Study of Membrane Proteins by Double-Quantum Coherence ESR: Gramicidin A

Many spectroscopic methods in Structural Biology are based on measurement of distances on the scale of interest, providing constraints used to resolve structures and thus shed light on the mechanisms underlying the related functions. In protein research, these constraints are used to reduce the number of degrees of freedom available to the polypeptide chain, to orient and dock proteins that participate in supramolecular assemblies, and to characterize the time evolution of these structures during their function. The approach to structural studies, based on site directed mutagenesis, spin labeling, and modern pulsed ESR techniques, e.g. those based on detection of double-quantum coherence (DQC), is particularly well-suited in the context of structural genomics. DQC yields distances over a range of at least 15 to 60 Å with high throughput, and it requires only nanomole amounts of protein that are difficult to produce in amounts needed for other structural methods. It is well-suited for resolving structures of “difficult systems” such as membrane proteins, large protein complexes, and RNA using a small number of long-distance constraints. By “difficult systems” we mean proteins (or other biomolecules) that are difficult to crystallize for structure determination by X-ray crystallography. In many instances determining one or several large distance is all that is needed to resolve structural and functional issues. This point, illustrated by the example of Gramicidin A (GA) in multi-bilayer vesicles of the phospholipids DPPC and DMPC. It shows just one of many benefits brought by DQC ESR into structural work on membrane proteins and their complexes. GA, spin labeled at its C-terminus with a nitroxide label, was studied by DQC ESR, which provided strong evidence that in DMPC membrane GA mainly takes on the form of a dimer with a well-defined distance of 30.9±0.5 Å between the electron spins at each end. This interspin distance is in good agreement with estimates for the head-to-head dimer of spin-labeled GA (the channel-forming conformation), which matches the hydrophobic thickness of the DMPC bilayer quite well. It is worth emphasizing that knowledge of just a single large distance was adequate for the task. By comparison, no DQC signal that could be associated with dimers was found in DPPC membranes, supporting the point that hydrophobic mismatch significantly reduce dimer formation. Nevertheless, pulsed ESR data indicate that there is an aggregation of GA that is yet to be investigated.