Article Images (0)
×![]()
featured_play_listSave article to folder
X
Create new folder
Full Text
PDF
Table of Contents
Abstract
H2O/OH
Abstract
Nature’s water splitting catalyst, an oxygen-bridged tetra-manganese calcium (Mn4O5Ca) complex, sequentially activates two substrate water molecules generating molecular O2. Its reaction cycle is composed of five intermediate (Si) states, where the index i indicates the number of oxidizing equivalents stored by the cofactor. After formation of the S4 state, the product dioxygen is released and the cofactor returns to its lowest oxidation state, S0. Membrane-inlet mass spectrometry measurements suggest that at least one substrate is bound throughout the catalytic cycle, as the rate of Olabeled water incorporation into the product O2 is slow, on a millisecond to second timescale depending on the S state. Here, we demonstrate that the Mn4O5Ca complex poised in the S0 state contains an exchangeable hydroxo bridge. Based on a combination of magnetic multiresonance (EPR) spectroscopies, comparison to biochemical models and theoretical calculations we assign this bridge to O5, the same bridge identified in the S2 state as an exchangeable fully deprotonated oxo bridge [Pérez Navarro et al. Proc. Natl. Acad. Sci. U.S.A. 2013 110, 15561]. This oxygen species is the most probable candidate for the slowly-exchanging substrate water in the S0 state. Additional measurements provide new information on the Mn ions that constitute the catalyst. A structural model for the S0 state is proposed that is consistent with available experimental data and explains the observed evolution of water exchange kinetics in the first three states of the catalytic
1 Nature’s water splitting catalyst, an oxygen-bridged tetra-manganese calcium cofactor (Mn4O5Ca), is found in the transmembrane protein super-complex photosystem II (PSII).1-8 Biological water-splitting chemistry is driven by the reaction center of PSII, a multi-chlorophyll assembly, found buried in the center of the transmembrane region. Visible-light excitation generates a chargeseparated state with the electron donor (P680•+) subsequently acting as the chemical oxidant for the water splitting reaction, successively extracting electrons from the Mn4O5Ca cofactor. 9 P680•+ and the Mn4O5Ca cofactor are coupled via a redox-active tyrosine residue YZ (D1-Tyr161), which acts as a single electron relay.10,11 The action of the cofactor is to accumulate the four oxidizing equivalents needed to drive the water-splitting reaction. As such it moves through a cycle of five distinct redox states that differ by one-electron oxidation, termed the Si states (i = 0–4, Fig. 1A).12 The release of dioxygen is followed by the spontaneous decay of the S4 state back to the lowest (most reduced) S state, S0. Stable regeneration of the S0 state is thought to involve the rebinding of one substrate water based on substrate exchange data, which monitor the uptake of isotope-labeled water into the product O2 molecule by mass spectrometry.13,14 The Mn4O5Ca cofactor, as visualized using synchrotron radiation X-ray diffraction (SR-XRD),1 and recently using femtosecond X-ray free electron laser diffraction (XFEL),2 adopts a distorted chair conformation with a Mn3O4Ca cubane unit forming the base of the chair (Fig. 1B). The fourth, outer Mn is attached to the base of the chair via the oxygen bridges O4 and O5. The new XFEL structure2 displays Mn-Mn distances consistent with extended X-ray absorption fine structure (EXAFS) constraints, which was not seen in SR-XRD structures due to photochemical reduction during data collection.15-19 The short Mn-O distances resolved for the oxygen bridges O1-O4 support assigning these linkages as oxo (O2-) bridges. In contrast, the comparatively longer Mn3/Mn4-O5 distances can be interpreted in terms of O5 representing either an oxo (O2-) or a μ-hydroxo (OH-) bridge. A straightforward protonation assignment, however, cannot be made as the O5 ACS Paragon Plus Environment 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 Figure 1. (A) Light (h)-driven S-state cycle of the OEC indicating oxidation by YZ , proton release and substrate water binding. (B) DFT cluster model of the OEC in the ST = 1/2 conformation of the S2 state including its immediate surroundings. 20 Mn ions are shown in purple, O in red, Ca in yellow, C in light grey, N in blue and H (mostly omitted for clarity) in white. (C) Top: Schematic representation of the inorganic Mn4O5Ca core in the XFEL structure including the Jahn–Teller axes (magenta bars) of Mn1 III and Mn4 III in the proposed S1 state. 2 Bottom: The two conformational isomers of the S2 state Mn4O5Ca core, differing in the connectivity of O5 and the Mn III position. 20 bridge sits along the MnIII Jahn–Teller (JT) axis of Mn4 (and Mn1), see Fig. 1C. It remains a subject of debate as to whether XFEL structures reported for the resting (S1) state of the cofactor 2,21,22 solely represent this state. This is in part because of postcrystallization and (the lack of) pre-flash treatments that supposedly affect S-state synchronization. In case the crystals have been pre-flashed,22 it is however not clear if the kinetics of S-state synchronization23,24 are identical for the partially dehydrated crystal preparations used in crystallographic studies as compared to solution samples, where S-state synchronization has been studied in depth. This must be a concern as solvent water has been implicated in the proton-electron-coupled oxidation of D2Tyr160 (YD). 25 It is thus conceivable that the reported XFEL structures represent an admixture of centers poised in both the S1 and S0 states, i.e. containing an S0 population of ≈25% of centers. In addition, questions have been raised as to whether light atoms in the vicinity of metal atoms can be accurately determined, with the position of O5 in particular being subject to debate.26 This further compounds the difficulty in assigning the protonation state of S1 (and S0) based on structural constraints. The ambiguity with regard to the protonation state of O5 is particularly problematic for establishing the substrate’s interaction with the catalyst in its resting state. O5 is a likely candidate for the first substrate water that binds to the catalyst,27-29 with its assignment as a potential substrate site based on its structural lability: O5 displays fast exchange kinetics with solvent water,27 and has been shown to adopt two ligand binding motifs in the S2 state (Fig. 1C, bottom).20 It has been suggested that the latter property, namely coordination flexibility, may be important for second substrate inclusion and activation of the catalyst.6,30,31 If O5 represents a μ-hydroxo bridge in the S1 state, then it could represent a water molecule in the S0 state as the S0-S1 transition is coupled to proton release.3,32,33 Alternative protonation assignments for O5 would instead require O5 to represent a μ-hydroxo or μoxo bridge in S0, and several experimental and computational studies have considered such protonation patterns.19,34-41 The presence of a μ-hydroxo bridge in S0 has been suggested earlier by EXAFS based on a fitted increased Mn-Mn distance in S0 as compared to S1, however at a comparatively low experimental distance resolution.42,43 In each of these scenarios, catalyst regeneration, following O2 release, involves the binding (and deprotonation) of substrate water for the next reaction cycle. Here we demonstrate that the Mn4O5Ca complex poised in the lowest S state (S0) does indeed contain a μhydroxo (OH-) bridge. With the aid of DFT calculations, new multifrequency/multiresonance EPR data are used to deduce the geometric structure of the S0 state, constrain the local oxidation states of all four Mn ions, the bridging network (connectivity) of the cofactor and its protonation state. This study, in conjunction with published membrane-inlet mass spectrometry data, strongly supports