Totally Synthetic by Paul H. Docherty, 18 April 2010
Total Synthesis of Actinophyllic Acid
C. L. Martin, L. E. Overman, J. M. Rohde, J. Am. Chem. Soc. 2010, 132, 4894-4906.
I don’t think there is a single academic whose syntheses I look forward to more than Larry Overman. I’m not sure exactly why, but I think the main reason is the overall standard of his work, as well as his tendency to write fascinating and informing full-papers. He’s also one of those professors who ‘owns’ a particular type of natural product; ‘Overman molecules’ are of course, alkaloids.
This paper not only discusses a 'first synthesis' in a racemic form, but also a second generation, and more interestingly, an asymmetric synthesis, which I’ll be focusing on. So why all this effort? Well, Actinophyllic acid was identified in a medium-throughput screen, targeting thrombin-activatable fibrinolysis inhibitors. These hits interupt the biological processes that control the fine balance between blood clotting and blood thinning - clearly a worthy pursuit.
So to the first piece of exciting chemistry - the transformation of a racemic route into asymmetry. The key to their route was the synthesis of a chiral 3-hydroxy-piperidine - which may seem simple, but required serious determination. Initial attempts focused on asymmetric epoxiation (Shi, Jacobsen), but were met with appalling and poor enantioselectivity. Next the group tried an organocatalytic hydroxylation of an aldehyde. This solved the enantioselectivity problem (96%), but with poor efficiency, managing about 30% yield. Latterly, they tried a Noyori reduction of an enone, which when ozonolysed and cyclised, gave the product in an impressive yield and enantioselectivity. But that is a very large catalyst!
The product of this reaction, an acetoxy aminal, was used immediately in a Lewis-acid promoted coupling with an indole. Their efforts in producing the enantiomerically enriched starting material paid-off here, as the acetoxy group controlled the reaction, allowing the group to achieve both an excellent yield and diastereomeric excess.
Removal of the acetal protecting group (using Dibal-H, which didn’t touch the t-butyl malonate), and oxidation of the revealed hydroxyl group set the group up for the next reaction.
This impressive C-C bond formation works by formation of a pair of enolates (using LDA), and then an oxidative coupling with a rather complex looking iron reagent. However, Overman explains that it’s rather easily formed by simply combining ferrous chloride with DMF. Adding this to the di-enolates allowed formation of the tetracyclic products in excellent yield…
A cerium chloride mediated addition of vinyl magnesium bromide was somewhat complex, as when a dimethyl-malonate moiety was present further reaction couldn’t be prevented. However, using a t-butyl malonate was quite successful, allowing construction of the final intermediate after borohydride reduction and hydrolysis of the remaining ester.
The came the final reaction, and coup de grace, an aza-Cope-Mannich cascade which generates the final target in cracking yield. The reaction works by firstly addition of formaldehyde to the amine, forming an iminium ion. This then performs a Cope [3+3] rearrangement, generating a enol, which collapses in a Mannich reaction to form a C-C bond and complete the synthesis in its enantiomerically enriched form.
There is even more in this paper that I haven’t discussed - two racemic syntheses and a discussion of biosynthesis, and that the aza-Cope-Mannich may be biomimetic. Add all that to a great discussion, as well as a neat synthesis, and we get what I come to expect from Overman.