Photosynthesis Research 72: 131–146, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.

131

Minireview

The structure and function of CP47 and CP43 in Photosystem II Terry M. Bricker∗ & Laurie K. Frankel Department of Biological Sciences, Division of Biochemistry and Molecular Biology, Louisiana State University, Baton Rouge, LA 70803, USA; ∗ Author for correspondence (e-mail: [email protected]; fax: +1-225-578-7258) Received 10 October 2001; accepted in revised form 14 January 2002

Key words: Photosystem II, CP47, CP43, chlorophyll-protein

Abstract This Minireview presents a summary of recent investigations examining the structure and functions of the Photosystem II chlorophyll-proteins CP47 and CP43, updating our previous review which appeared in 1990 (TM Bricker, Photosynth Res 24: 1–13). Since this time, numerous studies have clarified the roles of these chlorophyll-proteins within the photosystem. Biochemical, molecular and structural studies (electron and X-ray diffraction) have demonstrated the close association of these components with the photochemical reaction center of the photosystem and with the extrinsic oxygen evolution enhancer proteins.

Introduction CP47 and CP43 (CPa-1 and CPa-2), encoded by the psbB and psbC genes, respectively, are chlorophyllproteins which serve as the proximal antennae for Photosystem II (PS II), providing a conduit for excitation energy transfer from the exterior antennae of the photosystem (the intrinsic light-harvesting chlorophyllproteins in green algae and higher plants, and the extrinsic phycobilosomes in most cyanobacteria and red algae) to the reaction center core. In addition to their roles as proximal antennae proteins for the photosystem, both CP47 and CP43 appear to interact with proteins associated with the site of water oxidation. Three extrinsic proteins [33 (manganese-stabilizing protein), 24 and 17 kDa in higher plants and green algae and 33, 17 (cytochrome c550) and 12 kDa in cyanobacteria and the Rhodophyta] appear to act as enhancers of the oxygen evolution process. While these components are not absolutely required for oxygen evolution (Bricker 1992), they are necessary for high rates of water oxidation under physiological ionic conditions (Bricker and Ghanotakis 1996; Seidler 1996). Biochemical and genetic evidence indicates that CP47 and CP43 may help form the binding site(s) for these extrinsic proteins to PS II. This assertion appears to

be supported by the recent 3.8 Å crystal structure of the Synechococcus elongatus photosystem (Zouni et al. 2001). Additional functions for these proteins have been probed in site-directed mutagenesis studies that indicate that both proteins may be involved more directly in the oxygen-evolving process.

CP47 and CP43 structure In early studies, it was hypothesized that both CP43 and CP47 possess six transmembrane α helices (Vermaas et al. 1987; Bricker 1990). These helices were separated by five extrinsic loop domains (designated extrinsic loops A–E). The B and D loops were predicted to be exposed to the stroma while loops A, C and E were predicted to be exposed to the lumenal space. Sayre and Wrobel-Boerner (1994) confirmed this hypothesis for CP43. These authors utilized antipeptide antibodies which were generated against the N- and C-termini of the protein and extrinsic loops A, C, D and E. After limited trypsin hydrolysis and peptide mapping of thylakoid membranes, they concluded that CP43 contained six transmembrane helices, that the N- and C-termini were stromally exposed and that the extrinsic loops A, C and E were exposed to the

132

Figure 1. Topological models of CP47 and CP43. (A) Model of CP47 based on a consensus sequence for 66 species. (B) Model of CP43 based on a consensus sequence for 78 species. For both CP47 and CP43, sequences were obtained by a BLAST search and alignment of the GeneBank Database. Conserved amino acid residues are indicated by their single letter code; open circles, unconserved residues; closed circles, conservatively replaced hydrophobic residues; open squares, conservatively replaced hydrophilic residues, −, D or E; +, R or K.

