CLERCQ. Can. J. Chem. 66, 1258 (1 988). Treatment of selectively 3',5'-protected P -D -~y l~f~r a n~~y l nucleosides (4) with phenyl chlorothionocarbonate and DMAP followed by hydrogenolysis of the resulting (2'-0-phenoxythiocarbonyl) phenyl thionocarbonate esters (6) with tributylstannane/AIBN, and deprotection, gave 2'-deoxy-P-D-threo-pentofuranosyl nucleosides (7). Formation of a by-product bis(nuc1eosid-2'-y1)thionocarbonate dimer (8) was detected in the uracil nucleoside reaction sequence. Its subsequent reduction provides one explanation for "reversion to starting alcohol" in Barton-type deoxygenation reactions. Only the guanine 2'-deoxynucleoside analogue (76) Chem. 66, 1258Chem. 66, (1988.La rtaction des nucltosides de P-D-xylofurannosyle prottgts stlectivement en 3',5' (4) avec du chlorothionocarbonate de phCnyle et du DMAP, suivie d'une hydrogtnolyse des esters thionocarbonates de (0-phtnoxythiocarbonyl-2') de phtnyle (6) avec le tributylstannane/AIBN et d'une dtprotection, conduit aux nuclCosides du dtsoxy-2' P-D-thre'o-pentofurannosyle (7). Dans la suite de rtactions impliquant le nucltoside d'uracile, on a dttectt la formation du dimtre du thionocarbonate de bis-(nuclCosidy1-2') (8) comme sous-produit. Sa rtduction substquente permet d'expliquer la ccrtversion vers I'alcool de dtpart~ qui est observte dans des rtactions de dtsoxygtnation du type de Barton. L'analogue dCsoxy-2' nucltoside de la guanine (7b) est le seul i prtsenter une activitt antivirale (faible) (contre le virus herpts simplex de type 1).[Traduit par la revue]Methods for the synthesis of P-D-xylofuranosyl nucleosides via base-sugar coupling (ref. 1, and references therein) and for selective 2'-deoxygenation of nucleosides (2) have been developed by us previously. We now report combination of these approaches to give 2'-deoxy-P-D-threo-pentofuranosyl nucleosides (7) and their evaluation for antiviral activity in several cell systems.In the pioneering studies of the Cambridge group, Michelson and Todd (3) reported preparation of the 02,C3'-cyclonucleoside of thymidine and its hydrolysis to give 1-(2-deoxy-P-Dthreo-pentofuranosy1)thy mine (7e) in 1955. Fox and Miller (4) and Horwitz et al. (5) later described careful characterization of this product. An analogous procedure was employed (6) for the preparation of 1-(2-deoxy-P-D-threo-pentofuranosy1)uracil ( 7 4 .Goodman and co-workers (7) first prepared 9-(2-deoxy-P-Dthreo-pentofuranosy1)adenine (7a) (and its 3'-deoxy isomer) by nucleophilic ring opening (oxirane) of 9-(2,3-anhydro-P-D1yxofuranosyl)adenine with benzylthiolate followed by desulfurization. Analogous (acid-catalyzed) ring opening with bromide followed by hydrogenolysis with tributylstannane gave alternative access to the isomeric adenine 2'-(7a) and 3'-deoxythreo nucleoside products (8). Ikehara's 8-mercaptoadenineo or the previous paper in this series, see ref. 23.'~uthor to whom inquiries may be addressed at the Department of Chemistry, Brigham Young University, Provo, Utah 84602, U.S.A. S8,C2'-cyclonucleoside desul...