We propose to use nuclear magnetic resonance (NMR) spectroscopy together with theoretical chemistry, to obtain information about the electrostatic structures of proteins. Specifically, we wish to solve a twenty year old problem - that of the origin of the chemical shift nonequivalencies generated in proteins due to folding into their native conformations. We will use computed and experimental values of the dipole and quadrupole shielding polarizabilities and hyperpolarizabilities together with calculated (integrated field energy, finite difference Poisson-Boltzmann, protein-dipole Langevin-dipole, and distributed multipole representation) electric field strengths and electric field gradients at individual protein sites to help explain experimentally determined 1H, 13C, 15N, 17O and 19F chemical shifts. Additional work will also be carried out to more fully understand related infra-red vibrational frequency data on CO-myoglobin (conformational substates). The shielding polarizabilities will be computed using analytical, derivative Hartree-Fock methods, with large basis sets. The effects of gauge origin and basis set choice (from double zeta to triple zeta, triply polarized with diffuse valence and f functions) will be investigated, using amino-acid or amino-acid models (eg 2-methyl imidazole). Realistic amino-acid potential surfaces will be incorporated into the integrated field energy method, to improve upon the general point charge description currently used in integrated field energy method, to improve upon the general point charge description currently used in electrostatic modelling. Experiments and calculations on selected amino acids and peptides will be performed in order to substantiate and help develop some aspects of the calculations. Particular emphasis will be placed on investigating the effects of charge on shielding tensors (in e.g.Li+, zwitterionic and HC1 salts of L and DL amino-acids, and simple peptides). Experiments on proteins, in particular low temperature 13C and 17O NMR investigation of MbCO substates, as well as 13C, 15N and 19F NMR of normal and mutant myoglobins, prepared via site-directed mutagenesis, will provide a unique data base for myoglobin, which will be interpreted primarily in terms of electrostatic interactions. In addition to explaining the chemical shift problem in proteins, this work should lead to a much better understanding of the forces involved in protein folding, as well as being of relevance to e.g. some enzyme mechanisms, electron transfer and CO/O2 ligand binding. The health-relatedness of this project is that it will lead to a much better understanding of the structures of a series of proteins in their normal state, a prerequisite to understanding abnormal systems, e.g. the hemoglobinopathies. Given the current great interest in solving the structures of proteins (and nucleic acids) by using 2D, 3D and 4D NMR methods, our work should place on a firmer footing the very basis for the success of these methods-the generation of large (e.g. up to about 25 ppm for 15N) chemical shift non-equivalencies in proteins due to folding into their native conformation.
Showing the most recent 10 out of 16 publications
