Tricyclic nucleosides derived from D-glucose. Synthesis and conformational behaviour

Article abstract of DOI:10.1039/b006082h

Two anomeric nucleosides with tricyclic carbohydrate moieties, 3 and 14, are synthesised in 11 steps from diacetone-D-glucose, taking advantage of a stereoselective Grignard reaction, a stereoselective dihydroxylation and a regioselective tandem ring-closing procedure. The configuration of 3 is confirmed by measuring the ³J_HH coupling constants in connection with molecular modelling and ab initio calculations, as these exclude alternative tricyclic nucleoside structures. The conformational preference for 3 is described. The furanose ring is found to be in an O-4′-endo conformation and the γ torsion angle in the +ap range.

Full text of DOI:10.1039/b006082h

View Article Online / Journal Homepage / Table of Contents for this issue
Tricyclic nucleosides derived from D-glucose. Synthesis and
conformational behaviour
Poul Nielsen,* Michael Petersen and Jens Peter Jacobsen
1
Department of Chemistry, University of Southern Denmark, Odense University,
5230 Odense M, Denmark
Received (in Cambridge, UK) 27th July 2000, Accepted 15th September 2000
First published as an Advance Article on the web 31st October 2000
Two anomeric nucleosides with tricyclic carbohydrate moieties, 3 and 14, are synthesised in 11 steps from
diacetone--glucose, taking advantage of a stereoselective Grignard reaction, a stereoselective dihydroxylation
and a regioselective tandem ring-closing procedure. The configuration of 3 is confirmed by measuring the 3JHH
coupling constants in connection with molecular modelling and ab initio calculations, as these exclude alternative
tricyclic nucleoside structures. The conformational preference for 3 is described. The furanose ring is found to be
in an O-4Ј-endo conformation and the γ torsion angle in the ϩap range.
Introduction
Conformationally restricted nucleosides have attracted
considerable attention as monomers in oligonucleotide
analogues1–8 and as potential antiviral agents.9–12 Especially,
nucleosides with bi-2–7 and tricyclic8 carbohydrate moieties
have been synthesised in order to restrict the conformational
flexibility of nucleosides into conformations which are
ideal for nucleic acid recognition. As a prime example, LNA
(Locked Nucleic Acid) has recently been introduced and
shown to display unprecedented recognition of comple-
mentary DNA and RNA.4 Nucleoside analogues involving
the nucleobase in conformationally fixed polycyclic structures
have also been reported.13 Bicyclic nucleosides have demon-
strated some activity against antiviral enzymes,9 e.g. HIV
reverse transcriptase,11 and can, furthermore, serve as tools
in the evaluation of conformational preferences and struc-
ture–activity relations in nucleoside/nucleotide converting
enzymes.11,12
thetic strategy potentially leading to both of the two possible
The bicyclic nucleoside 1 has been synthesised and incorpor-
ated into oligodeoxynucleotides.2 A fully modified sequence
exhibited moderately enhanced affinity towards a comple-
mentary RNA strand when compared with an unmodified
oligodeoxynucleotide sequence. On the other hand a decreased
affinity towards a complementary single stranded DNA strand
was observed.2 The smaller and more constrained bicyclic
nucleoside 2 has also been incorporated into oligodeoxynucleo-
tides demonstrating a slight improvement over 1 concerning
affinity towards complementary RNA-sequences.3 Further-
more, an improvement in affinity towards single stranded
DNA-sequences was also observed. In a conformational study
we have suggested 1 to prefer a C-1Ј-exo conformation14 with a
pseudorotation angle14 of P ≈ 129Њ but with a slight degree of
conformational freedom in the bicyclic system.3 However, an
X-ray crystallographic study of 1 in a DNA dodecamer duplex
has shown that the bicyclic nucleoside residues have O-4Ј-endo
conformations14 with pseudorotation angles of P = 94 6Њ.15
Likewise, a recent NMR study has confirmed this conform-
ation of 1 in a duplex structure.16 In this paper, our earlier
conformational conclusions3 are revised.
β-configured tricyclic nucleosides 3 and 4.† However, only the
(7ЈR)-isomer† 3 was obtained, and the synthesis of this as well
as its α-configured anomer 14 is hereby presented. The con-
formation of 3 was established by use of NMR spectroscopy,
molecular modelling and ab initio calculations.
† In all the tricyclic nucleoside structures 3, 4, 14, 15, 16 we use the
same numbering scheme as exemplified below with 3; C-1Ј–C-6Ј repre-
sent the atoms C-1–C-6 in the original -glucose, and C-7Ј is the new
bridgehead atom connecting C-3Ј and C-8Ј, the additional CH2 posi-
tioned in either of the additional rings. H-6Љ and H-8Љ are positioned
on the β-face of the nucleoside structure. Imaginary structures (com-
pounds never synthesised) are throughout this report shown in italics.
Note, however, that atom locants in the NMR spectra (Experimental
section) follow the systematic nomenclature used for the tricycles 3,
11–14.
In order to induce further conformational restriction in the
bicyclic structure of 1 the introduction of a 5Ј-CH2–O-3Љ bridge
generating tricyclic nucleosides was planned. We chose a syn-
3706
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713
This journal is © The Royal Society of Chemistry 2000
DOI: 10.1039/b006082h

View Article Online
cyclic compound 11. Even though a tricyclic methyl furanoside
Results
structure was confirmed by NMR and MS data, the exact
configuration of the new C-7Ј-stereocentre and the distribution
of the fused ring systems were not finally solved at this stage.
Using the same reaction sequence on the α-anomer 10 gave the
tricyclic compound 12. Coupling of either 11 or 12 to thymine
using a modified Vorbrüggen method gave in both cases a
mixture of anomeric nucleosides 13 (β:α ≈ 3:1) in 52% yield.
After removal of the benzyl groups, using hydrogenation at
atmospheric pressure followed by chromatographic separation,
the two tricyclic nucleosides 3 and 14 were isolated in 55 and
18% yield, respectively. The major isomer was later determined
to be the β-anomer 3 (vide infra).
Chemical synthesis
Retrosynthetic analysis of 3 and 4, both with -altrose con-
figuration, suggested -glucose as the carbohydrate precursor
possessing a convenient stereochemistry. Thus, diacetone--
glucose was chosen as a very cheap and readily available
starting material demanding conversion of the configuration
at the C-2 and C-3 positions. Efficient conversion to the known
3-C-vinyl--allose derivative 517 was performed in three steps
and 75% overall yield using oxidation followed by a stereoselec-
tive vinylmagnesium-bromide-mediated Grignard reaction and
selective deprotection of the primary acetonide (Scheme 1).
The stereoselectivity observed for the dihydroxylation reac-
tions on 9 and 10 was ascribed to control from the allylic
oxygen18 in combination with a steric effect from the large C-4
substituent (Fig. 1). Subsequently, the primary hydroxy group
could not find a position for attacking either C-2 or C-6, while
the secondary hydroxy group was placed in a position favour-
able for attacking C-6 and forming the six-membered ring in
a bicyclo[4.3.0] intermediate. The remaining primary hydroxy
group in this system was thereby organised towards a nucleo-
philic attack at C-2 forming the tricyclic methyl furanosides 11
and 12.
