The specific chemical synthesis and characterization of ribonucleoside-O'-benzyl ethers

Much interest has been displayed in recent years in the synthesis of oligonucleotides of specific sequence and length. The synthesis of these molecules by chemical means require properly block nucleoside derivatives, the blocking groups of which can be selectively removed under mil conditions. The commonly used blocking groups are removed under acidic or basic conditions. Several researchers have proposed the use of the benzyl ether as a blocking group since it is stable to both acidic and basic conditions and can readily be removed under mild hydrogenolysis. The synthesis of several O-benzylnucleosides has been previously reported but the synthetic procedures are not general applicability and the yields are usually low. To allow further investigation of the O-benzyl blocking group in nucleoside and nucleotide chemistry, a generally applicable one-step synthesis of 2' - 3'-0-benzylribonucleosides have been developed. The reaction of adenosine, inosine guanosine, uridine, or cytidine with phenyl diazomethane in methanol containing a catalytic amount of stannous chloride proceeds to give good yields of a mixture of the corresponding 2'- and 3'- 0-benzylnucleosides. The reaction is selective for the cis-diol with none of the 5'-isomer observed as a product. N-alkylation is of little significance. Only in the benzylation of guanosine was a number of other products observed, probably resulting from N-alkylation. Debenzylation may be accomplished readily by hydrogenolysis. The Debenzylation of 0-benzylcytidine was complicated by a concurrent 10 to 20% reduction of the 5, 6-double bond of the pryimidine ring. Model studies demonstrate that in the case of the 2'0-benzylnucleosides, the phenyl group can stack with the heterocyclic moiety in a manner very similar to that demonstrated for dinucleoside phosphates. This intra-molecular stacking is impossible for the 3'-0-benzylnucleosides. The analogous stacking of dinucleoside phosphates caused a significant hypo-chromic shift in the uv spectra. The 2'-0-benzylnucleosides all exhibit a 150 to 29% hypochromicity while the 3'-0-benzylnucleosides have uv spectra that are very similar to those of the parent nucleosides. The hypochromicity of the 2'-isomer provides a unique and very facile means for distinguishing between the two isomers. The pmr spectra may be used to distinguish between the 2'- and 3-0-benzylnucleosides. Previous studies have shown that the chemical shift and coupling constant of H-1' can be used for identification of the isomers. The results here are in general agreement with this empirical rule. The OH proton signals can also be used to confirm the assignments. The pmr spectra also provide evidence of the stacked conformation that occurs in aqueous solution. The pmr spectra show an up-field shift of the protons most intimately involved in the stacking. This results from the magnetic anisotropy of the neighboring ring system. Temperature studies show the benzene ring must be oriented primarily over the purine H-2 proton since the chemical shift of this proton is most affected as the molecule is caused to un-stack at elevated temperatures. The pmr signal for the anomeric proton of 2'-0-benzyladenosine and 2'-0-benzylinosine in D2o appear as pentets instead of the anticipated doublets which are observed for the compounds in DMSO-d6. This multiplicity is temperature-dependent with the signal becoming a doublet as the temperature is raised to 70°C. The multiplicity is explained as resulting from virtual coupling, H-2' and H-3' having essentially identical chemical shifts.