Haouamine A

Baran, Burns, Krylova, Hannoush. JACS, 2009, ASAP. DOI: 10.1021/ja903745s. 

Oh, we’re going way back here… a whole three weeks before I started blogging!  This brings us all the way back to Baran’s first conquest of haouamine A (and his eleventh paper), racemic, and in this JACS. It was followed up by an Angewandte two years later, but I’ve apparently only covered Weinreb’s synthesis, also back in 2006.  Part of the reason for all this interest is that the molecule seems to exhibit an unusual isomerism.  Baran realised after his initial work that thee were two isomers of the target, configurationally identical, and sharing the same stereochemistry about the 6,5-ring junction.  This meant that the stereochemical divergence must be unusual – with two candidates – atropisomerism of the bent phenol or slow pyramidal inversion at the THP nitrogen.  The only way to probe this effectively was to attempt a synthesis of the two atropisomers, and compare with the isolate.

The synthesis, yet again, is racemic, and proceeds from an intermediate in their first synthesis.  As I didn’t cover that paper, I’ll describe it’s synthesis firstly to make the route more transparent, and secondly ’cause it’s pretty sweet.  The SM was produced very quickly, using a standard enolisation and trapping with an alkylating agent to generate the quaternary center.  The ketone used was then hydroxylamine-d, and treated with a source of electrophilic bromine, causing a 5-exo-trig cyclisation and formation of a nitrone.  Reduction of the nitrone produced the intermediate I’ve shown, which after tickling with a bunsen gave an aziridine.  Rearrangement of this allowed ring-expansion to give a tetrahydropyridine-N-oxide, reduced with a bit of indium.  Nice work!

A simple (but well executed) Suzuki coupling allowed completion of the macrocyclisation precursor (and an Appel reaction too…), which was simply de-Boc-ed and treated with base to prompt alkylative cyclisation in a pretty decent yield.  Interestingly, a stereochemical bias was found where none would be expect – even if slight.

The key reaction was the oxidation of this cyclohexenone to a fully-aromatic system – a reaction with a wealth of possible reagents.  However, some of the more obvious choices, such as manganese dioxide or palladium prove ineffective, so they leapt on a Mukaiyama protocol.  Initial developed to introduce ?,?-unsaturation to ketones, this one-pot procedure did a pretty decent job of oxidising the chiral cyclohexenone to a planar-chiral phenol, returning a respectable mass-balance.

A few more steps were requred to complete the targets, which Baran mentions was also done with enantiomerically pure material (perhaps after doing a bit of semi-prep chiral HPLC).  This gave them both atrop-isomers, a few crystal structures, and an opportunity to biologically profile both.  As it turns out, their IC₅₀s were with in analytical error…