Oxidation potentials and electron donation to photosystem II of manganese complexes containing bicarbonate and carboxylate ligands

The oxidation potentials of MnII in aqueous solutions of bicarbonate, formate, acetate and oxalate are reported as a function of concentration and compared to the rate of photooxidation of these solutions by the Mn-depleted water-oxidizing complex of photosystem II (apo-WOC-PSII) from peas.

Although all the carboxylate species lower considerably the oxidation potential of MnII, only bicarbonate stimulates the electron transfer from MnII to apo-WOC-PSII.

On the basis of the electrochemical data it is proposed that the unique capability of Mn-bicarbonate complexes to be photooxidized by PSII could be due to four possible reasons: (i) significantly larger decrease in the oxidation potential of MnII (down to 0.52 V); (ii) electroneutrality of the functional electron transfer complex; (iii) the more favorable energetics reflected in the two pKa values for H2CO3/HCO3 and HCO3/CO32− and greater number of proton transfer sites; and (iv) multiple composition possibilities for the MnIII photo-product as MnIII(HCO3)3, MnIII(HCO3)(CO32−) and MnIII(HCO3)2(OH) (due to the high Lewis acidity of MnIII (pK < 1).


Photosynthetic water oxidation produces molecular oxygen in photosystem II (PSII) of green plants and cyanobacteria.

Absorption of light in PSII antenna and reaction center (RC) of PSII induces charge separation between molecules of excited chlorophylls (P680) and pheophytin (free-base chlorophyll), with subsequent electron transfer steps to the first quinone acceptor QA.

The electron vacancy arising in P680+ is filled up by an electron donated ultimately by a water molecule (the terminal reductant of PSII).1,2

Substrate water is bound to an enzymatic water-oxidizing complex (WOC) comprised of Mn4Ca1Clx, redox-active tyrosine (Yz) and possibly bicarbonate.

The WOC serves the purpose of a redox catalyst (electron/proton reservoir) to deliver electrons one at a time to P680+ until four electron vacancies are accumulated resulting in oxidation of two H2O molecules to O2.

It is known that bicarbonate (HCO3) ions are needed for maximum activity of PSII (for recent reviews, see refs. 3–6 and references therein).

The stimulating effect of HCO3 on PSII was primarily ascribed to the processes taking place on the donor side of PSII7 and a model including bicarbonate as a mediator for photosynthetic water oxidation has been given7,8 which, however, was in contradiction with the results of isotopic experiments.9

In addition, strong evidence is known for the action of bicarbonate on the electron acceptor side of PSII, where it promotes efficient re-oxidation of QA.10

This role was supported by a number of compelling data, including evidence for binding of bicarbonate directly to the non-heme Fe located between QA and QB11,12 where it accelerates electron transfer between the quinones and may also serve as proton donor to the reduced QB.

Further, Xiong et al13. have suggested involvement of R-257 on D-1 protein in bicarbonate/formate binding through their site-directed mutagenesis studies.

These studies have led to a widely accepted model (for review, see ref. 4).

Recent data have shown a role for bicarbonate on the donor side, with specific evidence indicating a function in assembling the inorganic cofactors within apo-WOC (WOC-depleted of Mn, Ca and Cl cofactors), but also for sustaining oxygen-evolution activity.14–22

It was shown that after removal of all four manganese ions from subchloroplast PSII preparations, an effective reactivation of electron transfer14,15,17–20 and oxygen evolution18,22,23 was observed only if a nearly stoichiometric amount of MnII (2–4 atoms of MnII per PSII reaction center) was added together with bicarbonate.

Bicarbonate also specifically restores electron transport on the donor side of PSII containing Mn-sufficient, intact, WOC that has been modified with formate14–18 or by removal of bicarbonate from the medium.14,15,17–21

From comparison of the light-induced FT-IR spectra originating from the donor side of O2-evolving PS II and measured in the presence of (12C)NaHCO3 or (13C) NaHCO321 with the FT-IR spectrum ascribed to the S1/S2 – transition24 it has been postulated that bicarbonate acts as a bridging ligand between the redox-active Mn and Ca2+ within the WOC.21

It has been shown25 that formate modifies the S1-state of the WOC by reduction to the S0-state.

