Shaking Up Small Molecule Binding

I’ll be honest I thought I knew a fair bit about small molecules binding to proteins. If someone asked me what a phenyl ring in a molecule was doing I could talk earnestly about the entropic effect of displacing those water molecules, stacking interactions, Van der Waals forces, maybe even pi-charge interactions. I could have also talked about hydrogen-bonding and I would have certainly mentioned the hydrophobic effect (mind you that is more complicated than it looks sometimes) and how a molecule rotates (i.e. the less carbon chains and more rings the better).

One thing I certainly would not have mentioned was how individual bonds vibrate, but what do I know? 2 recent papers talking about deuterium effecting how compounds smell  and another using the IR spectra of nitrile groups to explain the observed binding affinity of a family of HIV drugs demonstrate how what I think I know and what I actually know are sometimes a disappointing distance apart.

I first heard about the deuterium-smell effect on the news, not my usual source of science information admittedly, but the paper in PLoS ONE is very interesting (and has also now been reviewed here). The group had speculated previously about the possibility that your olfactory receptors could discriminate different molecules based on how they vibrate and devised an experiment in order to test their hypothesis by synthesising dueterated variants of strong smelling musk compounds below (Dueterium, usefully, has twice the mass of hydrogen making for a powerful kinetic-isotope effect reducing the vibration frequency). If this sounds far-fetched at first; fish and insects have been shown to distinguish between deuterated compounds by smell; however that isn't to say the theory isn't somewhat controversial.

The test used trained perfumers as well as “normal” guys and girls, interestingly they subjects could not tell the difference between acetophenone and d-acetophenone, but were well able to distinguish between musks. The paper speculates that the musks (being on the large side for smelly molecules) potentially only activate a few receptors which are sensitive to a particular vibrational frequency, and that many groups resonate in that region i.e. CH₂ groups. All that from changing the vibrational frequency of a molecule through a kinetic isotope effect; more work to do here yet to nail down the hypothesis, but really interesting all the same.

Fine you may think, but how can bond vibration help you work out what is going on in an actual small molecule inhibitor? Well it turns out it can be important even here, according to a group reporting in Nature Chemistry. They examined an HIV1-protease inhibitor, rilpivirine, which contains two nitrile groups at opposite ends of the molecule.

In solution the nitrile groups produced identical IR spectra, although the exact resonance frequency altered depending on the solvent. When bound to the protein however the nitrile groups produced distinct IR spectra in such a manner that it suggested one group exists in a water phobic region while the other is interacting with the bulk solvent via a water channel.

What turns out to be even more interesting is that they use this information to figure out why this class of compounds is unusually resistant to changes within the binding site. That nitrile group interacting with the water channel turns out to anchor the entire molecule making it resistant to mutations (which they carry out and test) that expand the binding site weakening interactions else ware.  That potentially means you could use this knowledge to expand this class of molecules in order to make a series of inhibitors more resistant to mutation of the binding site.

So there you go; shaking a few bonds around turns out to be really important and can even tell you something useful.