Rapid Report pubs.acs.org/biochemistry
Reorganization of Substrate Waters between the Closed and Open Cubane Conformers during the S2 to S3 Transition in the Oxygen Evolving Complex Matteo Capone,† Daniele Bovi,‡ Daniele Narzi,‡ and Leonardo Guidoni*,‡ †
Dipartimento di Chimica, Sapienza Università di Roma, p.le A. Moro 5, 00185 Roma, Italy Dipartimento di Scienze Fisiche e Chimiche, Università degli studi dell’Aquila, Via Vetoio 2, Coppito, 67100 L’Aquila, Italy
‡
S Supporting Information *
ABSTRACT: A crucial step in the mechanism for oxygen evolution in the Photosystem II complex resides in the transition from the S2 state to the S3 state of Kok−Joliot’s cycle, in which an additional water molecule binds to the cluster. On the basis of computational chemistry calculations on Photosystem II models, we propose a reorganization mechanism involving a hydroxyl (W2) and a μ2-oxo bridge (O5) that is able to link the closed cubane S2B intermediate conformer to the S3 open cubane structure. This mechanism can reconcile the apparent conflict between recently reported water exchange and electron paramagnetic resonance experiments, and theoretical studies.
W
ater splitting in photosynthetic organisms occurs in the Mn4CaO5 core of the oxygen evolving complex in the integral membrane complex Photosystem II,1 which may serve as inspiration for the design of environmentally friendly materials for artificial photosynthesis and solar fuel production.2−4 The four oxidizing equivalents necessary for the water splitting reaction subsequently accumulate on the Mn4CaO5 cluster through five (S0−S4) subsequent oxidation states known as Kok−Joliot’s cycle.5 In recent years, using X-ray crystallography and extended X-ray absorption fine structure (EXAFS) experiments, atomic details of Photosystem II and in particular of the Mn4CaO5 cluster have been revealed at increasing levels of resolution.6−10 Albeit a radiation damage-free structure of Photosystem II in the S1 state is now available10 at a resolution of 1.95 Å, the exact positions of the Mn4CaO5 cluster atoms are still a matter of debate.11 The best characterized state of the catalytic cycle, from both the experimental and theoretical points of view, is undoubtedly the S2 state. In particular, between the S2 and S3 states, a large amount of spectroscopic data is available to identify possible intermediate steps and the relative catalytic mechanism. Two distinct signals characterizing the S2 state have been detected by electron paramagnetic resonance (EPR) spectroscopy12−14 and assigned to two interconvertible conformers by quantum chemistry calculations.15,16 The two conformers, sketched in the top panel of Figure 1, differ in both the geometric and magnetic points of view, one being an open cubane structure with a low-spin ground state (namely SA2 state, associated with a multiline EPR signal) and the other a closed cubane structure with a high-spin © XXXX American Chemical Society
Figure 1. Open cubane (A on left) and closed cubane (B on right) proposed models for the S2 to S3 transition along Kok−Joliot’s cycle (top panel). Quantum mechanical model (bottom) of the Mn4CaO5 cluster in Photosystem II considered in our density functional theory plus Hubbard correction calculations. The model represents the SB3 state.