All these facts suggest that HCO3 ion not only involved in the PS II acceptor side, but is also a necessary component for optimal functioning of the WOC.

Although it is chemically reasonable and consistent with available data to suggest that bicarbonate binds as a ligand to either Mn or Ca, we do not yet know where HCO3 binds within the WOC, nor how it functions.

Recent X-ray diffraction data at 3.5 Å resolution of a PSII complex isolated from a cyanobacterium have postulated26 suitable binding sites for (bi)carbonate on both the acceptor and donor sides of PSII.

The authors have suggested (in accordance with an earlier assumption)21 that on the donor side (bi)carbonate binds to the high affinity Mn site and also to Ca based on the absence of sufficient protein ligands,26 though they have no direct support of this.

It was shown in a previous work27 that upon addition of NaHCO3 to an aqueous solution of MnII the potential for oxidation of MnII to MnIII was shifted from 1.19 to 0.63 V at fixed pH 8.3 as a result of formation of (bi)carbonate complexes.

These data provided a basis for suggesting a role for bicarbonate in promoting oxidation of Mn in the WOC.

However, there is contradictory information in the literature about the chemical composition of Mn-bicarbonate complexes in solution.

Lesht and Bauman reported a stability constant for [MnII(HCO3)]+ (K1 = 18.6 at 298 K) using a thermodynamic method.28

Smith and Martel summarized literature data for K1 equal to 63 and 2.8 at ionic strengths of solutions equal to 3.0 and 0.0 (298 K), respectively.29

Sychev and co-workers30,31 summarized the prior literature and proposed successive formation of [MnII(HCO3)]+ and MnII(HCO3)2 complexes and reported their stability constants (K1 = 11 and K2 = 3.7) using an indirect assay method based solely on the HCO3 concentration dependence of the rate of Mn-dependent dismutation of exogenous hydrogen peroxide (at pH 7.0).

Using the same assay method, Stadtman et al. confirmed that HCO3 is required for catalytic dismutation of H2O2 by MnII and showed that three equivalents of HCO3 are required to create the catalyst that is active in the dismutation reaction that forms O2 and water.32

The values of the stability constants for the bicarbonate complexes with MnIII and MnIV and their composition are not known, presumably owing to their instability.

It is clear that more detailed knowledge of the complexation of MnII and MnIII with bicarbonate is required for understanding the thermodynamics consequences.

The aim of the present work is to determine the electrochemical potentials for the oxidation of Mn-complexes of bicarbonate, acetate and formate and correlate them with their capability to donate electrons to Mn-depleted PSII (apo-WOC-PSII).

Methods and materials

Subchloroplast membrane fragments enriched in PSII (BBY-particles) and containing 200–220 chlorophyll molecules per PSII reaction center were prepared from pea (Pisum sativum L.) chloroplasts, as described previously.33

A complete (>95%) removal of Mn from the membrane fragments was carried out using the treatment with 1 M Tris-HCl (pH 8.0) plus 0.5 M MgCl2.34

The preparations were stored in liquid nitrogen or at −80 °C at a Chl concentration of 2 mg ml−1 after the addition of 10% glycerol.

A partial removal of bicarbonate from BBY preparations was achieved, as described earlier14 by a 200-fold dilution of concentrated (2 mg ml−1) PSII preparations into the medium (100 mM MES-NaOH buffer, pH 6.2, 35 mM NaCl) depleted of endogenous bicarbonate by means of 60 min flushing with air (which was freed from CO2 by passage through a solution of 50% NaOH and 20 cm layer of ascarite) and subsequent 10-min incubation at 20 °C.

The kinetics of the photoinduced change of chlorophyll fluorescence yield (ΔF) (λ > 660 nm) related to photoreduction of the primary quinone electron acceptor of PSII, QA, were measured in a tightly closed 10-mm cuvette at 20 °C using a phosphoroscopic set-up.34

Voltammetry curves for MnII oxidation were recorded with a standard potentiostat (polarographic analyzer model PA-3; LP, Praha) using a three-electrode cell, as described earlier.27

A hemispheric Pt electrode with visible surface of 0.01 cm2, and soldered in glass, was used as the working electrode.

A Pt plate (1 × 1 cm) was used as counter-electrode.

