Hydrazine bonanza

I became aware of André Beauchemin’s really interesting Angewandte paper that appeared the other day. André’s lab at the University of Ottawa has been pushing the frontiers of synthesis with some imaginative use of heteroatom-heteroatom bonds as platforms for atom transfer chemistry. The Angewandte paper I refer to deals with novel amphoteric reagents, something that has been of particular interest to my lab for a number of years. Take a look at the hydrazine isocyanate André has developed:



The isocyanate molecule shown above is a kinetically amphoteric reagent with the electrophilic carbonyl group of the isocyanate carbon placed several Angstroms away from its nucleophilic hydrazine nitrogen. Due to this constellation of reactive loci in the dense area of space, it is possible to add an amine nucleophile to the carbonyl carbon and quickly follow it up by ring formation “around the corner”. This is accomplished using Andre’s previously reported Cope-type chemistry. The resulting molecules are novel and interesting from the diversity/medicinal chemistry points of view. For me though, it is the amphoteric nature of the centerpiece that is of particular note.

Incidentally, back in 2006, when Ryan Hili and I were thinking about the best way to call our aziridine aldehyde dimers, we really liked the term “amphoteric” due to the origins of the word (“both of two” in Greek). We felt that naming the molecule “bifunctional” would not be descriptive enough. I am not a fan of the term “bifunctional” to begin with… What does it mean? I even get confused with “biweekly”: you think it is unambiguous, but some people use the word to refer to an event taking place once in two weeks, while others – two times a week. I think this is confusing…

2013 Nobel Prize in Chemistry

This one has been long overdue… As you all likely know, the 2013 Nobel Prize has been awarded to Karplus, Levitt, and Warshel for their ground-breaking theoretical work that has enabled everything from modelling the way enzymes carry out chemical transformations to investigating receptor/ligand interactions in drug design (http://www.nobelprize.org/nobel_prizes/chemistry/). Professor Ariel Warshel, whom I know personally (more on that soon), is well known for using a combination of quantum and classical mechanics to tackle the seemingly insurmountable challenges of complex protein simulation. Interestingly, and despite the nature of this work, one of Ariel’s points is that a great question is the most important thing in his branch of science. The computing power is secondary. He has maintained over the years that it is erroneous to assume that the amazing things possible now are there due to the emergence of really strong computers. In Ariel’s view, he was able to pose and answer profound questions 35 years ago as well as he is able to do so today. It is all about which questions you ask, I suppose.

In our lab, we feel the significance of the work credited with this year’s Nobel particularly strongly these days as we delve more and more into building a bridge between protein/small ligand crystallography and the design of chemical reactions that allow us to build better protein probes. Dr. Conor Scully, a research associate in our lab, has been the main driving force behind this project and the announcement of this year’s Nobel Prize really means a lot to all of us who use computers in efforts to understand chemistry.

I do have a bit of a personal connection with one of the winners (Ariel Warshel). When I came to the US in 1992, I had no training in synthesis. During my undergraduate years in Moscow I worked on developing an algorithm for enumerating reaction mechanisms with the help of graph theory (very obscure). So… I had a fellowship (which I declined) to go to Oxford and work with Prof. Graham Richards. I also had an offer from USC. Los Angeles sounded really exciting to me despite the 14 hour plane trip and the infamous Rodney King riots that burnt half of downtown in 1992 (my father said: “Where the hell are you going?”). But there I came. With my background, I started investigating the research advisors and Prof. Warshel was one of the people I spoke with at length. He is a real gentleman, originally from Israel, who was really nice to me, and it really meant a lot. However, synthetic and physical organic chemistry ultimately attracted me more and I joined the labs of Prakash and Olah. When I was watching the USC press conference dedicated to Warshel earlier today, Olah (the 1994 Nobel laureate) made a nice comment at the end, congratulating Ariel. He also said that they both arrived to USC at about the same time (end of 1970’s) and saw a significant transformation of the university. There was a cool moment during his commentary: Olah mentioned that football is great but it is research that should drive a university. This is one of his pet peeves and I fully agree with his stand. I am not sure you all know, but a football coach is typically the highest paid position at a major University in the US.

