Oligonucleosides with a nucleobase-including backbone: Part 7. syn and anti conformations of a (5′-8)-ethynediyl-linked adenosine dimer

Article abstract of DOI:10.1002/1522-2675(200203)85:3<699::AID-HLCA699>3.0.CO;2-N

The conformational analysis of 7 was carried out in (D₆)DMSO and in mixtures of (D₆)DMSO and CDCl₃ to evaluate the syn/anti conformations, relevant to the pairing propensity of this type of nucleotide analogue. The HO-C(5′

Full text of DOI:10.1002/1522-2675(200203)85:3<699::AID-HLCA699>3.0.CO;2-N

Helvetica Chimica Acta ± Vol. 85 (2002)
699
Oligonucleosides with a Nucleobase-Including Backbone
Part 7
syn and anti Conformations of a (5' ± 8)-Ethynediyl-Linked Adenosine Dimer
by Punit Kumar Bhardwaj and Andrea Vasella*
Laboratorium f¸r Organische Chemie, ETH-Hˆnggerberg, HCI, CH-8093 Z¸rich
The conformational analysis of 7 was carried out in (D
to evaluate the syn/anti conformations, relevant to the pairing propensity of this type of nucleotide analogue.
)DMSO,
6 6 3
)DMSO and in mixtures of (D )DMSO and CDCl
The HOÀC(5') of unit I and of unit II of the dimer 7 form an intramolecular H-bond to N(3). In (D
6
the C(5')ÀOH ¥¥¥ N(3) H-bond in unit I is partially broken, while that in unit II persists to a larger extent. The
syn conformation prevails for unit I and particularly for unit II. The furanosyl moieties adopt predominantly a 2'-
endo conformation that is largely independent of the solvent.
Introduction. ± In spite of extensive modifications of oligonucleotides, all known
analogues are of type A (Fig.) i.e., they maintain the structural entities of backbone and
nucleobase [1]. This has led to the question of whether this structural invariant is a
necessary prerequisite for the formation of homo and/or hetero duplexes. To answer
this question, we have designed oligonucleotide analogues of type B (Fig.) that do not
1
differentiate between backbone and nucleobase ). The representatives 1 and 2 of this
type of oligonucleotide analogues (Fig.) were considered capable of forming
2
duplexes ). An ethynediyl-linked uridine-derived hexamer 1 (n ˆ 4) [2] and an
analogous adenosine-derived tetramer 2 (n ˆ 2) [3] have been synthesized. These
oligomers did not pair; they may be too short. Before synthesizing longer oligomers, we
wanted to analyse the conformations of a simple representative of 2. The preferred
conformation of oligonucleotides is an essential factor determining their pairing ability;
and the preferred conformation of polynucleotides corresponds to that of the related
mononucleosides and mononucleotides [4 ± 8].
The anti conformation [9] is a prerequisite for pairing of DNA and RNA [10± 12]; it
appears to be also required for the pairing of type 1 and 2 oligomers of C(8)-substituted
purine and C(6)-substituted pyrimidine units [2]. Although it is known that at least
some C(8)-substituents on purines and C(6)-substituents on pyrimidines induce a syn
conformation [13], we speculated that the sterically undemanding alkynyl substituent
may be compatible with a significant population of an anti conformation. It has been
reported that a single 2'-deoxy-8-(prop-1-yn-1-yl)adenosine moiety in an oligodeox-
ynucleotide hybridises with the corresponding base in the complementary strand, albeit
(
in this case) not as well as the unmodified adenosine moiety [14]. This suggests that
1)
2)
For the sake of simplicity, we have designated these analogues as −oligonucleotide analogues with a
nucleobase-including backbone×, while, strictly speaking, these new systems do not possess a −backbone×.
According to Maruzen model studies and Macromodel calculations using the Amber* force-field [2].

7
00
Helvetica Chimica Acta ± Vol. 85 (2002)
NH2
N
O
NH2
N
O
N
N
HN
HN
N
N
N
N
N
N
N
N
N
N
O
O
NH2
N
O
NH2
N
O
N
N
HN
HN
N
N
O
O
A
B
O
N
O
N
O
HN
HN
NH
OH
OH
O
O
N
O
HO
O
O
O
HO OH
HO OH
HO OH
n
1
NH2
N
NH2
N
NH2
N
N
N
N
HO
HO
N
N
N
N
N
HO
N
O
O
O
HO OH
HO OH
HO OH
n
2
Figure. Schematic representation of backbone and nucleobases of known oligonucleotides and analogues (A),
oligonucleotides with a nucleobase-including backbone (B), and ethynediyl-bridged uridine (1) and adenosine
(
2) oligomers with a nucleobase-including backbone
such a C(8)-alkynylated adenosine moiety can adopt a (somewhat disfavoured) anti
conformation for pairing. The propensity for hybridisation is also influenced by the
conformation of the ribosyl ring and the orientation of the substituent at C(4') [15][16].
The rotation about the C(1')ÀN(9) and C(4')ÀC(5') bonds is influenced by the
intramolecular H-bond from HOÀC(5') to N(3) of purines [17 ± 24]. This H-bond is
incompatible with the anti/gt conformation that appears to be required for pairing of
analogues 1 and 2 [2].
The increased acidity of the propargylic HOÀC(5') group in monomeric C(5')-
alkynylated analogues of adenosine favours the C(5')ÀOH ¥¥¥ N(3) H-bond [17][24].
To see if this H-bond is similarly favoured in analogues of type 2, we intended to
analyse the conformation of the dimer 7 (Scheme), the silylated analogue of the
simplest representative of the oligoadenosines 2. The dimer 7 has the advantage of
combining C(8)-substituted and C(8)-unsubstituted adenosyl moieties (units I and II),

Helvetica Chimica Acta ± Vol. 85 (2002)
701
Scheme
NH2
N
N
HO
Me3Si
a), b)
N
I
N
O
NHBz
N
O
O
N
N
HO
Me3Si
N
4
O
d)
6
NH2
O
O
N
O
N
HO
H
N
N
3
c)
O
O
5
NH2
NH2
NH2
NH2
N
N
N
N
N
N
N
N
HO
HO
HO
HO
N
N
N
Me3Si
N
N
R
R
N
N
N
O
O
O
O
e)
O
O
OH OH
HO HO
O
O
II
I
6
7 R = Me3Si
R = H
f)
8
NHBz
OH
N
N
HO
CH3
N
Me3Si
R
N
O
CH3
Me3Si
9
O
O
1
0 R = H
1
1 R = Cl
a) LDA, THF; then NIS. b) MeNH
2
, THF; 75%. c) THF, MeNH
2
; Bu
4
NF, THF; 70%. d) [Pd
2
(dba)
3
], CuI,
P(fur) , toluene/Et N 1:1; 86%. e) CF
3
3
3
CO H/H O 1:2, THF; 88%. f) Bu
2
2
4
NF ¥ 3 H O, THF; 54%.
2
the unsubstituted adenosyl moiety corresponding to unit I, allowing assessment of the
effect of the C(8)-substituent of unit II.
Results and Discussion. ± 1. Synthesis. The dimer 7 (Scheme) was prepared from
adenosine via 3 [24], which served as common intermediate for the synthesis of the iodo
derivative 4 and the alkyne 5 (cf. [3][25]). Iodination of 3 followed by debenzoylation
provided 4 in 75% yield, while debenzoylation of 3 followed by desilylation yielded
70% of 5. Sonogashira coupling [25] of 4 and 5 to 6 (86%) followed by deprotection of

