Totally Synthetic has already touched upon the recent structure debate about Maitotoxin in which the latest piece is penned by Nicolaou in Angew. Chem. But since I am planning to present the controversy for a group meeting, I thought it may be worth summarising the main points.
1. In 2006, Spencer and Gallimore speculated (DOI: 10.1002/anie.200504284) that the original structure of Maitotoxin may be wrong for the J/K ring junction, where the stereochemistry is opposite to what it is at all the other ring junctions, not to mention at all the ring junctions for lots of other similar "polyether ladders"
2. Their reasoning was based on biosynthetic considerations. According to their analysis, all the other rings of Maitotoxin and those of other polyether ladders can be envisaged to arise from sequential trans epoxide openings, either all (R,R) or all (S,S). For the J/K ring of Maitotoxin, there would be a deviation from this path, since formation of those two rings would necessitate two consecutive epoxides, an (R,R) followed by an anomalous (S,S) (or vice versa). This is not in line with their hypothesis, since they reason that a single epoxidase enzyme would be most convenient for forging all (R,R) or (S,S) epoxides. Suddenly having only one expoxide that's different from all the others would put nature in a fix, and maybe require a separate epoxidase for just that one epoxide, not a comfortable situation according to Ockham's Razor. Which of course says that the simplest explanation is the most plausible...or in the Bishop's original words, "entities should not be unnecessarily multiplied". In this case, those words would be pretty accurate, because one would need an "unnecessary multiplication" of epoxidases, one for forging one epoxide, and another for all the others.
3. Enter Nicolaou (DOI: 10.1002/anie.200604656). He first bypasses all the biosynthetic epoxidase stuff, and goes straight to a simpler fact; Spencer and Gallimore's structure does not match the NOE data! Then he comes up with two other structures, one of which is the original one and naturally satisfies all the NOE data, and the other one which satisfies more NOE data than the Spencer structure, but still retains Spencer's hypothesis about the J/K ring stereochemistry. However, even this structure does not match all the NOE data. The structure choice he makes by using quantum chemical NMR C13 chemical shift calculations, in which the difference between calculated and observed chemical shifts is the least for the original structure. In the end, Nicolaou comes up with an alternative biosynthesis pathway involving a C-glycoside type structure, that does not need this pair of anomalous epoxides. Naturally, this biosynthetic pathway is for the original structure.
There are several possibilities here;
1. Either the NMR data and therefore Nicolaou's analysis is wrong OR
2. Nature does not follow Ockham's Razor, and the biosynthesis of the J/K ring junction does not have to be in line with that for all the other rings including those of dozens of other polyether ladders. That is, there actually could be two different epoxidases, or some totally different mechanism for the formation of the J/K rings.
To me, both these possibilities seem likely. However, we must not forget that the structure of Maitotoxin was derived very painstakingly over a long period of time by Yasumoto, Kishi and others. Unless there is evidence to the contrary, I don't see any reason to doubt their NOE data. In fact, I am a bit surprised; in their drive towards biosynthetic explanations, did Spencer and Gallimore simply forget to see whether their proposed structure matches the NOE data or not? Clearly, if the NOE data is correct, it does not. At the same time, it's also clear that the signals in this particular region strongly overlapped in the NMR spectrum.
On the other hand, Ockham's Razor is really a human invention. "Simple" and "complicated" are frequently functions only of our minds. Nature in the past has proven to be too clever for human beings to always be able to pin it into some predetermined or standard pattern. What we called complicated before has often turned out to be the simplest and most elegant explanation in hindsight (relativity? quantum mechanics??). So Maitotixin may well again provide us with the message that Nature loves exceptions as well as unity.
So either the NOE data will be revised and/or we will discover a new mode of biosynthesis. In any case, we are the ones who gain.
Sixty-four years later: How Watson and Crick did it
21 hours ago in The Curious Wavefunction