The oxygen-evolving complex and the mechanism of water oxidation by Photosystem II"Of all the biochemical inventions in the history of life, the machinery to oxidize water — photosystem II — using sunlight is surely one of the grandest." (Sessions, A. et al, 2009)
1The origin of the OEC is an enigma.
There are several hypotheses for the origin of the OEC. One involves binuclear-manganese active sites, including ribonucleotide reductase, catalase and arginase.
Without question, the most fascinating and complicated aspect of Photosystem II is the oxygenevolving activity (Vinyard et al., 2013). However, this remains one of the most active research areas in the entire field of photosynthesis, so the picture is changing rapidly, and the complexity of the process is such that there is still much to learn. The chemical reaction carried out by the OEC is the oxidation of water to molecular oxygen :
2H2O → O2 + 4H+ + 4 e−
The thermodynamics of this redox process are formidable. Water is an extremely poor electron donor, which is as expected, because oxygen is such a powerful electron acceptor. The redox potential for the half-reaction shown in the equation above is +0.82 V at pH 7, and is somewhat higher in the acidic environment in which water oxidation takes place. In order to oxidize water, it is essential that an even stronger oxidant be available. This is provided by the oxidized reaction center chlorophyll P680+, which has a redox potential estimated to be as high as +1.2 V. The electrons that re-reduce P680+ ultimately come from water, and the protons that result from water oxidation are expelled into the thylakoid lumen. They are released into the lumen because of the vectorial nature of the membrane and the fact that the OEC is localized on the interior surface of the thylakoid. These protons are eventually released from the lumen to the stroma through the process of ATP synthesis. In this way, the electrochemical potential formed by the release of protons during water oxidation contributes to ATP formation. The chemistry of water oxidation is inherently a four-electron process, as four electrons must be extracted to make a single molecule of oxygen. The photochemistry of the reaction center takes place one electron at a time, so there is a fundamental mismatch between the intrinsic chemistries of these two processes. One could imagine several possible solutions to this problem, including at one extreme the cooperation of four reaction centers that all contribute to oxidize a single OEC, or at the other extreme, the storage of four oxidizing equivalents by a single reaction center/OEC working alone, which then oxidizes water to oxygen in a concerted step.
The latter possibility was given dramatic support in a classic series of experiments performed in the 1960s, first by Pierre Joliot and slightly later by Bessel Kok. Their experiments were a variation on the flashing light oxygen measurements made in the Emerson and Arnold experiment. However, these workers devised extremely sensitive electrochemical methods capable of detecting oxygen produced in a single flash.When this technique was applied to a sample that had been allowed to dark-adapt for several minutes, a remarkable pattern of oxygen production was observed, as shown in Fig. 7.5 (Kok et al., 1970).
Little or no oxygen is produced on the first two flashes; a large amount is produced on the third, with lesser maxima with a period of four, until the oxygen production finally damps to a constant value by about the twentieth flash. Kok and coworkers proposed a schematic model (Fig. 7.5) explaining these observations. The model consists of a series of five states, known as S0 to S4, which represent successively more oxidized forms of the OEC. Light flashes advance the system from one S state to the next, until state S4 is reached. State S4 produces O2 without further light input and returns the system to S0. Occasionally, a center does not advance to the next S state upon flash excitation, and, less frequently, a center is activated twice by a single flash. These “misses” and “double hits” cause the synchrony achieved by dark adaptation to be lost, and the oxygen yield eventually damps to a constant value. After this steady state has been reached, a complex has the same probability of being in any of the states S0 to S3 (S4 is unstable and occurs only transiently) and the yield of O2 becomes constant.
States S2 and S3 decay in the dark, but only as far back as S1, which is stable in the dark. Therefore, after dark adaptation, approximately three-fourths of the oxygen-evolving complexes appear to be in state S1, and one-fourth in state S0. This distribution of states explains why the maximum yield of O2 is observed after the third of a series of flashes given to dark adapted chloroplasts. This S state mechanism formally explains the observed pattern of O2 release, but not the chemical nature of the S states or the actual chemical mechanism of the process. It has been known for many years that Mn is an essential cofactor in the water-oxidizing process, and for many years it was suspected that the S states represent successively oxidized states of an Mn-containing enzyme.
This hypothesis has been confirmed by a variety of experiments, most notably X-ray absorption and EPR studies, both of which detect the Mn directly. Analytical experiments indicate that four Mn atoms are associated with each oxygenevolving complex. Other experiments have shown that Cl− and Ca2+ ions are essential for O2 evolution, although their precise mechanistic roles have not yet been determined. A structural model for the Mn cluster is shown in Fig. 7.6.