7
02
Helvetica Chimica Acta ± Vol. 85 (2002)
the product by treatment with trifluoroacetic acid gave 7 (88%) as a solid that decomposes
at 1808 and that is poorly soluble in H O and MeOH, but soluble in DMSO, mixtures of
2
CDCl and at least 21% of DMSO, and in a ca. 8 :2 mixture of CHCl and MeOH.
3
3
Treatment of 7 with tetrabutylammonium fluoride afforded the desilylated dimer 8
(
(
54%). The propargyl alcohol 9 was synthesized as described by Kitano et al. [26].
1
2
. Conformational Analysis. The H-NMR spectra of 7 were recorded in
D )DMSO and in CDCl containing increasing amounts of (D )DMSO. As described
6
3
6
below, the minimal content of (D )DMSO in CDCl ensuring solubility of 7 (21%) is
6
3
sufficiently low to allow detection and analysis of the intramolecular H-bonds of
interest, while higher amounts of DMSO are expected to disrupt weak H-bonds and to
impair H-bonds of medium strength [27][28].
1
The H-NMR signals of 7 in (D )DMSO (Table 1) were assigned as follows. The s×s
6
at 8.34, 8.16, and 8.13 ppm resonating at lowest fields were assigned to one of HÀC(8/
I), HÀC(2/I), or HÀC(2/II). Exchange with D O showed that the signals at 7.72, 7.38,
2
7
.02, 6.90, 5.52, 5.47, 5.39, and 5.38 ppm correspond to six OH and two NH groups, the
2
broad s−s at 7.72 and 7.38 integrating for two H each and thus corresponding to NH2
groups. Upon addition of D O, the HÀC ddd×s at 4.91, 4.75, 4.33, and 4.29 ppm changed
2
to dd×s and the broad HÀC t×s at 4.89 and 4.55 ppm to d×s. These findings and the
analogy to related monomers [24] evidence that the ddd×s at 4.91 (J ˆ 7.5, 6.5, and
5
.2 Hz) and 4.75 ppm (J ˆ 7.8, 6.5, and 5.4 Hz) correspond to HÀC(2'), those at 4.33
(
J ˆ 5.4, 3.2, and 1.0Hz) and 4.29 ppm ( J ˆ 5.3, 3.1, and 1.2 Hz) to HÀC(3'), and the t×s
at 4.89 and 4.55 ppm to HÀC(5'). The d×s at 5.95 (J ˆ 7.8 Hz) and 5.96 (J ˆ 7.5 Hz), and
the dd×s at 4.14 (J ˆ 4.3, 1.0) and 4.00 (J ˆ 3.4, 1.2) did not change upon addition of
D O; the former were assigned to HÀC(1') and the latter to HÀC(4') of the two units.
2
Table 1. 1H-NMR Chemical Shifts [ppm] and Coupling Constants [Hz] of 7 in Varying Concentrations of
(
D
6
)DMSO in CDCl
3
Amount of 21%
)DMSO
28%
40%
60%
80%
100%
(
D
6
Unit I Unit II Unit I Unit II Unit I Unit II Unit I Unit II Unit I Unit II Unit I Unit II
a
8.14 ) 8.13 a
a
8.19 ) 8.13 a
a
HÀC(2)
HÀC(8)
HÀC(1')
HÀC(2')
HÀC(3')
HÀC(4')
HÀC(5')
8.13
8.01
5.89
8.08
±
6.08
8.12
8.03
5.90
8.08
±
6.07
8.13 ) 8.11 a
)
)
)
8.16
8.34
5.95
8.13
±
5.96
a
a
a
8.07 ) ±
8.24 ) ±
8.26 ) ±
5.94
4.69
4.46
6.04
4.77
4.34
4.26
5.95
6.00
4.81
4.304.36
5.96
4.72
4.32
6.00
4.83
4.704.77
4.71
4.77
4.39
4.33
4.71
4.39
4.75
4.29
4.14
4.89
4.91
4.55
4.33
4.85
4.404.54
4.18
4.57
4.33
4.06
4.55
4.17
4.56
7.73
5.31
5.20
8.1
4.13
4.55
7.57
4.21
4.87
4.08
4.54
4.19
4.87
4.00
4.55
4.85
7.51
5.22
5.15
7.5
4.85
7.32
HOÀC(5') 7.65
7.75
7.16
7.307.12
7.21 2 7.06.90
b
b
b
b
b
b
b
b
b
HOÀC(2')
)
)
)
)
)
)
)
)
)
)
5.47
5.39
7.8
5.52
5.38
7.5
HOÀC(3') b)
b)
b
)
b
)
b
)
b
)
b
J(1',2')
7.2
8.1
5.3
7.5
5.3
7.8
7.5
7.8
5.2
7.8
7.8
b
J(2',3')
J(3',4')
J(4',5')
J(5',OH)
J(2',OH)
J(3',OH)
)
5.5
5.6
5.5
5.2
5.05.6
5.4
1.2 c
4.3
5.3
1.01.2
3.4
b)
b
)
1.0 c) 1.0 c) 1.2
1.0 c
)
1.5 c
)
1.2 c
)
1.2 c
2.8
)
)
2.01.9
2.01.9
2.2
2.8
1.5
7.2
3.5
2.2
3.1
2.5
3.4
b
b
b
)
)
)
)
3.4
1.9
4.4
2.5
4.4
2.5
4.7
6.5
3.2
3.1
6.5
3.1
b
b
b
b
b
b
b
6.5
3.7
)
)
)
)
)
)
)
b)
b)
b
b
)
b
)
b
)
b
)
b
)
a) Assignments may be interchanged, b) Not determined, c) The peak width at half height was taken.

