An isoxazole approach to the coleophomones

Coleophomones, metabolites reported first in a Japanese patent in 1998 and then by Merck in 2000 from a Coleophoma fungus, have antifungal and antibacterial properties and inhibit bacterial cell-wall transglycosylase and human heart chymase. Their unique molecular architecture contains a cyclic tricarbonyl moiety embedded in an ll-membered macrocycle. In our synthetic strategy the reactive ~,~-tricarbonyl structure is masked as an isoxazole to allow elaboration elsewhere in the molecule. We have investigated two isoxazole based approaches to these targets: 1) via an isoxazole ring with ketone or ester substituents; and 2) via a reduced benzisoxazole. The formation of isoxazole building blocks for the two approaches was by nitrile oxide l,3-dipolar cycloaddition. Several substituted phenylisoxazole diesters were made from approach 1 successfully, however, the attempt for 6-membered ring closure failed. For approach 2, a number of benzisoxazoles with different substituted phenyl groups were made in moderate to good yield. The side chain additions towards these building blocks have also been elaborated, and the formation of II-membered macrocycles have been attempted by ring closing metathesis (RCM) with Grubbs II catalyst. Precursors for total synthesis of coleophomone A, B, C and D have been made. Irradiation reactions have been used to prepare oxime precursors for coleophomone D and use of co-solvent (HMAP and DMPU) was applied in the synthesis to alkylate the benzisoxazole ring.