Mapping molecular accessibility and intermolecular interactions between ribonuclease A and paramagnetic small molecules using nuclear magnetic relaxation /

by Teng, Ching-Ling.

Molecular accessibility to protein surfaces and interiors is fundamental to molecular recognition and biological functions. Combining nuclear spin relaxation and high resolution nuclear magnetic resonance, the relaxation rate induced by small paramagnetic molecules at each structurally distinct protein nuclear site can be measured simultaneously. Because the electron-nuclear dipolar contribution to the nuclear relaxation is inversely proportional the sixth power of the intermoment distance, the changes in the proton relaxation rates report the close proximity, i.e., molecular accessibility, of the paramagnetic molecules to the observed protein proton sites. In our experiments, ribonuclease A (RNase A) is used as the model protein. Molecular oxygen and both neutral and charged nitroxide molecules, 4-hydroxy-TEMPO, 4-amino-TEMPO, and 4-carboxy-TEMPO, are paramagnetic probe molecules. Experiments showed that oxygen has remarkably different accessibility both at the protein surface and in the protein interior. Oxygen is found to penetrate into the loosely-packed protein interior and has higher affinity to protein surface crevices that are located between secondary structures. Neutral nitroxide molecules can not penetrate the protein interior and do not have any significant surface associations with RNase A. Some surface and interior sites of specific associations are observed for charged nitroxide molecules as a result of electrostatic interactions. Computer simulations showed the differences in molecular accessibility to the protein surfaces and interiors can not be explained by the differences in intermolecular contact accessibility that may be deduced from published high-resolution structures. Rather, molecular accessibility is greatly affected by intermolecular interactions, which may be deduced by modeling the differences between the measured paramagnetic relaxation rates and the predicted relaxation rates calculated at the hard-sphere limit. By this method, we find oxygen has higher occupancy at the protein-water interface than in the bulk water. The strengths of the intermolecular interactions between oxygen and the protein are dependent upon the protein surface topology and have a variance about ± 0.5 RT. Direct measurements of the effective electrostatic potentials at the water-protein interface are made by detecting changes in molecular accessibility for different charged nitroxide molecules, which assume no approximation for the charge distributions, the dielectric constant or the static protein structure.