assigning O5 as the first substrate of the biological water reaction. Consequences for product release and catalyst regeneration are discussed. PSII core complex preparations from wild-type Thermosynechococcus elongatus44 and from a mutant in which YD had been replaced by a phenylalanine45 were isolated as described earlier.46,47 PSII preparations were stored at −80 °C or 77 K (liquid N2) until use. All work was conducted in the dark or under dim green light. For procedures used to form the S0 state see Supporting Information section S1. X-band (≈9 GHz) continuous wave (CW) EPR spectra were recorded at 4 K using a Bruker ELEXSYS E500 spectrometer equipped with an Oxford Instruments Ltd. ESR 900 liquid He-flow cryostat and an ITC503 temperature controller. Q-band (34 GHz) pulse EPR and 55Mn-Davies ENDOR experiments were Page 2 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3 performed at 4.8 K on a Bruker ELEXSYS E580 spectrometer equipped with a homebuilt TE011 microwave cavity, 48 a CF935 liquid He cryostat, an ITC503 temperature controller (Oxford Instruments Ltd.) and an ENI 5100L radio frequency (RF) amplifier. W-band (≈94 GHz) measurements were performed at 5 K using a Bruker ELEXSYS E680 EPR spectrometer. All experiments were carried out employing a homebuilt ENDOR microwave cavity,49,50 which comprised a solenoid of Teflon-coated silver wire integrated into a commercial W-band ENDOR probe head (Bruker). To ensure broadband microwave excitation and to minimize distortions, the loaded quality factor QL was lowered to 700 to obtain a microwave frequency bandwidth of 130 MHz. For further details on the EPR experiments, see Supporting Information section S2. Spectra were fitted assuming an effective spin ST = 1/2 ground state (see Supporting Information section S5.2). The basis set that describes the spin manifolds consisting of one electron and n interacting nuclear spins can be built from the product of the eigenstates of the interacting spins: nnii mImImIM 1121 (1) Here, M refers to the electronic magnetic sublevel, ±1/2; I takes the values 5/2 for 55Mn and 17O, and 1 for 14N and 2H; mi takes the values −Ii, 1−Ii, ..., Ii−1, Ii. The spin manifolds can be described by the following spin Hamiltonian: SGBH ˆˆ 0e i iiiiiii IQIIASIBg ˆˆ 0n,n (2) It contains (i) the Zeeman term for the total electron spin, (ii) the hyperfine and (iii) nuclear Zeeman terms for either the metal 55Mn or the ligand 14N, 17O or 2H nuclei and (iv) the nuclear quadrupole interaction (NQI) term for the 14N or 2H nuclei (I = 1). The NQI splitting is not resolved in the EPR, 55Mn-ENDOR and 17O Electronelectron double resonance (ELDOR)-detected NMR (EDNMR) spectra.51 Spectral simulations were performed numerically using MATLAB® (R2010a, The MathWorks, Natick, MA, USA), a vector-based linear algebra package, and the EasySpin toolbox.52 For further information on data processing, details of the simulations and theory, see Supporting Information sections S3, S4 and S5, respectively. 2.4 Density functional theory (DFT) calculations. The computational models consist of 239–240 atoms and are obtained directly from the large set of possible S0 state structures reported by Krewald et al.40. Additional calculations of hyperfine interaction (HFI) and NQI parameters were carried out using established methods,20,28,29,53-60 on models with a spin doublet ground state based on the lowest energy broken-symmetry solution of each structure. These employed the TPSSh functional,61 which is known to perform well for magnetic and spectroscopic properties.54,62-65 All calculations were performed with ORCA66 using the zeroth order regular approximation (ZORA)67,68 for inclusion of scalar relativistic effects along with ZORA-TZVP (Mn, Ca, O, N) and ZORA-SVP (C, H) basis sets.69,70 Tight SCF convergence criteria and integra- tion settings were applied (Grid6 and IntAcc 6.0 in ORCA nomenclature), with more dense radial integration grids for Mn, N and O of 11, 9 and 9, respectively. The calculations used the RIJCOSX approximation with GridX6,71 and the decontracted basis sets of Weigend as auxiliary bases.72 The electrostatic effect of the protein was Page 3 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4 simulated with a conductor-like screening model with a dielectric constant of 8.0.73 The one-center approximation was applied and the spin-orbit coupling was evaluated with the effective potential/complete mean field approach. Picture change effects were taken into account in calculations of EPR properties. In this study we performed all measurements on PSII isolated from thermophilic cyanobacteria, the same type of material used in current crystallographic studies.21 Since all EPR/ENDOR data for the S0 state had thus far been collected on PSII isolated from higher plants (spinach), a full EPR/55MnENDOR characterization was first performed prior to characterizing the exchangeable oxygen ligands of the Mn4O5Ca complex, to ensure the properties of the S0 state complex are the same in both species.34,74-80 These data (Fig. 2) constrain the electronic structure of the cofactor, mapping out the contribution of each Mn ion to the ground spin state. It can be readily observed that these data, particularly the Q-band 55Mn-ENDOR, obtained in cyanobacteria are essentially the same as seen in earlier higher plant studies,34 showing the cofactor in the S0-state to be in an effective total spin ST = 1/2 ground state. Importantly though, owing to the higher intrinsic activity of the cyanobacterial preparations normalized to total protein concentration, orientationally selective ENDOR data could be obtained, further constraining the 55Mn HFI Table 1. Effective G and 55Mn HFI Tensors Ai Used for the Simulations of the EPR and ENDOR Spectra of the Cyanobacterial S0 (Fig. 3) and S2 29 States and the S0 State in Spinach PSII 34.a G Ai / MHz b A1 A2 A3 A4 S0 x 2.003 327 262 221 148 y 1.965 314 217 188 164 z 1.960 377 276 266 232 iso c 1.976 339 252 225 181 S0 34 x 2.009 320 270 190 170 y 1.855 320 270 190 170 z 1.974 400 200 280 240 iso c 1.946 347 247 220 193 S2 29 x 1.989 350 214 214 173 y 1.978 329 195 184 157 z 1.956 321 282 282 251 iso c 1.974 333 230 227 194 a All G and Ai tensors are collinear. b The numbering of the Ai (i = 1–4) tensors is in descending order of the Ai,iso values and does not correspond to the numbering of the Mni atoms. For an assignment of the Ai tensors to the Mni atoms, see Supporting Information Table S2. The isotropic components are the averages of the individual values: Giso = (Gx + Gy + Gz)/3 and Ai,iso = (Ai,x + Ai,y + Ai,z)/3. tensors (Fig. 2C-E, Table 1). The 55Mn HFI is derived from the coupling of the electron spin of the cofactor with the local nuclear spin of each Mn ion. The effective HFI parameters Ai provide a means to accurately describe the exchange coupling topology of the complex, which is to say, they allow an estimation of the magnitude of the magnetic interactions between the adjacent Mn ions of the complex (see Refs. 31,81). The 55Mn HFIs have been characterized in detail for the multiline EPR signal of the S2 state. 