Introduction

Photosystem II (PSII) is the membrane-associated redox enzyme whose active site consists of intrinsic and extrinsic proteins and three inorganic cofactors (four Mn, one Ca2+, and one Cl−). These ions form a cluster that is the site of the H2O oxidation reaction, which resides in the oxygen evolving complex (OEC). The OEC cycles through four light-driven advancements of its redox state, called the S-states (S n , where n = 0-4, and S1 is the dark stable state) (Nelson and Yocum 2006; McConnell 2008). The presence of both Ca2+ and Cl− in the vicinity of the Mn cluster is known to be essential for structure and function of the OEC (Miqyass et al. 2007; Popelkova and Yocum 2007; Yocum 2008). Calcium is a prominent cofactor found in the active site of a number of enzymes; its role in PSII has been extensively studied. The available data support both structural and functional roles in the OEC, in which Ca2+ likely functions as a binding site for substrate H2O and as a Lewis acid that along with Mn catalyzes the formation of the O–O bond (Pecoraro et al. 1998; Vrettos et al. 2001; McEvoy and Brudvig 2004). Calcium is required for advancement of the OEC through the S-states (Boussac and Rutherford 1988; Miqyass et al. 2008) and its binding is S-state dependent; Ca2+ is released from the OEC in the S2 oxidation state (Boussac and Rutherford 1988; Miqyass et al. 2008). On account of this phenomenon, PSII samples depleted of PsbP and PsbQ extrinsic subunits by salt washing exhibit low rates of O2 evolution activity in steady state assays. This is due to loss of the PSII-bound metal during turnovers of the S-states, even though salt-washed PSII samples initially retain some Ca2+ after removal of the extrinsic subunits (Miqyass et al. 2007, 2008; Ghanotakis et al. 1984a). Information on the position of Ca2+ in the OEC with respect to the Mn atoms in the cluster comes from the most recent X-ray crystal structure of PSII (Umena et al. 2011) and from earlier EXAFS experiments on PSII crystals (Yano et al. 2006). The latter study produced an estimate of a Mn–Ca distance of ~3.5 Å, while the former results produced a range of distances from Ca2+ to the individual Mn atoms (3.4–3.8 Å) that agree with the average Mn–Ca distance calculated from the EXAFS data.

Of the three extrinsic proteins associated with the function of the OEC (Roose et al. 2007; Enami et al. 2008), the largest is PsbO, or the manganese-stabilizing protein, which is found in PSII of all organisms, prokaryotic and eukaryotic, that have been examined thus far (Enami et al. 2008; Popelkova et al. 2003a; De Las Rivas et al. 2007; Williamson 2008). The protein is intrinsically disordered (Lydakis-Simantiris et al. 1999) and undergoes folding during functional assembly when it binds to PSII (Popelkova and Yocum 2011). Among numerous mutants of PsbO, some have been found to be defective in functional reconstitution into PSII, a process that is required for efficient H2O oxidation to occur. These mutations replaced R151, D157, or R161 in the large flexible loop of the protein (Popelkova et al. 2006, 2009), or caused a deletion that produced proteins lacking the first 15 or18 residues at the PsbO N-terminus (Popelkova et al. 2002a, 2003b).

Analysis of the X-ray crystal structures of cyanobacterial PSII indicates that the PsbO polypeptide harbors a binding site for an additional Ca2+ ion in PSII. According to the crystal structures, the PsbO-associated Ca2+ atom in the two cyanobacteria are in distinctly different locations in PsbO. The Glu54, Glu114, and His231 residues were proposed to ligate this Ca2+ in T. elongatus PsbO (Murray and Barber 2006), while Thr138, Asn200, and Val201 are ligands to the additional Ca2+ in T. vulcanus PsbO (Kawakami et al. 2011). In fact, PsbO-associated Ca2+ is one of the three additional Ca2+ ions that were found in cyanobacterial PSII at distances of 36–45 Å from the Mn4CaO5 cluster (Kawakami et al. 2011). In contrast, one Ca2+ per OEC and up to four Ca2+ ions per LHCII have been found in PSII isolated from spinach (Ädelroth et al. 1995; Grove and Brudvig 1998).

