Redox tuning over almost 1 V in a structurally conserved active site: lessons from Fe-containing superoxide dismutase

Abstract

Metalloenzymes catalyze some of the most demanding reactions in biochemistry, thereby enabling organisms to extract energy from redox reactions and utilize inorganic starting materials such as N 2 and CH 4. Bound metal ions bring to enzymes greater chemical versatility and reactivity than would be possible from amino acids alone. However the host proteins must control this broad reactivity, activating the metal for the intended reaction while excluding the rest of its chemical repertoire. To this end, metalloproteins must control the metal ion reduction midpoint potential ( E m), because the E m determines what redox reactions are possible. We have documented potent redox tuning in Fe- and Mn-containing superoxide dismutases (FeSODs and MnSODs), and manipulated it to generate FeSOD variants with E ms spanning 900 mV (21 kcal/mol or 87 kJ/mol) with retention of overall structure. This achievement demonstrates possibilities and strategies with great promise for efforts to design or modify catalytic metal sites. FeSODs and MnSODs oxidize and reduce superoxide in alternating reactions that are coupled to proton transfer, wherein the metal site is believed to cycle between M3+ x OH- and M2+ x OH2 (M = Fe or Mn). Thus the E m reflects the ease both of reducing the metal ion and of protonating the coordinated solvent molecule. Moreover similar E ms are achieved by Fe-specific and Mn-specific SODs despite the very different intrinsic E(m)s of high-spin Fe3+/2+ and Mn3+/2+. We provide evidence that E(m) depression by some 300 mV can be achieved via a key enforced H-bond that appears able to disfavor proton acquisition by coordinated solvent. Based on 15N-nuclear magnetic resonance (NMR), stronger H-bond donation to coordinated solvent can explain the greater redox depression achieved by the Mn-specific SOD protein compared with the Fe-specific protein. Furthermore, by manipulating the strength and polarity of this one H-bond, with comparatively minor perturbation to active site atomic and electronic structure, we succeeded in raising the E m of FeSOD by more than 660 mV, apparently by a combination of promoting protonation of coordinated solvent and providing an energetically favorable source of a redox-coupled proton. These studies have combined the use of electron paramagnetic resonance (EPR), NMR, magnetic circular dichroism (MCD), and optical spectrophotometry to characterize the electronic structures of the various metal sites, with complementary density functional theoretical (DFT) calculations, NMR spectroscopy, and X-ray crystallography to define the protein structures and protonation states. Overall, we have generated structurally homologous Fe sites that span some 900 mV, and have demonstrated the enormous redox tuning accessible via the energies associated with proton transfer coupled to electron transfer. In this regard, we note the possible significance of coordinated solvent molecules in numerous biological redox-active metal sites besides that of SOD.