ground state (namely SB2 state, associated with an EPR signal at g = 4.1).15 Recent experimental data provide evidence that the latter conformer should be an intermediate state toward the formation of the S3 state,17,18 although another study has previously proposed the opposite hypothesis.19 Quantum Received: July 14, 2015 Revised: October 3, 2015
A
DOI: 10.1021/acs.biochem.5b00782 Biochemistry XXXX, XXX, XXX−XXX
Biochemistry
Rapid Report
starting and ending points of our path, respectively. To describe the structural rearrangements between these two conformers, we have used the minimum energy path (MEP) approach based on the nudged elastic band method as implemented in the CP2K package.33,34 We applied the improved tangent nudged elastic band (IT-NEB)35 on a set of 32 replicas initially obtained by linear interpolation between the closed SB3 and open SA3 cubane optimized conformers. As a starting point of our investigation, we consider the equilibrium between the open cubane SA2 and the closed cubane SB2 states. Once Tyr-Z is oxidized (i.e., in the S2+ state), the radical tyrosyl group is re-reduced by the Mn4 ion, leading to an electronic configuration with all four manganese ions in an oxidation state of IV20 consistent with a S3 state.36 Under these conditions, the equilibrium between open and closed conformers is inverted, favoring the SB2 closed cubane structure.20 Because between the S2 and S3 states one water is binding and one proton is released from the cluster,23 we added a hydroxyl group to Mn4 (see the right sketch in the top panel of Figure 1). This model will hereafter be termed SB3 and represents the starting point of the pathway that is leading to the open cubane SA3 conformer, built consistently to that proposed by EPR spectroscopy14 and theoretical calculations37 (see the right sketch in the top panel of Figure 1). Our results indicate that the minimum energy path between these two conformers is achieved through an exchange between W2 and O5 as the ligand to Mn3, assisted by a proton transfer (Figure 2). In the first part of the reaction pathway, the Mn3− OW2 distance decreases whereas the orientation of the OW2-H hydroxyl group changes forming a hydrogen bond with the O5 atom. Interestingly, near the transition state, the undercoordinated Mn3 lowers its oxidation state from IV to about III as shown by the spin population reported in the bottom panel of Figure 2. At the same time, Mn1 acquires a partial
mechanics/molecular mechanics (QM/MM) calculations have also provided additional information about the energetics and the electronic structure of these two states, indicating that the open cubane structure SA2 is more stable in the S2 state, whereas the closed cubane SB2 conformer can facilitate the electron transfer toward Tyr-Z once its tyrosyl group is oxidized by light reaching its radical state S2+.20 The key role of the two intermediate states is also supported by evidence that between S2 and S3 an additional water molecule binds to the cluster, likely favored by the pentacoordinated, water-exposed Mn4 conformation in the closed cubane structure.17,21,22 At variance with the necessity of a closed cubane SB2 intermediate to reach the S3 state, recent EPR investigations conclude that the S3 state has an open cubane geometry21 with all manganese atoms in a formal oxidation state of IV, sketched as SA3 in the top panel of Figure 1. In summary, a puzzling picture is emerging for what concerns the interconversion between the open and closed conformers along the pathway from S2 to S3 states.23 In particular, as already pointed out by Boussac et al.,18 it is not clear if the passage between the SB2 closed cubane conformer and the SA3 open cubane can be supported or excluded from the thermodynamical and kinetic point of view. It has to be pointed out that still different hypotheses about the nature of the substrate oxygen exist 24 based on controversial experimental and theoretical studies.25−27 In particular, on the basis of isotopic substrate water exchange measurements, it was suggested that μ-oxo species are unlikely to represent the exchange of the slow site in Photosystem II (PSII) and therefore the substrate oxygen.25,26 Other studies21,28,29 suggest a different framework, identifying O5 as one of the substrates. The model reported here is based on this latter assumption,21,28,29 although it is not consistent with the previously mentioned data.24−27 In the present work, on the basis of quantum chemistry calculations, we report a mechanism that can reconcile the apparently contradictory experimental and computational data, providing a full coherent pathway of the geometrical changes occurring along this sector of Kok−Joliot’s cycle. The gas phase models used in our calculations (see Figure 1) were built starting from the recent crystal structure,10 using a protocol similar to that used in our previous works.20,30 It has to be pointed out that the X-ray structure is representative of the S1 state while in the work presented here we are modeling the S2 to S3 passage. In this regard, the position of the O5 atom found in the crystallographic structure may significantly differ from the position in the following steps of Kok−Joliot’s cycle, also because of a possible different protonation