Fig. 1 Suggested mechanism explaining the formation of the tricyclic
methyl furanoside 11.
Determination of the configuration
In order to determine the exact configuration of the two syn-
thesised anomeric tricyclic nucleosides 3 and 14, intensive
NMR investigations, molecular modelling and ab initio calcu-
lations were performed. 1H and 13C NMR signals were assigned
using 1D- and 2D-NMR methods. This confirmed, in conjunc-
tion with MS data, tricyclic nucleoside structures. Due to a con-
siderable signal overlap neither 1D- nor 2D-NOE experiments
were able to completely elucidate the exact configurations.
However, a conclusive NOE-contact between H-1Ј and H-4Ј
confirmed the β-configuration of the major product (nucleoside
3), and the lack of an NOE contact between H-1Ј and H-4Ј
indicated the α-configuration of the minor product (nucleoside
14). Furthermore, the lack of an NOE-contact between H-6
and H-7Ј indicated the C-7Ј-configuration to be the one in 3
rather than the one in 4. Nevertheless, the exact configuration
of the major product could not be completely assigned, and
four possible tricyclic β-nucleoside structures 3, 4, 15 or 16
had to be considered. However, the structure 3 seemed most
likely to be identical to the major product due to preliminary
NOE results as well as mechanistic arguments (vide supra,
Scheme 1 Reagents and conditions: i, CrO3, Ac2O, pyridine, CH2Cl2;
ii, vinylMgBr, ether, THF; iii, 80% AcOH; iv, TrCl, pyridine; v, BnBr,
NaH, DMF; vi, 20% HCl in aq. MeOH; vii, MsCl, pyridine; viii, OsO4,
N-methylmorpholine N-oxide, aq. pyridine, t-BuOH; ix, NaH, DMF;
x, thymine, N,O-bis(trimethylsilyl)acetamide, TMS triflate, CH3CN;
xi, H2, Pd(OH)2–C, EtOH.
Selective tritylation of the primary hydroxy group to give 6 and
benzylation of the two remaining hydroxy groups to give 7
in 63% overall yield was followed by treatment with methanolic
hydrochloric acid to give an anomeric mixture of detritylated
methyl furanosides 8 in 76% yield. The two hydroxy groups
were converted to methylsulfonic esters and the two anomers
9 and 10 were separated. A standard dihydroxylation of the
double bond in the β-anomer 9 using osmium tetraoxide
and N-methylmorpholine N-oxide as cooxidant did not, as
expected, yield two diastereomeric diols but a more complex
mixture of products. This mixture was, without further purifi-
cation, treated with sodium hydride to give in 48% yield only
one major product which was later determined to be the tri-
3
Fig. 1). Fortunately, measurements of JHH coupling constants
in conjunction with molecular modelling did, as shown below,
successfully establish the major tricyclic nucleoside product
to be 3.
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713
3707

View Article Online
Table 1 1H NMR experimental data for 3 and 14a
3
14
δ/ppm
3JHH/Hz
δ/ppm
3JHH/Hz
H-1Ј
H-2Ј
H-4Ј
H-5Ј
H-6Ј
H-6Љ
H-7Ј
H-8Ј
H-8Љ
6.11
4.39
4.04
4.22
4.02
3.72
3.90
4.01
3.95
H-1Ј
H-2Ј
H-4Ј
H-5Ј
H-6Ј
H-6Љ
H-7Ј
5.68
4.72
4.53
4.08
3.83
3.69
3.97
H-1Ј,H-2Ј
H-4Ј,H-5Ј
H-5Ј,H-6Ј
H-5Ј,H-6Љ
H-7Ј,H-8Ј
H-7Ј,H-8Љ
4.7
3.1
H-1Ј,H-2Ј
H-4Ј,H-5Ј
H-5Ј,H-6Ј
H-5Ј,H-6Љ
H-7Ј,H-8Ј
H-7Ј,H-8Љ
1.8
2.5
7.7
6.6
8.0
9.1
3–5b
<2b
3.2
H-8Ј/H-8Љ
≈4.05
<1.5
a Measured in CD3OD at 500 MHz. b Coupling constants estimated from an ABX system.
Fig. 2 JH1ЈH2Ј as a function of the furanose pseudorotation angle P for
3, 4, 15 and 16. (A, B, C) Karplus curves calculated with puckering
amplitude of 42Њ, 38Њ and 34Њ, respectively. It is evident that the
observed coupling constant (4.7 Hz) yields two possible conformational
ranges.
Measurement of 3JHH coupling constants and determination of
allowed ranges for the torsion angles
Fig. 3 (a) JH4ЈH5Ј as a function of θH4ЈH5Ј for 3, 4, 15 and 16. The
observed coupling constant (3.1 Hz) allows four ranges of θH4ЈH5Ј
(b) JH5ЈH6Ј (——) and JH5ЈH6Љ (----) as functions of θH5ЈH6Ј for 3, 4, 15
and 16. The observed coupling constants (JH5ЈH6Ј = 8.0 Hz and JH5ЈH6Љ
7.7 Hz or vice versa) give rise to two possible geometries.
.
3
The JHH coupling constants of 3 and 14 were determined as
given in Table 1. Since no change in the coupling constants was
observed for any of the nucleosides in the temperature range
from Ϫ50 to ϩ50 ЊC we believe that each nucleoside exists in
only one conformer and not in a dynamic equilibrium between
two or more conformers. All subsequent molecular modelling
calculations were performed under this assumption. We derived
the Karplus relationships between JH1ЈH2Ј and the pseudo-
rotation angle P of the furanose ring.19–21 Similarly, we derived
the relationships between JH4ЈH5Ј, JH5ЈH6Ј, JH5ЈH6Љ, JH7ЈH8Ј and
JH7ЈH8Љ and the corresponding torsion angles for the four
theoretically possible β-configured isomers 3, 4, 15 and 16.†
The Karplus dependence between JH1ЈH2Ј and P is identical
for all these isomers and for 1 as well. Likewise, the Karplus
relationships for JH4ЈH5Ј, JH5ЈH6Ј and JH5ЈH6Љ are identical for 3, 4,
15 and 16. The Karplus relationships are shown in Figs. 2–4.
Since the H-6Ј/H-6Љ and H-8Ј/H-8Љ proton pairs are geminal,
the difference in the torsion angles θH5ЈH6Ј and θH5ЈH6Љ as well as
in the torsion angles θH7ЈH8Ј and θH7ЈH8Љ must be close to 120Њ
in both cases. Thus, only a few of the regions are allowed by
the experimental vicinal coupling constants. These regions are
marked out in the Figures and discussed below.
=
The configuration of 3
As demonstrated by Figs. 2–4, the measured three-bond
coupling constants gave fair amount of information,
a
sufficient to establish the configuration and conformation of 3
with a high degree of confidence.
First, the alternative isomer 4 was considered. This is a rather
strained entity as can be gauged from simple model building.
Consequently, an unconstrained 3-21G* geometry optimisation
yielded only one possible geometry for this configuration
(Table 2). This geometry does not comply with the experimental
coupling constants for the major product 3, e.g. the pseudo-
rotation angle of the furanose is 39Њ, yielding JH1ЈH2Ј ≈ 7 Hz
(Fig. 2).