34,79,82,83 Earlier work has shown that while these coupling parameters do not in principle correspond to a unique geometric structure, together with constraints derived from X-ray diffraction and EXAFS, it is best described in terms of a single structural motif, an open cubane-type structure.20,29,31,40,55-57 To rationalize the EPR/55Mn-ENDOR data of this S2 state form, the exchange coupling topology needs to fulfill two criteria: it must (i) render the complex low total spin (ST = 1/2), and (ii) ensure all Mn ions contribute equally to the ground electronic state. This latter property is described in terms of spin projection coefficients, ρi, which in this instance are all approximately 1 (with only that of Mn1 being larger, Table S1). These two criteria are best explained by a set of three alternating net antiferromagnetic/ferromagnetic coupling interactions between Mn1III, Mn2IV, Mn3 IV and Mn4 IV (S2 A in Fig. 3).20,29,31,40,56 As the same spin state and similar HFI parameters (Table 1) are observed for the S0 state, a similar coupling topology of predominately antiferromagnetic coupling interactions is therefore expected. There remains though an ambiguity as to the precise individual oxidation states of the four Mn ions. 3.2 The Mn1-His332-imino-N interaction: a local probe for the electronic structure. In addition to 55Mn HFIs, the HFI and NQI of first coordination sphere ligands provides a means to assess the electronic structure of the cofactor (see Ref. 81). Ligand HFIs have the advantage that they provide site information, namely the local oxidation states of the Mn ion to which they are bound. The imino-N of His332 ligated to Mn11,2 is one Page 4 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 5 such example. In the low-spin S2 state, the large HFI associated with the imino-N28,29,84-88 requires Mn1 to carry the largest spin projection coefficient. This means it must represent the Mn of lowest oxidation state, and thus it can be assigned to the +III oxidation level28,29,87,88 as in the S2 state, the cofactor contains only Mn ions poised at +III or +IV level.20,29,31,40,57 A similar approach can be used to characterize the oxidation level of Mn1 in the S0 state. Q-band 14N three-pulse electron spin echo envelope modulation (ESEEM, Figs. 4, Figs. S3-S5) and 14N hyperfine sublevel correlation (HYSCORE, Fig. 5) were used to characterize the imino-N His322 ligand in the S0 state. Fig. 4 shows ESEEM spectra for the S0 state, compared with the S2 29 state. The Fourier-transformed ESEEM spectra from both S states are very similar, containing the same three features: a set of lines centered at frequencies below 2.5 MHz (νβ = νn – |Aiso|/2), single-quantum transitions between 5 and 9 MHz (να = νn + |Aiso|/2) and less intense double-quantum resonances around 15 MHz (ν2α = 2νn + |Aiso|). This requires the 14N His332 hyperfine and quadrupole interactions to be approximately the Figure 4. Q-band three-pulse 14 N-ESEEM light-minus-dark spectra of the imino-N His322 ligand in PSII poised in the S0 state. For comparison, the S2 state signal is also shown, see Ref. 29 . A: time-domain traces; B: corresponding Fourier transforms. Black lines represent baseline-corrected experimental spectra, red lines spin Hamiltonian-based simulations. Fit parameters are listed in Table 2. “sq” and “dq” refer to the position of single and double-quantum transitions, respectively. The full set of field- and -dependent threepulse ESEEM spectra are shown in Figs. S3-S5. Experimental parameters: microwave frequencies: 33.965 GHz (S0), 34.037 GHz (S2); magnetic fields: 1245 mT (S0), 1250 mT (S2); shot repetition times: 0.5 ms (S0), 1 ms (S2); microwave pulse lengths (/2): 16 ns (S0), 12 ns (S2); : 260 ns; ΔT: 48 ns (S0), 100 ns (S2); temperatures: 4.8 K (S0), 5.2 K (S2). ing EPR/ 55 Mn ENDOR spectra (Table 2). The Euler rotation angles [α, β, γ] of the NQI relative to the A tensors are [23, 111, 15] in the S0 and [20, 12, 0] for the S2 29 state simulations. b Aiso is defined as the average of the principal components of the HFI tensor (Table S3). c Adip is defined in terms of T1, T2, and T3 as Adip = (T1 + T2)/2 = −T3/2. T1, T2, and T3 represent the three principal components of the HFI tensors minus Aiso and of the NQI tensors and are labeled such that |T1| ≤ |T2| ≤ |T3|. d The rhombicity is defined by Aη or η = (T1 − T2)/T3, respectively. same in both the S2 and S0 state. Corresponding Q-band HYSCORE data allow the 14N HFI and NQI parameters to be further constrained. In the 2D-HYSCORE surface, the three features that the Q-band ESEEM spectra comprise appear as cross peaks in the (+,+) quadrant (Fig. 5), which spread inwards towards the diagonal as opposed to extending parallel to the frequency axis. Both these features are consistent with the 14N HFI being slightly smaller in S0 than in the S2 state, i.e. the HFI is further away from the “cancellation condition” (see Ref. 29). Fitted spin Hamiltonian parameters for collective simulation of the ESEEM and HYSCORE data are listed in Table 2. The magnitude of the isotropic HFI Aiso and the NQI (|e2Qq/h|) are slightly smaller than in the S2 state, while the anisotropic (dipolar) HFI component Adip is slightly larger as compared to the values seen for the S2 state. Nevertheless, the strong similarity of the hyperfine and quadrupole interactions in both the S0 and S2 states, and comparison to 14N ligands in model complexes89-94 assign Mn1 the same oxidation state (+III) and ligand field (5-coordinate square pyramidal) in both the S0 and S2 states (Supporting Information section S9.2). Importantly, the large Aiso value demonstrates clearly that the Mn1 ion must carry the largest spin projection factor of the complex,89-93 similar to the situation in the S2 state. This means that Mn1 must represent a manganese ion of lowest (or equal lowest) oxidation state of the Mn ions in the cluster, confirming that no MnII ion, which would then exhibit the largest spin projection factor, can be present in the S0 state. With the oxidation state pattern III,III,III,IV it remains unclear where the MnIV position is. Chemical modeling suggests Mn1 and Mn4 have similar redox potentials,20 as either site can stabilize the only MnIII in the S2 state. Thus, the most likely candidate Page 5 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 6 Figure 5. (+,+) quadrants (1, 2 > 0 MHz) of the Fourier-transformed Q-band 14 N-HYSCORE experimental spectra (top) and spin Hamiltonian-based simulations (bottom) of PSII in the S0 state. The varying magnetic-field positions correspond to g = 1.798– 2.032. “sq” and “dq” point out regions of single- and double-quantum transitions, respectively. The optimized simulation parameters are listed in Table 2. Experimental parameters: microwave frequencies: 33.966 GHz (1194 mT, 1245 mT), 33.964 GHz (1298 mT, 1350 mT); magnetic fields: 1194–1350 mT; shot repetition time: 0.5 ms; microwave pulse length (/2): 16 ns; : 260 ns; Δ 96 ns; temperature: 4.8 K. for the MnIV site in S0 is either Mn2 or Mn3 (Fig. 3). As recently demonstrated in our laboratory,27-29,51 W-band EDNMR is the method of choice to characterize interactions of 17O with the Mn tetramer of the OEC. Fig. 6 shows 17O-EDNMR spectra of the single-quantum region for the S0 and the S2 state 29 after H2 17O buffer exchange in the S1 (dark) state. It can be immediately observed that the 17O spectral profile of the S0 state, not present in nonenriched buffer (Fig. S6), is similar in appearance to that of the S2 state, 27-29 albeit slightly broader and somewhat better resolved. Additional 17O resonances can be observed in the double-quantum region (Fig. S6). A full spin Hamiltonian-based analysis of 17O-EDNMR data was not performed as spectra could only be collected and simulated (Fig. 6, Table S4) at the EPR signal maximum due to the fast magnetic relaxation properties of the S0 state (faster than the S2 state). Nevertheless, as also seen for the S2 state, 27 three 17O HFIs are required to reproduce the entire 17O signal envelope including the double-quantum region (Fig. S6): (i) a large coupling of ≈10 MHz, (ii) an intermediate coupling of ≈4 MHz and (iii) matrix couplings of <1 MHz. Likewise, we assign these three couplings to (i) an exchangeable oxygen bridge, (ii) (a) terminal oxygen ligand(s) and (iii) matrix water including water bound to the Ca2+ ion. Since for both the S0 and the low-spin S2 state spectra, 17O exchange occurred in the S1 state, their large similarity consequently implies that the labelled oxygen bridge position is identical in the two states, i.e. O5 (see Discussion). The protonation state of the exchangeable oxygen bridge is less clear, i.e. μ-oxo or μ-hydroxo. We do however stress that this signal is not consistent with a bridging water ligand. In the model system Mn catalase, terminal water bound along the JT axis of MnIII displays only a small 17O HFI (<1 MHz).95 If any of the oxygen bridges in the complex represents a µ-hydroxo, the proton of the bridge should exhibit a comparatively strong electron-nuclear HFI. To observe such species, PSII samples can be resuspended in 2H2O buffer to isolate all exchangeable, solvent-accessible protons, i.e. potential substrate sites. 