in the following
buffer: 60 mM Tris-HCl (pH 7.5), 60 mM KCl, 0.8 mM MgCl2.
Absorbance was monitored at 260 nm in the temperature range from
20 to 95 °C using an UV spectrophotometer equipped with a Peltier
temperature programmer with the heating rate of 1 °C/min. Prior to
measurements, the samples (1 μM of AON and 1 μM cDNA or RNA
mixture) were preannealed by heating to 80 °C for 5 min followed by
slow cooling to 21 °C and 30 min equilibration at this temperature. The
value of Tm is the average of two or three independent measurements. If
the error of the first two measurements is > ( 0.3 °C, the third
measurement was carried out to check if the error is indeed within
(0.3 °C; otherwise, it is repeated.
32P Labeling of Oligonucleotides. The oligoribonucleotides
and oligodeoxyribonucleotides were 50-end labeled with 32P using T4
polynucleotide kinase, [γ-32P]ATP, and the standard protocol. Labeled
AONs and the target RNA were purified by a Nucleotide Removal Kit,
and specific activities were measured using a counter.
SVPDE Degradation of AONs 1ꢀ8. Stability of the AONs
toward 30-exonucleases was tested using phosphodiesterase I from
Crotalus adamanteus. All reactions were performed at 3 μM DNA
concentration (50-end 32P labeled with specific activity 80 000 cpm) in
100 mM Tris-HCl (pH 8.0) and 15 mM MgCl2 at 21 °C. An exonuclease
concentration of 6.7 ng/μL was used for digestion of oligonucleotides.
Total reaction volume was 30 μL. Aliquots (3 μL) were taken at proper
time points and quenched by addition of stop solution (4 μL)
[containing 0.05 M EDTA, 0.05% (w/v) bromophenol blue, and 0.05%
(w/v) xylene cyanole in 80% formamide]. Reaction progress was mon-
itored by 20% denaturing (7 M urea) PAGE and autoradiography.
MichaelisꢀMenten Kinetics of SVPDE Digestion of AONs
9ꢀ16. AONs (concentration varied from 1 to 100 μM) and SVPDE
(for AONs 9ꢀ12, 0.000055 unit SVPDE used; for AONs 13ꢀ16,
0.0000825 unit SVPDE used) were mixed in a buffer containing100 mM
Tris-HCl, pH 8.0, 100 mM NaCl, and 15 mM MgCl2 at 21 °C, with a
total reaction volume of 30 μL. Aliquots (3 μL) were taken at proper
time points and quenched by addition of stop solution (4 μL)
[containing 0.05 M EDTA, 0.05% (w/v) bromophenol blue, and
0.05% (w/v) xylene cyanole in 80% formamide]. Reaction progress
was monitored by 20% denaturing (7 M urea) PAGE and autoradio-
graphy. The total percentage of integrated AON was plotted against time
points to give the initial velocities (v0) of the reactions (see Supporting
Information part IV). Values of the kinetic parameters (KM and Vmax) in
4429
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
this method were determined directly from v0 versus concentration of
substrate (C0) plots (see Supporting Information part IV) using a
correlation Program, where the correlation equation was y = ax/(b þ
x), where a = Vmax and b = KM. kcat was obtained from this equation:
kcat = Vmax/E0, where E0 corresponds to the concentration of enzyme
used in the reaction. All values are averages and standard deviations of
3ꢀ4 independent experiments.
DCM, dichloromethane; DMTr, (4,40-dimethoxytrityl); DIPEA,
(diisopropylethylamine); EtOAc, (ethyl acetate)
’ REFERENCES
(1) (a) Castanotto, D.; Rossi, J. J. Nature 2009, 457, 426–433.
(b) Kurreck, J. Eur. J. Biochem. 2003, 270, 1628–1644.
(2) (a) Watts, J. K.; Deleavey, G. F.; Damha, M. J. Drug Discovery
Today 2008, 13, 842–855. (b) Cook, P. D. In Current Protocols in Nucleic
Acid Chemistry; John Wiley & Sons, Inc.: New York, 2000. (c) Shukla, S.;
Sumaria, C. S.; Pradeepkumar, P. I. ChemMedChem 2010, 5, 328–349.
(3) (a) Obika, S.; Nanbu, D.; Hari, Y.; Morio, K.-I.; In, Y.; Ishida, T.;
Imanishi, T. Tetrahedron Lett. 1997, 38, 8735–8738. (b) Obika, S.;
Nanbu, D.; Hari, Y.; Andoh, J.-I.; Morio, K.-I.; Doi, T.; Imanishi, T.