The other alternative isomers 15 and 16 represent a more
complex situation since these ring systems contain more con-
formational freedom than does 4. Therefore, we carried out
a number of geometry optimisations at the 3-21G* level either
3708
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713

View Article Online
Table 2 Calculated data on the four theoretically possible tricyclic β-nucleosides.a All angles are in degrees
P
Φmax
θH4ЈH5Ј
θH5ЈH6Ј
θH5ЈH6Љ
θH7ЈH8Ј
θH7ЈH8Љ
γc
3
85
201
201
37
38
38
55
Ϫ53
Ϫ42
Ϫ23
47
16
Ϫ141
Ϫ73
Ϫ103
38
0
12
Ϫ85
Ϫ125
Ϫ113
178b
62
78
4
39
22
Ϫ26
54
Ϫ65
44
172
92
15
175
161
178
91
188
171
23
18
24
30
32
25
Ϫ42
Ϫ38
Ϫ42
69
Ϫ12
Ϫ6
84
88
83
؊30
؊30
؊30
Ϫ39
Ϫ35
Ϫ40
؊148
؊148
؊148
Ϫ55
62
85
60
62
60
85
81
86
80
؊38
؊64
Ϫ57
؊64
؊38
Ϫ163
107
115
16
85
66
81
81
82
64
38
36
39
44
40
46
Ϫ23
Ϫ25
57
19
43
83
49
Ϫ64
؊30
؊30
؊30
Ϫ39
Ϫ68
Ϫ178
؊148
؊148
؊148
Ϫ69
40
؊112
Ϫ44
؊112
40
165
؊82
126
Ϫ171
126
؊82
95
89
176
137
166
146
26
a Gaussian94 geometry optimisations carried out at the 3-21G* level except b which was carried out at the 6-31G* level. Torsion angles constrained
in a given calculation are shown in bold. c γ is defined as the C-3Ј–C-4Ј–C-5Ј–O torsion angle.
Table 3 Torsion angles (Њ) allowed by Karplus analysis of the
experimental coupling constantsa
15 ϩ 16
θH5ЈH6Ј
θH5ЈH6Љ
Ϫ148
c
Assignment 1b
Assignment 2
Ϫ30
138
p.g.
n.p.
15
15
θH7ЈH8Ј
θH7ЈH8Љ
c
Assignment 1b
Assignment 2
136
Ϫ38
Ϫ64
116
Ϫ100
n.p.
p.g.
p.g.
n.p.
85
60
Ϫ120
16
θH7ЈH8Ј
θH7ЈH8Љ
c
Assignment 1b
Assignment 2
40
Ϫ133
Ϫ112
63
Ϫ82
103
126
p.g.
d.n.c.
p.g.
Ϫ54
d.n.c.
a θH5ЈH6Ј- and θH5ЈH6Љ- (as well as θH7ЈH8Ј- and θH7ЈH8Љ-) values are derived
from separate Karplus curves and, hence, do not differ by exactly 120Њ.
b Assignment 1 denotes the assignment of the H-6Ј/H-6Љ and H-8Ј/H-8Љ
proton pairs listed in Table 1, which subsequently was confirmed to be
the correct assignment. c p.g. = possible geometry; n.p. = not possible
due to covalent geometry; d.n.c. = calculation did not converge to a
minimum.
The geometries allowed by Karplus analysis of the experi-
mental coupling constants are listed in Table 3. Subsequently,
we carried out Gaussian geometry optimisations with either the
H-5Ј–H-6Ј/H-6Љ fragment constrained, the H-7Ј–H-8Ј/H-8Љ
fragment constrained, or both these fragments constrained to
the possible geometries listed in Table 3. The results of these
calculations are included in Table 2. None of the calculations
yielded a geometry which can encompass all the experimental
data, e.g. the pseudorotational angles P found for five of the
six geometries of 15 are in the 161–188Њ range which do not
comply with either of the two possible conformational ranges
depicted in Fig. 2. The sixth geometry of 15, as well as the six
Fig. 4 (a) JH7ЈH8Ј (——) and JH7ЈH8Љ (----) as functions of θH7ЈH8Ј
calculated for 3 and 16. The observed coupling constants (JH7ЈH8Ј
3.2 Hz and JH7ЈH8Љ < 1.5 Hz or vice versa) give rise to four possible
geometries. (b) JH7ЈH8Ј (——) and JH7ЈH8Љ (----) as functions of θH7ЈH8Ј
=
calculated for 4 and 15. The observed coupling constants (JH7ЈH8Ј
=
3.2 Hz and JH7ЈH8Љ < 1.5 Hz or vice versa) give rise to four possible
geometries.
with no constraints or with one or two CH2CH fragments
constrained in the geometry allowed by the coupling constants.
The situation is further complicated as we, a priori, cannot
perform a stereoselective assignment of the geminal H-6Ј/H-6Љ
and H-8Ј/H-8Љ proton pairs. However, as shown below the
calculated geometries of 16, can be excluded by either θH4ЈH5Ј
,
θH5ЈH6Ј and/or θH7ЈH8Ј torsion angles not complying with the
possible ranges depicted in Figs. 3 and 4.
In contrast, unconstrained geometry optimisations of the
structure 3 at both the 3-21G* and 6-31G* levels yielded nearly
3
analysis of the JHH coupling constants allows stereoselective
assignment.
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713
3709

View Article Online
Fig. 5 A stereoview presenting the determined conformation of 3.
identical structures (Table 2, footnote b) in perfect agreement
Both five-membered rings of the tricyclic nucleoside 3 are in
with the experimental data (Table 1 and Figs. 2–4). Thus, we
conclude that only the structure 3 is commensurate with our
experimental data of the major tricyclic nucleoside product
obtained.
envelope conformations (as found for 1), whereas the six-
membered ring has a twisted boat conformation. The furanose
ring of the tricyclic nucleoside 3 prefers the O-4Ј-endo conform-
ation as in the bicyclic nucleoside 1. This is an unusual and
high-energy conformation in unmodified nucleosides.14 How-
ever, this sugar conformation might be favourable for nucleic
acid recognition as found for oligodeoxynucleotides containing
1,2 2,3 or the 2Ј-fluoro-arabinodeoxynucleoside analogues.15,24
On the other hand, the C-4Ј–C-5Ј torsion angle γ is found to
be in the ϩap range (Table 2). This is probably unfavourable for
duplex formation since in both A- and B-type duplexes γ is
in the ϩsc range.14 Bicyclic nucleosides restricting this torsion in
the ϩap/ϩac range have been studied by Leumann and co-
workers6,7 who showed that the corresponding oligodeoxy-
nucleotides prefer Hoogsteen-association over Watson–Crick
base-pairing.7 Subsequently, the ability to recognise comple-
mentary DNA and RNA was improved with tricyclic nucleo-
sides in which γ has been shown to be in the less unfavourable
ϩac range.8
The conformations of 3 (and 1 revised)
Two regions of the pseudorotation cycle are permitted by the
experimental coupling constant for 3 JH1ЈH2Ј = 4.7 Hz, one close
to P = 90Њ and one in the vicinity of P = 200Њ (Fig. 2). The
unconstrained geometry optimisations gave structures with
P ≈ 90Њ (vide supra) and to test if the P ≈ 200Њ domain is feasible
we carried out two constrained geometry optimisations with
P = 200Њ (constraining two of the endo cyclic torsion angles).