2H interactions in the S0 state were examined in PSII from higher plants at low frequency (X-band ESEEM), with modeling suggesting a large coupling, which could be consistent with a µ-hydroxo bridge.96-98 In previous experiments on Mn/Fe metallocofactors,99 we have been best able to resolve large 2H interactions by Qband ESEEM/HYSCORE spectroscopy, which partially suppresses matrix water contributions. We note that in addition to protons of potential substrates, all O- and Nbound 1H are expected to quantitatively exchange during the incubation time used (2 hours). For the S2 state, the 1H and 2H hyperfine splittings seen in Davies and Mims ENDOR, respectively, were shown to be too small to originate from a hydroxo bridge.27 Three-pulse ESEEM data of S0 state PSII samples after 2H2O buffer exchange are shown in Fig. 7. Note that the data represent a ratio of raw ESEEM traces collected on Page 6 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 7 Figure 6. W-band EDNMR spectra of H2 17 O-exchanged PSII samples in the S0 (A) and in the S2 29 (B) state. Black traces depict the single-quantum region of background-corrected experimental spectra; superimposed red traces represent spin Hamiltonian-based simulations. Colored lines represent a decomposition of the simulations showing contributions from the individual 14 N and 17 O nuclei. The optimized parameter sets are listed in Table S4. For experimental parameters and double-quantum regions, see Fig. S6. samples in 2H-labeled and non-labeled (no 2H signal) buffer to suppress the background 14N His332 signal. The relevant spectra are collected on the high-field edge of the S0 multiline spectrum. This is because all samples contain a small [Mn(H2O)6] 2+ contamination, which, owing to its slower magnetic relaxation, can strongly perturb our S0 data, as demonstrated in the Supporting Information: In Fig. S7, three-pulse ESEEM spectra measured at two field positions within the S0 multiline signal envelope are shown. The lower field position (1194 mT, g = 2.035) overlaps with the intense component of the hexaquo-Mn2+ signal whereas at the higher field position (1326 mT, g = 1.832), the same as used in Fig. 7, the Mn2+ signal is outside this region. Under conditions optimized to best visu- alize the S0 multiline (ST = 1/2, /2 = 12 ns), the Fourier transforms of ESEEM spectra collected at both these field positions superimpose sharp 2H and broader 14N (His332) signals centered around their respective Larmor frequencies. Importantly though, at the lower field, the 2H peak is twice as intense, suggesting it is representative of both 2H ions of the S0 state and of the background Mn 2+ complex. This can be shown by repeating the same experiment now under conditions optimized to best visualize the hexaquoMn2+ signal (S = 5/2, /2 = 8 ns). The corresponding Fourier transforms resolve now only a sharp 2H signal at the low field position, representative of the water ligands of the Mn2+ complex, with no corresponding peak at the high field position. Hence, ESEEM measurements at the high field position can be considered free of any Mn2+ contribution. Fourier transforms of the three-pulse ESEEM data, from which the 14N signal from the His332 has been removed by taking the ratio of 2H-labeled and non-labeled spectra in the time domain, are shown in Fig. 7B. Closer inspection of these spectra reveals that in addition to the 2H ‘matrix’ peak centered at the 2H Larmor frequency, there is a second spectral feature in the form of shoulders whose appearance is dependent on the τ value used (black arrows). This behaviour can be reproduced by spin Hamiltonian-based simulations using a HFI tensor A = [1.91 0.35 0.57] MHz and NQI of |e2Qq/h| = 0.27 MHz, η = 0.61 (see also Figs. S8, S9, Table S5). This indicates that this feature comes from a more strongly coupled 2H nucleus (A‖( 2H) = 1.91 MHz ≙ A‖( 1H) = 12.4 MHz). The simulation included the matrix component which was parameterized using values typical of water/hydroxo Mn terminal ligands (i.e. hexaquo-Mn2+-like,51 Table S5). The ratio of the two species was 1:3 (strongly coupled to terminal ligands, i.e. W1/W2) at the high field position and, due to the Page 7 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 8 Figure 8. Section of the (+,+) quadrant of (A), (C) the Fourier-transformed Q-band 2 H-HYSCORE spectrum of a 2 H2Oexchanged PSII sample in the S0 state, measured at the highfield edge (g = 1.832) of the corresponding Q-band multiline EPR spectrum and (B), (D) a simulation thereof (1:3 ratio of strongly and weakly coupled 2 H nuclei), once (A), (B) as contour plots and once (C), (D) as 3D surfaces. Panel (E) shows a 3D surface of the simulation of only the three weakly coupled nuclei (see Fig. S10B, C) The full frequency space of the (+,+) quadrant is depicted in Fig. S10A. Experimental parameters: see Fig. S10. additional hexaquo-Mn2+, 1:6 at the low field position (Fig. S8). To confirm that there was a second, strongly coupled 2H species present in our ESEEM data, 2D HYSCORE measurements (Figs. 8A, S10A) were performed at the high-field edge of the S0 multiline spectrum. These data reveal an intense 2H cross peak centered at the 2H Larmor frequency (n = 8.67 MHz) of width (≈1 MHz) consistent with terminal water ligands. Zooming in, a second structure is seen underneath this strong feature, a broadened cross peak of structure consistent with simulations of the three-pulse ESEEM data. The cross peak is both broadened along the diagonal and at 90° to the diagonal suggesting that this 2H nucleus displays a substantial NQI, consistent with the simulations, shown in Fig. 8B. The larger HFI and NQI of this species support assigning it to a µ-hydroxo bridging ligand. |e2Qq/h| = 0.27 MHz inferred from simulation is on the higher end, yet within the limits from the empirical model of Soda and Chiba100,101 for asymmetrically hydrogen-bonded deuterons (0.31 MHz). In the earlier X-band ESEEM experiments, the time-domain traces were simulated employing larger dipolar (and smaller isotropic) HFI constants (|Adip| ≈ 0.8-0.9 MHz, Aiso ≈ 0.3-0.4 MHz) 96-98 than in our simulations (|Adip| ≈ 0.48 MHz, Aiso ≈ 0.94 MHz), however, ignoring 2H NQI terms. Our Q-band ESEEM/HYSCORE data show that a comparatively large NQI is instead crucial to correctly reproduce the line shape and width. We note that the hyperfine splitting, while being distinctly larger than that of the terminal water ligands, is at least two times smaller than that seen for µ-hydroxo ligands in similar, dimeric systems. The larger HFI in these simpler exchange-coupled systems comes about because one of the metal ions carries a large spin projection (ρ ≈ 2).99 As outlined in section 3.2, Mn1, the Mn that ligates His332, carries the largest spin projection, and as such, a protonated bridge that involves Mn1 should have a spectral signature most like the simpler model systems. As the experimental value is lower, it is likely that the protonated bridge does not involve Mn1, but instead is located at the other end of the complex, i.e., it is ligated to Mn3 and/or Mn4 which carry spin projections ρ ≈ 1. This would then assign either O4 or O5 as the location of the µhydroxo bridge. The EPR and double resonance measurements described above require the S0 state to have (i) an electronic ground-state spin of ST = 1/2, (ii) the oxidation states MnIII3Mn IV, (iii) a MnIII ion in an approximately 5-coordinate square-pyramidal ligand field