Nature of S-States in the Oxygen-Evolving Complex Resolved by High-Energy Resolution Fluorescence Detected X-ray Absorption Spectroscopy

Photosystem II, the water splitting enzyme of photosynthesis, utilizes the energy of sunlight to drive the four-electron oxidation of water to dioxygen at the oxygen-evolving complex (OEC). The OEC harbors a Mn4CaO5 cluster that cycles through five oxidation states Si (i = 0–4). The S3 state is the last metastable state before the O2 evolution. Its electronic structure and nature of the S2 → S3 transition are key topics of persisting controversy. Most spectroscopic studies suggest that the S3 state consists of four Mn(IV) ions, compared to the Mn(III)Mn(IV)3 of the S2 state. However, recent crystallographic data have received conflicting interpretations, suggesting either metal- or ligand-based oxidation, the latter leading to an oxyl radical or a peroxo moiety in the S3 state. Herein, we utilize high-energy resolution fluorescence detected (HERFD) X-ray absorption spectroscopy to obtain a highly resolved description of the Mn K pre-edge region for all S-states, paying special attention to use chemically unperturbed S3 state samples. In combination with quantum chemical calculations, we achieve assignment of specific spectroscopic features to geometric and electronic structures for all S-states. These data are used to confidently discriminate between the various suggestions concerning the electronic structure and the nature of oxidation events in all observable catalytic intermediates of the OEC. Our results do not support the presence of either peroxo or oxyl in the active configuration of the S3 state. This establishes Mn-centered storage of oxidative equivalents in all observable catalytic transitions and constrains the onset of the O–O bond formation until after the final light-driven oxidation event.


Quantification of the S-states of Samples Using EPR Spectroscopy
The progression of PS II samples to the next S-transition upon illumination is not perfect, thus quantification of the population of PS II centers of the sample in each S-state is essential.EPR spectroscopy was employed following the protocol of Messinger et al. 1 The multiline signal of S2 state at g ≈ 2 was measured in all samples in order to monitor the evolution of the S2 population after each visible light flash.In Figure S1, representative spectra of a series of samples after 1-5 laser flashes are presented; they are subtraction spectra minus the S1 background.In the S1 state there was no sign of multiline spectrum that would be originated from any residual S2 state caused by the preflash.All spectra were normalized to the PS II content with respect to the gx signal of cytochrome c550 of the extrinsic subunit of PS II at 2300 G.The multiline intensity for each sample was estimated by the average of the peak-to-peak intensity of the lines indicated in Figure S1a.
The intensity of the multiline after 1 flash, i.e. in the S2 state, was assumed to be 100% and the multiline after 2-5 flashes was calculated as a percentage of the intensity of S2.The average of the multiline intensity of all samples after each number of flashes is presented in Table 1 and in Figure S1b as blue circles.In order to calculate the S-state content of samples after each flash a model including misses, double hits, initial S2 population and/or blocking at S2 was used.Different combinations of the above parameters were tested in order to fit the multiline intensity calculated using this model with the experimental; residuals of the experimental and calculated multiline intensity were used as an estimate of the error of the fitting.Some representative trials are presented in Table 1.First, a miss factor of 10% was used; higher miss factors were also tried but increased the error.It is noted that the same miss factor for all S-transitions was used; deviations that may exist are out of the resolving capability of the HERFD experiment.Consideration of double hits did not improve the fitting.Double hits are not expected anyway when laser is used for illumination of the samples.

S4
After 3 and 4 flashes there is 20% of multiline intensity that cannot be explained only by misses (see Table 1 after the first flash.The absense of delay in S-state advancement during the S1→S2 transition confirms that after dark adaptation the samples are synchronized in the S1 state with 100% oxidized YD.The blockage at S2 state is too high.Exogenous electron acceptor has been added but it may not be integrated well in all PS II centers.To the end, the population of each S-state after each laser flash is presented in Table 2.After "0, 1, 2, 3 flashes" there is 100% S1, 94% S2, 74% S3, and 70% S0 respectively.The error of each fitting was calculated as the sum of residuals between calculated and experimental multiline intensity, as in Messinger et al.

Heterogeneity of the S2 State
It was mentioned in the main text that we avoided using glycerol in the samples in order to achieve a homogeneous S3 that has been already characterized by EPR, hyperfine spectroscopy and DFT. 2 This choice has the drawback that a minority of centers of S2 state represents the high spin form.
In order to quantify this minor S2 population of our samples we compared the present S2 with methanol treated S2 (Figure 2).Methanol converts the minor component of S2 (high spin) to the major component (low spin) that is represented by the multiline signal.In the methanol treated sample, an increase in the intensity of the S2 multiline signal is observed by 20%, which implies that the untreated sample contains 20% of the minor high-spin component of S2.Thus, the S2 state of this study represents the low-spin form in 80% of centers.

