The ongoing phase I clinical trial with LMC's 6-amino-9-(2,3-dideoxy-2-fluoro-beta-D-threo-pentofuranosyl)-9-H- purine (betaF-ddA) has heightened the interest in developing alternative and more economical synthesis for this compound. Currently, the methods in use are very expensive, and this could jeopardize the future development of this compound towards a possible clinical use. The crux of the problem remains the introduction of the fluorine atom with the desired stereochemistry, a step which is currently performed at a very early stage of the synthesis from expensive carbohydrate precursors. In the present investigation, we are exploring new chemistry to introduce the fluorine atom at a later stage, and starting with less expensive materials, such as adenosine. Adenosine has been chosen as the ideal starting material, and two approaches have been devised to reach our goal. The problems associated with the direct beta-fluorination of adenosine are well known in the literature. These are, elimination to give dideoxydidehydro derivatives and depurination. Our current approach from adenosine has centered around the construction of 9-(5-0-monomethoxytrityl-beta-D-glycero-pent-3-enofuranosyl)adenine which was obtained in excellent yield from adenosine in two steps. Fluorination experiments on this substrate are ongoing. The second approach form is centered around the synthesis of the accessible 6-amino-9-(2,3-dideoxy-2-fluoro-beta-D-erythro- pentofuranosyl)-9H-purine, the inactive isomer of betaF-ddA, which was obtained by a direct fluorination without elimination. The inversion of the fluorine stereochemistry from this compound is planned to proceed via formation of the corresponding 6-amino-9-(5-0-monomethoxytrityl-2-fluoro-2,3-dideoxy-beta-D-glyc ero- pent-2- enofuranosyl)adenine, followed by catalytic hydrogenation of the double bond to give the desired threo-configuration.
