Convergent Synthesis of the Polyene Macrolide (-)-Roxaticin

TLDR: Roxaticin this paper was synthesized from polyol tetraacetonide 5, which was prepared by a threefold convergent route: sequential alkylation of dibromide 3 with cyanohydrin acetonides 2 and 4 followed by stereoselective reductive decyanation, and acid-catalyzed deprotection.

Abstract: (-)-Roxaticin has been synthesized from polyol tetraacetonide 5, which was prepared by a threefold convergent route. Each of the optically pure building blocks (2, 3, and 4) was prepared using a Noyori asymmetric hydrogenation. Sequential alkylation of dibromide 3 with cyanohydrin acetonides 2 and 4 followed by stereoselective reductive decyanation gave tetraacetonide 5. The initial approach to roxaticin using a 1-methylcyclopropyl ether in a key protection step was unsuccessful due to the instability of the polyene chain to oxydative deprotection. A 1,3-benzodithiolan-2-yl (BDT) ether performed well in a model study and was used in the roxaticin system. Protection of the roxaticin precursor as a BDT ether followed by elaboration of the polyene using Wollenberg's method gave a tetraenal. The macrocyclic ring was closed using an intramolecular Horner-Emmons Wittig reaction, and acid-catalyzed deprotection completed the synthesis of roxaticin. Our synthesis of roxaticin illustrates a first generation approach to the highly convergent synthesis of polyene macrolide antibiotics that should ultimately by useful for preparing stereochemical and structural analogs