Ciguatoxin

Isobe, Hamajima. ACIEE, 2009, EarlyView. DOI: 10.1002/anie.200805996. 

Also: 10.1055/s-2004-817769 , 10.1021/jo980088n , 10.1021/jo034021y, 10.1021/ol0600741 ,10.1016/j.tet.2007.03.012 , 10.1016/S0040-4020(03)00873-1, 10.1016/S0040-4020(02)00044-3… and many more.

As you can perhaps tell by the doi listings, this synthesis has been ongoing for quite some time.  And figuring out what was done, when, and how they did it has taken the best part of five hours now – the literature trail is like a tape worm, and as transparent as London tap water…  This seems to be a common theme for syntheses of marine polyethers, as Crimmin’s synthesis of brevetoxin earlier this year also demonstrates.  However, I’ve done my best to figure this one out, and here’s what I reckon is a full retrosynthesis.

Right, so the key disconnection is down the middle of the F-ring, splitting it into two similar sized fragments.  This is done about the F-G junction, so the G ring goes in at the same time, using an alkynylation to link them.  Then the F ring was installed by Co-complex displacement of an alcohol, then the dissubstituted acetylene was oxidised by removal of a cobalt complex formed earlier.  The G ring was then closed onto the resulting ketone, and the initial hemiketal reduced with BF₃.OEt₂ / Et₃SiH.

There’s a bit of an actual disconnect here, as I’ve no idea how the E ring was installed.  One of the papers describes the synthesis of an E’FGH system, but the E’ ring described is the 6-membered analogue.  And they buy it in.  Moving on a bit, the D ring was created using the same approach mentioned earlier – alkynylation, Co-complex formation, followed by displacement of a propargyl alcohol.  This brings the fragment back to a fused bis-pyran, and to more familiar strategies.  (An oxidative cyclisation formed the C ring as a hemiketal, with BF₃.OEt₂ / Et₃SiH used to reduce again).  Eventually, this brings us back to the SM, methyl-alpha-D-glucoside.

The other large fragment is a more lengthy synthesis, as befits it’s more complex nature (more medium rings…).  Formation of the H ring is done late-stage, through a particularly neat 1,4-addition.  Prior to that, an asymmetric delivery of hydrogen to an exocyclic methylene on the I ring imparts the methyl stereocenter, with formation of the ring done using Co complexes as before.  Union of the two (quite unequally sized) pieces was done using a Sonogashira coupling.

I’m not sure how they get from the starting material (a DHP) to the material show – they reference a previous paper, but I can’t find the actual steps used for this anywhere.  Any ideas where it’s detailed?

The other piece is far better documented, with a typical spiroketalisaiton used to put in the L, M system in, using a SAD to install a glycol on the medium ring.  That ring was build using the now over-familiar Co complex approach, again unifying similar sized chunks with acetylene chemistry.  Taking the synthesis right-back to the starting material, however, unearths a flaw, in that elaboration of the sugar derivatives is far from efficient.  Take the left-handside, a pyran evolved from a methyl alpha-D-glucopyranoside derivative – twenty one steps are required to get to it’s retrosynthetic parent.  Urgh.  However, as they start 1.2 kg, I guess they get the material through…

I want to write more about this synthesis – afterall, even though it’s enormously inefficient, it’s an epic feat.  I’ll put on a bit of mechanistic action about the Co chemistry, and perhaps some of the couplings, but I’d like to have readers input on this one, as it’s quite a task.  What do you like /dislike about this synthesis?