Tetrahedron Lett. 1998, 39, 5401–5404. (c) Imanishi, T.; Obika, S.
J. Synth. Org. Chem. Jpn. 1999, 57, 969–980. (d) Singh, S. K.; Nielsen, P.;
Koshkin, A. A.; Wengel, J. Chem. Commun. 1998, 455–456. (e) Koshkin,
A. A.; Singh, S. K.; Nielsen, P.; Rajwanshi, V. K.; Kumar, R.; Meldgaard,
M.; Olsen, C. E.; Wengel, J. Tetrahedron 1998, 54, 3607–3630. (f) Singh,
S. K.; Wengel, J. Chem. Commun. 1998, 1247–1248. (g) Koshkin, A. A.;
Rajwanshi, V. K.; Wengel, J. Tetrahedron Lett. 1998, 39, 4381–4384.
(h) Koshkin, A. A.; Nielsen, P.; Meldgaard, M.; Rajwanshi, V. K.; Singh,
S. K.; Wengel, J. J. Am. Chem. Soc. 1998, 120, 13252–13253.
(4) (a) Kumar, S.; Hansen, M. H.; Albaek, N.; Steffansen, S. I.;
Peterson, M.; Nielsen, P. J. Org. Chem. 2009, 74, 6756–6769.
(b) Plashkevych, O.; Chatterjee, S.; Honcharenko, D.; Pathmasiri, W.;
Chattopadhayaya, J. J. Org. Chem. 2007, 72, 4716–4726 and references
cited therein.
Stability Studies in Human Blood Serum. AONs at 2 μM
concentration (50-end 32P labeled with specific activity 80 000 cpm)
were incubated in 10 μL of human blood serum (male AB) at 21 °C
(total reaction volume was 36 μL). Aliquots (3 μL) were taken at proper
time points and quenched with 4 μL of stop solution [containing 0.05 M
EDTA, 0.05% (w/v) bromophenol blue, and 0.05% (w/v) xylene
cyanole in 80% formamide], resolved in 20% polyacrylamide denaturing
(7 M urea) gel electrophoresis, and visualized by autoradiography.
RNase H Digestion Assay. Target 0.1 μM RNA (specific activity
80 000 cpm) and AON (2 μM) were incubated in a buffer containing
20 mM Tris-HCl (pH 7.5), 20 mM KCl, 10 mM MgCl2, 0.1 mM EDTA,
and 0.1 mM DTT at 21 °C in the presence of 0.08 U E. coli RNase H1.
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 slow cooling to 21 °C and 30 min equilibration at this
temperature. Total reaction volume was 30 μL. Aliquots of 3 μL were
removed after 5, 10, 15, 30, and 60 min, and the reactions were
terminated by mixing with stop solution [containing 0.05 M EDTA,
0.05% (w/v) bromophenol blue, and 0.05% (w/v) xylene cyanole in
80% formamide]. The samples were subjected to 20% 7 M urea PAGE
and visualized by autoradiography. Pseudofirst-order reaction rates
could be obtained by fitting the digestion curves to single-exponential
decay functions.
(5) Petersen, M.; Nielsen, C. B.; Nielsen, K. E.; Jensen, G. A.;
Bondensgaard, K.; Singh, S. K.; Rajwanshi, V. K.; Koshkin, A. A.; Dahl,
B. M.; Wengel, J.; Jacobsen, J. P. J. Mol. Recognit. 2000, 13, 44–53.
(6) (a) Kaur, H.; Babu, B. R.; Maiti, S. Chem. Rev. 2007,
107, 4642–4697. (b) Zhou, C.; Chattopadhyaya, J. Curr. Opin. Drug.
Discovery Dev. 2009, 12, 876–898.
’ ASSOCIATED CONTENT
(7) (a) Singh, S. K.; Kumar, R.; Wengel, J. J. Org. Chem. 1998, 63
(26), 10035–10039. (b) Seth, P. P.; Vasquez, G.; Allerson, C. A.;
Berdeja, A.; Gaus, H.; Kinberger, G. A.; Prakash, T. P.; Migawa,
M. T.; Bhat, B.; Swayze, E. E. J. Org. Chem. 2010, 75, 1569–1581.
(c) Prakash, T. P.; Siwkowski, A.; Allerson, C. R.; Migawa, M. T.; Lee, S.;
Gaus, H. J.; Black, C.; Seth, P. P.; Swayze, E. E.; Bhat, B. J. Med. Chem.
