Watching Proteins Move with Site-Directed Spin Labeling

Since the first crystal structure of a protein by Kendrew in 1957 until a few years ago, proteins were viewed as essentially rigid molecules that interacted with other macromolecules and ligands by Òlock and keyÓ principles. Research of the last decade, primarily from NMR spectroscopy, has radically changed this view. It is now becoming apparent that proteins are dynamic structures, and that dynamics is an essential aspect of protein function. In the confines of a crystal lattice, the dynamic modes are damped, and the static images of crystallography cannot provide the information required to deduce mechanisms of function.

NMR spectroscopy has identified two functional dynamic modes of small proteins in solution, namely backbone fluctuations that occur on the pico- to nanosecond range, and conformational switching, characterized by motions in the micro- to millisecond range. Despite this success, NMR methods have practical and fundamental limitations that have so far prevented their application to high molecular weight and membrane-bound proteins, and one could argue that these are the most interesting cases. Thus, new and general experimental strategies are needed to advance the field, and Site-Directed Spin Labeling EPR (SDSL/EPR) meets the requirements with flying colors.

In SDSL/EPR, methods of genetic engineering are used to replace one or two native amino acids with a nitroxide side chain, an example of which is shown in the Figure The Figure also shows the electron density map of the side chain (designated R1) obtained from X-ray diffraction data of spin-labeled T4 Lysozyme crystals. The EPR spectrum of R1 at a particular site in a protein encodes information on the motion of the nitroxide ring, and this in turn reflects the entire set of dynamic modes of the protein, including: (1) rotational diffusion of the protein; (2) torsional oscillations about the bonds in R1 (side chain internal motions); (3) local backbone fluctuations, and (4) conformational switching.

Using a combination of EPR lineshape analysis, site-directed mutagenesis and X-ray crystallography of spin labeled proteins, significant progress has been made in separating contributions from the four dynamic modes. In particular, models have been developed that permit mapping of £\-helical backbone fluctuations throughout a protein structure, and to monitor conformational switching modes in real time. The models have been verified on simple proteins that have been previously investigated by NMR and other techniques, and then used to explore more interesting and complex systems. For example, application of SDSL/EPR led to the discovery of the conformational switch underlying activation of Rhodopsin, the photoreceptor protein of the vertebrate retina. Highly flexible sequences in Rhodopsin have been found that are involved in the interaction with Transducin, another element of the visual signal transduction system. In addition, SDSL/EPR has revealed putative conformational switches underlying the activation of Transducin. These recent studies underscore the ability of SDSL/EPR to map backbone dynamics, conformational switching and to identify protein-protein interactions in systems of an arbitrary degree of complexity. The future of EPR in proteomics research looks very bright indeed.