In addition to the requirement for Ca2+ in steady state assays created by removal of the PsbP and PsbQ extrinsic subunits from PSII (Ghanotakis et al. 1984a), removal of these polypeptides also increased exposure of the OEC to attack by added artificial reductants that generate Mn+2 and cause an inhibition of H2O oxidation (Ghanotakis et al. 1984b). It was suggested on the basis of these results that one function of PsbP and PsbQ is to protect the OEC from reduction by plastohydroquinone. Other experiments used artificial reductants as probes of the structure of the OEC to show that Mn atoms in the OEC are sensitive to reduction by large species, such as hydroquinone (H2Q) or N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD), as well as by the small reductant hydroxylamine (NH2OH) (Ghanotakis et al. 1984b; Kuntzleman et al. 2004). In the absence of the PsbP and PsbQ subunits, it was shown that Ca2+ binding to PSII retards NH2OH inhibition of O2 evolution activity by promoting retention of reduced Mn2+ in the OEC (Mei and Yocum 1991). Further characterization of the stabilizing function of Ca2+ in SW-PSII that employed the large hydrophobic reductant hydroquinone (H2Q) identified a unique Mn site in the Ca2+-reconstituted metal cluster that is not susceptible to H2Q but is sensitive to NH2OH reduction (Mei and Yocum 1992; Riggs et al. 1992). Taken together, these results obtained with SW-PSII preparations suggested that Ca2+ participates in stabilization and organization of manganese-binding ligands in the OEC. This hypothesis is consistent with current models for the structure of the OEC that are based on cyanobacterial crystal structures of the enzyme (Guskov et al. 2009; Umena et al. 2011), where Ca2+ and Mn atoms of the cluster share one bridging carboxyl and several bridging oxo ligands.

The advent of PSII crystal structures has also formed the basis of proposals concerning possible channels in PSII leading to and from the OEC (Guskov et al. 2009; Ho and Styring 2008; Umena et al. 2011). The ability of reducing agents of varying size and hydrophobicity to gain access to the OEC and react with PSII Mn atoms may be a reflection of the existence of these channels. Reductant access is, in some cases, regulated by Ca2+. In addition to the results on the effects of Ca2+ on hydroquinone reactivity with the OEC (Mei and Yocum 1992), it has been shown that the presence of the pair of methyl groups on N,N-dimethylhydroxylamine (DMHA) is sufficient to produce a Ca2+-specific block of access by this reductant to the Mn cluster (Kuntzleman et al. 2004; Kuntzleman and Yocum 2005). Table 1 presents the relative protective effect of Ca2+ along with other ions against inhibition of O2 evolution activity of PSII samples by DMHA. Because inhibition of activity by dark-incubation with DMHA was attenuated in samples containing OEC Ca2+ binding site surrogates (Sr2+, Cd2+), but not Mg2+ or Na+, the data in Table 1 demonstrate that the experiments on Ca2+-specific interference with Mn reduction by DMHA (Kuntzleman et al. 2004; Kuntzleman and Yocum 2005) probe the OEC-associated Ca2+ site, rather than additional Ca2+ sites associated with LHCII or other PSII subunits. In this respect, little or no interference by Ca2+ with reduction is observed with NH2OH, but retention of Mn2+ in or near the OEC is enhanced by the presence of Ca2+ (Mei and Yocum 1991).

Table 1 Protective effect of cations on loss of O2 evolution activity in SW-PSII
Full size table

