Totally Synthetic by Paul H. Docherty, 29 September 2008
Total Synthesis of Cassaine
S. Phoenix, M. S. Reddy, P. Deslongchamps, J. Am. Chem. Soc. 2008, 130, 13989-13995.
Cassaine is quite interesting from a biological perspective, as it acts as an inhibitor of Na+,K+-ATPase, just like digitoxin. However, its structure is pretty different.
The synthesis by Pierre Deslongchamps relies upon a swift construction of the tricycle using a trans-annular Diels-Alder (TADA) process, where a larger macrocycle performs an intramolecular Diels-Alder reaction. However, they had to make that initial macrocycle first. The synthesis starts with an Evans-type aldol reaction to generate the two contiguous stereocenters shown in the starting material below. Seven steps, using relatively well known chemistry, allowed construction of the rest of that substrate. Then Deslongchamps used an interesting methodology developed by Magid to transform the allylic alcohol into the corresponding chloride using hexachloroactetone and triphenyl phosphine. It’s not really very different to a standard Appel reaction, but works well, on a somewhat sensitive substrate.
Using this alkyl chloride, they did a displacement by treating it with a 1,3-dicarbonyl and a base. The crown ether was apparently very important; a 60% yield was the result in its absence. In an evolved route, this reaction was done on an allyl ester, as in the cyclisation step, a Stille coupling, the palladium catalyst could also promote decarboxylation, resulting in an intermediate macrocycle.
Here’s a snippet from the supporting information:
The next problem is a lack of clarity as to what they actually did next. They note that at 90°C, the formation of the macrocycle also resulted in side product of a Diels-Alder reaction, so warming to 123°C did the business and gave the product. However, I’m not entirely sure if this was done immediately following the Stille coupling, or whether the intermediate was isolated and then heated…
Either way, they got their tricycle in the end. Elaboration of this compound into the target is summarised in the partial retroanalysis above. Some smart chemistry used to finish the work. Hydroboration of the cyclohexenone was facilitated after a bit of screening using thexylborane. The configuration of the macrocycle seemed to be enough to steer the boron reagent to complete selectivity, and also helped methylation at the C-14 position. Lastly, there was some nice chemistry in the penultimate step, as they were able to do a carbonylative coupling of a vinyl triflate to give the conjugated ester, and epimerise the acidic C-8 position to the desired face in one step. A very interesting total synthesis!