The secondary hydroxy group in the 2-position is located
near a phosphodiester group on the strand containing 2
(Figure 1).
The synthesis of protected 5-hydroxymethyl-2′-deoxyuri-
dine glycosyl acceptor 12 has already been described.8,9
However, the rather low yields and high costs for starting
materials in these syntheses led us to explore a different route.
5-Hydroxymethyluracil (6), obtained in one step from
uracil,10 was coupled with tetraacetylribose 5 to give 5-hy-
droxymethyluridine (7) in 79% yield (Scheme 1).11 To
differentiate the two primary hydroxy groups, 7 was reacted
with dimethoxytrityl chloride. After removal of the acetyl
groups, the ribose was selectively protected at the 3′- and
5′-hydroxyls with TIPDS to give 10. Barton-McCombie
reduction of the 2′-hydroxy group gave deoxynucleoside 11
in high yield. The DMT protecting group could then easily
be removed to give the glycosyl acceptor 12 in 43% overall
yield from 5-hydroxymethyluracil (6).
Since we felt that the synthesis of a suitable 2,6-
diaminoglucose donor from a commercially available
monosaccharide would require eight to ten steps, we chose
instead to start from a natural product. The aminoglycoside
neomycin B (15) contains a 2,6-diamino-2,6-dideoxy-D-
glucose subunit and is readily available. The glycosidic
linkage between the ribose residue and the aminocyclohexitol
2-deoxystreptamine (2-DOS) can be cleaved selectively
under acidic conditions to give neamine hydrochloride 16.12
The amino functions in neamine were first protected as
trifluoroacetamides and then the hydroxy groups were
esterified with acetic anhydride to give 17 in 67% yield
(Scheme 2). This neamine derivative could be hydrolyzed
with HBr in acetic acid to give a mixture of the desired
glycosyl donor 18 and the 2-DOS byproduct 19.13
The coupling of glycosyl bromide 18 and uridine deriva-
tive 12 proceeded smoothly under mercury salt activation
to give diaminoglucosylated nucleoside 13 in 84% yield.
Removal of the silyl protecting groups and standard ma-
nipulations14 then provided phosphoramidite building block
14. With this, oligonucleotides 22 and 26 (Table 1) were
prepared on a 1.3 µmol scale by automated solid-phase
synthesis.15,16
Figure 1. The glucose moiety of nucleoside analogue 2 is located
within the major groove of double stranded DNA.3 By replacing
the hydroxy groups at the 2- and 6-position of the glucose by
ammonium groups, the two oligonucleotide strands could be
clamped together through electrostatic interactions.
Incorporating 2 into oligonucleotide duplexes reduces their
thermal stability. The replacement of ordered water in the
major groove by hydrogen-bonding contacts between the
glucose and nucleobases leads to a favorable entropic
contribution which, however, cannot fully compensate for
the smaller enthalpy of duplex formation.7 This seems to be
a tolerable tradeoff for the pathogens mentioned, to protect
their genome from degradation.
To improve the thermal stability and yet retain the desired
protection from nucleases, we sought to synthesize oligode-
oxynucleotides (ODNs) containing nucleoside analogues 1
and 4. The amino groups, positively charged at neutral pH,
should improve binding to a complementary strand by
electrostatic interactions (Figure 1). However, 4 when built
into ODNs behaves in a similar manner to that of 2: a
decrease in melting temperature is observed.8 Most likely,
the 2-ammonium group of 4 is located near a phosphodiester
group of the strand it is attached to and thus does not affect
duplex formation. Here, we report the synthesis of ODNs
containing 1 which show the anticipated behavior.
Replacing one thymidine residue with 2,6-diamino-â-D-
glucose-modified 1 at a central position in the reference
(7) Bertolini, R.; Hunziker, J. Unpublished data.
(8) Hunziker, J. Bioorg. Med. Chem. Lett. 1999, 9, 201.
(9) de Kort, M.; Ebrahimi, E.; Wijsman, E. R.; van der Marel, G. A.;
van Boom, J. H. Eur. J. Org. Chem. 1999, 2337.
(10) Cline, R. E.; Fink, R. M.; Fink, K. J. Am. Chem. Soc. 1959, 81,
2521.
(11) All new compounds were fully characterized by UV, IR, 1H and
13C NMR, and mass spectroscopy.
(12) Park, W. K. C.; Auer, M.; Jaksche, H.; Wong, C.-H. J. Am. Chem.
Soc. 1996, 118, 10150.
(13) The mixture of these two compounds could only be separated by
silica gel chromatography with substantial loss of 18. Nevertheless, the ease
of preparation of 18 outweighted the poor yields of this strategy.
(14) Conolly, B. A. In Oligonucletides and Analogues; Eckstein, F., Ed.;
Oxford University Press: Oxford, 1991; pp 155-183.
(15) Oligonucleotides 20-28 were prepared on a Pharmacia Gene
Assembler Plus using the standard protocol. The coupling time for 14 was
changed from 2 to 6 min. Average coupling yields for aminoglycoside-
modified nucleosides (>98%) were in line with unmodified phosphora-
midites. The oligonucleotides were purified by reverse-phase HPLC followed
by ion-exchange HPLC. The integrity of the isolated oligonucleotides 20-
28 was subsequently confirmed by MALDI-ToF mass spectrometry: m/z
(monoanion, H+-form) 22, calcd 3155.0, found 3154.9; 26, calcd 3363.3,
found 3364.4. Matrix conditions as described in the following: Pieles, U.;
Zu¨rcher, W.; Scha¨r, M.; Moser, H. E. Nucleic Acids Res. 1993, 21, 3191.
1694
Org. Lett., Vol. 2, No. 12, 2000