X-ray Absorption, Mössbauer, and EPR Studies of the Dinuclear Iron Center in the Hydroxylase Component of Methane Monooxygenase

The dinuclear iron center of the hydroxylase component (protein A) of methane monooxygenase isolated from M. capsulatus (Bath) has been studied by EXAFS, Mössbauer, and EPR spectroscopy. Highly purified hydroxylase, which exhibits a featureless visible absorption spectrum above 300 nm, has been obtained. Analysis of the Fe K-edge EXAFS data for the hydroxylase in the oxidized Fe(III)Fe(III) state gives an Fe⋯Fe distance of 3.42 Å and an average first shell Fe-O/N distance of 2.04 Å. Three of four oxidized hydroxylase samples investigated were photoreduced by the X-ray beam to give an EPR detectable mixed-valent Fe(II)Fe(III) form of the enzyme with g_av = 1.83. For these mixed-valent samples, an Fe⋯Fe distance of 3.41–3.43 Å and an average first shell Fe-O/N distance of 2.06–2.09 Å were determined. The EXAFS data for the photoreduced M. trichosporium OB3b hydroxylase are similar to those obtained for the M. capsulatus (Bath) samples, suggesting that the two proteins contain similar diiron cores. The absence of a short Fe-O distance indicates that the hydroxylase does not contain an oxo bridge. In addition, the results of the first shell fits suggest that the iron atoms are coordinated to more oxygen than nitrogen donor ligands. From a study of the power saturation behavior of the EPR spectrum obtained for the photoreduced M. capsulatus (Bath) sample, an antiferromagnetic exchange coupling constant J = −32 cm⁻¹ was measured. This value is consistent with the presence of a monodentate ligand such as hydroxo, alkoxo, or monodentate carboxylato bridging the Fe(II)Fe(III) center and with the presence of at least one other bridging group, most likely a syn,syn bidentate carboxylate. For Fe(II)Fe(II) hydroxylase EXAFS samples, 85% in the fully reduced state as judged by Mössbauer spectroscopy, no resolvable Fe⋯Fe interaction could be detected, and an average first shell Fe-O/N distance of 2.15 Å was obtained. The Fe K-edge EXAFS spectrum of the diferric model complex [Fe₂O(O₂CH)4(BIPhMe)₂] was also obtained and compared to the protein data. Contrary to our earlier work, we have found that the best fit Fe⋯Fe distance of the hydroxylase samples is dependent on the model compound used to determine parameters for data analysis. The potential consequences of this finding to the analysis of EXAFS data obtained for other iron oxo proteins is discussed.