The two starting structures were low-energy conformations
from a molecular dynamics calculation. However, neither of the
conformations generated (Table 2) agreed with the measured
coupling constants. Therefore, we conclude that the conform-
ation obtained by the unconstrained geometry optimisation
indeed is the conformation dictated by the experimentally
observed data. A view of the determined conformation is
presented in stereo in Fig. 5.
No conclusions about the torsional angle χ describing the
glycosidic bond14 can be drawn from this study. However,
no factors preventing the preferable anti conformation are
expected.
We have previously studied the conformations of 1.3
However, this study was performed using a general Karplus
equation19 and forcefield methods.3 To draw any conclusions on
the resemblance of 1 to 3, an analysis of this bicyclic nucleoside
was performed at the same theoretical level as 3, i.e. a Karplus
equation parametrised for nucleotides20 and ab initio quantum
mechanical calculations. It is evident from Fig. 2 that two
regions of the pseudorotation cycle are allowed by the experi-
mental coupling constant JH1ЈH2Ј = 4.3 Hz.2 A geometry opti-
misation at the 3-21G* level yielded a conformation with
P ≈ 90Њ, in accordance with the measured coupling constant,2
X-ray crystallographic data,15 and an NMR high-resolution
structure.16 It is noteworthy that P = 90Њ is a ‘high-energy’ O-4Ј-
endo furanose conformation,14 in this case probably driven by
the propensity of the 2Ј,3Ј-fused tetrahydrofuran ring to adopt
an envelope conformation.
The other target tricyclic nucleoside 4, which was not
obtained from the present synthetic strategy, seems to be more
conformationally restricted than 3. Thus, incorporation of 4 in
oligonucleotide sequences might be more favourable for high-
affinity nucleic acid recognition since our calculations (Table 2)
have shown the furanose ring to prefer a C-3Ј-endo (N-type)14
conformation. Bicyclic nucleosides restricted in this conform-
ation have demonstrated very high affinity towards both com-
plementary RNA and DNA.1,4 Furthermore, the γ-torsion of
4 seems to be restricted in the favourable ϩsc/ϩac range.
The tricyclic α-nucleoside 14 was not intensively evaluated
but the JHH coupling constants (Table 1) indicate that all
torsion angles seem to be in the same range as for 3. Only the
H-1Ј–H-2Ј angle differs due to the α-configured structure.
Conclusions
Discussion
We have synthesised the tricyclic nucleoside 3 as well as its
α-anomer 14 using an efficient and stereoselective method. The
synthesis of 4 demands a different synthetic strategy in which
conversion of the 7Ј-configuration must be accomplished.
However, 4 seems to be restricted in a more favourable con-
formation than does 3 concerning incorporation into nucleic-
acid-recognising oligonucleotides. Therefore, efforts towards
the synthesis of 4 and evaluation of both the conformationally
restricted tricyclic nucleosides 3 and 4 as monomers in oligo-
nucleotide sequences and as biologically active compounds will
be performed in due course.
It has to be taken into account that all our calculations were
performed as gas-phase calculations and at a restricted theor-
etical level. However, geometry optimisations tend to converge
at a theoretically lower level than energies do.22 Our level of
calculations taken into consideration, we refrain from discus-
sions of energies as these are prone to be inaccurate. The geom-
etry of 3 found seems reasonable, e.g. the five-membered rings
have puckering amplitudes of 37Њ and 38Њ, respectively. This is
normally observed in furanose and tetrahydrofuran rings in
general.14,23
3710
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713

View Article Online
4.25 (1H, d, J 3.6, 2-H), 5.35 (1H, dd, J 10.9 and 1.6, CH᎐
Experimental
᎐
CHH), 5.59 (1H, dd, J 17.3 and 1.6, CH᎐CHH), 5.73 (1H, d,
᎐
All commercial reagents were used as supplied. Light petroleum
refers to the fraction with distillation range 60–80 ЊC. 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 Bruker AC250, a Varian Gemini 2000 or a Varian
Unity 500 spectrometer. 1H NMR spectra were recorded at 250,
300 or 500 MHz and 13C NMR spectra were recorded at 62.5 or
75 MHz. Values for δ are in ppm relative to tetramethylsilane
as internal standard. J-Values are in Hz. 1H–1H COSY spectra
J 3.6, 1-H), 5.87 (1H, dd, J 17.3 and 10.9, CH᎐CH ), 7.19–7.46
᎐
2
(15H, m, ArH); δC (CDCl3) 26.36, 26.52, 64.89, 69.77, 78.95,
80.64, 83.64, 86.61, 103.53, 112.94, 116.59, 126.95, 127.75,
128.56, 134.73, 143.64.
3,5-Di-O-benzyl-1,2-O-isopropylidene-6-O-triphenylmethyl-3-C-
vinyl-ꢀ-D-allofuranose 7
A 60% oily dispersion of sodium hydride (782 mg, 19.6 mmol)
was suspended in anhydrous DMF (17 cm3) and the mixture
was cooled to 0 ЊC. A solution of 6 (4.02 g, 8.22 mmol) in
anhydrous DMF (6 cm3) was added dropwise over a period of
45 min. The mixture was stirred at 50 ЊC for 1 h and then cooled
to 0 ЊC. Benzyl bromide (2.34 cm3, 19.6 mmol) was added
dropwise and the reaction mixture was stirred at room temper-
ature for 16 h. The solvent was removed by distillation under
reduced pressure and the residue was treated with water (50
cm3), neutralised with aq. HCl, and extracted with CH2Cl2
(2 × 100 cm3). The combined extracts were washed with satur-
ated aq. NaHCO3 (150 cm3) and then dried (Na2SO4). The
solvent was removed by distillation under reduced pressure and
the residue was purified by silica gel column chromatography
(light petroleum–ethyl acetate 9:1, v/v) to give the product 7
(3.87 g, 70%) as a solid (Found: C, 78.80; H, 6.66. C44H44O6
requires C, 79.02; H, 6.63%); δH (CDCl3) 1.37 (3H, s), 1.60 (3H,
s), 3.40 (2H, d, J 4.4), 3.77 (1H, m), 4.42 (1H, d, J 6.8), 4.56
(1H, d, J 11.8), 4.57 (1H, d, J 11.0), 4.60 (1H, d, J 3.7), 4.69
(1H, d, J 11.0), 4.77 (1H, d, J 11.6), 5.20 (1H, d, J 17.9), 5.30
(1H, d, J 11.3), 5.74 (1H, d, J 3.8), 5.78 (1H, dd, J 18.1 and
11.5), 7.18–7.46 (25H, m); δC (CDCl3) 26.61, 26.83, 64.33,
67.09, 71.98, 77.30, 80.79, 81.67, 85.53, 86.59, 103.90, 112.69,
117.92, 126.85, 127.14, 127.21, 127.57, 127.74, 127.81, 128.14,
128.21, 128.96, 135.95, 138.79, 139.02, 144.30; PDMS m/z 691
[M ϩ Naϩ].
