Magic Angle Spinning NMR Structure Determination of Proteins from Pseudocontact Shifts

Abstract

Magic angle spinning solid-state NMR is a unique technique to study atomic-resolution structure of biomacromolecules which resist crystallization or are too large to study by solution NMR techniques. However, difficulties in obtaining sufficient number of long-range distance restraints using dipolar coupling based spectra hamper the process of structure determination of proteins in solid-state NMR. In this study it is shown that high-resolution structure of proteins in solid phase can be determined without the use of traditional dipolar−dipolar coupling based distance restraints by combining the measurements of pseudocontact shifts (PCSs) with Rosetta calculations. The PCSs were generated by chelating exogenous paramagnetic metal ions to a tag 4-mercaptomethyl-dipicolinic acid, which is covalently attached to different residue sites in a 56-residue immunoglobulin-binding domain of protein G (GB1). The long-range structural restraints with metal-nucleus distance of up to ∼20 Å are quantitatively extracted from experimentally observed PCSs, and these are in good agreement with the distances backcalculated using an X-ray structure model. Moreover, we demonstrate that using several paramagnetic ions with varied paramagnetic susceptibilities as well as the introduction of paramagnetic labels at different sites can dramatically increase the number of long-range restraints and cover different regions of the protein. The structure generated from solid-state NMR PCSs restraints combined with Rosetta calculations has 0.7 Å root-mean-square deviation relative to X-ray structure.

Structure determination of uniformly 13C, 15N labeled protein using qualitative distance restraints from MAS solid-state 13C-NMR observed paramagnetic relaxation enhancement

“…The unlabeled GB1-wt and U-C, N single cysteine mutants were purified using the same protocol, except the buffers used for the purification of the single cysteine mutants contained 5 mM DTT. The cell suspension was incubated in water for 10 min at 80 °C and the supernatant was loaded onto a HiTrap DEAE FF column (Li et al 2013 ). The purities of the samples were checked by SDS-PAGE.…”

NMR of sedimented, fibrillized, silica-entrapped and microcrystalline (metallo)proteins

“…The degree of the optimal paramagnetic dilution depends on how many metals contribute to the observed PCS, i.e. on the anisotropy of the paramagnetic susceptibility and on the position of the metal ion within the molecule: in the case of cobalt MMP-12 [131] and of lanthanide-tagged GB1 [132], 30% and 15% of paramagnetic proteins were found to be the optimal compromise. An alternative approach is using fully paramagnetic labeled samples which ensure both the largest signal amplitude and resolution in the NMR spectra, but the analysis of the PCSs so available Fig.…”

“…To date, there is only a single report of a highly resolved spectrum of a lanthanide-tagged protein [132]. The difficulty in the observation of well resolved SSNMR spectra of proteins containing paramagnetic lanthanides might be due to several reasons, among which incomplete binding, and small fluctuation in the first coordination sphere of the metal.…”

Recent Advances in Magic‐Angle Spinning Solid‐State NMR of Proteins

“…[198] While the initial investigations of PCSs relied on the presence of endogenous metal-binding sites in metalloproteins, in a recent study, Yang and co-workers demonstrated that structurally useful PCS could also be obtained from exogenous paramagnetic ions, which were chelated to a 4-mercaptomethyl-dipicolinic acid tag, and covalently attached to cysteines introduced by site-directed mutagenesis at different sites in GB1. [199] By using different metal ions and labeling sites, a sufficient number of PCS restraints could be obtained to determine the structure of GB1.…”

“…[202] Due to their long-range nature, PRE and PCS measurements often yield information on both intra-and intermolecular geometries. [196,203,199] In some cases, the intra-and intermolecular effects may be observable in a single sample, [196,203,199] and the two contributions can be differentiated by studying diluted samples. [196,203] In cases when dilution is not possible, an approximate knowledge of the intermolecular interface is required to choose an optimal mutation site for the incorporation of a paramagnetic tag.…”