A Multi-Modal Approach for the Investigation of Complex Protein Systems via Site-Directed Spin-Labeling

Abstract

The electron-nuclear spin dual-resonance method, ODNP (Overhauser dynamic nuclear polarization), characterizes the hydration layer of liquid-state samples at room temperature and can target membrane-embedded, membrane-interacting, and intrinsically disordered proteins in their native environment with ease. Past studies have correlated residue-by-residue measurements with MD simulations, however we seek a relatively a priori method for transforming experimental data into 3D “map” of hydration layer properties. While spin physics delineates the active volume of an ODNP measurement at a distance reaching 1-1.5 nm from the spin label, typical ODNP studies target dynamic liquid-state samples. In such samples, the covalently attached spin label itself typically moves over distances of up to 1.3 nm. To resolve the resulting spatial ambiguity, we turn to Rosetta; specifically exploiting its Monte Carlo algorithms to rapidly predict full ensembles of atomistic, minimal-energy structures approximating the solution state. These ensembles allow us to visualize the conformational space that the spin label in each position explores, structure by structure. ESR (electron spin resonance) and PRE (paramagnetic relaxation enhancement) NMR offer, respectively, information on spatial distributions and dynamics of spin label rotamers, and spin-label nitrogen-backbone distance information that can validate or disprove the ensembles that Rosetta predicts. The choice of a biochemical system composed of a protein in a lipid membrane with both solvent-exposed and membrane-occluded surface areas reinforces the thorough development of our methodology. We employ the photoactive membrane protein, proteorhodopsin, and begin to construct a map of the hydration layer spanning the lipid bilayer by spin-labeling along the full length of the membrane-spanning F-helix. We describe our progress towards a multi-modal study that employs Rosetta, PRE NMR, and ODNP to construct such a map.