The total synthesis of (+)-coccinine via an intramolecular 2-azaallyl anion cycloaddition.

Abstract

The total synthesis of the (+)-coccinine, the non-natural enantiomer of the Amaryllidaceae alkaloid ($-$)-coccinine, was achieved using an intramolecular 2-azaallyl anion cycloaddition as the key step. Studies toward the synthesis began with an approach to the parent 5,11-methanomorphanthridine ring system which was based on a cycloaddition onto a tethered allenyl anionophile. Due to the inefficiency of this process, an alternate route was developed, relying on an epoxide opening reaction using a vinyl sulfide nucleophile. This approach was successful, efficiently providing an enantiomerically enriched 2-azaallyl anion cycloaddition precursor, which was carried through the key cycloaddition step in the synthesis. Treatment of the cycloaddition precursor with n-butyllithium at ${-}78\sp\circ$C resulted in formation of (3S,3aS,5S,6S,7aR)-3-(1,3-benzodioxolan-5-yl)-3a-phenylthio-5-methoxy-6-benzyloxyoctahydroindole in yields ranging from 35-45%. This perhydroindole is the product of an intramolecular cycloaddition between the 2-azaallyl anion derived from tin-lithium exchange via n- butyllithium, and vinyl sulfide portions of the molecule. The cycloaddition product was then carried through the remaining steps of the synthesis to provide (+)-coccinine. The completion of this project represents the first total synthesis of an un-natural enantiomer of the 5,11-methanomorphanthridine family of Amaryllidaceae alkaloids. The original synthetic target in this synthesis was the C-3 epimer of coccinine, ($-$)-montanine. It was determined, however, that completion of the proposed route did not result in ($-$)-montanine, but a diastereomer resulting from cycloaddition occurring in the reverse sense to that originally predicted. Although a cis-cycloaddition product was obtained, it was the unexpected cis-adduct. The stereochemistry of the cycloaddition product was verified by X-ray crystallographic analysis of a subsequently prepared mesylate. This result may be the explained if the intermediate 2-azaallyl anion reacts via a boat-like transition-state rather than the predicted chair-like transition-state. As a consequence of this finding, the original route was appended with an epimerization strategy, allowing for completion of the synthesis of (+)-coccinine.