Protein structure by NMR

Proteins are polypeptides, formed from the twenty naturally occurring amino acids. There are two types of local conformation of the polypeptide chain: a form in which the backbone is folded into a helix (helix ), and a form in which the backbone is folded to form a more or less flat (pleated) surface (sheet ).

Protein nuclear magnetic resonance (NMR) mainly studies the spatial folding of the polypeptide chain. In this, it differs essentially from NMR on organic molecules, which must determine the covalent bond graph. An important feature of protein NMR spectra is their extreme complexity, due to the large number of observable nuclei in the sample. The various steps involved in protein NMR structure determination are: sample preparation, acquisition and processing of the NMR signal produced by the observed nuclei, analysis and assignment of the spectra, calculation of atom coordinates from the parameters measured on the NMR spectra.

A protein can be studied using homonuclear or heteronuclear NMR, depending on whether the proton nuclei are observed alone or together with carbon and/or nitrogen nuclei. NMR experiments enable observation of scalar and dipolar coupling between spins. Observation of magnetization transfer, via dipolar coupling, produces nuclear Overhauser effects (nOe) between nuclei located less than 5 Å apart, and enables distances between observed nuclei to be estimated. But homonuclear NMR has certain limitations for the study of proteins in solution, due to the rate of peak superposition on spectra and poor magnetization transfer between protons by scalar coupling. These problems, which made it difficult to use NMR to study proteins larger than a hundred residues, were reduced by using the NMR properties of nuclei other than the proton. In fact, different nuclei resonate in different frequency ranges, making it possible to resolve chemical shift superpositions. In addition, heteronuclear coupling constants improve the sensitivity of scalar coupling....