Cortistatins A (J, K and L) Pt. IV
Myers, Flyer, Si. Nat. Chem., 2010, AOP. DOI: 10.1038/nchem.794
Cortistatin is one of those targets like platensimycin that has attracted an almost unhealthy synthetic lust in the last few years, with three previous appearances in these interwebs (Baran, Nicolaou, Shair). The reasons for both are the same – both have remarkable biological profiles, and both are of challenging complexity. However, both are now what I’d call mature targets – in that the best disconnections have been tried, and that there are unavoidable methods that feature in most routes. However, that doesn’t mean that the science is over – now it’s time to refine the chemistry to the nth degree, and allow for maximum flexibility. And that’s more-or-less what Andy Myers has got going on here – a synthesis of four family members from a common intermediate – 5 – shown below.
Myers’ route to 5 is all about contruction of the 7-member B ring – using metathesis to close it, and a metal-coupling to build the acyclic intermediate. This leaves a relatively simple A ring intermediate, and the more complex C,D ring fragment. Myers’ doesn’t take this compound back to the Aldrich bottle, but starts instead with a literature compound containing much of the complexity. Keen organocatalysers will probably recognise this compound’s heritage in the Hajos–Parrish enedione, with several steps required for elaboration. I really like these few steps leading to the coupling component – a neat conjugate addition of triphenylphosphine and protection in the first step, then the world’s smallest Wittig coupling in the second. The third step is something I didn’t even know was possible in one step – a concomitant deprotection / triflation sequence. Nice stuff!
As I said, this triflate was coupled with the A ring fragment, and the the exocyclic alkene RCM-ed to form the seven member ring. An epoxidation installs the free hydroxyl shown in my next scheme, with a fairly forcing hydrogenation taking care of the styrene unsaturation. Then it’s time for an oxidative cyclisation (also used, and credited to, Sarpong), dearomatising the phenol and installing the oxo-bridge. Three more steps were required to get to 5 – firstly a neatly selective hydrosilyation / reduction to reduce the less substituted alkene. A bromination with NBS diastereoselectively installed a bromine, which was displaced by the desired azide in excellent yield. Reduction of the ketone was more problematic, though, as typical substrate-controlled methods were unselective, or leant towards the wrong diastereomer. A bit of CBS reduction did the trick, though – which some might frown at as ‘cheating’. I’d take the good yield and go to the pub, though…
From this intermediate Myers goes on to make four Cortistatins – two for the first time. I’m not going into the details, ’cause I don’t find the chemistry quite as appealing. However, I’ve summarised the route effiency below – and it’s hard not to be impressed with the flexibility of his route. Nice stuff, and a must-read paper!