1
were recorded for compounds 3, 8 and 14. H NOE difference
1
spectra were recorded for compounds 3 and 14 and a H–13C
COSY spectrum was recorded for compound 3. FAB mass
spectra were recorded in positive-ion mode on a Kratos
MS50TC spectrometer, and plasma desorption mass spectra
(PDMS) on an Applied Biosystems Biopolymer Mass Analyzer
BIO-ION 20R. Microanalyses were performed at The Micro-
analytical Laboratory, Department of Chemistry, University of
Copenhagen.
1,2-O-Isopropylidene-3-C-vinyl-ꢀ-D-allofuranose17
5
A stirred suspension of chromium trioxide (14.8 g, 136 mmol)
in anhydrous CH2Cl2 (228 cm3) was cooled to 0 ЊC and a
mixture of anhydrous pyridine (24.8 cm3) and Ac2O (14.8 cm3)
was added. Diacetone--glucose (21.0 g, 80.6 mmol) was
added in small portions over a period of 30 min and the mixture
was stirred for another 3 h and poured into ethyl acetate
(800 cm3). The mixture was filtered through a layer of silica, the
filter was rinsed with ethyl acetate (1000 cm3), and the filtrate
was concentrated in vacuo. The residue was re-dissolved in
anhydrous THF (500 cm3) and this solution was cooled to 0 ЊC.
A 1 M solution of vinylmagnesium bromide in THF (150 cm3;
150 mmol) was added dropwise and the mixture was stirred for
42 h. Water (600 cm3) was added and the solution was neutral-
ised with 4 M AcOH. The solvent was partly removed in vacuo
and the residue was extracted with ether (3 × 600 cm3). The
combined organic extracts were washed with saturated aq.
NaHCO3 (800 cm3) and then dried (MgSO4). The solvent was
removed by distillation under reduced pressure and the residue
was purified by silica gel column chromatography (CH2Cl2–
MeOH 99:1, v/v) to give the intermediate, which without fur-
ther purification was re-dissolved in 80% AcOH (250 cm3).
After stirring of the solution for 16 h and concentration in
vacuo, the residue was coevaporated with toluene (3 × 200 cm3)
and purified by silica gel column chromatography (CH2Cl2–
MeOH 95:5, v/v) to give the product 5 (14.9 g, 75%) as a white
solid, δH (CDCl3) in accord with literature data;17 δC (CDCl3)
26.33, 26.40, 63.98, 70.36, 79.27, 80.38, 83.35, 103.46, 113.02,
116.56, 134.23.
Methyl 3,5-di-O-benzyl-3-C-vinyl-D-allofuranoside 8
A solution of 7 (3.87 g, 5.79 mmol) in methanolic HCl (20%
w/w; 77 cm3) and water (11 cm3) was stirred at room temper-
ature for 16 h. After neutralisation with NaHCO3(s), the solu-
tion was extracted with CH2Cl2 (2 × 200 cm3). The combined
extracts were washed with water (200 cm3) and then dried
(Na2SO4). The solvent was removed by distillation under
reduced pressure and the residue was purified by silica gel col-
umn chromatography (CH2Cl2–MeOH 99:1, v/v) to give the
anomeric mixture 8 (1.75 g, 76%) as a clear oil (Found: C, 68.91;
H, 6.78. C23H28O6 requires C, 68.98; H, 7.05%); δH (CDCl3) 2.12
(dd, J 7.7 and 5.2, 6-OH α), 2.24 (dd, J 8.4 and 4.9, 6-OH β),
2.98 (d, J 8.0, 2-OH β), 3.14 (d, J 10.0, 2-OH α), 3.47 (s, OCH3
β), 3.48 (s, OCH3 α), 3.44–3.55 (m, 5-H), 3.81–3.92 (m, 6-H),
4.08 (dd, J 7.9 and 3.6, 2-H β), 4.14 (dd, J 10.0 and 4.8, 2-H α),
4.34–4.69 (m, CH2Ph, 4-H), 4.91 (d, J 3.5, 1-H α), 4.94 (d, J 4.7,
1,2-O-Isopropylidene-6-O-triphenylmethyl-3-C-vinyl-ꢀ-D-
allofuranose 6
1-H β), 5.35–5.45 (m, CH᎐CH α), 5.42 (dd, J 11.1 and 1.1,
᎐
2
CH᎐CHH β), 5.55 (dd, J 17.5 and 1.2, CH᎐CHH β), 6.04
᎐
᎐
To a stirred solution of 5 (2.16 g, 8.77 mmol) in anhydrous
pyridine (21 cm3) was added chloro(triphenyl)methane (3.66 g,
13.1 mmol). The reaction mixture was stirred for 2 days at room
temperature and treated with another portion of triphenyl-
chloromethane (1.22 g, 4.4 mmol), and after being stirred for
2 h the mixture was quenched with water (50 cm3) and extracted
with CH2Cl2 (2 × 100 cm3). The combined extracts were washed
with saturated aq. NaHCO3 (150 cm3) and then dried (Na2SO4).
The solvent was removed by distillation under reduced pressure
and the residue was purified by silica gel column chromato-
graphy (CH2Cl2) to give the product 6 (4.02 g, 94%) as a white
solid (Found: C, 72.95; H, 6.65. C30H32O6ؒ0.25H2O requires C,
73.07; H, 6.64%); δH (CDCl3) 1.34 (3H, s, CH3), 1.59 (3H, s,
CH3), 2.45 (1H, d, J 3.6, 5-OH), 3.07 (1H, br s, 3-OH), 3.29–
3.32 (2H, m, 6-H2), 3.79 (1H, m, 5-H), 4.04 (1H, d, J 11.3, 4-H),
(dd, J 17.5 and 11.0, CH᎐CH α), 6.11 (dd, J 17.6 and 11.2,
᎐
2
CH᎐CH β), 7.26–7.36 (m, Ph); δ (CDCl ) 55.47, 56.74, 60.93,
᎐
2
C
3
61.34, 66.41, 66.52, 71.13, 71.51, 75.69, 78.01, 78.68, 79.07,
82.65, 83.01, 85.37, 101.76, 109.50, 116.99, 118.63, 126.88,
127.22, 127.48, 127.83, 127.86, 127.89, 127.94, 128.00, 128.30,
128.54, 128.57, 128.59, 132.68, 135.37, 137.70, 137.79, 138.05,
139.16; FAB-MS m/z 423 [M ϩ Naϩ].
Methyl 3,5-di-O-benzyl-2,6-bis(O-methylsulfonyl)-3-C-vinyl-D-
allofuranoside 9 and 10
To a stirred solution of 8 (2.41 g, 6.01 mmol) in anhydrous
pyridine (18 cm3) at 0 ЊC was added dropwise methanesulfonyl
chloride (2.28 cm3, 30.1 mmol). The reaction mixture was
stirred for 30 min at room temperature, quenched with ice-cold
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713
3711

View Article Online
water (200 cm3), and extracted with CH2Cl2 (2 × 250 cm3). The
combined extracts were washed with saturated aq. NaHCO3
(2 × 200 cm3) and then dried (MgSO4). The solvent was
removed by distillation under reduced pressure and the residue
was purified by silica gel column chromatography (CH2Cl2) to
give three fractions as clear oils: the pure β-anomer 9 (1.464 g,
44%), the pure α-anomer 10 (1.302 g, 39%) and a mixture of
anomers (0.412 g, 12%).
used without further purification in the next step, δH (CDCl3)
3.43 (3H, s), 3.66–4.08 (6H, m), 4.40 (1H, m), 4.49 (1H, s), 4.52–
4.67 (4H, m), 4.98 (1H, s), 7.26–7.37 (10H, m); δC (CDCl3)
54.59, 59.48, 68.44, 68.51, 70.88, 71.09, 73.64, 78.91, 89.12,
91.01, 108.20, 127.08, 127.15, 127.81, 127.94, 128.04, 128.57,
137.83, 138.17; FAB-MS m/z 421 [M ϩ Naϩ].