The relationship between PsbO and the Ca2+ site in the OEC is still not completely understood (Popelkova and Yocum 2011). An earlier study showed that the presence of PsbO in PSII significantly lowers the requirement for Ca2+ in activity assays, which was interpreted to indicate that PsbO is essential for optimal association of the metal with the OEC (Bricker 1992). Recently, it has been shown that PsbO has no effect on Ca2+ retention under continuous illumination and that the Ca2+ K D in the steady state is about 0.15 mM for PSII samples reconstituted with mutated PsbO (Popelkova et al. 2006, 2008, 2009). The latter observation indicates that despite a fast exchange of Ca2+ at its site in higher S-states (Miqyass et al. 2008), the Ca2+ binding affinity during cycling of the OEC is relatively high. In the dark, Ca2+ at a high (10 mM) concentration in the incubation buffer protects the Mn cluster against inhibition by DMHA in the absence of PsbO (Popelkova et al. 2008). However, there are no experimental data to show that this would be the case at significantly lower concentrations of Ca2+. The study presented here carries out these experiments by examining PSII samples reconstituted with various mutants of PsbO whose ability to bind to PSII has been affected. Although the protective role of Ca2+ against dark-reduction of OEC Mn by an artificial exogenous reductant such as DMHA is not a physiological function of the metal in the OEC, the method allows one to probe the relationship between PsbO and the OEC-associated Ca2+ binding site. As is shown here, at low Ca2+ concentrations, a significant effect of PsbO on OEC-associated Ca2+ is revealed. These results are consistent with the hypothesis that PsbO may be required to facilitate the formation of the OEC-associated Ca2+ binding site (Loll et al. 2005).

Materials and methods

Preparation of site-directed mutants of PsbO and PSII membranes

The recombinant PsbO proteins used for the experiments presented here were characterized in previous studies and reported in (Popelkova et al. 2002a, b; 2003b; 2006; 2009). These references (see also Table 2) also provide details of their isolation and purification. Intact PSII, SW-PSII, and UW-PSII membranes were prepared as described in (Popelkova et al. 2002b). The use of urea-NaCl treatment to remove PsbO for these experiments has the added advantage of avoiding the addition of Ca2+ to PSII that might occur if PsbO had been extracted with 1 M CaCl2 (Ono and Inoue 1983).

Table 2 PsbO binding properties in SW-PSII and UW-PSII reconstituted with WT PsbO or the mutated PsbO proteins
Full size table

Reconstitution of UW-PSII with recombinant PsbO and incubation of PSII samples with EDTA and DMHA

PsbO-depleted PSII was reconstituted with PsbO for 1 h in the dark at 4°C in reconstitution buffer containing 37 mM MES (pH 6), 100 μg/ml BSA, 0.3 M sucrose (SigmaUltra, 99.5% GC, Sigma®), 2% betaine (w/v), and 40 mM Cl−. The Chl concentration was 200 μg/ml and 5 mol of PsbO/mol of PSII was used for reconstitution to insure optimal rebinding of the protein. SW-PSII was treated as reconstituted samples, except that no PsbO was added to the incubation mixture.

In the experiment designed to study the inhibitory effect of EDTA, the sample aliquots were incubated with various EDTA concentrations (0, 0.05, 0.1, 0.3, and 0.5 mM) in the presence (0.1 mM) or absence of Ca2+ at 4°C in darkness for 18 h. In the experiments comparing the ability of Ca2+ to protect the Mn cluster against inhibition using the method described in (Kuntzleman et al. 2004; Kuntzleman and Yocum 2005)), sample aliquots were incubated with EDTA (0.1 mM) and various Ca2+ concentrations (0.1, 0.3, 0.5, or 2.4 mM) in either the presence or absence of DMHA (10 mol/mol PSII) at 4°C in darkness for 18 h. Assuming an S1 oxidation state of 2 Mn3+/2 Mn4+ and about 1 μM reaction centers for 200 μg Chl/ml, this concentration of DMHA represents an amount of reductant that is sufficient to produce 4 Mn2+ per reaction center.

Activity assays

Steady state activities of samples were determined in the assay buffer (0.4 M sucrose, 50 mM MES (pH 6), 100 μg/ml BSA, 600 μM DCBQ as the electron acceptor, and saturating concentrations of Ca2+ (20 mM) and Cl− (100 mM). All O2 evolution assays were conducted in saturating light at 25°C. The fraction of centers inhibited by DMHA (Fig. 1) or EDTA (Fig. 2) was calculated as:

$$ F_{\text{I}} = \left( {A_{\text{C}} - A_{\text{R}} } \right)/A_{\text{C}} $$
(I)

where F I is the fraction of centers inactivated, A R is the activity of the sample aliquot incubated in the presence of DMHA or EDTA, and A C is the control activity of the identical sample aliquot incubated in the absence of DMHA or EDTA.