Helvetica Chimica Acta ± Vol. 85 (2002)
703
The HÀC(1') d at 5.95 ppm couples with the HÀC(2') ddd at 4.75 ppm (J ˆ 7.8, 6.5, and
5
4
.4 Hz). Similarly, the HÀC(1') d at 5.96 ppm couples with the HÀC(2') ddd at
.91 ppm (J ˆ 7.5, 6.5, and 5.2 Hz). The DQFCOSY spectrum allowed to divide the
signals into two sets, each one corresponding to a unit of the dimer. At this stage, the
signals of 7 could not be assigned to a specific unit, while the assignment of the
HÀC(5') signals of 8 (the desilylated analogue of 7) to unit I or II is unambiguous,
considering that only the HÀC(5') signal (ddd) at 4.55 ppm but not that at 4.89 ppm
(
br. t) shows a propargylic coupling (2.5 Hz). The assignment of the two sets of signals
to the individual units of 7 is based on the expectation that the relative chemical-shift
values will not be affected by desilylation. The small chemical-shift differences for 7
and 8 (Dd ꢀ 0.03 ppm) are in agreement with this expectation. The assignment of the
br. t at 4.89 ppm (J(5',OH) ˆ 4.7, J(4',5') ˆ 4.3 Hz) to HÀC(5'/I) of 7 and of the br. t at
4
.55 ppm (J(4',5') ˆ 3.4, J(5',OH) ˆ 3.1 Hz) to HÀC(5'/II), and the analogous assign-
ment of their coupling partners is corroborated by the ROESY spectra of 7 (see
below). The shift to lower fields for HÀC(5'/I) correlates with the electron-acceptor
effect of the C(7') adenosyl substituent (c.f. [25]). This identification of the HÀC(5')
signals allowed an unambiguous assignment of all CH and OH signals of the individual
d-allofuranosyl units of 7.
In the ROESY spectrum of 7 in (D )DMSO (Table 2), one finds cross-peaks
6
between the HOÀC(5'/I) d at 7.02 ppm and the s×s at 8.16 and 8.34 ppm, and similarly
between HOÀC(5'/II) (6.90ppm) and the s at 8.13 ppm. One also finds cross-peaks
between the s at 8.34 ppm and the HÀC(1'/I) d (5.95 ppm), the HÀC(2'/I) ddd
(
4.75 ppm), and the HÀC(3'/I) ddd (4.33 ppm), and finally cross-peaks between the
HÀC(1'/II) d at 5.96 ppm and both the HÀC(4'/I) dd (4.14 ppm) and the HÀC(3'/I)
ddd (4.33 ppm).
The cross-peak between HOÀC(5'/II) and the s at 8.13 ppm allows to assign this s
to HÀC(2/II). Similarly, the cross-peaks of HOÀC(5'/I) and those of the s at 8.34 ppm
show that the s×s at 8.16 and 8.34 ppm correspond to HÀC(2/I) and HÀC(8/I),
respectively. The cross-peaks between HÀC(8/I), and HÀC(1'/I) (5.95 ppm), HÀC(2'/
I) (4.75 ppm), and HÀC(3'/I) (4.33 ppm) confirm the assignment of the two sets of
signals to units I and II, respectively. The interaction between HOÀC(5'/I) and both
HÀC(2/I) and HÀC(8/I) evidences the syn and anti conformation for unit I of 7, in
keeping with the interactions of HÀC(8/I) with HÀC(1'/I), HÀC(2'/I), and HÀC(3'/
I).
The evidence for both the syn and the anti conformations raises the question about
their relative populations. The integrals of the cross-peaks correlate with the distance
between HOÀC(5') and HÀC(2) or HÀC(8), and with the population of the syn/anti
conformations. These H ¥¥¥ H distances depend also on the conformation of the
furanose ring and on the orientation of the C(4')-substituent. A comparison of the
J(1',2')/J(3',4') ratios [29 ± 32] for the solutions of 7 in (D )DMSO (7.8 (unit I) and 6.2
6
(
unit II)) and in 28% (D )DMSO in CDCl (7.5 (unit I) and 8.1 (unit II)) evidences
6
3
that both ribosyl rings of 7 prefer the 2'-endo (S) conformation. It also suggests that this
preference shows different solvent dependences for the two units, the 2'-endo
conformation of unit I being more strongly favoured in (D )DMSO than in 28%
6
(
D )DMSO in CDCl , while the 2'-endo conformation for unit II appears to be more
6 3
favoured for solutions in 28% (D )DMSO in CDCl than in (D )DMSO.
6
3
6

7
04
Helvetica Chimica Acta ± Vol. 85 (2002)
3 6
)DMSO in CDCl and in (D )DMSO, and the
Table 2. Cross-Peaks in the ROESY Spectrum of 7 in 30% (D
6
Integrals of Some Selected Cross-Peaks
Unit I
ROE with
HÀC(1')
HÀC(2')
HÀC(3')
HÀC(4')
HÀC(5')
HOÀC(5')
HÀC(8)
HÀC(2)
HOÀC(2')
HÀC(4'/I), HÀC(3'/I), HÀC(2'/I), HÀC(8/I)
HÀC(1'/I), HOÀC(5'/I), HÀC(8/I)
HÀC(5'/I), HOÀC(5'/I), HÀC(8/I), HÀC(1'/I), HÀC(1'/II)
HÀC(5'/I), HÀC(1'/I), HÀC(1'/II), HOÀC(5'/I)
HÀC(4'/I), HÀC(3'/I), HÀC(2'/I), HOÀC(5'/I), HÀC(8/I)
HÀC(4'/I), HÀC(3'/I), HÀC(2'/I), HÀC(8/I), HÀC(5'/I), HÀC(2/I)
HÀC(3'/I), HÀC(2'/I), HÀC(1'/I), HOÀC(5'/I), HÀC(5'/I)
HOÀC(5'/I)
a
)
HÀC(3'/I), HÀC(2'/I), HÀC(1'/I)
HOÀC(3') a)
HÀC(4'/I), HÀC(3'/I), HÀC(2'/I), HÀC(1'/I), HÀC(1'/II)
Unit II
ROE with
HÀC(1')
HÀC(5')
HÀC(2')
HÀC(3')
HÀC(4')
HOÀC(5')
HÀC(2)
HÀC(4'/II), HÀC(4'/I), HÀC(3'/II), HÀC(3'/I), HÀC(2'/II)
HOÀC(5'/II)
HÀC(1'/II), HOÀC(5'/II)
HOÀC(5'/II), HÀC(1'/II)
HÀC(1'/II), HOÀC(5'/II)
HÀC(4'/II), HÀC(3'/II), HÀC(2'/II), HÀC(5'/II), HÀC(2/II)
HOÀC(5'/II)
HOÀC(2')
HÀC(3'/II), HÀC(3'/I), HÀC(2'/II), HÀC(1'/II)
HOÀC(3') a)
HÀC(4'/II), HÀC(3'/II), HÀC(2'/II), HÀC(1'/II)
Cross-peaks
ROE (integral of the cross-peaks)
(
D
6
)DMSO in CDCl
3
(3 : 7)
6
(D )DMSO
HÀC(2/I) with HOÀC(5'/I)
HÀC(8/I) with HOÀC(5'/I)
HÀC(2/II) with HOÀC(5'/II)
HÀC(8/I) with HÀC(1'/I)
HÀC(8/I) with HÀC(2'/I)
HÀC(8/I) with HÀC(3'/I)
15.7
6.6
29.7
112.2
19.3
9.8
9.6
12.1
24.6
124.4
65.0
11.8
a) The cross-peaks for HOÀC(2'/I) and HOÀC(3'/II) overlap in 30% (D
6
3
)DMSO in CDCl and the cross-
peaks for HOÀC(3'/I) and HOÀC(3'/II) overlap in (D
6
)DMSO.
The relevant distances between HOÀC(5'/I) and either HÀC(2/I) or HÀC(8/I)
were estimated with the help of molecular modeling (Macromodel, Amber* force-field,
gas phase). For this, we subjected a model in a 2'-endo conformation deduced from the
coupling constants to energy minimisation, and checked for local minima by rotation
about the C(4')ÀC(5'), C(5')ÀO(5'), and C(1')ÀN(9) bonds. This led to a distance of
2
.83 ä between HOÀC(5'/I) and HÀC(2/I), and 1.85 ä between HOÀC(5'/I) and
HÀC(8/I) for the energetically preferred gg conformer. For a 1:1 syn > anti
equilibrium (all else being equal), one thus expects a larger integral for the cross-
peak between HOÀC(5'/I) and HÀC(8/I) than for the cross-peak between HOÀC(5'/
I) and HÀC(2/I).
3
The integrals (Table 2) ) for the cross-peaks of 7 in 30% (D )DMSO in CDCl are
6
3
6.6 for HOÀC(5'/I)/HÀC(8/I) and 15.7 for HOÀC(5'/I)/HÀC(2/I). Other integrals
3)
Lack of an appropriate reference cross-peak prevented volume integration.

