Solid-state magic-angle spinning NMR methods for tensor measurements and protein structure refinement using chemical shift tensors

Solid-state NMR (SSNMR) possesses the unique ability to measure structurally dependent anisotropic properties that are difficult to measure directly using solution NMR techniques, including chemical shift and dipolar tensors. Chemical shifts are the most fundamental parameters measured in NMR spectroscopy and have been leveraged for macromolecular structure determination, refinement and validation including details of active site chemistry in enzymes. Here we present multidimensional SSNMR experiments to measure tensors quantities in the streptococcal beta-1 immunoglobulin binding domain of protein G (GB1). Experimental techniques presented include: three-dimensional experiments that recouple 13C or 15N chemical shift anisotropy (CSA) pseudostatic powder lineshapes, slow magic angle spinning (MAS) analysis of highly-13C,15N-enriched solid proteins using 2D heteronuclear correlation at 750 MHz 1H frequencies, rotational resonance (R2) lineshape analysis, and distance information using the REDOR and TEDOR heteronuclear dipolar recoupling pulse sequences. Data from these experiments reveals important structural information, reporting directly upon secondary structure, hydrogen bonding, and electrostatics. This work culminates in the first structure of a solid protein solved and refined using Calpha CST data to constrain backbone conformation. This is achieved by comparing the experimental CST elements to ab initio chemical shielding calculations as a function of local conformational degrees of freedom. To this end, a customized CST force field was generated and used as a restraint class in the XPLOR-NIH simulated annealing algorithm. The addition of this restraint class substantially improved the precision (∼0.3 A backbone RMSD) and accuracy (∼1.1 A related to the crystal structure) of structure determination for GB1. These results demonstrate that de novo structure calculations utilizing Calpha CST data can yield atomic-resolution structures of solid proteins.