Fig. 1
figure 1

Effect of EDTA on DMHA inhibition of O2 evolution activity in SW-PSII and in UW-PSII reconstituted with WT PsbO. Samples were incubated at 4°C in darkness for 18 h without added Ca2+. The DMHA concentration was 10 mol/mol PSII and when present, the EDTA concentration was 0.1 mM. Columns depict averages of three experiments and vertical bars represent standard deviations

Full size image
Fig. 2
figure 2

Inhibition of PSII activity in SW-PSII and in UW-PSII reconstituted with WT PsbO by incubation with various EDTA concentrations at 4°C in darkness for 18 h in the presence or absence of Ca2+. No DMHA was present in incubation mixtures

Full size image

In the experiments to characterize interference by Ca2+ with a slow deactivation process (Table 3) or DMHA inhibition (Fig. 3), the fraction of centers protected by Ca2+ was calculated as:

$$ F_{\text{P}} = 1- \left( {A_{\text{C}} - A_{\text{R}} } \right)/A_{\text{C}} $$
(II)

where F P is the fraction of centers protected, A R is the activity of the sample aliquot incubated at the defined Ca2+ concentration in the presence (Fig. 3) or absence (Table 3) of DMHA, and A C is the control activity of the sample incubated in the absence of DMHA at the highest Ca2+ concentration (0.6 or 2.4 mM) that eliminated any inhibition due to Ca2+ deficiency.

Table 3 Effect of increasing Ca2+ concentration on O2 evolution activity in SW-PSII and in UW-PSII reconstituted with WT PsbO or the mutated PsbO proteins shown in the Table
Full size table
Fig. 3
figure 3

Effect of increasing Ca2+ concentration on the fraction of PSII centers protected against DMHA reduction in SW-PSII and in UW-PSII reconstituted with WT PsbO, D157E PsbO, R151G PsbO, ΔK14 M PsbO, or ΔT15 M PsbO. Samples were incubated with 0.1 mM EDTA at 4°C in darkness for 18 h. The DMHA concentration was 10 mol/mol of PSII. Equation II was used for calculation of the fraction of protected PSII centers. Points are the averages of three to four independent experiments and vertical bars at each point give standard deviations

Full size image

Results and discussion

Effect of residual Ca2+ on DMHA inhibition

A high concentration of Ca2+ (~10 mM) in a PSII sample can protect the Mn cluster against reductant attack and inhibition by DMHA (Kuntzleman et al. 2004; Kuntzleman and Yocum 2005), which reduces between three and four Mn, inhibits O2 evolution activity, and releases the Mn2+ from the OEC of Ca2+-depleted SW-PSII. Calcium blocks the reaction between DMHA and the PSII Mn cluster (Kuntzleman et al. 2004). While NH2OH and CH3NHOH will reduce the Mn cluster to inhibit O2 evolution in intact PSII preparations, the reaction between DMHA and the OEC requires removal of PsbP and PsbQ by salt washing to expose the site of reduction (Mei and Yocum 1993). In addition to removal of PsbP and PsbQ from PSII, salt washing creates a requirement for Ca2+ in steady state assays of activity. However, as noted in the “Introduction” section, PSII samples subjected to this treatment retain Ca2+ in the OEC active site that is released upon illumination (Boussac and Rutherford 1988; Miqyass et al. 2008). The extent to which this residual functionally bound Ca2+ can affect access of DMHA to the Mn cluster is illustrated by the results in Fig. 1, which show the effect of EDTA on the fraction of inhibited PSII reaction centers formed by the reaction with DMHA when no exogenous Ca2+ was added to incubation mixtures containing SW-PSII samples lacking PsbP and PsbQ. As can be seen, in the absence of EDTA only ~30% of reaction centers in SW-PSII and ~55% of centers in UW-PSII reconstituted with recombinant WT PsbO are sensitive to inhibition. In contrast, when 0.1 mM EDTA is present during the incubation period, more than 90% of the reaction centers in both PSII preparations are inhibited by exposure to DMHA. To determine if inhibition by EDTA alone, rather than an enhancement of reduction by DMHA, is the origin of the increase in the population of inactivated PSII centers, an independent experiment was carried out in the absence of DMHA in which the formation of inhibited PSII centers was assayed in the presence of several EDTA concentrations with or without added Ca2+ (Fig. 2). As this figure shows, the presence of EDTA alone at concentrations up to 0.5 mM in a sample fails to exhibit an inhibitory effect on O2 evolution activity of PSII preparations containing either native or recombinant WT PsbO in the presence or absence of Ca2+. When combined with the results in Fig. 2, the data in Fig. 1 suggest that 0.1 mM EDTA in the incubation mixture increased the fraction of reaction centers inhibited by DMHA by depletion of contaminating Ca2+ from the incubation buffer and residual Ca2+ from PSII. In the absence of EDTA, residual Ca2+ is a trace contaminant in, for example, sucrose (≤0.001% Ca2+ according to Sigma® for UltraPure grade sucrose; see also (Ädelroth et al. 1995)) in the incubation buffer, and as such can interfere with the reaction between DMHA and the Mn cluster. Addition of EDTA to the incubation buffer completely abolished this protective effect of residual Ca2+ and increased the population (to >90%) of PSII centers that are inactivated by DMHA (Fig. 1).

