Pivotal role of the redox-active tyrosine in driving the water splitting catalyzed by photosystem II

Abstract

Photosynthetic water oxidation is catalyzed by the Mn₄Ca cluster in photosystem II (PSII). The nearby redox-active tyrosine (Y_Z) serves as a direct electron acceptor of the Mn₄Ca cluster and it forms a low-barrier H-bond (LBHB) with a neighboring histidine residue (D1-His190). Experimental evidence indicates that Y_Z oxidation triggers changes in the hydrogen bonding network that precede proton abstraction from the Mn₄Ca cluster. In order to characterize such changes, we compare ab initio molecular dynamics simulations of different states of the catalytic cycle of PSII with dynamics of isolated tyrosine models (namely, p-cresol) in different oxidation states. The systematic comparison of the H-bond networks in different simulated systems suggests that the Y_Z oxidation leads to a water hydration pattern which is more similar to that of the neutral p-cresol rather than that of the p-cresol anion. Our simulations also reveal the twofold nature of the interactions between Y_Z and the Mn₄Ca cluster. Firstly, the Y_Z oxidation triggers rapid structural changes of the H-bond pattern in the proximity of the cluster which have been observed to propagate on the ps time scale on the Ca²⁺ hydration shell up to other water molecules in the proximity of the cluster. Secondly, it is clear that Y_Z interacts with the Mn₄Ca cluster also through Coulombic interactions mediated by CP43-Arg357 through the remaining positive charge of the pair. Our results are able to identify, for the first time, the structural rearrangements guided by the oxidation of Y_Z necessary for the evolution of the water splitting reaction in PSII. Based on these findings, we propose a mechanism of structural changes which is functional towards the progression of the catalytic cycle in PSII.

---

Introduction

The presence of oxygen in the atmosphere is required for the survival of all life forms that breathe, and it has been ultimately generated by photosynthetic water oxidation occurring in plants, algae and cyanobacteria.^4–6 Elucidation of the chemical mechanism of water oxidation is crucial for improvement and development of solar fuel systems and artificial photosynthesis devices.^7–10 The role of photosystem II (PSII), the water splitting complex embedded in the thylakoid membrane, is to obtain the electrons from water oxidation for subsequent CO₂ fixation to produce sugars.^11–16 Water oxidation splits water into protons and oxygen molecules, and it takes place in the water-oxidizing center (WOC) located on the luminal side of the D1 protein in PSII (Fig. 1A).^17–20

Fig. 1 The atomic structure of (A) amino acid ligands and (B) the hydrogen bond network around the Mn₄Ca cluster in the XFEL structure at a resolution of 1.95 Å (PDB code: 4UB6) with the direction of three channels linking to the bulk reported previously^1–3 (green arrows). The interacting amino acids and the water molecules are selectively shown as the QM region.

Several theoretical and experimental studies have provided considerable details regarding the electron transfer routes and rates in PSII.^5,21–27 The light-induced excited state of the reaction center of chlorophylls (i.e., the special pair of chlorophyll moieties, P680 and Chl_D1) releases an electron to the electron acceptor side. P680⁺ that is generated extracts an electron from a redox-active tyrosine, Y_Z. The resulting can oxidize the Mn₄Ca cluster and convert back to the reduced form on a time scale that ranges from 20 μs to 2 ms.²⁸ Finally, the Mn₄Ca cluster oxidizes two water molecules, passing through five intermediate states (S_i → S_i+1, i = 0–4), which is referred to as the Kok–Joliot cycle.^29,30

The redox reaction of a tyrosine side chain involves a proton-coupled electron transfer (PCET) reaction. A tyrosine side chain releases a phenolic proton upon its oxidation due to the extremely low pK_a (the pK_a value is approximately −2).³¹ When Y_Z is oxidized by P680⁺, it releases a proton from the phenolic group to a neighboring histidine residue, D1-His190 (Fig. 1B).^32–35 The recently reported high-resolution X-ray structure of photosystem II reveals an extremely short-distance H-bond (∼2.5 Å) between the Nτ atom of D1-His190 and the phenolic oxygen of Y_Z.^17–19 The QM/MM data also revealed that Y_Z and His190 form a low-barrier H-bond (LBHB).³⁶ This LBHB greatly facilitates the rapid PCET reaction of Y_Z due to the lower barrier proton transfer from the phenolic group to D1-His190. This is a basic mechanistic feature of the redox reaction of Y_Z.

Another redox-active tyrosine, Y_D, located symmetrically in the D2 protein of PSII, is also able to donate an electron to P680⁺. Y_D functions as a secondary electron transfer pathway and shows competitive electron transfer with Y_Z. The redox reaction rate of Y_D is much slower than that of Y_Z, the Y_D radical form being stabilized by a different arrangement of the Tyr–His pair and the surrounding water molecules.^33,34 The H-bond structure and the PCET mechanism of Y_D have been elucidated in recent years. A theoretical study by Saito et al. predicted that the phenolic proton of Y_D is released to the bulk via an H-bonding network upon Y_D oxidation.³⁷ Proton release to the bulk upon Y_D oxidation has indeed been demonstrated by Fourier transform infrared (FTIR) spectroscopy.³⁸ The difference in the proton transfer reactions of the two tyrosines, Y_Z and Y_D, underlines their functional difference. These two proton-release reactions are characterized by the H-bond directions caused by the different protonation states of their neighboring histidine residues. Although data regarding the difference in PCET mechanisms of the two tyrosines have been reported previously, the factors determining the H-bond properties, such as the strength and the number of H-bonds of the tyrosine side chains, are still a matter of debate.

Although, recently, the different steps of the Kok–Joliot cycle have attracted a lot of attention and are studied actively,^39–49 the role of Y_Z in water oxidation is still one of the main issues that need to be elucidated for the mechanism of water oxidation. The X-ray structure of PSII also shows that Y_Z is able to interact with the Mn₄Ca cluster via H-bond networks, comprising several water molecules, and that it is located in the H-bond network linking the cluster to the luminal surface.^17–19 Electron paramagnetic resonance (EPR) has been used to examine the magnetic and electronic interactions between Y_Z and the Mn₄Ca cluster by detecting the split signal arising from the interaction of the spin states of Y_Z and the cluster itself. These EPR studies have shown the weakness of this interaction.^50–53 It has also been postulated that the positive charge of triggers the proton release reaction from substrate water through electrostatic interactions with the Mn₄Ca cluster^54–57 and that the proton of the phenolic oxygen of Y_Z is involved in the proton transfer from substrate water.⁵⁸ However, a consensus has not yet been reached regarding the role of Y_Z in water oxidation.

Our previous study also showed that the redox states of Y_Z control the equilibrium between the two isomers of the Mn₄Ca cluster in the S₂ → S₃ transition.⁵⁹ However, because the previous study focused on the role of the two isomers of the cluster in the S₂ → S₃ transition, the structural effects of the Y_Z oxidation on the cluster and the possible influence on the H-bond network around Y_Z were not analyzed in detail. In addition, in order to elucidate the mechanism of Y_Z involvement in water oxidation, a detailed study should include a systematic comparison between the H-bond properties of Y_Z and those of a basic tyrosine side chain.

So far, few studies have focused on the difference in the H-bond properties in each state of a free tyrosine molecule, although the structural and spectroscopic properties of each state have been well established.^31,60–66 One of the reasons for this is the complexity of the system due to the presence of several H-bond patterns of a phenolic oxygen of a tyrosine side chain. A tyrosine can form an H-bond donor and/or acceptor, while a deprotonated tyrosine forms only an H-bond acceptor but with a different oxidation state. Thus, elucidating the H-bond properties of each state of a tyrosine is an essential and necessary step to understand the role of a key reactive tyrosine. This information can strongly contribute to clarifying the fundamental molecular mechanism of the protein functions as well as to predicting its function and structural changes from the structural information.

In the present study, we performed quantum mechanical-molecular dynamics (QM-MD) simulations of p-cresol in aqueous solution, which is a simple model of a tyrosine side chain, and quantum mechanics/molecular mechanics-molecular dynamics (QM/MM-MD) simulations of Y_Z and the Mn₄Ca cluster in PSII. Since H-bonding is a dynamic process, the use of molecular dynamics simulations is crucial to pinpoint changes in the H-bond pattern, since shallow local minima can be easily overcome during the simulated time.

