A cyclic dinucleotide with a four-carbon 5′-C-to-5′-C connection; Synthesis by RCM, NMR-examination and incorporation into secondary nucleic acid structures

Article abstract of DOI:10.1039/b603830a

A 5′-C-allylthymidine derivative was prepared from thymidine by the application of a stereoselective allylation procedure and its 5′(S)-configuration was confirmed. From this nucleoside derivative, appropriately protected building blocks were prepared and

Full text of DOI:10.1039/b603830a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
A cyclic dinucleotide with a four-carbon 5ꢀ-C-to-5ꢀ-C connection; synthesis by
RCM, NMR-examination and incorporation into secondary nucleic acid
structures
Pawan K. Sharma,† Birgitte H. Mikkelsen, Mikkel S. Christensen, Katrine E. Nielsen, Claus Kirchhoff,
Søren L. Pedersen, Anders M. Sørensen, Kirsten Østergaard, Michael Petersen and Poul Nielsen*
Received 14th March 2006, Accepted 25th April 2006
First published as an Advance Article on the web 17th May 2006
DOI: 10.1039/b603830a
A 5ꢀ-C-allylthymidine derivative was prepared from thymidine by the application of a stereoselective
allylation procedure and its 5ꢀ(S)-configuration was confirmed. From this nucleoside derivative,
appropriately protected building blocks were prepared and coupled using standard phosphoramidite
chemistry to afford a dinucleotide with two 5ꢀ-C-allylgroups. This molecule was used as a substrate for
a ring-closing metathesis (RCM) reaction and after deprotection, a 1 : 1 mixture of E- and Z-isomers of
a cyclic dinucleotide with an unsaturated 5ꢀ-C-to-5ꢀ-C connection was obtained. Alternatively, a
hydrogenation of the double bond and deprotection afforded a saturated cyclic dinucleotide. An
advanced NMR-examination confirmed the constitution of this molecule and indicated a restriction in
its overall conformational freedom. After variation of the protecting group strategy, a phosphoramidite
building block of the saturated cyclic dinucleotide with the 5ꢀ-O-position protected as a pixyl ether and
the phosphate protected as a methyl phosphotriester was obtained. This building block was used in the
preparation of two 14-mer oligonucleotides with a central artificial bend due to the cyclic dinucleotide
moiety. These were found to destabilise duplexes, slightly destabilise bulged duplexes but, to some
extent, stabilise a three-way junction in high Mg2+-concentrations.
conformationally restricted dinucleotide20–27 and trinucleotide
Introduction
structures.22 We formulated a general strategy (see Fig. 1) by which
The construction of chemically modified nucleic acid fragments
terminal double bonds are incorporated by various methods and
with altered conformational behaviour is a powerful tool in
on various positions in dinucleotides that are subsequently used
chemical biology and in the development of nucleic acid based
as substrates for RCM-reactions. These are envisioned to mimic,
therapeutics and diagnostics.1,2 For instance, oligonucleotides
modulate and even stabilise other secondary nucleic acid structures
containing nucleoside monomers with restricted carbohydrate
than duplexes like bulges or three-way junctions (Fig. 1).
moieties can form duplexes with significantly increased stability2–4
and are currently under intensive investigation as antisense
therapeutics.5,6 On the other hand, conformationally restricted
models of other secondary structural elements7 have gained much
less attention,8,9 though a few restricted dinucleotides have been
synthesised and investigated,10–15 e.g. as a model of the anticodon
loop in a bacterial tRNA.12
Recently, we have applied the ring-closing metathesis (RCM)
methodology16–18 using the ruthenium-based precatalysts devel-
oped by Grubbs and co-workers,19§ in the construction of
Nucleic Acid Center‡, Department of Chemistry, University of Southern
Fig. 1 General RCM-based strategy towards cyclic dinucleotides and
secondary nucleic acid structures like bulges and three-way junctions. B =
a nucleobase, R = varying protecting groups.
Denmark, 5230 Odense M, Denmark. E-mail: pon@chem.sdu.dk
† On leave from Kurukshetra University, Kurukshetra-136 119, India.
‡ Nucleic Acid Center is funded by the Danish National Research
Foundation for studies on nucleic acid chemical biology.
§ RCM and metathesis reactions in general have been recently reviewed.16–18
Grubbs’ 2nd generation catalyst (Mes = 2,4,6-trimethylphenyl, Cy =
cyclohexyl):19
A number of cyclic dinucleotides have been obtained by this
general strategy.20–29 As the first example, four stereoisomeric
dinucleotides with unsaturated 7-membered rings in the internu-
cleoside linkage were obtained from RCM-reactions on substrates
with an allyl phosphotriester linkage and an adjacent 5ꢀ-C-vinyl
group.20,21 Also cyclic dinucleotides in which the ring is based on
the combination of an allyl phosphotriester and a 5-allyluracil
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2433–2445 | 2433
©

View Article Online
moiety have been efficiently obtained giving 14-membered ring
systems.22,23,25 A trinucleotide with two allyl phosphotriester
linkages has been used as an efficient RCM-substrate and a
cyclic trinucleotide with a 13-membered ring was obtained albeit
as a mixture of 8 stereoisomers.22 Recently, a series of 2ꢀ-C to
phosphate connections revealing cyclic dinucleotides with 7–10-
membered rings have been obtained in medium to high yields.27 In
the mentioned examples, phosphotriester internucleoside linkages
were obtained revealing two problems; (1) the chirality of the
phosphorus leads to stereoisomeric products and (2) the allylic
phosphotriester moiety demonstrates a high basic lability.23 The
latter problem, however, has been significantly reduced by satu-
ration of the allylic moiety using a tandem RCM-hydrogenation
protocol30 in which Grubbs’ catalyst is active in both reactions.23,26
Hence, one example has been more comprehensively studied,
that is a cyclic dinucleotide with a 2ꢀ-C-(CH2)4O–P connection.26
The major RP isomer was found to be chemically stable and
incorporated at a central position in two 14-mer oligonucleotide
sequences. A large destabilisation of duplex structures was,
expectedly, observed, but a stabilisation of a three way-junction
with a complementary RNA-hairpin was also revealed.26
Scheme 1 Reagents and conditions: a, allylMgBr, CeCl3, THF, 47%;
b, allyltrimethylsilane, BF3:OEt2, CH2Cl2, 73% 5; c, (i) TBAF,
THF, (ii) TIPDSCl2, Pyr, 52%; d, (i) TBAF, THF, (ii) TBDM-
SCl, imidazole, DMF, 61%
3 and 2% 7; e, (i) BzCl, Pyr, 71%,
In order to obtain more chemically stable and convenient
molecules, cyclic dinucleotides with an intact, charged and achiral
phosphodiester internucleoside linkage have been approached.
Two adjacent 5-allyluracil moieties have been combined in a cyclic
dinucleotide with a very large 20-membered ring, however in this
case in a relatively low yield.25 In a more promising approach,
we investigated dinucleotide substrates on which terminal double
bonds are placed on the 4ꢀ-C or 5ꢀ-C positions. In the preliminary
study,24 a 5ꢀ-C-allyl group was found to be significantly more
reactive than an 4ꢀ-C-vinyl group, and a fully protected cyclic
dinucleotide with an unsaturated four-carbon 5ꢀ-C-to-5ꢀ-C con-
nection was obtained. In the present paper, the development of an
appropriate protecting strategy, and the successful preparation
and NMR-analysis of a saturated cyclic dinucleotide and its
incorporation into oligonucleotides is presented.
(ii) TBAF, THF, 63%; f, (i) DMT-triflate, 2,6-lutidine, CH2Cl2, (ii)
TBAF, THF, 58%; g, NC(CH2)2OP(N(iPr)2)2, 4,5-dicyanoimidazole,
CH3CN, 80%; T = thymin-1-yl, TBDMS = tert-butyldimethylsilyl,
TBDPS = tert-butyldiphenylsilyl, TIPDS = tetraisopropyldisilyl, DMT =
4,4ꢀ-dimethoxytrityl, CE = 2-cyanoethyl.
allyltrimethylsilane33 was repeated to give only a single isomer 5
in 71% overall yield. We also attempted this reaction with the
TBDMS-protected derivate 1 but in this case, a 9 : 1 epimeric
mixture of 3 and 4 was obtained in 78% yield.
In the original paper,33 5 was reported to be the 5ꢀ(S)-isomer.
This was determined by the conversion of 5 to another product
which has been alternatively obtained from a different precursor
determined by X-ray crystallography.33 In order to confirm this
important determination in a simple way, we converted the 9 :
1 mixture of 3 and 4 to the conformationally restricted 3ꢀ,5ꢀ-
disiloxane protected derivative 6 which was subsequently exam-
ined by NMR spectroscopy in an NOE-difference experiment.
Large mutual contacts between H-5ꢀ and H-4ꢀ as well as the
lack of contacts between H-5ꢀ and H-3ꢀ/H-6 confirmed the 5ꢀ(S)-
configuration of 3. This way of determining 5ꢀ-configuration has
been used before with other 5ꢀ-alkylated thymidine derivatives
where both isomers have been examined.21,31,38,40 Wang and Mid-
dleton determined the configurations of 3 and 4 by the conversion
to the two 3ꢀ,5ꢀ-O-di(t-butyl)silyl derivatives and subsequent NOE-
difference spectroscopy.32 However, direct comparison of our
NMR-data of 3 and 4 with the reported data32 was not reliable due
to intense spectral overlap. Finally, the 5ꢀ(S)-configuration of 5 was
confirmed, indirectly, by converting 5 to 3. Hence, a desilylation
and a subsequent treatment with TBDMSCl and imidazole gave
the isomer 3 as the only major product in 61% yield. This also
proved an extraordinary difference in reactivity between the two
secondary 3ꢀ- and 5ꢀ-alcohols as the 5ꢀ-silylated isomer 7 was
obtained as the minor product in only 2% yield.
Results and discussion
Chemical synthesis
Construction of 5ꢀ-C-allylthymidine building blocks. The in-
troduction of a 5ꢀ-C-allyl group via the 3ꢀ-silylated thymidine
aldehyde derivatives 1 and 2 (Scheme 1) has been demonstrated
before in the literature. Thus, a Grignard reaction with 1 using
allylMgCl and CuCN has afforded both 5ꢀ-epimers 3 and 4 as a 1 :
1 mixture in 63% yield.31,32 On the other hand, a stereoselective pro-
cedure starting with 2 applying allyltrimethylsilane and BF3:OEt2
afforded exclusively the 5ꢀ(S)-configurated 5ꢀ-O-TMS-protected
analogue of 5 in an impressive 95% yield after silylation.33 We
decided to repeat and evolve these procedures and to confirm the
configuration of the products.
Firstly, we found that the oxidation step for obtaining 1 or 2 was
most reliably and efficiently performed by the use of Dess–Martin
periodinane34,35 in both cases compared to alternative IBX,36
Swern or Moffatt oxidations.37,38 Subsequently, we attempted to
improve the Grignard reaction by the use of allylMgBr and CeCl3
In the further preparation of dinucleotides, we applied the 5ꢀ(S)-
configurated compound 5 and considered different protecting
group strategies. In our preliminary study, we prepared the
39
but we obtained a similar 1 : 1 epimeric mixture of 3 and 4 in only
45% yield. On the other hand, the stereoselective method using
2434 | Org. Biomol. Chem., 2006, 4, 2433–2445
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
benzoyl ester for the 5ꢀ-position and obtained 8 after desilylation
(Scheme 1).24 However, the benzoyl group is not orthogonal
to the cyanoethyl group that is planned as a protecting group for
the phosphodiester moiety. We decided, therefore, to investigate
the DMT-group, which might be envisioned as a permanent
protecting group useful in the eventual incorporation of the
dinucleotide into oligonucleotide sequences. Tritylation of
the secondary 5ꢀ-OH group of 5 was hereafter accomplished.
However, the use of DMT-chloride and pyridine as a standard
condition turned out to be insufficient, and the DMT-protection
was performed by the use of DMT-triflate prepared according
to a procedure by Leumann and co-workers.41 Subsequent
desilylation gave the 3ꢀ-O-deprotected derivative 9 in a reasonable
overall yield. In order to prepare a dinucleotide by standard
phosphoramidite chemistry, 5ꢀ-phosphoramidite derivative 10
was prepared from 5 in a good yield.