133 lumen. Topological models for both CP47 and CP43 based on their consensus sequences are shown in Figure 1. These models are based on the analysis of 66 and 78 sequences, respectively. The results obtained by Sayre and Wrobel-Boerner (1994), at least with respect to the number of transmembrane α-helices, have been confirmed by electron diffraction (Rhee et al. 1998) and X-ray crystallographic studies (Zouni et al. 2001). The six-helix bundles (trimer of dimers) which have been proposed to arise from CP47 and CP43 flank the regions of electron density assigned to the D1–D2 heterodimer, with CP47 being located adjacent to D2 and CP43 being adjacent to D1. In these studies, PS II was observed to crystallize as a dimer exhibiting C2 symmetry. CP47 appears to be located near the interface of the two PS II monomers while CP43 is located at the periphery of the dimeric supercomplex. This arrangement could explain the differential solubilization properties of CP43 and CP47. A number of studies have shown that CP43 is much easier to remove from PS II than is CP47 (Akabori et al. 1988; Yamaguchi 1988; Takahashi and Satoh 1988; Ghanotakis et al. 1989). Additionally, the photosystem can assemble in the absence of CP43 albeit in greatly reduced quantities and with restricted function (Rogner et al. 1991). The proposed positioning of CP43 at the periphery of the dimeric PS II supercomplex could allow PS II assembly in its absence. The location of CP43 could also have biological consequences with respect to the turnover of the D1 protein. It is well known that the D1 protein is easily damaged and rapidly replaced during photoinactivation (for reviews see Prasil et al. 1992; Aro et al. 1993; Yamamoto 2001). Given the locations of CP43 and the D1 protein, it is possible to imagine mechanisms by which CP43 might be transiently removed from the complex, the damaged D1 replaced, followed by re-association of CP43 with the repaired PS II complex. Within the Zouni crystal structure (Zouni et al. 2001), 12 and 14 regions of electron density assigned to chlorophyll (chl) are associated with CP43 and CP47, respectively. This corresponds well to the number of histidyl residues located in the transmembrane regions of the proteins which had been hypothesized to serve as chlorophyll ligands (11 and 12 for CP43 and CP47, respectively) although it is unclear at this time if there is a 1:1 correspondence between histidyl residues and chl in these proteins (see below). Other residues (Asn and Gln), as well as water and backbone carbonyls, can also serve as chl ligands (Tronrud et al.

1986; Zuber et al. 1987). Earlier biochemical studies had suggested that there were either 10–12 chl (Tang and Satoh 1984; Barbarto et al. 1991) or 20–25 chl (de Vitry et al. 1984; Yamaguchi et al. 1988) associated with each of these proteins. Thus, the current crystal structure appears to confirm the conclusions of 10–12 chl of the former studies. It should be noted, however, that at the current resolution, not all of the chlorophyll molecules present may be unequivocally recognizable. At a 3.8 Å resolution, the polypeptide backbone in regions which do not exhibit regular and repeating structural elements (α-helix and β-sheet) cannot be traced. Consequently most of the structure of CP43 and CP47, including all of the extrinsic loop domains, cannot be identified at this time. In the published PS II structure, a large amount of unassigned electron density is present at the lumenal face of the complex. Analysis of the hydropathy plots of the major intrinsic proteins of the PS II complex (CP47, CP43, D1, and D2) allows one to predict that the majority (60–70%) of the unassigned electron density arises from the extrinsic loops A, C and E of the CP47 and CP43 proteins (Bricker and Ghanotakis 1996). Given the positioning of the electron density assigned to the extrinsic proteins (manganese-stabilizing protein and cytochrome c550), it seems likely that these components interact with the extrinsic loops of CP47 and CP43. This assertion is supported, at least with respect to the manganese-stabilizing protein, by a substantial body of biochemical and molecular data which has accumulated over the last decade.

CP47 biochemical studies: interaction of CP47 with the oxygen-evolving site A large body of evidence has now accumulated from the work of a number of laboratories which indicates that the large extrinsic loop E of CP47 (257W– 450 W in spinach) interacts closely with the oxygenevolving complex. Most of these studies have been performed on PS II membranes isolated from spinach. Early reports indicated that removal of the manganesestabilizing protein and the chloride-insensitive manganese associated with the oxygen-evolving site leads to a conformational change which exposes a domain of loop E (360P–391 S) to the monoclonal antibody reagent FAC2 (Bricker and Frankel 1987; Frankel and Bricker 1989). It was unclear, however, if this putative conformational change occurs in CP47 or in proteins closely associated with CP47 (such as D1 or D2).