By comparing the results of the QM/MM simulations of Y_Z in PSII and the simple solvated p-cresol models, we are able to identify that the specificity of the H-bond properties of oxidized Y_Z is to form an anion state that promotes the formation of three H-bonds. The present simulations also show the S-state dependence of the H-bond structure of Y_Z and the rearrangements of the H-bond structure occurring around both Y_Z and the Mn₄Ca cluster upon Y_Z oxidation. These rearrangements suggest that an arginine residue, CP43-Arg357, is involved in water oxidation by control of the H-bond structure around the O4 atom of the cluster. Based on the data of the structural effect upon Y_Z oxidation and the H-bond properties of Y_Z, we discuss the possible proton transfer mechanism from O5 in concert with Y_Z re-reduction with the rate-limiting electron transfer to .

Methods

The computational setup for the QM-MD simulations of p-cresol was based on the classical MD simulation. The initial conditions for the classical MD simulations were created by TLEAP software in the AMBER program package. The p-cresol molecules were modeled using the GAFF force field⁶⁷ and RESP charges calculated at the level of HF/6-31(d) by Gaussian09 software.⁶⁸ We added 120 water molecules around the p-cresol moiety in a cubic boundary box. The full systems were treated at the quantum level.

For the full QM-MD simulations of p-cresol in aqueous solution, we first performed individual classical MD simulations for each protonation or oxidation state (p-cresol, anionic p-cresol⁻, and oxidized p-cresol˙) to obtain the initial structures and set the boundary conditions using AMBER12 software.⁶⁹ The MD simulations in the NPT ensemble were performed for 5.2 ns (2 fs per step) at 298 K using a Langevin heat bath after the NVT ensemble simulation for 100 ps. The equilibrated structures and the cubic boundary box were used for the QM-MD simulations, using CP2K ver.4.1 software.^70,71 The QM-MD simulations in the NVT ensemble were performed for 80 ps (0.5 fs per step) at 298 K, using the PBE functional⁷² and the DZVP-MOLOPT-SR-GTH basis set⁷³ with the GTH-PBE potential.^74,75 The cutoff value for the plane-wave expansion in the QM calculation equaled 320 Rydberg.

The calculation system for the QM/MM-MD simulations of WOC in PSII was based on previous studies.^76,77 The initial structures were taken from the structure with PDB ID: 3ARC. Water molecules, ions and protons were added to the systems and equilibrated by means of classical MD simulations as previously described in ref. 76 and 77. The whole QM/MM system consisted of the D1, D2, and CP43 proteins, as well as the redox co-factor and water molecules located within these proteins. The protein environment topology and other cofactors were followed by the AMBER99SB⁷⁸ and GAFF force fields, respectively, as previously described.⁷⁹ The QM region of the Mn₄Ca cluster in the S₁, S₂, and S₃ states was constructed as described in our previous study.⁸⁰ Such a region consists of 206 atoms, which includes the Mn₄Ca cluster, the side chains of its ligands (D1-Asp170, D1-Glu189, D1-His332, D1-Glu333, D1-Ala344, D1-Asp342, and CP43-Glu354), additional residues (D1-Asp61, D1-Tyr161, D1-His190, CP43-Arg357, D1-His337, D1-Ile60, D1-Ser169, and D1-Gly171), the Cl⁻ ion and its ligand D1-Asn181, and a further 17 water molecules. The QM/MM-MD simulations of WOC in each oxidation state (S₁, S₂, S₂Y_Z⁺, and S₃) were performed as described previously,^59,76,80,81 using the PBE+U functional with the Hubbard correction of 1.16 eV for Mn ions⁸² and the DZVP-MOLOPT-SR-GTH basis set with the GTH-PBE pseudopotential. A cubic cell of 28 Å on each side and a cutoff for the plane-wave expansion of 320 Rydberg were applied for the QM calculations. The QM/MM-MD simulations were performed in the NVT ensemble with a thermal bath at 298 K using a Nosè–Hoover thermostat^83,84 for 20 ps (0.5 fs per step) after the equilibration, with the exception of S₂Y_Z⁺. The MD simulation of the S₂Y_Z⁺ model started from the equilibrated S₂ structure and was performed for 8.5 ps (0.5 fs per step) with the same conditions as the other S states. The total spin of the Mn₄Ca cluster was set to S = 0, 1/2, and 3 in the S₁, S₂, and S₃ states, respectively. The spin configuration of the Mn₄Ca cluster was simulated by means of broken-symmetry DFT, based on the results of previous works consistently with the high oxidation paradigm.^26,46 Mulliken spin populations of Mn ions in QM/MM-MD simulations of each S state are reported in Table S1 (ESI†).

Results

Models of p-cresol, p-cresol⁻, and p-cresol˙ in aqueous solution

A tyrosine side chain can assume three different forms: i.e., neutral, high pH (anionic deprotonated), and oxidized (radical deprotonated). To estimate the H-bond properties of a tyrosine side chain, we used QM-MD simulations of solvated p-cresol, p-cresol⁻, and p-cresol˙ models, in aqueous solution. Fig. 2 shows the density maps of the solvation shell of the oxygen atoms of water around the phenolic oxygen for each p-cresol form. In the p-cresol model, two regions of oxygen atoms were observed in the H-bond donor and acceptor sides of the phenolic oxygen. Compared with the p-cresol, three regions of oxygen atom density were arranged around the phenolic oxygen in the p-cresol⁻, while two regions were observed in the p-cresol˙. The two density maps of the oxygen atoms of water molecules in the p-cresol and the p-cresol˙ overlap despite the different protonation states. The different densities for the two deprotonated states (i.e., the p-cresol⁻ and the p-cresol˙) are to be expected: the anion form has a higher affinity for accepting an H-bond than the radical form due to the negative charge of the phenolic oxygen. It was noted that the two regions in the radical state were arranged perpendicularly with respect to the direction of the plane of the phenolic ring. This arrangement of the water density in the p-cresol˙ indicates that the tyrosyl radical behaves as an H-bond acceptor with a C=O double bond in the phenolic group.

Fig. 2 Density maps of the oxygen atoms of water molecules located within the H-bond distance (∼3.5 Å) in the neutral (A), the anion (B), and the radical (C) forms of p-cresol with the same isovalue of isosurface as determined by QM-MD simulations.

Additionally, we report the calculated radial distribution functions of the oxygen and hydrogen atoms of water molecules around the phenolic oxygen in the ESI†, Fig. S1. The peak positions of the first hydration shell in the p-cresol⁻ model are 2.60 Å and 1.50 Å in the O_p-cresol–O_water and the O_p-cresol–H_water, respectively, which are shortened from those of the p-cresol (2.65 Å and 1.85 Å in the O_p-cresol–O_water and O_p-cresol–H_water, respectively) and the p-cresol˙ (2.70 Å and 1.75 Å in the O_p-cresol–O_water and O_p-cresol–H_water, respectively) models. This indicates, as expected, that the H-bonds are stronger in the anion than in the other forms. To quantitatively measure the structural properties around each model, we calculated the O–O and O–H coordination numbers, obtained by a definite integral of the radial distribution function up to its first minimum (Table 1). The coordination numbers of the O_p-cresol–O_water in the first hydration shell in p-cresol⁻ and p-cresol˙ were 3.00 and 1.96, respectively, which are consistent with the corresponding density maps. The coordination number of the O_p-cresol–O_water of the p-cresol (2.33) instead was slightly larger than two, consistently with its oxygen density map. On the other hand, the coordination number of the O_p-cresol–H_water in p-cresol⁻ was identical to that of the O_p-cresol–O_water, whereas that in p-cresol˙ was slightly decreased. This indicates that the H-bonds of p-cresol⁻ are more rigid than those of p-cresol˙. In the p-cresol case, there are several possible H-bond patterns, because p-cresol can become any of the H-bond donor, acceptor, and donor–acceptor forms. The difference in coordination numbers between O_p-cresol–O_water and O_p-cresol–H_water (0.85) suggests the relatively stable H-bond on the donor side of p-cresol and the formation of an H-bond acceptor without donating an H-bond in ∼15% of the simulation. Interestingly, the shape of the O_p-cresol–O_water in the second hydration shell in p-cresol˙ was significantly different from those of p-cresol and p-cresol⁻. The raising of the second hydration shell began from 3.15 Å to the long-distance region, which was a shorter distance than those of p-cresol (3.25 Å) and p-cresol⁻ (3.20 Å). The peak position of the O_p-cresol–O_water of the second hydration shell in p-cresol˙ (4.15 Å) was also a shorter distance than those of p-cresol (4.60 Å) and p-cresol⁻ (4.35 Å).