All potentials were measured against the saturated calomel electrode and then were recalculated versus the normal hydrogen electrode.

The accuracy of the potential measurements was equal to 10 mV.

A solution of 0.1 M LiClO4 in water was used as the background.

Oxygen was not removed during the measurements.

Adsorption of the oxidized products accompanied by the appearance of a new anodic wave at 0.68 V (that increased with the number of scans), was a characteristic feature of the measurements of MnII oxidation curves.

Therefore, recordings of the voltage–current curves for MnII on Pt were carried out only upon the first scan.

The Pt working electrode was cleaned with filter paper after each scanning into the anodic region.

Reagent grade NaHCO3, NaHCO2, CH3COONa and LiClO4 were used in this work using doubly-distilled water with initial pH 6.5.

Addition of acetate (NaCH3COO), formate (NaHCO2) or oxalate (Na2(COO)2) to water led to gradual alkalization of the solution up to pH 8.5.

Dissolving of bicarbonate (NaHCO3) in the range of concentrations used in this study produces solutions having a constant pH equal to pH 8.3.

Accordingly, the dominant species in solution is the bicarbonate anion (pKa1 = 6.3, pKa2 = 10.3).

Control measurements to examine the influence of pH were performed with MnSO4 in unbuffered water by addition of HCl or KOH.

The potential–current curves for oxidation of MnII were found to be independent of pH in the 5.0 to 8.5 range (less than 10 mV change for the whole region).

However, above pH 9 the oxidation potential begins to shift due to complex formation with hydroxide.

This shift is in agreement with the onset of formation of [Mn(OH)]+ (pKa for Mn2+aq is 10.5).

Therefore, in the present experiments at pH lower than 8.35, Mn2+aq, and not [Mn(OH)]+, is the predominant MnII species detected electrochemically in equilibrium with exogenous anions.


Effect of bicarbonate and other carboxylates on reactivation of electron transport from Mn2+ to PSII reaction centers

It has been shown earlier that complete removal of Mn from PSII membrane fragments results in a 15–20-fold decrease of the photoinduced changes of chlorophyll fluorescence yield (ΔF) (due to loss of electron transfer from the WOC to PSII reaction centers) that can be restored by Mn2+ added at a catalytic (0.1–0.2 μM) concentration.34

The capability of added Mn2+ to restore photoinduced ΔF in Mn-depleted PSII preparations considerably increases upon the addition of bicarbonate.15,16

Fig. 1 shows that (in accordance with the previous publications) the photoinduced ΔF lost due to removal of Mn from BBY preparations (curve 1) is efficiently reactivated if 0.2 μM MnCl2 (which corresponds to 4 Mn atoms per one PSII reaction center) is added together with 5 mM NaHCO3 (curve 4), while 0.2 μM MnCl2 added alone induces a very low reactivation (curve 3).

This activation by bicarbonate in the presence of 0.2 μM MnCl2 (curve 4) saturates at 1–2 mM bicarbonate and 50% saturation occurs at 100–200 μM NaHCO3 (data not shown).

By contrast, formate does not facilitate the restoration of ΔF (curve 5); on the other hand, bicarbonate added to the sample containing 0.2 μM MnCl2 and 5 mM formate causes a considerable restoration of ΔF (curve 6), similar to that observed upon the addition of bicarbonate in the absence of formate (curve 4).

Similar effects are seen for acetate: no reactivation of ΔF upon its addition together with MnCl2 (curve 7) and subsequent reactivation with bicarbonate (curve 8).

The effects are shown for 5 mM concentration of formate and acetate but they are similar over a wide range of concentrations (10 μM–10 mM) of these carboxylates.

Voltammetry of MnIII complexes

The formation of MnIII complexes with different ligands was measured under electrochemical oxidation of MnII to MnIII.

If the electrode reaction is reversible, and the amount of a ligand is so great that its concentration near the surface of the electrode is equal to that in bulk solution, then recording the dependence of the oxidation potential on the logarithm of concentration of the ligand [eqn.

