I’m still wading through the bumper-Tetrahedron issue flagged-up in the comments, but halfway through reading Amos Smiths spongistatin epic I noticed a new Danishefsky paper in JACS. So whilst I man-handle Amos’ tour-d’force into bite-sized-chunks, here’s a smart slice of typical Danishefsky.
Initial work on the target was published in a rather nice Tet. Lett. last year, and that’s where the synthetic action starts. A well known Robison Annulation created a 6,5-fused enone in a racemic fashion, which Danishefsky cyclopropanated in a rather interesting way. Enolisation and silylation gave them an activated alkene, which was treated with a carbene derived from dichloroethane. This method isn’t new by any means, but I like how effective it was. Treatment of the somewhat volatile chloro-cyclopropane with silver nitrate allowed an oxidative cleavage of the cyclopropane in a endo fashion to give the ring-expanded product. Nice.
A couple of steps later, and it was time to create another ring, and like the previous modus-operandi, Danishefsky firstly installed a small ring and then ring-expanded. In this case, treatment of trichloroacetyl chloride with zinc and a bit of ultrasound cause formation of a ketene, 1 which did a [2+2] cycloaddition to give a dichloro cyclobutane as a single diastereomer. A bit more zinc (and a few protons) allowed reduction of the gem-dichloride (without rupturing the cyclobutane, which I would have been worried about), which was ring-expanded to a gamma-lactone using a baeyer villiger oxidation. All in a rather spiffing yield.
[However, I can't help but wonder if there might have been a faster route... using manganese triacetate and acetic acid. A good chunk of my readers will remember that I did my PhD on this stuff, mostly doing radical cyclisation chemistry with malonates. However, the prototype reaction is with olefins and acetic acid, in which the acetyl radical is generated, and adds to the olefin to form a gamma-lactone. However, I think it would add to the wrong end of the olefin, resulting in a diastereomeric product. Worth a pop, though...]
Not much more needing done, to be fair. Oxidation of the freed alcohol to the ketone and alkylation using cerium chloride and the requisite Grignard reagent appended the required sidechain for tricholomalide B, and using a spot of base, that could be converted into tricholomalide A. Neat and efficient work, if racemic (though Danishefsky seems to favour racemic syntheses just now…).
[1] I wonder if one could apply other sources of sound to get this to go. My Pinky and Perky record (unfortunately in my Mum’s loft) gets pretty high-pitch.
No comments:
Post a Comment