The photosynthetic protein complex photosystem II oxidizes water to molecular oxygen at an embedded tetramanganese-calcium cluster. Resolving the geometric and electronic structure of this cluster in its highest metastable catalytic state (designated S3) is a prerequisite for understanding the mechanism of O-O bond formation. Here, multifrequency, multidimensional magnetic resonance spectroscopy reveals that all four manganese ions of the catalyst are structurally and electronically similar immediately before the final oxygen evolution step; they all exhibit a 4+ formal oxidation state and octahedral local geometry. Only one structural model derived from quantum chemical modeling is consistent with all magnetic resonance data; its formation requires the binding of an additional water molecule. O-O bond formation would then proceed by the coupling of two proximal manganese-bound oxygens in the transition state of the cofactor.

Photosystem II enriched membranes were depleted of Ca2+ and the 17- and 23-kDa polypeptides by treatment with NaCl and EGTA. The 17- and 23-kDa polypeptides were then reconstituted. This preparation was incapable of O2 evolution until Ca2+ was added. An EPR study revealed the presence of two new EPR signals. One of these is a modified S2 multiline signal with an isotropic g value of 1.96 with at least 26 hyperfine peaks (average spacing 55 G) distributed over approximately 1600 G. The other is a near-Gaussian signal with an isotropic g value of 2.004, which is attributed to a formal S3 state. Experiments involving the interconversion of these signals and the effect of Ca2+ and Sr2+ rebinding provide evidence for these assignments. From these results the following conclusions are drawn: (1) These results are consistent with our earlier demonstration that charge accumulation is blocked after formation of S3 when Ca2+ is deficient. (2) Binding of the 17- and 23-kDa polypeptides to photosystem II in the absence of Ca2+ results in the perturbation of the Mn cluster. This is taken as a further indication that the Ca2+-binding site is close to or even an integral part of the Mn cluster. (3) The S3 signal may arise from an organic free radical interacting magnetically with the Mn cluster. However, other possible origins for this signal, including the Mn cluster itself, must also be considered.

Water binding to the Mn(4)O(5)Ca cluster of the oxygen-evolving complex (OEC) of Photosystem II (PSII) poised in the S(2) state was studied via H(2)(17)O- and (2)H(2)O-labeling and high-field electron paramagnetic resonance (EPR) spectroscopy. Hyperfine couplings of coordinating (17)O (I = 5/2) nuclei were detected using W-band (94 GHz) electron-electron double resonance (ELDOR) detected NMR and Davies/Mims electron-nuclear double resonance (ENDOR) techniques. Universal (15)N (I = ½) labeling was employed to clearly discriminate the (17)O hyperfine couplings that overlap with (14)N (I = 1) signals from the D1-His332 ligand of the OEC (Stich Biochemistry 2011, 50 (34), 7390-7404). Three classes of (17)O nuclei were identified: (i) one μ-oxo bridge; (ii) a terminal Mn-OH/OH(2) ligand; and (iii) Mn/Ca-H(2)O ligand(s). These assignments are based on (17)O model complex data, on comparison to the recent 1.9 Å resolution PSII crystal structure (Umena Nature 2011, 473, 55-60), on NH(3) perturbation of the (17)O signal envelope and density functional theory calculations. The relative orientation of the putative (17)O μ-oxo bridge hyperfine tensor to the (14)N((15)N) hyperfine tensor of the D1-His332 ligand suggests that the exchangeable μ-oxo bridge links the outer Mn to the Mn(3)O(3)Ca open-cuboidal unit (O4 and O5 in the Umena et al. structure). Comparison to literature data favors the Ca-linked O5 oxygen over the alternative assignment to O4. All (17)O signals were seen even after very short (≤15 s) incubations in H(2)(17)O suggesting that all exchange sites identified could represent bound substrate in the S(1) state including the μ-oxo bridge. (1)H/(2)H (I = ½, 1) ENDOR data performed at Q- (34 GHz) and W-bands complement the above findings. The relatively small (1)H/(2)H couplings observed require that all the μ-oxo bridges of the Mn(4)O(5)Ca cluster are deprotonated in the S(2) state. Together, these results further limit the possible substrate water-binding sites and modes within the OEC. This information restricts the number of possible reaction pathways for O-O bond formation, supporting an oxo/oxyl coupling mechanism in S(4).