Many of the amino acid residues that serve as ligands for the Mn cluster come from the D1 protein, as well as the C terminal carboxyl terminus of the D1 protein. Consequently, the OEC is located off center towards the D1 side of the core reaction center complex (Fig. 7.3). An additional protein that is important in the oxygen evolution process is a 33 kDa peripheral membrane protein (PsbO protein). This protein is found in all oxygenic photosynthetic organisms. Removal of this protein does not abolish oxygen evolution activity entirely, but it destabilizes the Mn cluster and perturbs the system.
Two additional peripheral proteins, with masses of 23 and 17 kDa (PsbP and PsbQ), are part of the complex that stabilizes the OEC in eukaryotic organisms, but are not found in cyanobacteria. Instead, two different proteins, PsbU and PsbV, are found in this position. All these extrinsic proteins, as they are often called, help to stabilize the OEC and protect it against damage, but they do not directly bind the Mn cluster . There is one intermediate electron carrier between P680 and the OEC. After P680 is oxidized in the primary photochemistry, it is re-reduced in tens of nanoseconds by electron transfer from a tyrosine amino acid side chain (Y161) from the D1 protein, thus forming a tyrosine radical . This tyrosine residue is known as YZ, and it serves to connect the OEC to P680 (Fig. 7.5). The YZ radical can be observed by EPR. A second tyrosine in the symmetry-related position in the D2 protein is known as YD. It appears not to be involved in mainstream electron flow, but, surprisingly, it is usually found in the oxidized radical form. Many mechanisms have been put forward to explain how the OEC accomplishes the difficult chemistry of oxygen evolution. Most of these mechanisms involve the essential role of the tyrosine in a coupled proton and electron (hydrogen atom) transfer process . This proton coupled electron transfer (PCET) mechanism stresses the importance of maintaining charge neutrality in the OEC. Otherwise, the energetics of the removal of additional electrons from a highly positively charged center found in the higher S states would be prohibitively difficult. Current ideas about the detailed chemical mechanism of oxygen evolution incorporate information from the structure of the OEC, biochemical data, theoretical calculations, and principles of inorganic chemistry .
The OEC in PS II contains a heteronuclear Mn4CaO5 cluster (Figure 1, inset) which catalyzes the water oxidation reaction ,that couples the four-electron oxidation of water with the one-electron photochemistry occurring at the PS II reaction center, P680. The OEC cycles through five intermediate S states (S0 to S4, known as the Kok cycle) that corresponds to the abstraction of four successive electrons from the OEC (Figure 1)
The dark stable S1 state is the first oxidized state and subsequent illumination leads to the formation of the S2 and S3 states. Once four oxidizing equivalents are accumulated
(S4 state), a spontaneous reaction occurs that results in the release of O2 and the formation of the most reduced state, the S0 state. Upon further light excitation, the initial S1 state is formed once again, and the catalytic cycle is resumed. Given the importance of PS II in maintaining life and the anticipated role of lightinduced water-splitting for building a renewable energy economy, understanding the structure of the Mn4CaO5 catalyst and the mechanism of the water oxidation reaction is considered to be one of science’s grand challenges . Although details of the chemistry involved in water oxidation are slowly emerging, the mechanism of the reaction is not yet clear. In this chapter, we describe results from X-ray spectroscopy and diffraction studies, especially the use of time-resolved X-raymethods for room temperature studies using the recently introduced X-ray lasers. We will also describe the use of membrane inlet mass spectrometry for the elucidation of the mechanism of water-oxidation and its utility for time-resolved X-ray spectroscopy and diffraction measurements.
Speaking of photosynthesis, Japanese scientists have achieved the imaging of the “Crystal structure of oxygen-evolving photosystem II at a resolution of 1.9?Å,” zooming in almost twice as far as previous studies. Their paper, published in Nature,1 spoke of the reactor as “indispensable for sustaining life on Earth.” It includes detailed drawings of the 20 subunits involved with numerous molecular contacts.