Construction of cyclic dinucleotides. A dinucleotide 11 was ob-
tained by the coupling of the alcohol 9 with the phosphoramidite
10 using standard coupling conditions, i.e. 1H-tetrazole as the
activator (Scheme 2). Oxidation of the intermediate phosphite to
a phosphotriester was performed by tert-butyl hydrogenperoxide
instead of the standard iodine solution in order to avoid reactions
with the double bonds. This afforded the dinucleotide 11 in a 3 :
1 mixture of phosphorous epimers. This surprising ratio might
be rationalised from the fact that 10 is a sterically hindered
phosphoramidite of a secondary alcohol. The dinucleotide 11
with two terminal double bonds was hereafter examined as a
substrate for RCM reactions. The DMT-group was found to be
completely stable when using the standard metathesis conditions,
i.e. 5–10 mol% of Grubbs’ 2nd generation catalyst§ in refluxing
dichloromethane. Only when heating the reaction mixture in other
solvents and/or under microwave conditions, was some detrityla-
tion observed. This result is in contrary to former results in our
group, where RCM reactions with DMT-protected substrates have
been impossible due to detritylation and subsequent quenching of
the catalyst.22 Hence, the cyclic dinucleotide 12 was obtained as a
mixture of four stereoisomers (ratio: ∼8 : 6 : 4 : 1) in a high yield
after a successful ring-closing reaction.
In order to obtain a saturated cyclic dinucleotide, a tandem
RCM-hydrogenation procedure30 was attempted. We have recently
obtained very successful results with other dinucleotide substrates
by the application of Grubbs’ 2nd generation catalyst§ as a
combined metathesis–hydrogenation catalyst, where the latter
hydrogenation was carried out under a 1000 psi pressure.23,25
However, when this was attempted with the present dinucleotide
substrate 11, only the detritylated, ring-closed but still unsaturated
product 13 was obtained. Standard hydrogenation of the cyclic
dinucleotide 12 was unsuccessful after attempting Pd/C or
Pd(OH)2/C at atmospheric or high pressure, thus, detritylation
was the dominating result. However, performing the detritylation
after the RCM reaction under standard acidic conditions to give
13 and, subsequently, performing a standard hydrogenation under
a high pressure afforded the dinucleotide 14 in a reasonable yield.
No starting material was left as indicated by MALDI-MS. An
attempt of performing both the RCM and the hydrogenation
after detritylation was also attempted. Hence, the deprotected
dinucleotide 15 was obtained after a standard acidic detritylation
of 11 and used as a substrate for RCM. However, this was less
Scheme 2 Reagents and conditions: a, (i) 1H-tetrazole, CH3CN, (ii)
t-BuOOH, toluene, 63%; b, Grubbs’ 2nd gen. catalyst§, CH2Cl2, 74%;
c, CF3COOH, Et3SiH, CH2Cl2, 86% 13, 80% 15; d, H2 (1000 psi),
Pd(OH)2/C, CH3OH, 63%; e, (i) TBAF, THF, (ii) 32% aq. NH3.
efficient, and a mixture of ring-closed products and unidentified
side-products was obtained.
Finally, a complete deprotection of 13, and 14, to give the cyclic
dinucleotide compounds 16 and 17, respectively, was performed
by desilylation followed by treatment with saturated ammonia
and HPLC-purification. 16 was isolated as a 1 : 1 mixture of E/Z
isomers as judged from 31P NMR. The equimolar ratio of E/Z
isomers in 16 compared with the ∼8 : 6 : 4 : 1 ratio of isomers in 12
reflects that whereas the major phosphorus epimer of 11 affords
a 6 : 8 mixture of E : Z products in the RCM-reaction, the other
epimer affords the opposite 4 : 1 mixture of E : Z products. The
saturated cyclic dinucleotide 17 was obtained as a single isomer
and subsequently examined by NMR spectroscopy and dynamic
simulations (see below).
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2433–2445 | 2435
©

View Article Online
Preparing the cyclic dinucleotide for incorporation into oligonu-
cleotides. A clear experience from the RCM-hydrogenation
protocol is that the DMT-group cannot be used as a permanent
protecting group. Furthermore, attempts to reprotect the saturated
cyclic dinucleotide 14 failed even using the more active DMT-
triflate41 as a tritylation agent. An alternative to the DMT group
is the pixyl group, which can be used in standard automated
oligonucleotide synthesis. Pixylation is usually more efficient than
DMT-protection.40 However, preliminary attempts to convert 12
into a 3ꢀ-O-phosphoramidite building block demonstrated other
difficulties in removing the TBDPS-group without affecting the
cyanoethyl-group. Therefore, a complete revision of the protect-
ing group strategy was obviously necessary in order to obtain
an efficient incorporation of 17 into oligonucleotides. Methyl
phosphotriesters have been used before for phosphate protection
in oligonucleotide synthesis being much more stable than the
corresponding cyanoethyl protected phosphotriesters,42 though
they demand a different deprotection procedure.42,43
dinucleotide was a good substrate for the RCM-reaction. However,
this proved most efficient in 1,2-dichloroethane as the solvent and
a partial depixylation was observed under the reaction conditions.
Thus, both pixylated and depixylated fractions were collected in
a combined 69% yield. These fractions were recombined for the
following hydrogenation, which proceeded very efficiently using
Adams’ catalyst under pressure leading to a saturated but, this
time, completely depixylated intermediate. Therefore, a repixy-
lation of the reaction mixture was introduced. This was again
very efficient and, hence, the pixylation has been efficient where
a DMT-protection has failed. After a straightforward desilylation
using TBAF and acetic acid showing no problems with either the
pixyl or the methyl phosphotriesters, the appropriately protected
cyclic dinucleotide 21 was obtained in 27% yield over the four steps
from 20. The use of Et3N–3HF or TBAF without acid both failed
to give a pure removal of the TBDPS-group. Phosphitylation of 21
gave the phosphoramidite 22 as the building block for introducing
the cyclic dinucleotide moiety into oligonucleotide sequences.
Two new building blocks for constructing the dinucleotide were
prepared from the key intermediate 5 (Scheme 3). Pixylation40 was
indeed remarkably more efficient than DMT-protection affording
18 in quantitative yield after desilylation. A methyl protected
phosphoramidite 19 was also efficiently obtained. The coupling
of 18 and 19 afforded the dinucleotide 20 in a good yield. This
NMR analysis and modelling
In order to further investigate the structural and dynamic prop-
erties of the cyclic dinucleotide 17, a 1000 ps molecular dynamics
simulation was performed. During the simulation, time averaged
coupling constant and distance restraints were employed. The
restraints were obtained from 1D 1H NMR and NOESY spectra.
Due to severe overlap of especially the linker protons, no restraints
were used in this part of the molecule. In order to analyze and
compare the results, a similar simulation was performed for the
analogous unmodified dinucleotide.
The molecular dynamics simulations showed that compared to
the unmodified dinucleotide, the cyclic dinucleotide is considerably
more restricted in its motions. With respect to whether the
dinucleotide is more prone to adopt a bent configuration, however,
the simulations yielded inconclusive results. A clear result from the
simulation is, that the “upper” O–C5ꢀ–C4ꢀ–C3ꢀ torsional angle of
17 remains in a trans conformation throughout the simulation,
whereas the corresponding angle of the unmodified dinucleotide
remains in the gauche+ conformation that is also usually found
3
ꢀ
ꢀ
coupling constant of 10.1 Hz
in duplexes. The very large JH4
H5
observed for 17 confirms the preferred trans conformation. The
3
ꢀ
ꢀ
simulation of 17 as well as the observed JH1
constants
H2
indicated for both nucleoside parts in the dinucleotide the same
preference for S-type conformations as normally seen for 2ꢀ-
deoxynucleosides. Fig. 2 shows an energy minimized snapshot
taken during the simulation of the cyclic dinucleotide.
Synthesis and evaluation of oligonucleotides
Two 14-mer oligonucleotide sequences each with the cyclic
dinucleotide 22 incorporated once in the middle were con-
structed. These contain either solely standard 2ꢀ-deoxynucleotides
(24) or a mixture of standard 2ꢀ-deoxynucleotides and LNA-
monomers (26), i.e. a DNA and an LNA sequence, respectively
(Table 1). Thus, LNA is defined as an oligonucleotide sequence
containing one or several of the LNA-monomers, which are
nucleosides with a bicyclic carbohydrate moiety, locked in the
N-type conformation.6,44 Hereby, the LNA-monomers secure an
overall A-type or A-type-like duplex conformation as well as a
Scheme 3 Reagents and conditions: a, (i) pixyl-Cl, pyridine, (ii) TBAF,
THF, 99% (2 steps); b, CH3OP(Cl)N(iPr)2, EtN(iPr)2, CH2Cl2, 75%; c,
(i) 1H-tetrazole, CH3CN, (ii) t-BuOOH, toluene, 75%; d, (i) Grubbs’ 2nd
gen. catalyst§ (CH2Cl)2, (ii) H2 (1000 psi), PtO2, CH3OH, (iii) pixyl-Cl,
2,6-lutidine, CH2Cl2, (iv) TBAF, AcOH, THF, 27% (4 steps); e, EtN(iPr)2,
NC(CH2)2OP(Cl)N(iPr)2, CH2Cl2, 37%. Pixyl = 9-phenylxanthen-9-yl.
2436 | Org. Biomol. Chem., 2006, 4, 2433–2445
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
tection protocol following the completed synthesis, a treatment
with disodium 2-carbamoyl-2-cyanoethylene-1,1-dithiolate43 was
introduced in order to remove the methyl protecting group of the
phosphate.42,43 After completed deprotection, the constitution and
purity of 24 and 26 were confirmed by MALDI-MS (see Table 1)
and ion-exchange HPLC.
The hybridisation properties of the oligonucleotides with
complementary DNA and RNA sequences were evaluated by
thermal stability experiments (Table 2 and 3). First, the duplexes
formed between 23–26 and their fully matching DNA and RNA
compliments were studied (Table 2). As expected, large drops in
duplex stability were observed. Comparing 23 and 24, the cyclic
dinucleotide induced a drop in Tm of 16 ◦C in a DNA–DNA
duplex, and, comparing 25 and 26, 20 ◦C in an LNA-modified du-
plex. The latter must be A-type like and is, as expected, significantly
more stable due to the LNA-monomers.6,44 Also the DNA–RNA
duplex was significantly destabilised by the cyclic dinucleotide with
a drop in Tm of 14 ◦C. These results are comparable to the results
obtained with our firstly studied cyclic dinucleotide (with a 2ꢀ-C-
(CH2)4O–P connection)26 and demonstrate the expected distortion
of a standard nucleic acid duplex structure by the introduction of
a bending cyclic dinucleotide moiety.
Fig. 2 An energy minimized snapshot of the dinucleotide 17 taken during
the simulation.
Table 1 Prepared oligodeoxynucleotide sequences and their MS-data
ODN sequencesa
MW (found/calc.)b
23
24
25
26
5ꢀ-GCTCACTTCTCCCA
—
ꢀ
=
5 -GCTCAC[T T]CTCCCA
4186.8/4188.5 [MH]+
—
5ꢀ-GCLTCLACLTTCLTCLCCLA
ꢀ
L
L
L
L
L
L
4438.9/4440.6 [MH]+
=
5 -GC TC AC [T T]C TC CC A
Next, the hybridisation properties of oligonucleotides 23–26
with bulged DNA and RNA complements were studied (Table 2,
see also Fig. 1). Thus, an additional guanosine residue was
incorporated in between the two adenosines opposite the cyclic
dinucleotide residue. The subsequent Tm measurements revealed
a smaller but still significant drop in stability of 5.2 to 11.4 ◦C
when comparing the modified sequences, 24 and 26, with their
unmodified counterparts, 23 and 25. These results are again com-
parable to the results obtained in our former study,26 however, the
destabilisation is slightly more pronounced with the present cyclic
structure. Nevertheless, the observation that the introduction of
a bulge in a duplex reduces the stability with 8–12 ◦C whereas
the same bulge in distorted duplexes containing 17 reduces the
stability by only 0.5–3 ◦C, is still valid.
a
L
=
[T T] refers to the incorporation of 22. C refers to the LNA 5-
methylcytidine monomer. b HiResESI-MS positive mode.
