Lyconadin A & B
Smith and Beshore. JACS, 2007, ASAP. DOI: 10.1021/ja070336+
A deceptively simple looking structure, lyconadin A and the related lyconadin B were the targets of this paper by Amos Smith. The biological profile was somewhat simple, showing moderate cytotoxicity against epidermoid carcinoma KB cells, but the pentacyclic structure is reason enough to interest us, right?
The retosynthesis might not be too revealing, so I guess it’s time to look at the forward – and surprise (!!!) – it’s the cyclisations that I found most interesting. The initial cyclisation precursor was produced from two fragments, coupling by deprotonation of a hydrazone and addition of an alkylhalide. Anyway, the cyclisation goes via the planned aldol (acid mediated, as base led to majority polymerisation), but proceeded further still, performing a conjugate addition to provide the seven member ring.
Getting this product in only one step impressed the group as well as me, thought the were disappointed to learn that the protonation of the enol had occurred on the undesired face, and thus the C-12 centre required epimerisation. The method they used for this was really smart – they removed the Cbz protecting group, allowing it form the hemiaminal under acidic conditions. Reduction of this with borohydride then gave overall epimerisation and the desired product.
The same N-C bond was reformed shortly afterwards, after elimination of the hydroxyl resulting from reduction of the ketone and (a particularly nice) aminoiodation. With a few further functional group manipulations, they were able to perform a Michael addition of propiolamide to provide the final cyclisation material. Then, in one pot, they performed a decarboxylation, olefin isomerisation, and cyclocondensation – amazing! However, it should be noted that for lyconadin B (which required prior reduction of the olefin in the cyclisation material), those conditions were deemed less suitable, and they used LiCl instead (more like a Krapcho? is the TMAA is mimicking LiCl to decarboxylate?). Still, a damn fine synthesis, if a little steppy (27 steps).