The results in Figs. 1 and 2 establish the efficacy of EDTA as a reagent for the removal of residual Ca2+ in the assay system used here that does not, by itself, cause an inhibition of PSII activity in samples containing WT PsbO. Therefore, all of the experiments reported below (Table 3; Fig. 3) were carried out in the presence of 0.1 mM EDTA in the incubation buffer to assure either total Ca2+ depletion in the sample or control of the Ca2+ concentration by addition of defined amounts of the metal to a sample. It can also be seen in the data of Fig. 1 that a UW-PSII sample reconstituted with WT PsbO exhibits a greater sensitivity to inhibition by DMHA than does a SW-PSII sample when EDTA is omitted from the incubation mixture. The significantly different sensitivity of the reconstituted sample to inhibition in the presence of limiting concentrations of Ca2+ could indicate that in vitro the reassembly of PsbO into PSII does not completely repair modifications to OEC-associated Ca2+ binding that are created when the protein is removed from PSII. If so, this would indicate that PsbO may regulate the ability of PSII-bound Ca2+ to protect the Mn cluster in the dark against inhibition by DMHA. To examine this putative role of PsbO in more detail, mutated PsbO’s with distinctly different PSII binding properties were used to reconstitute PSII complexes for further characterization with respect to the ability of Ca2+ to protect the OEC against inhibition by DMHA reduction.

Mutations of PsbO that affect PSII binding and/or functional assembly also affect the ability of Ca2+ to block inhibition of activity by reduction

Functional reconstitution of PsbO into PSII involves correct folding of the protein that occurs upon its initial binding to specific PSII binding sites (Popelkova and Yocum 2011). In the experiments described here, PsbO-depleted PSII was reconstituted with various PsbO mutants that are defective in binding and/or functional assembly. This procedure yielded five different samples. In addition to a control PSII sample containing two native PsbO subunits bound with a high affinity and functionally assembled in vivo (SW-PSII), the reconstituted samples were: (1) a sample where two recombinant PsbO subunits rebind with a high affinity, and functionally reassemble into PSII (UW-PSII reconstituted with WT PsbO), (2) a sample where the two recombinant PsbO subunits rebind with a high affinity, but exhibit a defect in functional assembly into PSII (UW-PSII reconstituted with D157E), (3) a sample where the two recombinant PsbO subunits rebind with a low affinity, and exhibit impaired functional folding into PSII (UW-PSII reconstituted with R151G), (4) a sample where one recombinant PsbO subunit rebinds with a high affinity, and folds into PSII (UW-PSII reconstituted with ΔK14 M), and (5) a sample where a single recombinant PsbO subunit rebinds with a low affinity, and exhibits defective functional assembly into PSII (UW-PSII incubated with ΔT15 M). Table 2 presents the PsbO binding properties for each sample along with the references to studies that characterize the corresponding PsbO mutants.