Table 1 Coordination numbers in the first hydration shell of the oxygen and the hydrogen atoms of water from the phenolic oxygen of each p-cresol model and of the oxygen (nitrogen) and the hydrogen atoms of water and D1-His190 from those of Y_Z and in the S₁, S₂, S₂Y_Z⁺, and S₃ states

	Coordination number		Average of H-bond number
	Ophenol–Owater^a	Ophenol–Hwater^a
a The coordination numbers in the YZ and cases include the nitrogen and the hydrogen atoms of D1-His190 H-bonding with the phenolic oxygen of YZ.
p-Cresol	2.33	1.48	2.29
p-Cresol^−	3.00	3.00	2.99
p-Cresol˙	1.96	1.77	1.66
S1YZ	2.92	1.71	2.71
S2YZ	3.00	1.99	2.98
	2.99	2.15	2.06
S3YZ	3.00	2.00	2.95

---

H-bond network around Y_Z

For PSII, we performed QM/MM-MD simulations around the Mn₄Ca cluster, including Y_Z and its H-bond network in the S₁, S₂, S₂Y_Z⁺ and S₃ states. Fig. 3 shows the evolution along the simulations of the two distances O–H and N–H in the H-bond between Y_Z and D1-His190 (d1 and d2 in Fig. 3) in each state, as well as the number of H-bonds of Y_Z. It should be noted that we collected the structural information in the S₁, S₂, and S₃ states after an equilibration time, whereas the trajectories of S₂Y_Z⁺ started from the equilibrated S₂ structure. In other words, the initial structure (0 ps) of S₂Y_Z⁺ corresponds to the last frame of the S₂ dynamics.

Fig. 3 Time evolution of the number of H-bonds of the phenolic oxygen (red dots) of Y_Z and the distances of O–H (blue lines) and N–H (green lines) in S₁, S₂, S₂Y_Z⁺, and S₃, and the designation of the distance of O–H (shown at the bottom by a blue arrow) and N–H (shown at the bottom by a green arrow).

In the simulation, Y_Z formed two or three H-bonds with D1-His190 and water molecules in the S₁ state (Fig. 3; S₁), whereas Y_Z in the S₂ and S₃ states tended to form three H-bonds (Fig. 3; S₂ and S₃). In order to compare such results with those of p-cresol, the coordination numbers of the first hydration shell of oxygen, nitrogen, and hydrogen atoms of water and D1-His190 from the phenolic oxygen of Y_Z and were also calculated for each simulated S state and are shown in Table 1. The coordination numbers of the O_Yz–H_water in the first hydration shell in Y_Z in the S₂ and S₃ states were 1.99 and 2.00, respectively, which indicated that the water molecules, W4 and W7, formed rigid H-bonds with Y_Z as H-bond donors. In contrast to the S₂ and S₃ states, the coordination number of the O_Yz–H_water in the first hydration shell in Y_Z in the S₁ state (1.71) was lower than those of the S₂ and S₃ states, which was consistent with the partial formation of two H-bonds in the S₁ state (Fig. 3; S₁).

In S₂Y_Z⁺, the formation of two H-bonds was the most stable conformation (Fig. 3; S₂Y_Z⁺), which is consistent with the coordination number of the O_Yz–H_water(H_His190) in the first hydration shell in in the S₂Y_Z⁺ state (2.15) in Table 1. It has to be pointed out that the reason for the same coordination number of O_Yz–O_water(N_His190) shown by S₂Y_Z and is the presence of W4 within the H-bond distance from the phenolic oxygen even after breaking the H-bond between Y_Z and W4. So albeit similar coordination numbers, the two above mentioned states are characterized by a different number of H-bonds, as evident from Fig. 3.

Compared to S₂, the O–H and N–H distances were fully switched around in S₂Y_Z⁺ (Fig. 3; S₂Y_Z⁺). Switching the two distances indicated that the PCET reaction of Y_Z was reproduced well in the S₂Y_Z⁺ model. When Y_Z was oxidized, the proton of the phenolic group was released to D1-His190. The result for the number of H-bonds of in S₂Y_Z⁺ was in good agreement with the simulation of p-cresol in aqueous solution. It was noted that the distances of the O–H and N–H groups in the S₁ state were dependent on the number of H-bonds. The difference between the two distances increased when the number of H-bonds decreased (Fig. 3; S₂). This indicates that the strength of the H-bond between Y_Z and D1-His190 is dependent on the number of H-bonds of the phenolic oxygen of Y_Z.

The distributions of the O–H and N–H distances for the H-bond between Y_Z and D1-His190 in the neutral Y_Z neutral state are also reported in the ESI†, Fig. S2. The distances between the two peaks of O–H and N–H in the S₁ and S₃ states were 0.36 Å and 0.37 Å, respectively, whereas for S₂ the distance was 0.22 Å. This shorter distance in the S₂ state indicates that the H-bond between Y_Z and D1-His190 is stronger. The overlapping region of the histograms of O–H and N–H in the S₂ state was larger than that of the S₁ and S₃ states. The difference in the overlapping region indicates that switching of N–H and O–H bond distances is more frequent in the S₂ state than in the other states. The differences in the number of H-bonds and the proton shuttle between Y_Z and D1-His190 indicate that the H-bond structure in Y_Z does depend on the S-state of the Mn₄Ca cluster.

Structural effect of Y_Z oxidation on the Mn₄Ca cluster, ligands and H-bond network

Decreasing the number of H-bonds of the phenolic oxygen from S₂ to S₂Y_Z⁺ leads to the structural effects of the H-bond structures not only around Y_Z but also around the Mn₄Ca cluster. This is due to the presence of a water network linking between Y_Z and the Mn₄Ca cluster. To determine the effect of the conformational changes of the H-bond network upon Y_Z oxidation, we compared the H-bonds and the ligand structure of the Mn₄Ca cluster. Fig. 4 shows the distribution histograms of the distances of Ca–Mn and Ca–O in the WOC. There was a clear difference in the shape for the distance between Ca²⁺ and W4 H-bonded with Y_Z, in which the peak position increased from 2.33 to 2.43 Å upon Y_Z oxidation. Modifications of the ligand structure of the Ca²⁺ ion were also observed. The histograms of Ca–Glu189 and Ca–Asp170 in S₂Y_Z⁺ were slightly shifted to the shorter distance region, while that of Ca–Ala344 was broader toward the region of longer distance. In the Ca-oxo distances, the Ca–O1 histogram became sharper and shifted slightly to the shorter distance region, while that of Ca–O5 became broader and its peak position was shifted from 2.57 to 2.63 Å in the S₂Y_Z⁺ state. As in the case of the Ca²⁺ ion, the coordination structures of Mn1 and Mn4 were also partially changed upon Y_Z oxidation. The distance distribution histograms between Mn1 and its amino acid ligands in S₂Y_Z⁺ were slightly downshifted to shorter distance although those between Mn1 and oxo-bridges were mostly the same as shown in the ESI†, Fig. S3. In the case of Mn4, the histogram of Mn4–Glu333 in S₂Y_Z⁺ was broader toward the region of longer distance, whereas those of Mn4–O4 and Mn4–O5 were shifted to shorter distance by 0.02 and 0.04 Å, respectively (Fig. S4, ESI†). These modifications show that the ligand coordination of the Mn cluster is slightly modified by Y_Z oxidation. Notably, the histogram of the Ca–Mn4 distance was broadened and its peak position was shifted from 3.80 to 3.88 Å upon Y_Z oxidation (Fig. 4A).

Fig. 4 Distributions of the distance between the Ca²⁺ ion and Mn ions (A), Ca²⁺ and oxygen atoms of carboxylate ligands (B), Ca²⁺ and μ-oxo (C), and Ca²⁺ and water ligands (D) in the S₂ (dotted line), and the S₂Y_Z⁺ (solid line) state.

