THE TOTAL SYNTHESIS OF MUAMVATIN

Abstract

Muamvatin (30) is a polypropionate natural product isolated from Siphonaria normalis by Ireland et al. in 1986. Muamvatin (30) is made from eight propionate units and contains an extraordinary trioxaadamantane ring system. This ring system exists in only one other naturally occurring polypropionate known as caloundrin B. Regarding the rare muamvatin trioxaadamantane ring system, it was hypothesized this ring system may not be formed via an enzymatic process and the actual natural product could be the linear precursor ent-71 which cyclizes to muamvatin (30) during isolation. The first total synthesis of muamvatin (30) by Paterson et al. confirmed its absolute and relative configuration, but the ambiguity regarding the origin of the trioxaadamantane ring system in this molecule remains unresolved. This work describes two approaches to make the linear precursor ent-71 from triol ketone 153. The carbon skeleton of muamvatin was synthesized through two iterative diastereoselective aldol reactions. In the first approach, “the thiopyran route”, the diene moiety of aldehyde 73 required protection to avoid reduction during desulfurization. Although use of the tircarbonyliron complex was successful, the trihydroxy ketone revealed upon desulfurization was unstable and spontaneously cyclized to bicyclic acetal 156. Molecular mechanics revealed that the relative configurations embedded in C3, C7, and C8 dramatically effected the stability of the corresponding bicyclic acetal. With that lesson learned, the fully assembled linear precursor 197 was made in our second approach “the acyclic route”. The oxidation state of the backbone oxygens were manipulated via an unusual chemoselective double Swern oxidation. Finally, revealing the sensitive 5-hydroxy-3,7,9-trione functionality formed the precursor 202. Efficient cyclization of precursor 202 and removal of the protecting group at C11-OH produced the desired natural product 30. The cyclization conditions tested on the linear precursor 202, suggested that although the cyclization to the trioxaadamantane is strongly favored thermodynamically, the process is very slow and unlikely to occur during the isolation process. Thus, formation of the trioxaadamantane ring system could be an enzyme-mediated process as was concluded for caloundrin B.