Protein structure is critically linked to an understanding of protein function. Currently there are three approaches to determining 3-D protein structure: nuclear magnetic resonance (NMR), x-ray diffraction, and computational prediction of structure. NMR examines proteins in solution; x-ray diffraction measures scattering from a solid protein crystal; and computational approaches predict structure by finding similarities in amino acid sequence and substructures between the unknown and known protein structures. At this time, about 80% of the known protein structures were determined using x-ray diffraction.
7.1 Nuclear Magnetic Resonance
Nuclear magnetic resonance measures the coupling of atoms across chemical bonds and short distances under the influence of a magnetic field. An NMR instrument is shown in Fig. 7.1. A typical NMR experiment with a 600-MHz field takes 4 to 5 weeks for data collection and is limited to proteins with fewer than 360 amino acids (40 kDa). Data analysis that once took months can now occur in a single day using new algorithms. New NMR instruments with fields as high as 1 GHz and more sensitive techniques allow faster analysis of larger proteins, but there is still a size limit of a few hundred amino acids.
The principles behind NMR are similar to those used in the medical procedure of magnetic resonance imaging (MRI). In fact, MRI is really a specific application of the more general NMR principles. MRI usually concentrates on hydrogen atoms. Not only does hydrogen have the strongest magnetic response, it is the most common atom in biological systems. NMR spectroscopy includes hydrogen, carbon, nitrogen, and other elements (see Table 7.1). Only certain isotopes of these elements can be detected in NMR systems. Fortunately, these are some of the most common atoms in proteins. If a specific isotope is needed or an increased abundance of one isotope rather than another, organisms can be fed a restricted diet that includes the isotope. For NMR, these isotopes do not need to be radioactive.