To understand why we have to go back to our premise: How proteins work is closely related to how they are shaped. This is no doubt true, but we are leaving something out. Protein function is also strongly dependent on how they move. Thats right, proteins don't just sit there looking pretty (as it appears in our snapshots). They also 'fold up' and 'unfold'. Even when they are just hanging around in a nice comfortable environment their shape is constantly fluctuating with everything from small very fast movements to big, slow movements. These motions are not random. They are, in fact, function critical (that is to say, if a protein can't move properly, it can't work, at least not very well) and they have defined start and end points. Lets say that the most stable shape (the one we see in our snapshot) is the start point. The end point for a particular motion (or set of motions) will be a different shape that is less stable than the starting shape and so it won't hang around long; it'll quickly go back to the starting shape. This means that at any one time, in a sample of a gazillion proteins, only a small fraction will be in our end point shape. We call this a 'weakly populated structure' and, as you might imagine, these structures are very hard to detect. And here's the kicker: There is mounting evidence that in many cases it in fact weakly populated structures and not the lowest energy 'snapshot' structure that is the source of biological activity! It should come as little surprise, then, that our structure based drug design efforts have yielded lackluster results; we've been targeting the wrong shapes!

So we're in a bit of a pickle. It appears that if we're to understand how proteins work well enough to custom design drugs, we need information about shapes that are weakly populated and/or short-lived. Here's where our super-cool Mass Spectrometry and Nuclear Magnetic Resonance techniques come in. Mass Spectrometry, as the name suggest, is a way of detecting molecules according to their mass. We use Electrospray Mass Spectrometry which has the additional benefit that we can distinguish - in a very rough way - different protein shapes by looking at how charged they become in the electrospray process. We can (and do) combine this with H/D exchange to get a more detailed idea of the shape. Mass Spectrometry (MS) makes a really good tool for our work for two fundamental reasons: 1) It is highly sensitive, so we can detect even weakly populated structures. The world record for MS sensitivity of 'big' molecules is 14 zeptomoles (10^-21 moles) [1]. 2) It is highly selective, so we can distinguish analytes that differ from each other only slightly. The world record for MS selectivity is a difference of .00045 Da, which is smaller than the mass of an electron! [2]. We use a technique that I introduced in my PhD with Dr. Lars Konermann that facilitates Time-resolved Electrospray MS. This allows us to use our highly sensitive and selective Mass Spectrometer to study rapid processes (like protein motions, which occur mostly on the millisecond timescale)

For some time now, we've been able to take highly detailed, 3 dimensional 'snapshots' of proteins. There are a number of ways of getting these 3-D pictures, but by far the most common is to analyze the shadows produced by protein crystals (X-ray crystallography). One thing that quickly became 'crystal clear' (pun intended) from our snapshots is that how proteins work is very closely related to how they are shaped. Ahah, we said! Now if we want to 'custom design' small molecule drugs targeted at specific proteins, all we have to do is find a small molecule that fits nicely into the protein and has the right chemistry to stick there! This approach is called 'structure based drug design', and I think you'll agree it's a pretty cool idea. But it's far from perfect. In fact, if we're honest, our success in structure based drug design has been mixed at best.

Nuclear Magnetic Resonance (NMR) is, among other things, a way of studying proteins at atomic resolution. NMR can be used to get protein 'snapshots' just like X-ray crystallography, but there are also all kinds of experiments that give information about what we're interested in - motion. These include the NMR version of H/D exchange, experiments with funny acronyms like CLEANEX, unfortunate acronyms like STD and unimformative acronyms like CPMG relaxation dispersion. That last one, incidentally (CPMG relaxation dispersion) was pioneered by Dr. Lewis Kay's group at the University of Toronto. NMR is not as sensitive or selective as MS, but it can provide us with the highest detail information about how proteins move. Thats why we use 'em both!