Kapakahines B and F

Baran, Newhouse, Lewis. JACS, 2009, ASAP. DOI: 10.1021/ja901573x. 

It’s always nice to be surprised by a synthesis dropping from nowhere… even though the commenting team here at Tot. Syn. mentioned an impeding Baran synthesis of an Indole-containing natural product.  And here it is, breaking a unmentioned Tot. Syn. rule – I don’t do peptides.  I’ve never mentioned this before, and no one has ever complained, so I guess I’m not the only one!  Kapakahine B gets out of this quandry by also featuring a rather tasty tetracyclic indole centered moiety, and a selectivity issue built into the macrocycle.

The synthesis starts with a rather nice bit of Knochel, allowing addition of an alkyne to the primary iodides without disturbing the acidic chiral center.  What’s going on here is firstly an insertion of a copper/zinc complex into the alkyliodide, and then attack into the haloacetylene.  Nice, as this acetylene leaves them perfectly set for a bit of indole building later.

The next bit of chemistry is distinctly reminiscent of work done by Baran and Newhouse last year, working towards psychotrimine.  As I said in that previous post, “their key reaction … involved an oxidative coupling of an indole with an aniline, and then cyclisation of the resulting imine with a pendant amide to form the aminal centre”.  And this trick works as well as in the previous synthesis, with a similar yield, and in this case, with impressive induction of stereochemistry.

After doing a Larock indole-formation with both the fragments above, the group had the right-had intermediate shown below.  The group’s bet was that under amide-forming conditions, the aminal center will equibrilate between the two forms on the top of the figure.  The key point is that one has a free, primary amine, whilst the other has a secondary amide (with the nitrogen lone-pair tied-up into the pi-system).  This means that although the 6,5,6 system is less favoured, it reacts in the cyclisation reaction far faster, pulling the equilibrium towards it.

And damn, their bet paid-off.  A 70% yield, with the majority in their favour is a fantastic success, using non-too-special reagents.  (BTW, HOAt = 1-Hydroxy-7-Azabenzotriazole).  What more can be said?  The rest of the synthesis is peptide-bashing, allowing completion of two family members in 12 and 14 steps.  Top banana.