Our paper describing the total synthesis of Teleocidins B-1–B-4 is now published in JACS! I started working on this project in November 2017 and completed the synthesis in June 2018. I then spent another 6 months optimizing the route to increase the yield and decrease the step count.
In the biosynthesis of these molecules the indolactam core is transformed into Teleocidin by a Friedel-Crafts type reaction with a terpene fragment. However, nature’s synthetic strategy provides a mixture of Teleocidins B-1-B-4 which is something we wanted to avoid. When I started to evaluate our own synthetic strategy, I immediately identified the introduction of the two quaternary carbons to the indolactam core in a stereocontrolled fashion as the major challenge (even one sterically hindered quaternary carbon can be difficult!). I knew we would have to establish the relative stereochemistry independently because the terpene fragment and amino acid moieties are quite distal.
I also realized we needed to install a functional handle on the indole moiety in a highly regioselective fashion to facilitate coupling with the terpene fragment.With a bit of luck (and a lot of hard work) we were able to overcome these challenges and you can find the details in the paper/SI. I wanted to use this blog post to provide further insight into some of the key reactions which led us to success.
In the beginning, we had to fight against nature…
The first step in our synthesis is an electrochemical amination between our indole scaffold and valine. After optimization, I tried to scale up the reaction using some conventional large glassware which I bought from our local Japanese supermarket. However, almost no desired product was observed and even worse, I lost 6 grams of starting material along with the nickel, ligand and my time. I was so shocked that I couldn't even find my tongue! Phil suggested to try again but this time using the ElectraSyn carousel. The switch in setup facilitated reactivity and with a little optimization the desired product was formed in 51% yield.
I also want to highlight one of the unexpected problems we faced in this synthesis which is not described in the paper. While the desired indolactam is commercially available, it costs $274 per mg (Sigma) so it was a bit outside of our budget! We first tried to make the 9-membered ring by hydrolysis of the valine methyl ester followed by condensation using HATU, but the desired product was isolated in just 19% yield. However, after plenty of screening we found that the use of LDA was crucial for this reaction and we managed to obtain 2.1 grams of the protected indolactam! That was a really great day in the lab.
In April 2018 I was joined by Kosuke, a very talented visiting student from Osaka (Japan), who really helped me to optimize the route to what is now shown in the paper. Yuzuru, an amazing graduate student, also helped me by optimizing a key reaction and preparing a late-stage intermediate. I want to thank the team and especially Phil who was always ready with a lot of suggestions. I can definitely say he never compromises in the quest for an ideal total synthesis.
Hugh and Teleocidin team