Totally Synthetic by Paul H. Docherty, 31 August 2008
Total Synthesis of Gonyautoxin 3
J. V. Mulcahy, D. Du Bois, J. Am. Chem. Soc. 2008, 130, 12630-12631.
Gonyautoxin 3 is a dinoflagellate neurotoxin that features some very strange functional groups. Unlike other large families of natural products with common functionalities, there doesn’t seem to be a universal strategy that could synthesise the majority of the family. This means we get to see a whole variety of strategies towards these targets - such as palau’amine.
Du Bois’ approach starts with the use of L-serine as the starting material, and source of asymmetry. Smart choice, as in only four steps they had dihydropyrrolo[1,2]pyrimidin-1(2H)-one product shown below. It was the cyclisation step that caught my eye, treating the aldehyde with allylamine to get the imine, and then Lewis acid mediated cyclisation. Great result; the yield is reasonable, but the d.r. is excellent.
A few steps further into the synthesis comes another sweet cyclisation; this time using a rhodium catalyst to effect formation of the final 5-membered ring. Again, the yield was excellent and the d.r. apparently complete. Mechanistic insights are somewhat less clear. Du Bois suggests an initial formation of an aziridine, and then opening of the allylic aziridine with acetic acid to give the product. He then reduced the N,O-acetal using an interesting reagent combination: Et3SiH and BF3·OEt2.
With the ring system complete, they attempted a ketohydroxylation, but this proved problematic. Instead, they did a standard Upjohn Oxidation, protected the less hindered hydroxyl and then oxidised the remaining. The completion of the synthesis was quite sweet; Pd catalysed deprotection of all protecting groups with hydrogen, acetic acid and then some ammonia gave an α-hydroxy ketone hydrate. Then, addition of DMF·SO3 resulted in selective sulfation of the C11 alcohol, and the target. A nice total synthesis!