The particular part of the reactor that splits water molecules and combines oxygen atoms into the O2 gas we breathe they said is “one of nature’s most fascinating and important reactions.” Understanding Photosystem II may help humans to mimic plants’ ability to split water efficiently at ambient temperatures, leading to renewable energy for a multitude of applications. The ability lives all around us if we can tap into its secrets. 3Crystal structure of oxygen-evolving photosystem II at a resolution of 1.9 Å
The electron densities of the four manganese atoms and the single calcium atom in the oxygen-evolving complex were well defined and clearly resolved, and the electron density for the calcium atom was lower than those of the manganese atoms, allowing us to identify the individual atoms unambiguously 4,5 (Fig. 2a).Figure 2 | Structure of the Mn4CaO5 cluster. a, Determination of individual atoms associated with the Mn4CaO5 cluster. The structure of the cluster was superimposed with the 2Fo2Fc map (blue) contoured at 5s formanganese and calcium atoms, and with the omit map (green) contoured at 7s for oxygen atoms and water molecules. b, Distances (in a˚ngstro¨ms) between metal atoms and oxo bridges or water molecules. c, Distances between each pair of manganese atoms. d, Distances between the manganese and the calcium atoms. e, Stereo view of the Mn4CaO5 cluster and its ligand environment. The distances shown are the average distances between the two monomers. Manganese, purple; calcium, yellow; oxygen, red; D1, green; CP43, pink.
In addition, five oxygen atoms were found to serve as oxo bridges linking the five metal atoms from the omit map (Fig. 2a). This gives rise to a Mn4CaO5 cluster. Of these five metals and five oxygen atoms, three manganese, one calcium and four oxygen atoms form a cubane-like structure in which the calcium and manganese atoms occupy four corners and the oxygen atoms occupy the other four. The bond lengths between the
oxygens and the calcium in the cubane are generally in the range of 2.4–2.5A˚ , and those between the oxygens and manganeses are in the range of 1.8–2.1A˚ (Fig. 2b). However, the bond length between one of
the oxygens at the corner of the cubane (O5) and the calcium is 2.7A˚ , and those between O5 and the manganeses are in the range of 2.4–2.6A˚ . Owing to these differences in bond lengths, the Mn3CaO4 cubane is not an ideal, symmetric one. The fourth manganese (Mn4) is located outside the cubane and is linked to two manganeses (Mn1 and Mn3) within the cubane by O5 and the fifth oxygen (O4) by a di-m-oxo bridge. In this way, every two
adjacent manganeses are linked by di-m-oxo bridges: Mn1 and Mn2 are linked by a di-m-oxo bridge via O1 and O3, Mn2 and Mn3 are linked via O2 and O3, and Mn3 and Mn4 are linked via O4 and O5. The calcium is linked to all four manganeses by oxo bridges: to Mn1 via the di-m-oxo bridge formed by O1 and O5, to Mn2 via O1 and O2, to Mn3 via O2 and O5, and to Mn4 via the mono-m-oxo bridge formed by O5. The whole structure of the Mn4CaO5 cluster resembles a distorted chair, with the asymmetric cubane serving as the seat base and the isolated Mn4 and O4 serving as the back of the chair. The cubane-like structure has been reported previously 4,9–12, but the oxo bridges and exact distances among the individual atoms could not be determined at the medium resolution achieved previously 4.
The distances among the four manganeses determined for monomer A are 2.8A˚ (Mn1–Mn2), 2.9A˚ (Mn2–Mn3), 3.0A˚ (Mn3–Mn4; 2.9A˚ for monomer B), 3.3A˚ (Mn1–Mn3) and 5.0A˚ (Mn1–Mn4) (Fig. 2c). The distances between the calcium and the four manganeses are 3.5A˚ (Ca–Mn1), 3.3 A˚ (Ca–Mn2), 3.4A˚ (Ca–Mn3) and 3.8 A˚ (Ca–Mn4) (Fig. 2d; for the corresponding distances inmonomer B and the average distances between the two monomers, see Supplementary Table 3). These distances are largely different fromthose reported in the previous crystal structures3–6; however, they are comparable to those reported from extended X-ray absorption fine structure studies 13,14 if we consider that there is an error of 0.16A˚ in the distances determined from the X-ray structural analysis (Methods).
