Generalized Model-Free Spectral Density Analysis Applied to Rhodopsin Activation in Membranes

Abstract

Although molecular structures of G-protein-coupled receptors (GPCRs) are becoming increasing available from X-ray crystallography, understanding their functions requires information about molecular dynamics in membranes. Here we use rhodopsin as a model to illuminate general features of GPCR activation. With solid-state 2H NMR spectroscopy we obtain experimental data pertinent to both structure and dynamics. Experimentally, order parameters and relaxation rates are the two observables of solid-state 2H NMR experiments. We propose that the local dynamics of the retinylidene ligand are coupled to large-scale fluctuations of the transmembrane helices of rhodopsin, leading to activation of the receptor. To study the structural dynamics of retinal bound to rhodopsin, we start with an irreducible representation of the correlation function in terms of mean-squared amplitudes and correlation times [1]. The mean-squared amplitudes are related to the orientational order parameter, while the irreducible correlation times include the preexponential factor and energy barrier. To bridge the generalized model-free theory with experimental measurements, we separated the relaxation rates into spectral densities by applying Redfield theory. The spectral densities are Fourier transformation partners of the irreducible correlation functions. By fitting theoretical spectral densities to experimental data we can readily obtain the values of preexponential factors and activation energies [2]. We are currently applying our generalized model-free method to interpret the behavior of active Meta-II rhodopsin. Our aim is to establish if the local fluctuations of the ligand initiate the structural changes of rhodopsin to understand the activation mechanisms of GPCRs in general. Moreover, the results from our generalized model-free analysis method can be used in molecular dynamics (MD) simulations without the limitations of simplified motional models. [1] M.F. Brown (1982) JCP 77, 1576-1599. [2] A.V. Struts et al. (2011) NSMB 18, 392-394.