The putative initial adduct of ferrous superoxide reductase (SOR) with superoxide has been alternatively formulated as ferric-peroxo or ferrous-superoxo. The ~600-nm UV-vis absorption band proposed to be assigned to this adduct (either as sole intermediate in the SOR catalytic cycle, or as one of the two intermediates) has recently been interpreted as due to a ligand-to-metal charge transfer, involving thiolate and superoxide in a ferrous complex, contrary to an alternative assignment as a predominantly cysteine thiolate-to-ferric charge transfer in a ferric-peroxo electromer. In an attempt to clarify the electromeric formulation of this adduct, we report a computational study using a multiconfigurational complete active space self-consistent field (MC-CASSCF) wave function approach as well as modelling the UV-vis absorption spectra with time-dependent density functional theory (TD-DFT). The MC-CASSCF calculations disclose a weak interaction between iron and the dioxygenic ligand and a dominant configuration with an essentially ferrous-superoxo character. The computed UV-vis absorption spectra reveal a marked dependence on the choice of density functional - both in terms of location of bands and in terms of orbital contributors. For the main band in the visible region, besides the recently reported thiolate-to-superoxide charge transfer, a more salient, and less functional-dependent, feature is a thiolate-to-ferric iron charge transfer, consistent with a ferric-peroxo electromer. By contrast, the computed UV-vis spectra of a ferric-hydroperoxo SOR model match distinctly better (and with no qualitative dependence on the DFT methodology) the 600-nm band as due to a mainly thiolate-to-ferric character - supporting the assignment of the SOR "600-nm intermediate" as a S=5/2 ferric-hydroperoxo species.
Keywords: CASSCF; Electromerism; Non-heme iron; SOR; Superoxide reductase; TD-DFT.
Copyright © 2016 Elsevier Inc. All rights reserved.