(I)], it is possible to determine the composition of the complex and the equilibrium stability constant for reaction:35,36 MnIIe + pX ⇔ MnIIIXpwhere E1/2me and E1/2com are the half-wave oxidation potentials for MnII and its complex with ligand X, respectively; Kst is the stability constant of the complex; CX is the concentration of ligand X in solution, q and p is the number of ligands in complexes of MnII and MnIII, respectively; n is the number of electrons involved in the redox reaction.

In the case of prior complex formation with MnII the oxidation process is: MnII Xqe + (pq) X ⇔ MnIIIXp.

In this case, the shift of the oxidation potential of MnII depends on the ratio of the stability constants for oxidized (Kox) and the reduced (Kred) forms, as well as on the concentration of the ligand as given by eqn. (II):35,36

In both cases, the number of ligands in the complex can be determined from the slope of the dependence of the oxidation potential on log CX.

By extrapolation of the curve to log CX = 0, the standard reduction potential E0 of the complex and its stability constant can be determined:

The voltage–current curves obtained by voltammetry method on the Pt-electrode are seen as a wave with a peak characterized by the value of the peak potential, Ep.

For reversible one- or two-electron processes, the difference between Ep and E1/2 is equal to 28 and 14 mV, respectively.37

We used the experimental Ep values instead of E1/2 for plotting of eqns. (1) and (2).

The first stage of MnII oxidation in acidic (pH < 1) solutions is a reversible reaction given by eqn. (1):MnII(H2O)6 − e ⇔ MnIII(H2O)6which is characterized by E0 = 1.51 V.38,39

Since the hydrolysis constant (pKa) for MnIII is close to 0, the hydrolysis reaction of MnIII with the formation of [MnIII(OH)2+] takes place at pH > .038

This reaction is followed by further hydrolysis to form Mn2O3(aq) which is a precursor to forming the insoluble solid, Mn2O3(s), as summarized in the following equations:2MnII − 2e + 2H2O ⇔ 2MnIII(OH)2+ + 2H+2MnIII(OH)2+ + 2H2O ⇒ MnIV(OH)4 + MnII + 2H+MnII + MnIV(OH)4 ⇒ MnIIMnIVO3 + H2O + 2H+2MnII − 2e + 3H2O ⇒ MnIIMnIVO3 + 6H+

Due to these hydrolysis reactions the wave for MnII oxidation at pH > 0 in the absence of ligands corresponds increasingly to eqns. (1)–(5).

The hydrolysis reactions are responsible for the irreversibility of reactions (2)–(4).

However, in the presence of ligands (X) that bind to MnIII more strongly than water the hydrolysis reactions of MnIII are suppressed.

Therefore, in the presence of the ligands, oxidation of MnII occurs in accordance with reaction (1).

The potential for oxidation is no longer equal to 1.51 V, but shifts to lower potentials.

Fig. 2 shows the cyclic voltage–current curve of MnSO4 (5 × 10−4 M) oxidation (curve 1) on the background of 0.1 M LiClO4 in water (curve 0).

The oxidation wave of MnII aqua-cations with the maximum at 1.18 V is clearly seen on the anode branch of the curve (in good agreement with previous measurements of the potential of MnII oxidation in water solutions).27

A small cathode wave at 1.12 V and additional two peaks at 0.96 and 0.28 V are observed upon the return scan, reflecting the presence of all the reactions, e.g., (1–5) of MnII oxidation in water.

The anode peak of MnII oxidation is shifted to lower potentials (0.82 instead of 1.18 V) upon the addition of ca.

100 mM acetate, and the cathode peaks at 0.76 V and at 0.59 V are observed upon the return scan in this case.

The difference between the anode (Epa) and the first cathode (Epk) peaks for MnII is equal to 60 mV, both in the presence and absence of acetate thus indicating a reversible transfer of one electron per MnII during the oxidation step.

The peak current of the cathode peak is much lower than the anode peak current, which indicates that a subsequent transformation of the oxidation product into another product takes place.

This oxidation product (the peak at 0.59 V) is attributed to the disproportionation reaction eqn. (3).

The same picture is observed in the case of formate, oxalate and bicarbonate complexes of MnIII (not shown).

The reversibility of the electron transfer was further confirmed by the data on independence of the peak potential difference between Epa and Epk on the scan rate in the region from 10 mV s−1 to 50 mV s−1.

