Cossy, FerriÃ©, Reymond and Capdevielle. Org. Lett., 2007, ASAP. DOI: 10.1021/ol070670a.
Another synthesis of this popular target, and whilst not mind-blowing, certainly a nice strategy for one of my favourite targets. However, I should point out that the synthesis is actually only a “formal”, as they didn’t bolt-on the oxazole sidechain. Rather, they focused on the macrolide core, using three allylations and a crotylation to install four of the hydroxyl units.
The rest of the molecule was constructed using a metathesis to add the ester unit to the top polyol fragment. They then closed the nearby hydroxyl onto the resultant enone to provide the -cis pyran; nice work. Also noteworthy was the Mukaiyama style coupling of the lower terpeneoid fragment to the top fragment.
So, to the allylation / crotylations. They used a TADDOL derived Ti reagent, allowing both enantiomeric and diastereo control in the initial crotylation. Credit, however, was given to Hafner of Ciba Geigy (and by inference, the legend that is Seebach), and more of this chemistry is discussed in a rather nice JACS paper.
They later used the related allylation substrate shown below to do two of the three allylations, again with good diastereoselectivity. The remaining allylation was done in a more traditional sense, using substrate control to add allyltrimethylsilane with tin tetrachloride.
Again, not particularly novel, but a damn nice implementation of existing chemistry.