2010, 53, 1636–1650. (d) Hari, Y.; Obika, S.; Ohnishi, R.; Eguchi, K.;
Osaki, T.; Ohishi, H.; Imanishi, T. Bioorg. Med. Chem. 2006, 14,
1029–1038. (e) Rahman, S. M. A.; Seki, S.; Obika, S.; Yoshikawa, H.;
Miyashita, K.; Imanishi, T. J. Am. Chem. Soc. 2008, 130, 4886–4896.
(8) Pradeepkumar, P. I.; Cheruku, P.; Plashkevych, O.; Acharya, P.;
Gohil, S.; Chattopadhyaya, J. J. Am. Chem. Soc. 2004, 126, 11484–11499.
(9) Honcharenko, D.; Varghese, O. P.; Plashkevych, O.; Barman, J.;
Chattopadhyaya, J. J. Org. Chem. 2006, 71, 299–314.
S
Supporting Information. NMR data for all the new
b
synthesized compounds; PAGE pictures as well as degradation
curves, showing the cleavage kinetics of AONs by SVPDE, blood
serum. This material is available free of charge via the Internet at
http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: þ46-18-4714577. Fax: þ46-18-554495. E-mail: jyoti@
boc.uu.se.
Author Contributions
§RamShankar Upadhayaya, Sachin Gangadhar Deshpande, and
Qing Li have contributed equally to this work.
(10) Koizumi, M. Curr. Opin. Mol. Ther. 2006, 8, 144–149.
(11) Morita, K.; Takagi, M.; Hasegawa, C.; Kaneko, M.; Tsutsumi,
S.; Sone, J.; Ishikawa, T.; Imanishi, T.; Koizumi, M. Bioorg. Med. Chem.
2003, 11, 2211–2226.
(12) Varghese, O. P.; Barman, J.; Pathmasiri, W.; Plashkevych, O.;
Honcharenko, D.; Chattopadhyaya, J. J. Am. Chem. Soc. 2006, 128,
15173–15187.
(13) (a) Albæk, N.; Petersen, M.; Nielsen, P. J. Org. Chem. 2006,
71, 7731–7740. (b) Enderlin, G.; Nielsen, P. J. Org. Chem. 2008, 73,
6891–6894.
(14) (a) Srivastava, P.; Barman, J.; Pathmasiri, W.; Plashkevych, O.;
Wenska, M.; Chattopadhyaya, J. J. Am. Chem. Soc. 2007, 129,
8362–8379. (b) Zhou, C.; Plashkevych, O.; Chattopadhyaya, J. Org.
Biomol. Chem. 2008, 6, 4627–4633. (c) Zhou, C.; Liu, Y.; Andaloussi,
M.; Badgujar, N.; Plashkevych, O.; Chattopadhyaya, J. J. Org. Chem.
2009, 74, 118–134. (d) Xu, J. F.; Liu, Y.; Dupouy, C.; Chattopadhyaya, J.
J. Org. Chem. 2009, 74, 6534–6554. (e) Li, Q.; Yuan, F.; Zhou, C.;
Plashkevych, O.; Chattopadhyaya, J. J. Org. Chem. 2010, 75, 6122–6140.
(f) Liu, Y.; Xu, J.; Karimiahmadabadi, M.; Zhou, C.; Chattopadhyaya, J.
J. Org. Chem. 2010, 75, 7112–7128. (g) Zhou, C.; Plashkevych, O.;
’ ACKNOWLEDGMENT
Generous financial support from the Swedish Natural Science
Research Council (Vetenskapsradet), the Swedish Foundation for
ꢃ
Strategic Research (Stiftelsen f€or Strategisk Forskning), the EU-FP6
funded RIGHT project (Project no. LSHB-CT-2004-005276), and
Department of Science and Technology, India (Grant no.VII-
PRDSF/109/05-06), is gratefully acknowledged. Authors also
acknowledge the contribution of Yogesh Walunj, Sachin Borude,
and Tushar Gawali for scaleup of some of the reactions.
’ ABBREVIATIONS
Bn, (benzyl); Ac, (acetyl); TMSOTf, (trimethylsilyl triflate); THF,
(tetrahydrofuran); AIBN, [2,20-azobis(2-methyl-propionitrile)];
4430
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431

The Journal of Organic Chemistry
ARTICLE
Chattopadhyaya, J. J. Org. Chem. 2009, 74, 3248–3265. (h) Seth, P. P.;
Allerson, C. R.; Berdeja, A.; Siwkowski, A.; Pallan, P. S.; Gaus, H.;
Prakash, T. P.; Watt, A. T.; Egli, M.; Swayze, E. E. J. Am. Chem. Soc. 2010,
132, 14942–14950.