significantly increased duplex stability.6,44 Unmodified DNA and
LNA sequences, 23 and 25, used for comparison, are the same as
those used in our former studies.26,45
The incorporation of 22 into oligonucleotides was performed by
automated solid-phase synthesis using a standard protocol includ-
ing a prolonged coupling time for 22 of 20 minutes, a subsequent
double capping procedure and a slightly prolonged coupling time
for the proceeding standard phosphoramidite. A coupling yield
of >95% was detected for 22. The pixyl group was removed by a
standard acidic treatment after the coupling of 22. In the depro-
Table 2 Hybridisation data for the prepared oligonucleotides with DNA/RNA-complementsa
Fully matched complementsb Complements with G-bulgesc
DNA
RNA
DNA
RNA
23
24
25
26
54.4
60.4
43.1
49.8
38.1 (−16.3)
46.5 (−13.9)
36.6 (−6.5)
44.6 (−5.2)
79.7
>90
69.4
84.9
58.6 (−20.1)
80.5 (<−9.5)
58.0 (−11.4)
77.4 (−7.5)
Complements with an intrastrand stem-loopd,e
DNA
RNA
0 mM
0 mM
5 mM
10 mM
5 mM
10 mM
23
24
25
26
26.5
30.2
31.5
38.4
42.7
44.0
25.7 (−0.8)
30.0 (−0.2)
32.0 (+0.5)
38.2 (−0.2)
42.6 (−0.1)
44.5 (+0.5)
54.5
57.1
58.8
74.1
78.0
>80
51.9 (−2.6)
55.6 (−1.5)
57.5 (−1.3)
73.5 (−0.6)
77.4 (−0.6)
>80
a Melting temperatures (Tm values/◦C) obtained from the maxima of the first derivatives of the melting curves (A260 vs. temperature) recorded in a
medium salt buffer (Na2HPO4 (15 mM), NaCl (100 mM), EDTA (0.1 mM), pH 7.0) using 1.0 lM concentrations of each strand. All Tm values are given
as averages of double determinations. DTm values are given in brackets. b DNA 3ꢀ-CGAGTGAAGAGGGT, RNA 3ꢀ-CGAGUGAAGAGGGU. c DNA
3ꢀ-CGAGTGA-G-AGAGGGT, RNA 3ꢀ-CGAGUGA-G-AGAGGGU. d DNA 3ꢀ-CGAGTGA-CGCGTTTTCGCG-AGAGGGT, RNA 3ꢀ-CGAGUGA-
CGCGUUUUCGCG-AGAGGGU, bold sequences are stem-loops. e Increasing concentration of Mg2+ by addition of MgCl2.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2433–2445 | 2437
©

View Article Online
Table 3 Hybridisation data for the oligonucleotides 23 and 24 with bulged intrastrand stem-loop DNA-complementsa,b
Intrastrand stem-loop with
Intrastrand stem-loop with
Intrastrand stem-loop with
Intrastrand stem-loop with
1 × C bulge from 5ꢀ-end.c
1 × C bulge from 3ꢀ-end.d
2 × C bulge from 5ꢀ-end.e
2 × C bulge from 3ꢀ-end.f
0 mM
5 mM
10 mM 0 mM
5 mM
10 mM
0 mM
5 mM
10 mM
0 mM
5 mM
10 mM
h
h
23
24
17.0
n.d.g
20.7
n.d.g
25.0
n.d.g
26.0
23.9
30.3
27.6
32.1
29.4
n.d.g
n.d.g
—
—
—
—
26.3
22.9
31.2
27.3
33.1
28.7
h
h
(−2.1)
(−2.6)
(−2.7)
(−3.4)
(−3.9)
(−4.4)
a See Table 2, note a. b Increasing concentration of Mg2+ by addition of MgCl2. c 3ꢀ-CGAGTGA-CGCGTTTTCGCG-C-AGAGGGT. d 3ꢀ-CGAGTGA-C-
CGCGTTTTCGCG-AGAGGGT. e 3ꢀ-CGAGTGA-CGCGTTTTCGCG-CC-AGAGGGT. f 3ꢀ-CGAGTGA-CC-CGCGTTTTCGCG-AGAGGGT, bold
sequences are stem-loops. g n.d. = no well-defined transition. h — = not determined.
The stability of a three-way junction (TWJ) was also studied.
This is composed of a standard stable stem-loop sequence with
two single stranded regions being complementary to the oligonu-
cleotides 23–26 with the cyclic dinucleotide in the branching
point (see Fig. 1). First, we studied a TWJ in which no bulges
are included, i.e. all nucleobases are intended to participate in
Watson–Crick base-pairing or, alternatively, to leave two unpaired
nucleobases on opposite positions. This TWJ was in general not
very stable with a Tm of 26.5 ◦C in a complete DNA-context
The present results cannot, unfortunately, demonstrate any
significant positive influence by the introduction of the cyclic
moiety of 17 into secondary nucleic acid structures. However,
the tolerance of the cyclic structure within one of the TWJ’s
studied (DTm = +0.5 ◦C, Table 2) indicates that a TWJ being
more significantly stabilised by 17 might be found. A reason for
the present results could be that the right sequence context has
not been found. Thus, no intriguing difference was observed with
the bulged TWJ’s (Table 3) either. Also the sequence constitution
can be discussed. Only seven base pairs were used in each duplex
flanking the TWJ. A more well-defined structure might have been
obtained by longer flanking duplex sequences. On the other hand,
this was attempted by using the inherent stabilisation obtained by
using an LNA-sequence. The relative differences between modified
and unmodified TWJ’s are undoubtedly similar despite the higher
level of thermal stability.
It is well known that TWJ’s are very flexible and inhomogen
in their structure.46 Therefore, the idea of inducing some con-
formational restriction by a cyclic dinucleotide and thereby a
better defined structure was intriguing. Nevertheless, the applied
conformational restriction is not very strong. The 11-membered
ring introduced here is large and relatively flexible. Nevertheless,
the philosophy was to modulate/decrease the conformational
freedom to some extent leaving an oligonucleotide preferring a
bent to a linear organisation. Leaving the double bond unsaturated
as in 16 could be a preferable alternative concerning the degree
of conformational restriction. On the other hand, two isomers
would have to be handled which seems to be a difficult task. A
simple alteration of 17 could be to invert one or both of the 5ꢀ-
configurations. If the 5ꢀ(R)-configuration (as in compound 4) has
been used, completely different results might have been obtained.
The O–C5ꢀ–C4ꢀ–C3ꢀ torsional angle is strongly dependent on the
5ꢀ-configuration and, as confirmed by the NMR-examination,
more distorted compared to a standard B-type duplex with a
5ꢀ(S)-configuration than with a 5ꢀ(R)-configuration. There is no
doubt from the results of incorporating 17 into standard duplexes
(Table 2) that the duplexes are indeed significantly distorted.
Another important aspect of incorporating the cyclic dinucleotide
structure 17 into nucleic acid structures is its hydrophobic nature.
Thus, the butylene chain might exclude structural water and
thereby actually oppose any gain of conformational restriction.
Furthermore, the cyclic structure might force phosphates into
proximity giving a high charge density. This might explain why
the positive influence on stability of adding Mg2+ ions is relatively
higher for the modified compared to the unmodified TWJ’s.
◦
and 38.4 C as a DNA–RNA hybrid, though significantly more
stable with six LNA-monomers (Table 2). A 5 mM concentration
of Mg2+ was in general found to increase the thermal stability
by 3–5 ◦C, and a smaller further stabilisation with a 10 mM
Mg2+ concentration of approx. 2 ◦C was observed. The TWJ
was in general found to be unaffected or slightly destabilised by
the cyclic dinucleotide concerning the overall thermal stability.
Thus, small drops in Tm of 0.6–2.6 ◦C were seen when comparing
25 with 26, i.e. an LNA-sequence context. On the other hand,
even smaller differences were observed when comparing 23 and
24 tending to a small increase in stability of 0.5 ◦C with 24 when
increasing the Mg2+ concentration. For the unmodified sequences,
some variations in the Tm’s compared to our former measurements
of the same TWJ26 were seen. This is due to relatively broad
transitions reflecting a not very well-defined structure of the
TWJ.
In order to study a more stable or at least more well-
defined structure, we incorporated an additional bulge into the
TWJ. Hence, one or two additional cytosines were inserted into
the target DNA-sequence on either side of the stem-loop (see
Table 3). Thus, the optimal constitution of natural TWJ’s seems
to be involving two costacking bases in a bulge at one of the three
strands.46,47 In our first investigation, only DNA-compliments were
prepared. As evident from the Tm’s, the unmodified TWJ’s with
one or two cytosines in a bulge from the 3ꢀ-end of the target
stem-loop display the same stability as the first TWJ (Tm
≈
26 ◦C). Indeed, more sharp transitions and thereby more well-
defined structures were observed. The cyclic dinucleotide, on the
other hand, destabilises this in both cases with the drop in Tm
◦
of 2.1–4.4 C (comparing 23 and 24) increasing with the Mg2+-
concentration. The TWJ’s with the cytosine(s) in a bulge from the
5ꢀ-end of the target stem-loop are significantly less stable, and with
the cyclic dinucleotide incorporated, no well-defined transitions
could be observed. Due to the general destabilisation induced by
the cyclic dinucleotide, no further studies with RNA-compliments
nor with the LNA-sequences 25 and 26 were undertaken.
2438 | Org. Biomol. Chem., 2006, 4, 2433–2445
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Two other modifications have been introduced in the same
sequence context; (1) another cyclic dinucleotide (with the 2ꢀ-
C-(CH2)4O-P connection in a 9-membered ring)26 and (2) a
nucleoside with an additional base (2ꢀ-deoxy-2ꢀ-C-(2-(thymin-1-
yl)ethyl)uridine)45 incorporated at the same position as 17 (via 22)
in sequence 24 and 26. The former demonstrated results which
are more or less comparable to the data presented herein for 17.
The destabilisation of duplexes and bulged duplexes was slightly
more pronounced in the case of 17, and the stabilisation of the
TWJ with no C-bulges that was actually seen in our former
study26 was less pronounced with 17. In the other approach, in
which the stabilisation of TWJ’s was attempted by an additional
nucleobase instead of a large ring and with sequences thereby being
one nucleoside shorter in the branching point, more promising
results have been obtained.45 Here, small stabilisations of the
TWJ’s were in general seen in the DNA–DNA and DNA–
RNA contexts compared to either two thymidines (as in 23) or
one in the branching point.45 Furthermore, much more relative
stabilisation of the TWJ in an LNA–DNA context was gained by
the addition of Mg2+ indicating a more dynamic structural context.
Thus, the introduction of an additional binding potential in the
oligonucleotide (an additional nucleobase45 or an intercalating
moiety48) seems to be a more powerful approach for targeting
a stem-loop and to stabilise a TWJ than the introduction of a
large cyclic moiety bending the oligonucleotide. The only way of
improving the latter approach seems to be a target oriented design
of the bend needed for the specific TWJ.
assigned C-1ꢀ. However, compound names for bicyclic compounds
are given according to the von Baeyer nomenclature. Fast-atom
bombardment mass spectra (FAB-MS) were recorded in positive
ion mode on a Kratos MS50TC spectrometer and MALDI mass
spectra as well as accurate mass determinations were performed
on an Ionspec Ultima Fourier Transform mass spectrometer. Mi-
croanalyses were performed at The Microanalytical Laboratory,
Department of Chemistry, University of Copenhagen.
Preparation of 5ꢀ(S)-C-allyl-3ꢀ-O-(tert-
butyldimethylsilyl)thymidine (3) and 5ꢀ(R)-C-allyl-3ꢀ-O-(tert-
butyldimethylsilyl)thymidine (4)
The aldehyde 1 (466 mg, 1.31 mmol) was coevaporated with
anhydrous CH2Cl2 and dissol◦ved in the same solvent (10 cm3).
The solution was stirred at 0 C. Allyltrimethylsilane (1.05 cm3,
6.58 mmol) and BF3:OEt2 (0.820 cm3, 6.67 mmol) were added,
whereby the reaction mixture went from weak yellow to orange-
red. The reaction mixture was stirred for 1 h at 0 ◦C and then
diluted with CH2Cl2 (60 cm3). The solution was washed with a
saturated aqueous solution of NaHCO3 (2 × 25 cm3) and brine
(25 cm3). The organic phase was dried (MgSO4) and the solvent
was removed under reduced pressure. The residue was purified
by silica gel column chromatography using EtOAc and petroleum
ether (1 : 1 v/v) as the eluent to give the product (407 mg, 78%)
as a white foam and a 9 : 1 mixture of the 5ꢀ-epimers 3 and 4,
respectively; Rf 0.46 (EtOAc and petroleum ether, 3 : 1 v/v); dH
(300 MHz; CDCl3; Me4Si) (major isomer 3) 8.73 (1H, br s, NH),
7.52 (1H, d, J 1.0 Hz, H-6), 6.18 (1H, t, J 6.8 Hz, H-1ꢀ), 5.84
(1H, m, H-7ꢀ), 5.23–5.15 (2H, m, H-8ꢀ), 4.50 (1H, m, H-3ꢀ), 3.82
(1H, m, H-5ꢀ), 3.76 (1H, m, H-4ꢀ), 2.43–2.28 (3H, m, H-2ꢀ, H-6ꢀ),
2.17 (1H, ddd, J 13.3, 6.2, 3.4 Hz, H-2ꢀ), 1.92 (3H, d, J 1.0 Hz,
CH3), 0.90 (9H, s, C(CH3)3), 0.08 (6H, s, Si(CH3)2). dC (75 MHz;
CDCl3; Me4Si) (major isomer 3) 163.7 (C-4), 150.3 (C-2), 137.2
(C-7ꢀ), 134.1 (C-6), 118.8 (C-8ꢀ), 110.9 (C-5), 88.8, 86.8 (C-1ꢀ, C-4ꢀ),
73.0, 70.1 (C-3ꢀ, C-5ꢀ), 40.3, 39.0 (C-6ꢀ, C-2ꢀ), 25.7 (C(CH3)3), 17.9
(C(CH3)3), 12.6 (CH3), −4.6, −4.8 (Si(CH3)2); MS(FAB) m/z 397
(M + H+).