The H-bond structure around the Mn₄Ca cluster was also modified upon Y_Z oxidation. The conformational changes in the H-bond structure around the manganese cluster are shown in Fig. 5 and 7. The stable distances between O1 and Wb, which are indicated in Fig. 5C, were approximately 4.6 Å and 2.0 Å, respectively, in S₂, whereas the distance was more than 8 Å in S₂Y_Z⁺. This difference was a result of the conformational changes in the H-bond network around Y_Z due to W4 forming an H-bond with Wa instead of Y_Z in S₂Y_Z⁺ (Fig. 7). Due to the formation of an H-bond between W4 and Wa, the O–O distance between W4 and D1-Glu189 decreased significantly from approximately 3.0 to approximately 2.7 Å. The H-bond pattern around CP43-Arg357 and D1-Asp61 also changed upon Y_Z oxidation (Fig. 6). We did not show the information for S₂ in Fig. 6 because the structural properties in S₂ were almost the same as those of S₂Y_Z⁺ before the 5 ps point. In S₂, Wc H-bonded with D1-Asp61 and O4 of the Mn₄Ca cluster as an H-bond donor, while Wd H-bonded with Wc and CP43-Glu354 as an H-bond donor. The S₂Y_Z⁺ simulation showed the rearrangement of this H-bond network upon Y_Z oxidation. In S₂Y_Z⁺, Wd formed an H-bond with D1-Asp61 as soon as the H-bond between Wc and D1-Asp61 broke, and Wc then formed an H-bond with Wd as an H-bond donor (Fig. 7). At the same time, the H-bond structure of CP43-Arg357 was remarkably modified. Wd was located at the H-bond distance from a nitrogen of CP43-Arg357 in S₂, whereas Wc was close to the nitrogen of CP43-Arg357 upon the conformational change at 5 ps in S₂Y_Z⁺. Coupled with the conformational changes of the water network, CP43-Arg357 fully formed an H-bond with the C-terminus of D1-Ala344. The structural changes around CP43-Arg357 might not be directly related to the rearrangement of the H-bond network around Y_Z. The CP43-Arg357 electrostatic-mediated changes seemed to start at different times (5 ps) whereas the H-bond changes already happened (about 2 ps). In addition there is no direct link between the H-bond network and CP43-Arg357.

Fig. 5 Time evolution of the distances between O1 and a hydrogen atom of Wb shown in C (blue arrow) in the S₂ (A) and the S₂Y_Z⁺ (B) state. Distributions of the distances between the oxygen atom of D1-Glu189 coordinated with the Ca²⁺ ion of the Mn₄Ca cluster and that of W3 (green line in D and represented in E as a blue arrow) or W4 (red line in D and represented in E as a red arrow) in S₂ (dotted line) and S₂Y_Z⁺ (solid line).

Fig. 6 Time evolution of the distances in the S₂Y_Z⁺ state between the oxygen atom of D1-Asp61 H-bonded with W1 and hydrogen atoms of water molecules (A), between the nitrogen atom of CP43-Arg357 H-bonded with water molecules and oxygen atoms of water molecules (B), between the hydrogen atom of CP43-Arg357 H-bonded with carboxylate ligands and oxygen atoms of carboxylate ligands coordinated with the Ca²⁺ ion, D1-Ala344, and D1-Asp170 (C). The designation of each distance is shown on the right with arrows, in which the colors correspond to the lines on the left.

Fig. 7 The H-bond structure around Y_Z (top), CP43-Arg357 (middle) and O4 (bottom) of the Mn₄Ca cluster in neutral Y_Z and the radical forms.

Discussion

In the present study, we first determined the H-bond patterns of p-cresol, a tyrosine side chain model, in aqueous solution. The QM-MD simulations of the three different protonation or oxidation states of p-cresol revealed the possible H-bond patterns in aqueous solution. The neutral p-cresol mostly formed two H-bonds (three partially formed H-bonds) with water molecules as an H-bond donor–acceptor (Fig. 2A). The H-bond structure was consistent with the vibrational insight of the p-cresol in H₂O (D₂O) solution.⁶⁵ The conformation and the strength of the H-bonds in the radical form after the PCET reaction (Fig. 2C and Fig. S1, ESI†) were very similar to those of the neutral form due to the similarity of the charge distributions on the phenolic groups in the two states. Thus, in the case of the oxidation of a neutral tyrosine, a phenolic proton was released to an H-bond acceptor without major conformational changes. In contrast to the oxidation of the neutral form, there were significant differences in the H-bond properties of the oxidation of p-cresol⁻ despite similarities in the atomic structure (Fig. 2B and Fig. S1, ESI†). The difference indicated that conformational changes occurred in the H-bond structure of a tyrosine anion upon its oxidation. The differences of the strength and the number of H bonds between the anion and the radical forms were mainly caused by those of the charge distribution of the phenolic group. Remarkably, the structural conformation of the H-bonded water molecules was determined by the phenolic C–O form, which was strongly influenced by the electronic structure of the phenolic ring. The anionic p-cresol⁻ formed a single bond on the phenolic C–O group with three H-bonds arranged as the sp3 conformation (Fig. 2B), whereas the oxidized p-cresol˙ formed a partial double bond on the phenolic C=O group, which could not form three H-bonds with the sp3 conformation (Fig. 2C). The difference of the C–O bond was consistent with the significant upshift of the CO stretching frequency upon the oxidation of anionic tyrosine and p-cresol molecules.⁸⁵

The H-bond structure of Y_D was satisfactorily consistent with the model simulation of p-cresol. The H-bond structure of Y_D as revealed by the X-ray structure^17–19 was very similar to that of neutral p-cresol in aqueous solution. For the neutral Y_D form, both proximal (∼2.8 Å from Y_D) and distal (∼4.3 Å from Y_D) positions were stable in the QM/MM simulations.^37,86 A recent EPR study has indicated that, at equilibrium, a water molecule near Y_D can be located both proximally and distally.⁸⁷ The H-bond pattern of Y_D and the proximal water corresponded to the results of the density map of neutral p-cresol (Fig. 2A). In the Y_D˙ state, the phenolic proton was released into the bulk upon its oxidation, and Y_D then became a neutral radical. The QM/MM simulations also predicted that the water would move to the distal site in Y_D˙ and that Y_D˙ would then only H-bond with D2-His189.^37,86 The water movement was consistent with the decreasing H-bond affinity from neutral p-cresol to oxidized p-cresol˙, which was estimated by decreasing the coordination number and longer distance of the peak position of the radial distribution functions of the O_p-cresol–O_water in the first hydration shell (Table 1). The proximal position was not, however, stable in Y_D and Y_D˙, although the density maps and the radial distribution function showed that neutral and radical forms of p-cresol were able to form two H-bonds. This may be due to D2-Arg180 being located in the H-bond distance from the distal water because the H-bond between the distal water and D2-Arg180 was more stable than that between Y_D˙ and the proximal water.

In contrast to Y_D, a complete agreement regarding the H-bond structure between Y_Z and the neutral p-cresol model in the QM-MD simulation was not obtained. D1-His190 and two water molecules were located within the H-bond distance from the phenolic oxygen of Y_Z in the X-ray structure.^17–19 To understand its specificity and the respective contribution to water oxidation, we performed QM/MM-MD simulations based on our previous models (see Methods). Neutral Y_Z formed three H-bonds for a long time in the simulation (Fig. 3), which was not the same tendency of the neutral p-cresol tending to form two H-bonds in aqueous solution. The coordination numbers of both of the O_Yz–H_water and O_Yz–O_water in the first hydration shell in Y_Z are also higher than that of p-cresol in aqueous solution (Table 1). The comparison between the neutral solvated p-cresol and the neutral Y_Z demonstrates that the protein environment plays a crucial role in triggering the conformational changes upon oxidation. Since the protein hydrogen bond network of neutral Y_Z is not optimal, as found by solvated tyrosine simulations, the structural changes upon oxidation are amplified. One of the keys for the elucidation of the specificity of the number of H-bonds in neutral Y_Z was the formation of the LBHB between Y_Z and D1-His190. Our simulation showed the shuttling of a proton via this short H-bond (Fig. S2, ESI†). This proton shuttling suggests that Y_Z momently formed an anion state due to transient deprotonation. The partial anion strongly contributed to stabilizing three H-bonds with W4, W7, and D1-His190 in the neutral Y_Z state because the H-bond distance between Y_Z and D1-His190 depended on the number of H-bonds (Fig. 3; S₁). The influence of the H-bond of W4 on the LBHB between Y_Z and D1-His190 was consistent with the results of the QM/MM study by Saito et al.³⁶

