Scheme 1
Figure 2. Hydrophobic nucleoside analogues.
more dramatic stabilization came when several phenoxazine
nucleotides were clustered, which allowed for maximum
hydrophobic stacking of the tricyclic ring system upon
hybridization.
at -20 °C in dichloromethane gave exclusively the N3
product 7. The structures of the â-N1 and â-N3 products
were unambiguously determined by long-range 1H-13C
heteronuclear correlated 2-D NMR (HMBC). For the N3
product (7, Figure 3), the anomeric proton (H1′, δ ) 6.53
Alloxazine ribonucleosides have been previously prepared
by the glycosylation of ribose tetraesters with alloxazine to
give predominately the â-N1 product 8, with â-N3 (7) and
the N1,N3-diribosyl products as minor components.5 Pfleider-
er has also studied the glycosylation of 3,5-di-O-p-toluoyl-
R-D-2-deoxyribosyl chloride with a number of persilylated
lumazine derivatives.6 The N1 regioisomer was formed
exclusively; however a mixture of anomers was usually
obtained. Interestingly, the glycosylation of persilylated
alloxazine was regio- and stereoselective, giving 3b as the
only product after deprotection. We report here glycosylation
conditions to give the â-N1 or â-N3 ribonucleosides regi-
oselectively and the subsequent conversion of the â-N3
isomer to the 2′-deoxynucleoside.
In analogy with the glycosylation of guanine,7,8 we
reasoned that glycosylation at N3 may be the kinetic product,
which is converted to the thermodynamically favored N1
product via the N1,N3-bis-glycosylated intermediate, which
has been observed.5 Using 1,2,3,5-tetraacetylribose (6) as the
donor, Vorbru¨ggen glycosylation9 with silylated alloxazine
at room temperature in acetonitrile using tin tetrachloride as
a Lewis acid gave a mixture of N1 and N3 glycosylation
products. Prolonged reaction times favored the N1 product
(8), consistent with our initial hypothesis; after 2 h, 8 was
the exclusive product (Scheme 1). Glycosylation for 15 min
Figure 3. Structure determination of the â-N3 and â-N1 regioi-
1
somers by H-13C heteronuclear correlation.
ppm) showed three-bond coupling to both carbonyl carbons
of the base (C2, C4, δ ) 148.8, 159.2 ppm); for the N1
product (8), the anomeric proton (H1′, δ ) 6.80 ppm)
showed three-bond coupling to only one carbonyl carbon
(C2, δ ) 149.1 ppm) and to the 10a bridgehead carbon (δ
) 144.7 ppm).5,6b,10
We have reported a streamlined synthesis of â-2′-deoxy-
ribonucleosides via the glycosylation of 9 followed by
selective deoxygenation of the 2′-m-(trifluoromethyl)benzoate
via a photoinduced electron-transfer (PET) mechanism using
carbazoles as the photosensitizer.11 Glycosylation of 9 with
persilylated alloxazine gave a 1:1 mixture of the â-N3 and
â-N1 products in 70% combined yield (Scheme 2); these
(5) Ienaga, K.; Pfleiderer, W. Chem. Ber. 1977, 110, 3449.
(6) (a) Maurinsch, Y.; Pfleiderer, W. Nucleosides Nucleotides 1996, 15,
431. (b) Ro¨sler, A.; Pfleiderer, W. HelV. Chim. Acta 1997, 80, 1869.
(7) (a) Dudycz, L. W.; Wright, G. E. Nucleosides Nucleotides 1984, 3,
33. (b) Wright, G. E.; Dudycz, L. W. J. Med. Chem. 1985, 27, 175. (c)
Garner, P.; Ramakanth, S. J. Org. Chem. 1988, 53, 1294.
(8) For solutions of the regiochemical problem of guanine glycosylation,
see: (a) Robins, M. J.; Zou, R.; Guo, Z.; Wnuk, S. F. J. Org. Chem. 1996,
61, 9207. (b) Jenny, T. F.; Benner, S. A. Tetrahedron Lett. 1992, 33, 6619.
(9) (a) Vorbru¨ggen H. F.; Ruh-Pohlenz, C. Org. React. 2000, 55, 1. (b)
Vorbru¨ggen H. Acc. Chem. Res. 1995, 28, 509. (c) Vorbruggen, H. Acta
Biochim. Pol. 1996, 43, 25.
(10) Ewers, U.; Gu¨nther, H.; Jaenicke, L. Chem. Ber. 1974, 107, 876.
(11) (a) Park, M.; Rizzo, C. J. J. Org. Chem. 1996, 61, 6092. (b) Wang,
Z.; Rizzo, C. J. Tetrahedron Lett. 1997, 38, 8177. (c) Prudhomme, D. R.;
Wang, Z.; Rizzo, C. J. J. Org. Chem. 1997, 62, 8257.
228
Org. Lett., Vol. 2, No. 2, 2000