These considerations allow us to consider the MnII oxidation process in the presence of ligands as reversible and thus we used eqns. (I) or (II) to determine the composition, the stability constant and standard potential E0 for the complexes from the dependence of Ep on log CX taking into account that for the free aqua complex of MnII, E0me = 1.51 V.38

Since the value of the stability constants is well known for MnII-oxalate and MnIII-oxalate complexes,40 we started first by considering the effects of the oxalate (Ox2−) on the oxidation of MnII in order to check the validity of using eqns. (I) and (II) to estimate these values.

Upon addition of sodium oxalate the wave of MnaqII oxidation (1.18 V in 0.1 M LiClO4) is shifted to negative potentials.

In the concentration region from 2.15 × 10−3 M to 7.11 × 10−3 M (Fig. 3, curve 1) the dependence of the peak potential on log CX is linear and the slope of this dependence (ΔE/Δlog CX) is equal to 58 mV, indicating participation of just one ligand in the electrode reaction:MnIIe + Ox2− ⇔ [MnIII(Ox2−)]+.Extrapolation of this linear part of the dependence to log CX = 0 gives the value for E0, equal to 0.92 V. Using eqn. (I), the logarithm of the stability constant of the MnIII-oxalate complex is equal to 10 ± 0.01 which is in agreement with the literature data for log Kst of [MnIII(C2O42−)]+ equal to 9..9840

At oxalate concentrations above 7.11 × 10−3 M, a new linear section with a slope of 126 mV is observed (Fig. 3, curve 1) that reveals an additional electrode process.

The slope of this section reveals participation of two ligands in the electrode reaction.

Extrapolation of the linear dependence of Ep to log CX = 0 gives the E0 value equal to 0.77 V. The same value is obtained using eqn. (II) and in the available literature data, as follows.

The dissociation constant of Mn complexes with Ox2− are 1.5 × 10−4 for MnII(Ox2−) and 2.72 × 10−17 for MnIII[(Ox2−)2].40

Then the logarithm of Kox/Kred will be equal to 12.75 that [according to eqn. (II)] gives ΔE0 equal to 0.75 V and E0com equal to 0.76 V. This (calculated) value agrees with the value of E0 obtained from our experimental work (0.77 V).

We conclude that the electrode reaction consists in this case of two steps:MnII + Ox2− ⇔ MnII(Ox2−)MnII(Ox2−) − e + Ox2− ⇔ [MnIII(Ox2−)2]The agreement with the literature data, as well as the strict linear dependence of EP on log CX, demonstrates the accuracy of our measurements and confirms the suitability of using the eqns. (I) and (II).

Upon the addition of acetate ions (Fig. 2, curve 2), the oxidation peak is also shifted to lower potentials.

A linear dependence of the potential versus the logarithm of CH3COONa concentration is observed over the whole range of acetate concentration from 1 × 10−3 M to 0.15 M (Fig. 3, curve 2).

The slope is equal to 120 mV which by eqn. (I), gives p = 2 for n = 1 and thus corresponds to a 1∶2 ratio of Mn∶acetate, formulated as [MnIII(CH3COO)2]+:MnIIe + 2CH3COO → [MnIII(CH3COO)2]+.Extrapolation of the potential dependence to log CX = 0 gives E0 = 0.69 V for this process and enables calculation of the stability constant for acetate binding, Kst = 7.91 × 1013 at T = 21.5 °C.

Essentially identical behaviour is seen for oxidation of MnII in the presence of formate.

In this case, the shift of EP is characterized by a linear dependence on the logarithm of the formate concentration between 3 × 10−3 and 1.6 × 10−1 M, as shown in Fig. 3 (curve 3).

The slope is equal to 131 ± 5 mV, which according to eqn. (I) gives p ≅ 2 at n = 1, corresponding to a reaction stoichiometry of 1∶2:MnIIe + 2HCO2 → [MnIII(HCO2)2]+.Extrapolation of the data to log CX = 0 gives E0 = 0.775 V for MnIII(HCO2)2+ and a corresponding stability constant Kst equal to 2.86 × 1012.

A different behaviour is seen for the dependence of the voltage–current curves for MnII oxidation in the presence of NaHCO3 (Fig. 4).