in the Mn1 position, as in the low-spin S2 state, and (iv) a protonated oxo bridge, which could be O4 or O5. The S2 state contains two interconvertible structures, 20 in which all µ-oxo bridges are unprotonated, with H2O in the W1 position and OH− in the W2 position (Fig. 1C, Fig. 9).56 In the S1-S2 transition, one electron is lost, whereas in the S0-S1 transition, one electron and one proton are lost from the catalytic center.32,33 Thus, an S0 state model that could lead to the interconvertible S2 forms must have one additional proton and two more electrons. There are then three possible protonation sites: the µ-oxo bridges O4 and O5 and the terminal OH− ligand W2, all of which were examined recently.40 Only three models were found that exhibit the correct ST = 1/2 ground state at the protonation level that corresponds to the spectroscopically consistent S2 state models: S0-A, S0-B and S0-C, (Fig. 9). S0-A and S0-B have the same protonation pattern as a model proposed by Siegbahn35,36 (protonated O5, W1 = H2O, W2 = OH−), while the protonation pattern of S0-C resembles that proposed by Saito et al.41 in a QM/MM study of the deprotonation pathways during the S0-S1 transition. We note that computational models for the S0 state with a different total number of protons have also been proposed in the literature.37-39 These are not explicitly treated here but have been evaluated previously.40 Only one model with an additional proton compared to models S0A, S0-B and S0-C was predicted to have a ground state of ST = 1/2, albeit it was among the least energetically favorable in that set of isomers. This is model S0-D, in which O5 is a water ligand. Note that in this circumstance, O5 Page 8 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 9 Figure 9. Top: 3D and schematic representations of DFT models for the S0 state. Orientations of JT axes are indicated by magenta bars. Middle: Mn-Mn distances in Å. Bottom: Mn oxidation state distributions and exchange coupling constants in cm −1 (using the −2J convention for the Heisenberg-Dirac-van Vleck Hamiltonian). All models have a spin ST = 1/2 ground state and an ST = 3/2 first excited state. Relative energies are given for the three isomeric forms that carry the same number of protons. cannot represent the large 17O HFI of the exchangeable bridge, which would instead be assigned to another oxygen bridge, e.g. O4. Fig. 9 shows 3D depictions of the Mn4CaO5W2 cores, followed by schemes showing the JT axis orientations, then Mn–Mn distances and finally the computed exchange coupling constants. The four models share the same oxidation state distribution with Mn2 representing the MnIV ion. S0-A and S0-B have the same protonation pattern, but they differ in the direction of the Mn4 JT axis: it is oriented approximately perpendicular to the plane spanned by Mn3, O4 and Mn4 in S0-A, while it lies along the W1-Mn4-O5 vector in S0-B. It is not strictly correct to describe S0-A and S0-B as open/closed cubane isomers in analogy to the S2 state since the Mn1 JT axis leads to long Mn1-O5 distances (>3 Å) in both structures. In S0-C, O4 is protonated and the Mn3 JT axis is oriented along O3Mn3-O4, leading to a different exchange coupling topology (ferromagnetic coupling between Mn2 and Mn3 as opposed to antiferromagnetic coupling in S0-A and S0-B). In S0-D, all JT axes point towards the doubly protonated O5, similar to model S0-B. The isomers S0-A, S0-B and S0C are relatively close in energy, with S0-A being energetically favored (no energetic comparison can be made with S0-D as it is not an isomer). Overall, all models, irrespective of their protonation state, contain three short (2.74–2.96 Å) Mn-Mn distances, and one longer Mn-Mn distance of >3.3 Å, consistent with EXAFS constraints.42,43 While all four models are structurally similar, their magnetic properties differ.40 The 55Mn HFIs for S0-A, S0-B and S0-C are all similar and consistent with experimental data. In contrast, S0-D exhibits too small 55Mn couplings, well outside experimental bounds (up to ≈90 MHz), excluding that O5 is doubly protonated, i.e. a water molecule. The His332 14N HFI differs among the remaining subset of structures (Table 2). While the DFT calculations systematically underestimate the experimental Adip and Aη values, which represent the pronounced HFI tensor anisotropy, the isotropic coupling strength serves as a sensitive probe for the electronic structure of the models. S0-A exhibits the largest Aiso, close to experiment (and to 5.8 MHz as calculated for the low-spin S2 state) 29, but S0-B also shows reasonable agreement. The value calculated for S0-C however is too small, owing to the small spin projection on Mn1 (Table S2) arising from its different exchange-coupling topology. Also the experimental NQI parameters are best matched by those computed for S0-A. To summarize, the calculations presented here favor an S0 structure which contains a singly protonated O5, and disfavor a doubly protonated O5 or a singly protonated O4. The EPR results described above constrain the structure and protonation state of S0. As expected, its geometric structure is similar to the low-spin S2 state, but with the Page 9 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 10 O5 present as a μ-hydroxo bridge. The structure however is more complicated in that it contains three MnIII ions, leading to a number of energetically close Jahn-Teller isomers. This may explain the strong species dependence of S0 state EPR signals and their sensitivity to small molecules,78,102 which suggests that the cofactor can access multiple magnetic states. Nevertheless, while structural details remain ambiguous, we present clear experimental evidence that the cofactor contains an exchangeable oxygen bridge, which appears to be protonated. In the following, while we cannot definitively rule out O4 as the hydroxo bridge, we rationalize the assignment of the exchangeable oxygen bridge to O5, the same as in the S2 state, and, as a consequence, as one of the substrates of the water splitting reaction. 4.1 Assigning O5 as the exchangeable oxygen bridge. In the S2 state, the large 17O HFI was assigned to a single exchangeable oxygen (oxo) bridge, the O5 bridge.2729 This result was based on site perturbation(s) of the cofactor – it was seen that changing the immediate environment around the O5 bridge altered the 17O HFI. In our experiments on both the S2 and S0 states, 16O/17O exchange takes place during incubation in H2 17O buffer in the S1 state prior to flash-induced S-state advancement. Thus, oxygen sites exchangeable in S1, such as O5, should be 17O-labeled both in S2 and S0 state experiments. This is also the case if the exchangeable site represents a substrate, which has then been replenished upon O2 release by a water molecule from its surrounding. Hence, the large 17O HFI of a labeled oxygen bridge observed in the S0 state can be assigned to O5. 4.2 O5 as the µ-hydroxo bridge. The question that then needs to be asked is: if O5 is an exchangeable hydroxo bridge as opposed to an oxo bridge, would we expect its HFI to remain approximately the same? We have partially addressed this question recently in a study of HFI constants of bridging ligands in model systems.95 Model complexes and computational modeling predict that the HFI of an oxygen bridge in exchange-coupled Mn dimers should increase upon protonation. While this is counterintuitive considering that protonation should lead to a lowered covalency of the Mn−µO bond, the larger HFI can be rationalized by an increase of s-orbital character found for the Mn−µO bond and thus of spin-core polarization. In silico, the coupling is expected to increase by a factor of two for simple dimer systems assuming no change occurs in the oxidation states of both Mn ions. Clearly, here we do not see such a large change; the maximum would be a ≈10% increase as compared to S2 state data. We suspect that the oxidation state change of the two Mn ions that ligate this oxygen (i.e. Mn4 and Mn3), resulting in a lowering of covalency of the Mn−µO bonds, could possibly act to outbalance the effect of bridge protonation. 