The Role of Mixed Amine/Amide Ligation in Nickel Superoxide Dismutase

Abstract

Superoxide dismutases (SODs) utilize a ping-pong mechanism in which a redox-active metal cycles between oxidized and reduced forms that differ by one electron to catalyze the disproportionation of superoxide to dioxygen and hydrogen peroxide. Nickel-dependent SOD (NiSOD) is a unique biological solution for controlling superoxide levels. This enzyme relies on the use of cysteinate ligands to bring the Ni(III/II) redox couple into the range required for catalysis (∼300 mV vs. NHE). The use of cysteine thiolates, which are not found in any other SOD, is a curious choice because of their well-known oxidation by peroxide and dioxygen. The NiSOD active site cysteinate ligands are resistant to oxidation, and prior studies of synthetic and computational models point to the backbone N-donors in the active site (the N-terminal amine and the amide N atom of Cys2) as being involved in stabilizing the cysteines to oxidation. To test the role of the backbone N-donors, we have constructed a variant of NiSOD wherein an alanine residue was added to the N-terminus (Ala0-NiSOD), effectively altering the amine ligand to an amide. X-ray absorption, electronic absorption, and magnetic circular dichroism (MCD) spectroscopic analyses of as-isolated Ala0-NiSOD coupled with density functional theory (DFT) geometry optimized models that were evaluated on the basis of the spectroscopic data within the framework of DFT and time-dependent DFT computations are consistent with a diamagnetic Ni(II) site with two cysteinate, one His1 amide, and one Cys2 amidate ligands. The variant protein is catalytically inactive, has an altered electronic absorption spectrum associated with the nickel site, and is sensitive to oxidation. Mass spectrometric analysis of the protein exposed to air shows the presence of a mixture of oxidation products, the principal ones being a disulfide, a bis-sulfenate, and a bis-sulfinate derived from the active site cysteine ligands. Details of the electronic structure of the Ni(III) site available from the DFT calculations point to subtle changes in the unpaired spin density on the S-donors as being responsible for the altered sensitivity of Ala0-NiSOD to O₂.

Figure 1.

Active site of NiSOD (PDB code 1T6U) in the a) reduced and b) oxidized states.

Figure 2.

Room temperature UV-Vis of anaerobic and air exposed Ala0-NiSOD.

Figure 3.. Mass spectral data and analysis for Ala0-NiSOD_ox.

(a) The ESI-MS spectrum of Ala0-NiSOD_ox before the addition of 5mM TCEP (b) The overlaid spectra of Ala0-NiSOD_ox before and after the addition of 5mM TCEP.

Figure 4.

DSC thermogram of Ala0-NiSOD compared with WT-NiSOD.

Figure 5.

Disappearance of O₂⁻ monitored at 260 nm in the presence and absence of 10 μM Ala0-NiSOD. Conditions: pulse-radiolytic generation of 30 μM (bottom curves) or 45 μM O₂⁻ (top curves) in 10 mM phosphate buffer pH = 8.5, 30 mM formate and 5 μM EDTA.

Figure 6.

Ni K-edge XANES spectra of anaerobic Ala0-NiSOD (blue) and dithionite-reduced WT-NiSOD (black) from ref.

Figure 7.

Ni K-edge EXAFS spectra of Ala0-NiSOD (blue) and best fit from Table 2 (red).

Figure 8.

Geometry optimized active site models of S=0, Ni(II)-bound WT and Ala0 NiSOD: a) WT^red, b) Ala0^red, and c) Ala0-dep^red. Asterisks (*) denote the atoms that were frozen during geometry optimizations.

Figure 9.

Low temperature absorption (top) and MCD (bottom) spectra of Ala0-NiSOD prepared under anaerobic conditions.

Figure 10.

Low temperature absorption (top) and MCD (bottom) spectra of Ala0-NiSOD following long-term exposure to air.

Figure 11.

TD-DFT computed absorption spectra for a) WT^red b) Ala0^red, and c) Ala0-dep^red. All spectra were red-shifted by 10,000 cm⁻¹ to facilitate a direct comparison with the experimental absorption spectrum. The EDDM for the key transition of Ala0^red is shown at the top of the corresponding spectrum, where blue and white represent areas of loss and gain of electron density.

Figure 12.

DFT-calculated MO diagram and isosurface plots of relevant filled (blue/white) and empty (red/white) MOs of Ala0^red. Note that the HOMO/LUMO gap is not drawn to scale.

Figure 13.

Geometry optimized active site models of S=1/2 Ni(III)-bound WT and Ala0 NiSOD: A) WT^ox and b) Ala0^ox. Asterisks (*) denote the atoms that were frozen during geometry optimizations.

Figure 14.

DFT-calculated MO diagram and isosurface plots of relevant filled (blue/white) and empty (red/white) β-MOs of Ala0^ox. Inset: unpaired spin density plot, where white and yellow lobes show positive and negative spin density, respectively. Note that the HOMO/LUMO gap is not drawn to scale.

Figure 15.

DFT-computed compositions and isosurface plots of a) the HOMO of WT^red and Ala0^red and b) the β-LUMO of WT^ox and Ala0^ox.