Heterogeneity of the S3 State
W-band EPR/EDNMR and DFT revealed an open cubane structure of the S3 state with four 6coordinate Mn IV in the thermophilic cyanobacterium T. vestitus; a new water that is not present in S2 is bound to Mn1. 2 The EPR spectrum is shown in green in Figure S2b.Three perturbed configurations of this motif were observed in glycerol treated PS II from cyanobacterium Synechocystis sp by 130 GHz EPR. 3 A form of the S3 state with a Mn ion to represent highly anisotropic hyperfine couplings was observed by W-band EPR/EDNMR in methanol or glycerol treated or Sr-substituted samples from T. vestitus. 4The aforementioned S3 forms represent a total spin of S = 3.A high spin S3 form (S = 6) was observed by Q-band EPR spectroscopy in methanol treated PS II from spinach and also as a major component in untreated PS II, 5 as first predicted by quantum chemical studies. 6 Figure S2b representative spectra of the two configurations of the S3 state in T. vestitus are presented.The first observed high-field EPR signal of the S3 ("narrow") is shown in green.Upon addition of glycerol or methanol in the samples, a "wide" component is observed (marked with the red arrow) together with the "narrow" signal.The populations of each component depend on the percentage of methanol of glycerol added.We used glycerol-free samples that represent only the "narrow" S3 (green in Figure S2b).S3.

Figure S1 .
Figure S1.(a) Representative EPR spectra after 1-5 flashes.All spectra are subtracted minus the dark S1 state in order to eliminate the gy component from cytc550 that interferes with the multiline signal.The peaks used for measuring the multiline intensity after each flash are marked by dashed lines.EPR parameters: modulation frequency: 100 kHz, modulation amplitude: 7.5 G, microwave frequency: 9.6 GHz, microwave power: 8 mW, sweep time: 84 s, sweep field: 5000 G, average of 4 scans, temperature: 10 K. (b) Experimental multiline signal intensity after 1-5 flashes (blue circles) in comparison with the calculated multiline intensity fitted with parameters: miss factor = 0.06 and deficiency at S2 = 0.15.Blue circles are the average of multiline intensity of all samples; the standard deviation is marked by bars.

Figure S2 .Figure S4 .S9 5 .S10 6 .Figure S6 .Figure S7 .
Figure S2.Inhomogeneity of the S2 (a) and S3 (b) states as indicated by X-band and W-band EPR spectroscopy respectively.The green spectra in both figures are the homogeneous S2 and S3 configurations, while the black represent mixtures (see text).In the samples used herein the S2 is inhomogeneous (black in panel a), while the S3 is homogeneous (green in panel b).(a) Comparison of the S2 state in untreated and methanol-treated PS II.The 20% increase of the multiline intensity upon addition of methanol indicates that 20% of the population in the untreated S2 represents a high spin form, not observable by X-band.EPR parameters as in Figure S1.(b) Wband ESE-detected field-swept spectra of the untreated S3 (green), S3 with 10% glycerol and 3% methanol (black).Cytochrome signals are marked with grey lines.The untreated S3 signal (narrow) is marked by a green arrow while the recently revealed configuration of S3 (wide) by a red arrow.Panel S2b adapted from Chrysina et al.PNAS 2019, 116 (34), 16841-16846.Copyright 2019 National Academy of Sciences

8 .Figure S10 .
Figure S10.Fitting of the pre-edge experimental peaks (black) with Voigt curves (red).The individual curves used for the fitting are presented in blue and the residuals in orange.Fitted parameters are listed in TableS3.

Figure S13 .
Figure S13.Calculated XAS spectra of all models for each state shown in different colors compared to experiment (black curves).A 1.1 eV broadening and a constant shift of 36.3 eV have been applied to the calculated spectra.

Figure S14 .
Figure S14.Comparison of experimental S2−S1 data (black line) with the calculated S2−S1 difference spectra for models   , and   , (blue line) and with the calculated S2−S1 difference spectra for model   , and a mixture of 80%   , and 20%   , (red line).The effect of 20%   , admixture on the difference spectra is negligible and both S2−S1 difference spectra reproduce experiment.

Figure S16 .
Figure S16.Assignment of the calculated pre-edge XAS spectrum based on the NTOs associated with the transitions for models    and   , .Assignment of transitions for   , is similar to that of    .The local Mn4 1s → 3d transition at ~6540 eV is more intense in   , .

Figure S17 .Figure S18 .
Figure S17.Assignment of the calculated pre-edge XAS spectrum based on the NTOs associated with the transitions for model    .

Figure S19 .
Figure S19.Assignment of the calculated pre-edge XAS spectrum based on the NTOs associated with the transitions for model   . .

Figure S20 .
Figure S20.Assignment of the calculated pre-edge XAS spectrum based on the NTOs associated with the transitions for model    .

Figure S21 .S26 12 .
Figure S21.Assignment of the calculated pre-edge XAS spectrum based on the NTOs associated with the transitions for model    .
). Consideration of a population of PS II centers blocked at S2 state that do not progress further upon illumination improved the fitting.In these centers the acceptor side of PS II is

Table S1 .
Calculated multiline intensity in comparison with the experimental intensity of the multiline signal of the S2 state using different combinations of parameters.All parameters are given as percentages (%).

Table S2 .
Calculated S-state content after a given number of flashes.