The Nicolaou/Li Synthesis of Tubingensin A

The complex indole diterpene alkaloids, isolated both from Aspergillus sp. and from Eupenicillium javanicum, display a wide range of physiological activity. K. C. Nicolaou of Scripps/La Jolla and Ang Li, now at the Shanghai Institute of Organic Chemistry, conceived (J. Am. Chem. Soc. 2012, 134, 8078. DOI: 10.1021/ja302765m) a divergent strategy for the assembly of these alkaloids, that enabled syntheses both of Anomine (not illustrated) and of Tubingensin A (3). A key step in the assembly of the carbocyclic skeleton of both alkaloids was the radical cyclization of 1 to 2, establishing the second of the two alkylated quaternary centers of 3.

The starting point for the preparation of 1 was commercial pulegone (4). Methylation followed by acid-mediated retro aldol condensation delivered the enantiomerically-pure 2,3-dimethyl cyclohexanone 5. To maximize yield, the subsequent Robinson annulation was carried out over three steps, formation of the silyl enol ether, condensation of the enol ether with methyl vinyl ketone 6, and base-mediated cyclization and dehydration of the 1,5-diketone to give 7. The secondary hydroxyl group was introduced by exposure to Oxone of the methyl dienol ether derived from 7. The mixture of diastereomers from the radical Ueno-Stork cyclization of 1 was equilibrated to the more stable 2 by exposure to acid.

The authors took advantage of the regioselective enolization of 2, preparing the silyl enol ether, that could then be condensed with formaldehyde to give 10. This hydroxy ketone was carried on to 11 over four steps, commencing with silylation and proceeding through Wittig condensation, desilylation and oxidation. The addition of the Grignard reagent 12 to the aldehyde 11 gave a secondary alcohol, that was readily dehydrated to the diene 13. The diene resisted thermal cyclization, but on exposure to CuOTf at room temperature it was smoothly cyclized and oxidized to 14.

The elaboration of the sidechain had already been worked out in the Anominine synthesis. The free lactol derived from 14 resisted many nucleophiles, but vinyl magnesium bromide did add. Bis acetylation of the resulting diol followed by Pd-mediated ionization and reduction of the allylic acetate, and reductive removal of the residual acetate, delivered the terminal alkene 15. Metathesis with isobutylene gave 16, that was deprotected to give Tubingensin A (3).

A key early intermediate in this synthesis was the 1,5-diketone that was cyclized to 7. There is a good chance, for instance following the protocol of Silas P. Cook of Indiana University (J. Am. Chem. Soc. 2012, 134, 13577. DOI: 10.1021/ja3061479), that 7 could be prepared by direct enantioselective conjugate addition/modified Robinson annulation, from 2-methyl cyclohexenone.