Corey and Surendra. JACS, 2008, ASAP. DOI: 10.1021/ja802730a.
Steroids!Â Or at least something close to that structure.Â This little beastie is actually classed as a ‘pentacyclic triterpenoid’ according to Corey, but it’s still a resolutely old-school target.Â The synthesis is formal – what we’re targeting here is an intermediate in Johnson and Irelands synthesis way back in ’70.Â That synthesis is definitely a tour de force, but somewhat lacking in efficiency with 32 steps.Â Corey reckons that a biomimetic approach, stitching the ring systems together like a zip, might be a bit quicker.
Bam! Straight in with the good stuff.Â The diene starting material can be made easily from farnesyl acetate, using a Corey-Zhang oxidation (that’s another Corey reaction for the Named Reactions book…), though you’ll to read the SI to see this.Â A spot of 2-propenyllithium did the expected carbonyl addition, which was followed by a Brook rearrangement to give the corresponding allylic anion which trapped the aryl bromide to give the product.Â Sweet chain of reactions.
Next up was the key biomimetic cationic cyclisation, using a bit of Lewis acid to promote it.Â This lead to formation of the psuedo A, B and C rings (were it actually a steroid), but they needed two more steps to complete the manueouver – protection of the free hydroxyl and a bit of acid to complete the final ring.Â This apparently chews-off the freshly installed TBS group again too…
I’d like to think that the group went down to the pub at this point and had a good few pints before deciding ‘Nah… I’m sure we could make that even faster’, but who knows how things work in the Corey group.Â Anyway, another approach was apparently brewing, using a similar substrate and key step.Â However, the order in which the rings are formed has been turned over, with the ‘D-ring’ going in first.
This time, though, the LA didn’t work quite as selectively, and they got two products out of the reaction.Â However, more LA didn’t convert between these species, meaning that their origins must be divergent.Â Corey ascertained that the common parent to both structures is the cation shown below.Â In the expected pathway, we get addition of the proximal double bond to the cation, forming a new ring and a resulting benzylic cation (nice and stable).Â However, a different pathway can result in benzylic stabilisation – a [1,5] proton shift.Â The group proved this by making the deuterated substrate shown and followed the deuterium on it’s way.Â If you’re surprised to see this, you’re not the only one; Corey guardedly states that this might be the first example of this rearrangement.Â Also interesting was the fact that there is no control over the formation of that stereocentre – so the mechanism proceeds on either face of the dihydronaphthalene.
Very interesting work.Â The synthesis is efficient (relying on recursive use of similar reactions to build the substrates) and impressive – as is their dissection of the mechanistic quandry.Â But not my favourite Corey synthesis – that honour goes to his work on the prostaglandins, to which I (for obvious reasons) have a strong affinity.Â I’m also amazed by his Ecteinascidin 743 and Ginkgolide B syntheses.Â What about you folks?
Surendra, K., Corey, E.J. (2008). Rapid and Enantioselective Synthetic Approaches to Germanicol and Other Pentacyclic Triterpenes. Journal of the American Chemical Society DOI: 10.1021/ja802730a