Conclusion
In summary, our RCM-based strategy for introducing large
internucleotide connections in dinucleotides has proved efficient
in the construction of a cyclic dinucleotide with a 5ꢀ-C-(CH2)4-
5ꢀ-C connection forming an 11-membered ring. Thus both
the deprotected cyclic dinucleotide 17 and its corresponding
phosphoramidite building block 22 were isolated in reasonable
overall yields. In the right sequence context, oligonucleotides
containing 17 might stabilise secondary nucleic acid structures. In
general, the present RCM-based strategy can with the appropriate
effort in modelling and design be a general tool towards the
development of oligonucleotide structures that are preorganised
for recognising RNA secondary structures. New nucleic acid based
molecular structures based on this concept can reveal a new impact
for modulating RNA function for future therapeutic uses and
nanobiotechnology.
Preparation of 5ꢀ(S)-C-allyl-3ꢀ-O-(tert-
butyldiphenylsilyl)thymidine (5)
The aldehyde 2 (6.25 g, 13.1 mmol) was dissolved in anhydrous
CH2Cl2 (150 cm3) and stirred at 0 ◦C. Allyltrimethylsilane
(10.4 cm3, 65.4 mmol) and BF3:OEt2 (8.3 cm3, 65.4 mmol) were
added, whereby the reaction mixture went from weak yellow to
orange-red. The reaction mixture was stirred for 5 h at 0 ◦C and
then quenched by the dropwise addition of a saturated aqueous
solution of NaHCO3 (60 cm3). The layers were separated and the
organic phase was dried (MgSO4). The solvent was removed under
reduced pressure, and the residue was purified by silica gel column
chromatography using EtOAc and petroleum ether (1 : 2 v/v) as
the eluent to give the product (5.00 g, 73%) as a white foam; Rf
0.55 (EtOAc and petroleum ether, 3 : 1 v/v); mp 83–84 ◦C. dH
(300 MHz; CDCl3; Me4Si) 8.75 (1H, br s, NH), 7.69–7.36 (11H,
m, H-6, Ph), 6.30 (1H, dd, J 6.1, 8.1 Hz, H-1ꢀ), 5.63 (1H, m, H-7ꢀ),
5.11–5.02 (2H, m, H-8ꢀ), 4.48 (1H, m, H-3ꢀ), 3.83 (1H, br s, H-4ꢀ),
3.11 (1H, m, H-5ꢀ), 2.30–2.11 (4H, m, H-2ꢀ, H-6ꢀ), 1.86 (3H, d, J
1.1 Hz, CH3), 1.09 (9H, s, C(CH3)3). dC (75 MHz; CDCl3; Me4Si)
Experimental
All reagents were used as supplied. When necessary, all reactions
were performed under an atmosphere of nitrogen. Column
chromatography was carried out on glass columns using Silica gel
60 (0.040–0.063 mm). NMR spectra were obtained on a Varian
1
Gemini 2000 spectrometer. H NMR spectra were recorded at
300 MHz, 13C NMR spectra were recorded at 75,5 MHz and 31
P
NMR spectra were recorded at 121.5 MHz. Values for d are in ppm
relative to tetramethylsilane as an internal standard or 85% H3PO4
as an external standard. Assignments of NMR-signals when given
are based on 2D spectra and follow standard carbohydrate and
nucleoside style; i.e. the carbon atom next to a nucleobase is
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2433–2445 | 2439
©

View Article Online
163.7 (C-4), 150.3 (C-2), 137.1 (C-7ꢀ), 135.8, 135.7 (Ph), 134.1 (C-
6), 133.4, 133.1, 130.1, 130.0, 127.9 (Ph), 118.5 (C-8ꢀ), 110.8 (C-5),
88.9 (C-4ꢀ), 86.8 (C-1ꢀ), 74.6 (C-3ꢀ), 70.0 (C-5ꢀ), 39.9, 38.9 (C-6ꢀ,
C-2ꢀ), 26.9 (C(CH3)3), 19.0 (C(CH3)3), 12.5 (CH3). MS(MALDI)
m/z 543 (M + Na+), HRMS 543.2288 (calcd for C29H36N2O5SiNa
543.2286).
6.5, 3.4 Hz, H-2ꢀ), 1.92 (3H, s, CH3); MS(MALDI) m/z 305 (M +
Na+). This diol (202 mg, 0.715 mmol) was dissolved in anhydrous
DMF (5 cm3), imidazole (227 mg, 3.33 mmol) and TBDMSCl
(344 mg, 2.28 mmol) were added. The mixture was stirred at room
temperature for 4 h, then EtOAc (20 cm3) was added and washed
with brine (2 × 20 cm3). The organic phase was dried (Na2SO4)
and the solvent was removed under reduced pressure. The residue
was purified by silica gel column chromatography using MeOH
and CH2Cl2 (1 : 50 v/v) as the eluent to give the products as white
solids. 7 (4.9 mg, 1.7%); Rf 0.82 (MeOH and CH2Cl2, 1 : 14 v/v);
dH (300 MHz; DMSO-d6; Me4Si) 11.34 (1H, br s, NH), 7.57 (1H,
br s, H-6), 6.18 (1H, m, H-1ꢀ), 5.81 (1H, m, H-7ꢀ), 5.25 (1H, m,
3ꢀ-OH), 5.16–5.05 (2H, m, H-8ꢀ), 4.18 (1H, m, H-3ꢀ), 3.89 (1H, m,
H-5ꢀ), 3.78 (1H, m, H-4ꢀ), 2.50 (1H, m, H-6ꢀ), 2.40 (1H, m, H-6ꢀ),
2.25 (1H, m, H-2ꢀ), 2.05 (1H, m, H-2ꢀ), 1.77 (3H, d, J 1.1 Hz, CH3),
0.87 (9H, m, C(CH3)3), 0.12 (3H, s, SiCH3), 0.09 (3H, s, SiCH3).
3 (188 mg, 66%); Rf 0.55 (MeOH and CH2Cl2, 1 : 14 v/v); dH
(300 MHz; DMSO-d6; Me4Si) 11.30 (1H, br s, NH), 7.87 (1H, s,
H-6), 6.17 (1H, dd, J 8.0, 6.5 Hz, H-1ꢀ), 5.84 (1H, m, H-7ꢀ), 5.21
(1H, d, J 5.2 Hz, 5ꢀ-OH), 5.11–5.05 (2H, m, H-8ꢀ), 4.44 (1H, m,
H-3ꢀ), 3.69 (1H, br s, H-4ꢀ), 3.65 (1H, m, H-5ꢀ), 2.29–2.19 (2H, m,
H-6ꢀ), 2.15 (1H, m, H-2ꢀ), 2.05 (1H, ddd, J 13.7, 6.5, 2.9 Hz, H-2ꢀ),
1.77 (3H, s, CH3), 0.87 (9H, m, C(CH3)3), 0.08 (6H, s, SiCH3);
dC (75 MHz; CDCl3; Me4Si) 163.6 (C-4), 150.3 (C-2), 137.2 (C-
7ꢀ), 134.1 (C-6), 118.8 (C-8ꢀ), 110.9 (C-5), 88.9, 86.9 (C-1ꢀ, C-4ꢀ),
73.0, 70.1 (C-3ꢀ, C-5ꢀ), 40.3, 39.1 (C-6ꢀ, C-2ꢀ), 25.7 (C(CH3)3), 18.0
(C(CH3)3), 12.6 (CH3), −4.6, −4.8 (Si(CH3)2); MS(MALDI) m/z
419 (M + Na+).
Preparation of 5ꢀ(S)-C-allyl-3ꢀ,5ꢀ-O-(1,1,3,3-tetraisopropyl-1,3-
disiloxandiyl)thymidine (6)
Compounds 3 and 4 (9 : 1 mixture) (66.0 mg; 0.167 mmol) were
dissolved in anhydrous THF (3.0 cm3) and a 1 M solution of
TBAF in THF (0.360 cm3) was added. The mixture was stirred for
2.5 h at room temperature and then concentrated under reduced
pressure. The residue was dissolved in EtOAc (20 cm3) and washed
with a saturated aqueous solution of NaHCO3 (2 × 10 cm3) and
brine (10 cm3). The combined aqueous phases were extracted
with more EtOAc (3 × 10 cm3). The combined organic phases
were dried (MgSO4) and the solvent was removed under reduced
pressure. The residue was coevaporated with anhydrous pyridine
and redissolved in the same solvent (1.0 cm3). The solution was
stirred at 0 ◦C and 1,3-dichloro-1,1,3,3-tetraisopropyldisiloxane
(0.111 cm3, 0.347 mmol) was added. The solution was stirred
at room temperature for 18 h and concentrated under reduced
pressure. The residue was dissolved in CH2Cl2 (30 cm3) and washed
with water (15 cm3), a saturated aqueous solution of NaHCO3
(2 × 15 cm3) and brine (15 cm3). The organic phase was dried
(MgSO4) and the solvent was removed under reduced pressure.
The residue was purified by silica gel column chromatography
using MeOH and CH2Cl2 (1 : 50 v/v) as the eluent to give the
product (91 mg, ∼52% calculated from an estimated content of
hydrolysed disiloxane reagent) as an oil; Rf 0.56 (MeOH and
CH2Cl2, 1:9 v/v); dH (300 MHz; CDCl3; Me4Si) 8.91 (1H, br s,
NH), 7.49 (1H, d, J 1.1 Hz, H-6), 6.06 (1H, dd, J 7.4, 1.6 Hz, H-
1ꢀ), 5.84 (1H, m, H-7ꢀ), 5.23–5.04 (2H, m, H-8ꢀ), 4.42 (1H, m, H-3ꢀ),
4.04 (1H, m, H-5ꢀ), 3.67 (1H, dd, J 8.4, 2.8 Hz, H-4ꢀ), 2.39–2.62
(3H, m, H-2ꢀ, H-6ꢀ), 2.21 (1H, ddd, J 13.3, 7.4, 1.6 Hz, H-2ꢀ), 1.92
(3H, d, J 1.1 Hz, CH3), 1.17–0.90 (28H, m, SiCH, CH(CH3)2).
dC (75 MHz; CDCl3; Me4Si) 163.9 (C-4), 150.1 (C-2), 135.2 (C-
7ꢀ), 133.7 (C-6), 118.1 (C-8ꢀ), 110.3 (C-5), 85.1, 83.6 (C-1ꢀ, C-4ꢀ),
68.4, 67.3 (C-3ꢀ, C-5ꢀ), 39.8, 38.1 (C-6ꢀ, C-2ꢀ), 17.6, 17.5, 17.3,
17.2, 17.1, 17.0, 16.9, 16.9 (CH(CH3)2), 13.5, 13.1, 12.8, 12.6, 12.5
(CH(CH3)2, CH3); MS(FAB) m/z 525 (M + H+).
Preparation of 5ꢀ(S)-C-allyl-5ꢀ-O-benzoylthymidine (8)
Compound 5 (200 mg, 0.384 mmol) was dissolved in anhydrous
pyridine (4 cm3) and added benzoylchloride (0.09 cm3, 0.77 mmol).