134 Protein crosslinking studies Protein crosslinking data have documented the formation of crosslinked products between extrinsic loop E and the manganese-stabilizing protein. The homobifunctional reagent dithiobis(succinimidyl)propionate (DTSP) crosslinks lysyl residues which lie within 14 Å of each other. In early studies, the laboratories of I. Enami and our laboratory demonstrated a CP47-manganese-stabilizing protein interaction with this reagent (Enami et al. 1987; Bricker et al. 1988). The site of this crosslinking interaction involves the domain 418 K–438K of the extrinsic loop E of CP47 (Queirolo 1992). The location of the domain(s) on the manganese-stabilizing protein participating in this association were not identified. The water-soluble carbodiimide 1-ethyl-3-(3dimethylaminopropyl)carbodiimide (EDC) has also been used to investigate this interaction (Bricker et al. 1988; Enami et al. 1991). EDC crosslinks amino groups to carboxyl groups which are in van der Waals contact (Hackett and Strittmatter 1984). Peptide mapping of the crosslinked complex indicated that the domain 364 E–440D on CP47 was crosslinked to the domain 1 E–76K of the manganese-stabilizing protein (Odom and Bricker 1992). It should be noted that an additional domain on CP47 other than the large extrinsic loop E has also been implicated in manganese-stabilizing protein binding. Ohta et al. (1995) have suggested that additional EDC crosslinked products occur between the manganese-stabilizing protein and CP47. These authors demonstrated that residues 87 D and/or 94 E of extrinsic loop A of CP47 interact with the manganesestabilizing protein residues 60 K–130 K. We hypothesize that a number of points of interaction exist between CP47 (and possibly other PS II components) and the manganese-stabilizing protein and that this multiplicity of interactions yields the strong binding observed for manganese-stabilizing protein to PS II (Leuschner and Bricker 1996). Site-specific labeling and limited proteolysis studies In addition to these protein crosslinking studies, sitespecific labeling and limited proteolysis has been used to map the interaction between the extrinsic loop E and the manganese-stabilizing protein. The reagent N-hydrosuccinimidobiotin (NHS-biotin) labels amino groups on CP47 only in the absence of the manganesestabilizing protein (Frankel and Bricker 1990). Peptide

mapping studies (Frankel & Bricker, 1992) indicated that two biotinylated domains (304K–321K and 389 K–419 K) which lie within loop E are labeled in the absence of the extrinsic component. Limited proteolysis studies performed in the presence and absence of the manganese-stabilizing protein indicate that the manganese-stabilizing protein shields CP47 from proteolytic attack (Bricker and Frankel 1987; Hayashi et al. 1993). Studies performed by Hayashi et al. (1993) which utilized Lys C were particularly informative. These investigators found that both oxygen-evolving capacity and manganese-stabilizing protein rebinding decreased in parallel with proteolytic cleavage at 389K in the extrinsic loop E of CP47. They concluded that residues on CP 47 in the vicinity of 389 K were required for the binding of the manganese-stabilizing protein and oxygen evolution. These conclusions are strongly supported by site-directed mutagenesis studies (see below).

Post-translational modifications of CP47 It has recently been suggested that CP47 (along with D1 and/or D2) may contain quinocofactors that can be covalently modified with [14C]-labeled amines under reducing conditions (Ouellette et al. 1998; Anderson et al. 2000). These experiments were performed in spinach PS II membranes. The labeling did not take place in the presence of the 24 and 17 kDa extrinsic components. In the absence of the 24 and 17 kDa proteins, labeling occurred either in the presence or absence of the manganese cluster. Interestingly, the labeling which occurred in the presence of an intact manganese cluster was suppressed in the presence of chloride but not in the presence of sulfate. Removal of the manganese-stabilizing protein and manganese associated with the oxygen-evolving site led to an increase in the level of labeling. Additionally, increased levels of denaturation of the PS II complex led to increased labeling of the protein components, suggesting that the putative cofactors resided in a sterically restricted domain. Labeling of the three PS II components appeared to be at a 1:1 stoichiometry. Interestingly, the formation of stable labeled products was more favorable under illumination. This suggests that under illumination, an endogenous reductant was produced which stabilized the covalent modification or that light-induced conformational changes take place which facilitate the labeling reaction. It was hypothesized that these, as yet unidentified cofactors, may be

135 involved in electron transport at or near the oxygenevolving site. At this time, the locations of the putative quinocofactors have not been mapped on any of the labeled PS II components.

