If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported in part by the Deutsche Forschungsgemeinschaft (SFB 498, Berlin), the UniCat Cluster of Excellence (Unifying Concepts in Catalysis, Berlin), the German Ministry of Education and Research (Bio-H2, 03SF0318C; and H2 Design Cell, 03SF0355D), and the European Union (SOLAR-H2 516510). The on-line version of this article (available at http://www.jbc.org) contains supplemental “Materials and Methods,” Figs. S1–S7, Tables S1–S3, Equations S1–S6, and additional references. 1 Supported by a fellowship from the Alexander-von-Humboldt Foundation.
In oxygenic photosynthesis, solar energy drives the oxidation of water catalyzed by a Mn4Ca complex bound to the proteins of Photosystem II. Four protons are released during one turnover of the water oxidation cycle (S-state cycle), implying thermodynamic limitations at low pH. For proton concentrations ranging from 1 nm (pH 9) to 1 mm (pH 3), we have characterized the low-pH limitations using a new experimental approach: a specific pH-jump protocol combined with time-resolved measurement of the delayed chlorophyll fluorescence after nanosecond flash excitation. Effective pK values were determined for low-pH inhibition of the light-induced S-state transitions: pK1 = 3.3 ± 0.3, pK2 = 3.5 ± 0.2, and pK3 ≈ pK4 = 4.6 ± 0.2. Alkaline inhibition was not observed. An extension of the classical Kok model facilitated assignment of these four pK values to specific deprotonation steps in the reaction cycle. Our results provide important support to the extended S-state cycle model and criteria needed for assessment of quantum chemical calculations of the mechanism of water oxidation. They also imply that, in intact organisms, the pH in the lumen compartment can hardly drop below 5, thereby limiting the ΔpH contribution to the driving force of ATP synthesis.
). Its evolutionary development, ∼3.5 billion years ago, has boosted life on Earth by facilitating the efficient use of water as a source of electrons and protons for the synthesis of energy-storing carbohydrates and biomass in general. The long-standing scientific interest in photosynthetic water oxidation has recently been invigorated by the vision of future technological systems that, akin to plants and cyanobacteria, use water as a substrate for light-driven formation of energy-rich and storable compounds (H2 or other fuel materials) (
). We believe that insights into energetics and reaction mechanisms of the natural paragon could provide inspiration and guidelines for the development of new technologies (
In oxygenic photosynthesis, solar energy drives the oxidation of water and the accumulation of “energized” electrons by reduction of quinones (Fig. 1). In this process, four electrons and four protons are removed from two molecules of water (
). The reaction is denoted as water oxidation because the electrons are removed from the oxides (O2−) of two “substrate” water molecules, followed by O–O bond formation and dioxygen (O2) release. In photosynthesis, the process is catalyzed by a Mn4Ca complex bound to the proteins of Photosystem II (PSII)
). The light-driven oxidation of one specific tyrosine residue (YZ) in PSII provides the oxidant (YZ·+) and thus the driving force. YZ·+ is formed and reduced 4 times per released O2 molecule; four protons are released into the aqueous phase. Accordingly, the following net equation describes the process of photosynthetic water oxidation (Equation 1).
). Photon absorption is followed by formation of the primary radical pair P680+Phe−. Subsequently, at the (electron) acceptor side of PSII, Phe− reduces the primary quinone acceptor, QA; at the donor side of PSII, P680+ is reduced by a redox-active tyrosine denoted as YZ. These processes are completed within <1 μs. Subsequently, YZox oxidizes the catalytic center of water oxidation comprising an inorganic Mn4Ca(μ-O)n core, coordinating residues and bound water molecules, whereas at the acceptor side, an electron is transferred from QA to QB, the secondary quinone acceptor. Schematically indicated are the binding of two water molecules and the release of four protons. In intact organisms, this proton release results in acidification of the lumen compartment, which is separated from the stroma by the thylakoid membrane. At the acceptor side, the protonation of QB causes alkalization of the stroma. In plants and cyanobacteria, the thylakoid pH gradient resulting from lumen acidification and stroma alkalization contributes to the driving force of ATP synthesis. Chl, chlorophyll.