Helvetica Chimica Acta ± Vol. 85 (2002)
705
are 112.2 for HÀC(1'/I)/HÀC(8/I), 19.3 for HÀC(2'/I)/HÀC(8/I), and 9.8 for
HÀC(3'/I)/HÀC(8/I); the integral is 29.7 for HOÀC(5'/II)/HÀC(2/II). These values
evidence a higher population of the syn conformer, and a higher population of the syn
conformer in unit II than in unit I. The larger integral for the HÀC(1'/I)/HÀC(8/I)
cross-peak than for the cross-peak of HÀC(8/I) with HÀC(2'/I) and HÀC(3'/I)
confirms that unit I adopts preferentially the syn conformation [33].
Assuming that the syn conformation is correlated with an intramolecular
C(5')ÀOH ¥¥¥ N(3) H-bond that is expected to be at least partially disrupted in
(
D )DMSO [27][28], we also measured the integrals of the cross-peaks for 7 in this
6
solvent. The expected shift of the syn > anti equilibrium for 7 upon changing the
solvent from 30% (D )DMSO in CDCl to (D )DMSO is evidenced by the change of
6
3
6
the integrals for the cross-peaks HOÀC(5'/I)/HÀC(8/I) (6.6 ! 12.1) and HOÀC(5'/I)/
HÀC(2/I) (15.7 ! 9.6). For solutions of 7 in (D )DMSO, an increase of the integrals
6
for the HÀC(1'/I)/HÀC(8/I) (112.2 ! 124.4) and HÀC(2'/I)/HÀC(8/I) (19.3 ! 65.0)
cross-peaks evidences a change in the preferred orientation of the base moiety about
the glycosidic bond, shortening the average distance between HÀC(8/I) and HÀC(1'/I)
and, particularly, between HÀC(8/I) and HÀC(2'/I). That the integral for the HÀC(8/
I)/HÀC(1'/I) (124.4) is larger than that for the HÀC(8/I)/HÀC(2'/I) (65.0) cross-peak
suggests that the syn conformation for unit I is (still) preferred for solutions of 7 in
(
D )DMSO [33], but much less so than for solutions in 30% (D )DMSO in CDCl .
6 6 3
The small difference between the integrals for the HOÀC(5'/II)/HÀC(2/II) cross-
peak (29.7 ! 24.6) in the two solvents evidences that unit II adopts predominantly the
syn conformation even in (D )DMSO.
6
To characterize the intramolecular C(5'/I)ÀOH ¥¥¥ N(3) H-bond, we analysed the
J(H,OH), d(OH), and Dd/DT values. There is ample evidence that a comparison of
1
these parameters for H-NMR spectra of alcohols in CDCl and in (D )DMSO allows
3
6
characterisation of intramolecular H-bonds [27][28].
The J(5',OH) values for 7 in (D )DMSO and in mixtures of CDCl and (D )DMSO
6
3
6
are listed in Tables 1 and 3. Their interpretation requires information about the
J(H,OH) value for a freely rotating OH group of an aliphatic propargylic alcohol, and
about the J(H,OH) value for a HOÀC(5') group of an appropriate model compound
with a completely persistent intramolecular H-bond to N(3). The former value is
J(H,OH) ˆ 5.6 Hz for 9, while 10 and 11 were taken as model compounds for purine
nucleosides with a persistent intramolecular C(5')ÀOH ¥¥¥ N(3) H-bond in CDCl [24].
3
The C(8)-unsubstituted 10 (J(5',OH) ˆ 1.5 Hz) was expected to be a suitable model
compound for unit I of 7, and the C(8)-Cl-substituted 11 (J(5',OH) ˆ 0.5 Hz) [34] was
similarly expected to be a suitable model for unit II of 7, leading to two scales for the
J(H,OH) values, one ranging from 1.5 to 5.6 Hz and the other one ranging from 0.5 to
5.6 Hz, corresponding to 100% and 0% persistence of the H-bond, respectively. The
persistencies of the C(5')ÀOH ¥¥¥ N(3) H-bonds derived by interpreting J(5',OH)
according to these scales are shown in Table 3. To check for the consistency of these
values, we correlated them with the conformation about the C(4')ÀC(5') bonds of 7, as
expressed by J(4',5'). For this, we calculated J(4',5') by taking into account the
population of the gg conformers corresponding to the persistency of C(5')ÀOH ¥¥¥ N(3)
H-bonds as deduced from J(5',OH) and the population of the conformations of the
non-H-bonded tautomers. These tautomers were assumed to adopt a 1:1:1 equilibrium

7
06
Helvetica Chimica Acta ± Vol. 85 (2002)
4
of the freely rotating gg/gt/tg conformers ) (J(gg) ˆ 2.1, J(gt) ˆ 2.1 and J(tg) ˆ
5
9
.3 Hz) ). The values for J(4',5') correlating with the C(5')ÀOH ¥¥¥ N(3) H-bond were
1
taken from the H-NMR spectra of 10 (1.5 Hz) and 11 (2.0Hz). The difference
between these values evidences the influence of the C(8)ÀCl substituent on the
conformation of the H-bonded tautomer of these nucleosides.
Table 3. Persistency of H-Bonds in 7, as Deduced from J(HÀC(5', OH). Correlation of the persistency with the
C(4')ÀC(5') gg conformation (c.f. text for the calculation of J(4',5')).
With 10 as model compound
(
D
6
)DMSO
Persistency [%]
Unit I Unit II
J(5'/,OH/I) J(5',OH/II) J(4',5'/I)
Observed Calculated a
J(4',5'/II)
in CDCl [%]
3
Calculated b
)
)
Observed Calculated
2
4
6
8
8
0
0
0
83
54
30
30
22
100
903.4
76
76
61
2.2
1.5
2.0
2.0
2.1
2.2
4.1
4.1
4.3
1.9
1.5
1.9
2.8
2.9
3.1
1.8
4.4
4.4
4.7
2.5
2.5
3.1
3.1
3.4
4.3
3.5
3.5
3.8
2.5
2.8
3.4
2.2
2.2
2.6
1
(
00
With 11 as model compound
D
6
)DMSO
[%]
Persistency [%]
Unit I Unit II
J(5'/,OH/I) J(5',OH/II) J(4',5'/I)
J(4',5'/II)
in CDCl
3
Observed Calculated Observed Calculated c
Calculated b
) )
2
4
6
8
8
67
80
72
61
61
50
2.2
3.4
4.4
4.4
4.7
1.5
1.9
2.5
2.5
3.1
2.02.9
2.8
3.1
3.4
4.3
1.9
2.6
2.2
043
024
024
3.4
3.8
3.8
2.2
2.5
2.8
2.7
3.02.9
3.02.9
2.5
1
00
18
4.03.4
3.3
3.3
a) Calculated with J(4',5') ˆ 1.5 Hz for the H-bonded tautomer and a 1 : 1 : 1 ratio (gg/gt/tg) of the rotamers. b) Calculated with
c
J(4', 5') ˆ 1.5 Hz for the H-bonded tautomer and a 1 : 1 : 1.4 ratio (gg/gt/tg) of the rotamers. ) Calculated with J(4',5') ˆ 2.0Hz for
the H-bonded tautomer and a 1 : 1 : 1 ratio (gg/gt/tg) of the rotamers.
The results of these calculations are listed in Table 3. As expected, use of 10 as
model compound leads to a smaller difference between the experimental and
calculated J(4',5') values for unit I; similarly, use of 11 as model compound leads to
a better correspondence between the experimental and calculated J(4',5') values for
unit II of 7. This evidences that the combination of the persistencies for units I and II
indicated in italics in Table 3 is more reliable than values derived from using a single
model compound. The correspondence between experimental and calculated J(4',5')
values for unit I (with 10 as model compound) is better for solutions in the less polar
solvent mixtures, while the correspondence for unit II is better for solutions in
(
(
D )DMSO. That J(4',5') (obs.) for unit I of 7 in (D )DMSO is larger than J(4',5')
calc.) suggests that the assumption of an equidistribution of rotamers for the non-H-
6
6
bonded tautomer of 7 has to be corrected in favour of a larger proportion of the tg
rotamer. Assuming a gg/gt/tg ratio of 1:1:1.4 results in a much smaller difference
between J(4',5') (obs.) and J(4',5') (calc.) for solutions in (D )DMSO, a good
6
4)
5)
While the gg and gt conformers are preferred over the tg conformers in the natural nucleosides, introduction
of the ethynyl substituent (d-allo-isomer) leads to the destabilization of the gg and gt conformers by
electronic (loss of stabilizing gauche interaction) and steric effects.
Using the extended Karplus equation of Altona and Sundaralingam [31][32][35] as implemented in the
Macromodel V 6.0programme.