The present simulation in S₂Y_Z⁺ reproduced not only the proton transfer from Y_Z to D1-His190 upon Y_Z oxidation but also breaking of the H-bond between and W4 (Fig. 7), which was suggested by the QM/MM simulation reported by Noguchi and co-workers.⁵⁸ This H-bond pattern of corresponds to the H-bond property of oxidized p-cresol˙ in aqueous solution. The changing H-bond pattern upon Y_Z oxidation induced direct connection of the water chain from the Ca²⁺ ion of the Mn₄Ca cluster to the bulk surface via the “large channel system”^1–3 (shown in Fig. 1B as channel A) due to the formation of the new H-bond of W4 with Wa (Fig. 7). This causes the breaking of H-bond between Wb and O1 in the Mn₄Ca cluster. A recent study by Sakashita et al.⁸⁸ based on MD simulations showed that this channel has an elevated degree of water accessibility due to its large radius. Thus, the direct connection of the Ca²⁺ ion and this channel upon Y_Z oxidation leads to higher accessibility of bulk water to the Mn₄Ca cluster. Moreover, the structural changes in the H-bond network around Y_Z cause W4 to come in the proximity of D1-Glu189 at the same time that the new H-bond network forms. The changes are indicative of the proton transfer pathway from the O5 site of the Mn₄Ca cluster to the bulk via D1-Glu189. Our previous study showed that D1-Glu189 can be involved in the proton transfer from the O5 site of the Mn₄Ca cluster in the S₃ → S₄ transition.⁸⁹ Our data strongly support the notion that the proton located at the O5 site can be released via D1-Glu189 and W4 coupled with Y_Z reduction (Fig. 8). This means that this mechanism is used in the S₀ → S₁ transition in particular because it promoted the proton release from the O5 site. Since cannot form three H-bonds, the proton release into the bulk has to be coupled with the electron transfer to . In particular, the recent time-resolved infrared measurement showed a fast single phase with a small kinetic isotope effect (H/D) in the S₀ → S₁ transition, which strongly suggested the PCET with electron transfer as the rate-limiting step.⁹⁰ Thus, we propose that the proton release mechanism via D1-Glu189 from the O5 site is effective in such a PCET reaction, and therefore cannot precede the oxidation of the Mn₄Ca cluster by .

Fig. 8 Possible proton transfer mechanism from the site of O5 of the Mn₄Ca cluster to the bulk. In the first step, Y_Z is oxidized by P680⁺ with structural changes in the H-bond network around Y_Z (blue arrows). After this, the proton of O5 is oriented toward the oxygen atom of D1-Glu189 (red arrow). In the electron transfer step from Mn to , a proton between D1-Glu189 and O5 moves to W4 and the proton forms an H-bond with Y_Z with a concerted proton transfer relay analogous to the Grotthuss mechanism from W4 into the bulk via the large channel system (green).

The comparison between the dynamics of different S-states reports that the H-bond structure of Y_Z is also dependent on the S-state of the Mn₄Ca cluster (Fig. S2, ESI†). In particular, the H-bond between Y_Z and W4 was not rigid in the S₁ state, while the proton shuttling on the H-bond between Y_Z and D1-His190 was more frequent in the S₂ state than in the other states. This was due to the different charge distribution on the Mn₄Ca cluster by different oxidation states of Mn ions because the topologies of the S₁ and S₂ models were the same. Changing the H-bond interaction of phenolic oxygen upon Mn oxidation was consistent with the FTIR study, which detected the structural coupling between Y_Z and the Mn₄Ca cluster in the S₁ and S₂ states using isotope labeling.⁹¹ The Mn₄Ca cluster in the S₂ and S₃ states had excess positive charge since the proton release pattern was 0:1:2:1 for the S₁ → S₂, S₂ → S₃, S₃ → S₀, and S₀ → S₁ transitions, respectively.^92–94 The partial anion of neutral Y_Z in the S₂ state was more stable than that in the S₁ state due to the Coulombic interaction between Y_Z and the excess positive charge on the Mn₄Ca cluster. Although the Mn₄Ca cluster in the S₃ state also had excess positive charge, the H-bond structure of Y_Z in the S₃ state was slightly different from that in the S₂ state. This may have been caused by the structural and/or screening effect due to binding an additional hydroxide (i.e. O6) in the S₃ state. The stabilization of the LBHB and the three-H-bond state of Y_Z contributed to the rapid conformational changes of the re-reduction of Y_Z in the S₁ → S₂ and S₂ → S₃ transitions. Thus, the excess positive charge on the Mn₄Ca cluster may play a key role in controlling the H-bond interaction of the phenolic oxygen of Y_Z.

The H-bond network around O4 is a part of the “narrow channel”^1–3 shown in Fig. 1B as channel B although this network and that around Y_Z are divided. In addition, the hydrogen bond distances among the water cluster between Y_Z and the Mn₄Ca cluster and its interatomic distance are virtually unchanged (Tables S1 and S2, ESI†). As a remarkable difference between the S₂ and S₂Y_Z⁺ states, the moiety has a remaining positive charge by its oxidation. Thus, it is reasonable that the hydrogen bond rearrangements around O4 arose from the charge repulsion of the positive charges of CP43-Arg357 and . This interaction mechanism strongly suggests that CP43-Arg357 modified the H-bond structure of the water network around O4, coupled with the redox reaction of Y_Z. The role of CP43-Arg357 is a key issue for the elucidation of the water oxidation mechanism. The CP43-Arg357Lys point mutation has very low oxygen-evolving activity (∼18% of wild-type).⁹⁵ These results mean that CP43-Arg357 not only functioned as a positive charge around the Mn₄Ca cluster but that it also stabilized the structure around the Mn₄Ca cluster through the guanidine group. Our previous study showed that the water network around D1-Asp61 was involved in the delivery of the substrate water in the S₂ → S₃ transition.⁹⁶ The QM/MM study by Saito et al. suggested that the water network around O4 serves as the proton transfer pathway in the S₀ → S₁ transition.⁹⁷ Thus, we propose that CP43-Arg357 plays a key role in controlling the H-bond structure around O4 to facilitate the proton transfer and/or the water movement in the S-state transitions. However, further careful studies are needed to obtain a more detailed understanding of the function of CP43-Arg357 in water oxidation.

Recent X-ray studies gave more insight into the intermediate structures in the water oxidation cycle.^98,99 The simulations here presented reproduced well the metal–metal distances in the Mn₄Ca cluster in the stable S₁, S₂, and S₃ states (Table S2, ESI†). In contrast, the Mn1–Mn4 distances obtained by the simulation of the Y_Z-oxidized S₂ state showed different behavior from the transient state at 150 μs of the S₂ → S₃ transition.⁹⁹ The simulation of the S₂Y_Z⁺ state, that preceded the insertion of a water molecule, provided a virtually identical Mn1–Mn4 distance to that in the S₂ state. On the other hand, the X-ray transient structure showed a longer Mn1–Mn4 distance by ∼0.2 Å than that of the structure after single illumination. This longer distance is maintained in the stable structure after the second flash illumination. This indicates that the significant structural changes in the Mn₄Ca cluster to insert a new water molecule (O6) between Mn1 and Mn4 occur before 150 μs in the S₂ → S₃ transition. This mismatch arose from the different time scales between the X-ray measurement (>100 μs) and the dynamics simulation (∼10 ps).

The hydrogen bond distances among the water molecules around the Mn₄Ca cluster in the recent X-ray structures^17,19,98,99 were also satisfactorily reproduced by the present simulations in the stable S states (Table S3, ESI†). In terms of the S₂ → S₃ transition, the X-ray structure after 150 μs from the second illumination provided a longer O–O distance between Wb and O1 (3.68 Å and 4.00 Å),⁹⁹ which indicated the absence of the hydrogen bond. This appears in good agreement with our simulations for the S₂Y_Z⁺ state. In contrast with the hydrogen bond between Wb and O1, the hydrogen bond between W4 and Wa in the transient structure (3.45 Å and 3.40 Å) is longer than that in the S₂Y_Z⁺ simulation (2.95 Å). This difference suggests that further hydrogen bond rearrangement takes place around Y_Z on the time scale of microseconds.