In addition to the five oxygens, four water molecules (W1 to W4) were found to be associated with the Mn4CaO5 cluster, of which W1 and W2 are coordinated to Mn4 with respective distances of 2.1 and 2.2A˚ , and W3 and W4 are coordinated to the calcium with a distance of 2.4A˚ . No other water molecules were found to associate with the other three manganeses, suggesting that some of the four waters may serve as the substrates for water oxidation.In addition to the direct ligands of the Mn4CaO5 cluster, we found that D1-Asp 61, D1-His 337 and CP43-Arg 357 are located in the second coordination sphere and may have important roles in maintaining the structure of the metal cluster, in agreement with various reports showing the importance of these three residues in maintaining the oxygen-evolving activity15–19. One of the guanidinium g-nitrogens of CP43-Arg 357 is hydrogen-bonded to both O2 and O4 of the Mn4CaO5 cluster, whereas the other is hydrogen-bonded to the carboxylate oxygen of D1-Asp 170 and to that of D1-Ala 344. The imidazole e-nitrogen of D1-His 337 is hydrogen-bonded to O3. These two residues may thus function to stabilize the cubane structure of the metal cluster as well as to provide partial positive charges to compensate for the negative charges induced by the oxo bridges and
carboxylate ligands of the metal cluster. The carboxylate oxygen of D1-Asp 61 is hydrogen-bonded to W1, and also to O4 indirectly through another water molecule, suggesting that this residue may also contribute to stabilizing the metal cluster. Furthermore, D1-Asp 61 is located at the entrance of a proposed proton exit channel involving a chloride ion (Cl21; see below), suggesting that this residue may function in facilitating proton exit from the Mn4CaO5 cluster5,20,21.
The most significant structural feature of the Mn4CaO5 cluster, which may be important for elucidating the mechanism of the water-splitting reaction, is its distorted chair form. The large distortion is principally caused by the existence of the calcium and O5 in the Mn4CaO5 cluster, as described above. The apparently longer distances between O5 and metal atoms suggested that the corresponding bonds are weak, and that O5 may therefore have a lower negative charge than the valence of 22 expected for normal oxygen atoms in oxo bridges. This in turn suggests that O5 may exist as a hydroxide ion in the S1 state and may provide one of the substrates for dioxygen formation. Because both W2 and W3 are within the hydrogen-bond distances to O5, one of these two waters may provide another substrate. Because the transition between S0 and S1 is fastest in the Kok cycle, the proton released during this transition may be accepted by D1-Tyr 161 (also termed YZ), which is deprotonated by means of protoncoupled electron transfer (PCET; see below).W3is closer to YZ than is O5 (Fig. 3a)
and may be a more favourable candidate than O5 as the proton-releasing group. Thus, W3 rather than O5 may be a hydroxide ion in the S1 state, suggesting that O–O bond formation may occur between W2 and W3. In any case, our results suggest that the O–O bond formation occurs in two of the three species O5, W2 and W3.Conclusion
The high-resolution structure of PSII reveals the geometric arrangement of the Mn4CaO5 cluster as well as its oxo bridges and ligands, and four bound water molecules. This provides a basis for unravelling the mechanism of water splitting and O–Obond formation, one of nature’s most fascinating and important reactions. In addition, our determination of the precise arrangement of amino-acid side chains and cofactors gives us a solid structural understanding of energy migration, electron transfer and water-splitting reactions taking place within PSII.Structure of manganese–calcium cluster in PSII
In 2011, Umena et al. 7 reported the crystal structure of manganese–calcium cluster of PSII at an atomic resolution using a sliding-oscillation method to reduce the X-ray dose illuminated on a unit volume of the crystal. In this structure, one calcium and four manganese ions are bridged by five oxygens (Fig. 1).
Four water molecules were also found in this structure; and two of them may serve as the substrate for water oxidation. The whole structure of the Mn4CaO5 cluster resembles a distorted chair, with the asymmetric cubane serving as the seat base and the isolated Mn4 and O4 atoms serving as the back of the chair1 (Fig. 1). The structural models from structure refinement of computational models based on EXAFS simulations by Brudvig’s group and Siegbahn’s group25 are also close to the structure found in the 1.9A˚ crystal structure of Umena et al.The origin of the OEC is an enigma.
Sauer and Yachandra proposed a possible evolutionary origin for the Mn4 cluster of the photosynthetic water oxidation complex from natural MnO2 precipitates in the early oceans. In the past fewyears, there has been a tremendous surge in research on the synthesis of various manganese compounds aimed at simulating the OEC of PSII but very few inorganic complexes have been claimed to be functional models for the OEC.
It appears that an effective mechanism
coupled to structural changes in the Mn4Ca cluster, with the optimum positions of associated ligands, during the Kok’s S-state cycle, for water to reach the Mn may operate in PSII. The functional PSII is characterized by a branched water supply structure with multiple control points
A number of experiments demonstrate a requirement for Ca2+ for the functional assembly and stability of the OEC.