Equation 1 implies that the driving force (which corresponds to the overpotential in electrochemical water oxidation) is diminished by the presence of the reaction products, namely dioxygen and protons, and approaches zero at high product concentrations (high pO2 and low pH). Equation 1 thus implies a principal thermodynamic limitation but leaves open paths and mode of product backpressure (or product inhibition). Experimental studies are required to determine how energetics, efficiency, and reaction kinetics of photosynthetic water oxidation relate to the diminished driving force at increased product levels. The influence of elevated O2 concentrations has been investigated previously (
). In time-resolved x-ray experiments, no inhibition of the O2 evolution transition (S3 → S0 + O2) was detected, up to pO2 of 16 bar, suggesting that the Gibbs free energy of any reaction intermediate is >130 meV higher than for the product state (S0 + O2) (
) discovered, for excitation with a sequence of saturating flashes of light, a periodic flash number dependence of the oxygen evolution yield. To explain this periodicity, in 1970, the paradigmatic Kok cycle was proposed (see Fig. 2A), which describes the sequential accumulation of 4 oxidizing eq (by the manganese complex) before O2 formation (
). These studies represent major progress but could not provide a complete and clear-cut characterization of the thermodynamic limitations at high proton concentration.
FIGURE 2S-state cycle model and pH-jump experiment.A, starting in the dark-stable S1-state, each saturating flash of visible light initiates a sequence of ET steps and eventually drives the Si → Si+1 transition. Each Si → Si+1 transition involves oxidation of the manganese complex, a protein-bound Mn4Ca(μ-O)n complex and its ligand environment. Water oxidation and O2 formation follow S4-state formation induced by the third flash applied to dark-adapted PSII. Later, we will extend the Kok S-state cycle model by taking into account explicitly also the removal of four protons from the manganese complex. B, the flash number dependence of the O2 formation yield was detected for PSII particles on a Clark-type electrode (Elec.; ●). The amplitude of the chlorophyll DF at ∼2 ms after a saturating laser flash (DFms) provides an alternative measure of the relative O2 yield (○). The maximal O2 yield at the third and seventh flashes is explained by the S-state cycle in A. r.u., relative units. C, experimental pH-jump protocol. D, representative data. The DF amplitude at 2 ms after a laser flash application (DFms) is shown for the flashes of the “probe phase” (at pH 6.4). The changed pH of 4.2 was present only during application of one specific flash of the sequence, namely during the first (S1 → S2), second (S2 → S3), third (S3 → S0), and fourth (S0 → S1) flashes. There was no pH change in the control data. Inhibition of the S3 → S0 transition at pH 4.2 resulted in a shift of the maximum from the seventh to eighth flashes (arrow). a.u., arbitrary units.
A crucial experimental difficulty lies in discrimination between the various influences on PSII that relate directly or indirectly to the proton concentration, namely (i) influence on the efficiency (or yield) of individual S-state transitions, (ii) acceleration of S-state decay in the dark interval between saturating flashes, and (iii) transient or (iv) irreversible deactivation of PSII (damage). Using a new experimental approach, we will discriminate between points i, ii, iii, and iv and explore the thermodynamic limitations of PSII water oxidation in a quantitative way. The results support an extension of the classical Kok reaction cycle (
) were suspended in Buffer I (500 mm glycine betaine, 15 mm NaCl, 5 mm MgCl2, 5 mm NaCl2, 1.25 mm MES, 1.25 mm MOPS, and 1.25 mm HEPPS, pH 6.4). The addition of Buffer II (Buffer I with the three buffering substances at 200 mm instead of 1.25 mm) changed the pH from 6.4 to a desired value in the range between 3.0 and 8.9. Buffer III (600 mm) shifted back the pH to 6.4. The individual pH-jump protocols are schematically shown in supplemental Fig. S1 (A–H).
The delayed fluorescence (DF) transients were recorded after nanosecond flash excitation (∼5 ns, 532 nm, ≥2 mJ cm−2) (
). To obtain a low-noise signal that exhibits the typical flash number dependence of dioxygen formation (DFms), the DF amplitude was averaged from 1 and 3 ms after the respective flash (
). For estimation of the efficiency of S-state transition at different pH values, the DFms pattern for the flashes applied after the pH jump was simulated on the basis of an adapted Kok model. To determine the pK values of the individual S-state transitions, the pH dependence of the resulting transition efficiencies was simulated assuming a titration curve behavior (supplemental Equation S5). For further details, see supplemental “Materials and Methods.”