Helvetica Chimica Acta ± Vol. 85 (2002)
707
correspondence of these values for 28% (D )DMSO in CDCl , but a worse correlation
6
3
for the intermediate solvent mixtures. An increasingly higher proportion of the tg
rotamers is also evidenced by the J(4',5') value for analogues of 10 and 11 possessing an
increasingly bulkier substituent at C(8) [24][25]. The coincidence of the observed and
calculated values for 7 in 28% (D )DMSO in CDCl reflects the appropriate choice of
6
3
10 as model. The choice of 11, which possesses a Cl instead of an alkynyl substituent at
C(8), as model compound for unit II is less fortunate, as suggested by the difference
between observed and calculated values for 7 in 28% (D )DMSO in CDCl . A better, if
6
3
not fully satisfactory, coincidence between observed and calculated values for J(4',5')
results by assuming J(4',5') ˆ 1.5 Hz for the H-bonded tautomer also of unit II, and a
similar distribution of rotamers as for unit I. Thus, the comparison of J(5',OH) and
J(4',5') values indicates that 10 is an appropriate model for unit I, that J(5',OH) is a
valid parameter for the persistency of the intramolecular H-bond in this unit, and that
the tg rotamer is favoured. It is not possible to deduce to what extent the difference
between the observed and calculated J(4',5') values for unit II of 7 is due to an
underestimation of the persistency of the intramolecular H-bond and/or the preference
of the tg rotamer. The Cl substituent at C(8) of 11 and the alkynyl substituent at C(8) of
7
lead to slightly different conformations of the H-bonded tautomers.
This interpretation agrees qualitatively with the solvent dependence of the chemical
shift (Tables 1 and 4). Remarkably, the largest change of the chemical shift for
HOÀC(5') of the two units upon increasing the concentration of (D )DMSO is for the
6
change from 28 to 40% for unit I, and from 80% to pure (D )DMSO for unit II, in
6
6
agreement with the more persistent C(5')ÀOH ¥¥¥ N(3) H-bond in unit II ).
Table 4. Difference of the Chemical Shift Values [ppm] of HOÀC(5') of Units I and II of 7 for Solutions in
(
D
6
3 6
)DMSO and in CDCl with Increasing Amounts of (D )DMSO and Correlation between the Change in the
Chemical Shifts [ppm] of HOÀC(5') upon Increasing the Amount of (D
6
)DMSO in CDCl
3
(
D
6
)DMSO in CDCl
3
[%]
Chemical shift of HOÀC(5')
D1d a)
D2d b)
Unit I
Unit II
Unit I
Unit II
2
2
4
6
8
1
8
0
0
0
7.65
7.51
7.32
7.16
7.12
7.02
7.75
7.73
7.57
7.30
7.21
6.90
0.1
0.22
0.25
0.14
0.09
0.14
0.19
0.16
0.04
0.10
0.02
0.16
0.27
0.09
0.31
1
00
À 0.12
a) D d ˆ d(HOÀC(5'/II)) À d(HOÀC(5'/I)). ) D d ˆ Change in the chemical shift upon increasing the amount
1
b
2
6 3
of (D )DMSO in CDCl .
The Dd/DT values for the two HOÀC(5') of 7 were determined for solutions in
pure (D )DMSO and in 28% (D )DMSO in CDCl and (Table 5). A reference value
6
6
3
for fully solvated propargylic alcohols was obtained from Dd/DT for HOÀC(3) of 9
26] in the same solvents (Table 5). The Dd/DT values for HOÀC(3) of 9 and for
[
6)
The difference between the chemical shifts (Table 4) for HOÀC(5') of these units is maximal at a
concentration of 40% (D )DMSO in CDCl ; this appears to be the concentration of (D )DMSO that
induces the largest difference in the persistency of intramolecular H-bonds of these OH groups.
6
3
6

7
08
Helvetica Chimica Acta ± Vol. 85 (2002)
HOÀC(5'), HOÀC(3'), and to a lesser extent also HOÀC(2') of 7 in (D )DMSO, are
6
similar (Table 5), implying that HOÀC(5') is mostly solvated in (D )DMSO. This is at
6
variance with the results discussed above, suggesting that Dd/DT values are a more
complex parameter of intramolecular H-bonds than coupling constants and chemical-
shift values. The Dd/DT value for HOÀC(3) of 9 in 28% (D )DMSO in CDCl is
6
3
1
0.9 ppb/K, while Dd/DT values for HOÀC(3') and HOÀC(2') in this solvent mixture
are smaller (5.1 to 7.7, resp.); intermediate values characterize Dd/DT for HOÀC(5') in
this solvent mixture (8.8 ppb/K for unit I and 7.6 for unit II). This means that the large
Dd/DT for 9 reflects the solvent effect on a propargylic OH group, while the Dd/DT
values for HOÀC(5') evidence partially persistent C(5')ÀOH ¥¥¥ N(3) H-bonds for 7 in
this solvent mixture. Again, a higher persistency is deduced for the H-bond in unit II
than in unit I.
Table 5. Dd/DT Values [ppb/K] for HOÀC(5'), HOÀC(2') and HOÀC(3') of 7 and HOÀC(3) of 9 in CDCl
)DMSO, and in a Mixture of 28% of (D )DMSO in CDCl
3
,
(
D
6
6
3
(
D
6
)DMSO Dd/DT Values
in
CDCl
7
9
3
[%]
HOÀC(5'/I) HOÀC(5'/II) HOÀC(2'/I) HOÀC(2'/II) HOÀC(3'/I) HOÀC(3'/II) HOÀC(3)
2.5
8.8
6.0
0
8
±
±
±
±
±
±
2
7.6
5.6
5.1
4.8
5.5
5.2
6.4
5.4
7.7
5.6
10.9
5.7
1
00
Force-field calculations (Macromodel, Amber*, gas phase) for unit I of the dimer 8,
starting with a structure possessing a C(5'/I)ÀOH ¥¥¥ N(3) H-bond, a 2'-endo
conformation and a HÀC(5'/I)ÀC(4'/I)ÀH dihedral angle of 628 (corresponding to
J(4',5') ˆ 2.1 Hz), constraining unit II suggest a c angle of ca. 508 for unit I,
corresponding to a syn conformation.
These observations and interpretations show that the syn conformation is strongly
preferred for 7, particularly for unit II that is relevant to the oligonucleotides of type 2.
They lead to an unfavourable prediction for the pairing abilities of these analogues.
We thank the Swiss National Science Foundation and F. Hoffmann-La-Roche AG, Basel, for generous
support and Dr. B. Bernet for numerous critical and useful comments.
Experimental Part
General. Solvents were distilled before use: THF from K/benzophenone, CH
were run under Ar. Qual. TLC: precoated silica-gel plates (Merck silica gel 60 F254); detection by spraying with
mostain× (400 ml of 10% aq. H SO , 20 g of (NH Mo O, 0.4 g of Ce(SO ) and heating. Flash
chromatography (FC): silica gel Merck 60 (0.04 ± 0.063 mm). Optical rotations: 1-dm cell at 258 and 589 nm. FT-
2 2 2
Cl from CaH . Reactions
−
2
4
4
)
6
7
O
24 ¥ H
2
4 2
)
1
13
IR: 1 ± 2% soln. in the indicated solvent. H- and C-NMR: at 200, 300, 400, or 500 MHz, and 50, 75, 100, or
25 MHz, resp. MS: matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-
TOF; indol-3-acrylic-acid (IAA), 0.05m in THF or a-cyano-4-hydroxycinnamic acid (CCA) 0.05m in MeCN/
EtOH/H O and 2,5-dihydroybenzoic acid (DHB) 0.05m in THF for high-resolution (HR) MALDI-MS.
-[6,7-Dideoxy-2,3-O-isopropylidene-7-C-(trimethylsily)-b-d-allo-hept-6-ynofuranosyl]-8-iodoadenine (4).
At À 788, a soln. of (i-Pr) NH (2.5 ml, 18 mmol, distilled from CaH ) in THF (20ml) was treated dropwise with
.6m BuLi in hexane (11.2 ml, 18 mmol), stirred for 15 min, warmed to 08 for 15 min, cooled to À 788, treated
dropwise with a soln. of 3 (1.3 g, 2.5 mmol) in THF (15 ml), stirred for 2.5 h, treated dropwise with a soln. of N-
iodosuccinimide (NIS; 3.1 g, 14 mmol) in THF (15 ml), stirred for 1.5 h, treated with sat. NH Cl soln. (10ml),
1
2
9
2
2
1
4