Previous theoretical studies proposed the formation of Mn1-oxidized form (i.e. closed cubane structure) of the Mn₄Ca cluster in the S₂ → S₃ transition.^{59,76,96,100,101} In this proposed model, the equilibrium between Mn4- and Mn1-oxidized forms in the S₂ states is inverted upon Y_Z oxidation, favoring the Mn1-oxidized structure in a closed cubane fashion. The present study showed that Y_Z oxidation induced modifications of the coordination and the geometry of the Mn₄Ca cluster. The coordinate bonds between Mn1 and amino acid ligands strengthened upon Y_Z oxidation (Fig. S3, ESI†), which can contribute to a stabilization of the Mn1-oxidized S₂ state. On the other hand, the increased distance between Ca²⁺ and Mn4 (Fig. 4 with a broadening of the distribution shown upon Y_Z oxidation in the S₂ state), which indicates the respective weakening of the bond Ca–Mn4 with higher flexibility, may promote the reduction of Mn4, also triggered by the presence of the positive moiety. Remarkably, the Mn4–O5 distance distribution is slightly broadened upon Y_Z oxidation, despite shortening the average value. The observation is consistent with our proposed S₂ → S₃ transition model. Thus, these structural changes upon Y_Z oxidation suggest that radical may act not only as a mere electron acceptor, but can induce conformational changes in the cluster promoting the transition between isomers and can have a functional role in the progress of the catalytic cycle. The formation of the closed cubane structure was not revealed as an intermediate by the recent X-ray study.⁹⁹ Also in the present simulations we did not observe spontaneous shifts of the O5 position from the open to the closed cubane conformation in the S₂Y_Z⁺ state. This is not surprising since the energy barrier (∼10 kcal mol⁻¹) present between the two conformers cannot be overcome in the simulated time.⁵⁹ Thus, the present results do not exclude our previous hypothesis, also corroborated by the evidence that direct water insertion between Mn1 and O5 provides a much higher barrier (∼20 kcal mol⁻¹) than the W2 insertion between Mn4 and O5 (∼8 kcal mol⁻¹).⁹⁶

It has been proposed that the role of Y_Z and D1-His190 in water oxidation involves the release of a proton from substrate water, especially in the S₂ → S₃ transition.^34,54,56 In particular, an FTIR study has indicated that the proton between and protonated D1-His190 is involved in the proton transfer of water oxidation due to the low NH stretching frequency as an extremely broad band.⁵⁸ The possibility of proton transfer via Y_Z in water oxidation cannot be fully excluded because our simulation also reproduced a strong H-bond between and protonated D1-His190 with proton shuttling. Furthermore, a recent theoretical study by Saito et al. proposed the oxidized and deprotonated D1-His190 instead of Y_Z, with the rotation of D1-Asn298.¹⁰² D1-Asn298 is involved in the large channel system and the H-bond network from Y_Z to PsbV-Lys129.¹⁷ All of the mutations of D1-Asn298 that have been tested exhibited very low O₂ evolution activity,¹⁰³ while an FTIR study using the D1-Asn298Ala mutation has shown that the efficiency of the S₂ → S₃ transition was lowered and the S₃ → S₀ transition was inhibited by this mutation.¹⁰⁴ Thus, it seems clear that D1-Asn298 is a key amino acid involved in the mechanism of water oxidation. Therefore, further dynamic simulations with a larger QM region than that of the present study are necessary to elucidate the involvement of Y_Z and its H-bond network.

The data in the present study add further structural information regarding the rearrangement of the H-bond network around the Mn₄Ca cluster upon Y_Z oxidation. Y_Z is able to interact with not only the large channel system but also the narrow channel (Fig. 6). The interaction mechanism of the two channels with Y_Z is, however, completely different. The large channel system directly interacts with Y_Zvia the H-bond network, whereas the narrow channel interacts with Y_Zvia CP43-Arg357 by Coulombic interactions. Even though the length of our ab initio molecular dynamics is restricted to the ps time scale we have clearly observed both the initial movements that are triggered by Y_Z oxidation in the large and narrow water channels.

The effect of on the H-bond network around the Mn₄Ca cluster may be crucial for water oxidation because the proton transfer and the water movements are coupled with or are triggered by the Y_Z redox reaction, especially in the S₂ → S₃ and S₃ → S₀ transitions.^54–57 In the S₂ → S₃ transition, the rearrangement of the H-bond network, including the water movement, occurs before the concerted reduction of Y_Z and the proton transfer.¹⁰⁵ The present data and the mechanism of water movement in the S₂ → S₃ transition, which we previously reported,⁹⁶ strongly support the notion that the rearrangements and the water movement in the S₂ → S₃ transition are triggered by the H-bond rearrangement upon Y_Z oxidation. As with the S₂ → S₃ transition, the first proton transfer in the S₃ → S₀ transition takes place before the reduction of Y_Z.^54,106,107 We previously proposed a proton transfer mechanism whereby D1-Glu189 mediates transfer of a proton from O5–O6 of the Mn₄Ca cluster in the S₃ → S₄ transition.⁸⁹ The present study supports this mechanism because the ligand structure of the Ca²⁺ ion of the Mn₄Ca cluster is modified slightly upon Y_Z oxidation.

Finally it has to be pointed out that the QM/MM-MD simulation of the S₂Y_Z⁺ state reported here captured only beginning of the structural changes upon Y_Z oxidation due to its reduced length (∼10 ps). Nonetheless, such conformational changes may be a trigger for further changes (i.e. proton transfer and water movements), which require longer times (i.e. ns to μs).

Conclusions

In the present study, we performed QM MD simulations of p-cresol as a model of a tyrosine side chain in aqueous solution in different oxidation/protonation states and QM/MM MD simulations of Y_Z in its protein environment for different states of the Kok–Joliot cycle. From the systematic comparison between simulated systems, we assessed the specificity of the H-bond properties of Y_Z in PSII with respect to the tyrosine side chain model. We found that oxidized formed two H-bonds, whereas neutral Y_Z tended to form three H-bonds. The structure of Y_Z with three H-bonds was not in agreement with the model simulations using p-cresol, whereas the H-bond conformation was almost the same as that of the anionic p-cresol⁻ model. This difference was due to the transient formation of the anion state as a result of the LBHB between Y_Z and D1-His190. Furthermore, the H-bond structure of Y_Z was slightly changed by the S state transitions. In particular, the Mn₄Ca cluster of the S₂ state clearly lowered the barrier in the LBHB. The Y_Z simulation also showed that the redox reaction of Y_Z affected the two main channels linking to the bulk and the Mn₄Ca cluster. Our results clearly indicated that Y_Z is actively involved in the water movement in the S₂ → S₃ transition and in the proton transfer along the S₀ → S₁ transition. Last but not least, this study revealed that the proton transfer mechanism can occur in a concerted manner with the re-reduction, only after the radicalization of Y_Z and the respective change in the H-bond network. Based on these results, we proposed a model of structural changes involving H-bond network rearrangements around the Mn₄Ca cluster and the coordination sphere of Y_Z, occurring upon the tyrosine oxidation, and functional towards the progression of the catalytic cycle. Our study corroborates the hypothesis that Y_Z plays a key role in the Kok–Joliot cycle not only as a mere electron acceptor, but also by triggering structural changes required for the progress of the cycle itself.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We acknowledge PRACE infrastructure (project id: Pra16-3574) for computational time. SN was supported by a JSPS Overseas Research Fellowships.

