If you ever worked with both computer modelers and synthetic chemists on a given project, you know that these two breeds have little in common. My personal perspective is that of an organic chemist, so I will allow myself to make some chemistry-centric comments. Let’s face it: modelers rarely provide insights that are synthetically meaningful. But I don’t want to accuse them of not knowing what can or cannot be made. Instead, tonight is about subtleties that are intuitively clear to practitioners of synthesis and might be critical to understanding and predicting the biological activity of small molecules. Take tautomerism: there is little more foundational to organic chemistry than this concept. The question is whether or not tautomerism can be relevant to docking of small molecules into receptor sites. I have dealt with many modelers and have yet to hear their appreciation of this concept, but a fairly recent paper describing a series of DNA gyrase inhibitors by Chan and Gwynn of GlaxoSmithKline provides an insight. The authors described a range of tautomeric forms of the barbituric acid portion of their inhibitor (QPT-1) and determined the corresponding co-crystal structures. In docking experiments they evaluated eight tautomers of QPT-1 and assigned them to subtly different binding sites. A tantalizing possibility is for compounds such as QPT-1 to adopt nuanced tautomeric shapes that facilitate productive interactions with the ligand-binding pocket. I think we need to pay way more attention to tautomers, especially when it comes to omnipresent amides.
An interesting paper by Lumb and colleagues recently appeared in J. Med. Chem. By way of background, approximately 15 years ago, Brian Shoichet published a provocative article, in which he presented evidence that promiscuous inhibitors often work by forming aggregates:
This was a landmark paper because it unveiled a plausible cause of false positives in high throughput screening. The formation of such false positives had nothing to do with any special protein/small molecule interaction but was, instead, a non-specific effect. The community took this to heart and went a bit overboard, as we always do, seeing potential aggregators in all sorts of settings. Lumb and colleagues present evidence that aggregate formation and specific molecular level interaction may not be mutually exclusive. This work provides an example of aggregation-based inhibition of a protein–protein interaction involving tumor necrosis factor α. The reported capacity of small molecule ensembles to mimic protein surfaces is super cool. Unfortunately, the pdb file has not been released yet, so forgive me for not providing my usual pretty picture to accompany a structural biology post. You do have an option to see some nice views of the crystallographically characterized complex in the paper. There are 5 (!) inhibitor molecules that clump together into an aggregate in this crystal, replacing an entire protein subunit. I am not sure that there is a general lesson here, but I love seeing exceptions from the rules.
I have not been able to write my posts for a while, which is partly due to travelling and partly to my attempts to finish up some papers while carrying on with teaching duties. I visited UC Davis and UC Merced two weeks ago. Dean Tantillo (Davis) and Ben Stokes (Merced) made this visit into a memorable stay. In the middle of that week came a dreadfully sad moment when I found out about the passing of George A. Olah, one of my PhD mentors. A lot has been said about George, particularly in recent weeks. His science has had transcending impact on organic chemistry and not a week goes by without me mentioning some of his insights in my 2nd year organic chemistry class. In particular, I refer to the concept of stable ion conditions. I cannot think of a more lucid idea that explains how an electron-poor intermediate such as carbocation can be made to persist in the liquid phase. I fondly recall our group meetings, where George’s lightning fast mind was on display. He was a towering figure and a true gentleman, who always came to lab impeccably dressed with a suite and tie, which is not something we do these days. Apart from this air of elegance, George was always eager to share his wisdom with students, who benefited tremendously from his wit. I could use many superlatives in describing George, but, upon reflection, one thing stands out: he was one of those people whom you do not want to disappoint. I will miss Professor Olah dearly.
George A. Olah (1927 – 2017)
Deciphering the structures of random reaction by-products is rarely on top of our “to do” list. This avoidance is understandable because constant pressure to move toward the stated goal makes us focused, which is both good and bad. Noticing reaction by-products is how one makes important breakthroughs, yet the chances of finding something truly spectacular in the “noise” are not in one’s favor.