Helvetica Chimica Acta ± Vol. 85 (2002)
709
and allowed to warm to 238. After evaporation, a soln. of residue in AcOEt was washed with cold sat. aq.
NaHCO soln. and brine, dried (Na SO ), and evaporated. After filtration through a short pad of silica gel
hexane /AcOEt 1:1) and evaporation, a soln. of the residue in THF (20ml) was treated with 8 m MeNH soln.
1.75 ml, 14 mmol) in THF (20ml), stirred for 2 h, and evaporated. A soln. of the residue in AcOEt was washed
3
2
4
(
(
2
with H
solid. M.p. 1628 (dec.). R
IR (CHCl
368w, 1318w, 1285m, 1267m, 1252m, 1154w, 1093m, 1045w, 1003w, 970w, 959w, 90 w8 , 849m. 1H-NMR
300 MHz, CHCl ); 5.21 (t,
): 8.17 (s, HÀC(2)); 7.65 (s, HOÀC(5')); 6.00 (d, J ˆ 5.4, HÀC(1')); 5.95 (br. s, NH
J ˆ 5.4, HÀC(2')); 5.14 (dd, J ˆ 5.4, 1.0, HÀC(3')); 4.70( t, J ꢁ 1.7, HÀC(4')); 4.53 (d, J ˆ 1.7, HÀC(5')); 1.71, 1.42
2
O and brine, dried (Na
2
SO
4
), and evaporated. FC (hexane/AcOEt 1 :1) gave 4 (1.042 g, 75%). Off-white
2
5
f
(CHCl
3
/MeOH 95 :5) 0.37. [a]
D
ˆ À122.4 (c ˆ 1, CHCl
3 3
). UV (CHCl ): 266 (24000).
3
): 3522w, 3473w, 3410w, 3169w, 3007w, 2180w, 1634s, 1597w, 1580m, 1528w, 1485w, 1445w, 1419w,
1
(
3
2
1
3
(
1
2s, Me
23.62 (s, C(5)); 114.25 (s, Me
C(4')); 81.86 (d, C(2')); 80.88 (d, C(3')); 63.71 (d, C(5')); 27.84, 25.55 (2q, Me
2
C); 0.20 (s, Me
3
Si). C-NMR (75 MHz, CDCl
C); 101.75 (d, C(8)); 99.51 (s, C(6')); 96.18 (d, C(1')); 91.99 (s, C(7')); 87.27 (d,
C); À 0.22 (q, Me Si). HR-
3
): 154.95 (s, C(6)); 152.48 (d, C(2)); 149.65 (s, C(4));
2
2
3
‡
‡
MALDI-MS: 530.0718 ([M ‡ H] ; calc. 530.0721).
9
-(6,7-Dideoxy-2,3-O-isopropylidene-b-d-allo-hept-6-ynofuranosyl)adenine (5). At 258, a soln. of 3 (1.2 g,
.3 mmol) in THF (20ml) was treated with 8 m MeNH in THF (2.3 ml, 19 mmol), stirred for 2 h, and
evaporated. A soln. of the residue in THF (20ml) was treated with 1 m Bu NF in THF (2.8 ml, 2.8 mmol), stirred
for 2 h, and evaporated. A soln. of the residue in AcOEt (100 ml) was washed with H O and brine, dried
), and evaporated. FC (AcOEt/hexane/MeOH 10:9 :1) gave 5 (0.55 g, 70%). White solid. M.p. 2108
2
2
4
2
(
Na
2 4
SO
2
5
(dec). R
f
(CHCl
3
/MeOH 95 :5) 0.27. R
/MeOH 3 :2): 260(14628). IR (KBr): 3288 s, 3134s, 1676s, 1608s, 1576m, 1512w, 1474m,
423m, 1381m, 1341m, 1301m, 1283w, 1264m, 1222m, 1155w, 1103m, 1082m, 1068m, 991m, 967w, 938w, 90 9w ,
f
(AcOEt/hexane/MeOH 10:9 :1) 0.22. [ a]
D
ˆ À136.7 (c ˆ 0.525, CHCl
3
/
MeOH 3 :2). UV (CHCl
1
8
3
1
56w, 798m. H-NMR (300 MHz, (D
6
)DMSO): 8.24 (s, HÀC(8)); 8.15 (s, HÀC(2)); 7.32 (br. s, NH
2
); 6.26 (d,
J ˆ 5.3, HOÀC(5')); 6.12 (d, J ˆ 3.1, HÀC(1')); 5.26 (dd, J ˆ 3.1, 5.9, HÀC(2')); 5.04 (dd, J ˆ 6.2, 1.9, HÀC(3'));
4
.39 (dd, J ˆ 5.3, 2.2, HÀC(5')); 4.16 (dd, J ˆ 1.8, 5.3, HÀC(4')); 3.28 (d, J ˆ 2.2, HÀC(7')); 1.25, 1.37 (2s, Me
2
C).
6
)DMSO): 156.38 (s, C(6)); 152.85 (d, C(2)); 148.95 (s, C(4)); 139.79 (d, C(8)); 119.12 (s,
1
3
C-NMR (75 MHz, (D
C(5)); 113.26 (s, Me
2
C); 90.12 (d, C(1')); 88.18 (d, C(4')); 83.23 (d, C(2')); 82.94 (s, C(6')); 81.11 (d, C(3')); 76.39
‡
(
s, C(7')); 61.29 (d, C(5')); 26.97, 25.09 (2q, Me
2
C). HR-MALDI-MS: 332.1355 ([M ‡ H] ; calc. 332.1359).
9
-[6,7-Dideoxy-2,3-O-isopropylidene-7-C-(trimethylsilyl)-b-d-allo-hept-6-ynofuranosyl]adenin-8-yl-(8 ! 7')-
-(6,7-dideoxy-2,3-O-isopropylidene-b-d-allo-hept-6-ynofuranosyl)adenine (6). soln. of (0.15 g,
.45 mmol), 4 (0.31 g, 0.59 mmol), [Pd (dba) ] (16.2 mg, 0.018 mmol), CuI (8.6 mg, 0.045 mmol), and P(fur)
N/toluene 1:1 (5 ml) was stirred for 18 h at 238. Evaporation and FC
/MeOH 95 :5) gave 6 (0.288 g, 86%). Yellowish brown solid. M.p. 1808 (dec) . R (CHCl /MeOH 95 :5)
/MeOH 3 :2): 266 (12487), 245 (19483). IR (KBr):
9
0
A
5
2
3
3
(
(
6.2 mg, 0.027 mmol) in degassed Et
CHCl
3
3
f
3
2
5
0
3
1
.2. [a]
D
ˆ À106.8 (c ˆ 1, CHCl
3 3
/MeOH 3 :2). UV (CHCl
333m, 3188m, 2988m, 1650m, 1600m, 1577m, 1479w, 1418w, 1376m, 1330m, 1302m, 1250m, 1216m, 1156w,
1
089m, 970w, 850m, 798w. H-NMR (300 MHz, (D
6
)DMSO): 8.33 (s, HÀC(8/I)); 8.16 (s, HÀC(2/I), HÀC(2/
); 6.68 (d, J ˆ 5.8, HOÀC(5'/I)); 6.36 (d, J ˆ 5.2, HOÀC(5'/II)); 6.24 (d,
J ˆ 2.7, HÀC(1'/I)); 6.12 (d, J ˆ 3.2, irrad. at 5.45 ! s, HÀC(1'/II)); 5.45 (dd, J ˆ 3.2, 6.2, irrad. at 5.09 ! d, J ˆ
.2, HÀC(2'/II)); 5.41 (dd, J ˆ 2.7, 5.9, HÀC(2'/I)); 5.20( dd, J ˆ 6.0, 2.5, HÀC(3'/I)); 5.09 (dd, J ˆ 6.0, 2.3, irrad.
at 4.08 ! d, J ˆ 6.0, HÀC(3'/II)); 4.84 (t, J ˆ 6.0, HÀC(5'/I)); 4.42 (dd, J ˆ 5.3, 6.6, HÀC(5'/II)); 4.30( dd, J ˆ
.1, 6.2, irrad. at 4.84 ! d, J ˆ 2.1, HÀC(4'/I)); 4.08 (dd, J ˆ 2.2, 6.7, irrad. at 4.42 ! d, J ˆ 2.2, HÀC(4'/II)); 1.57,
2 2
II)); 7.66 (br. s, NH ); 7.37 (br. s, NH
3
2
1
1
3
.50, 1.36, 1.30 (4s, 2 Me
II)); 153.85, 152.90(2 d, C(2/I), C(2/II)); 148.87, 148.14 (2s, C(4/I) C(4/II)); 140.02 (d, C(8/I)); 132.37 (s, C(8/
II)); 119.18, 119.05 (2s, C(5/I), C(5/II)); 113.47, 113.36 (2s, 2 Me C); 105.51 (s, C(6'/II)); 95.27 (s, C(6'/I)); 90.45,
9.88 (2d, C(1'/I), C(1'/II)); 89.20( s, C(7'/II)); 87.98, 87.94 (2d, C(4'/I), C(4'/II)); 83.38, 82.00 (2d, C(2'/I), C(2'/
2 3 6
C); 0.07 (s, Me Si). C-NMR (75 MHz, (D )DMSO): 156.41, 156.35 (2s, C(6/I), C(6/
2
8
II)); 81.68, 81.40(2 d, C(3'/I), C(3'/II)); 73.55 (s, C(7'/I)); 61.87, 61.81 (2d, C(5'/I), C(5'/II)); 26.94, 25.11 (2q,
‡
Me
2
C); 25.16 (2q, Me
9
2
C); À 0.34 (q, Me
3
Si). HR-MALDI-MS: 755.2747 ([M ‡ Na]; calc. 755.2698).
-[6,7-Dideoxy-7-C-(trimethylsily)-b-d-allo-hept-6-ynofuranosyl]adenin-8-yl-(8 ! 7')-9-(6,7-dideoxy-b-d-
al lo-hept-6-ynofuranosyl)adenine (7). A soln. of CF COOH (2.5 ml) in H O (5 ml) was treated at 258 with a
soln. of 6 (0.1 g, 0.14 mmol) in THF (12 ml), stirred for 10 min, and evaporated. The suspension of the residue in
THF (1 ml) was treated with 8m MeNH in THF (0.5 ml), stirred for 10 min, and evaporated. FC (CHCl /MeOH
/MeOH 85 :15) 0.14. [a]
/MeOH 3 :2): 267 (8880), 300 (sh, 6921). IR (KBr): 3323m, 3201m, 2184w, 1652m,
616m, 1479w, 1427w, 1377w, 1333w, 1250w, 1204m, 1131m, 1075m, 956w, 847m, 798w. H-NMR (300 MHz,
3
2
2
3
2
5
8
0
1
6 :14) gave 7 (0.0783 g, 88%).White solid. M.p. 1808 (dec). R
.51, DMSO). UV (CHCl
f
(CHCl
3
D
ˆ À20.4 (c ˆ
3
1
1
3
(
D
6
)DMSO): see Table 1; additionally, 0.11 (s, Me
3 6
Si). C-NMR (75 MHz, (D )DMSO): 156.47, 156.43 (2s, C(6/
II), C(6/I)); 153.30, 152.61 (2d, C(2/I), C(2/II)); 149.23, 148.05 (2s, C(4/I), C(4/II)); 140.13 (d, C(8/I)); 133.24 (s,