(3R)- and (3S)-(1R,4S,7R,10R,11R)-10,11-Dibenzyloxy-3-
(thymin-1-yl)-2,5,8-trioxatricyclo[5.3.1.04,11]undecane 13
Methyl 3,5-di-O-benzyl-2,6-bis(O-methylsulfonyl)-3-C-vinyl-
ꢁ-D-allofuranoside 9. δH (CDCl3) 2.97 (3H, s), 3.01 (3H, s), 3.48
(3H, s), 3.67 (1H, m), 4.33–4.71 (7H, m), 5.08 (1H, d, J 3.1),
5.19 (1H, d, J 3.1), 5.51 (1H, d, J 11.8), 5.65 (1H, d, J 17.7), 6.05
(1H, dd, J 17.8 and 11.2), 7.22–7.34 (10H, m); δC (CDCl3)
37.59, 38.92, 56.68, 67.42, 68.02, 71.75, 76.81, 83.22, 83.47,
84.79, 106.53, 119.69, 127.08, 127.52, 128.03, 128.08, 128.44,
128.57, 137.18, 137.18, 138.59; FAB-MS m/z 579 [M ϩ Naϩ].
A mixture of 11 (100 mg, 0.251 mmol) and thymine (64 mg,
0.51 mmol) was dried and dissolved in anhydrous CH3CN (3.9
cm3). The mixture was treated with N,O-bis(trimethylsilyl)-
acetamide (BSA) (0.38 cm3, 1.5 mmol) and stirred under reflux
for 15 min. After cooling of the mixture to 0 ЊC, TMS triflate
(96 mm3, 0.48 mmol) was added dropwise and the solution was
stirred at room temperature for 24 h and then at 60 ЊC for
another 24 h. The reaction mixture was quenched with satur-
ated aq. NaHCO3 (10 cm3) and extracted with CH2Cl2 (3 × 15
cm3). The combined extracts were dried (Mg2SO4) and the
solvent was removed by distillation under reduced pressure.
The residue was purified by silica gel column chromatography
(CH2Cl2–MeOH 98:2, v/v) to give the product 13 as a clear oil
and a mixture of anomers (β:α ≈ 3:1; 64.4 mg, 52%) which was
used without further purification in the next step, δC (CDCl3)
12.10, 12.48, 63.58, 63.86, 67.16, 67.63, 70.94, 71.38, 71.52,
71.99, 72.87, 73.73, 74.53, 79.49, 80.17, 83.65, 83.70, 85.87,
92.60, 93.69, 97.75, 109.30, 110.46, 127.22, 127.37, 127.93,
128.05, 128.29, 128.31, 128.46, 128.57, 128.63, 128.70, 128.79,
136.97, 137.31, 138.01, 138.75, 139.77, 150.38, 150.49, 163.82,
164.13; FAB-MS m/z 493 [M ϩ Hϩ].
Methyl 3,5-di-O-benzyl-2,6-bis(O-methylsulfonyl)-3-C-vinyl-
ꢀ-D-allofuranoside 10. (Found: C, 53.82; H, 5.78. C25H32O10S2
requires C, 53.94; H, 5.79%); δH (CDCl3) 2.92 (3H, s), 3.05 (3H,
s), 3.51 (3H, s), 3.74 (1H, m), 4.31–4.74 (7H, m), 5.11 (1H, d,
J 4.5), 5.14 (1H, d, J 4.3), 5.46 (1H, s, J 17.6), 5.51 (1H,
d, J 11.3), 5.99 (1H, dd, J 17.8 and 11.1), 7.23–7.34 (10H, m);
δC (CDCl3) 37.64, 39.27, 56.22, 67.43, 68.58, 72.09, 76.07, 78.96,
80.72, 83.35, 101.27, 119.14, 127.43, 127.56, 128.17, 128.32,
128.46, 128.67, 137.43, 137.43, 138.78; FAB-MS m/z 579
[M ϩ Naϩ].
(1R,3R,4S,7R,10R,11R)-10,11-Dibenzyloxy-3-methoxy-2,5,8-
trioxatricyclo[5.3.1.04,11]undecane 11
To a solution of 9 (1.43 g, 2.57 mmol) in tert-butyl alcohol (28.6
cm3) were added N-methylmorpholine N-oxide (2.10 g, 18.1
mmol), pyridine (1.44 cm3, 18.8 mmol), water (1.56 cm3) and a
2.5 w/w% solution of OsO4 in tert-butyl alcohol (0.144 cm3).
The solution was stirred under reflux at 80 ЊC for two days and
quenched at room temperature with 20% aq. sodium bisulfite
(10.5 cm3). The mixture was diluted with water (100 cm3) and
extracted with CH2Cl2 (2 × 300 cm3). The combined extracts
were washed with saturated aq. NaHCO3 (200 cm3) and then
dried (MgSO4). The solvent was removed by distillation under
reduced pressure and the residue was purified by silica gel col-
umn chromatography (CH2Cl2–MeOH 98:2, v/v) to give a clear
oil (1.152 g), which was dried and dissolved in anhydrous DMF
(7 cm3). This solution was stirred at 0 ЊC and a 60% oily
dispersion of NaH (300 mg, 7.2 mmol) was added. The mixture
was stirred at room temperature for four days. The reaction
mixture was quenched with water (100 cm3), saturated with
NaCl, and extracted with CH2Cl2 (2 × 200 cm3). The combined
extracts were dried (MgSO4) and the solvent was removed by
distillation under reduced pressure. The residue was purified
by silica gel column chromatography (CH2Cl2–MeOH 99:1,
v/v) to give the product 11 (490 mg, 48%) as a clear oil which
was used without further purification in the next step,
δH (CDCl3) 3.66 (3H, s), 3.40–4.19 (6H, m), 4.41 (1H, d, J 4.2),
4.51–4.77 (5H, m), 4.92 (1H, d, J 4.0), 7.26–7.39 (10H, m);
δC (CDCl3) 58.02, 58.36, 67.76, 68.44, 71.21, 71.42, 73.07, 77.10,
84.15, 89.55, 104.79, 127.15, 127.92, 127.97, 128.03, 128.56,
128.62, 137.65, 137.93; FAB-MS m/z 421 [M ϩ Naϩ].
