Corey, Shenvi. JACS, 2009, ASAP. DOI: 10.1021/ja901400q.
It was Corey that first brought this website properly off-topic back in 2006 with his synthesis of Tamiflu, which of course, isn’t a natural product (but is thusly derived). It’s a similar situation with this target, 7-methylomuralide, which is a derivative of muralide, a target quite popular with Corey (with syntheses back in ’98 and ’03). In this case, a simplification was made to ‘racemise’ the C-7 methyl group by converting it into a gem-dimethyl group – and retaining most of the activity in the process. In this case, the activity is ‘potent proteasome inhibition’ – a useful tool for biologists and biochemists, and may cause apoptosis of cancer cells. It’s also a neat, compact target!
Key to the synthesis is a rather neat preparation of beta-keto lactam using a protected glycine and dimethylmalonyl dichloride. It was then their plan to do an aldol condensation with isobutyraldehyde, basically completing the carbon skeleton in only a couple of steps! However, this approach came slightly unstuck, as the reaction couldn’t be driven to completion when using their initial substrate, which was PMB protected.
They then moved to using carbonyl protecting groups, and the Boc group, initially. This, at least, gave them the desired product, but unfortunately predominately as the wrong diastereoisomer. A further attempt, using the Troc protecting group (trichloroethoxycarbonyl), was thankfully pretty awesome in turning over that selectivity, returning a 11:1 d.r.
Next up was an asymmetric variant of the reaction, using a chiral naphthalene-sulfonyl cyclohexanol directing group developed by the group for asymmetric Diels-Alder chemistry. This choice led to pretty damn nice selectivity (and of course, planarity at the ‘alpha,alpha-position’), setting two new stereocenters in one go.
It only took a few more steps to finish the prep – a nice bit of tandem reduction / deprotection was used to prevent retro-aldol, leaving them with only removal of the auxiliary, and beta-lactone formation. Neat work!
The paper is spoiled somewhat by a bit of mis-referencing – of Corey’s own work! Reference ten, which should lead to a paper about the chiral auxiliary and some asymmetric Diels-Alder work… doesn’t. However, given that the same authors, Corey and Sarakinos, published in the same journal twice in one year, I guess they can be forgiven! If you dig out the correct paper, Org. Lett., 1999, 1, 1741–1744, you’ll see a rather nice prep of the auxiliary. Opening of cyclohexene oxide with thionaphthalene, enzymatic resolution and oxidation gives them one enantiomer, with the other only requiring one more step. Nice stuff!
I liked this synthesis a lot, and think the auxiliary directed aldol is a good approach. However, I wonder if there’s any scope for a chiral-Lewis acid directed synthesis – although there are perhaps too many Lewis basic sites.