Bicarbonate binding to the non-heme iron of photosystem II, investigated by Fourier transform infrared difference spectroscopy and 13C-labeled bicarbonate.

Rainer Hienerwadel,Catherine Berthomieu
DOI: https://doi.org/10.1021/bi00050a008
IF: 3.321
1995-12-01
Biochemistry
Abstract:The binding site of the non-heme iron of photosystem II (PS II) is investigated by light-induced Fourier tranform infrared (FTIR) difference spectroscopy on Tris-washed membranes. The non-heme iron is oxidized (Fe3+) in the dark with ferricyanide and reduced (Fe2+) after light-induced charge separation by electron transfer from the semiquinone anion QA-. EPR experiments and IR modes of ferri- and ferrocyanide show that the electron donor side of PS II is reduced in less than 2 s after a flash and that ferricyanide reoxidizes the non-heme iron with a half-time of approximately 20 s. Recording FTIR spectra before and 2 s after flash illumination thus results in the Fe2+/Fe3+ difference spectrum. This spectrum shows band shifts and intensity changes of IR modes from ligands and neighboring residues of the non-heme iron. The IR modes of bicarbonate are revealed by comparison of Fe2+/Fe3+ spectra obtained on PS II membranes with 12C or 13C isotope labeled bicarbonate in H2O and in 2H2O. The nu as(CO) and nu s(CO) modes of bicarbonate in the Fe2+ state are assigned at 1530 +/- 10 and 1338 cm-1, respectively. The low frequency of the nu as(CO) mode is taken as experimental evidence that bicarbonate is a ligand of the non-heme iron. Furthermore, the small frequency difference (192 cm-1) between the nu as(CO) and nu s(CO) modes as compared to even hydrogen-bonded ionic bicarbonate strongly indicates that bicarbonate is a bidentate ligand of the non-heme iron in PS II. Upon iron oxidation, the bicarbonate modes are largely affected. The nu s(CO) mode is assigned at 1228 cm-1, while the nu as(CO) mode is tentatively assigned at 1658 +/- 20 cm-1. The strong up- and downshifts of the nu as and nu s(CO) modes of bicarbonate upon iron oxidation results in a frequency difference of 430 +/- 20 cm-1 that is not only explained by the increased charge on the iron but indicates that bicarbonate is a monodentate ligand of the oxidized iron. The sensitivity of the nu s(CO) mode of bicarbonate to 1H/2H exchange in both the Fe2+ and Fe3+ states and the presence in the Fe2+ state of a delta (COH) mode at 1258 cm-1 confirm that bicarbonate and not carbonate is the iron ligand and further exhibits hydrogen bond(s) with the protein. The 13C isotope-sensitive modes of bicarbonate are not affected by 15N labeling of the PS II membranes. 15N sensitive signals at 1111/1102 and 1094 cm-1 are assigned to side chain modes from histidine ligands of the iron. The latter signal is proposed to account for a histidine ligand that deprotonates upon iron oxidation. The involvement of protein peptide groups and side chains in the hydrogen-bond network around the iron is also discussed.
biochemistry & molecular biology
What problem does this paper attempt to address?