Prior to experiments to determine the extent to which Ca2+ can block reduction of the OEC by DMHA in each reconstituted sample (see Eq. II in the “Materials and methods” section), the O2 evolution activity of an identical sample incubated in the absence of DMHA was determined. Table 3 shows the effect of increasing Ca2+ concentration in the various DMHA-free incubation mixtures on the activities of various reconstituted PSII preparations. The initial values at 0 mM Ca2+ in Table 3 indicate that in the presence of 0.1 mM EDTA without added Ca2+, some activity is lost over the 18 h incubation period; a UW-PSII sample loses 100% of its activity under these conditions (data not shown). Table 3 shows that with increasing Ca2+ concentrations in DMHA-free incubation mixtures, the O2 evolution activities show modest increases, indicating that Ca2+ binding near the Mn cluster can hinder its slow deactivation. Based on previous results (Popelkova et al. 2008), this deactivation consists principally of loss of Mn from PSII centers. The greatest effect of Ca2+ is seen in the UW-PSII preparations reconstituted with ΔK14 M and ΔT15 M, while for the samples containing native or recombinant WT PsbO, this increase is negligible as reflected in the fraction of protected PSII centers given in Table 3. A possible explanation for this difference among PsbO proteins could be that in PSII reconstituted with mutated PsbO’s, the equilibrium between bound and free Ca2+ is shifted in favor of the free metal as opposed to PSII containing WT PsbO, where OEC-associated Ca2+ is present mostly in the bound form. Added Ca2+ can bind and confer some structural stability to the Mn4Ca cluster. This creates a more significant effect on the equilibrium between bound and free Ca2+ (and in turn on the activity) in PsbO mutant-reconstituted samples than it does in samples containing WT PsbO. In addition, some interference by added Ca2+ with limited inhibition of the OEC by 0.1 mM EDTA cannot be excluded in mutant-reconstituted PSII samples. It has to be emphasized that the effect of this slow destabilization/inhibition on the activity of the DMHA-free samples caused by incubation conditions with EDTA without Ca2+ is relatively minor when compared to the substantial loss of activity that is observed under similar conditions with DMHA. In the presence of DMHA without added Ca2+, about 2 to 17% of PSII centers retain activity depending on the PsbO protein used to reconstitute a PSII sample (see Fig. 3), as compared to between 43 and 89% of PSII centers that retain activity in the absence of DMHA without added Ca2+ (see Table 3). Moreover, SW-PSII is the most active and stable PSII preparation of all samples presented here and DMHA is able to inhibit this sample rapidly (6 mols of DMHA per mol PSII inhibit about 50% of SW-PSII centers without Ca2+ in the presence of EDTA after 4 h in darkness (data not shown)). All of these observations suggest that, under the conditions used here, DMHA inhibition predominates in inactivation of the Mn cluster over the much slower deactivation process caused by dark-incubation with EDTA alone. Thus, the 18 h results without Ca2+ shown in Fig. 3 must be due primarily to the action of DMHA.

