Comparison of Rates and Kinetic Isotope Effects Using PEG-Modified Variants and Glycoforms of Glucose Oxidase:  The Relationship of Modification of the Protein Envelope to C−H Activation and Tunneling

An earlier investigation of the temperature dependencies of rates and kinetic isotope effects (KIEs) in glucose oxidase (GO) used variants that differed in the extent of glycosylation at the surface of the protein. Kohen et al. [Kohen, A., Jonsson, T., and Klinman, J. P. (1997) Biochemistry 36, 2603−2611] presented evidence that the KIE on the Arrhenius prefactor varied as a function of protein modification, concluding that the degree of hydrogen tunneling at the active site was dependent on changes in mass at the surface. We now examine GO proteins containing polyethylene glycol (PEG) at their surface and a more extensively glycosylated form of GO, to distinguish simple mass effects from other sources of altered catalytic behavior. One PEG variant was created by modifying deglycosylated GO with short PEG chains (average of 350 Da each), while another contained a smaller number of long PEG chains (average of 5000 Da each). The light (146 kDa) and heavy (211 kDa) PEG variants and the hyperglycosylated variant display isotope effects on the Arrhenius prefactor that are similar (A D/A T = 0.55−0.62), while the unperturbed wild-type GO (WT-GO) is found to have an A D/A T that is reassessed as being close to unity. It appears that any modification of the protein surface away from that of the wild type gives rise to altered behavior for hydrogen transfer in the active site. We have also compared the effect of enthalpies of activation on both k cat/K M and k cat for the variants, introducing a new method to extract the k cat/K M rate constant and enthalpy of activation for the tritiated substrate from competitive KIE experiments. We find similar trends in ΔH ⧧ for both competitive and noncompetitive parameters and a smaller trend in k cat than reported earlier. Correlations are observed between A D/A T and both the enthalpies of activation and the thermal melt temperatures (T M) of the GO isoforms. In addition to the present study, there are now a number of examples where a perturbation of enzyme structure away from that of the wild type causes the observed KIE to become more temperature-dependent. The implications of these findings are discussed in the context of hydrogen tunneling and the relationship of protein structure and dynamics to this process.