A charge polarization model for the metal-specific activity of superoxide dismutases

EPR shows that the angular position of the histidine-171 ligand likely plays an important role in metal-selective activities of Mn/FeSODs.


Data collection
MnSOD camSOD Date collected 24 Values in parentheses are for the high-resolution shell. Table 1.

Structural survey of Mn, cambialistic and FeSODs used for entries in
Sixty-four structures of the Mn, cambialistic and FeSODs found in the PDB database were examined. Each structural element, calculated from the structures, were assumed to be statistically independent. PDB accession codes used in the analysis of the crystallographic data. Many of the structures were of proteins carrying a mutation. However, the maximum standard deviations, ±10º for dihedral angles, ±0.12 Å for the Mn(II)-Water bond-lengths and ±0.07Å for the other bond-lengths, were only modestly larger than the variation seen among the wild-type S. aureus, E. coli and H. sapiens MnSODs. The averages and standard deviations given in Table 1 show a conservative estimate of the variation in the structural parameters.

EPR Spectroscopy.
The 94 GHz EPR and ELDOR-NMR spectra were obtained at 5 K using a Bruker Elexsys II 680 EPR spectrometer equipped with a "power upgrade 2" and an Oxford Instruments CF935 flow cryostat. The EPR spectra were obtained by measuring the amplitude of a standard twopulse Hahn echo (t p (π)=24 ns and interpulse time of 400 ns) as a function of the magneticfield.
The ELDOR-NMR spectra were obtained by measuring the integrated two-pulse echo intensity detected at ν obs as a function ν ex ( Figure S3) . They were typically taken overnight.
Microwave cavity tuning, and the lengths and powers of the excitation and detection pulses Figure S3. The ELDOR-NMR experiment. A pulse with frequency ν ex and duration t ex excites a forbidden transition inverting the populations of the two states. The excitation of the forbidden transition reduces the intensity of the electron spin echo (from black to red) detected at ν obs . The ELDOR-NMR spectrum is the echo are listed below. The choices in these parameters were dictated by efficiency and spectra resolution. Spectra were typically taken overnight. 3. π-pulse time measured at the ν obs +ν NMR for SQ and ν Detect +2ν NMR for DQ spectra. 4. π-pulse time at observation frequency (ν obs ).
The normalization methods of the ELDOR-NMR are also listed in Table S3. For SQ 14 N, it was sufficient to use the center transition which took into account number of scans, detection gain and sample concentration. For the 1 H SQ, the center transition was too far to use for normalization. Normalization accounting for number of scans, sample concentration, excitation power and detection gain resulted in about 10% variation from sample to sample and among the three proteins. It was found that the best basis for comparison was to simply normalize the spectra to the height of the 1 H matrix. The 1 H and 14 N DQ spectra were sensitive to excitation power. This was exacerbated by the need to over-couple the cavity, which was difficult to do in a reproducible way, and the variation in the microwave power was evident that the shape of the Mn(Fe)SOD spectrum was different.
All ENDOR spectra were recorded using the method of Davies. [4] They were obtained with a 200-ns microwave preparation pulse followed by a 16-µs radiofrequency pulse and two-pulse spin-echo detection (12-and 24-ns microwave pulses separated by 400 ns) and with a shoot repetition time of 11 ms. Spectra typically took overnight.

Determination of D and E values from spectra.
The zero-field D and E values were obtained directly from the spectra using the three D nn,-5/2 (nn=xx, yy and zz) field positions shown in Figure S4. These positions are given with respect to ν/gβ (where ν is the microwave frequency, g=2.0010 [5] and β=13.996246 GHz/T). The values could be 'read-off' using the first order equations: However, since |D|/ν was significantly large, the equations for the field positions to secondorder in D and E were required to achieve greater accuracy. They were derived from the work of Bir [6] and are: The D zz,-3/2 field position was also useful. First, the first-order estimate of E was obtained from the difference of the D xx,-5/2 and D yy,-5/2 field positions. This E value was used to estimate D to second-order using the difference, D zz,-3/2 -D zz,-5/2 , and the corresponding second-order equations. The second-order D value was then used to obtain a second-order estimate of E.
The last two steps were repeated after which the values became self-consistent. The use of differences rather than absolute field positions removed systematic errors arising the measurement of the magnetic-field.
Analytical expressions for the orientations contributing to the D -3/2,yy magnetic-field position. The orientations of the magnetic-field with respect to the zero-field interaction that contribute to the D -3/2,yy magnetic-field position can be derived from Eq.1. For the m s =-5/2↔-3/2 transition, they are: the m s =-3/2↔-1/2 transition: and m s =+1/2↔+3/2 transition: For the ENDOR calculations, the second-order versions of the expressions were used. These did not lend themselves to simple analytical solution like the three above and required numerical solutions.

Calculation of ENDOR and ELDOR-NMR spectra. The hyperfine tensors and their
orientations obtained from the DFT calculations were used to calculate the ENDOR and ELDOR-NMR spectra. They were used to solved the spin Hamiltonian: The D N are functions of D, E, θ zf and zf and are given in reference [6] . The zero-field D and E values were those obtained from field-swept spin echo spectra described above. For the 14 (Table S4). The genetic insert was cloned into NcoI/XhoI (NEB) digested pET29a E. coli expression vector (Novagen). After sequencing (GATC Biotech, Germany), site-directed mutagenesis with sodB-qc1_F and sodB-qc1_R primer pair was used to correct for an introduced Gly2 residue (insertion of G5, C6 bases) and revert to the wild type sequence, which was confirmed by sequencing.

Elemental analysis of recombinant proteins by inductively coupled plasma massspectrometry (ICP-MS)
Analysis of the metal content of purified proteins was performed by ICP-MS using a Thermo x-series instrument operating in collision cell mode as previously described (Garcia et al., 2017). Stoichiometry of metal to protein was determined by comparing quantified analytes with the protein concentration (A 280nm ).

SOD activity assay
SOD activity was assessed qualitatively in-gel by nitrotetrazolium blue chloride negative staining of purified protein samples resolved on non-denaturing 12% polyacrylamide gels [6] as previously described (Garcia et al., 2017).

Preparation of protein samples for EPR
The purified protein samples for EPR experiments were treated with 10 mM EDTA at room temperature for 30 minutes to remove free manganese contamination. EDTA was removed by . The data were processed using XDS [22] and scaled using Aimless (Evans, 2013). The phase problem was solved by molecular replacement using Molrep [23] , using the search model 2RCV. The pdb models were completed by iterative cycle of refinement in Refmac [24] in tandem with manual model rebuilding in COOT [25] . The PDB codes for Mn(Mn)SOD and Mn(cam)SOD are 5N56 and 5N57, respectively. The data collection and refinement statistics are summarized in Table S1. All crystallographic images were generated using PyMOL Molecular Graphics System, Version 1.8 (Schrödinger, LLC).