Crystallography (and increasingly, electron microscopy) can provide beautiful and detailed maps of proteins as well as other important macromolecules. What I’ve left out in my earlier discussion is the fact that these pictures only offer snapshots of the protein at various moments in time. Think Empire Strikes Back, where Han Solo is frozen in Carbonite, and remains stuck in a state of suspended animation for all time (or at least until Return of the Jedi). This is sort of like a crystal structure of a molecule.
Of course, all of the interesting things about a protein can be traced back to not only its structure but also its dynamics. A protein’s motion dictates how it functions as an enzyme, how it modulates structural changes, how it interacts with partners, and so on. There are certainly crystallographic methods aiming to capture the dynamic behavior of a protein (e.g. time-resolved studies, XFEL). Nevertheless, these techniques require a hardiness of crystals that isn’t always possible. There is a growing list of emerging biophysical techniques which can examine protein conformational changes. In fact, this meeting has had fascinating presentations about all of them. One technique that is regularly used in our group is Electron Paramagnetic Resonance Spectroscopy, which makes use of free radicals to probe protein motions. Since most proteins lack intrinsic free radicals, we introduce them by way of a covalent probe, known as a spin label. The free radical spin label then absorbs magnetic radiation depending upon its environment. This in turn can provide information about stable conformational changes made by the protein, depending on the placement of the label. Examples of this include changes made by transport proteins to facilitate substrate delivery (which I presented about) and domain motions within proteins during enzymatic activity.
While traditional EPR technique uses only one spin label to study protein motion, a few EPR techniques of burgeoning importance utilize multiple (usually two) spin labels to make distance measurements. This is based on the degree to which the free radicals interact. The information provided by these techniques is a distribution of possible distances of the labels from each other, from which we can infer how the protein behaves in different conditions.
Sounds great, right? An inherent error in all EPR measurements is the degree of rotational freedom these spin labels have (think of a tether on a cone). All small changes in distances thus come with this caveat: any change in signal you observe can be attributed simply to the spin label side chain motions. That limits the resolution of this technique, but still makes it a sensitive probe for conformational changes of 15-20 Angstroms for continuous wave techniques, and 20-80 Angstroms for pulsed techniques (like DEER, or double electron-electron resonance). Those measurements are extremely useful: in addition to providing information about protein dynamics, they are an excellent complement to crystallographic data, where proteins are often rigid and can adopt unnatural conformations. This allows for a more complete picture about how a protein actually functions.
Well, not so fast. Florida State’s Peter Fajer delivered a remarkable presentation to close Tuesday night’s workshop (Distance Measurements by D.E.E.R.) titled, “Do Spin Labels Lie?”. Using simulations, he demonstrated that the most commonly used spin label, MTSSL, experiences rotations of around 20 angstroms while attached to proteins. This doesn’t even include the protein’s backbone motions. The observation is pretty incredible, and calls into question years of measurements made by researchers in the field on a variety of systems. I’m certainly going to follow up on this work and reassess my data and interpretation. I’m amazed that this research was not done sooner, and kudos to Fajer and his group for looking into this phenomenon. His recommendation was to use shorter labels or bifunctional labels, which bind to two nearby cysteines and experience limited rotations. Fascinating. I now truly appreciate what “disruptive” means – this research is the embodiment of the word!