Tracking Reactive Water and Hydrogen-Bonding Networks in Photosynthetic Oxygen Evolution

ConspectusIn oxygenic photosynthesis, photosystem II (PSII) converts water to molecular oxygen through four photodriven oxidation events at a Mn₄CaO₅ cluster. A tyrosine, YZ (Y161 in the D1 polypeptide), transfers oxidizing equivalents from an oxidized, primary chlorophyll donor to the metal center. Calcium or its analogue, strontium, is required for activity. The Mn₄CaO₅ cluster and YZ are predicted to be hydrogen bonded in a water-containing network, which involves amide carbonyl groups, amino acid side chains, and water. This hydrogen-bonded network includes amino acid residues in intrinsic and extrinsic subunits. One of the extrinsic subunits, PsbO, is intrinsically disordered. This extensive (35 Å) network may be essential in facilitating proton release from substrate water. While it is known that some proteins employ internal water molecules to catalyze reactions, there are relatively few methods that can be used to study the role of water. In this Account, we review spectroscopic evidence from our group supporting the conclusion that the PSII hydrogen-bonding network is dynamic and that water in the network plays a direct role in catalysis. Two approaches, transient electron paramagnetic resonance (EPR) and reaction-induced FT-IR (RIFT-IR) spectroscopies, were used. The EPR experiments focused on the decay kinetics of YZ• via recombination at 190 K and the solvent isotope, pH, and calcium dependence of these kinetics. The RIFT-IR experiments focused on shifts in amide carbonyl frequencies, induced by photo-oxidation of the metal cluster, and on the isotope-based assignment of bands to internal, small protonated water clusters at 190, 263, and 283 K. To conduct these experiments, PSII was prepared in selected steps along the catalytic pathway, the S_<i>n</i> state cycle (<i>n</i> = 0–4). This cycle ultimately generates oxygen. In the EPR studies, S-state dependent changes were observed in the YZ• lifetime and in its solvent isotope effect. The YZ• lifetime depended on the presence of calcium at pH 7.5, but not at pH 6.0, suggesting a two-donor model for PCET. At pH 6.0 or 7.5, barium and ammonia both slowed the rate of YZ• recombination, consistent with disruption of the hydrogen-bonding network. In the RIFT-IR studies of the S state transitions, infrared bands associated with the transient protonation and deprotonation of internal waters were identified by D₂O and H₂¹⁸O labeling. The infrared bands of these protonated water clusters, W_<i>n</i>⁺ (or <i>n</i>H₂O­(H₃O)⁺, <i>n</i> = 5–6), exhibited flash dependence and were produced during the S₁ to S₂ and S₃ to S₀ transitions. Calcium dependence was observed at pH 7.5, but not at pH 6.0. S-state induced shifts were observed in amide CO frequencies during the S₁ to S₂ transition and attributed to alterations in hydrogen bonding, based on ammonia sensitivity. In addition, isotope editing of the extrinsic subunit, PsbO, established that amide vibrational bands of this lumenal subunit respond to the S state transitions and that PsbO is a structural template for the reaction center. Taken together, these spectroscopic results support the hypothesis that proton transfer networks, extending from YZ to PsbO, play a functional and dynamic role in photosynthetic oxygen evolution