At first (at concentration of NaHCO3 higher than 6 × 10−3 M) the peak of MnII oxidation is shifted to lower potentials (Fig. 4, curves 2–4), which is analogous to the data obtained with acetate and formate.

However, at NaHCO3 concentrations higher than 2 × 10−2 M a new MnII oxidation wave appears around 0.70 V. It begins to emerge as a shoulder above 3 × 10−2 M and becomes the only peak observed at bicarbonate concentration of 8.3 × 10−2 M (Fig. 4, curves 4–8).

Upon further increase of NaHCO3 (0.1 M and higher) the shift of the oxidation potential of MnII is still observed, however, the height of the oxidation wave begins to decrease progressively due to the precipitation of a white suspension of MnCO3.

Fig. 3 (curves 4 and 5) shows the dependence of the MnII oxidation potential on the logarithm of bicarbonate (BC) concentration.

The dependence reveals three linear components with different slopes reflecting participation of as many as three precursors in the oxidation.

We discuss these in sequence.

(i) The linear dependence of the MnII oxidation potential on log CX at concentration higher than 3 × 10−3 M, NaHCO3 (Fig. 3, curve 4) reflects the oxidation of free aqua MnII (in analogy with the other anions) since it extrapolates back to the free MnII potential.

The slope of this dependence is equal to 173 ± 5 mV, in contrast to the slope for oxalate, acetate and formate, and thus by eqn. (I) corresponds most closely to p ≅ 3 for n = 1.

This ratio suggests participation of three HCO3 anions in this electrode reaction:MnII − e + 3HCO3 ⇔ MnIII(HCO3)3E0 is equal to 0.67 V for this reaction (determined by extrapolation to log C = 0).

This value gives a stability constant for MnIII(HCO3)3Kst = 1.73 × 1014 [T = 21 °C].

It is clearly seen (Fig. 3, curve 4) that the plot of the dependence of MnII oxidation potential on log CX is free of any breaks, thus indicating that a single species is detected, formulated as the electro-neutral species MnIII(HCO3)3.

(ii) The voltage–current curves in Fig. 4 show that a second precursor is revealed (as a “new” wave of oxidation) at increasing bicarbonate concentrations that is more easily oxidized (appears at less positive potential) than the previous species.

The experimental dependence of the oxidation potential of this new wave on log CX (from −1.4 to −1.08) reveals a linear region with slope equal to 60 ± 2 mV (Fig. 3, curve 5), suggesting the involvement of one additional bicarbonate ion per electron in the oxidation of this new complex.

Extrapolation of the data in Fig. 3 (curve 5) to log CX = 0 gives the E0 = 0.61 V (complex I).

(iii) Upon further increase of the NaHCO3 concentration (to 0.1 M and higher) there is a break in the dependence of EP on log CX (Fig. 3, curve 5) with a linear slope equal to 120 ± 10 mV, suggesting the involvement of two additional bicarbonate ions per electron in the oxidation of this new complex.

In this case, the slope extrapolates to E0 = 0.52 V at log CX = 0.

The appearance of this new section is evidently related to oxidation of another MnII-BC complex (complex II).

Analysis of the dependence of oxidation EP on log CX for both MnII-BC complexes (using eqn. (II)) is complicated since it requires knowledge of the stability constants of the initial MnII complexes.

In the literature the values represented for the stability constant, Kst, of complex MnII(HCO3)+ are different: 18628. and 63..129

Sychev et al30,31. reported the formation of two complexes, MnII(HCO3)+ and MnII(HCO3)2 having stability constants equal to 11 and 40.7, respectively.

This considerable uncertainty in the values of Kst requires further work.

A similar dependence of EP on log CX is observed, if the titration is performed at a constant concentration of bicarbonate while changing the MnII concentration (Fig. 5).

In this case at fixed bicarbonate concentration, the wave associated with complexes I and II appears first followed by formation of the MnaqII oxidation wave at higher MnII concentration, e.g., in reverse sequence to that observed in Fig 4.

After recording the background curve (Fig. 5, curve 0), 0.05 M NaHCO3 was added and the background in the presence of NaHCO3 was again measured (curve 01).

The titration shows the step-by-step addition of MnSO4 (from 55 μM to 1 mM).