7
10
Helvetica Chimica Acta ± Vol. 85 (2002)
C(8/II)); 119.49 (2s, C(5/I), C(5/II)); 104.74 (s, C(6'/II)); 94.84 (s, C(6'/I)); 89.64, 88.74 (2d, C(1'/I), C(1'/II));
8
7
9.24 (s, C(7'/II)); 87.83, 87.67 (2d, C(4'/I), C(4'/II)); 73.61 (s, C(7'/I)); 73.13, 71.82 (2d, C(2'/I), C(2'/II)); 70.60,
0.32 (2d, C(3'/I), C(3'/II)); 62.80, 62.74 (2d, C(5'/I), C(5'/II)); À 0.30 (q, Me
3
Si). HR-MALDI-MS: 675.2035
‡
(
[M ‡ Na] ; calc. 675.2072).
9
-(6,7-Dideoxy-b-d-allo-hept-6-ynofuranosyl)adenin-8-yl-(8 ! 7')-9-(6,7-dideoxy-b-d-allo-hept-6-ynofura-
nosyl)adenine (8). A soln. of 7 (0.03 g, 0.046 mmol) in THF (10 ml) was treated with 1m Bu NF ¥ 3 H O in THF
/MeOH 85 :16) gave 9 (0.014 g, 54%).
4
2
(0.05 ml, 0.055 mmol), stirred for 30 min, and evaporated. FC (CHCl
3
1
White solid. R
I)); 8.15 (s, HÀC(2/II)); 7.76 (br. s, NH
HOÀC(5'/II)); 5.97 (d, J ˆ 8.0, HÀC(1'/II)); 5.94 (d, J ˆ 8.0, HÀC(1'/I)); 5.54 (d, J ˆ 6.8, HOÀC(2'/II)); 5.50
f
(CHCl
3
/MeOH 7:3) 0.32. H-NMR (500 MHz, (D
6
)DMSO): 8.34 (s, HÀC(8)); 8.16 (s, HÀC(2/
2
); 7.44 (br. s, NH
2
); 7.04 (d, J ˆ 4.8, HOÀC(5'/I)); 6.85 (d, J ˆ 3.3,
(
d, J ˆ 6.8, HOÀC(2'/I)); 5.44 (d, J ˆ 3.8, HOÀC(3'/I)); 5.42 (d, J ˆ 4.0, HOÀC(3'/II)); 4.94 (ddd, J ˆ 8.0, 6.8,
5
2
4
.0, HÀC(2'/II)); 4.89 (br. t, J ꢁ 4.6, HÀC(5'/II)); 4.75 (ddd, J ˆ 8.0, 6.8, 5.2, HÀC(2'/I)); 4.55 (ddd, J ˆ 4.5, 3.3,
.5, HÀC(5'/I)); 4.33 (ddd, J ˆ 5.2, 3.8, 1.0, HÀC(3'/I)); 4.28 (ddd, J ˆ 5.0, 4.0, 1.0, HÀC(3'/II)); 4.14 (dd, J ˆ
.5, 1.0, HÀC(4'/I)); 4.02 (dd, J ˆ 4.5, 1.0, HÀC(4'/II)); 3.35 (d, J ˆ 2.5, HÀC(7')). MALDI-MS: 581 ([M ‡
‡
‡
H] ), 603 ([M ‡ Na] ).
4
-Methyl-1-(trimethylsilyl)pent-1-yn-3-ol (9). Prepared according to the procedure of Kitano et al. [26].
1
Distillation gave 9 (1.57 g, 87%). Colourless liquid. B.p. 758/17 mbar, ([36]: 998/30Torr). H-NMR (300 MHz,
CDCl
3
) : 4.15 (t, J ˆ 5.6, HÀC(3)); 1.87 (oct., J ˆ 5.1, HÀC(4)); 1.76 (d, J ˆ 5.6, HOÀC(3)); 1.01, 0.99 (2d, J ˆ
1
5
.0, Me
2
CH); 0.18 (s, Me
3
Si). H-NMR (300 MHz, (D
6
)DMSO)) : 5.31 (d, J ˆ 5.6, HOÀC(3)); 3.96 (t, J ˆ 5.6,
13
HÀC(3)); 1.68 (oct., J ˆ 5.1, HÀC(4)); 0.90, 0.87 (2d, J ˆ 5.0, Me
2
CH); 0.13 (s, Me
3
Si). C-NMR (75 MHz,
Si).
CDCl ): 105.75 (s, C(2)); 90.27 (s, C(1)); 68.43 (d, C(3)); 34.56 (d, C(4)); 18.22, 17.59 (2q, 2 Me); 0.04 (s, Me
3
3
REFERENCES
[
1] E. Uhlmann, A. Peyman, Chem. Rev. 1990, 90, 543; K. Dueholm, P. Neilsen, New J. Chem 1997, 21, 19; G.
Haaima, A. Lhose, O. Buchardt, P. E. Neilsen, Angew. Chem., Int. Ed. 1996, 35, 1939; M. Beier, F. Reck, T.
Wagner, R. Krishnamurthy, A. Eschenmoser, Science 2000, 283, 699; M. Dobler, A. Eschenmoser, Helv.
Chim. Acta 1992, 75, 218; U. Diedrichsen, D. Weicherding, Synlett 1999, S1, 913; A. Egger, C. Leumann,
Synlett 1999, S1, 913.
[
[
[
[
[
[
[
[
2] S. Eppacher, N. Solladi e¬ , B. Bernet, A. Vasella, Helv. Chim. Acta. 2000, 83, 1311.
3] H. Gunji, A. Vasella, Helv. Chim. Acta 2000, 83, 3229.
4] M. Sundaralingam, Biopolymers 1969, 7, 821.
5] S. Arnott, D. W. L. Hukins, Nature 1969, 224, 886.
6] D. J. Cram, Chemtech 1987, 17, 120.
7] E. T. Kool, Chem. Rev. 1997, 97, 1473.
8] W. P. Jencks, Adv. Enzymol. Relat. Areas Mol. Biol. 1975, 43, 219.
9] W. Saenger, −Principles of Nucleic Acid Structure×, Springer Verlag, Berlin, 1984, p. 21.
[
[
10] T. Wada, N. Minamimoto, Y. Inaki, Y. Inoue, J. Am. Chem. Soc. 2000, 122, 6900.
11] S. M. Gryaznov, D. H. Lloyd, J. K. Chen, R. G. Schultz, L. A. DeDionisio, L. Ratmeyer, W. D. Wilson,
Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 5798.
[
[
[
[
[
[
[
[
[
12] O. Almarsson, T. C. Bruice, Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5542.
13] W. Saenger, −Principles of Nucleic Acid Structure×, Springer Verlag, Berlin, 1984, p. 69 ± 78.
14] B. Catalanotti, A. Galeone, L. G. Paloma, L. Mayol, A. Pepe, Bioorg. Med. Chem. Lett. 2000, 10, 2005.
15] L. Kvaerno, J. Wengel, Chem. Commun. 2001, 1419.
16] W. Saenger, −Principles of Nucleic Acid Structure×, Springer Verlag, Berlin, 1984, p. 79.
17] S. Fuji, T. Fujiwara, K. Tomita, Nucleic Acids Res. 1976, 3, 1985.
18] F. Seela, H. Debelak, H. Reuter, G. Kastner, I. A. Mikhailopulo, Nucleosides Nucleotides 1998, 17, 729.
19] S. T. Rao, M. Sundaralingam, J. Am. Chem. Soc. 1970, 92, 4963.
20] V. Swaminathan, M. Sundaralingam, J. B. Chattopadhyaya, C. B. Reese, Acta Crystallogr., Sect. B 1980, 36,
828.
[
[
21] D. Plochocka, A. Rabczenko, D. B. Davis, Biochem. Biophys. Acta 1977, 476, 1.
22] L. H. Koole, H. Deboer, J. W. Dehaan, C. A. G. Haasnoot, P. Vandael, H. M. Buck, J. Chem. Soc., Chem.
Commun. 1986, 362.
[
[
23] T. Udea, Y. Nomoto, A. Matsuda, Chem. Pharm. Bull. 1985, 33, 3263.
24] H. Gunji, A. Vasella, Helv. Chim. Acta 2000, 83, 1.

