Metal Ion Substitution at the Catalytic Site of Horse-Liver Alcohol Dehydrogenase: Results from Solvent Magnetic Relaxation Studies. 1. Copper(II) and Cobalt(II) Ions

Abstract

The influence of paramagnetic Cu2+ and CO2+ ions, substituted for Zn2+ ions at the catalytic sites of native alcohol dehydrogenase from horse liver (EC 1.1.1.1), on the nuclear magnetic spin-lattice relaxation rates of solvent water and substrate (CH3OD) protons was studied as a function of magnetic field strength. For the Cu2+ ions, the data can best be fit to a model in which the resulting “blue copper” center (type I) of the enzyme is characterized by inner sphere coordinated water or substrate, both more strongly bound in the binary complex of protein with coenzyme and displaced from the ternary complex with pyrazole. In the binary complex with pyrazole, a pentacoordinated species is indicated; thus, the coordination number is reduced upon formation of the enzyme- pyrazole-coenzyme ternary complex. Although the Cu2+-enzyme is able to bind coenzyme, thereby distorting its metal-binding site, it cannot discriminate significantly between alcohol substrates and water. The resulting relatively weak binding of alcohol is sufficient to explain the observed absence of enzymatic activity of the Cu2+-substituted protein under the usual experimental conditions. This is the first example of a blue copper protein for which the Cu2+ ion is accessible to solvent. The correlation times for the paramagnetic dipolar interaction between the solvent protons and the Cu2+ ion are unusually short, presumably due to a strong spin-orbit interaction of the electronic spins of the Cu2+ ions with their thiol-sulfur ligands. The magnetic spin-lattice relaxation rates of both solvent water protons and solvent methanol methyl protons were also measured for solutions of the native enzyme, the enzyme with Zn2+ ions removed from the catalytic sites, and with CO2+ ions specifically substituted for Zn2+ ions at the catalytic sites. We could detect no paramagnetic contribution from the CO2+ ions to the magnetic relaxation rate of the solvent water and methanol protons, despite attempts to enhance the detection of paramagnetic effects by altering a variety of experimental parameters, including temperature and ionic content of the solvent, and by the addition of coenzyme and inhibitors. There are small differences in the diamagnetic contributions to the relaxation rates of the native, demetalized, and CO2+-substituted enzymes that change sign with magnetic field; these small variations can readily be mistaken for true paramagnetic effects when analysis of the relaxation data is limited to the high values of magnetic field strength usually used for measurements of relaxation enhancement. As a result, previous high-field data require reinterpretation: any paramagnetic effects that may be present are small and not easily separable from a variety of small diamagnetic effects that depend on solvent composition. The reason seems to be an unusually short correlation time for the interaction between solvent protons and the CO2+ ions, due to a strong spin-orbit interaction of the electronic spins of the CO2+ ions, as was found for the Cu2+-substituted enzyme. © 1981, American Chemical Society. All rights reserved.