state. The hexacoordination of all manganese ions in the S3 state has been reached by adding a hydroxyl group to Mn4 in the closed cubane conformer and to Mn1 in the open cubane conformer. The quantum mechanical models include 224 atoms (see Figure 1): the Mn4CaO5 cluster, the side chains of its ligands (Asp170, Glu189, His332, Glu333, Ala344, Asp342, and CP43Glu354) and the nearest residues (Asp61, Tyr161, His190, CP43-Arg357, His337, Ile60, and Ser169), the five water molecules (or hydroxyl group) coordinated to the Ca2+ and the Mn4 ion, the Cl− ion and its ligand (Asn181), and 12 additional crystallographic water molecules. The quantum chemistry calculations were performed with density functional theory plus Hubbard correction31 using PBE32 as exchange-correlation functional and following our previous setup.30 The SB3 (closed) and SA3 (open) structures were initially optimized and used as
Figure 2. Minimum energy path (top) of the OW2/O5 reorganization along the passage between the SB3 closed cubane and the SA3 open cubane states of Kok−Joliot’s cycle. Mulliken spin population (bottom) on the Mn and O centers along the reaction coordinate. B
DOI: 10.1021/acs.biochem.5b00782 Biochemistry XXXX, XXX, XXX−XXX
Biochemistry
Rapid Report
radical character, decreasing its spin population to a minimum value of 2.7 units in the proximity of the transition state. The energy profile also indicated that the proton transfer between OW2 and O5 should succeed the energy maximum along the path. The comparison with the spin density profiles also shows that such passage nicely correlates with the maximum radical character of O5 (pink line in the bottom panel of Figure 2). In principle, it might be possible that a second small transition state is hidden in the energy plateau preceding such proton transfer. The refinement of such a hypothetical transition state would necessitate additional and more detailed studies. Moreover, its determination might not be of physical interest because the depth of the well and the height of the barrier would be anyway rather small. It is also worth mentioning that near the transition state the proton on His337 is almost shared between the O3 atom and the imidazole nitrogen, thus indicating that these two conformers are nearly isoenergetic. In a previous study,20 we found that the oxidation of the Mn4 ion by the tyrosyl-Z in the SB+ 2 state is modulated by the concomitant proton exchange occurring between the Asp61 and W1 and between the Tyr-Z and His190. We therefore also monitor the protonation state of such residues for all the replicas considered in the MEP calculation, finding that Asp61 always maintains its deprotonated state whereas the Tyr-Z its protonated state. In summary, beyond the local structural rearrangements around the O5 ligands, no other major changes in the electronic structure and proton distribution occur during the transition between the SA3 and SB3 states. According to the energy profile, we can estimate the kinetic barrier between the SB3 and SA3 conformers (ΔE⧧) to be ≈12 kcal/mol. This barrier is similar to that evaluated for the SA2 to SB2 interconversion16 and corresponds to microsecond kinetics, which is compatible with experimental data.38 From the thermodynamic point of view, the SA3 state is largely stabilized over the SB3 state by ΔE ≈ 15 kcal/mol, similar to what was proposed by previous calculations.21,37 Our results clearly show the possibility of a transition from the S 2B closed cubane intermediate to the experimentally determined SA3 open cubane structure, thus reconciling the apparently contradictory experimental and computational data. In the proposed mechanism, a concerted switch between OW2 and O5 is possible on a time scale that is compatible with the experimental data describing the S2 to S3 transition. It is interesting to point out that within this process O5 and OW2 remain good candidates for the formation of the O−O bond, because the S3 state is reached without assuming the arrival of an additional water molecule directly binding to Mn1 from the cavity, as proposed previously.39 Our proposed mechanism is supported by isotopic labeling experiments40 on water exchange, which indicates that the two substrate water molecules are already prebound in the S2 state, therefore likely being W2 and O5. Nevertheless, other isotopic labeling experiments suggest an opposite interpretation.41 The overall mechanism for the S2 to S3 transition emerging from both experimental data and theory is summarized in Figure 3. The Mn4CaO5 cluster undergoes a sort of oscillating pathway between the closed and open cubane conformation, and some reasons for such a nonlinear route can be identified. First, in the SB2 closed cubane conformer, the pentacoordinated Mn4 is more inclined to bind an additional water molecule with respect to the pentacoordinated Mn1 in an SA2 open cubane conformer. In the latter, the binding of an additional water is indeed hindered by the hydrophobic pocket that originated from
Figure 3. Oscillating mechanism between open and closed conformers during the S2 to S 3 transition in Kok−Joliot’s cycle. The thermodynamic equilibrium between the two conformers is inverted two times along the pathway triggered by Mn4CaO5 reduction and water binding.
