UC Berkeley

The ophiobolins are a family of sesterterpene natural products that share a 5-8-5 tricyclic core structure, complex oxidation patterns, and at least six stereogenic centers. Originally isolated from fungi responsible for crop blights, and long known for intriguing phytotoxic activities, the ophiobolins have recently received additional attention following the discovery of significant anti-cancer potential, in particular the observation that treatment with ophiobolin A leads to non-apoptotic cell death in glioblastoma multiforme. Previous total syntheses of ophiobolins have featured stepwise construction of the key bonds of the natural products and have required lengthy step counts, in comparison to the biosynthetic pathway which features a single cationic cyclization cascade. We hypothesized that the natural biosynthetic precursors of the ophiobolins would be valuable synthetic intermediates, but that an abiotic cyclization strategy would be possible if we synthetically “reprogrammed” these intermediates to engage in radical, as opposed to cationic, reaction pathways. Specifically, on a monocyclic intermediate, we considered that an 8-endo-5-exo radical cyclization cascade to create two additional rings and several stereogenic centers could be a key step in the synthesis (Chapter 1).

Initial work on a model system (Chapter 2) demonstrated the feasibility of this 8-endo-5-exo radical cascade, but we discovered during our first experiments that the configuration of the C10 center was set incorrectly. However, a change in the radical tether of the substrate changed the selectivity of the reaction, and we found that cyclization precursors with a specific relative configuration between C6 and C11 cyclized to form the desired 5,8-trans ring junction of the C10 and C11 stereogenic centers. Mechanistic possibilities for this change in selectivity are discussed.

We next turned our attention to the synthesis of ()-6-epi-ophiobolin N (Chapter 3). A nine-step total synthesis of ()-6-epi-ophiobolin N was completed, utilizing ()-linalool and trans,trans-farnesol as affordable terpene-based building blocks. The synthesis featured (i) a three-component coupling reaction to rapidly assemble our cyclization precursor, (ii) a reductive cyclization cascade that constructed the B and C rings and correctly set three stereogenic centers (at C10, C14, and C15), and (iii) the development of a novel, TADDOL-derived monothiol as a polarity-reversal catalyst capable of reversing the inherent diastereoselectivity of hydrogen atom transfer to an intermediate radical at the C15 position. Furthermore, we demonstrated that the cyclization cascade can be performed under a number of experimental conditions, altering the fate of the intermediate C15 radical and leading to a number of different final products from the reaction.

Finally, we explored the synthesis of the flagship and most complex member of the family, ophiobolin A. While studies are still ongoing, we have to date completed a small-scale total synthesis of (+)-6-epi-ophiobolin A (Chapter 4), which required us to develop methods to formally epimerize the C3 stereogenic center and to create a tetrahydrofuran (THF) ring on the side chain. A two-step protocol featuring a 1,4-reduction of a hindered enone allowed us to address the C3 stereogenic center. The THF ring was then efficiently created via a redox-neutral radical cyclization mediated by copper, followed by a sequence including singlet oxygen ene reaction, hydroformylation, and Grignard addition to an oxonium ion intermediate. Optimization of this final sequence, and the development of an epimerization reaction at the C6 stereogenic center to facilitate the synthesis of ophiobolin A, is currently being studied and will be reported in due course.