A convenient route to the anti-HIV active compound, 9-(2,3-dideoxy-2fluoro-13-D-threo-pentofuranosyl)adenine (1, ~FddA) started with the facile introduction of fluorine at C2' from the a-side of protected 9-(l$-D-arabinofuranosyl)adenine (ara-A). Inversion of the stereochemistry at C2' was accomplished via a stable vinyl intermediate ( ), which underwent stereoselective reduction of the double bond to give the desired 2'-F-threo isomer with the opposite I~-fluoro stereochemistry. Published by Elsevier Science Ltd.
9-(2,3,-¢fideoxy
currently in clinical trial, that has shown potent in vitro and in vivo activity against HIV in experimental models. 1,2 One of the most promising aspects of fl-FddA is its effectiveness against HIV strains resistant to zidovudine (AZT), zalcitabine (dideoxycytine, ddC), and even didanosine (dideoxyinosine, ddI). 3 Several
~.N NH2
F 1 (~-FddA) methods of synthesis of I~-FddA have been reported, and, in each case, the effective inlroduction of the fluorine atom at the "up", or 13-position of the aglycone has proven to be a challenge. The convergent approaches consisting of coupling a fluorinated sugar with a purine base have progressively improved the yield of the I~fluorosugar precursor, but have failed to optimize the coupling step.l,4, 5 This low efficiency purine coupling stands in sharp contrast to the more efficient glycosylation of pyrimidines employing the same fluorosugar. 4
The more attractive linear approaches, where one introduces the fluorine directly onto the 13-position of a purine nucleoside, also give poor yields.6, 7 A clever, but lengthy modification of a linear approach employing bulky 3',5 '-di-O-trityl groups which changed the sugar conformation and reduced the preponderance of elimination products gave better yields, s-t0 However, this procedure appears to be impractical for a large scale synthesis.