On the status of ferryl protonation

We examine the issue of ferryl protonation in heme proteins. An analysis of the results obtained from X-ray crystallography, resonance Raman spectroscopy, and extended X-ray absorption spectroscopy (EXAFS) is presented. Fe–O bond distances obtained from all three techniques are compared using Badger’s rule. The long Fe–O bond lengths found in the ferryl crystal structures of myoglobin, cytochrome c peroxidase, horseradish peroxidase, and catalase deviate substantially from the values predict by Badger’s rule, while the oxo-like distances obtained from EXAFS measurements are in good agreement with the empirical formula. Density functional calculations, which suggest that Mössbauer spectroscopy can be used to determine ferryl protonation states, are presented. Our calculations indicate that the quadrupole splitting (ΔEQ) changes significantly upon ferryl protonation. New resonance Raman data for horse-heart myoglobin compound II (Mb-II, pH 4.5) are also presented. An Fe–O stretching frequency of 790 cm−1 (shifting to 754 cm−1 with 18O substitution) was obtained. This frequency provides a Badger distance of rFe–O = 1.66 Å. This distance is in agreement with the 1.69 Å Fe–O bond distance obtained from EXAFS measurements but is significantly shorter than the 1.93 Å bond found in the crystal structure of Mb-II (pH 5.2). In light of the available evidence, we conclude that the ferryl forms of myoglobin (pKa ⩽ 4), horseradish peroxidase (pKa ⩽ 4), cytochrome c peroxidase (pKa ⩽ 4), and catalase (pKa ⩽ 7) are not basic. They are authentic FeIVoxos with Fe–O bonds on the order of 1.65 Å.