Prof. Yoshyto Kishi, 1937-2023
Last week saw the death of one of the biggest names synthetic organic chemistry, Yoshyto Kishi. He was of course a longtime professor at Harvard, and during his long career he and his group prepared molecules that would once have been thought impossible for human beings to make (for more biographical details, see here). He took his doctorate at Nagoya, and then was a post-doc with R. B Woodward. That’s a generation passing - even the youngest chemists who worked directly with Woodward are at retirement age or beyond.
There is no better illustration of Kishi’s scientific fearlessness than palytoxin, shown at right (here’s the final paper in the sequence, which came out in 1994). It’s an extremely potent toxin from some species of tropical coral, and it works by locking a key sodium-potassium channel in a completely open position and letting those ions pour through it in both directions. That is incompatible with life - life depends on gradients and stuff being more on one side of a membrane than another - and palytoxin will kill any cells it comes into contact with.
You can know no chemistry whatsoever and appreciate what an accomplishment it was to make such a compound by hand, from the ground up, and the more chemistry you know the harder it looks. Palytoxin has 64 chiral centers, which right off gives you a theoretical two-to-the-sixty-fourth stereoisomers (a bit short of twenty million million million), and that’s before you consider double-bond geometric isomers and the like. It takes extraordinary care and planning to figure out how to make such a thing, and extraordinary resourcefulness to get such a plan to actually come through, because you can be sure that some of those ideas are going to blow up on you and leave you completely stranded, aground in the tropical sun with no supplies at, like, step thirty-seven.
But Kishi’s research group was able to synthesize a whole list of compounds at this highest level of difficulty, many of them similarly toxic. They reported a racemic synthesis of tetrodotoxin in 1972, eight years after R. B. Woodward’s group worked out the structure, and a chiral synthesis wasn’t achieved for another thirty years. The poison-dart frog compound batrachotoxin A was synthesized in 1998, and the anticancer agent halichondrin B (from a marine sponge) in 1992. This work led to the most synthetically complex small-molecule drug ever, eribulin, approved by the FDA in 2010 for metastatic breast cancer. The first route to that one came out in 2001, with an improved route published in 2009, well after Kishi was an emeritus professor. And there are certainly more.
You can’t make such molecules without having to discover some new chemistry, and an example of that is the Nozaki-Hiyama-Kishi reaction, sometimes known as the NHK in what is a Japanese culture in-joke. That coupling came originally from Nozaki and Hiyama, who described it as working with chromium(II) chloride. Other groups had mixed success with the reaction, for reasons that were not clear. Kishi’s lab tried to use this reaction in the palytoxin synthesis (more obvious choices having failed!) and discovered that catalytic amounts of nickel were required to get it to work, and around that same time Nozaki et al. also realized that the variability was down to the source of the chromium chloride. If you used really, really pure stuff, only the best, the reaction failed. The cheaper reagent had trace amounts of nickel in it, and that one worked.
Through my entire career as a chemist, Kishi’s name has been synonymous with work like this, ferociously complex molecules that needed ferociously difficult syntheses to make. In fact, I would say that Kishi’s work (along with K. C. Nicolaou’s) established in many people’s minds the idea that basically every natural product (no matter how fearsome) could be made synthetically if you were willing to throw enough time, effort, money, and brainpower into it. That might not be completely accurate, but it’s pretty close to the truth at this stage.
And that’s what’s turned many people (I’m one of them) to thinking about what the next thing is in organic chemistry past total synthesis. That’s been a driver of the field from the earliest days, and during its heroic era (mid-to-late 20th century) it advanced the science greatly. But the days of “We’re going to make this molecule because it’s so fiendishly complex and no one thinks it can be done” are over. Kishi’s own work reflects this - note the halichondrin/eribulin syntheses, where the point went from being able to make halichondrin at all to being able to make something of it in the real world. You’ve always had to think about what you’re going to make, how you’re going to make it, and why, but the answers to those questions have changed over the years. If you’re doing total synthesis now, you have to pick your targets more carefully, and see if you can work in a way that can translate more easily outside the journal pages. That has its own challenges (does it ever).
Prof. Kishi worked in both eras, the earlier ars gratia artis style of total synthesis and the later applied-science one. I do not think that we shall see his like again.
About the author
Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He’s worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer’s, diabetes, osteoporosis and other diseases.