Helvetica Chimica Acta ± Vol. 85 (2002)
25] H. Gunji, A. Vasella, Helv. Chim. Acta 2000, 83, 2975.
26] Y. Kitano, T. Matsumoto, F. Sato, Tetrahedron 1988, 44, 4073.
27] B. Bernet, A. Vasella, Helv. Chim. Acta 2000, 83, 995.
28] B. Bernet, A. Vasella, Helv. Chim. Acta 2000, 83, 2055.
29] D. B. Davies, S. S. Danyluk, Biochemistry, 1974, 13, 4417.
30] W. Saenger, −Principles of Nucleic Acid Structure×, Springer Verlag, Berlin, 1984, p. 45.
31] C. Altona, M. Sundaralingam, J. Am. Chem. Soc. 1973, 95, 2333.
32] C. Altona, M. Sundaralingam, J. Am. Chem. Soc. 1972, 94, 8205.
33] H. Rosemeyer, G. Toth, B. Golankiewicz, Z. Kazimierczuk, W. Bourgeois, U. Kretschemer, H. P. Muth, F.
Seela, J. Org. Chem. 1990, 55, 5784.
711
[
[
[
[
[
[
[
[
[
[
[
34] S. Eppacher, A. Vasella, unpublished results.
35] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M. Lipton, C. Caufield, G. Chang, T.
Hendrickson, W. C. Still, J. Comput. Chem. 1990, 11, 440.
[36] G. Deleris, J. Dunogues, R. Calas, Tetrahedron Lett. 1976, 2449.
Received December 10, 2001