With this, I once again salute this year’s real heroes of science!

Electrophiles that seemingly defy logic

Here is one of my favourite subjects: electrophiles. They are often used as reactive intermediates in synthesis, which is common knowledge.  The vast majority of electrophiles are toxic to humans for a very clear reason: our bodies are largely composed of nucleophiles. Indeed, it is difficult to name an amino acid residue in our proteins that is electrophilic in nature. Apart from the ones that are neutral (leucine, alanine and the like), amino acids are nucleophilic (cysteine, arginine, lysine, serine, etc). Therefore, one should expect irreversible reactivity between exogenous electrophilic species and proteins. In fact, there are many drugs that act by way of covalent modification of the nucleophiles found at protein active sites. We are also told (rather emphatically, in all sorts of MSDS sheets we get from Aldrich!) to avoid contact with electrophilic chemicals. Alkyl halides are on top of some of those lists and many of these chemicals are banned by the Montreal Convention. It may come as a surprise, therefore, that some odd balls pop up here and there and, on the surface, there is no explanation for their relatively benign nature. Take Splenda as an example. This artificial sweetener contains sucralose, whose structure is shown below. This stuff is served at Starbucks, not to mention a ton of other establishments. One look at this molecule was enough to make me cringe when I first saw it. I instinctively considered all of my cysteines as potential targets! However, sucralose is benign and does not covalently inactivate cysteines. In fact, it is an alosteric modulator of T1R receptor, which explains its effect as a sweet taste-enhancing additive. There was a great PNAS paper on that some years ago (see the link below).

There is another lesson in this story and it relates to the foundation of Sn2 chemistry, the rate of which is rather sensitive to steric bulk around the reactive site. This is why we teach our students that neopentyl systems are lousy electrophiles. So, there you have it, folks: Splenda is quite ok for you despite the chloromethylene and chloromethyl groups in it. Next time you put it in your coffee, rest assured that your cysteines will remain intact due to relative rates and the good old physical organic chemistry.



Boron in aromatic molecules

Here’s my Monday, October 7, 2013’s two cool facts about boron heterocycles.

First of all, boron is tricky and it is not always obvious which factors control stability. Take, for instance, a recent theoretical JACS paper by Professor Roald Hoffmann of Cornell University. The manuscript contains a lot of very interesting data on boron heterocycles and is, in many ways, a call to arms for those who are interested in this subject. However, how would you predict the whopping 24 kcal/mol difference in stability (calculated relative energies) between the following two molecules? I realize they are not the same, but 24 kcal/mol? This means that NBN motifs are special…


A lot of really imaginative work in the area of aromatic boron heterocycles has been done by Prof. Liu of Boston College (formerly of the University of Oregon). I am not going to say anything about his nice synthesis of aromatic boron-containing heterocycles. Instead, I will focus on properties and showcase a co-crystal structure between azaborine and T4 lysozyme. The hydrophobic cavity of this enzyme accepted azaborine, a view of which I made using PyMol (pdb code 3hh3, below). It can be clearly seen that the azaborine molecule binds in two conformations which relate to each other by way of a “flip”. Given the amount of aryl groups in biologically active molecules, these boron heterocycles are certainly interesting! How about using them as fragments? Someone should…china



The economy of means in targeting complex interfaces

Who says that molecules have to be complicated in order to effectively interfere with protein-protein interactions in cells? Here is a fascinating recent example of a small molecule developed in the lab of Professor Ahn at the UT Southwestern, Dallas, Texas. This paper, published in Nature Communications by Ahn and his collaborator Ganesh Raj, is a lesson to all of us who think that mimicry of complex epitopes using small molecules is a lofty goal.  Well, it certainly is a lofty one, but it is also reachable! This and other examples (Hamilton’s great work on polyaromatic molecules comes to mind) exemplify the state of the art in reductionist systems feasible with smart design. Look at the helical pitch (below) and its mimicry by a trivial amide compound. I am willing to bet there is a lot of completely unexplored molecules that are also small, perhaps a bit more architecturally advanced than the one shown, yet capable of interrogating non-helical epitopes, including disordered ones. In Ahn’s example, the IC50 of 40nm was achieved in efforts to disrupt specific protein-protein interactions involving LXXLL motifs. The use of such simple molecules in targeting androgen receptor-coreceptor interactions has been demonstrated to have clinical value. Kudos to Ahn and co-workers who designed these wonderful molecules!

