Organic Chemistry Portal
Reactions >> Total Syntheses

Totally Synthetic by Paul H. Docherty, 8 October 2007

Total Synthesis of Pasteurestin A

Mulzer

M. Kögl, L. Brecker, R. Warras, J. Mulzer, Angew. Chem. Int. Ed. 2007, 46, 9320-9322.

DOI: 10.1002/anie.200703457

A second paper in Angewandte from the Mulzer lab in a few weeks, this synthesis shares a little with the previous post on Platensimycin. Both natural products contain tightly functionalised polycyclic systems, and both are potent antibiotics. However Pasteurestin A&B are only active against bovine illness. This synthetic work consists of the first synthesis and also confirmation of assignment, so let's get into the chemistry.

The chemistry really gets going with an auxiliary controlled Reformatsky addition, using an unsaturated aldehyde. This tin-mediated process gave them an interesting result, in that the stereochemistry of the hydroxyl was (S-), not (R-) as in the aldol reaction with this system. They presumed that this must be due to the low temperature at which they ran the reaction, so they repeated it at RT, and found that the product rearranged before they could isolate it.

A few simple synthetic transformations then gave them the starting materials for the interesting Vollhardt [2+2+2] cycloaddition, which went in moderate yield to form three rings in one reaction. Very efficient! The next step was interesting too, but for different reasons… the selectivity here is impressive, maybe based on delivery of one electron to give the allyl anion, which is then protonated to give the most stable alkene.

The next few transformations are far more easily understood, and complete Pasteurestin B in a synthetically pleasing manner. Hydroboration of the remaining alkene and oxidation gave the ketone shown, which was selectively deprotonated, trapped with carbon dioxide and methylated to give the methyl ester. This was then deprotonated again, selenated and the selenide eliminated to give the desired enone.

Deprotection then gave Pasteurestin B in what was admittedly a rather linear route, but they were able to use a similar approach to get Pasteurestin A, showing the flexibility of the chemistry. Neat work.

Selected Comments

9 October, 2007 at 0:07, Andrei says:
Ok…I used this Li/NH3 reduction on dienes, hoping to have a 1,4 but ..surprise I just reduced only one of the double bonds…I think it’s the most stable one that survives but depends a lot on the conformation. In their case we have a lot of strain release, in mine just to much strain introduced. Working at low temperature helps also the kinetic conditions.
9 October, 2007 at 2:08, jimbo says:
Bulky acid, low temp… my guess is that it’s a kinetic phenomenon. Although it very well may also be the thermodynamic product, I think you would have to argue that the protonation step is reversible before invoking a thermodynamic equilibrium. I can’t really imagine tBuO- yoinking a proton off of any of the other possible olefinic products at -78°C… then again, I’m not exactly a walking pKa table.
9 October, 2007 at 17:02, Elwoodcity says:
What is the copper doing in the vollhardt cyclization? I’m not familiar with this reaction. Does anyone know, or is it simply “copper additives improved the yield so we added it” chemistry?
9 October, 2007 at 20:19, aa says:
Elwood: I believe the copper is added in a second step (one-pot). First you irradiate with Co, to give a Co-complexed arene. Then add Cu (II) to oxidize the Cobalt and get it off the benzene.
10 October, 2007 at 13:06, ElwoodCity says:
Or a cyclohexadiene, in this case. I imagine you could do it with three alkynes to give an arene, right?
10 October, 2007 at 15:19, excimer says:
elwood: you sure can, though that cyclotrimerization is typically done with Co2(CO)8.
10 October, 2007 at 15:44, Liquidcarbon says:
I think strain release explains the selectivity of Li/NH3 reduction, too.
Does LAH make enolate from bromoisobutyrate? Why don’t they use conventional deprotonation?