Nevertheless, the results in Table 3 and especially in Fig. 3 show that even at low concentrations (0.1–2.4 mM) Ca2+ blocks inhibition of O2 evolution activity by reduction of the Mn cluster. As can be seen, the extent to which Ca2+ can block this reduction is affected by the binding properties of PsbO. Since inhibition by DMHA (Fig. 3) is robust relative to minor or negligible slow destabilization with EDTA alone (Table 3), the data in Fig. 3 allowed for a detailed analysis of the effect of PsbO stoichiometry, binding affinity, and the ability of PsbO to functionally assemble into PSII on the ability of Ca2+ to block inhibition of the OEC by DMHA. As Fig. 3 shows, 0.1 mM Ca2+ blocks DMHA access to ~70% reaction centers in SW-PSII, and to ~40–50% of centers in UW-PSII reconstituted with WT, D157E, R151G, or ΔK14 M PsbO. In comparison, in UW-PSII reconstituted with ΔT15 M, only ~20% of centers are protected by this Ca2+ concentration. This is consistent with data showing that ΔT15 M PsbO binds very weakly to PSII (Popelkova et al. 2003b). At a low (0.3 mM) Ca2+ concentration, the effects of PsbO mutations on the ability of Ca2+ to protect the Mn cluster against dark-inhibition by DMHA are most obvious. At this Ca2+ concentration, ~95% of reaction centers are protected in SW-PSII and in UW-PSII reconstituted with recombinant WT PsbO, while about 80, 70, or 60% of centers are protected by Ca2+ in UW-PSII reconstituted with ΔK14 M, D157E, or R151G PsbO, respectively, and only 50% of centers are protected in the sample reconstituted with ΔT15 M PsbO. Figure 3 also shows that 0.6 mM Ca2+ in the incubation mixtures is sufficient to obtain maximal Ca2+ protection of the Mn cluster against inhibition by DMHA in samples reconstituted with functionally assembled PsbO (i.e., native or recombinant WT PsbO, or mutated ΔK14 M PsbO). In comparison, samples reconstituted with PsbO that is defective in its functional assembly into PSII have about 80–90% of centers protected by 0.6 mM Ca2+, and a higher (~2.4 mM) concentration of Ca2+ is needed by these samples to observe maximum interference by Ca2+ with DMHA reduction of the OEC. Table 4 summarizes the data from Fig. 3 and further demonstrates the effect of PsbO on the ability of Ca2+ to block DMHA inhibition of the OEC. Using the averaged data in Fig. 3 produces the results presented in Table 4 that show that the Ca2+ concentrations required for 50 and 90% protection of the Mn cluster increase with the extent of the defect in PsbO binding and folding into PSII.

Table 4 Calcium concentrations required for 50 or 90% protection of the Mn cluster against inhibition by DMHA in SW-PSII and in UW-PSII reconstituted with WT PsbO or the mutated PsbO proteins shown in the Table
Full size table

Several conclusions can be drawn from the results in Fig. 3 and Table 4:

  1. (1)

    Native in vivo binding and assembly of WT PsbO, as in SW-PSII, enables Ca2+ protection against DMHA dark-reduction to a greater extent than in vitro functional reconstitution of recombinant WT PsbO into UW-PSII preparations (see Fig. 3a), in agreement with the results in Fig. 1.

  2. (2)

    The positive effect of PsbO on the ability of Ca2+ to block inhibitory reduction of the Mn cluster increases with increasing binding affinity of the protein to PSII (compare R151G with D157E PsbO in Fig. 3b, and ΔK14 M with ΔT15 M PsbO in Fig. 3c).

  3. (3)

    A single PsbO subunit that functionally reassembles and restores substantial OEC activity in PSII (see ΔK14 M PsbO, Fig. 3c) also produces greater Ca2+ interference with DMHA reduction than do two PsbO subunits that are defective in their functional folding into PSII (see D157E PsbO, Fig. 3b). This suggests that there is a very close relationship between the functionally assembled PsbO protein in PSII and the Ca2+ binding site in the OEC. Since the PsbO subunit cannot function as a Ca2+ ligand in the OEC (Miqyass et al. 2007; Popelkova and Yocum 2011; Umena et al. 2011), the data presented here indicate that PsbO may be required to facilitate the formation of the OEC-associated Ca2+ binding site. The protein appears to shift the binding equilibrium between free and bound Ca2+ in the OEC in favor of the bound metal. This conclusion is consistent with the results of the experiments on cyanobacterial PSII (Loll et al. 2005) that were interpreted to indicate that the Ca2+ binding site is created upon binding of PsbO to PSII, and with recent work by Bricker and Frankel (2008) that showed inefficient retention of Ca2+ in an A. thaliana mutant lacking PsbO-1, but not the PsbO-2 protein. The data presented here do not allow one to elucidate the exact mechanism by which PsbO affects the Ca2+ binding site in the OEC, but it can be hypothesized that PsbO may play a role in retention of Ca2+ in its site, for example, through positive effects on the Ca2+ ligation environment whose spatial structure might be regulated by PsbO binding to sites on the intrinsic subunits of PSII.

