Distance Determination in Multimeric Membrane Protein by Symmetry Constrained Analysis of Double Electron-Electron Resonance Spectroscopy

Abstract

Distance determinations are often a core strategy to decipher the molecular mechanism by which membrane proteins execute their biological function. EPR spectroscopy and pulsed Double Electron-Electron Resonance (DEER) methods are unparalleled tools for the measurement of probe-based long-range distances in protein. However, distance determination from echo intensity modulation in pulsed DEER experiments is a moderately ill-posed problem. Here, we show that by using protein symmetry-based geometrical constraints in homo-oligomeric membrane proteins, we are able to greatly facilitate the fitting solution. We modeled the distance distribution with 2 Gaussians, with each mean distance identified by its Gaussian peak. During the fit process, the mean distance ratio was constrained to be within a specified tolerance of the theoretical distance ratio determined by the symmetry of the oligomeric assembly (1:1.4 for tetramers; 1:1.6 for pentamers). We compared this approach against the classical Tikhonov regularization, concluding that our method stabilizes a solution that is much easier to interpret in molecular modeling terms. Importantly, we also show that when the qualities of the dipolar evolution signal increases (more observed periods and higher SNR), there is little or no difference in the distance distribution obtained by our method vs. Tikhonov regularization. This suggests that the geometrically constrained fit does not artificially distort the distance distribution. Our approach was validated on 2 different ion channels of different oligomeric states: CorA, a homopentameric Mg2+ channel and KcsA, a homotetrameric K+ channel. In all cases, the distances obtained by DEER are in excellent agreement with respective crystal structures.