Probabilistic Approach to NMR-Based Determination of Accurate Local Conformation and Three-Dimensional Structure of Proteins in Solution

Abstract

We have recently shown that a probabilistic approach to NMR-based protein structure determination can yield higher quality structures than the more commonly used deterministic approaches. In addition to NMR data, the probabilistic approach uses a priori information on empirical distribution functions for backbone conformations, generated from the high resolution X-ray structures in the Protein Data Bank. This approach leads to greater accuracy in the determination of local conformations, and hence to greater precision and accuracy in the spatial structure built on the basis of these local conformations. In this review, we describe the application of this approach to two specific types of structural problems: (i) comparison of protein structures at the level of local conformations, which is particularly useful for large proteins for which NMR assignments exist, but for which spatial structures in solution have not been determined; cytochrome c is used as an example; and (ii) determination of three dimensional structures of proteins in solution, using accurate local conformations based on NOE data, and a build-up strategy that works well even with sparse NOE data sets; the C-terminal tryptic fragment of human plasma fibronectin is used as an example.