Totally Synthetic by Paul H. Docherty, 6 April 2008
Total Synthesis of Callipeltoside C
MacMillan
J. Carpenter, A. B. Northrup, C. deMichael, J. J. M. Wiener, S.-G. Kim, D. W. C. MacMillan, Angew. Chem. Int. Ed. 2008, 47, 3568-3572.
This time a total synthesis of a macrolide, and as David MacMillan is involved, it is time for organocatalysis too. The callipeltoside family of natural products has been reasonably popular, with syntheses from the usual suspects: Trost, Evans, Paterson, and Panek. However, all of those syntheses were of callipeltoside A, so this synthesis marks the first of callipeltoside C, including a reassignment of stereochemistry of the sugar fragment. Retroanalysis:
So far reasonably predictable/sensible disconnection - and you’ll note no L-proline so far. It’s also worth pointing out here that the dienyne fragment was made using the route developed by Evans. However the fragment synthesis is where MacMillan’s favourite amino acid come in, with several organocatalysed aldol reactions.
The first of those nice little organocatalytic aldol reaction is used for the construction of the tetrahydropyran (THP) using a derivative of the ever-popular Roche ester as substrate. However, what I liked most was their use of a Semmelhack reaction to close the THP and leave a methyl ester as a useful functional handle (MacMillan references two papers. The original paper is here; Marshall’s improved conditions for acetylenic systems are here.) In their hands this went with great diastereomeric control, and left a heavily functionalised THP after only two steps. Now that’s smart work!
The next aldol reaction to look at is of the prodigious direct aldehyde-aldehyde variety. A catalytic amount of L-proline this time results in a great e.e. and decent yield of the semi-protected triol. Reaction with silyl enol-ether completes the hexose framework and leaves a differentially protected sugar (the two TIPS protected hydroxyls will react differently). Of course, we read all about this sugar synthesis in Science a few years back.
It took them a few more steps to complete the sugar synthesis (even a second time as the originally proposed stereochemical configuration of the sugar was incorrect). Conditions developed by Tietze allowed them to connect this with the aglycon.
As with many syntheses, the last step was deprotection of the silyl groups - but not with TBAF or HF.amine here. MacMillan uses tris(dimethylamino)sulfonium difluorotrimethylsilicate (TASF) - useful if there’s any water sensitivity as the reagent is completely anhydrous.
An interesting synthesis, showcasing much of MacMillan’s work - and notably only using unmolested proline as the catalyst.
Selected Comments
There is a cheap alternative of TASF actually: When you mix anhydrous Bu4NCN with 0.2 equivs of C6F6 in THF you get perfectly anhydrous TBAF solution that is also stable at RT (the sideproduct is hexacyanobenzene)