As a research advisor, I love learning about crude NMRs. But I am also aware that our resources and bandwidth to pay attention to what comes our way are not limitless. This inevitably leads to (un)healthy skepticism. A recent paper in Organic Process Research and Development stopped in my tracks the other day because I clearly recall hearing about allene formation (by NMR) in one of our peptide coupling reactions. This observation was made while analyzing the reaction by-products and I dismissed it as “OK, the NMR might suggest there is something like an allene but it is unlikely, so let’s move on”. In the case of Knapp and colleagues, the acid used in the coupling process did lead to an allene. There is even a crystal structure of the product reported (it is a complex structure, so I am avoiding it here). The mechanism of allene formation is cool and involves ketene formation. I am not saying that definitive identification of a similar structure in our reaction several years ago would have affected the line of research we had been pursuing at that time, but I just wish we all had more time to dwell on what’s brewing in our flasks. Incidentally, I think it was BASF that has been pushing T3P as a fairly inexpensive reagent for amino acid coupling. It is a nice little molecule.
Here is a question for tonight: how many cases are out there where a single hydrogen bond dominates the regiochemical outcome of a chemical transformation? Let’s say we are comparing the protio version to its methylated congener, which is arguably the smallest steric perturbation that does not produce a confounding effect. Here is a case from Paul Carlier’s lab published some time ago in Organic Letters. If you look at the two epoxide-containing anilines depicted below, you will note that they differ by a methyl group. Their behavior is strikingly dissimilar, demonstrating how a single hydrogen bond can affect the reactive conformation. The reactions were run neat, by the way. Both outcomes are driven by the logic of trans-diaxial epoxide ring opening, yet the NH-to-O interaction flips the reactive conformation in the first case compared to the second one. If you are looking for a powerful demonstration of the Fürst-Plattner rule, there is probably no better way to show how a relatively weak interaction can control the reaction outcome.
I have always been interested in how intermediates with supposedly well-understood behaviour take an occasional detour. At the moment, I am collecting papers of this kind with the goal of writing a review article at some point. There is a lot to talk about here, particularly because the “roads less travelled” are often influenced by subtle structural changes in reaction components. The trouble is that it is not easy to find these cases. Let me illustrate what I mean by using a process you all probably know – the Staudinger reduction of azides. The example below is not new, but is nonetheless instructive because it shows a rare departure from the accepted reaction course. Depending on the group on the phosphorus centre, the reduction of the indole-derived azide proceeds to the well expected iminophosphorane outcome or to the less common triazine heterocycle, whose ring system features three contiguous nitrogen atoms. The mechanism is fascinating, especially if you are programmed to see nitrogen extrusion whenever phosphorus meets azide. I thought this reduction was largely predictable, but I was wrong. Let me know if you are aware of other interesting cases.
Last December, Dr. Steve Ritter of the Chemical and Engineering News asked me to comment on a paper from the lab of Prof. Petr Beier of the Czech Academy of Sciences. I gladly did (http://cen.acs.org/articles/94/web/2016/12/Fluorinated-azides-click-make-triazoles.html?type=paidArticleContent) and I just want to share my thoughts with you in the event you have not seen the Beier paper.
Every now and then we need a reminder of a rather straightforward kind: if we have trouble making a bond, just reverse the darn polarity of reagents! It is remarkable how infrequently this way of thinking pops into our heads, and I am judging from years of experience. Indeed, unless you are into radical reactions, there are always at least two ways to make a bond by a polar mechanism. In the example described by the Prague team in the Angewandte article, the curious CF3N3 molecule was the target. Attempts to forge the C-N bond using CF3I as the electrophile led nowhere, whereas using the fluoroalkyl portion as the nucleophile delivered CF3N3 and other uncommon azides without a glitch. I know this stuff ultimately relates to the well-known umpolung arguments, but those of us who are in the business of making bonds would still rather search “closer to the lamp post” than reverse reagent polarity. I am convinced that there are a lot of other previously “unmakable” molecules that might be made using this simple logic. We should keep this in mind.