The reaction mixture was stirred at room temperature for 8 h and
then quenched by the addition of water (0.5 cm3). The mixture was
diluted with EtOAc (20 cm3) and washed with a saturated aqueous
solution of NaHCO3 (2 × 10 cm3) and water (10 cm3), dried
(MgSO4) and concentrated under reduced pressure. The residue
was purified by column chromatography using MeOH and CH2Cl2
(1 : 19 v/v) as the eluent to give 5ꢀ(S)-C-allyl-5ꢀ-O-benzoyl-3ꢀ-O-
(tert-butyldiphenylsilyl)thymidine (175 mg, 71%) as a white solid;
found C 69.30, H 6.41, N 4.57, C36H40N2O6Si requires C 69.20,
H 6.45, N 4.48; mp 75–76 ◦C; dH (300 MHz; CDCl3; Me4Si) 1.09
(9H, s, C(CH3)3), 1.73 (1H, m, H-2ꢀ), 1.84 (3H, d, J 1.1, CH3),
2.43–2.21 (3H, m, H-6ꢀ, H-2ꢀ), 4.07 (1H, m, H-4ꢀ), 4.33 (1H, m,
H-3ꢀ), 4.79 (1H, m, H-5ꢀ), 5.09–5.00 (2H, m, H-8ꢀ), 5.68 (1H, m,
H-7ꢀ), 5.98 (1H, dd, J 5.6, 8.4, H-1ꢀ), 7.34–7.48 (9H, m, H-6,
Ph), 7.69–7.53 (5H, m, Ph), 7.90–7.85 (2H, m, Ph), 8.83 (1H, br s,
NH); dC (75 MHz; CDCl3; Me4Si) 12.4 (CH3), 19.0 (C(CH3)3), 26.9
(C(CH3)3), 35.8 (C-6ꢀ), 40.7 (C-2ꢀ), 72.6 (C-5ꢀ), 73.8 (C-3ꢀ), 84.7 (C-
1ꢀ), 87.2 (C-4ꢀ), 111.1 (C-5), 118.8 (C-8ꢀ), 127.9, 128.0, 128.5, 129.3,
130.1 (Ph), 132.4, 132.8, 133.3 (Ph, C-7ꢀ), 135.0, 135.0, 135.6, 135.7
(C-6, Ph), 150.2 (C-4), 163.7 (C-2), 165.6 (PhCO); MS(MALDI)
m/z 647 (M + Na+), HRMS 647.2559 (calcd for C36H40N2O6SiNa
647.2548). This compound (295 mg, 0.460 mmol) was dissolved in
anhydrous THF (15 cm3) and TBAF (1.0 M in THF, 0.46 cm3) was
added. The reaction mixture was stirred at room temperature for
45 min and then a suspension of DOWEX in pyridine, water and
MeOH (3 : 1 : 1 v/v/v, 10 cm3) was added. The mixture was stirred
Preparation of 5ꢀ(S)-C-allyl-3ꢀ-O-(tert-
butyldimethylsilyl)thymidine (3) and
5ꢀ(S)-C-allyl-5ꢀ-O-(tert-butyldimethylsilyl)thymidine (7)
Compound 5 (402 mg, 0.77 mmol) was dissolved in anhydrous
THF (12 cm3) and added a 1 M solution of TBAF in THF
(1.25 cm3). The mixture was stirred for 80 min at room temperature
and then concentrated under reduced pressure. The residue was
purified by silica gel column chromatography using MeOH and
CH2Cl2 (1 : 20 v/v) as eluent to give 5ꢀ(S)-C-allylthymidine
(172 mg, 92%) as a white solid; Rf 0.64 (MeOH and CH2Cl2,
1 : 9 v/v); dH (300 MHz; CDCl3; Me4Si) 7.56 (1H, br s, H-6), 6.24
(1H, dd, J 7.3, 6.5 Hz, H-1ꢀ), 5.85 (1H, m, H-7ꢀ), 5.23–5.16 (2H, m,
H-8ꢀ), 4.58 (1H, m, H-3ꢀ), 3.91 (1H, dd, J 3.0, 2.3 Hz, H-4ꢀ), 3.85
(1H, m, H-5ꢀ), 2.44–2.34 (3H, m, H-2ꢀ, H-6ꢀ), 2.29 (1H, ddd, J 13.7,
2440 | Org. Biomol. Chem., 2006, 4, 2433–2445
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
for 8 h and then filtered. The filter was rinsed with pyridine, water
and MeOH (3 : 1 : 1 v/v/v, 2 cm3), and the combined filtrates were
concentrated under reduced pressure. The residue was purified by
column chromatography using EtOAc and petroleum ether (1 : 1
v/v) as the eluent to give the product (117 mg, 63%) as a white
solid; found C 62.19, H 5.81, N 7.32, C20H22N2O6 requires C 62.17,
H 5.74, N 7.25; mp 167–168 ◦C; dh (300 MHz; DMSO-d6; Me4Si)
1.75 (3H, s, CH3), 2.21–2.04 (2H, m, H-2ꢀ), 2.65–2.46 (2H, m,
H-6ꢀ), 3.95 (1H, m, H-4ꢀ), 4.30 (1H, m, H-3ꢀ), 5.03–5.19 (2H, m,
H-8ꢀ), 5.35 (1H, m, H-5ꢀ), 5.47 (1H, d, J 4.7, OH), 5.83 (1H, m,
H-7ꢀ), 6.20 (1H, t, J 6.7, H-1ꢀ), 7.43 (1H, s, H-6), 7.60–7.50 (2H,
m, Ph), 7.69 (1H, m, Ph), 7.92–8.06 (2H, m, Ph), 11.34 (1H, br s,
NH); dc (75 MHz; DMSO-d6; Me4Si) 12.1 (CH3), 35.5 (C-6ꢀ), 39.3
(C-2ꢀ), 70.4 (C-3ꢀ), 72.8 (C-5ꢀ), 83.7 (C-1ꢀ), 86.3 (C-4ꢀ), 109.7 (C-5),
118.3 (C-8ꢀ), 128.8, 129.1, 129.6 (Ph), 133.5, 133.5 (C-7ꢀ, Ph), 135.4
(C-6), 150.3 (C-4), 163.6 (C-2), 165.4 (PhCO); MS(MALDI) m/z
409 (M + Na+).
Preparation of 3ꢀ-O-(tert-butyldiphenylsilyl)-5ꢀ-
(cyanoethoxy(diisopropylamino)phosphinyl)-5ꢀ(S)-C-
allylthymidine (10)
Compound 5 (1.95 g, 3.74 mmol) was coevaporated with
anhydrous CH3CN (3 × 40 cm3) and redissolved in an-
hydrous CH2Cl2 (50 cm3). To the mixture was added a
1.0 M solution of 4,5-dicyanoimidazole in CH3CN (2.5 cm3,
2.5 mmol) and 2-cyanoethyl N,N,Nꢀ,Nꢀ-tetraisopropyl phospho-
rdiamidite (1.3 cm3, 3.97 mmol). The reaction mixture was
stirred at room temperature for 4 h. An additional amount
of 2-cyanoethyl N,N,Nꢀ,Nꢀ-tetraisopropyl phosphordiamidite
(0.5 cm3, 1.52 mmol) was added and the reaction mixture
was stirred for another 18 h. The solution was diluted with
CH2Cl2 (200 cm3), washed with a saturated aqueous solution of
NaHCO3 (2 × 50 cm3) and brine (50 cm3). The combined organic
phases were dried (MgSO4) and the solvent was removed under
reduced pressure. The residue was purified by silica gel column
chromatography using EtOAc and petroleum ether (1 : 1 v/v) as
eluent to give the product (2.2 g, 80%) as a white foam; Rf 0.59
(EtOAc and petroleum ether, 3 : 1 v/v). dH (300 MHz; CDCl3;
Me4Si) 8.57 (m, NH), 7.69–7.56 (m, H-6ꢀ, Ph), 7.49–7.35 (m, Ph),
6.53–6.48 (m, H-1ꢀ), 5.63–5.51 (m, H-7ꢀ), 5.10–5.03 (m, H-8ꢀ), 4.46
(d, J 5.1 Hz, H-3ꢀ), 4.36 (d, J 5.1 Hz, H-3ꢀ), 4.01–3.97 (m, H-
4ꢀ), 3.80–3.58 (m, CH2OP), 3.53–3.23 (m, H-5ꢀ, HC(CH3)2), 2.58–
2.52 (m, CH2CN), 2.39–2.24 (m, H-2ꢀ, H-6ꢀ), 2.13–2.08 (m, H-6ꢀ),
1.98–1.82 (m, H-2ꢀ), 1.88 (s, CH3), 1.87 (s, CH3), 1.14–0.90 (m,
C(CH3)3, HC(CH3)3). dp (CDCl3) 151.14, 151.05. MS(MALDI)
m/z 743 (M + Na+), HRMS 743.3334 (calcd for C38H53N4O6PSiNa
743.3364).
Preparation of 5ꢀ(S)-C-allyl-5ꢀ-O-(4,4ꢀ-dimethoxytrityl)thymidine
(9)
Compound 5 (293 mg, 0.562 mmol) was dissolved in a mixture of
anhydrous 2,6-lutidine and anhydrous CH2Cl2 (1 : 1 v/v, 4 cm3).
DMT-triflate (532 mg, 1.18 mmol) was added and the reaction
mixture was stirred at room temperature. Additional amounts of
DMT-triflate were added after 18 h and 24 h (174 mg, 0.39 mmol
and 261 mg, 0.58 mmol, respectively). After 48 h the solution was
washed with a saturated aqueous solution of NaHCO3 (23 cm3).
The aqueous layer was extracted with CH2Cl2 (15 cm3), the
combined organic extracts were dried (MgSO4) and the solvent was
removed under reduced pressure. The residue was coevaporated
with toluene and redissolved in anhydrous THF (22.5 cm3) and
TBAF (1.0 M in THF, 1 cm3) was added. The reaction mixture
was stirred at room temperature for 3 h. An additional amount of
TBAF (1.0 M in THF, 1 cm3) was added and the reaction mixture
was stirred for another 2 h. Dowex 50 W × 2 in pyridine, MeOH
and water (3 : 1 : 1 v/v, 8 cm3) was added and the solution was
stirred at room temperature overnight. The resin was filtered off
and the filtercake was washed with pyridine, MeOH and water (3 :
1 : 1 v/v, 8 cm3). The filtrate was diluted with CH2Cl2 (15 cm3). The
combined organic phases were washed with water (7.5 cm3), dried
(MgSO4) and the solvent was removed under reduced pressure.
The residue was purified by silica gel column chromatography
using EtOAc and petroleum ether (3 : 4 v/v) and then MeOH and
CH2Cl2 (1 : 49 v/v) as eluents to give the product (190 mg, 58%)
as a beig◦e foam; Rf 0.30 (MeOH and CH2Cl2, 1 : 19 v/v); mp
118–120 C. dH (300 MHz; CDCl3; Me4Si) 8.79 (1H, br s, NH),
7.86 (1H, d, J 1.1 Hz, H-6), 7.46–7.24 (9H, m, Ph), 6.85–6.81 (4H,
m, Ph), 6.32 (1H, t, J 7.2 Hz, H-1ꢀ), 5.41 (1H, m, H-7ꢀ), 4.97–4.86
(2H, m, H-8ꢀ), 4.38 (1H, m, H-3ꢀ), 3.82 (1H, m, H-4ꢀ), 3.79 (6H, s,
OCH3), 3.41 (1H, m, H-5ꢀ), 2.45–2.19 (4H, m, H-2ꢀ, H-6ꢀ), 2.01
(1H, d, J 4.2 Hz, OH), 1.83 (3H, s, CH3). dC (75 MHz; CDCl3;
Me4Si) 163.7 (C-4), 158.7 (Ph), 150.2 (C-2), 145.8, 136.4, 136.1,
135.6, 133.9, 130.3, 130.2, 128.1, 127.8, 127.1 (C-6, C-7ꢀ, Ph), 117.8
(C-8ꢀ), 113.1 (Ph), 111.2 (C-5ꢀ), 87.3 (C(Ar)3), 86.7 (C-4ꢀ), 84.1 (C-
1ꢀ), 73.6 (C-5ꢀ), 71.6 (C-3ꢀ), 55.2 (OCH3), 41.0 (C-2ꢀ), 36.3 (C-6ꢀ),
12.4 (CH3). MS(MALDI) m/z 607 (M + Na+), HRMS 607.2425
(calcd for C34H36N2O7Na 607.2415).
Preparation of [5ꢀ-O-(5ꢀ(S)-C-allyl-3ꢀ-O-(tert-
butyldiphenylsilyl)thymidinyl]-[3ꢀ-O-(5ꢀ(S)-C-allyl-5ꢀ-O-(4,4ꢀ-
dimethoxytrityl)thymidinyl] cyanoethylphosphate (11)
The alcohol 9 (548 mg, 0.937 mmol) and the phosphoramidite 10
(1.288 g, 1.787 mmol) were mixed, coevaporated with a mixture of
anhydrous CH3CN and anhydrous CH2Cl2 (2 : 1 v/v, 3 × 30 cm3)
and redissolved in the same mixture (45 cm3). The solution was
added a 0.45 M solution of 1H-tetrazole in CH3CN (9.0 cm3,
4.05 mmol) and stirred at room temperature for 3 h. The reaction
mixture was cooled to 0 ◦C, added a 3.0 M solution of t-BuOOH
in toluene (1.5 cm3, 3.0 mmol) and stirred at room temperature
for 12 h. The solution was diluted with CH2Cl2 (100 cm3), washed
with a saturated solution of NaHCO3 (2 × 50 cm3) and brine
(50 cm3). The combined organic phases were dried (MgSO4) and
the solvent was removed under reduced pressure. The residue was
purified by silica gel column chromatography using MeOH and
CH2Cl2 (1 : 99–1 : 14 v/v) to give the product (716 mg, 63%) as
a white foam; Rf 0.27 (MeOH and CH2Cl2, 1 : 14 v/v); found C
64.38, H 6.08, N 5.62, C66H74O14N5PSi·H2O requires C 64.00, H
6.18, N 5.65. dH (300 MHz; CDCl3; Me4Si) 9.05 (br s, NH), 8.86
(br s, NH), 8.72 (br s, NH), 7.91 (s, H-6), 7.80 (s, H-6), 7.65–7.61
(m, Ph), 7.49–7.21 (m, H-6, Ph), 6.86–6.80 (m, Ph), 6.52 (dd, J 5.0,
9.0 Hz, H-1ꢀ), 6.47 (dd, J 5.1, 9.0 Hz, H-1ꢀ), 6.32 (dd, J 5.2, 9.0 Hz,
H-1ꢀ), 6.20 (t, J 7.2 Hz, H-1ꢀ), 5.58–5.19 (m, H-7ꢀ), 5.11–4.82 (m,
H-8ꢀ), 4.79–4.74 (m, T1-H-3ꢀ), 4.32–4.28 (m, T2–H-3ꢀ), 4.05–3.73
(m, H-4ꢀ, T2–H-5ꢀ, OCH2, OCH3), 3.33–3.29 (m, T1-H-5ꢀ), 2.42–
1.98 (m, H-2ꢀ, H-6ꢀ, CH2CN), 1.89–1.76 (m, H-2ꢀ, CH3), 1.14–1.00
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2433–2445 | 2441
©

View Article Online
(m, C(CH3)3). dp (CDCl3) −1.43, −1.71 (1 : 3). MS(MALDI) m/z
1242 (M + Na+), HRMS 1242.4612 (calcd for C66H74N5O14PSiNa
1242.4631).
