Total Synthesis of (+)-Sieboldine A
Stephen M. Canham, David J. France, and Larry E. Overman
Unfortunately I was very busy the last weeks with studying but now the last exam is written so I took the advantage and finished to review this nice paper from Overman et al..
Sieboldine A presents in my eyes a classical Overman target because of the rigid alkaloid structure ready for cool rearrangement chemistry. The compound itself inhibits electric eel AChE with an IC50 value comparable to that of Huperzine A (https://syntheticnature.wordpress.com/2009/11/23/total-synthesis-of-huperzine-a/). But the real interest for a synthetic chemist poses the unprecedented N-hydroxyazacyclononane ring which was unknown until the isolation and structure elucidation of Sieboldine A in 2003.
Retrosynthetically spoken the first step cleaves the sensitive N,O-acetal. The precursor derives from a Diels-Alder product which in turn was produced by a sweet pinacol-terminated cyclization.
The synthesis starts off with the known unsaturated lactone which was opened by diastereoselective Michael addition of methylcuprate and subsequent lactonization with iodine. Exhaustive reduction with LAH furnished a diol which was selectively monoprotected and oxidized to give the ketone shown.
The second intermediate was synthesized through a known route. Michael addition of tributyltin-cuprate complex on the alkyne and quenching the reaction with MeOH gave z-vinyl tributyltin ester. This was reduced with DIBAL-H, exposed to Mitsunobu conditions to produce the phenyl ether and converted to the iodide by halogen/metal exchange.
Next both intermediates were combined by reacting the iodide with sec-BuLi, transmetallate the lithium species with cerium trichloride and add to this the ketone (all at -78°C). Protection of the resulting alcohol, Swern oxidation of the terminal silyl ether (which was deprotected under the reaction conditions) and Seyferth-Gilbert homologation utilizing the Ohira-Bestmann reagent yielded the terminal alkyne ready for the first key step.
Exposure of this to a bit of gold and silver produced two of the four rings needed in a nice tandem Prins/pinacol rearrangement reaction.
The mechanism might look like this:
The gold attacks the terminal alkyne which in turn is attacked by the alkene through a 5-exo-dig cyclization. The resulting tertiary carbenium ion is neutralized by a pinacol type reaction of the TES-ether to give the vinylic gold intermediate which is protonated by i-PrOH.
Having most of the carbon skeleton in place the group turned their attention on the next key step. Ozonolysis of the exo-methylene group followed by neutral work—up with dimethylsulfide and subsequent phenolate elimination produced another exo-methylene group. This underwent a europium catalyzed Diels-Alder reaction with ethyl vinyl ether. Diastereoselective reduction of the ketone was followed by facial selective expoxidation with DMDO.
The resulting epoxide was opened in the presence of ethanethiol with BF3-etherate in a sweet Overman style reaction. Desilylation, Mitsunobu reaction with double protected hydroxylamine and removal of the nosyl protecting group furnished an odd looking hemiacetal. Next some carbohydrate chemistry was utilized which is completely new to me to close the last ring (if someone has access to the paper, mail me). Oxidation of the remaining alcohol and MOM-cleavage yielded at least (+)-Sieboldine in 5% yield over 20 steps in the longest linear sequence.
The Diels Alder reaction inspired me to propose the two mechanisms below. I am not sure which one is right but I favour the red one.
For a better understanding of the abbreviations some structures of the reagents possibly new to some readers:
I loved the synthesis as usual when reading Overman’s work. Especially the pinacol-terminated cyclization, the Diels-Alder reaction and the ring opening/ring closing cascade.
If you have any questions do not hesitate to ask them…