Site-directed mutagenesis studies on CP47 A number of laboratories have performed rather extensive site-directed mutagenesis studies on the psbB gene from Synechocystis sp. PCC 6803. These studies have principally examined domains and individual residues located in the large extrinsic loop E of the CP47 and putative chlorophyll ligands located in the transmembrane α-helices. A few sites in other locations in CP47 have also been probed by these techniques. Deletion mutants in the large extrinsic loop E Vermaas and coworkers have extensively probed the function of the extrinsic loop E by producing 13 partial deletion strains in which between 3 and 15 amino acid residues have been removed (Figure 2A). The phenotypes of these deletion mutants range from essentially unaffected (examples include (V392–Q394), (G333–I336), etc.) to obligately photoheterotrophic (examples include (I265–F268), (F311–N317), (D440–P447), etc.) (Eaton-Rye and Vermaas 1991; Haag et al. 1993; Gleiter et al. 1994, 1995). While the phenotypes of the obligately photoheterotrophic strains are difficult to interpret (see below), those of a number of moderately affected deletion mutants are intriguing. (A373–D380) and (R384– V392) both exhibit a marked decrease in the ability to bind the manganese-stabilizing protein (Gleiter et al. 1994). Additionally, these mutants exhibit altered thermoluminescence properties. The B- and Q-bands are upshifted by 7 ◦ C and 5 ◦ C, respectively. These results indicate an alteration in charge recombination properties of QB − S2 and QA − S2 , respectively. These mutants also exhibit rapid dark inactivation of the oxygen-evolving apparatus (Gleiter et al. 1995). While these authors did not explicitly determine a t1/2 for dark inactivation, their data indicate that it was between 30 min and 24 h. These results are very similar to those obtained in psbO strains lacking the manganese-stabilizing protein. Burnap and colleagues have demonstrated that psbO strains exhibit an 8– 9 ◦ C upshift in the B-band and 10–15 ◦ C upshift in the Q-band (Burnap et al. 1992). Additionally, psbO

dark- inactivates very rapidly (Burnap et al. 1996) with a t1/2 of 10 min (vs. 44 h for wild type). One interesting difference is that psbO strains cannot grow photoautotrophically at low calcium or chloride media concentrations (Philbrick et al. 1991). (A373–D380) and (R384–V392) do not exhibit this defect; however, photoactivation of these mutant strains (after dark inactivation) is accelerated by more than an order of magnitude in the presence of 10 mM CaCl2 (Gleiter et al. 1995) indicating a defect in the normal calcium and/or chloride interaction with PS II in these strains. Thus, the phenotypes of the deletion strains strongly resemble that of psbO but are not as extreme. These results, as well as those obtained for point mutations (see below), strongly suggest that the large extrinsic loop of CP47 acts as a binding site for the manganese-stabilizing protein. It should be noted that photoactivation of (A373– D380) and (R384–V392) was reported to proceed much more slowly in these strains than in wild type (Gleiter et al. 1995). psbO, however, has been reported to photoactivate much faster than wild type (Burnap et al. 1996). This apparent discrepancy seems to indicate a major difference in the CP47 deletion strains vis-à-vis psbO (Morgan et al. 1998). An alternative explanation, however, can be found upon examining the conditions used in the two photoactivation experiments. In the experiments with (A373–D380) and (R384–V392), the mutants and wild type were incubated in the dark for 48 h prior to photoactivation (Gleiter et al. 1995). The mutants were allegedly completely dark-inactivated by this incubation; however, given the extremely long t1/2 for dark inactivation of wild type strains [not reported in Gleiter et al. (1995); reported as 44 h in Qian et al. (1997)], it seems very likely that a substantial proportion of wild-type PS II reaction centers were not completely dark-inactivated and contained functional manganese clusters in these studies. Indeed, these authors state that period four oscillations were still discernable (although highly damped) even after 72 h of dark incubation. In the experiments with psbO, however, the control strains and the mutant were treated with 1 mM hydroxylamine, which extracts the manganese cluster. Under these conditions, nearly all of the PS II reaction centers from both the wild type and the mutant lacked functional manganese clusters. Thus, the two different photoactivation experiments are not comparable and the conclusion that the phenotypes of (A373–D380) and (R384–V392) are significantly different from

136

Figure 2.

