Home » Still In The RBF


28 January 2008 6,162 views 10 Comments

Hoppe, Becker, Bergander and Fröhlich, ACIEE, 2007, EarlyView. DOI: 10.1002/anie.200704678. Article PDF Supporting Information Group Website

We (that is Tot. Syn. at least) met Vigulariol just over a year ago, with a nice racemic synthesis by Steve Clark, covered on these very pixels. However, a recent asymmetric synthesis by Dieter Hoppe was brought to my attention (by one of the authors, no less – any questions, just get in line…), featuring one hell of a domino reaction!

I didn’t discuss the properties of the natural product in the previous post, so perhaps it’s time for a quick resume; it’s a cytotoxic marine diterpene of the cladiellin (eunicellin) family, of which I have a little familiarity. However, the cyctoxicity isn’t particularly potent, so let’s stick to the tasty chemical structure as a raison d’etre.

The synthesis effectively stems from a three-component-coupling, with each component the product of a short synthesis. Somewhat nicely, the central cyclohexene fragment (which was originally the product of a seven step synthesis) could be derived from a component of commercial eucalyptus oil. However, it is only present as 5%, so that must be a fun column!

The excitement starts with a substrated directed deprotonation of the cyclohexenol. The carbamate moiety ligates the lithium, which is also chelated to by the trans-tetramethyl-1,2-diaminocyclohexane. They then dump in the Ti, which performs a metal exchange, and then a Ti-mediated aldol with the aldehyde fragment, returning a reasonable 83:17 d.r. The mixture of diastereoisomers was then treated with Lewis Acid to promote the condensation onto the third fragment, the diethyl acetal. This set two further stereocentres with apparent complete control. Top banana.

Although we’re only a few steps in, they finished the synthesis with another four. A key RCM reaction tied the nine-member ring together – no mean feat, as several other groups working on this type of natural product has serious problems with RCM. Their RCM went in reasonable 45%, also yielding an eight-member ring analogue in which a methylene unit had been “snipped” out. This is annoying side-reaction of metathesis catalysts, as it destroys the SM and creates a mess, particularly in cross metathesis. Any-ho, next was a substrate controlled epoxidation with DMDO, which went in good yield, preparing them for the last cyclisation.

A bit of hydrogen loosened the benzyl protecting group, allowing THF formation, completing the carbocyclic skeleton of the natural product. All that remained was a Wittig methylenation of the ketone to complete the target. Although the yields for the reactions I’ve shown aren’t fantastic, that belies an incredible strategy for the synthesis of a complex target.

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)
Loading ... Loading ...


  • milkshake says:

    sorry for nitpicking: in the scheme you are missing some oxygens: first the OBn group in the aldol intermediate and then also the oxygen bridge of the THF ring, in all intermediates. (For few minutes I was scratching my head tryin to figure the Meerwein rearragement that turned one carbon from gem-diMe group into the -CH2- in cyclopentane ring bridge until I realised it should have been OBn and -O- bridge in a THF ring)

    Also columning 5% of eucalyptus oil and getting quickly gram quantities of intermediate in two steps rather than 6 is not that unreasonable. I did thesis on natural alkylphospholipids that are present in egg yolks and are quite air-sensitive so they have to be extracted freshly. Average egg yolk has about 4-5 grams of lipids which conatin about 1 gram of phospholipids which contain 0.2 g of phosphatidylethanolamines and 7% of these PE was the actual stuff that I wanted… anyway the large-scale extraction, fractionation and column separation took less than 3 days – but for awhile I was starting my day in the lab by cracking few dozens eggs…

  • TheEdge says:

    FWIW, the first step is a homoaldol reaction. There is no enolate, and the product isn’t a beta-hydroxy ketone equivalent. I’m impressed that they got as much control over the hydroxy center as they did. 4:1 isn’t bad at all. I’m curious to see if they explained the selectivity in the paper.

  • star says:

    i dont understand two things.One is that in final step they used 3.5 equ of base for wittig reaction still they got 93% yield. Other one is in RCM, they got 17% yield only for product 22.

  • earth23 says:

    @ star: 17% is their yield of the byproduct where n = 0. The actual compound that they want (n = 1) they got a 45% yield on.

  • Tot. Syn. says:

    I think I’ve fixed the structures, milkshake. Thanks for that. And, Edge, I stand corrected… thanks also!

  • Xiu says:

    Ya..I wonder how this paper ended up in Angew Chem ! I would say a JOC would have been more appropriate.

  • Liquidcarbon says:

    Paul, there’s another platensimycin (ACIE)… do you want it or will you let me write it up, ’cause I did one in your absence :)

  • Liquidcarbon says:

    oh noes, sorry, it’s actually platencin!

  • Tot. Syn. says:

    LC: You can have that one if you like :). The homo-allyl radical rearrangement is nice, but I think I’ve covered platensimycin and platencin enough here. Have fun!

  • stereoelectronis says:

    Hey Star. They need 3.5 equiv. wittig to quench the tertiary alcohol first and then for the formation of the ylide. That’s not as trivial as it seems.