RESULTS
In our experiment, we exploited a unique feature of water oxidation in PSII: saturating nanosecond flashes of green laser light (5 ns, 532 nm) can be used to drive the catalysts synchronously through its reaction cycle. Flashes 1–4 induce the transitions S1 → S2 (first flash), S2 → S3 (second flash), S3 → S0 (third flash), and S0 → S1 (fourth flash); subsequent flashes drive further rounds of the reaction cycle (Fig. 2A) (
). O2 is formed only in the course of the S3 → S0 transition. Therefore, a period of four oscillations in the flash number dependence of the O2 yield is observed, ideally with maxima at flashes 3, 7, and 11 (Fig. 2B) (
). The damping and (phase) shift of maxima at higher flash numbers reflect partial desynchronization explainable by the less-than-unity quantum efficiency of the individual transitions, which has been described in terms of a miss probability (
). In Fig. 2B, two alternative methods for detection of the flash number dependence of the O2 formation yield are compared. The closed circles relate to detection by an oxygen electrode, whereas the open circles describe the oscillations in the amplitude of the delayed fluorescence in the millisecond range after flash excitation (DFms) (
). We found that the flash number dependences of the two signals are identical. Because detection by an oxygen electrode is incompatible with our pH-jump protocol, we have chosen the DFms signal for detection of the O2 flash pattern.
The timing diagram in Fig. 2C illustrates our experimental approach (see also supplemental Fig. S1). In Phase I (S-state preparation at pH 6.4), the oxygen-evolving complex was advanced to the desired S-state by zero (S1), one (S2), two (S3), or three (S0) saturating flashes spaced by 700 ms, a period ensuring a maximal transition probability (i.e. a minimal value of the Kok miss parameter) (
). In Phase II (changed pH), the pH of the medium was changed by the addition and rapid mixing of Buffer II, facilitating a jump toward the chosen pH value (here in the pH range from 3 to 9). At the changed pH, a single laser pulse initiated the next S-state transition. Subsequently, the pH was brought back to pH 6.4 (by addition of Buffer III). In Phase III (probe flashes at pH 6.4), a train of laser pulses was applied (probe flashes) (Fig. 2C).
In Fig. 2D, typical results are shown: the O2 flash patterns in the probe phase (at pH 6.4) for preceding pH jumps down to 4.2. Any shift (delay) of the maximum of the oscillations indicates a lowered efficiency of the transition that was initiated by a single flash during the pH-jump phase. For the S3 → S0 transition (third flash) at the changed pH, the flash pattern was delayed by one flash, indicating an S3 → S0 efficiency that is close to zero at pH 4.2. A smaller but still sizeable delay was observed for a pH jump in the S2 → S3 transition, whereas the S1 → S2 transition was only marginally affected, and no delay was observed for S0 → S1. For application of the third (S3 → S0) and especially the fourth (S0 → S1) flashes, a decreased amplitude of the oscillations in the probe phase was observed, explainable by deactivation of a sizeable PSII fraction by the preceding low-pH jump.
Visual inspection revealed the order in the low-pH sensitivity of the individual transitions (S3 → S0 > S2 → S3 > S1 → S2 > S0 → S1) and partial deactivation (for third and fourth flashes at low pH). Curve fitting of the probe pattern on the basis of the classical Kok model facilitated quantification of (i) the relative transition probability (efficiency) at the changed pH (Fig. 3) and (ii) the active fraction of PSII (Fig. 4).
FIGURE 3pH dependence of the transition efficiency. The pH was different from the standard value of 6.4 only for the indicated flash-induced transition (protocol of Fig. 2C). The resulting flash pattern (exemplary data in Fig. 2D) was simulated using an adaptation of the Kok model, resulting in the transition probabilities shown at the changed pH (open circles). The black lines were obtained by calculation of standard pH titration curves for the indicated pK values (the pK corresponds to the half-inhibition pH). The data also were corrected for S-state decay and deactivation (red circles). Simulation of the corrected data by pH titration curves resulted in the red lines and pK values of 3.3 ± 0.3 (S1 → S2), 3.5 ± 0.2 (S2 → S3), and 4.6± 0.2 (S3 → S0). The steep pH dependence of the S3 → S0 transition is better simulated if cooperativity with a Hill coefficient of 2 is considered (n = 2, steeper red line; n = 1, less steep red line). For S0 → S1, the transition efficiency is pH-independent.