The most important lesson I learned from my mentor, Professor Barry Sharpless of Scripps, is that complex problems do not require complex solutions (as opposed to what many people in synthesis preach). The present case certainly underscores this notion.



Macrocycles vs linear molecules: not a clash (yet)

As I get further and further into fragment-based screening using protein crystallography/small molecules, I have been developing ideas about a rational way of constructing linear oligomers of heterocycles. There will be way more on that in my future posts. But first, some people might ask: “What? You? Thinking about the virtues of linear compounds with your long-standing interest in macrocycles? What is going on?” Well, hear me out. I am not suggesting that we abandon some of our favourite peptide-based rings. No way, we are fully committed to the cyclic peptide deal and we will stand by them.

Having said that (as Larry David of “Curb Your Enthusiasm” would say – check it out: http://www.youtube.com/watch?v=ENHVQ2gslp8), there are some things to consider in terms of how we design our molecules to interact with protein targets. Here is something I mentioned at my lab’s meeting 2 days ago: we simply do not think about enthalpy hard enough. This is happening for a good reason: this is a tougher nut to crack. The history of drug discovery tells us that the first, and often most obvious, thing to do in an effort to make a molecule that selectively interacts with a protein target is to consider entropy. Macrocycles, of course, provide a gateway to address this problem. Researchers would typically consider a peptide ligand, identify their favorite beta turn or helix (think about stapled peptides), and then go on to make a cyclized version that hopefully recapitulates the bent epitope. But take a look at how drugs in a particular class evolve over the years. I suggest the following paper:


It is curious, isn’t it? Figure 1 shows (and rather emphatically) that the entropic contribution (which forms the basis for our instinctive thoughts about making macrocycles) dominates at the beginning, when first molecules in a given class are being discovered and developed. After some time, the sophistication increases and you can see how enthalpy contribution “grows”. The underlying reason is that enthalpy is more difficult to address. The reasons are manifold with solvation/desolvation sitting at the center stage. To sum up, I am not thinking about abandoning macrocycles. I am just under the influence of some of the fragment-based crystallographic work we are involved in and I think that we might often be better off improving the enthalpic contribution in our molecules when we consider linear shapes. I am falling in love with those.

Amazed by microscopic reversibility

Here’s a synthetic post, dedicated to the talented folks in total synthesis who have a lot of really cool tricks up their sleeves. I am going to talk about just one (or two) steps in Tohru Fukuyama’s fairly recent lyconadine synthesis. The sequence goes through one of my favorite processes – a microscopic reverse of an electrocyclization… I mentioned a similar reaction in the past when I referred to our own work in electrocyclic ring-opening of bromoaziridines. Take a look at a clever use of electrocyclization (its microscopic reverse, that is) in efforts to create a complex 7-memebred ring. First of all, please note the dibromocyclopropane preparation. Wait – before we go any further: do you know whose process this is? If you are thinking of Prof. Makosza from Poland, you are correct as he is the man! I already commented on his vicarious aromatic substitution mechanism. The phase transfer-catalyzed dibromocyclopropanation hails from his lab as well. The yield here is not great, presumably since we are dealing with a fairly challenging substrate… Then comes the key step, which is carried out in pyridine at reflux. A ton of fun, no doubt, but the result is impressive – the ring system is set up and the benzyl group is gone… There is an elimination pathway that competes, but these are minor details. It is still an elegant sequence. I think one nice lesson here is to always remember the principle of microscopic reversibility, which is not simple when thinking about retrosynthesis, in my view!