A general and enantioselective approach to pentoses: a rapid synthesis of PSI-6130, the nucleoside core of sofosbuvir.

An efficient route towards biologically relevant pentose derivatives is described. The de novo synthetic strategy features an enantioselective α-oxidation reaction enabled by a chiral amine in conjunction with copper(II) catalysis. A subsequent Mukaiyama aldol coupling allows for the incorporation of a wide array of modular two-carbon fragments. Lactone intermediates accessed via this route provide a useful platform for elaboration, as demonstrated by the preparation of a variety of C-nucleosides and fluorinated pentoses. Finally, this work has facilitated expedient syntheses of pharmaceutically active compounds currently in clinical use.

Carbohydrates represent compounds of both vast abundance and fundamental biological importance. Within this class, five-carbon saccharides (pentoses) are perhaps most readily recognized as the structural monomers composing the backbones of DNA and RNA, which enable the replication, transcription, and translation of genetic information. In addition, the ribose derivative adenosine triphosphate (ATP) represents the molecular unit of energy, while a variety of other elaborated pentoses serve as cofactors crucial to enzyme function.[1] It is not surprising, therefore, that nucleoside frameworks are found at the core of many pharmaceutically active compounds and that significant research effort has been expended to gain synthetic access to non-natural pentose analogs.[2] The most common strategy to build enantioenriched nucleosides is to employ natural sugars as starting materials;[3] however, these protocols are typically protracted by the need to discriminate among four chemically similar hydroxyl groups, which further limits opportunities for the incorporation of unnatural moieties and stereochemical information. Clearly, an attractive alternative would involve a de novo synthetic sequence that rapidly and enantioselectively couples prefabricated fragments and is amenable to broad diversification of functional groups and nucleoside stereochemistry.[4−6]

In 2004, our group described a two-step synthesis of orthogonally protected hexoses applying an enantioselective proline-catalyzed aldol coupling followed by a Lewis acid-mediated, diastereoselective Mukaiyama aldol reaction (eq 1). This approach allows for the rapid and asymmetric construction of gluco-, manno-, and allo-configured carbohydrates from simple starting materials.[7] We questioned whether a similar strategy might provide access to their 5-carbon, nucleoside counterparts, beginning with the enantioselective catalytic production of an α,β-dioxygenated aldehyde (eq 2). By analogy to our hexose synthesis, we envisioned that this enantioenriched aldehyde could undergo aldol coupling to build the requisite nucleoside skeleton. Importantly, such a strategy would employ catalysis-derived starting materials in place of chiral pool precursors (e.g., isopropylidene-protected glyceraldehydes),[8] which have been shown to be poorly or nonselective in similar de novo nucleoside syntheses.[9] Moreover, our building blocks would be easily modified to provide a variety of differentially substituted products and would allow for preinstallation of protecting groups, thereby obviating the need for extraneous protection–deprotection sequences. Herein we describe the successful execution of these design ideals and outline a generic and enantioselective route to nucleoside architecture.