(15) Zhou, C.; Chattopadhyaya, J. J. Org. Chem. 2010, 75, 2341–
2349.
(16) Wenska, M.; Honcharenko, D.; Pathmasiri, W.; Chattopadhyaya,
J. Heterocycles 2007, 73, 303–324.
(17) Robins, M. J.; Zou, R.; Guo, Z.; Wnuk, S. F. J. Org. Chem. 1996,
61, 9207–9212.
(18) Buchini, S.; Leumann, C. J. Eur. J. Org. Chem. 2006, 3152–3168.
(19) (a) Julia, M. Acc. Chem. Res. 1971, 4, 386–392. (b) Beckwith,
A. L. J. Tetrahedron 1981, 37, 3073–3100. (c) Beckwith, A. L. J. Chem.
Soc. Rev. 1993, 143–151.
on the number of 5-Me(T) stacking interactions, and the contribution of
the base-pairing strength is relatively smaller. In contradistinction,
RNA/RNA duplexes are more stable than DNA/DNA duplexes because
of larger energy gain from the base-pairing in the former compared to the
latter:Acharya, P.; Cheruku, P.; Chatterjee, S.; Acharya, S.; Chattopad-
hyaya, J. J. Am. Chem. Soc. 2004, 126, 2862. A comparison of these two
linear plots ([ΔGobp]RR-DD/DR versus [ΔGostacking]RR-DD/DR as a
function of % A-T/U bp content) with opposite slope shows that with
the increasing content of A-T base pairs the stability of the DNA/DNA
duplex weakens over the corresponding DNA/RNA or RNA/RNA
duplexes ([ΔGostacking]RR-DD/RD), while the strength of stacking
([ΔGostacking]RR-DD/RD) of A-T rich DNA/DNA sequence increases
in comparison with A-U rich sequence in RNA/RNA duplexes. This
increased stacking contribution from T compared to U, in DNA/DNA,
followed by DNA/RNA over RNA/RNA duplex, comes from favorable
electrostatic CH/σ* interaction between the 5-methyl group of T with
the nearest-neighbor A in the AT-rich sequence.
(33) Recently, carba-LNA has been studied against various biologi-
cal targets to evaluate the true potential of carba-LNA (cLNA). Corey
et al. tested the broad spectrum of modified antisense oligonucleotides
(AON) targeted to patient-derived fibroblast cell lines GM04281
(69 CAG repeat mutant allele), GM04717 (41 CAG repeat mutant
allele), and GM04719 (44 CAG repeat mutant allele) involving a direct
comparison of IC50s (based on protein and RNA analysis) within LNA
and cLNA modified AON (with phosphodiester backbone). Corey;
Biochemistry 2010, 49, 10166–10178. These studies clearly showed that
cLNA is better than any other modification including LNA in terms of
selective inhibition of HTT. Quantification of Western blots upon dose
response revealed that LNA modified AON potencies range from IC50
values of 27ꢀ40 nM and selectivities of inhibition for mutant versus
wild-type HTT by as much as >3.7-fold, whereas cLNA modified AONs
were found to be more potent compared to their LNA counterpart with
IC50 of 15 nM and >6.6-fold selectivity over the wild-type. Corey et al.
also studied phosphorothioate modified AONs and revealed that cLNA-
PS and LNA-PS achieved comparable potent and selective inhibition of
HTT. Another study involved carba-LNA (cLNA) and LNA modified
siRNAs for silencing the target HIV-1 TAR RNA employing a standard
single cell replication model using HEK 293T cells:Suman; Med. Chem.
Commun. 2011, 2, 206. This study clearly showed that cLNA modified
siRNAs (IC50 = 0.54 nM) have 2-fold lower IC50 values than that of their
LNA counterpart (IC50 = 1.13 nM) with 3ꢀ4-fold higher stability in
human blood serum. Apart from these studies, Punit et al. have studied
the potential of carba-LNA (cLNA) versus LNA and methylene-cLNA
in cell culture (brain endothelial cells using electroporation) and animal
experiments (subchronic dosing schedule in mice).Punit; J. Am. Chem.
Soc. 2010, 132, 14942–14950. These experiments revealed that carba-
LNA were slightly less potent compared to LNA and methylene-cLNA
AONs. This particular set of experiments differs in their delivery
techniques as the electroporation method has been used to assist the
delivery of AONs, while the earlier studies have used lipid-based
transfacting agents; therefore, the transfection efficiency of carba-LNA
modified AONs should be studied in light of the electroporation
technique. It is also clear that it is not always straightforward to
understand why particularly modified AONs are more effective than
others, especially with such a wide variety of targets and experimental
conditions. A number of factors can complicate an oligonucleotide’s
ability to reach its target RNA and inhibit gene expression. Simple
explanations might include poor transfection or release from endosomes
after cell entry and cellular uptake.