Val185, whereas Mn4 is easily accessible to several different water molecules. Interestingly, along the reaction pathway, the equilibrium between the open and closed S2 conformers is first inverted upon Tyr-Z oxidation (transition to S2+),20 and subsequently inverted again in the S3 conformers after the binding of the additional water molecule. This double inversion of the thermodynamic stability between conformers A and B is therefore triggered in the first case by the incoming electron hole from Tyr-Z, which favors the closed conformer by a few kilocalories per mole.20 When the water binds and the proton is released, the thermodynamic stability is again inverted, passing through the transition state described in the work presented here, stabilizing the open conformer at the price of tens of kilocalories per mole. Each of these steps can be viewed as an event occurring in the catalytic cluster neighboring that is guiding the reaction toward the narrow funnel of water splitting, preventing back reactions through the tuning of the thermodynamic equilibrium between the open and closed cubane forms. A crucial structural feature in this mechanism relies on the specific distance between Mn1 and Mn4, which in turn generates the special site for O5, because of which the oscillating mechanism between the open and closed cubane structures is possible. This catalytic fingerprint requires that Mn4 be kept at a specific distance by its ligands and might be a unique feature that characterizes the Mn4CaO5 cluster of Photosystem II compared to analogous biomimetic structural models42 and inorganic Mn-based water-splitting catalysts.43,44 C
DOI: 10.1021/acs.biochem.5b00782 Biochemistry XXXX, XXX, XXX−XXX
Biochemistry
■
Rapid Report
(20) Narzi, D., Bovi, D., and Guidoni, L. (2014) Proc. Natl. Acad. Sci. U. S. A. 111, 8723−8728. (21) Cox, N., Retegan, M., Neese, F., Pantazis, D. A., Boussac, A., and Lubitz, W. (2014) Science 345, 804−808. (22) Askerka, M., Vinyard, D. J., Brudvig, G. W., and Batista, V. S. (2015) Biochemistry 54, 5783−5786. (23) Cox, N., Pantazis, D. A., Neese, F., and Lubitz, W. (2013) Acc. Chem. Res. 46, 1588−1596. (24) Shen, J.-R. (2015) Annu. Rev. Plant Biol. 66, 23−48. (25) Hillier, W., and Wydrzynski, T. (2008) Coord. Chem. Rev. 252, 306−317. (26) McConnell, I. L., Grigoryants, V. M., Scholes, C. P., Myers, W. K., Chen, P.-Y., Whittaker, J. W., and Brudvig, G. W. (2012) J. Am. Chem. Soc. 134, 1504−1512. (27) Siegbahn, P. E. (2013) J. Am. Chem. Soc. 135, 9442−9449. (28) Perez Navarro, M. P., Ames, W. M., Nilsson, H., Lohmiller, T., Pantazis, D. A., Rapatskiy, L., Nowaczyk, M. M., Neese, F., Boussac, A., Messinger, J., Lubitz, W., and Cox, N. (2013) Proc. Natl. Acad. Sci. U. S. A. 110, 15561−15566. (29) Rapatskiy, L., Cox, N., Savitsky, A., Ames, W. M., Sander, J., Nowaczyk, M. M., Rögner, M., Boussac, A., Neese, F., Messinger, J., and Lubitz, W. (2012) J. Am. Chem. Soc. 134, 16619−16634. (30) Bovi, D., Narzi, D., and Guidoni, L. (2014) New J. Phys. 16, 015020. (31) Dudarev, S. L., Botton, G. A., Savrasov, S. Y., Humphreys, C. J., and Sutton, A. P. (1998) Phys. Rev. B: Condens. Matter Mater. Phys. 57, 1505−1509. (32) Perdew, J. P., Burke, K., and Ernzerhof, M. (1996) Phys. Rev. Lett. 77, 3865−3868. (33) VandeVondele, J., Krack, M., Mohamed, F., Parrinello, M., Chassaing, T., and Hutter, J. (2005) Comput. Phys. Commun. 167, 103−128. (34) Hutter, J., Iannuzzi, M., Schiffmann, F., and VandeVondele, J. (2014) WIREs Comput. Mol. Sci. 4, 15−25. (35) Henkelman, G., and Jónsson, H. (2000) J. Chem. Phys. 113, 9978−9985. (36) Krewald, V., Retegan, M., Cox, N., Messinger, J., Lubitz, W., DeBeer, S., Neese, F., and Pantazis, D. A. (2015) Chem. Sci. 6, 1676− 1695. (37) Siegbahn, P. E. M. (2013) Biochim. Biophys. Acta, Bioenerg. 1827, 1003−1019. (38) Reza Razeghifard, M. R., and Pace, R. J. (1997) Biochim. Biophys. Acta, Bioenerg. 1322, 141−150. (39) Li, X., and Siegbahn, P. E. M. (2015) Phys. Chem. Chem. Phys. 17, 12168−12174. (40) Nilsson, H., Rappaport, F., Boussac, A., and Messinger, J. (2014) Nat. Commun. 5, 4305. (41) Vinyard, D., Khan, S., and Brudvig, G. (2015) Faraday Discuss., DOI: 10.1039/C5FD00087D. (42) Zhang, C. X., Chen, C. H., Dong, H. X., Shen, J. R., Dau, H., and Zhao, J. Q. (2015) Science 348, 690−693. (43) Bergmann, A., Zaharieva, I., Dau, H., and Strasser, P. (2013) Energy Environ. Sci. 6, 2745−2755. (44) Mattioli, G., Zaharieva, I., Dau, H., and Guidoni, L. (2015) J. Am. Chem. Soc. 137, 10254−10267.