Under these experimental conditions, the wave of the most stable Mn–bicarbonate complex at EP = 0.56 V appears first (curve 1).

The amplitude of this wave is linearly increased upon the increase of the MnII concentration.

Analysis of the EP shift for curves 1–6 yields a linear slope ΔE/Δlog CMn of 120 mV, followed by (curves 7–9) a shallower slope, ΔE/Δlog CMn approximately equal to ca.

60 mV.

At this concentration, one sees that a new current wave also appears at a potential of ca.

0.9 V. This peak shifts to more positive potentials with increasing concentrations of MnII (curves 7–12) with a slope ΔE/Δlog C equal to 180 mV.

In this region (curves 7–12) MnagII is oxidized by the reaction:MnII − e + 3HCO3 ⇒ MnIII(HCO3)3.The observed equilibria are the same as found in Fig. 4.


The data presented herein show that only bicarbonate but not carboxylate ligands (acetate, formate) is capable of stimulating the photooxidation rate of MnaqII by Mn-depleted apo-WOC-PSII.

This stimulation occurs using MnII concentrations equal to the native stoichiometry of 4 Mn/PSII and over a 1000-fold range in ligand concentrations.

The electrochemical characteristics of MnII oxidation offer a possible explanation for the stimulation of the photooxidation rate and its selectivity for bicarbonate.

The oxidation potential for MnIII/MnII decreases from 1.18 V for MnaqII (in 0.1 M LiClO4, pH independent in the region pH 5.0–pH 8.5) to less positive values upon formation of MnIII complexes with acetate, formate, oxalate and bicarbonate, respectively:MnIIe + 2CH3COO ⇔ [MnIII(CH3COO)2]+E0 = 0.69 V, (Kst = 7.9 × 1013)MnIIe + 2HCO2 ⇔ [MnIII(HCO2)2]+E0 = 0.775 V, (Kst = 2.9 × 1012)MnIIe + Ox2− ⇔ [MnIII(Ox2−)]+E0 = 0.92 V, (Kst = 1.0 × 1010)MnIIe + 3HCO3 ⇔ MnIII(HCO3)3E0 = 0.67 V, (Kst = 1.7 × 1014)In the bicarbonate stimulated oxidation process the stoichiometry of 3 bicarbonates reflects the total number of bicarbonate molecules involved in the process and thus includes several potential MnIII species differing in both ligation and deprotonation of bicarbonate and hydroxide.

Only bicarbonate exhibits two additional electrochemical processes corresponding to two more easily oxidizable MnII-BC complexes (denoted complexes I and II):MnII − BCcomplex Ie + HCO3 ⇔ MnIII(HCO3)3E0 = 0.61 VMnII − BCcomplex IIe + 2HCO3 ⇔ MnIII(HCO3)3E0 = 0.52 VThe formation of these MnII complexes with bicarbonate results in a maximum decrease of the oxidation potential of MnII that is probably important for their redox interaction with apo-WOC-PSII.

These complexes form at bicarbonate concentrations above 0.025 M, which is close to the upper concentration region, we could explore, owing to the onset of precipitation of MnCO3 at 0.1 M. It is possible that these MnII–BC complexes represent oligomeric Mnn(BC)m clusters.

However, the instability of the solutions to precipitation did not allow a wider study of the dependence on Mn and bicarbonate concentrations.

In case of acetate and formate, a positively charged complex, MnIIIL2, forms upon oxidation of MnaqII.

Only for bicarbonate does the neutral complex MnIII(HCO3)3 form upon oxidation of MnaqII..

The absence of breaks in the plot of EPvs. log of the concentration of acetate and formate shows that, in spite of the possibility of formation of MnII-complexes with the added ligands (log Kst for the MnII–(CH3COO)+ is equal to 1.2),40 there is no additional wave related to the oxidation of the MnIIL complexes.

These results suggest that these complexes are either not electroactive or are not stable in solution at the surface of the positively charged electrode.

For MnII-complexes with acetate, formate and bicarbonate the value of the stability constant is known to be ca.


The large difference in the stability constants for MnIIvs. MnIII complexes is a clear indicator of large differences in the kinetics of binding/dissociation and thus will be of importance for understanding the dynamics of Mn photooxidation by the WOC.

