Application of spherical harmonics for DEER data analysis in systems with a conformational distribution

Double electron-electron resonance (DEER) and other pulse electron paramagnetic resonance (EPR) techniques are valuable tools for determining distances between paramagnetic centres. DEER theory is well developed for a scenario where relative orientations of paramagnetic centres do not affect the DEER data. In particular, such theory enables a number of approaches for extracting distance distributions. However, in a more general case, when orientation selection effects become substantial, the analytical theory of DEER is less well developed, therefore quite commonly researchers rely on a comparison of some model-based simulations with experimental data. This work elaborates the theory of DEER with orientation selection effects, focusing on a scenario of a moderate conformational disorder, leading to an orientation distribution in a pair of paramagnetic centres. The analytical treatment based on expansions into spherical harmonics, provides important insights into the structure of DEER data. As follows from this treatment, DEER spectra with orientation selection can be represented as a linear combination of modified Pake pattern (MPP) components. The conformational disorder has a filtering effect on the weights of MPP components, specifically by significantly suppressing MPP components of higher degrees. The developed theory provides a pathway for model-based simulations of DEER data where orientation distribution is defined by analytical functions with parameters. The theory based on spherical harmonics expansions was also applied to develop an iterative processing algorithm based on Tikhonov regularization, which disentangles the distance and orientation information in a model-free manner. As an input, this procedure takes several DEER datasets measured at various positions of an EPR line, and outputs a distance distribution and orientation distribution information encoded in a set of coefficients related to the weights of MPP components. The model-based and model-free approaches based on the developed theory were validated for a nitroxide biradical and a spin-labelled protein. (C) 2020 Elsevier Inc. All rights reserved.