FIGURE 4Inactivation by ΔpH exposure of PSII residing in one specific S-state (●) or during laser flash application (○). ●, fraction of active PSII for an experiment in which the pH was changed for 3 s without application of any laser flash during these 3 s. The indicated S-state had been populated before at pH 6.4 (supplemental Fig. S1, E–H). ○, fraction of active PSII after application of a laser flash at the changed pH, inducing the indicated Si → Si+1 transition (supplemental Fig. S1, A–D). Inactivation of PSII was observed only if PSII in the S0-state was exposed to high proton concentrations (low pH). The error bars represent S.E. (three to seven repetitions per pH value).
The transition probabilities for the four S-state transitions at pH values ranging from 3 to 9 are shown in Fig. 3. Simulations based on the Henderson-Hasselbalch equation (supplemental Equation S5) resulted in effective pK values for the efficiency drop at low pH values of 3.3 (first flash, S1 → S2), 3.9 (second flash, S2 → S3), and 4.5 (third flash, S3 → S0), whereas the efficiency of the S0 → S1 transition (ϕS0→S1, fourth flash) was pH-independent. In the alkaline range, the efficiency was found to be maximal and pH-independent for all S-state transitions. (In Fig. 3, the numerical values of the transition probabilities of the fourth-flash transition at low pH are exactly equal to 100% and slightly lower at higher pH. This results merely from inadequacies of the used Kok model to describe S-state deactivation and transient inactivation of PSII.)
In the above experiments, the pH differed from the standard pH (6.4) not only during the flash-induced transition itself but also for 1.5 s before and after the flash. A discrimination between the accelerated S-state decay at low pH (
) and the pH sensitivity of the light-induced transition itself required additional experiments. To assess and quantify the influence of pH exposure in the dark, we used the protocol shown in Fig. 2C but without flash application in the pH-jump phase. Specifically for PSII poised in the S3-state, strong acceleration of the S-state decay was observed, a process describable by an effective pK of 3.7 (supplemental Fig. S2).
A reduction in the amplitude of the probe phase oscillations was observed only for low-pH exposure of PSII poised in the S0-state, indicating formation of an inactive S0-state (S0#). The extent of S0#-state formation upon dark exposure to low pH for 3 s is shown in Fig. 4, suggesting an apparent pK of 4.6. Notably, S0#-state formation was slowly reversed at the control pH (supplemental Fig. S4), indicating that formation of this state does not occur by irreversible “damage” to PSII. For all other S-states, the short time period of exposure to low and high pH (3 s) precluded transient or permanent inactivation of PSII, as confirmed by an O2 activity assay (supplemental Fig. S7).
Our finding of optimal efficiency of the S0 → S1 transition also at low pH (Fig. 4) is in conflict with previous reports (
). (See the detailed quantitative comparison with the previously obtained pK values under supplemental “Materials and Methods”). Previously, a clear separation between reduced transition efficiency and PSII inactivation in the S0-state had not been achieved. Therefore, PSII inactivation by S0#-state formation could mimic a reduced efficiency of the S0 → S1 transition in these earlier studies.
On the basis of our results on the pH dependence of S-state deactivation (pH jump without flash), refined simulations were approached that facilitated a correction for S-state decay of the data shown in Fig. 3 (see supplemental “Materials and Methods” for details). The thereby obtained pH dependence of the efficiency of the light-induced S-state transition itself (Fig. 3, red lines) represents a central result of our study. The half-inhibition pH of the S1 → S2 transition was determined to be 3.3 ± 0.3; for the S2 → S3 transition, the value was determined to be 3.5 ± 0.2. The low-pH efficiency drop of the S3 → S0 transition is characterized by an effective pK of 4.6 ± 0.2. We note that, for the S3 → S0 transition, a Hill coefficient of 2 (n = 2 in supplemental Equation S5) (Fig. 3, steeper red line) results in a better description than a Hill coefficient of 1. (Also, the first-flash data might be simulated better assuming that n = 2. However, at these extremely low-pH values, the data are less reliable because other pH-dependent processes might start to interfere. Therefore, we used a conservative simulation approach for this transition, i.e. an n value of unity.)
