Nature of oxygen activation in glucose oxidase from Aspergillus niger: the importance of electrostatic stabilization in superoxide formation.

Glucose oxidase catalyzes the oxidation of glucose by molecular dioxygen, forming gluconolactone and hydrogen peroxide. A series of probes have been applied to investigate the activation of dioxygen in the oxidative half-reaction, including pH dependence, viscosity effects, 18O isotope effects, and solvent isotope effects on the kinetic parameter Vmax/Km(O2). The pH profile of Vmax/Km(O2) exhibits a pKa of 7.9 +/- 0.1, with the protonated enzyme form more reactive by 2 orders of magnitude. The effect of viscosogen on Vmax/Km(O2) reveals the surprising fact that the faster reaction at low pH (1.6 x 10(6) M-1 s-1) is actually less diffusion-controlled than the slow reaction at high pH (1.4 x 10(4) M-1 s-1); dioxygen reduction is almost fully diffusion-controlled at pH 9.8, while the extent of diffusion control decreases to 88% at pH 9.0 and 32% at pH 5.0, suggesting a transition of the first irreversible step from dioxygen binding at high pH to a later step at low pH. The puzzle is resolved by 18O isotope effects. 18(Vmax/Km) has been determined to be 1.028 +/- 0.002 at pH 5.0 and 1.027 +/- 0.001 at pH 9.0, indicating that a significant O-O bond order decrease accompanies the steps from dioxygen binding up to the first irreversible step at either pH. The results at high pH lead to an unequivocal mechanism; the rate-limiting step in Vmax/Km(O2) for the deprotonated enzyme is the first electron transfer from the reduced flavin to dioxygen, and this step accompanies binding of molecular dioxygen to the active site. In combination with the published structural data, a model is presented in which a protonated active site histidine at low pH accelerates the second-order rate constant for one electron transfer to dioxygen through electrostatic stabilization of the superoxide anion intermediate. Consistent with the proposed mechanisms for both high and low pH, solvent isotope effects indicate that proton transfer steps occur after the rate-limiting step(s). Kinetic simulations show that the model that is presented, although apparently in conflict with previous models for glucose oxidase, is in good agreement with previously published kinetic data for glucose oxidase. A role for electrostatic stabilization of the superoxide anion intermediate, as a general catalytic strategy in dioxygen-utilizing enzymes, is discussed.