Home » Still In The RBF

Reidispongiolide A   

11 July 2007 5,090 views 9 Comments


Paterson, Ashton, Britton, Cecere, Chouraqui, Florence and Stafford. ACIEE, 2007, EarlyView. DOI: 10.1002/anie.200702178.

I think we’re going to do a bit of a macrolide spectacular this week, as there’s been quite a few publications of late. I’ve been privy to the progress on this molecule over the first two years of my Ph D, as our group shared meetings with the Paterson group. Therefore, I’d like to start by congratulating my colleagues back at Cambridge who worked on this impressive target: good job! The target has some interesting activity, as it interacts with Actin, promoting microfilament destabilisation and antiproliferation. As is quite common for this type of target, much can be taken from the retrosynthesis, so lets get to grips with it:


As can clearly be seen, this synthesis is heavily reliant on Aldol reactions, a discipline which the group has excelled in. Indeed, this synthesis utilises that reaction six times, deriving much of the stereochemistry. The varients they use include their lactate aldol, Mukaiyama and of course the boron aldol. Anyone familiar with Paterson’s work will be aware of the power of this reaction, but a highlight in this paper is the example below, where they were able to overturn the substrate induced selectivity in close to 2:1 preference for reagent control. The vinyl iodide moiety must have required some care, as they tend to decompose on exposure to light; I’m astonished they managed to carry it through so many steps!


A second impressive example is the final aldol (which I’m not redrawing cause you can figure it out yourself!), appending the C-30 to C-36 fragment. This was completed as the third-to-last step, working with elaborate fragments, and must have been quite a scary step to complete.

Once again, well done folks!

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)


  • carbazole says:

    Can anyone comment on the use of peroxide in the workup for those boron-aldol reactions? Is there a reason you can’t just hydrolyze the borate ester that’s formed? Also, I find it very frustrating when a supporting information includes no procedures, and only spectra. Especially for total synthesis, the fact that you made the thing doesn’t matter much, but the procedures used might be applicable in other forums with similarly complex substrates.

  • Spiro says:

    Hi Paul, good to see you back.

    I find it inept that the authors blame the aldehyde for the low diastereoselectivity in the mismatched coupling of 21 with 20. There is another actor to blame: the chiral promoter (DIP Cl), used over and over again. Had they tried a few of the common promoters for asymmetric aldol coupling, Science would have progressed, but what they did here was pointless.

    To Carbazole:
    Actually you get the borinic ester ROBR’2 . Peroxide workup will give you borate esters B(OR”)3, which, I guess, hydrolyze better and will end in the aqueous phase.

    And yes, I find it very frustrating too that there is no procedure. Especially for this article which I find unsteady on several aspects. If you look at the carbon NMR, you can assume that they either acquired it with 16 scans or that they never synthesized more than 1 mg of the target. That’s embarassing when you claim 0.9% overall yield for a demanded compound… I’m just saying…

  • TheEdge says:

    In addittion to the Felkin-Anh preference of the aldehyde, they’re also fighting 1,5-anti induction from the methoxy stereocenter on the ketone. The only way to really shut it down (while using a boron aldol) is to stick a silyl protecting group on the offending stereocenter. I’ve seen cases where a fully matched ketone and aldehyde will give the same (single) diastereomer with Bu2BOTf and both enantiomers of ipc2BOTf. I’m impressed that they got what they did.

    One more note for carbazole-the trimethylborate that is the product of the oxidative workup is volatile, so you can azeotrope it away from particularly “sticky” substrates. This can be handy if silica won’t clean up your material.

  • TheEdge says:

    I just realized that the Felkin product wrt the aldehyde is the desired diastereomer, so they don’t have a matched case and are just fighting 1,5 induction. They actually comment on this in the paper.

    Sorry about the confusion.

  • milkshake says:

    Trimethyl borate will not survive aqueous workup (it hydrolyzes realy easily). The main thing with peroxide workup is that the boronate with two isopinacolyl substituents is a high-molecular-weight greasball that slowly oxidises into other crap, streaks along and gums up your product – so you better destroy it before trying to purify your compound.

  • carbazole says:

    Thanks everyone. I’m trying one of these reactions with a Cy2B-enolate, and had been wondering why it was neccessary.

  • Spiro says:


    One more time?

  • aldo says:

    I think sometimes some credit is deserved where necessary. This group is expert at aldol reactions. Particularly boron-mediated aldol reactions. I work with aldols too and am also surprised to see such a good turn-over of selectivity on the aldol in question. Perhaps the avoidance here of “common asymmetric methods of aldol promotion” confirms or highlights the tendency of these beta-alkoxy/delta-olefinated substrates to eliminate into conjugation under the lewis acidic conditions used in most “common asymmetric aldols”. Also note the iodide in the case of a lithium aldol! It is HIGHLY unlikely that this group simply didn’t try these methods. I suggest you contact the author for an explanation as it seems an intriguing problem. As for the amount of compound produced, try not to criticise too much spiro…..I think most synthetic chemists will agree that a natural product is a natural product none-the less if isolated cleanly, which this clearly is. If spiro, you are indeed a natural product chemist, then you must too be capable and confidently familiar with manipulation of reactions on micro-gram scale. I know we are in our lab. Bear in mind that most in vitro biological activity assays generally require no more than half a mig, and that most total syntheses of compounds of this complexity rarely yield much more than this, especially on the first report of completion. I would like to stretch my congratulations to this group, which continues to be first in line on many complex macrolides.

  • aldo says:

    just a further comment….I think you may have mixed up the selectivities obtained. The Felkin-Anh product (reagent control) is down in the paper as the major product. Also, are aldols of the type in the side-chain appendage precedented with such a sensitive vinyl-formamide moiety present? that must have caused troubles.