Smith, Basu and Bosanac. JACS, 129, 14872 – 14874. DOI: 10.1021/ja077569l.
I felt like suggesting I was bullied into posting an entry on this synthesis, but the truth is that 1) it’s an awesome piece of work, and 2) I did ask for suggestions… So feeling suitably deferential, let’s look at the biological activity. Nothing particularly special in terms of the effect (cytotoxicity against the human cell lines P388 and lymphoid leukemia), but the numbers are fairly small: IC50 0.09 and 0.037 Î¼g/mL. However, if we’re fair – it’s that amazing 11,6,6,5- ring system that’s the real pull! Retro time…
So the larger ring was to be completed by RCM, reducing out the extraneous unsaturation, and the smallest ring by an enolate-type addition into a carbonate. This brings us back to a complex pair of linked six-member rings, which they intended to join using a (relatively) simple acetal formation, followed by Petasis-Ferrier rearrangement… ambitious! The linear fragment is a standard polyketide type of thing, and the heavily substituted cyclohexane resulted from a Rawal Diels-Alder… lets start from there:
After the DA, and a few functional group transformations, they were set for a short series of sweet reactions. First, a catalytic carbonylation, and then diastereoselective allylation -nice result. They then transformed the enol ether into a dimethyl acetal by treating with mCPBA in methanol, delivering the oxidant diastereoselectively, and providing the required tertiary alcohol. After a few more functional group transformations (required for optimal results in the next few steps), they were set for the unification of the two fragments.
First-up – formation of the acetal/lactone thing (apparently a “dioxanone”), methylenation of the carbonyl group and then that awesome rearrangement. The yield might not be stunning, but the result is really nice! Then time for a bit of carbonyl-chemistry; deprotonation of the free alcohol and enolate formation, followed by trapping of both with methyl chloroformate gave a pair of carbonates. Explaining what happens next is a task I’ll just quote: “biscarbonate… was subjected to chemoselective methanolysis to form an intermediate sodium enolate, derived from the enol carbonate, which in turn undergoes cyclization with the tertiary carbonate to furnish lactone”. Nice stuff… With the 1,3-dicarbonyl in place, methylation was easy, and also stereoselective.
With a bit of metathesis action and reduction of the alkene, the macrocycle was complete, leaving them with only a few function group transformations to complete the molecule. However, these steps seemed to trouble the group – the planed selenoxide eliminations were tricky, as they couldn’t form both the enolates required to trap PhSeCl. They overcame this by performing one elimination as usual, and the other by first completing a Grieco-Nishizawa elimination, and then oxidising the product (over a few steps) to the desired unsaturated acid.
Again, this synthesis is an awesome read, and full of interesting chemistry. Thanks for the tip, folks…