4.3 Assigning O5 as the slowly exchanging substrate Ws in the S0 and S2 states. The 17O signals observable in the S0 state originate either from oxygen species exchangeable in S1 or from a substrate newly bound after O2 formation and release. The fact that no additional 17O interactions are observed in S0 as compared to S2, thus limits the possible candidates for Ws to those 17O species that have been identified in S2 and thus precludes any oxygen bridges other than O5 from representing Ws. Mass spectrometry measurements have shown that the two substrate waters bound to the cofactor exchange at different rates with bulk water,13,14 demonstrating the two sites are not chemically equivalent. One substrate, termed the slowly exchanging water (Ws) exchanges on a seconds timescale, while the second substrate, termed the fast exchanging water (Wf) exchanges on a sub-second timescale. Rates for Ws have been measured in all S states, with rate constants of S0: ≈10 s -, S1: ≈0.02 s -1, S2: ≈2 s -1 in spinach thylakoid membranes at 10 °C.13,14 This requires that Ws is bound in all S states, including S0. The general trend is a slowing of the rate with the increasing net oxidation state of the cofactor. This is as expected because it is the acidity of bound oxygen that governs its exchangeability; if there is a high barrier to protonation of the oxygen ligand, the site is non-exchangeable.103,104 As the oxidation state of Mn changes from +III, to +IV, the acidity of a bridging oxygen ligand will increase dramatically (9-10 pKa units in [Mn2(μ-O)2(bpy)4] n+).104 The protonation of the exchangeable oxygen to yield a bound water molecule prior to exchange with a solvent water molecule is therefore energetically more costly in a MnIV compared with a MnIII ion, slowing the exchange rate. It is clear from the DFT calculations that the acidity of O5 is lower in the S0 state as compared to S2, with calculations favoring O5 being protonated as opposed to O4 or any other oxo bridge. Only the O5 bridge, but none of the exchangeable terminal water ligands W1-W4, changes its protonation state going from S0 to S1/S2 and thus assigning O5 as Ws, the exchange rate of which decreases, readily explains the results presented here. In addition, absolute rates of exchange favor assigning O5 to Ws. In model systems, terminal water ligands (H2O/OH) of Mn in the +III and +IV oxidation state and Ca all exchange with rates on a micro- to nanosecond timescale, much faster than that observed for Ws, but rather consistent with Wf. Historically, bridging oxygen ligands, which must represent oxo ligands in the higher S states (S2, S3), have been less favored as substrates of the reaction because these ligands exchange very slowly in model systems.103-105 This, however, is clearly not the case of the unique O5 bridge, which by virtue of its flexible coordination (Figs. 2C, 10)20,31 has more degrees of freedom and thus displays an enhanced exchange rate as compared to simpler models. Whether this flexibility simply overcomes steric constraints of water access to the O5 bridge or tunes bridge acidity to energetically lower substrate exchange transition states, we cannot say for certain. Although, as O4 is also accessible by solvent via a water channel terminating at Mn4,1,21 we favor an effect on bridge acidity at least partly contributing to the enhanced exchange rate. Our basis for invoking the structural flexibility of Ws/O5 as key to understanding its exchange rate is based on the observation that, while the rate of Ws exchange Page 10 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 11 Figure 10. Mn4O5Ca cluster models including directly bound H2O/OH - ligands in the states S0, S1 and S2 of the reaction cycle, visualizing Mn oxidation state changes, substrate binding, deprotonation and oxygen release events, also considering the results presented in this work. Furthermore, Mn3-Mn4 EXAFS distances 42,43 , as well as exchange rates k of Ws 13,14 , consistent with the assignment of Ws to O5, are shown for the individual states. slows upon oxidation of the cofactor, it is slower in the S1 state as compared to the S2 state. The flexible coordination of O5 may not be critical in the S0 state as it represents a hydroxo ligand. A protonated bridge, unlike a fully deprotonated bridge, should be fast exchanging compared with model systems.103,104 However, a similar mechanism as assumed for the equilibrium between S2 A and S2 B (Fig. 10, left), involving proton transfer between a terminal
H2O/OH
- ligand and O5, could be in effect in the S0 state for the interchange between the two protonation isomers S0-A and S0-B, which differ in energy by 4.5 kcal/mol. It is however noted again that the spin-coupling topology of S0-B leads to Mn spin projection factors (Table S2) that result in larger deviations from experimental 55Mn and His332 14N hyperfine couplings (Table 2) than that of S0A. Upon deprotonation to form the S1 state, the exchange rate of O5 dramatically slows down. While the cofactor does not display redox isomerism in the S1 state, recent experiments indicate the existence of two different S1 state forms,106 which could differ with regards to the JT axes of the two MnIII ions (compare S0 state models S0-A and S0-B in Fig. 9; see also ref. 107). The outcome of this work and of previous work on the S2 state, 20,27-29,55-58,60,81 combined with further information, especially on substrate exchange rates,13,14,108 has allowed a detailed picture of the first half of the catalytic cycle of the OEC to be developed. It is noted that this could only be achieved by the combination of experiments and theoretical model construction to provide essential selection constraints. In this way, EPR spectroscopy and DFT computations together yield a detailed, consistent picture of Mn oxidation states and ligand interactions of the OEC in the S0 and S2 states, the requirement for any mechanistic considerations. Here, the main results for the S0 state comprise (i) the experimental characterization of the Mn1-His332-imino-N interaction, which in combination with EPR/55Mn-ENDOR and DFT modeling enables the assignment of the oxidation states as Mn1IIIMn2IVMn3IIIMn4III and the site of oxygen bridge protonation as O5, as well as (ii) direct detection of an exchangeable oxygen bridge, identified as µO5-H. Its assignment as the first substrate is based on (i) the spectral similarities between S0 and S2 (Fig. 6), excluding any oxygen other than those 17O sites observable in both these states, of which only O5 is bound to both the Ca2+ ion and Mn, as shown for Ws by mass spectrometry. 109 (ii) O5 is the only oxygen ligand being deprotonated during the transition from S0 to S1/S2, consistent with the slowing of the Ws exchange. 