ASSOCIATED CONTENT
* Supporting Information S
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biochem.5b00782. A movie of the interconversion of the Mn4CaO5 cluster along the minimum energy path between the SB3 and SA3 states (MPG) A more detailed explanation of the computational methods and the coordinates (XYZ) of the replica sketched in Figure 2 (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Funding
Funds were provided by European Research Council Project n. 240624 within the VII Framework Program of the European Union. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We acknowledge PRACE infrastructure and the Caliban-HPC centre at the University of L’Aquila for the computational resources supplied.
■
REFERENCES
(1) McEvoy, J. P., and Brudvig, G. W. (2006) Chem. Rev. 106, 4455− 4483. (2) Kanan, M. W., and Nocera, D. G. (2008) Science 321, 1072− 1075. (3) Barber, J. (2009) Chem. Soc. Rev. 38, 185−196. (4) Faunce, T., et al. (2013) Energy Environ. Sci. 6, 1074−1076. (5) Joliot, P., and Kok, B. (1975) Bioenergetics of photosxynthesis, 387−412. (6) Loll, B., Kern, J., Saenger, W., Zouni, A., and Biesiadka, J. (2005) Nature 438, 1040−1044. (7) Dau, H., Grundmeier, A., Loja, P., and Haumann, M. (2008) Philos. Trans. R. Soc., B 363, 1237−1243. (8) Guskov, A., Kern, J., Gabdulkhakov, A., Broser, M., Zouni, A., and Saenger, W. (2009) Nat. Struct. Mol. Biol. 16, 334−342. (9) Umena, Y., Kawakami, K., Shen, J.-R., and Kamiya, N. (2011) Nature 473, 55−60. (10) Suga, M., Akita, F., Hirata, K., Ueno, G., Murakami, H., Nakajima, Y., Shimizu, T., Yamashita, K., Yamamoto, M., Ago, H., and Shen, J. R. (2014) Nature 517, 99−103. (11) Askerka, M., Vinyard, D. J., Wang, J., Brudvig, G. W., and Batista, V. S. (2015) Biochemistry 54, 1713−1716. (12) Boussac, A., Girerd, J.-J., and Rutherford, A. W. (1996) Biochemistry 35, 6984−6989. (13) Haddy, A. (2007) Photosynth. Res. 92, 357−368. (14) 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., and Messinger, J. (2011) J. Am. Chem. Soc. 133, 3635−3648. (15) Pantazis, D. A., Ames, W., Cox, N., Lubitz, W., and Neese, F. (2012) Angew. Chem., Int. Ed. 51, 9935−9940. (16) Bovi, D., Narzi, D., and Guidoni, L. (2013) Angew. Chem., Int. Ed. 52, 11744−11749. (17) Cox, N., and Messinger, J. (2013) Biochim. Biophys. Acta, Bioenerg. 1827, 1020−1030. (18) Boussac, A., Rutherford, A. W., and Sugiura, M. (2015) Biochim. Biophys. Acta, Bioenerg. 1847, 576−586. (19) Chrysina, M., Zahariou, G., Ioannidis, N., and Petrouleas, V. (2010) Biochim. Biophys. Acta, Bioenerg. 1797, 487−493. D
DOI: 10.1021/acs.biochem.5b00782 Biochemistry XXXX, XXX, XXX−XXX