Up to this point, the analysis had been based exclusively on the oscillation pattern of the millisecond fluorescence in the probe phase. Next, we turned to the DF decays detected within 10 μs to 60 ms after the laser flash was applied at the changed pH. These DF decays carry unique mechanistic information on the efficiency loss at low pH. The competition between the forward reaction (i.e. the S-state transition) and losses by charge recombination is a major determinant of the transition efficiency (or transition yield, Φs = 1 − ms). The DF decays reflect the competing charge recombination losses. Therefore, the time integral of the DF provides a relative measure of the probability for miss events by charge recombination (
In Fig. 5, the integrated delayed fluorescence is compared with the transition efficiency (from the red lines in Fig. 3, corrected for the S-state dark decay). For the S1 → S2 and S2 → S3 transitions, a good correlation was observed (Fig. 5, upper panel), implying that the efficiency drop at low pH is indeed explainable by competition between pH-dependent forward reactions and recombination processes. For S3 → S0, however, this correlation breaks down at low pH (Fig. 5, lower panel), suggesting that a pH limitation comes into play that does not result in charge recombination within the 60-ms time window of the DF measurement. Therefore, we propose that there are two deprotonation-limited processes in the S3 → S0 transition: the first one competing with fast charge recombination (pK slightly higher than 4.6) and the second one unrelated to rapid recombination losses (pK slightly below 4.6). The existence of two pH-dependent processes of similar pK also explains the steep pH dependence of the efficiency drop at low pH (Hill coefficient of 2) (Fig. 3).
FIGURE 5Comparison of the integrated DF (■ and ●) and transition efficiency (□ and ○) of transitions S1 → S2 (● and ○), S2 → S3 (■ and □), and S3 → S0 (● and ○). We note that that the open symbols represent the miss probability (Prob.; ms), which is related to the transition efficiency (Φs) according to ms = 1 − Φs. The integrated (Int.) DF represents a relative measure of the recombination losses occurring after the laser flash (within the data acquisition period of 10 μs to 60 ms). For all data points, the DF integral was scaled by the same factor. r.u., relative units.
In summary, we assessed the influence of proton concentrations ranging from 1 nm (pH 9) to 1 mm (pH 3) on (a) the efficiency of distinct S-state transitions, (b) the rate of S-state decay, and (c) deactivation by short-term exposure to extreme pH values. The experimental approach used facilitated full discrimination between the three above-mentioned types of influences. Accelerated S-state decay was observed only in the S3-state (pK ∼ 3.7) (supplemental Fig. S2). We did not detect any low-pH inhibition of the S0 → S1 transition itself; however, we discovered that low-pH exposure of PSII in the S0-state resulted in transient formation of an inactive state (S0#). Detailed characterization of this interesting state requires further investigation. In all other S-states, neither low- nor high-pH exposure (for 3 s) caused transient deactivation or lasting damage.
DISCUSSION
For three transitions between semistable states of the reaction cycle, effective pK values for the efficiency decrease in the light-induced transition at high proton concentrations have been determined: 3.3 ± 0.3 (S1 → S2), 3.5 ± 0.2 (S2 → S3), and 4.6 ± 0.2 (S3 → S0; Hill coefficient of ∼2). Reduced transition efficiencies in the alkaline pH range were not observed, inter alia proving the insensitivity of our experiment to deceleration of the quinone chemistry at the acceptor side of PSII.
The efficiency decrease at low pH relates to the competition between forward reactions (S-state transitions) and recombination processes. At low pH, the forward reactions are blocked, and the yield for recombination reactions is increased. Comparison with the integrated recombination fluorescence indeed suggests that lacking deprotonation at low pH results in increased charge recombination, thereby preventing the ET from the manganese complex to YZ·+ for all three pH-dependent transitions (Fig. 5). Specifically, in the S3 → S0 transition, a process may be involved for which the observed low-pH block is not explainable by charge recombination within the experimentally accessed time range of 10 μs to 60 ms after flash excitation. This process likely involves the blockade of a later step in the reaction sequence that leads from the S3-state to the S0-state. We propose that, due to this low-pH blockade, a state is formed that decays either by slow charge recombination, within hundreds of milliseconds, or by a similarly slow one-electron side reaction.