Notes and references

1. S. Vassiliev, T. Zaraiskaya and D. Bruce, Biochim. Biophys. Acta, 2012, 1817, 1671–1678.
2. F. M. Ho and S. Styring, Biochim. Biophys. Acta, 2008, 1777, 140–153.
3. K. Linke and F. M. Ho, Biochim. Biophys. Acta, 2014, 1837, 14–32.
4. J. P. McEvoy and G. W. Brudvig, Chem. Rev., 2006, 106, 4455–4483.
5. N. Cox, D. A. Pantazis, F. Neese and W. Lubitz, Acc. Chem. Res., 2013, 46, 1588–1596.
6. J. Barber, Q. Rev. Biophys., 2016, 49, e14.
7. D. Gust, T. A. Moore and A. L. Moore, Acc. Chem. Res., 2009, 42, 1890–1898.
8. V. Artero and M. Fontecave, Chem. Soc. Rev., 2013, 42, 2338–2356.
9. H. Dau, C. Limberg, T. Reier, M. Risch, S. Roggan and P. Strasser, ChemCatChem, 2010, 2, 724–761.
10. J. Barber, Sustainable Energy Fuels, 2018, 2, 927–935.
11. R. J. Debus, Biochim. Biophys. Acta, 1992, 1102, 269–352.
12. W. Hillier and J. Messinger, in Photosystem II: The Light-Driven Water: Plastoquinone Oxidoreductase, ed. T. J. Wydrzynski and K. Satoh, Springer, Dordrecht, The Netherlands, 2005, pp. 567–608.
13. J. Messinger, T. Noguchi and J. Yano, in Molecular solar fuels, ed. T. J. Wydrzynski and W. Hillier, Royal Society of Chemistry, Cambridge, UK, 2011, ch. 7, pp. 163–207.
14. G. Renger, in Photosynthesis: Plastid Biology, Energy Conversion and Carbon Assimilation, ed. J. J. Eaton-Rye, B. C. Tripathy and T. D. Sharkey, Springer, Dordrecht, The Netherlands, 2012, pp. 359–414.
15. A. Grundmeier and H. Dau, Biochim. Biophys. Acta, 2012, 1817, 88–105.
16. D. J. Vinyard, G. M. Ananyev and G. C. Dismukes, Annu. Rev. Biochem., 2013, 82, 577–606.
17. Y. Umena, K. Kawakami, J. R. Shen and N. Kamiya, Nature, 2011, 473, 55–60.
18. A. Tanaka, Y. Fukushima and N. Kamiya, J. Am. Chem. Soc., 2017, 139, 1718–1721.
19. M. Suga, F. Akita, K. Hirata, G. Ueno, H. Murakami, Y. Nakajima, T. Shimizu, K. Yamashita, M. Yamamoto, H. Ago and J. R. Shen, Nature, 2015, 517, 99–103.
20. I. D. Young, M. Ibrahim, R. Chatterjee, S. Gul, F. Fuller, S. Koroidov, A. S. Brewster, R. Tran, R. Alonso-Mori, T. Kroll, T. Michels-Clark, H. Laksmono, R. G. Sierra, C. A. Stan, R. Hussein, M. Zhang, L. Douthit, M. Kubin, C. de Lichtenberg, P. Long Vo, H. Nilsson, M. H. Cheah, D. Shevela, C. Saracini, M. A. Bean, I. Seuffert, D. Sokaras, T. C. Weng, E. Pastor, C. Weninger, T. Fransson, L. Lassalle, P. Brauer, P. Aller, P. T. Docker, B. Andi, A. M. Orville, J. M. Glownia, S. Nelson, M. Sikorski, D. Zhu, M. S. Hunter, T. J. Lane, A. Aquila, J. E. Koglin, J. Robinson, M. Liang, S. Boutet, A. Y. Lyubimov, M. Uervirojnangkoorn, N. W. Moriarty, D. Liebschner, P. V. Afonine, D. G. Waterman, G. Evans, P. Wernet, H. Dobbek, W. I. Weis, A. T. Brunger, P. H. Zwart, P. D. Adams, A. Zouni, J. Messinger, U. Bergmann, N. K. Sauter, J. Kern, V. K. Yachandra and J. Yano, Nature, 2016, 540, 453–457.
21. J. Kern and G. Renger, Photosynth. Res., 2007, 94, 183–202.
22. R. J. Debus, Biochim. Biophys. Acta, 2015, 1847, 19–34.
23. M. R. Blomberg, T. Borowski, F. Himo, R. Z. Liao and P. E. Siegbahn, Chem. Rev., 2014, 114, 3601–3658.
24. J. Barber, Q. Rev. Biophys., 2003, 36, 71–89.
25. J. Messinger, Phys. Chem. Chem. Phys., 2004, 6(20), 4764–4771.
26. N. Cox, M. Retegan, F. Neese, D. A. Pantazis, A. Boussac and W. Lubitz, Science, 2014, 345, 804–808.
27. A. Guskov, A. Gabdulkhakov, M. Broser, C. Glockner, J. Hellmich, J. Kern, J. Frank, F. Muh, W. Saenger and A. Zouni, ChemPhysChem, 2010, 11, 1160–1171.
28. G. T. Babcock, B. A. Barry, R. J. Debus, C. W. Hoganson, M. Atamian, L. McIntosh, I. Sithole and C. F. Yocum, Biochemistry, 1989, 28, 9557–9565.
29. P. Joliot, G. Barbieri and R. Chabaud, Photochem. Photobiol., 1969, 10, 309–329.
30. B. Kok, B. Forbush and M. McGloin, Photochem. Photobiol., 1970, 11, 457–475.
31. W. T. Dixon and D. Murphy, J. Chem. Soc., Faraday Trans. 2, 1976, 72, 1221–1230.
32. G. Renger, Biochim. Biophys. Acta, 2012, 1817, 1164–1176.
33. B. A. Diner and R. D. Britt, in Photosystem II: The Light-Driven Water: Plastoquinone Oxidoreductase, ed. T. J. Wydrzynski and K. Satoh, Springer, Dordrecht, The Netherlands, 2005, pp. 207–233.
34. S. Styring, J. Sjöholm and F. Mamedov, Biochim. Biophys. Acta, 2012, 1817, 76–87.
35. F. Rappaport, A. Boussac, D. A. Force, J. Peloquin, M. Brynda, M. Sugiura, S. Un, R. D. Britt and B. A. Diner, J. Am. Chem. Soc., 2009, 131, 4425–4433.
36. K. Saito, J. R. Shen, T. Ishida and H. Ishikita, Biochemistry, 2011, 50, 9836–9844.
37. K. Saito, A. W. Rutherford and H. Ishikita, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 7690–7695.
38. S. Nakamura and T. Noguchi, Biochemistry, 2015, 54, 5045–5053.
39. S. Luber, I. Rivalta, Y. Umena, K. Kawakami, J. R. Shen, N. Kamiya, G. W. Brudvig and V. S. Batista, Biochemistry, 2011, 50, 6308–6311.
40. M. Askerka, D. J. Vinyard, J. Wang, G. W. Brudvig and V. S. Batista, Biochemistry, 2015, 54, 1713–1716.
41. J. Wang, M. Askerka, G. W. Brudvig and V. S. Batista, ACS Energy Lett., 2017, 2, 2299–2306.
42. S. Petrie, R. Stranger and R. J. Pace, Phys. Chem. Chem. Phys., 2017, 19, 27682–27693.
43. S. Petrie, R. J. Pace and R. Stranger, Angew. Chem., Int. Ed., 2015, 54, 7120–7124.
44. W. Ames, D. A. Pantazis, V. Krewald, N. Cox, J. Messinger, W. Lubitz and F. Neese, J. Am. Chem. Soc., 2011, 133, 19743–19757.
45. H. Isobe, M. Shoji, S. Yamanaka, Y. Umena, K. Kawakami, N. Kamiya, J. R. Shen and K. Yamaguchi, Dalton Trans., 2012, 41, 13727–13740.
46. D. A. Pantazis, W. Ames, N. Cox, W. Lubitz and F. Neese, Angew. Chem., Int. Ed., 2012, 51, 9935–9940.
47. M. Askerka, J. Wang, D. J. Vinyard, G. W. Brudvig and V. S. Batista, Biochemistry, 2016, 55, 981–984.
48. M. Shoji, H. Isobe, T. Nakajima, Y. Shigeta, M. Suga, F. Akita, J. R. Shen and K. Yamaguchi, Faraday Discuss., 2017, 198, 83–106.
49. H. Isobe, M. Shoji, J. R. Shen and K. Yamaguchi, Inorg. Chem., 2016, 55, 502–511.
50. A. Haddy, Photosynth. Res., 2007, 92, 357–368.
51. K. G. V. Havelius, J. Sjöholm, F. M. Ho, F. Mamedov and S. Styring, Appl. Magn. Reson., 2009, 37, 151–176.