(3R)- and (3S)-(1R,4S,7R,10R,11R)-10,11-Dihydroxy-3-
(thymin-1-yl)-2,5,8-trioxatricyclo[5.3.1.04,11]undecane 3 and 14
A solution of 13 (62 mg, 0.13 mmol) in EtOH (1.0 cm3) was
stirred at room temperature and 20% Pd(OH)2–C (50 mg) was
added. The mixture was degassed with argon and placed in a H2
atmosphere. After being stirred for 36 h the mixture was directly
purified by silica gel column chromatography (CH2Cl2–MeOH
96:4, v/v) to give the two products 3 (21.8 mg, 55%) and 14 (7.1
mg, 18%) respectively.
(1R,3R,4S,7R,10R,11R)-10,11-Dihydroxy-3-(thymin-1-yl)-
2,5,8-trioxatricyclo[5.3.1.04,11]undecane 3. δH (CD3OD) 1.85
(3H, d, J 0.9, CH3), 3.72 (1H, dd, J 11.2 and 8.0, 9-HЉ), 3.90
(1H, dd, J 3.2 and 1.2, 7-H), 3.95 (1H, d, J 9.8, 6-HЉ), 4.01 (1H,
dd, J 10.0 and 3.2, 6-HЈ), 4.02 (1H, dd, J 10.6 and 7.6, 9-HЈ),
4.04 (1H, d, J 3.1, 1-H), 4.22 (1H, td, J 7.7 and 3.0, 10-H), 4.39
(1H, d, J 4.8, 4-H), 6.11 (1H, d, J 4.6, 3-H), 7.69 (1H, d, J 1.1,
thymine 6-H); δC (CD3OD) 12.44 (CH3), 65.29 (C-10), 66.75
(C-9), 75.70 (C-6), 80.89 (C-1), 83.11 (C-7), 85.08 (C-3), 87.92
(C-11), 89.58 (C-4), 109.48 (thymine C-5), 141.06 (thymine
C-6), 152.23 (thymine C-2), 166.61 (thymine C-4); FAB-MS
m/z 313 [M ϩ Hϩ]; FAB-HRMS Found: m/z, 313.1036.
C13H17N2O7ϩ requires m/z, 313.1031.
(1R,3S,4S,7R,10R,11R)-10,11-Dihydroxy-3-(thymin-1-yl)-
2,5,8-trioxatricyclo[5.3.1.04,11]undecane 14. δH (CD3OD) 1.88
(3H, s, CH3), 3.69 (1H, dd, J 10.4 and 9.1, 9-HЉ), 3.83 (1H, dd,
J 10.8 and 6.6, 9-HЈ), 3.97 (1H, m, 7-H), 4.04–4.07 (2H, m,
6-H2), 4.08 (1H, m, 10-H), 4.53 (1H, d, J 2.5, 1-H), 4.72 (1H, d,
J 1.7, 4-H), 5.68 (1H, d, J 1.9, 3-H), 7.42 (1H, d, J 0.8, thymine
6-H); δC (CD3OD) 12.80, 65.11, 65.91, 74.04, 81.89, 85.27,
88.63, 94.25, 95.74, 111.67, 139.52, 155.10, 170.05; FAB-MS
m/z 313 [M ϩ Hϩ].
(1R,3S,4S,7R,10R,11R)-10,11-Dibenzyloxy-3-methoxy-2,5,8-
trioxatricyclo[5.3.1.04,11]undecane 12
The same procedure as for preparation of 11 was used with
precursor 10 (1.28 g, 2.29 mmol), tert-butyl alcohol (25.5 cm3),
N-methylmorpholine N-oxide (1.87 g, 16.1 mmol), pyridine
(1.28 cm3, 16.1 mmol), water (1.39 cm3), a 2.5 w/w% solution
of OsO4 in tert-butyl alcohol (0.128 cm3), anhydrous DMF
(6 cm3), and a 60% oily dispersion of NaH (250 mg, 6.0 mmol)
to give the product 12 (382 mg, 42%) as a clear oil which was
ab initio Calculations
All ab initio quantum mechanical calculations were performed
using the Gaussian94 program.25 Geometry optimisations
3712
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713

View Article Online
6 (a) M. Tarköy and C. Leumann, Angew. Chem., Int. Ed. Engl., 1993,
were carried out at either the 3-21G* or 6-31G* levels using
the restricted Hartree–Fock procedure. Constrained geometry
optimisations were obtained by constraining the appropriate
torsion angles with the AddRedundant option. No restraints
were placed upon the glycosidic linkage χ between the furanose
ring and the nucleobase in any calculations.
32, 1432; (b) M. Bolli, H. U. Trafelet and C. Leumann, Nucleic Acids
Res., 1996, 24, 4660; (c) R. Meier, S. Grüschow and C. Leumann,
Helv. Chim. Acta, 1999, 82, 1813.
7 M. Bolli, J. C. Litten, R. Schütz and C. J. Leumann, Chem. Biol.,
1996, 3, 197.
8 (a) R. Steffens and C. J. Leumann, J. Am. Chem. Soc., 1997, 119,
11548; (b) R. Steffens and C. J. Leumann, J. Am. Chem. Soc., 1999,
121, 3249.
9 M. A. Siddiqui, H. Ford Jr., C. George and V. E. Marquez,
Nucleosides Nucleotides, 1996, 15, 235.
Measurement of three-bond coupling constants
1
1D H NMR spectra of the nucleosides studied were acquired
10 (a) N. C. Bar, R. Patra, B. Achari and S. B. Mandal, Tetrahedron,
1997, 53, 4727; (b) Q. Chao and V. Nair, Tetrahedron, 1997, 53, 1957;
(c) J. H. Hong, B. K. Chun and C. K. Chu, Tetrahedron Lett., 1998,
39, 225; (d) S. Obika, J. Andoh, T. Sugimoto, K. Miyashita and
T. Imanishi, Tetrahedron Lett., 1999, 40, 6465; (e) H. R. Moon,
H. O. Kim, M. W. Chun, L. S. Jeong and V. E. Marquez, J. Org.
Chem., 1999, 64, 4733; ( f ) C. Lescop and F. Huet, Tetrahedron,
2000, 56, 2995.
on a Varian Unity 500 MHz spectrometer. The nucleosides 3
and 14 were dissolved in CD3OD and spectra were obtained
in the temperature range from Ϫ50 to ϩ50 ЊC. Coupling con-
stants were measured as the splitting of multiplet components,
thereby limiting the accuracy to within 10% of the linewidth
(≈0.1 Hz).
11 V. E. Marquez, A. Ezzitouni, P. Russ, M. A. Siddiqui, H. Ford Jr.,
R. J. Feldman, H. Mitsuya, C. George and J. J. Barchi Jr., J. Am.
Chem. Soc., 1998, 120, 2780.
12 V. E. Marquez, P. Russ, R. Alonso, M. A. Siddiqui, S. Hernandez,
C. George, M. C. Nicklaus, F. Dai and H. Ford Jr., Helv. Chim.
Acta, 1999, 82, 2119.
Karplus relationships
3
Karplus relationships correlating JHH and torsion angles were
constructed employing a state-of-the-art generalised Karplus
equation [eqn. (1)] for nucleosides and nucleotides developed
13 (a) M.-O. Bévierre, A. De Mesmaeker, R. M. Wolf and S. M. Freier,
Bioorg. Med. Chem. Lett., 1994, 4, 237; (b) H. Urata, H. Miyagoshi,
T. Yumoto and M. Akagi, J. Chem. Soc., Perkin Trans. 1, 1999,
1833.
