CO dehydrogenase (CODH) is an environmentally crucial bacterial enzyme that oxidizes CO to CO2 at a Mo–Cu active site. Despite the close to atomic resolution structure (1.1 Å), significant uncertainties have remained with regard to the protonation state of the water-derived equatorial ligand coordinated at the Mo-center, as well as the nature of intermediates formed during the catalytic cycle. To address the protonation state of the equatorial ligand, we have developed a realistic in silico QM model (~179 atoms) containing structurally essential residues surrounding the active site. Using our QM model, we examined each plausible combination of redox states (MoVI–CuI, MoV–CuII, MoV–CuI, and MoIV–CuI) and Mo-coordinated equatorial ligands (O2−, OH−, H2O), as well as the effects of second-sphere residues surrounding the active site. Herein, we present a refined computational model for the Mo(VI) state in which Glu763 acts as an active site base, leading to a MoO2-like core and a protonated Glu763. Calculated structural and spectroscopic data (hyperfine couplings) are in support of a MoO2-like core in agreement with XRD data. The calculated two-electron reduction potential (E = −467 mV vs. SHE) is in reasonable agreement with the experimental value (E = −558 mV vs. SHE) for the redox couple comprising an equatorial oxo ligand and protonated Glu763 in the MoVI–CuI state and an equatorial water in the MoIV–CuI state. We also suggest a potential role of second-sphere residues (e.g., Glu763, Phe390) based on geometric changes observed upon exclusion of these residues in the most plausible oxidized states.
CO dehydrogenase Molybdenum–copper bimetallic site Density functional theory Quantum mechanics Computational model
This is a preview of subscription content, log in to check access.
This research was funded by generous financial contributions from Whitman College and the M. J. Murdock Charitable Trust. Special thanks to Dr. Robert Szilagyi (Montana State University, Bozeman, MT) for his tremendous assistance in setting up the computational server at Whitman College, and for providing comments and feedback during the preparation of this manuscript. We gratefully acknowledge the Max Planck Society for financial support of this work.