Biomimetic synthesis

Lately I've been getting really into biomimetic synthesis.

It's interesting to me because one criticism I hear a lot of total synthesis is that it doesn't ask questions. Or rather, the question is, "can we make this huge-ass natural product with eight bajillion stereocenters?" and the answer seems to always be "with enough depressed, overworked graduate students, yes?" It's been described to me as the "climbing the mountain" metaphor by a professor, whatever that means. Actually, it's mostly molecular biologists, I've noticed who go on this Theory of Science rant about the scientific method (observation, question, hypothesis, experiment, knowledge or some variant of that) and how synthetic organic chemistry doesn't follow it. It's a favorite rant of a physiology professor of mine's.

Then, as far as I can tell a lot of mechanistic organic papers are like "so dude, we noticed this stereochemistry forming and thought that was cool so we invoked a Hatree-Fock 31-G transition state model and then examined some MO secondary orbital effects and tried to rationalize the stereochemistry and found out that this other cool reaction also occurs so we examined that too and it occured in an x:x dr because of blah (probably)". I had a physical organic chemistry class where the professor was really into asking us questions like "is this paper hypothesis-driven or a data-driven?" and then he and I would get into long discussions in his office/email exchanges afterwards about it because I nearly always would claim that the paper read like a fishing expedition. But I digress.

Biomimetic synthesis is definitely hypothesis-driven, because it asks the question, "is this putative biosynthetic pathway plausible?". Furthermore, the stereoselectivity lends evidence as to whether the reaction is enzyme-mediated or spontaneous. If the natural product is racemic, it's probably spontaneous, but if it's enantioselective, it's probably not (but, as one person pointed out in a paper I read recently, the cell is a chiral environment, so perhaps that is leading to some stereoselective effects that are difficult to predict) . I read of this one example of a biosynthetic Diels-Alder reaction that only resulted in the exo product, despite the endo product being highly favored in the RBF. Or, if you can only get stereoselectivity with a Lewis acid, that indicates that there is probably an enzyme there performing the role of the Lewis acid. And this is really cool, because you can think like a synthetic chemist, but you can also ask questions about how the world works. And not just at an abstract theoretical MO level, but at something that's a little more discrete to me--a biosynthetic pathway leading to a metabolite structure.

I guess it's also appealing to me because I finally have enough of a background in pericyclic reactions (due to the aformentioned professor and the aformentioned physical organic class, which was called "Advanced Mechanistic Organic Chemistry") to make sense of all these 8π-6π electrocyclization cascades and Diels-Alder reaction selectivity nuances and Cope rearrangements and [1.7] hydride shifts. And some of the structural rearrangements are wicked. They are generally basically just pericyclic refoldings of polyenes derived from fatty acids or polyketides. It blows my mind that something so simple can re-arrange into such a vast structural diversity.

The other thing that occurred to me while reading these papers, is I recall (sorry, I'm too lazy for refs at the moment, although I really should link most of these papers) that in the past few years there have been several papers on "On Water" Diels-Alder reactions. Sharpless has done a bunch of work on this, and the principle is the same principle that drives most of structural biochemistry in terms of lipid structure and protein folding: the hydrophobic effect and solvent entropy. There have been some pretty impressive rate and stereoselectivity enhancements from doing Diels-Alder reactions on water, since the reagents are forced to be in close proximity, minimizing the entropy of finding one another in a reactive conformation relative to being dispersed in solution. Since many Diels-Alder (and other pericyclic) reactions require thermal conditions that are far from physiological (beyond weird thermophilic archea and such), it's interesting that there is such a rate enhancement from these on-water reactions, and I imagine that there's a lot of work in the future that needs to be done to elucidate exactly how the role of the cellular aqueus environment impacts the kinetics of these biosynthetic reactions.