(20) Ti, G. S.; Gaffney, B. L.; Jones, R. A. J. Am. Chem. Soc. 1982,
104, 1316–1319.
(21) Foldesi, A.; Yamakage, S.-I.; Nilson, F. P. R.; Maltseva, T. V.;
Chattopadhyaya, J. Nucleic Acids Res. 1996, 24, 1187–1194.
(22) Debart, F.; Rayner, B.; Degols, G.; Imbach, J.ꢀL. Nucleic Acids
Res. 1992, 20, 1193–1200.
(23) Foldesi, A.; Trifonova, A.; Dinya, Z.; Chattopadhyaya, J. Tetra-
hedron Lett. 1999, 40, 7283–7284.
(24) (a) Koshkin, A. A.; Fensholdt, J.; Pfundheller, H. M.; Lomholt,
C. J. Org. Chem. 2001, 66, 8504–8512. (b) Kumar, T. S.; Kumar, P.;
Sharma, P. K.; Hrdlicka, P. J. Tetrahedron Lett. 2008, 49, 7168–7170.
(25) D€orper, T.; Winnacker, E.-L. Nucleic Acids Res. 1983,
11, 2575–2584.
(26) The cause for the formation of P(V) was described; for
oxidation of P(III) to P(V) phosphoramidite under aqueous acidic
conditions is known: Sanghvi, Y. S.; Guo, Z.; Pfundheller, H. M.;
Converso, A. Org. Process Res. Dev. 2005, 9, 730–737.
(27) Beaucage, S. L.; Caruthers, M. H. In Current Protocols in Nucleic
Acid Chemistry; John Wiley & Sons: New York, 2003.
(28) Shaw, J. P.; Kent, K.; Bird, J.; Fishback, J.; Froehler, B. Nucleic
Acids Res. 1991, 19, 747–750.
(29) Miyashita, K; Abdur Rahman, S. M.; Seki, S.; Obika, S.;
Imanishi, T. Chem. Commun. 2007, 3765–3767.
(30) (a) Razzell, W. E.; Khorana, H. G. J. Biol. Chem. 1959,
234, 2105–2117. (b) Razzell, W. E. Methods Enzymol. 1963, 6, 236–258.
(31) Pradeepkumar, P. I.; Chattopadhyaya, J. J. Chem. Soc., Perkin
Trans. 2001, 2, 2074–2083.
(32) Why are AON/RNA duplexes more stable than their AON/
DNA counterpart? The total free energy of stabilization of a duplex has
at least two major components besides hydration: the free-energy of
base-pairing and free-energy of stacking. This interdependency of
stacking and hydrogen bonding has been shown to play a key role in
the total free-energy of stabilization of AON/DNA versus AON/RNA
versus RNA/RNA duplexes. In most isosequential duplexes (T is
switched for U in RNA), the order of stability follows the following
order RNA/RNA > DNA/RNA > DNA/DNA because of variable
contribution of stacking versus hydrogen bonding. The dissection of
the relative energetic contributions from the free-energy of base-pairing
and free-energy of stacking to the total free-energy of stabilization allows
us to understand the differences in the intrinsic nature of the electrostatic
forces that are responsible for the self-assembly of the heteroduplexes
(AON/RNA) compared to homoduplexes (DNA/DNA versus RNA/
RNA). The pKa differences between the monomeric nucleotides in both
20-deoxy (dN) and ribo (rN) series (N = A/G/C/T/U), as the model
donor and acceptor (in which stacking is completely eliminated), gives:
Acharya, P.; Cheruku, P.; Chatterjee, S.; Acharya, S.; Chattopadhyaya, J.
J. Am. Chem. Soc. 2004, 126, 2862. The strength of base-pairing energies
in different homo- and heteroduplexes:Chatterjee, S.; Pathmasiri, W.;
Chattopadhyaya, J. Org. Biomol. Chem. 2005, 3, 3911–3915. The high
correlation of ΔG°37 of the helix stability and the sum of ΔpKa values in a
duplex (ΣΔpKa) give a powerful tool to separate base-pairing contribu-
tion of RNA/RNA over DNA/RNA and DNA/DNA ([ΔGobp]RR-DD)
from the free energy of the total helix stability ([ΔGobp]RR-DD or DR).
The study shows that the stability of DNA/RNA duplexes is dependent
4431
dx.doi.org/10.1021/jo200073q |J. Org. Chem. 2011, 76, 4408–4431