13,14 This leads to the following reaction sequence (Fig. 10): (i) During the spontaneous transition from the transient state S4 to S0, the loss of four oxidation equivalents and release of O2 are followed by the uptake of Ws, incorporated at the O5 position as a µ-hydroxo bridge, and release of a proton. (ii) The light-driven transition to S1 proceeds most probably via oxidation of Mn3 III to Mn3IV and release of the proton bound to O5 (see Refs. 3,34,39,108). The proton-coupled electron transfer results in shortening of the Mn4III-Mn3IV distance42,43 and a significant decrease of the Ws exchange rate. (iii) Upon lightinduced oxidation of Mn4III to Mn4IV without release of a proton (see Refs. 3,36,108), the Mn4O5Ca arrives at the structurally flexible20 S2 state, enabling faster Ws exchange. For completing our knowledge of the catalytic cycle, lacking the transitions to S3 and S4, which involve the most im- Page 11 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 12 portant process of O-O bond formation, the next step will be to extend our approach to obtain a more precise picture of the S3 state than currently available, 22,110 including the proposed binding of the late substrate Wf using 17O labeling. PSII sample preparation; experimental details of the pulse EPR measurements; data processing: baseline correction and light-minus-dark subtraction; spectral simulations; theoretical background; multifrequency EPR and 55 Mn ENDOR spectra and simulations of the S2 and S0 states; the S2 state: EPR/ 55 Mn ENDOR simulation parameters, Mn exchange couplings, fine structure interactions and spin projections; electronic structure of the S0 state: spin projections, 55 Mn HFIs and Mn fine structure interactions; the Mn1–His332imino-N interaction: field- and -dependent Q-band 14 Nthree-pulse ESEEM and 14 N-HYSCORE experiments and interpretation of the simulation parameters; W-band ELDOR-detected NMR experiments; Interactions with exchangeable 1 H/ 2 H species: Q-band 2 H-three-pulse ESEEM and 2 H-HYSCORE experiments and simulations; general considerations on the experimental approach. This material is available free of charge via the Internet at http://pubs.acs.org. *E-mail: thomas.lohmiller@cec.mpg.de; nick.cox@anu.edu.au, nicholas.cox@cec.mpg.de. ¶Department of Chemistry, University of Bath, Bath BA2 7AY, United Kingdom The authors declare no competing financial interest. We dedicate this work to Prof. Dr. Karl Wieghardt on the occasion of his 75th birthday. Financial support was provided by The Max-Planck-Gesellschaft, the “Bioénergie” program of the Commissariat { l’Énergie Atomique et aux Énergies Alternatives, the program FRISBI, the EU SOLAR-H2 initiative (FP7 contract 212508) and the Cluster of Excellence RESOLV (EXC 1069) funded by the Deutsche Forschungsgemeinschaft. T.L. was supported by the Federal Ministry of Education and Research of Germany (BMBF) in the framework of the Bio-H2 project (03SF0355C). A.W.R. is supported by the Royal Society (Wolfson Merit Award) and by BBSRC Research Grant BB/K002627/1. N.C. is supported by the Australian Research Council (Future Fellowship FT140100834). (1) Umena, Y.; Kawakami, K.; Shen, J.-R.; Kamiya, N. Nature 2011, 473, 55. (2) Suga, M.; Akita, F.; Hirata, K.; Ueno, G.; Murakami, H.; Nakajima, Y.; Shimizu, T.; Yamashita, K.; Yamamoto, M.; Ago, H.; Shen, J.-R. Nature 2015, 517, 99. (3) McEvoy, J. P.; Brudvig, G. W. Chem. Rev. 2006, 106, 4455. (4) Cox, N.; Pantazis, D. A.; Neese, F.; Lubitz, W. Acc. Chem. Res. 2013, 46, 1588. (5) Yano, J.; Yachandra, V. Chem. Rev. 2014, 114, 4175. (6) Cox, N.; Pantazis, D. A.; Neese, F.; Lubitz, W. Interface Focus 2015, 5. (7) Krewald, V.; Retegan, M.; Pantazis, D. A. In Solar Energy for Fuels; Tüysüz, H., Chan, C. K., Eds.; Springer International Publishing: Cham, 2016, p 23. (8) Pérez-Navarro, M.; Neese, F.; Lubitz, W.; Pantazis, D. A.; Cox, N. Curr. Opin. Chem. Biol. 2016, 31, 113. (9) Joliot, P.; Barbieri, G.; Chabaud, R. Photochem. Photobiol. 1969, 10, 309. (10) Diner, B. A.; Britt, R. D. In Photosystem II: The Light-Driven Water:Plastoquinone Oxidoreductase; Wydrzynski, T., Satoh, K., Eds.; Springer: Dordrecht, 2005; Vol. 1, p 207. (11) Meyer, T. J.; Huynh, M. H. V.; Thorp, H. H. Angew. Chem., Int. Ed. 2007, 46, 5284. (12) Kok, B.; Forbush, B.; McGloin, M. Photochem. Photobiol. 1970, 11. (13) Hillier, W.; Wydrzynski, T. Phys. Chem. Chem. Phys. 2004, 6, 4882. (14) Hillier, W.; Wydrzynski, T. Coord. Chem. Rev. 2008, 252, 306. (15) Luber, S.; Rivalta, I.; Umena, Y.; Kawakami, K.; Shen, J. R.; Kamiya, N.; Brudvig, G. W.; Batista, V. S. Biochemistry 2011, 50, 6308. (16) Grundmeier, A.; Dau, H. Biochim. Biophys. Acta, Bioenerg. 2012, 1817, 88. (17) Galstyan, A.; Robertazzi, A.; Knapp, E. W. J. Am. Chem. Soc. 2012, 134, 7442. (18) Kurashige, Y.; Chan, G. K. L.; Yanai, T. Nat. Chem. 2013, 5, 660. (19) Vogt, L.; Ertem, M. Z.; Pal, R.; Brudvig, G. W.; Batista, V. S. Biochemistry 2015, 54, 820. (20) Pantazis, D.; Ames, W.; Cox, N.; Lubitz, W.; Neese, F. Angew. Chem., Int. Ed. 2012, 51, 9935. (21) Young, I. D.; Ibrahim, M.; Chatterjee, R.; Gul, S.; Fuller, F. D.; Koroidov, S.; Brewster, A. S.; Tran, R.; AlonsoMori, R.; Kroll, T.; Michels-Clark, T.; Laksmono, H.; Sierra, R. G.; Stan, C. A.; Hussein, R.; Zhang, M.; Douthit, L.; Kubin, M.; de Lichtenberg, C.; Pham, L. V.; Nilsson, H.; Cheah, M. H.; Shevela, D.; Saracini, C.; Bean, M. A.; Seuffert, I.; Sokaras, D.; Weng, T. C.; Pastor, E.; Weninger, C.; Fransson, T.; Lassalle, L.; Brauer, P.; Aller, P.; Docker, P. T.; Andi, B.; Orville, A. M.; Glownia, J. M.; Nelson, S.; Sikorski, M.; Zhu, D. L.; Hunter, M. S.; Lane, T. J.; Aquila, A.; Koglin, J. E.; Robinson, J.; Liang, M. N.; Boutet, S.; Lyubimov, A. Y.; Uervirojnangkoorn, M.; Moriarty, N. W.; Liebschner, D.; Afonine, P. V.; Waterman, D. G.; Evans, G.; Wernet, P.; Dobbek, H.; Weis, W. I.; Brunger, A. T.; Zwart, P. H.; Adams, P. D.; Zouni, A.; Messinger, J.; Bergmann, U.; Sauter, N. K.; Kern, J.; Yachandra, V. K.; Yano, J. Nature 2016, 540, 453. (22) Suga, M.; Akita, F.; Sugahara, M.; Kubo, M.; Nakajima, Y.; Nakane, T.; Yamashita, K.; Umena, Y.; Nakabayashi, M.; Yamane, T.; Nakano, T.; Suzuki, M.; Masuda, T.; Inoue, S.; Kimura, T.; Nomura, T.; Yonekura, S.; Yu, L. J.; Sakamoto, T.; Motomura, T.; Chen, J. H.; Kato, Y.; Noguchi, T.; Tono, K.; Joti, Y.; Kameshima, T.; Hatsui, T.; Nango, E.; Tanaka, R.; Naitow, H.; Matsuura, Y.; Yamashita, A.; Yamamoto, M.; Nureki, O.; Yabashi, M.; Ishikawa, T.; Iwata, S.; Shen, J. R. Nature 2017, 543, 131. (23) Styring, S.; Rutherford, A. W. Biochemistry 1987, 26, 2401. Page 12 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 13 (24) Messinger, J.; Renger, G. Biochemistry 1993, 32, 9379. (25) Saito, K.; Rutherford, A. W.; Ishikita, H. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 7690. (26) Askerka, M.; Vinyard, D. J.; Wang, J. M.; Brudvig, G. W.; Batista, V. S. Biochemistry 2015, 54, 1713. (27) Rapatskiy, L.; Cox, N.; Savitsky, A.; Ames, W. M.; Sander, J.; Nowaczyk, M. M.; Rogner, M.; Boussac, A.; Neese, F.; Messinger, J.; Lubitz, W. J. Am. Chem. Soc. 2012, 134, 16619. (28) Pérez Navarro, M.; Ames, W. M.; Nilsson, H.; Lohmiller, T.; Pantazis, D. A.; Rapatskiy, L.; Nowaczyk, M. M.; Neese, F.; Boussac, A.; Messinger, J.; Lubitz, W.; Cox, N. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 15561. (29) Lohmiller, T.; Krewald, V.; Pérez Navarro, M.; Retegan, M.; Rapatskiy, L.; Nowaczyk, M. M.; Boussac, A.; Neese, F.; Lubitz, W.; Pantazis, D. A.; Cox, N. Phys. Chem. Chem. Phys. 2014, 16, 11877. (30) Retegan, M.; Krewald, V.; Mamedov, F.; Neese, F.; Lubitz, W.; Cox, N.; Pantazis, D. A. Chem. Sci. 2016, 7, 72. (31) Krewald, V.; Retegan, M.; Neese, F.; Lubitz, W.; Pantazis, D. A.; Cox, N. Inorg. Chem. 2016, 55, 488. (32) Lavergne, J.; Junge, W. Photosynth. Res. 1993, 38, 279. (33) Schlodder, E.; Witt, H. T. J. Biol. Chem. 1999, 274, 30387. (34) Kulik, L. V.; Epel, B.; Lubitz, W.; Messinger, J. J. Am. Chem. Soc. 2007, 129, 13421. (35) Siegbahn, P. E. M. Acc. Chem. Res. 2009, 42, 1871. (36) Siegbahn, P. E. M. Biochim. Biophys. Acta, Bioenerg. 2013, 1827, 1003. (37) Isobe, H.; Shoji, M.; Yamanaka, S.; Umena, Y.; Kawakami, K.; Kamiya, N.; Shen, J. R.; Yamaguchi, K. Dalton Trans. 2012, 41, 13727. (38) Yamaguchi, K.; Yamanaka, S.; Isobe, H.; Saito, T.; Kanda, K.; Umena, Y.; Kawakami, K.; Shen, J. R.; Kamiya, N.; Okumura, M.; Nakamura, H.; Shoji, M.; Yoshioka, Y. Int. J. Quantum Chem. 2013, 113, 453. (39) Pal, R.; Negre, C. F.; Vogt, L.; Pokhrel, R.; Ertem, M. Z.; Brudvig, G. W.; Batista, V. S. Biochemistry 2013, 52, 7703. (40) Krewald, V.; Retegan, M.; Cox, N.; Messinger, J.; Lubitz, W.; DeBeer, S.; Neese, F.; Pantazis, D. A. Chem. Sci. 2015, 6, 1676. (41) Saito, K.; Rutherford, A. W.; Ishikita, H. Nat. Commun. 