Preparation of (3R/S)-(1S,5S,10S,11R,13R)-3-(2-cyanoethoxy)-
10-hydroxy-3-oxo-5-(3(S)-(tert-butyldiphenylsilyl)oxy-5(R)-
(thymin-1-yl)tetrahydrofuran-2(R)-yl)-13-(thymin-1-yl)-3-
phospha-2,4,12-trioxabicyclo[9.3.0]tetradecane (14)
Compound 13 (119 mg, 0.134 mmol) was dissolved in MeOH
(10 cm3) and Pd(OH)2/C (23.2 mg, 0.033 mmol) in MeOH (3 cm3)
was added. The reaction mixture was bubbled with H2 for 5 min.
and another 2 cm3 of MeOH was added. The reaction mixture was
placed in an autoclave at 1000 psi H2 at 55 ◦C for 16 h. The mixture
was filtered through celite and after washing the celite with MeOH
(50 cm3), the solvent was removed under reduced pressure. The
residue was purified by silica gel column chromatography using
MeOH and CH2Cl2 (1 : 14 v/v) to give the product (75 mg, 63%)
as a white foam; Rf 0.37, 0.41 (MeOH and CH2Cl2, 1 : 9 v/v);
dH (300 MHz; CDCl3; Me4Si) 9.37 (br s, NH), 9.32 (br s, NH),
9.14 (br s, NH), 8.98 (br s, NH), 7.69–7.63 (m, Ph), 7.50–7.40 (m,
H-6, Ph), 7.22 (br s, H-6), 7.07 (br s, H-6), 6.49 (dd, J 5.4, 8.5 Hz,
H-1ꢀ), 6.36 (t, J 6.9 Hz, H-1ꢀ), 6.09–6.02 (m, H-1ꢀ), 5.03 (T1-H-3ꢀ),
4.76 (T1-H-3ꢀ), 4.47 (T2-H-3ꢀ), 4.28–3.68 (m, T2–H-3ꢀ, H-4ꢀ, H-5ꢀ,
CH2O), 2.93 (br s, OH), 2.72–2.47 (m, H-2ꢀ, CH2OP), 2.35–2.28
(m, H-2ꢀ), 1.96–0.99 (m, H-2ꢀ, CH3, (CH2)4, C(CH3)3). dp (CDCl3)
−0.23, −4.79 (1 : 2.4). MS(MALDI) m/z 914 (M + Na+), HRMS
914.3130 (calcd for C43H54N5O12PSiNa 914.3168).
Preparation of (E,Z)-(3R/S)-(1S,5S,10S,11R,13R)-3-(2-
cyanoethoxy)-10-(4,4ꢀ-dimethoxytrityl)oxy-3-oxo-5-(3(S)-(tert-
butyldiphenylsilyl)oxy-5(R)-(thymin-1-yl)tetrahydrofuran-2(R)-
yl)-13-(thymin-1-yl)-3-phospha-2,4,12-
trioxabicyclo[9.3.0]tetradec-7-ene (12)
Compound 11 (338 mg, 0.277 mmol) was coevaporated with
anhydrous CH2Cl2 (3 × 25 cm3) and redissolved in the same
solvent (15 cm3). To the solution was added a solution of Grubbs’
2nd generation catalyst§ (16 mg, 0.019 mmol) in CH2Cl2 (5 cm3)
◦
and the reaction mixture was stirred at 40 C for 4 h. Then the
◦
temperature was lowered to 30 C and the reaction mixture was
stirred for 16 h. The solvent was removed under reduced pressure,
and the residue was purified by silica gel column chromatography
using MeOH and CH2Cl2 (1 : 49–1 : 9 v/v) to give the product
(246 mg, 74%) as a white foam; Rf 0.41 (MeOH and CH2Cl2,
1 : 14 v/v). dH (300 MHz; CDCl3; Me4Si) 8.57–8.28 (m, NH),
7.68–7.65 (m, Ph), 7.54–7.08 (m, H-6, Ph), 6.87–6.72 (m, H-
1ꢀ, Ph), 6.60–6.48 (m, H-1ꢀ), 6.02–5.95 (m, H-1ꢀ), 5.78–5.71 (m,
=
=
CH2CH CHCH2), 5.30–5.04 (m, CH2CH CHCH2), 4.93–3.61
(m, H-3ꢀ, H-4ꢀ, H-5ꢀ, CH2OP, CH3O), 2.64–1.49 (m, H-2ꢀ, CH3,
Preparation of [5ꢀ-O-(5ꢀ(S)-C-allyl-3ꢀ-O-(tert-
butyldiphenylsilyl)thymidinyl]-[3ꢀ-O-(5ꢀ(S)–C-allyl)thymidinyl]
cyanoethylphosphate (15)
=
CH2CN, CH2CH CHCH2), 1.09–1.08 (m, C(CH3)3). dp (CDCl3)
0.37, −1.03, −6.51, −8.06 (4 : 1 : 8 : 6). MS(MALDI)m/z 912 (M +
Na+ − DMT), HRMS 912.3003 (calcd for C43H52N5O12PSiNa
912.3012).
Compound 11 (48 mg, 0.039 mmol) was dissolved in CH2Cl2
(4 cm3). To the solution was added Et3SiH (13 lL, 0.079 mmol)
and a 1% solution of TFA in CH2Cl2 (0.23 cm3, 0.030 mmol),
whereby the reaction mixture went from weak yellow to orange-
red. The reaction mixture was stirred at room temperature for
20 min. The solution was diluted with CH2Cl2 (30 cm3) and washed
with a saturated solution of NaHCO3 (2 × 10 cm3). The organic
phase was dried (MgSO4) and the solvent was removed under
reduced pressure. The residue was purified by silica gel column
chromatography using MeOH and CH2Cl2 (1 : 33 v/v) to give
the product (29 mg, 80%) as a white foam; Rf 0.44 (MeOH and
CH2Cl2, 1 : 9 v/v); dH (300 MHz; CDCl3; Me4Si) 9.00 (br s, NH),
8.90 (br s, NH), 8.83 (br s, NH), 8.78 (br s, NH), 7.67–7.62 (m,
Ph), 7.54–7.39 (m, H-6, Ph), 6.55–6.50 (m, H-1ꢀ), 6.19–6.13 (m, H-
1ꢀ), 5.82–5.51 (m, H-7ꢀ), 5.27–4.96 (m, T1-H-3ꢀ, H-8ꢀ), 4.35–4.27
(m, T2–H-3ꢀ), 4.20–3.93 (m, H-4ꢀ, T2–H-5ꢀ, CH2OP), 3.84–3.73
(m, T1–H-5ꢀ), 2.76–2.53 (m, H-2ꢀ, H-6ꢀ, CH2CN), 2.41–2.14 (m,
H-2ꢀ, H-6ꢀ), 1.91–1.81 (m, H-2ꢀ, CH3), 1.10–1.07 (m, C(CH3)3).
dp (CDCl3) −1.42. MS(MALDI) m/z 940 (M + Na+), HRMS
940.3369 (calcd for C45H56N5O12PSiNa 940.3325).
Preparation of (E,Z)-(3R/S)-(1S,5S,10S,11R,13R)-3-(2-
cyanoethoxy)-10-hydroxy-3-oxo-5-(3(S)-(tert-
butyldiphenylsilyl)oxy-5(R)-(thymin-1-yl)tetrahydrofuran-2(R)-
yl)-13-(thymin-1-yl)-3-phospha-2,4,12-
trioxabicyclo[9.3.0]tetradec-7-ene (13)
Compound 12 (196 mg, 0.164 mmol) was dissolved in CH2Cl2
(10 cm3). To the solution was added Et3SiH (60 lL, 0.364 mmol)
and a 1% solution of TFA in CH2Cl2 (1.0 cm3, 0.129 mmol),
whereby the reaction mixture went from weak yellow to orange-
red. The reaction mixture was stirred at room temperature for
45 min. The solution was diluted with CH2Cl2 (100 cm3) and
washed with a saturated solution of NaHCO3 (2 × 40 cm3).
The organic phase was dried (MgSO4) and the solvent was
removed under reduced pressure. The residue was purified by
silica gel column chromatography using MeOH and CH2Cl2 (1 :
19 v/v) to give the product (126 mg, 86%) as a white foam;
Rf 0.34, 0.39 (MeOH and CH2Cl2, 1 : 19 v/v); dH (300 MHz;
CDCl3; Me4Si) 8.94–8.78 (m, NH), 8.63 (m, NH), 7.71–7.64
(m, Ph), 7.50–7.43 (m, Ph), 7.38–6.97 (m, H-6, Ph), 6.58–6.41
Preparation of (E,Z)-(1S,5S,10S,11R,13R)-3,10-dihydroxy-5-
(3(S)-hydroxy-5(R)-(thymin-1-yl)tetrahydrofuran-2(R)-yl)-3-oxo-
13-(thymin-1-yl)-3-phospha-2,4,12-trioxabicyclo[9.3.0]tetradec-
7-ene (16)
ꢀ
ꢀ
=
(m, H-1 ), 6.11–5.89 (m, H-1 ), 5.68–5.60 (m, CH2CH CHCH2),
ꢀ
ꢀ
=
5.44–5.26 (m, CH2CH CHCH2), 5.09–4.30 (m, H-3 , H-4 , H-
5ꢀ), 4.16–3.87 (m, H-4ꢀ, H-5ꢀ, CH2OP), 3.78–3.57 (m, T1–H-5ꢀ),
3.07 (br s, OH), 2.88 (br s, OH), 2.81–2.26 (m, H-2ꢀ, CH2CN,
Compound 13 (11.8 mg, 0.013 mmol) was dissolved in anhydrous
THF (1 cm3) and TBAF (1.0 M in THF, 0.030 cm3) was added to
the solution. After stirring for 1 h, the solvent was removed under
reduced pressure. The residue was redissolved in 32% NH3(aq) and
the solution was stirred at 55 ◦C for 20 h. The solvent was removed
ꢀ
=
=
CH2CH CHCH2), 2.20–1.78 (m, H-2 , CH3, CH2CH CHCH2),
1.10–1.07 (m, C(CH3)3). dp (CDCl3) 0.44, −0.76, −6.00, −7.46 (4 :
1 : 8 : 6). MS(MALDI) m/z 912 (M + Na+), HRMS 912.2998
(calcd for C43H52N5O12PSiNa 912.3012).
2442 | Org. Biomol. Chem., 2006, 4, 2433–2445
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
under reduced pressure, and the residue was subjected to HPLC-
purification (C18 column, buffer A (CH3CN and 0.1 M NH4HCO3,
1 : 19 v/v) and buffer B (CH3CN and 0.1 M NH4HCO3, 3 : 1
v/v)); dH (300 MHz; D2O; Me4Si) 7.95 (s, H-6), 7.85 (s, H-6),
7.58 (br s, H-6), 6.44 (m, H-1ꢀ), 6.30 (m, H-1ꢀ), 5.73–5.69 (m,
CDCl3; Me4Si) 164.1 (C-4), 152.2, 151.9, 150.5 (C-2, arom), 147.0,
135.8, 134.0, 131.9, 130.9, 130.1, 130.0, 128.1, 127.9, 127.8, 127.4,
123.9, 123.6, 123.1 (C-6, C-7ꢀ, arom), 118.0, 116.9, 116.7, 110.9
(C-8ꢀ, arom, C-5), 86.5, 84.3, 77.4, 72.6, 71.0 (C-1ꢀ, C-3ꢀ, C-4ꢀ, C-
5ꢀ, C(Ar)3), 40.6 (C-2ꢀ), 36.8 (C-6ꢀ), 12.9 (CH3). MS(MALDI) m/z
561 (M + Na+).
