Yamamoto, Li and Payette. JACS, 2007, ASAP. DOI: 10.1021/ja073590a.
I know what you are thinking; not that f*cking molecule again! However, this approach by Yamamoto is really quite elegant, and contains some amazing transformations. I’m not going over the biology again – it’s all here: it would also be appropriate to point out this is a formal synthesis, heading for an intermediate found in other syntheses containing the challenging tetracyclic core. I’ve also not bothered with a retrosynthetic analysis (such a slacker, I know…), but their work differs from previous approaches by not forming the ether by alcohol addition to an alkene, but via a Robinson annulation.
On with the synthesis… Their work kicks-off with an asymmetric Diels-Alder on methyl cyclopentadiene and methyl acrylate using their BLA (BrÃ¸nsted acid assisted chiral Lewis acid) catalyst; awesome result! This approach seems very “Corey” to me, and reminds me distinctly of his work on similar systems in his total synthesis of prostaglandin F2a (:)). They then deprotonated the ester, and did a nitrosoaldol to give the N-nitroso species. This was then decarboxylated to leave the desired ketone. Nice!
The next few steps to the Robinson precursor are nicely planned, and are executed with rather decent results, other than an unfortunately unselective cyanide addition to a lactol. I’m surprised at the lack of selectivity here, and also the choice of reagent: diethyl aluminium cyanide can’t have been at the front of the fridge! However, they explain that this was a LA-mediated addition, which makes some sense to me, but I think I’m not quite getting it…
A few more steps and they were ready to cyclise, using stoichiometric L-proline to control the addition to the Si face of the enone. This reinforced the substrate induce bias to achieve a 5:1 d.r., better than the 3:1 achieved using the unmatched D-proline case. The sodium hydroxide that follows then dehydrated (aldol style) to give the desired intermediate.
This paper was a good read, and an excellent synthetic plan. However, I wish it had been one of those fancy new three-page JACS communications…