This difference may explain the considerable variations in the bicarbonate concentration (from μM to mM) measured for optimal reconstitution of electron transfer rates from MnII to apo-WOC-PSII.14–22

These variations could reflect changes in the average oxidation state of manganese ions as they assemble to form the Mn4 cluster or upon the S-state transitions of the WOC.

The electrochemical data allow us to answer why the YZ˙ radical in PSII is capable of oxidizing MnII.

The E0 value for MnaqII is equal to 1.18 V, which is higher than the oxidation potential of the YZ˙ radical estimated to 0.95–1.0 V.41

We may conclude that MnII can be efficiently oxidized by YZ˙ only in the presence of ligands capable of shifting the MnII oxidation potential to a lower value and thus stabilizing the MnIII oxidation state.

Nevertheless, the effect of these anions on electron transfer from MnII to apo-WOC-PSII is quite different.

Acetate and formate are not active in stimulation of electron transfer from MnII to apo-WOC-PSII, although both anions considerably lower the reduction potential of MnIII/MnII.

Bicarbonate efficiently stimulates the process of MnII-dependent electron transfer to apo-WOC-PSII, and is also known to reverse the formate-induced inhibition of electron transport from MnII and O2 evolution of functionally active PSII with high specificity.14–18

Thus, bicarbonate is special and presumably binds together with MnII to apo-WOC-PSII forming a ternary complex that is capable of efficient photooxidation of additional MnII.

This ternary complex may form starting either with MnaqII or the MnII-BC complexes I and II as precursors.

However, the current results have not identified which one or all of these precursors is involved, nor the structure of these complexes.

One may consider the following differences as possible reasons for the advantage of bicarbonate vs. carboxylate anions in stimulating the photooxidation rate of MnII by apo-WOC-PSII:

(i) Only bicarbonate forms an electro-neutral complex MnIII(HCO3)3, while the other ligands form charged complexes: [MnIII(HCO2)2]+, [MnIII(CH3COO)2]+.

Net charge exerts influence on the local concentration of the MnII precursor near PSII.42

(ii) Since photooxidation of MnII with YZ˙ in apo-WOC-PSII releases a proton on the first and subsequent steps,42,43 ligands such as acetate, formate or oxalate, which have a low pK 2–4, may not be suitable for exchanging H+ with the medium at pH 6.5 (used in the PSII experiments), versus the higher proton affinity of bicarbonate (pKa = 6.4).

Bicarbonate is also amphoteric having both acidic and basic properties (carbonate pKa = 10.3).

(iii) On the basis of our electrochemical results, the bicarbonate stoichiometry is 3 for the formation of the MnIII-bicarbonate complex.

The high Lewis acidity of MnIII (pKa ≈ 1)39 requires that MnIII(HCO3)3 will ionize one or more HCO3 ligands to form MnIII(HCO3)2(OH)1 + CO2 or MnIII(HCO3)(CO32−) + CO2 at pH 8.3 used in the present studies.

Formation of these complexes may stimulate both the kinetics of incorporation of MnII into the apo-WOC and the subsequent assembly of other MnII ions during reconstitution of the Mn4Ox core of the WOC.

Thermodynamic data also confirm this proposal.

The energy required to dissociate OH from bicarbonate is considerably smaller (8.9 kcal mol−1) than the energy for dissociation of OH from water (21.4 kcal mol−1).44

(iv) The maximum decrease of the potential of MnII oxidation (down to 0.52 V) is observed for MnII–BC complex II, which forms at the highest bicarbonate concentration studied (>0.1 M).

We previously suggested that MnII–BC complexes were similar to the major MnII species in the oceans of the Archean period (>2.2 BYA), where estimates of the bicarbonate concentration are considerably higher than for the contemporary ocean (10–104 mM).44

The electrochemical potentials of MnII–BC complexes I and II (0.61 and 0.52 V) indicate that they are thermodynamically capable of serving as electron donors to reaction centers of both green and purple photoautotrophic bacteria (E0 ∼ 0.26–0.55 V).45

Recently, it has been shown that Mn–bicarbonate complexes can serve as electron donor to bacterial reaction centers genetically modified to shift the redox potential of the primary electron donor to more positive potentials46.