ꢀ
ꢀ
ꢀ
=
CH2CH CHCH2), 5.13–4.52 (m, H-3 , H-4 , H-5 ), 4.12–3.52 (m,
ꢀ
ꢀ
ꢀ
ꢀ
=
H-3 , H-4 , H-5 ), 2.64–2.23 (m, H-2 , CH2CH CHCH2), 1.97–
1.92 (m, CH3). dp (D2O) −0.39, −1.23 (1 : 1).
Preparation of 5ꢀ(S)-C-allyl-3ꢀ-O-(tert-butyldiphenylsilyl)-5ꢀ-
Preparation of (1S,5S,10S,11R,13R)-3,10-dihydroxy-5-(3(S)-
hydroxy-5(R)-(thymin-1-yl)tetrahydrofuran-2(R)-yl)-3-oxo-13-
(thymin-1-yl)-3-phospha-2,4,12-trioxabicyclo[9.3.0]tetradecane
(17)
(methoxy(diisopropylamino)phosphinyl)thymidine (19)
To a solution of 5 (2.75 g, 5.29 mmol) in anhydrous CH2Cl2
◦
(20 cm3) at 0 C was added diisopropylethylamine (1.81 cm3,
10.6 mmol) and (i-Pr)2NP(Cl)OCH3 (1.53 cm3, 7.93 mmol). The
reaction mixture was stirred at room temperature for 3 h and
quenched with methanol (5 cm3). The reaction mixture was diluted
with EtOAc (100 cm3), washed with a saturated aqueous solution
of NaHCO3 (3 × 30 cm3) and brine (3 × 30 cm3), dried (Na2SO4)
and concentrated under reduced pressure. The residue was purified
by silica gel column chromatography using 0.1% Et3N in EtOAc
and petroleum ether (1 : 3 v/v) to give the product as a white foam
(2.70 g, 75%) Rf 0.30 (EtOAc and petroleum ether, 3 : 1 v/v); dH
(300 MHz; CDCl3; Me4Si) 8.80–8.76 (1H, br s, NH), 7.75–7.59
(5H, m, H-6, Ph), 7.46–7.32 (6H, m, Ph), 6.60–6.45 (1H, m, H-1ꢀ),
5.65–5.45 (1H, m, H-7ꢀ), 5.10–4.95 (2H, m, H-8ꢀ), 4.58 (1/2H, d, J
5.1 Hz, H-3ꢀ), 4.50 (1/2H, d, J 5.1 Hz, H-3ꢀ), 3.99–3.94 (1H, m, H-
4ꢀ), 3.48–3.20 (3H, m), 3.25 (3H, d, J 13.5 Hz, OCH3), 3.07 (3H, d,
J 13.5 Hz, OCH3), 2.40–2.20 (3H, m), 2.00–1.80 (4H, m), 1.14–0.90
(21H, m). dP (CDCl3) 152.10, 150.44. MS(MALDI) m/z 704 (M +
Na+), HRMS 704.3228 (calcd for C36H52N3O6PSiNa 704.3255).
Compound 14 (25.4 mg, 0.029 mmol) was dissolved in anhydrous
THF (3 cm3) and TBAF (1.0 M in THF, 0.030 cm3) was added to
the solution. After stirring for 45 min the solvent was removed
under reduced pressure. The residue was redissolved in 32%
NH3(aq) and the solution was stirred at 55 ◦C for two days. The
solvent was removed under reduced pressure, and the residue was
subjected to HPLC-purification (C18 column, buffer A (CH3CN
and 0.1 M NH4HCO3, 1 : 19 v/v) and buffer B (CH3CN and 0.1 M
NH4HCO3, 3 : 1 v/v)); dH (500 MHz; D2O; Me4Si) 7.81 (s, H-6),
7.52 (s, H-6), 6.36 (t, J 6.5 Hz, H-1ꢀ), 6.26 (dd, J 7.5, 6.7 Hz,
H-1ꢀ), 4.76 (1H, m, T1-H-3ꢀ), 4.67 (1H, m, T2-H-3ꢀ), 4.43 (1H, m,
T2-H-5ꢀ), 4.04 (1H, dd, 10.1, 3.4 Hz, T1-H-4ꢀ), 3.99 (1H, m, T2-H-
4ꢀ), 3.82 (1H, m, T1–H-5ꢀ), 2.56–2.32 (4H, m, T1-H-2ꢀ, T2-H-2ꢀ),
2.09–1.93 (2H, m, (CH2)4), 1.96 (3H, s, CH3), 1.91 (3H, s, CH3),
1.87–1.62 (4H, m, (CH2)4), 1.46–1.38 (2H, m, (CH2)4). dp (D2O)
0.12.
Preparation of 5ꢀ(S)-C-allyl-5ꢀ-O-pixylthymidine (18)
Preparation of [5ꢀ-O-(5ꢀ(S)-C-allyl-3ꢀ-O-(tert-
butyldiphenylsilyl)thymidinyl]-[3ꢀ-O-(5ꢀ(S)-C-allyl-5ꢀ-O-
pixylthymidinyl] methylphosphate (20)
Compound 5 (2.20 g, 4.23 mmol) was coevaporated with an-
hydrous pyridine (3 × 15 cm3) and redissolved in anhydrous
pyridine (20 cm3). 9-Chloro-9-phenyl-9H-xanthen (pixyl chloride)
was added, and the reaction mixture was stirred for 12 h. The
solvent was removed under reduced pressure, and the residue
was dissolved in CH2Cl2 (50 cm3) and washed with a saturated
aqueous solution of NaHCO3 (2 × 20 cm3). The aqueous phase
was extracted with CH2Cl2 (3 × 20 cm3) and the combined
organic phases were dried (Na2SO4) and concentrated under
reduced pressure. The residue was coevaporated with toluene and
redissolved in anhydrous THF (20 cm3) and TBAF (1.0 M in THF,
7.8 cm3) was added. The reaction mixture was stirred for 12 h at
room temperature. The reaction mixture was partitioned between
water (100 cm3) and EtOAc (100 cm3). The separated aqueous
phase was extracted with EtOAc (2 × 25 cm3). The combined
organic phases were washed with brine (50 cm3), dried (Na2SO4)
and concentrated under reduced pressure. The residue was purified
by silica gel column chromatography using 0.1% Et3N in CH3OH
and CH2Cl2 (1 : 99 v/v) to give the product as a white foam (2.20 g,
99%); Rf 0.30 (MeOH and CH2Cl2, 1 : 19 v/v); dH (300 MHz;
CDCl3; Me4Si) 7.65 (1H, d, J 1.2 Hz, H-6), 7.48–7.20 (9H, m,
Ph), 7.09–6.99 (4H, m, Ph), 6.16 (1H, t, J 6.3 Hz, H-1ꢀ), 5.47–5.33
(m, 1H, H-7ꢀ), 4.85 (1H, dd, J 9.9, 1.5 Hz, H-8ꢀ), 4.71 (1H, dd,
J 17.1, 1.5 Hz, H-8ꢀ), 3.64–3.55 (2H, m, H-4ꢀ, H-3ꢀ), 3.40–3.25
(1H, m, H-5ꢀ), 2.22–1.93 (6H, m, H-2ꢀ, H-6ꢀ, CH3). dC (75 MHz;
To a solution of 18 (1.00 g, 1.90 mmol) and 19 (1.30 g, 1.90 mmol)
in CH3CN and CH2Cl2 (2 : 1 v/v, 50 cm3) was added 1H-tetrazole
(0.45 M in CH3CN, 4.65 cm3, 2.09 mmol) and the reaction mixture
was stirred for 12 h at room temperature. The reaction mixture
was cooled to 0 ◦C and tert-butylhydroperoxide (5.5 M in decane,
0.52 cm3, 2.85 mmol) was added. The reaction mixture was stirred
at room temperature for 3 h. CH3OH (10 cm3) was added and the
solution was diluted with CH2Cl2 (50 cm3). The reaction mixture
was washed with a saturated aqueous solution of NaHCO3 (3 ×
25 cm3) and brine (25 cm3), dried (Na2SO4) and evaporated under
reduced pressure. The residue was purified by silica gel column
chromatography using 0.1% Et3N in EtOAc and petroleum ether
(3 : 1 v/v) to give the product as a white foam (1.61 g, 75%): Rf
0.15 (EtOAc and petroleum ether, 3 : 1 v/v); dH (300 MHz; CDCl3;
Me4Si) 9.40–9.10 (m, NH), 7.80–6.90 (m, H-6, Ph), 6.56–6.49 (m,
H-1ꢀ), 6.19–6.09 (m, H-1ꢀ), 6.02–5.97 (m, H-1ꢀ), 5.53–5.26 (m, H-
7ꢀ), 5.15–4.62 (m, T1-H-3ꢀ, H-8ꢀ), 4.37–4.32 (m, T2-H-3ꢀ), 3.98–
3.26 (m, H-4ꢀ, H-5ꢀ, OCH2), 3.44 (d, J = 11.4 Hz, OCH3), 3.43 (d,
J = 11.4 Hz, OCH3), 2.40–1.65 (m, H-2ꢀ, H-6ꢀ, CH3), 1.10–1.06 (m,
C(CH3)3). dP (CDCl3) −0.01, −0.88 (1 : 1.8). MS(MALDI) m/z
1157 (M + Na+), HRMS 1157.4119 (calcd for C62H67N4O13PSiNa
1157.4104).
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2433–2445 | 2443
©

View Article Online
Preparation of (3R/S)-(1S,5S,10S,11R,13R)-3-methoxy-10-
pixyloxy-3-oxo-5-(3(S)-(tert-butyldiphenylsilyl)oxy-5(R)-(thymin-
1-yl)tetrahydrofuran-2(R)-yl)-13-(thymin-1-yl)-3-phospha-2,4,12-
trioxabicyclo[9.3.0]tetradecane (21)
then EtOAc and MeOH (99 : 1 v/v) to give the product as a white
foam (36 mg, 37%): Rf 0.40 (MeOH and CH2Cl2, 1 : 9 v/v); dP
(CDCl3) 150.40, 150.04, 150.00, 1.20, 0.99, −4.08.
NMR spectroscopy of 17
To a solution compound 20 (1.16 g, 1.02 mmol) in anhydrous
1,2-dichloroethane was added Grubbs’ 2nd generation catalyst
(60.8 mg, 0.072 mmol) and the reaction mixture was stirred at
60 ◦C for 88 h. The solvent was evaporated under reduced pressure
and the residue was purified by silica gel column chromatography
using a gradient of EtOAc (0–100%) in 0.1% Et3N in petroleum
ether to give the two ring-closed intermediates (308 mg, 27%,
Rf 0.15 (EtOAc); dP (CDCl3) 2.10, 0.64, −4.95, −6.30 (5 : 1.3 :
1 : 4.2); MS(MALDI) m/z 1129 (M + Na+), HRMS 1129.3815
(calcd for C60H63N4O13PSiNa 1129.3791) and (353 mg, 42%,
Rf 0.05 (EtOAc); MS(MALDI) m/z 873 (M + Na+), HRMS
873.2893 (calcd for C41H51N4O12PSiNa 873.2903)) as well as
unreacted starting material (110 mg, 10%, one isomer only). The
two ring-closed intermediates were combined in a solution in
anhydrous methanol (1 cm3) and added Adams’ catalyst (PtO2)
(60 mg, 0.26 mmol). The solution was degassed several times with
argon and placed under a hydrogen atmosphere. The reaction
mixture was stirred for 12 h at room temperature and filtered
through celite. The solvent was removed under reduced pressure
to give a crude saturated and depixylated intermediate (621 mg,
MS(MALDI) m/z 875 (M + Na+), HRMS 875.3077 (calcd for
C41H53N4O12PSiNa 875.3059)). To a solution of this intermediate
(100 mg, 0.111 mmol) in anhydrous CH2Cl2 was added 2,6-lutidine
(68 lL, 0.59 mmol) and pixyl chloride (51.5 mg, 0.18 mmol). The
reaction mixture was stirred at room temperature for 2 h, and
the solvent was removed under reduced pressure. The residue was
coevaporated with toluene to give the crude pixylated intermediate
(150 mg). To a stirred solution of this intermediate (100 mg,
0.09 mmol) in THF at 0 ◦C was added a mixture of glacial acetic
acid (10.3 lL, 0.18 mmol) and TBAF (1.0 M in THF, 271 lL).
The reaction mixture was stirred at room temperature for 3 h.
The reaction mixture was concentrated under reduced pressure,
and the residue was purified by silica gel column chromatography
using 0.1% Et3N in MeOH and CH2Cl2 (1 : 24 v/v) to give the
product as a white foam (25 mg, 27% from 20): Rf 0.33 (MeOH and
CH2Cl2, 1 : 9 v/v); dP (CDCl3) 1.35, −3.63 (3 : 1); MS(MALDI)
m/z 893 (M + Na+), HRMS 893.2704 (calcd for C44H47N4O13PNa
893.2769).
