Measurement of Structural Restraints

Abstract

Structure determination by high-resolution NMR spectroscopy in solution has traditionally relied on the use of Nuclear Overhauser Enhancement (NOE) derived distance restraints, complemented by dihedral restraints obtained from the analysis of J-couplings [1]. This approach has proven to be very successful in solving the three-dimensional structures of proteins up to the 20-30 kDa range [2]. More recently, additional sources of structural information have become available. These include chemical shifts [3,4], reliable information about hydrogen bonding [5], residual dipolar couplings (RDCs) [6,7] and small-angle X-ray (SAXS) [8,9] or neutron scattering (SANS) data [10].The different types of data are highly complementary and their combined use has led to strongly improved NMR structures. Since the information from NOE, J-couplings and chemical shifts is inherently short-range in nature, cumulative errors from such restraints can translate into inaccurate long-range structural behaviour [11]. This is particularly the case for non-globular biomolecules, such as larger DNA-molecules or protein (complexes) made up of loosely associated modules. In contrast, RDCs and SAXS/SANS data provide long-range information by reporting on the orientation of local groups relative to an overall molecular frame and constraining the overall shape of the molecule. The combination of both types of data has opened avenues for the accurate study of a much wider range of systems, in particular large complexes, dynamic domain-domain interactions and partially or completely unfolded proteins.