Search

By using near-UV circular dichroism (CD) and solvent proton nuclear magnetic relaxation dispersion measurements, three different conformational states have been detected in Ca(2+)-Mn(2+)-concanavalin A upon binding a variety of asparagine-linked carbohydrates. Two of these transitions have been described previously, one for the binding of monosaccharides such as methyl alpha-D-mannopyranoside and oligosaccharides with terminal alpha-Glc or alpha-Man residues, and the second for the binding of oligomannose and complex type carbohydrates (Brewer, C. F., and Bhattacharyya, L. (1986) J. Biol. Chem. 261, 7306-7310). The third transition occurs upon binding a bisected biantennary complex type carbohydrate with terminal GlcNAc residues. Temperature-dependent nuclear magnetic relaxation dispersion and CD measurements have identified regions of the protein near the two metal ion binding sites that are associated with the conformation changes, and Tyr-12, which is part of the monosaccharide binding site, as responsible for the CD changes. The results support our previous conclusions that the rotamer conformation of the (alpha 1,6) arm of bisected complex type oligosaccharides binds to concanavalin A with dihedral angle omega = -60 degrees whereas nonbisected complex type oligosaccharides bind with omega = 180 degrees (Bhattacharyya, L., Haraldsson, M., and Brewer, C. F. (1987) J. Biol. Chem. 262, 1294-1299). The present findings also explain the effects of increasing chain length of bisected complex type carbohydrates on their interactions with the lectin.

Bovine and human B erythrocyte carbonic anhydrases substituted with Co2+ cause an enhancement of the nuclear magnetic relaxation rate of solvent water protons (T1-1) at high pH that is decreased by the addition of carbonic anhydrase inhibitors such as azide and Ethoxzolamide. The part of T1-1 which can be inhibited by Ethoxzolamide is due to exchangeable protons located at the active site of the enzyme. This inhibitable part of T1-1 is pH-dependent with a pK of 7.0 ± 0.2 for the bovine cobalt enzyme and a pK of 8.2 ± 0.2 for the human B cobalt enzyme. From the magnetic field dependence of the inhibitable part of T1-1, a correlation time for the dipolar interaction of a proton with a cobalt electronic spin of 10-11 sec, a proton-cobalt distance of 2.5 to 2.9 A, and for the bovine enzyme a proton residence time, τm, small compared to 10-5 sec are calculated. We conclude from the pH dependence and the proton-cobalt distance that the proton which contributes to the inhibitable part of T1-1 is located either on a hydroxide ion bound to the Co2+ at the active site or on a water molecule with 1 proton hydrogen bonded to a nearby residue. In contrast to the similarity of the inhibitable part of T1-1 for the bovine and human B cobalt enzymes, there is a non-inhibitable, pH-independent contribution resulting from the cobalt substitution, relatively large for the bovine enzyme and very small for the human B enzyme, that is not understood. The contributions of the (diamagnetic) native zinc enzymes to T1-1 have also been measured. As in the case of apotransferrin (Koenig, S. H., and Schillinger, W. E., J. Biol. Chem., 244, 3283 (1969)), the results agree qualitatively but not quantitatively with theory.

We report measurements of the magnetic field dependence of the nuclear magnetic spin lattice relaxation rate T1-1 of solvent protons in solutions of the diamagnetic protein apotransferrin, as a function of temperature, concentration, and pH. The bulk of the data were taken over a range of magnetic field (4 Oe to 1000 Oe) much lower than that traditionally used in nuclear magnetic resonance experiments. The major result in that T1-1 is highly magnetic field-dependent; it is very large below 10 Oe, and decreases monotonically as the magnetic field is increased. The data are compared with the predictions of several models for the interaction of solvent protons with the protein molecules, among them the model (Daszkiewicz, O. K., Hennel, T. W., Lubas, B., and Szczepkowski, T. W., Nature, 200, 1006 (1963)) in which water molecules bind irrotationally to the protein surface. The theory for all of the models is very similar. The conclusions are that proton exchange between the bulk water and the protein surface occurs via the exchange of water molecules in the pH range 5 to 8. At higher pH values, direct proton exchange with basic amino acid residues may be important. The theory shows that exchanging water molecules need not be irrotationally bound to the protein; it is sufficient that the water molecules be restricted to rotate about an axis that maintains a fixed geometry with respect to the protein molecule. The number of water molecules involved is a small fraction of the number usually considered to be in the first hydration shell. From the temperature dependence of the data, it is shown that the variation of T1-1 with magnetic field is related to the rotational diffusion relaxation time of the protein molecules, for all models considered. Additionally, the lifetime on the protein molecules of the exchanging entities, whether entire water molecules or individual protons, is in the range 0.1 to 10 µsec.

To investigate the role of specifically bound bicarbonate in the metal-complexing function of transferrin, copper transferrin was prepared both in the presence and absence of bicarbonate ion. The magnetic field and temperature dependence of the solvent proton spin-lattice relaxation rate of solutions of the two complexes were determined. The molar relaxivity R (essentially, relaxation rate per protein concentration) for solutions of copper transferrin is very large at low fields (∼20,000 m-1 sec-1 at 4 Oe) and is highly field-dependent, decreasing to almost zero at 12 kOe. At low field, R is almost temperature-independent over the range 0–45°. Comparison of the results with theory indicates that protons are relaxed by a dipolar interaction with the unpaired electronic spin of Cu2+. The observed relaxivity results from the rapid exchange of a water molecule bound sufficiently close to the copper to be an axial ligand. The magnitude of R at low field for copper transferrin-HCO3 is about one-quarter of that for copper transferrin and also disperses almost to zero at high field. Above 25° the low field value of R has a strong positive temperature dependence. There is no significant variation in R over the pH range 7 to 9.5. The data suggest that, as in the absence of bicarbonate, the protons of water molecules are relaxed by a dipolar interaction with the electron spin of Cu2+, and that bicarbonate binding alters the water exchange rate.

Dispersion curves for phonons propagating in the [100] and [110] directions in an aluminum crystal at ~ 300°K have been measured in a neutron scattering experiment. These data together with the elastic constants obtained by Schmunk and Smith yield dispersion curves having average random errors of ~0·6 per cent, and estimated total errors of 1–2 per cent. Interplanar force constants were obtained by least squares analysis of the experimental dispersion curves. For phonons propagating in the [100] direction, forces due to atoms at least 5 planes distant must be included to obtain a fit to the experimental data, while for phonons in the [110] direction, at least 7 planes are required. This corresponds to an interaction having a range of 2·6 lattice spaces, and includes forces due to atoms out to at least the 15th neighbors.