McDonald and Tong. ACIEE, 2008, EarlyView. DOI: 10.1002/anie.200800749.
An intriguing approach to this molecule, McDonald has decided upon a biosynthetic route, in which the ‘molecular zipper’ idea is employed. This idea has been pondered upon and used for years (with a notable paper by Jamison in Science last year). However, impressive as the biomimetic transformation is, quite often the synthesis of the substrate is a lengthy and drawn-out affair. Not so in McDonald’s work. This is actually a second generation synthesis – a first attempt was published in JACS last year, and uses the same principles, but with a slightly clunky implementation.
The two main fragments in this synthesis are derived from natural sources. The top number resulted from an impressive chemoselective bis-epoxidation of farnesyl actetate under Shi conditions, whereas the bottom piece came from alkylation of geranyl bromide (and some other stuff…). That allyl silane certainly looks conspicuous, and those of you with a good named reaction knowledge will be ready for the funky Brook rearrangement / alkylation that generates the silyl enolether in moderate yield.
And bosh – ready for a bit of biomimetic chemistry. A bit of TMS-triflate, some bulky base and watch it go. (They suggest that the TBAF is only there to mop up a silyloxonium ion intermediate, but according to the supporting information isn’t strictly necessary – but does give a better yield.)
Next up – more epoxidations using the Shi catalyst, and a Wittig methylenation. Then we’re set for the final biomimetic cyclisations, but unfortunately in a less impressive yield. Several other isomers of the final target were also produced, both of which I believe could be converted to the desired product… thoughts? However, let’s not detract too much from what is an impressive achievement none-the-less.
I’d like to leave it there, but I can’t because of the supporting information. Or, rather, the lack of it. Why no spectra of the final synthetic sample compared to the isolation sample? They compare this with the data for their first generation synthesis, and I’m sure the work is kosher, but I would have though Angewandte would like some spectra in the SI…
It’s also worth noting their explanation for an ‘ent-’ synthesis – and it’s all about the Shi catalyst. This reagent (as you all know) is derived from naturally available fructose, but only as the abundant enantiomer. To get the correct stereochemistry for the natural product, they need to epoxidise with the catalyst derived from L-fructose, which is pretty damn expensive. However, Shi himself sought to rectify this situation with a short synthesis of the requisite catalyst a few years ago.
Tong, R., McDonald, F.E. (2008). Mimicking Biosynthesis: Total Synthesis of the Triterpene Natural Product Abudinol B from a Squalene-like Precursor. Angewandte Chemie International Edition DOI: 10.1002/anie.200800749