1D 500 MHz 1H NMR spectra were recorded at 10, 23 and
25 ◦C with 20032 points and a spectral width of 5000 Hz. A
phase sensitive NOESY spectrum with a mixing time of 700 ms
◦
was recorded at 23 C with 2 K points in the t2-dimension using
spectral widths of 5000 Hz in both dimensions and 48 transients
for each of the 600 t1-experiments. Phase sensitive DQF-COSY
spectra were recorded at 23 and 25 C with 4 K points sampled
in F2, 5000 Hz spectral widths in both dimensions and 512 t1-
experiments, each recorded with 48 transients. All spectra were
processed using NMRpipe.49 Cross peak intensities were measured
using the program Sparky.50
◦
Restrained molecular dynamics simulations on 17
MD calculations were performed using the AMBER 7.0 program
package51 on an SGI/O2 workstation. Distance restraints were
derived from the NOESY spectra using the ISPA approach.
Atomic charges around the modified phosphorus were calculated
using the restrained electrostatic potential (RESP) procedure.52
The unmodified dinucleotide and the cyclic dinucleotide 17 were
following a short energy minimisation subjected to 1000 ps
molecular dynamics simulation at 300 K. In case of the cyclic
dinucleotide, the calculations were performed both with and
without time averaged NMR restraints.
Preparation of oligodeoxynuclotides
Oligonucleotide synthesis was carried out by using an ExpediteTM
8909 nucleic acid synthesis system from PerSeptive Biosystems Inc.
following the phosphoramidite approach. Synthesis of oligonu-
cleotides 24 and 26 was performed on a 0.2 lmol scale by using
2-cyanoethyl phosphoramidites of standard 2ꢀ-deoxynucleosides
as well as the LNA-meC monomer in combination with the
modified phosphoramidite 22. The synthesis followed the regular
protocol employing standard CPG supports. However, for the
LNA monomer, a prolonged coupling time of 4 min was used,
and for 22, a manual coupling in 20 min followed by 2 × 45 s
capping was used. Coupling yields for 22 were >95%. The 5ꢀ-O-
DMT-ON oligonucleotides were treated with a 1 M solution of
disodium 2-carbamoyl-2-cyanoethylene-1,1-dithiolate43 in DMF
for 30 min, rinsed with ethanol and then removed from the solid
support by treatment with concentrated aqueous ammonia at
room temperature for 72 h, which also removed the protecting
groups. Purification using reversed-phase HPLC was performed
on a Waters Prep LC 4000 system using a Xterra prep MS C18; 10 lm;
7.8 × 150 mm column; buffer A: 0.5 M triethylammonium acetate,
pH 7.4; buffer B: CH3CN and water (3 : 1 v/v); 0–2 min 100% A,
2–40 min from 100% A to 70% B, 30% A, 40–50 min from 70% B,
30% A to 100% B. All fractions containing 5ꢀ-O-DMT-protected
oligonucleotide (retention time 25–30 minutes) were collected,
concentrated and diluted with water (1 cm3). The oligonucleotides
were precipitated by treatment with 100 lL 80% CH3COOH for
30 min. followed by the addition of 100 lL UHQ water, 50 lL
3 M CH3COONa (aq) and 600 lL 99.9% ethanol. The mixture
Preparation of (3R/S)-(1S,5S,10S,11R,13R)-
3-methoxy-10-pixyloxy-3-oxo-5-(3(S)-(2-
cyanoethoxy)(diisopropylamino)phosphin)oxy-5(R)-
(thymin-1-yl)tetrahydrofuran-2(R)-yl)-13-(thymin-1-yl)-3-
phospha-2,4,12-trioxabicyclo[9.3.0]tetradecane (22)
To a stirred solution of compound 21 (80 mg, 0.092 mmol) in
anhydrous CH2Cl2 were added diisopropylethylamine (23.6 lL,
0.138 mmol) and i-Pr2NP(Cl)OCH2CH2CN (24.6 lL, 0.11 mmol)
at 0 ◦C. The mixture was stirred at room temperature for 6 h
and then partitioned between CH2Cl2 (2 × 10 cm3) and a
saturated aqueous solution of NaHCO3 (10 cm3). The combined
organic phases were dried (Na2SO4) and concentrated under
reduced pressure. The residue was purified by silica gel column
chromatography using EtOAc and petroleum ether (1 : 2 v/v) and
2444 | Org. Biomol. Chem., 2006, 4, 2433–2445
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
◦
was left at −18 C for 1 h followed by centrifugation for 20 min
13 A. Kowalczyk, J. R. Carmical, Y. Zou, B. Van Houten, R. S. Lloyd,
C. M. Harris and T. M. Harris, Biochemistry, 2002, 41, 3109.
14 I. Le Cle´zio, J.-M. Escudier and A. Vigroux, Org. Lett., 2003, 5, 161.
15 I. Le Cle´zio, H. Gornitzka, J.-M. Escudier and A. Vigroux, J. Org.
Chem., 2005, 70, 1620.
16 A. Fu¨rstner, Angew. Chem., Int. Ed., 2000, 39, 3012.
17 R. R. Schrock and A. H. Hoveyda, Angew. Chem., Int. Ed., 2003, 42,
4592.
at 4 ◦C. The supernatant was removed and the oligonucleotide
was washed with cold 96% ethanol, dried and dissolved in water.
MALDI MS m/z (M + H) found/calcd: 24 4186.8/4188.5; 26
4438.9/4440.6.
Melting experiments
18 R. H. Grubbs, Tetrahedron, 2004, 60, 7117.
19 T. M. Trnka and R. H. Grubbs, Acc. Chem. Res., 2001, 34, 18.
20 A. M. Sørensen and P. Nielsen, Org. Lett., 2000, 2, 4217.
21 A. M. Sørensen, K. E. Nielsen, B. Vogg, J. P. Jacobsen and P. Nielsen,
Tetrahedron, 2001, 57, 10191.
22 P. Børsting, A. M. Sørensen and P. Nielsen, Synthesis, 2002, 797.
23 P. Børsting and P. Nielsen, Chem. Commun., 2002, 2140.
24 C. Kirchhoff and P. Nielsen, Tetrahedron Lett., 2003, 44, 6475.
25 P. Børsting, M. Freitag and P. Nielsen, Tetrahedron, 2004, 60, 10955.
26 P. Børsting, K. E. Nielsen and P. Nielsen, Org. Biomol. Chem., 2005, 3,
2183.
UV melting experiments were carried out on a Perkin-Elmer
Lambda 35 UV/VIS spectrometer with a PTP-6 Peltier system
and data were processed with Templab version 2.00 and UV
WINLAB version 2.85.04. Samples were dissolved in a medium
salt buffer containing Na2HPO4 (15 mM), NaCl (100 mM) and
EDTA (0.1 mM), pH 7.0 with 1 lM concentrations of the
two complementary sequences. The extinction coefficients were
calculated assuming the extinction coefficients for the dinucleotide
17 to equal two thymidines. The increase in absorbance at 260 nm
as a function of time was recorded while the temperature was
increased linearly from 10 to 90 ◦C at a rate of 1.0 ◦C min−1. The
melting temperature was determined as the local maximum of the
first derivatives of the absorbance versus temperature curve. All
melting curves were found to be reversible. For melting experi-
ments with Mg2+-concentrations of 5 or 10 mM, MgCl2 (1.0 lL
and 2.0 lL, respectively, of a 5.0 M solution) was added to the
samples.
27 P. Børsting, M. S. Christensen, S. I. Steffansen and P. Nielsen,
Tetrahedron, 2006, 62, 1139.
28 M. D. Reynolds, J. M. Dougherty and P. R. Hanson, Chem. Rev., 2004,
104, 2239.
29 F. Amblard, S. P. Nolan and L. A. Agrofoglio, Tetrahedron, 2005, 61,
7067.
30 J. Louie, C. W. Bielawski and R. H. Grubbs, J. Am. Chem. Soc., 2001,
123, 11312.
31 G. Wang and P. J. Middleton, Tetrahedron Lett., 1996, 37, 2739.
32 G. Wang and P. J. Middleton, Nucleosides Nucleotides, 1998, 17, 1033.
33 V. Banuls and J.-M. Escudier, Tetrahedron, 1999, 55, 5831.
34 D. B. Dess and J. C. Martin, J. Am. Chem. Soc., 1991, 113, 7277.
35 S. Abbas, R. D. Bertram and C. J. Hayes, Org. Lett., 2001, 3, 3365.
36 J. D. More and N. S. Finney, Org. Lett., 2002, 4, 3001.
37 A. K. Saha, T. J. Caulfield, C. Hobbs, D. A. Upson, C. Waychunas and
A. M. Yawman, J. Org. Chem., 1995, 60, 788.
Acknowledgements
The Danish National Research Foundation, the Danish Natural
Science Research Council and Møllerens Fond are thanked for
financial support. Prof. Otto Dahl is thanked for supplying the
dithiolate and for useful advice concerning the preparation of the
oligonucleotides. Mrs Birthe Haack and Suzy W. Lena are thanked
for technical assistance.
38 J. Fensholdt and J. Wengel, Acta Chem. Scand., 1996, 50, 1157.
39 P. Nielsen, K. Larsen and J. Wengel, Acta Chem. Scand., 1996, 50, 1030.
40 H. Trafelet, E. Stulz and C. Leumann, Helv. Chim. Acta, 2001, 84, 87.
41 M. Tarko¨y, M. Bolli and C. Leumann, Helv. Chim. Acta, 1994, 77, 716.
42 Y. L. Jiang, C. J. Wiederholt, J. N. Patro, K. Haraguchi and M. M.
Greenberg, J. Org. Chem., 2005, 70, 141.
43 B. J. Dahl, K. Bjergaarde, L. Henriksen and O. Dahl, Acta Chem.
Scand., 1990, 44, 639.
44 S. K. Singh, P. Nielsen, A. A. Koshkin and J. Wengel, Chem. Commun.,
1998, 455.
45 S. L. Pedersen and P. Nielsen, Org. Biomol. Chem., 2005, 3, 3570.
46 N. B. Leontis, W. Kwok and J. S. Newman, Nucleic Acids Res., 1991,
19, 759.
References
1 E. T. Kool, Chem. Rev., 1997, 97, 1473.
2 C. J. Leumann, Bioorg. Med. Chem., 2002, 10, 841.
3 P. Herdewijn, Biochim. Biophys. Acta, 1999, 1489, 167.
4 M. Meldgaard and J. Wengel, J. Chem. Soc., Perkin Trans. 1, 2000,
3539.
47 V. Thiviyanathan, B. A. Luxon, N. B. Leontis, N. Illangasekare, D. G.
Donne and D. G. Gorenstein, J. Biomol. NMR, 1999, 14, 209.
48 V. V. Filichev and E. B. Pedersen, Org. Biomol. Chem., 2003, 1, 100.
49 F. Delaglio, S. Grzesiek, G. W. Vuister, G. Zhu, J. Pfeifer and A. Bax,
J. Biomol. NMR, 1995, 6, 277.
50 T. D. Goddard and D. G. Kneller, SPARKY 3, University of California,
San Francisco, (http://www.cgl.ucsf.edu/home/sparky).
51 D. A. Case, D. A. Pearlman, J. W. Caldwell, T. E. Cheatham, III, J.
Wang, W. S. Ross, C. L. Simmerling, T. A. Darden, K. M. Merz, R. V.
Stanton, A. L. Cheng, J. J. Vincent, M. Crowley, V. Tsui, H. Gohlke,
R. J. Radmer, Y. Duan, J. Pitera, I. Massova, G. L. Seibel, U. C. Singh,
P. K. Weiner and P. A. Kollman, AMBER 7, University of California,
San Francisco, 2002.
5 J. Kurreck, Eur. J. Biochem., 2003, 270, 1628.
6 M. Petersen and J. Wengel, Trends Biotechnol., 2003, 21, 74.
7 P. Belmont, J.-F. Constant and M. Demeunynck, Chem. Soc. Rev., 2001,
30, 70.
8 C. R. Allerson and G. L. Verdine, Chem. Biol., 1995, 2, 667.
9 J. T. Goodwin, S. E. Osborne, E. J. Scholle and G. D. Glick, J. Am.
Chem. Soc., 1996, 118, 5207.
10 M. Sekine, O. Kurasawa, K. Shohda, K. Seio and T. Wada, J. Org.
Chem., 2000, 65, 6515.
11 M. Sekine, O. Kurasawa, K. Shohda, K. Seio and T. Wada, Eur. J. Org.
Chem., 2001, 1989.
12 K. Seio, T. Wada and M. Sekine, Helv. Chim. Acta, 2000, 83, 162.
52 C. I. Bayly, P. Cieplak, W. D. Cornell and P. A. Kollman, J. Phys. Chem.,
1993, 97, 10269.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2433–2445 | 2445
©