Totally Synthetic by Paul H. Docherty, 4 March 2007
Total Synthesis of Lepadiformine
Craig
J. J. Caldwell, D. Craig, Angew. Chem. Int. Ed. 2007, 46, 2631-2634.
Another elegant synthesis from Don Craig’s lab at Imperial College, focusing on lepadiformine, previously synthesised by the Kibayashi group amongst others. The biological profile, whilst interesting, is of no real efficacy, so I guess it’s popularity revolves around its interesting structure.
Their synthesis begins by formation of a exocyclic aziridine, from opening of the analogous epoxide with SES-amine and performing a Mitsunobu reaction on the resulting amino alcohol. Opening of this aziridine with a lithiated sulfone then gave the starting material (SM) for the following scheme:
Double-deprotonation of the SM allowed generation of the dianion, to which was added 2-benzyloxyethanal and then benzoyl chloride to quench. This gave the product, a spirocyclic pyrrolidine as a single diastereoisomer in a quite reasonable yield. Deprotection of the amine with TBAF, and the benzyl and acetal groups with BBr3 prompted imediate cyclisation onto the free aldehyde, producing the aminal, which was ready to be alkylated (remember it’s effectively at the aldehyde oxidation state).
Interestingly, addition of the unsaturated sidechain as a grignard reagent resulted in a decent yield of undesired diastereoisomer, but the acetylenic analogue reversed this result dramatically, resulting in addition to the desired Re face. The authors attribute this turnaround in selctivity to the difference in steric bulk of the two reagents. Conversion of this material to the natural product was relatively straight-forward, but required careful selection of a hydrogenation catalyst to reduce the acetylene. This done, the sulphonyl group was removed, and the alcohol sidechain epimerized by oxidation, deprotonation/reprotonation and reduction.
Selected Comments
I agree, since the addition is occurring from a psuedo concave face of the bicycle (although I don’t have a model in front for me to see how flat the iminium would be) the acetylide addition is probably directed. I guess I’m looking at the saturated quinoline system with the pyrrole fused in an angular fashion to generate the bowl shape. Interestingly, the addition of the acetylide also looks to occur from a pseudo axial position. Thoughts?
It seems to me that, immine formation would relieve some strain in that fused ring network allowing the alkoxide to effectively direct the addition. However, this obviously doesnt account for why the hexyl grignard was unsuccessful. I do think milkshake’s explanation is somewhat relevant, however…I am curious as to what would happen if they used the hexyl grignard with excess magnesium. I wonder if that would be able to induce a faster immine/alkoxide formation, therefore, allowing the ‘prefered’ grignard to add preferentially from the top face as opposed to the observed bottom face. Thoughts?
I am not sure but I think it is an “old school” trick to take the acetylide if the addition of an alkyl Grignard does not work (for steric or whatever reasons [and sterics DO MATTER!]) and reduce. Anyone references?
Sterics count: A values for acetylene vs. ethyl are known (0.52 kcal vs. 1.79 kcal respectively, from Eliel).
With the much-less reactive acetylide the aminal has maybe time to live around in the cleaved zwitterionic form (coordinated to Mg) and the intramolecular delivery acetylide assisted by the hydroxymethyl group anion brings about the desired stereochemistry. Whereas with the hexyl Grignard the organometalic kicks from behind as soon as the O-C bond is rupturing. So we would have SN1-like versus SN2-like scenario.