52. M. Chrysina, G. Zahariou, N. Ioannidis and V. Petrouleas, Biochim. Biophys. Acta, 2010, 1797, 487–493.
53. J. Sjoholm, S. Styring, K. G. Havelius and F. M. Ho, Biochemistry, 2012, 51, 2054–2064.
54. A. Klauss, M. Haumann and H. Dau, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 16035–16040.
55. A. Klauss, T. Sikora, B. Suss and H. Dau, Biochim. Biophys. Acta, 2012, 1817, 1196–1207.
56. F. Rappaport, M. Blanchard-Desce and J. Lavergne, Biochim. Biophys. Acta, 1994, 1184, 178–192.
57. P. E. M. Siegbahn, Phys. Chem. Chem. Phys., 2012, 14, 4849–4856.
58. S. Nakamura, R. Nagao, R. Takahashi and T. Noguchi, Biochemistry, 2014, 53, 3131–3144.
59. D. Narzi, D. Bovi and L. Guidoni, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 8723–8728.
60. A. Hinchliffe, Chem. Phys. Lett., 1974, 27, 454–456.
61. P. J. O'Malley, Chem. Phys. Lett., 2000, 325, 69–72.
62. J. Spanget-Larsen, M. Gil, A. Gorski, D. M. Blake, J. Waluk and J. G. Radziszewski, J. Am. Chem. Soc., 2001, 123, 11253–11261.
63. P. J. O'Malley, Biochim. Biophys. Acta, 2002, 1553, 212–217.
64. R. Ramaekers, J. Pajak, M. Rospenk and G. Maes, Spectrochim. Acta, Part A, 2005, 61, 1347–1356.
65. R. Takahashi and T. Noguchi, J. Phys. Chem. B, 2007, 111, 13833–13844.
66. W. J. McDonald and O. Einarsdottir, J. Phys. Chem. A, 2008, 112, 11400–11413.
67. J. Wang, R. M. Wolf, J. W. Caldwell, P. A. Kollman and D. A. Case, J. Comput. Chem., 2004, 25, 1157–1174.
68. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian Inc., Wallingford CT, 2009.
69. D. Case, T. Darden, T. Cheatham III, C. Simmerling, J. Wang, R. Duke, R. Luo, R. Walker, W. Zhang and K. Merz, AMBER12, University of California, San Francisco, USA, 2012.
70. J. VandeVondele, M. Krack, F. Mohamed, M. Parrinello, T. Chassaing and J. Hutter, Comput. Phys. Commun., 2005, 167, 103–128.
71. T. Laino, F. Mohamed, A. Laio and M. Parrinello, J. Chem. Theory Comput., 2005, 1, 1176–1184.
72. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
73. J. VandeVondele and J. Hutter, J. Chem. Phys., 2007, 127, 114105.
74. C. Hartwigsen, S. Goedecker and J. Hutter, Phys. Rev. B: Condens. Matter Mater. Phys., 1998, 58, 3641–3662.
75. S. Goedecker, M. Teter and J. Hutter, Phys. Rev. B: Condens. Matter Mater. Phys., 1996, 54, 1703–1710.
76. D. Bovi, D. Narzi and L. Guidoni, Angew. Chem., Int. Ed., 2013, 52, 11744–11749.
77. D. Narzi, G. Mattioli, D. Bovi and L. Guidoni, Chemistry, 2017, 23, 6969–6973.
78. V. Hornak, R. Abel, A. Okur, B. Strockbine, A. Roitberg and C. Simmerling, Proteins, 2006, 65, 712–725.
79. D. Narzi, E. Coccia, M. Manzoli and L. Guidoni, Biophys. Chem., 2017, 229, 93–98.
80. M. Capone, D. Bovi, D. Narzi and L. Guidoni, Biochemistry, 2015, 54, 6439–6442.
81. D. Bovi, M. Capone, D. Narzi and L. Guidoni, Biochim. Biophys. Acta, 2016, 1857, 1669–1677.
82. D. Bovi, D. Narzi and L. Guidoni, New J. Phys., 2014, 16, 015020.
83. S. Nosé, Mol. Phys., 1984, 52, 255–268.
84. S. Nosé, J. Chem. Phys., 1984, 81, 511–519.
85. C. Berthomieu, C. Boullais, J.-M. Neumann and A. Boussac, Biochim. Biophys. Acta, 1998, 1365, 112–116.
86. A. Sirohiwal, F. Neese and D. A. Pantazis, J. Am. Chem. Soc., 2019, 141, 3217–3231.
87. N. Ahmadova, F. M. Ho, S. Styring and F. Mamedov, Biochim. Biophys. Acta, 2017, 1858, 407–417.
88. N. Sakashita, H. C. Watanabe, T. Ikeda, K. Saito and H. Ishikita, Biochemistry, 2017, 56, 3049–3057.
89. D. Narzi, M. Capone, D. Bovi and L. Guidoni, Chemistry, 2018, 24, 10820–10828.
90. T. Shimizu, M. Sugiura and T. Noguchi, J. Phys. Chem. B, 2018, 122, 9460–9470.
91. T. Noguchi, Y. Inoue and X. S. Tang, Biochemistry, 1997, 36, 14705–14711.
92. C. F. Fowler, Biochim. Biophys. Acta, 1977, 462, 414–421.
93. E. Schlodder and H. T. Witt, J. Biol. Chem., 1999, 274, 30387–30392.
94. H. Suzuki, M. Sugiura and T. Noguchi, J. Am. Chem. Soc., 2009, 131, 7849–7857.
95. H. J. Hwang, P. Dilbeck, R. J. Debus and R. L. Burnap, Biochemistry, 2007, 46, 11987–11997.
96. M. Capone, D. Narzi, D. Bovi and L. Guidoni, J. Phys. Chem. Lett., 2016, 7, 592–596.
97. K. Saito, A. W. Rutherford and H. Ishikita, Nat. Commun., 2015, 6, 8488.
98. M. Suga, F. Akita, M. Sugahara, M. Kubo, Y. Nakajima, T. Nakane, K. Yamashita, Y. Umena, M. Nakabayashi, T. Yamane, T. Nakano, M. Suzuki, T. Masuda, S. Inoue, T. Kimura, T. Nomura, S. Yonekura, L. J. Yu, T. Sakamoto, T. Motomura, J. H. Chen, Y. Kato, T. Noguchi, K. Tono, Y. Joti, T. Kameshima, T. Hatsui, E. Nango, R. Tanaka, H. Naitow, Y. Matsuura, A. Yamashita, M. Yamamoto, O. Nureki, M. Yabashi, T. Ishikawa, S. Iwata and J. R. Shen, Nature, 2017, 543, 131–135.
99. J. Kern, R. Chatterjee, I. D. Young, F. D. Fuller, L. Lassalle, M. Ibrahim, S. Gul, T. Fransson, A. S. Brewster, R. Alonso-Mori, R. Hussein, M. Zhang, L. Douthit, C. de Lichtenberg, M. H. Cheah, D. Shevela, J. Wersig, I. Seuffert, D. Sokaras, E. Pastor, C. Weninger, T. Kroll, R. G. Sierra, P. Aller, A. Butryn, A. M. Orville, M. Liang, A. Batyuk, J. E. Koglin, S. Carbajo, S. Boutet, N. W. Moriarty, J. M. Holton, H. Dobbek, P. D. Adams, U. Bergmann, N. K. Sauter, A. Zouni, J. Messinger, J. Yano and V. K. Yachandra, Nature, 2018, 563, 421–425.
100. M. Shoji, H. Isobe and K. Yamaguchi, Chem. Phys. Lett., 2015, 636, 172–179.
101. I. Ugur, A. W. Rutherford and V. R. Kaila, Biochim. Biophys. Acta, 2016, 1857, 740–748.
102. K. Kawashima, K. Saito and H. Ishikita, Biochemistry, 2018, 57, 4997–5004.
103. H. Kuroda, N. Kodama, X. Y. Sun, S. Ozawa and Y. Takahashi, Plant Cell Physiol., 2014, 55, 1266–1275.
104. R. Nagao, H. Ueoka-Nakanishi and T. Noguchi, J. Biol. Chem., 2017, 292, 20046–20057.
105. H. Sakamoto, T. Shimizu, R. Nagao and T. Noguchi, J. Am. Chem. Soc., 2017, 139, 2022–2029.
106. T. Noguchi, H. Suzuki, M. Tsuno, M. Sugiura and C. Kato, Biochemistry, 2012, 51, 3205–3214.
107. A. Klauss, M. Haumann and H. Dau, J. Phys. Chem. B, 2015, 119, 2677–2689.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9cp04605d

---