Transient Inactivation by S0#-state Formation
In our study, a pK value of 4.6 was determined for the formation of the inactive S0#-state, which was induced by low-pH exposure for ∼3 s (Fig. 4). We note that the value of 4.6 reflects merely an apparent pK because, as indicated by the experiments of supplemental Fig. S4, low-pH exposure for 3 s was insufficient for reaching the equilibrium level of inactivation. The equilibrium pK may instead be close to 5.0 (supplemental Fig. S4). How does the protonation reaction associated with S0#-state formation in the dark (within seconds) relate to the transient S0+-state formation in the flash-induced S3 → S0 transition (see Fig. 6)? The formation of S0# at low pH and its decay at the control pH both took several seconds and thus are by ∼4 orders of magnitude slower than the formation and decay of S0+ (in the microsecond/millisecond domain). This observation strongly suggests that S0#+ and S0+ are not identical states of the manganese complex. Nonetheless, the same protonatable group or hydrogen bond cluster may be involved in both S0#+- and S0+-state formation.
FIGURE 6Basic reaction cycle of photosynthetic water oxidation. The classical Kok model is extended to describe electron and proton removal from the Mn4Ca complex and its ligand environment. The charge and oxidation states of the manganese complex are described by an S-state nomenclature in which the subscript indicates the number of accumulated oxidation equivalents and the superscript indicates the charge relative to the dark-stable S1-state (+, positive; and n, neutral). The states S1n (dark-stable), S2+, S3+, and S0n correspond the states S1, S2, S3, and S0 of the classical Kok cycle; S3n or S4+ may correspond to the Kok S4-state. Each proton is removed from the catalytic site; estimates of the corresponding pK values (pK1 to pK4) were obtained in this study. At low pH, the S0-state is reversibly inactivated by S0#-state formation. The interrelation between the protonation reaction associated with pK4 and pK# is still unclear. We note that the ET from the manganese complex to YZox and the long-distance proton movement toward the aqueous phase cannot proceed in a concerted mode. However, direct coupling of the ET to local proton shifts is conceivable and not covered by the scheme.
Extension of the Kok Cycle Model and Assignment of pK Values
The effective pK values determined here are likely related to deprotonation steps occurring at the catalytic site, i.e. the manganese complex and in its ligand environment. Any deprotonation at more distant sites would be unfavorable for the kinetics and energetics of the water oxidation chemistry (
). Assignment to specific deprotonation steps in the light-driven reaction cycle of photosynthetic water oxidation is facilitated by an extended S-state model that describes not only the ET steps (from the manganese complex to YZ) but also the essential deprotonation steps (proton transfer from the catalytic site to the aqueous phase) (Fig. 6) (
) experiments suggest that, in the S3 → S0 transition, a deprotonation step (S3+ → S3n, pK3) precedes ET, O–O bond formation, and a second deprotonation step (S0+ → S0n, pK4). Our results on the S3 → S0 transition imply effective values for pK3 and pK4 close to 4.6. Also in the S2 → S3 transition, proton removal from the manganese complex precedes the ET step (S2+ → S2n) and predicts inhibition at low-pH values; we found that pK2 is likely close to 3.5. The pH independence of the S0 → S1 transition is explainable by a rapid ET step (S0n → S1+) that precedes deprotonation (S1+ → S1n). How does low-pH blockage of the latter deprotonation steps affect the efficiency of water oxidation?
The S1 → S2 transition itself is not associated with proton release and thus is predicted to be pH-independent. The low-pH inhibition found here is nonetheless explainable within the framework model of Fig. 6. At low pH, the S1+-state is formed, thereby preventing induction of the S1n → S2+ transition by a laser flash. This conjecture is plausible and supported circumstantially by a severalfold increase in the initial level of the first-flash DF decay observable at pH values below 4 (supplemental Fig. S6).