2015, 6. (42) Robblee, J. H.; Messinger, J.; Cinco, R. M.; McFarlane, K. L.; Fernandez, C.; Pizarro, S. A.; Sauer, K.; Yachandra, V. K. J. Am. Chem. Soc. 2002, 124, 7459. (43) Haumann, M.; Müller, C.; Liebisch, P.; Iuzzolino, L.; Dittmer, J.; Grabolle, M.; Neisius, T.; Meyer-Klaucke, W.; Dau, H. Biochemistry 2005, 44, 1894. (44) Sugiura, M.; Boussac, A.; Noguchi, T.; Rappaport, F. Biochim. Biophys. Acta, Bioenerg. 2008, 1777, 331. (45) Sugiura, M.; Rappaport, F.; Brettel, K.; Noguchi, T.; Rutherford, A. W.; Boussac, A. Biochemistry 2004, 43, 13549. (46) Boussac, A.; Rappaport, F.; Carrier, P.; Verbavatz, J. M.; Gobin, R.; Kirilovsky, D.; Rutherford, A. W.; Sugiura, M. J. Biol. Chem. 2004, 279, 22809. (47) Ishida, N.; Sugiura, M.; Rappaport, F.; Lai, T. L.; Rutherford, A. W.; Boussac, A. J. Biol. Chem. 2008, 283, 13330. (48) Reijerse, E.; Lendzian, F.; Isaacson, R.; Lubitz, W. J. Magn. Reson. 2012, 214, 237. (49) Savitsky, A.; Grishin, Y.; Rakhmatullin, R.; Reijerse, E.; Lubitz, W. Rev. Sci. Instrum. 2013, 84. (50) Nalepa, A.; Möbius, K.; Lubitz, W.; Savitsky, A. J. Magn. Reson. 2014, 242, 203. (51) Cox, N.; Lubitz, W.; Savitsky, A. Mol. Phys. 2013, 111, 2788. (52) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42. (53) Orio, M.; Pantazis, D. A.; Petrenko, T.; Neese, F. Inorg. Chem. 2009, 48, 7251. (54) Pantazis, D. A.; Orio, M.; Petrenko, T.; Zein, S.; Bill, E.; Lubitz, W.; Messinger, J.; Neese, F. Chem. Eur. J. 2009, 15, 5108. (55) Pantazis, D. A.; Orio, M.; Petrenko, T.; Zein, S.; Lubitz, W.; Messinger, J.; Neese, F. Phys. Chem. Chem. Phys. 2009, 11, 6788. (56) Ames, W.; Pantazis, D. A.; Krewald, V.; Cox, N.; Messinger, J.; Lubitz, W.; Neese, F. J. Am. Chem. Soc. 2011, 133, 19743. (57) Cox, N.; Rapatskiy, L.; Su, J.-H.; Pantazis, D. A.; Sugiura, M.; Kulik, L.; Dorlet, P.; Rutherford, A. W.; Neese, F.; Boussac, A.; Lubitz, W.; Messinger, J. J. Am. Chem. Soc. 2011, 133, 3635. (58) Su, J. H.; Cox, N.; Ames, W.; Pantazis, D. A.; Rapatskiy, L.; Lohmiller, T.; Kulik, L. V.; Dorlet, P.; Rutherford, A. W.; Neese, F.; Boussac, A.; Lubitz, W.; Messinger, J. Biochim. Biophys. Acta, Bioenerg. 2011, 1807, 829. (59) Krewald, V.; Neese, F.; Pantazis, D. A. J. Am. Chem. Soc. 2013, 135, 5726. (60) Retegan, M.; Neese, F.; Pantazis, D. A. J. Chem. Theory Comput. 2013, 9, 3832. (61) Staroverov, V. N.; Scuseria, G. E.; Tao, J.; Perdew, J. P. J. Chem. Phys. 2003, 119, 12129. (62) Kossmann, S.; Kirchner, B.; Neese, F. Mol. Phys. 2007, 105, 2049. (63) Jensen, K. P. Inorg. Chem. 2008, 47, 10357. (64) Römelt, M.; Ye, S. F.; Neese, F. Inorg. Chem. 2009, 48, 784. (65) Pantazis, D. A.; McGrady, J. E.; Maseras, F.; Etienne, M. J. Chem. Theory Comput. 2007, 3, 1329. (66) Neese, F. WIREs Comput. Mol. Sci. 2012, 2, 73. (67) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993, 99, 4597. (68) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1994, 101, 9783. (69) Pantazis, D. A.; Chen, X. Y.; Landis, C. R.; Neese, F. J. Chem. Theory Comput. 2008, 4, 908. (70) Schäfer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829. (71) Neese, F.; Wennmohs, F.; Hansen, A.; Becker, U. Chem. Phys. 2009, 356, 98. (72) Weigend, F. Phys. Chem. Chem. Phys. 2006, 8, 1057. (73) Klamt, A.; Schüürman, D. J. Chem. Soc., Perkin Trans. 1993, 799. (74) Messinger, J.; Nugent, J. H. A.; Evans, M. C. W. Biochemistry 1997, 36, 11055. (75) Åhrling, K. A.; Peterson, S.; Styring, S. Biochemistry 1997, 36, 13148. (76) Messinger, J.; Robblee, J. H.; Yu, W. O.; Sauer, K.; Yachandra, V. K.; Klein, M. P. J. Am. Chem. Soc. 1997, 119, 11349. (77) Åhrling, K. A.; Peterson, S.; Styring, S. Biochemistry 1998, 37, 8115. (78) Deák, Z.; Peterson, S.; Geijer, P.; Åhrling, K. A.; Styring, S. Biochim. Biophys. Acta, Bioenerg. 1999, 1412, 240. Page 13 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 14 (79) Kulik, L. V.; Epel, B.; Lubitz, W.; Messinger, J. J. Am. Chem. Soc. 2005, 127, 2392. (80) Kulik, L. V.; Lubitz, W.; Messinger, J. Biochemistry 2005, 44, 9368. (81) Lohmiller, T.; Ames, W.; Lubitz, W.; Cox, N.; Misra, S. K. Appl. Magn. Reson. 2013, 44, 691. (82) Peloquin, J. M.; Campbell, K. A.; Randall, D. W.; Evanchik, M. A.; Pecoraro, V. L.; Armstrong, W. H.; Britt, R. D. J. Am. Chem. Soc. 2000, 122, 10926. (83) Charlot, M.-F.; Boussac, A.; Blondin, G. Biochim. Biophys. Acta, Bioenerg. 2005, 1708, 120. (84) Debus, R. J.; Campbell, K. A.; Gregor, W.; Li, Z. L.; Burnap, R. L.; Britt, R. D. Biochemistry 2001, 40, 3690. (85) Yeagle, G. J.; Gilchrist, M. L.; McCarrick, R. M.; Britt, R. D. Inorg. Chem. 2008, 47, 1803. (86) Yeagle, G. J.; Gilchrist, M. L.; Walker, L. M.; Debus, R. J.; Britt, R. D. Phil. Trans. R. Soc. B 2008, 363, 1157. (87) Milikisiyants, S.; Chatterjee, R.; Weyers, A.; Meenaghan, A.; Coates, C.; Lakshmi, K. V. J. Phys. Chem. B 2010, 114, 10905. (88) Stich, T. A.; Yeagle, G. J.; Service, R. J.; Debus, R. J.; Britt, R. D. Biochemistry 2011, 50, 7390. (89) Zheng, M.; Khangulov, S. V.; Dismukes, G. C.; Barynin, V. V. Inorg. Chem. 1994, 33, 382. (90) Schäfer, K. O.; Bittl, R.; Zweygart, W.; Lendzian, F.; Haselhorst, G.; Weyhermuller, T.; Wieghardt, K.; Lubitz, W. J. Am. Chem. Soc. 1998, 120, 13104. (91) Sinnecker, S.; Neese, F.; Lubitz, W. J. Biol. Inorg. Chem. 2005, 10, 231. (92) Stich, T. A.; Whittaker, J. W.; Britt, R. D. J. Phys. Chem. B 2010, 114, 14178. (93) Chatterjee, R.; Milikisiyants, S.; Lakshmi, K. V. Phys. Chem. Chem. Phys. 2012, 14, 7090. (94) Sinnecker, S.; Neese, F.; Noodleman, L.; Lubitz, W. J. Am. Chem. Soc. 2004, 126, 2613. (95) Rapatskiy, L.; Ames, W. M.; Perez-Navarro, M.; Savitsky, A.; Griese, J. J.; Weyhermuller, T.; Shafaat, H. S.; Hogbom, M.; Neese, F.; Pantazis, D. A.; Cox, N. J. Phys. Chem. B 2015, 119, 13904. (96) Aznar, C. P.; Britt, R. D. Phil. Trans. R. Soc. B 2002, 357, 1359. (97) Britt, R. D.; Campbell, K. A.; Peloquin, J. M.; Gilchrist, M. L.; Aznar, C. P.; Dicus, M. M.; Robblee, J.; Messinger, J. Biochim. Biophys. Acta, Bioenerg. 2004, 1655, 158. (98) Åhrling, K. A.; Evans, M. C. W.; Nugent, J. H. A.; Ball, R. J.; Pace, R. J. Biochemistry 2006, 45, 7069. (99) Shafaat, H. S.; Griese, J. J.; Pantazis, D. A.; Roos, K.; Andersson, C. S.; Popovic-Bijelic, A.; Graslund, A.; Siegbahn, P. E. M.; Neese, F.; Lubitz, W.; Hogbom, M.; Cox, N. J. Am. Chem. Soc. 2014, 136, 13399. (100) Soda, G.; Chiba, T. J. Chem. Phys. 1969, 50, 439. (101) Soda, G.; Chiba, T. J. Phys. Soc. Jpn. 1969, 26, 249. (102) Boussac, A.; Kuhl, H.; Ghibaudi, E.; Rogner, M.; Rutherford, A. W. Biochemistry 1999, 38, 11942. (103) Tagore, R.; Crabtree, R. H.; Brudvig, G. W. Inorg. Chem. 2007, 46, 2193. (104) Vinyard, D. J.; Khan, S.; Brudvig, G. W. Faraday Discuss. 2015, 185, 37. (105) Tagore, R.; Chen, H. Y.; Crabtree, R. H.; Brudvig, G. W. J. Am. Chem. Soc. 2006, 128, 9457. (106) Boussac, A.; Rutherford, A. W.; Sugiura, M. Biochim. Biophys. Acta, Bioenerg. 2015, 1847, 576. (107) Paul, S.; Neese, F.; Pantazis, D. A. Green Chem. 2017, 19, 2309. (108) Cox, N.; Messinger, J. Biochim. Biophys. Acta, Bioenerg. 2013, 1827, 1020. (109) Hendry, G.; Wydrzynski, T. Biochemistry 2003, 42, 6209. (110) Cox, N.; Retegan, M.; Neese, F.; Pantazis, D. A.; Boussac, A.; Lubitz, W. Science 2014, 345, 804. TOC Figure. Page 14 of 14 ACS Paragon Plus Environment Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Article Images (0)
×![]()
Bioz is the world’s most comprehensive AI search engine for scientific experimentation. The Bioz search engine offers researchers billions of data-driven product, technique, and protocol recommendations. Bioz taps into the latest advances in Artificial Intelligence (AI) – including Natural Language Processing (NLP), Machine Learning (ML) and Generative AI to mine and structure hundreds of millions of pages of complex and unstructured scientific papers. By harnessing the power of AI, Bioz provides researchers with an unprecedented amount of summarized scientific experimentation knowledge right at their fingertips, ultimately helping to speed up drug discovery and the rate of success in finding cures for diseases. Bioz is used by over 15 Million researchers from over 17,000 different universities and companies in 196 countries.