14 For definitions and nomenclature of nucleoside conformations and
torsions see W. Saenger, Principles of Nucleic Acid Structure,
Springer, New York, 1984.
15 G. Minasov, M. Teplova, P. Nielsen, J. Wengel and M. Egli,
Biochemistry, 2000, 39, 3525.
16 L. B. Jørgensen, P. Nielsen, J. Wengel and J. P. Jacobsen, J. Biomol.
Struct. Dyn., 2000, 18, 45.
3
3
(1)
JHH(θ) = ͚ Cm cos(mθ) ϩ ͚ Sn sin(nθ)
m = 0
n = 1
by Altona and co-workers,20 where the electronegativity of
the HCCH-fragment substituents is accounted for in the co-
efficients Cm and Sn. Altona and Sundaralingam21 have
developed a simple relationship relating the five ring torsion
angles of the furanose ring, θi, to the pseudorotation angle, P,
and the puckering amplitude, Φmax [eqn. (2)]. Thus, for a given
17 The synthesis of the 3Ј-C-vinyl compound 5 has been reported
before via an ethynyl Grignard addition and subsequent reduction
of the corresponding 3Ј-C-ethynyl compound [(a) D. C. Baker,
D. K. Brown, D. Horton and R. G. Nickol, Carbohydr. Res., 1974,
32, 299] followed by selective deprotection [(b) J. Marco-Contelles,
P. Ruiz, L. Martínez and A. Martínez-Grau, Tetrahedron, 1993, 49,
6669], as well as via a vinyl Grignard addition published without
experimental details [(c) J. M. J. Tronchet, M. Zsely, R. N. Yazji,
F. Barbalat-Rey and M. Geoffroy, Carbohydr. Lett., 1995, 1,
343].
θi = Φmax cos[P ϩ 144(i Ϫ 2)], i = 1, . . . 5
(2)
value of Φmax, we can deduce the value of θH1ЈH2Ј depending on
P and hence the Karplus relationship between JH1ЈH2Ј and P.
Acknowledgements
The Danish Natural Science Research Council is thanked for
financial support.
18 J. K. Cha and N.-S. Kim, Chem. Rev., 1995, 95, 1761.
19 D. B. Davies, Prog. Nucl. Magn. Reson. Spectrosc., 1978, 12,
135.
20 (a) L. A. Donders, F. A. A. M. de Leeuw and C. Altona, Magn.
Reson. Chem., 1989, 27, 556; (b) C. Altona, J. H. Ippel, A. J. A.
Westra Hoekzema, C. Erkelens, M. Groesbeek and L. A. Donders,
Magn. Reson. Chem., 1989, 27, 564; (c) C. Altona, R. Francke,
R. de Haan, J. H. Ippel, G. J. Daalmans, A. J. A. Westra Hoekzema
and J. van Wijk, Magn. Reson. Chem., 1994, 32, 670.
21 C. Altona and M. Sundaralingam, J. Am. Chem. Soc., (a) 1972, 94,
8205; (b) 1973, 95, 2333.
22 F. Jensen, Introduction to Computational Chemistry, Wiley,
Chichester, 1999.
23 B. Cadioli, E. Gallinella, C. Coulombeau, H. Jobic and G. Berthier,
J. Phys. Chem., 1993, 97, 7844.
24 (a) H. Ikeda, R. Fernandez, A. Wilk, J. J. Barchi Jr., X. Huang and
V. E. Marquez, Nucleic Acids Res., 1998, 26, 2237; (b) I. Berger,
V. Tereshko, H. Ikeda, V. E. Marquez and M. Egli, Nucleic Acids
Res., 1998, 26, 2473; (c) M. J. Damha, C. J. Wilds, A. Noronha,
I. Brukner, G. Borkow, D. Arion and M. A. Parniak, J. Am. Chem.
Soc., 1998, 120, 12976; (d) C. J. Wilds and M. J. Damha, Nucleic
Acids Res., 2000, 28, 3625.
References
1 (a) P. Herdewijn, Liebigs Ann., 1996, 1337; (b) P. Herdewijn,
Biochim. Biophys. Acta, 1999, 1489, 167.
2 (a) P. Nielsen, H. M. Pfundheller and J. Wengel, Chem. Commun.,
1997, 825; (b) P. Nielsen, H. M. Pfundheller, C. E. Olsen and
J. Wengel, J. Chem. Soc., Perkin Trans. 1, 1997, 3423.
3 N. K. Christensen, M. Petersen, P. Nielsen, J. P. Jacobsen, C. E.
Olsen and J. Wengel, J. Am. Chem. Soc., 1998, 120, 5458.
4 (a) S. K. Singh, P. Nielsen, A. A. Koshkin and J. Wengel, Chem.
Commun., 1998, 455; (b) A. A. Koshkin, S. K. Singh, P. Nielsen,
V. K. Rajwanshi, R. Kumar, M. Meldgaard, C. E. Olsen and
J. Wengel, Tetrahedron, 1998, 54, 3607; (c) S. Obika, D. Nanbu,
Y. Hari, J. Andoh, K. Morio, T. Doi and T. Imanishi, Tetrahedron
Lett., 1998, 39, 5401; (d) A. A. Koshkin, P. Nielsen, M. Meldgaard,
V. K. Rajwanshi, S. K. Singh and J. Wengel, J. Am. Chem. Soc.,
1998, 120, 13252; (e) V. K. Rajwanshi, A. E. Håkansson, M. D.
Sørensen, S. Pitsch, S. K. Singh, R. Kumar, P. Nielsen and J. Wengel,
Angew. Chem., Int. Ed., 2000, 39, 1656.
5 (a) K.-H. Altmann, R. Kesselring, E. Francotte and G. Rihs,
Tetrahedron Lett., 1994, 35, 2331; (b) K.-H. Altmann, R.
Imwinkelried, R. Kesselring and G. Rihs, Tetrahedron Lett., 1994,
35, 7625; (c) V. E. Marquez, M. A. Siddiqui, A. Ezzitouni, P. Russ,
J. Wang, R. W. Wagner and M. D. Matteucci, J. Med. Chem., 1996,
39, 3739; (d) G. Wang, E. Gunic, J.-L. Girardet and V. Stoisavljevic,
Bioorg. Med. Chem. Lett., 1999, 9, 1147; (e) V. K. Rajwanshi,
R. Kumar, M. Kofod-Hansen and J. Wengel, J. Chem. Soc., Perkin
Trans. 1, 1999, 1407; ( f ) M. Raunkjær, C. E. Olsen and J. Wengel,
J. Chem. Soc., Perkin Trans. 1, 1999, 2543.
25 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G.
Johnson, M. A. Robb, J. R. Cheeseman, T. A. Keith, G. A.
Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-Laham,
V. G. Zakrewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B.
Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y.
Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle,
R. Comperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees,
J. Baker, J. P. Stewart, M. Head-Gordon, C. Gonzalez and J. A.
Pople, Gaussian94 (Rev D.3), Gaussian Inc., Pittsburgh, PA,
1995.
J. Chem. Soc., Perkin Trans. 1, 2000, 3706–3713
3713