Alkaline pH results in an increased driving force for water oxidation, so, from the viewpoint of thermodynamics, high-pH inhibition is not anticipated and indeed not observed in our investigation. The findings reported here are consistently explainable within the framework provided by the reaction cycle of Fig. 6 (
), thus circumstantially supporting the proposed basic sequence of ET and proton removal steps.
Mechanistic Implications
Mostly due to the non-equilibrium character of the light-induced processes, the effective pK values determined here (pK1 = 3.3, pK2 = 3.5, and pK3 and pK4 close to 4.6) might differ from the actual pK of the deprotonating group, but likely by <1 pK unit (estimate on the basis of simple kinetic modeling). We tentatively assign pK1 to μ-hydroxo deprotonation and pK2 to deprotonation of a water species terminally ligated to Mn(IV) paralleled by μ-oxo bridge formation (
); pK4 might be related to deprotonation of a substrate water (after water ligation to manganese). However, also other protonatable groups could be involved, e.g. carboxylate side chains such as Asp61 of the D1 protein. Proton shifts directly coupled to the ET step could complicate the assignment. Definitive identification of the four deprotonating groups represents an important task for future work; computational chemistry could contribute (
The pK value of the deprotonating group (pKi) relates to the binding energy of the proton and facilitates calculation of the relative free energy drop (ΔGiH) associated with the respective deprotonation step (
), numerical calculation of the four ΔGiH values resulted in −4.6, +8.1, +4.1, and −2.4 kcal/mol (at pH 7), whereas the experimental pKi values suggest a clearly different set of energies (approximately −5, −5, −3, and −3 kcal/mol, according to Equation 2). In another study (
), the discrepancy between calculated and our experimentally determined values was smaller but still sizeable. This example illustrates that the four experimental pK values, even though still imprecise, represent a criterion for evaluation of computed reaction paths.
Water Oxidation in Intact Organisms
Product backpressure limits water oxidation by PSII already at surprisingly low proton concentrations. Between pH 4.5 and 5.5, reduced efficiency results from inhibition of two steps in the transition S3 → S0 + O2 (S3+ → S3n (pK3) and S0+ → S0n (pK4)), as well as from transient deactivation by S0#-state formation. Only at clearly lower pH values would inhibition of the S2 → S3 and S1 → S2 transitions and accelerated S3-state decay come into play, whereas the S0 → S1 transition is fully pH-independent.
For intact plants illuminated at saturating light intensities, the pH in the lumen compartment and thus at the donor side of PSII has been estimated to be 5.7 ± 0.5 (
). Accordingly, two steps in the transition S3 → S0 + O2 work close to their respective thermodynamic limitations, whereas a clear free energy surplus (ΔGH, Equation 2) characterizes all other S-state transitions.
Acidification of the lumen compartment by the PSII donor side provides an essential contribution to the driving force of ATP synthesis, the so-called proton-motive force (
). The described thermodynamic limitations of the S3 → S0 transition imply that the lumen pH cannot drop much below pH 5, thereby severely limiting the driving force for ATP synthesis. (A high trans-thylakoid ΔpH might also come from alkalization of the stroma. However, the comparatively large volume and buffer capacity of the stroma space likely prevent a pronounced increase in the stroma pH.) To compensate, plants apparently control their trans-thylakoid ion fluxes such that a major fraction of the proton-motive force stems from the membrane voltage formed upon illumination (
). In clear contrast, photosynthetic water oxidation is severely inhibited already at moderately increased proton concentrations. Our differentiated assessment of the thermodynamic low-pH limitations has provided insights into (i) the energetics and reaction cycle of the catalyst, i.e. the Mn4Ca complex of PSII; and (ii) the bioenergetics of oxygenic photosynthesis in general. Future progress could come from tracking the proton release in the time domain and from identification of the chemical identity of the deprotonating groups.
Acknowledgments
We thank Dr. M. Haumann for discussion and M. Fünning for preparing PSII particles.
51260-X&id=gr1.jpg){kind=link}
51260-X&id=gr2.jpg){kind=link}
51260-X&id=gr3.jpg){kind=link}
51260-X&id=gr4.jpg){kind=link}
51260-X&id=gr5.jpg){kind=link}
51260-X&id=gr6.jpg){kind=link}