  4. (4)

    Finally, functional reconstitution of the second PsbO subunit into PSII has an additional positive effect on the ability of Ca2+ to protect the OEC against reduction (compare WT PsbO and ΔK14 M PsbO in Fig. 3c). This finding is analogous to the recent observation that the first PsbO subunit enhances retention of the Cl− cofactor, while the second PsbO subunit optimizes it in order to maximize O2 evolution activity (Popelkova et al. 2008). Thus, the data in Fig. 3c indicate that not only Cl−, but Ca2+ as well requires the presence of both PsbO subunits in eukaryotic PSII for its optimal function in the OEC. Moreover, these data suggest that the absence of PsbO, rather than an effect of urea, is responsible for modifications to the OEC-associated Ca2+-binding site that are created upon PsbO removal from PSII by urea and that allow DMHA to access and reduce the Mn cluster.

The interaction between DMHA, Ca2+ and the OEC can be represented by the following equation:

$$ \begin{gathered} {\text{PSIIMn}}_{ 4} {\text{Ca}}_{\text{B}} + {\text{ DMHA}} \rightleftarrows {\text{PSIIMn}}_{ 4} + {\text{Ca}}_{\text{F}}^{2 + } + {\text{DMHA}} \rightleftarrows {\text{PSIIMn}}_{ 4} {\text{DMHA}} + {\text{Ca}}_{\text{F}}^{2 + } \to {\text{PSIIMn}}_{ 3} + {\text{ Mn}}^{ 2+ } + \hfill \\ {\text{DMHA}}^{ \cdot } + {\text{ Ca}}_{\text{F}}^{2 + } , \hfill \\ \end{gathered} $$

where the B and F subscripts on Ca2+ refer to the OEC-bound and solution forms of the metal, and DMHA is the one-electron oxidized form of DMHA. The equilibrium between the intact OEC and the DMHA-sensitive OEC depends on the properties of PsbO bound to PSII. This reaction is presented schematically in Fig. 4. As can be seen, in the control sample containing WT PsbO, access of DMHA through a hypothetical channel leading to the PSII Mn4Ca cluster is blocked by Ca2+, because the properties of the environment around the Ca2+ binding site favor the bound form of the metal (see Fig. 4a). In contrast, the presence of one or two mutated variants of PsbO in PSII alters Ca2+ ligation and shifts the binding equilibrium toward free Ca2+. This results in an increased fraction of PSII centers with unoccupied Ca2+ binding sites, which in turn allows DMHA to access and reduce Mn atoms in the OEC. As a consequence, Mn2+ is released from the OEC (see Fig. 4b).

Fig. 4
figure 4

Model for regulation by PsbO of Ca2+ interference with DMHA reduction of the eukaryotic OEC. PsbO subunits are shown along with the OEC active site (Mn, Ca2+, Cl−) and a hypothetical channel in PSII intrinsic subunits that leads to the Ca2+ binding site. a Two PsbO subunits bound and functionally assembled into PSII regulate the OEC-associated Ca2+ binding site so that the equilibrium between free and bound Ca2+ favors the bound form, as indicated by arrows. This in turn restricts the access of DMHA (green dots) in the channel to the OEC, which blocks reduction of Mn. b The presence of only one functionally assembled PsbO subunit (pink) or one or two defective PsbO subunits (gold) in eukaryotic PSII alters Ca2+ ligation in the OEC. This shifts the equilibrium in favor of free Ca2+ (indicated by arrows). The unoccupied Ca2+ binding site allows DMHA to access and reduce the Mn cluster. Following reduction of the cluster, the Mn2+ is released from the OEC

Full size image

Conclusion

The results presented here show that PsbO affects the ability of Ca2+ to protect the Mn cluster against inhibition by reduction in the dark, and that functional reconstitution of both PsbO subunits into eukaryotic PSII is necessary for maximum protection. This indicates that PsbO may be required in PSII as one of the factors that facilitate the formation of the OEC-associated Ca2+ binding site. In Arabidopsis, it has been shown that the presence of PsbO is required for accumulation of PSII and formation of the OEC (Yi et al. 2005). The data presented here show that the PsbO protein also appears to regulate the dynamic equilibrium between bound and free Ca2+ in the OEC by enabling the metal to bind and be retained in its functional site in the redox active Mn cluster under steady state illumination.