137 psbO with respect to photoactivation is unwarranted (Morgan et al. 1998) if based on this evidence alone. It should be noted that the phenotypes of deletion strains must be interpreted with caution (as should the results from any site-directed mutagenesis experiment). Combinational mutagenesis which restored the length and photoautotrophy of the obligately photoheterotrophic strain (D440–P447) (Tichy and Vermaas 1998) indicated that none of the individual residues within this deletion were critical for function. The reverted mutants did exhibit high rates of photoinactivation and a number of them exhibited an increased chloride requirement for photoautotrophic growth. It was unclear, however, if these effects were due to alterations of residues within the deletion or to the interaction of these residues with nearby sites (such as 448 R, see below). Additionally, site-directed modification of conserved charged residues which lie within a number of the lethal deletions of extrinsic loop E have yielded negative results (Putnam-Evans et al. 1996). This indicates that either the conserved uncharged residues which lie within these deletions are important for photoautotrophy or that the deletion induces large alterations in the global architecture of CP47 leading to a loss of function of this protein or other interacting members of the PS II complex. We favor the latter hypothesis. Point mutations Identification of a putative binding site for the manganese-stabilizing protein The biochemical studies discussed previously provide strong evidence that the large extrinsic loop E of CP47 interacts with the manganese-stabilizing protein. Sitedirected mutagenesis studies also strongly support this hypothesis. The arginyl residues at positions 384 and 385 have been the most intensively examined. Positive charges are conserved at these positions in all species examined. Replacement of these residues either individually or in tandem leads to a significant loss

in oxygen-evolving capability, an alteration in the quantum yield for oxygen evolution, a slowing in the rise time of oxygen release during the S3 →[S4 ]→S0 transition and an increase in the S2 decay lifetime (Putnam-Evans and Bricker 1992; Putnam-Evans et al. 1996). The pattern of these phenotypes is remarkably consistent with wild type > R385G > R384G > RR384385GG > RR384385EE. Further studies (Qian et al. 1997) indicated that RR384385EE exhibited weak binding of the manganese-stabilizing protein (RR384385GG exhibited moderate binding) and a t1/2 for dark inactivation of 9 h (22 h for RR384385GG and 44 h for wild type). Importantly, deletion of the psbO gene within the RR384385EE background induced only a small additional loss of oxygen-evolving capacity. This indicated that the mutation 384 Arg385Arg→384Glu385 Glu and deletion of the psbO gene affect the same parameter(s) governing the reaction mechanism for oxygen evolution. Our current hypothesis is that the arginyl residues at positions 384 and 385 form a binding site for the manganese-stabilizing protein to PS II. Other binding sites probably exist. We speculate that the manganesestabilizing protein is weakly bound to PS II via these other sites in the RR384385EE mutant and that this weak binding yields the intermediate phenotypes observed for oxygen evolution, dark inactivation, oxygen release (wild type, 3 ms; RR384385EE, 8 ms; psbO, 17 ms), and other parameters. The proximity of 384,385RR to the deletion strains (A373– D380) and (R384–V392) clearly suggests that the phenotypes of these strains arise from alterations in manganese-stabilizing protein binding, probably at the 384 Arg385Arg site. Indeed, the phenotype of these deletion strains is virtually indistinguishable from that of RR384385EE with respect to oxygen evolution rates, assembly of functional PS II reaction centers, etc. (Putnam-Evans et al. 1996) (for a discussion of apparent differences in photoactivation, see above).

Figure 2. Lumenally exposed loops of CP47 and CP43. The amino acid sequences for Synechocystis 6803 are shown. A. CP47. Green brackets indicate unaffected-mildly affected deletion strains; pink brackets indicate moderately affected deletion strains and red brackets indicate deletion strains that are obligate photoheterotrophs. Colored residues indicate positions of site-directed point mutations; green, unaffected-mildly affected; pink, moderately affected; red, obligate photoheterotrophs. Red line indicates domains crosslinked to the manganese-stabilizing protein with EDC, dark blue line indicates domain crosslinked to the manganese-stabilizing protein with DTSP. Cyan line indicates domains shielded from NHS-biotin modification by the manganese-stabilizing protein. Green line indicates the position of the antigenic determinant for the monoclonal antibody FAC2, which becomes exposed upon removal of the manganese-stabilizing protein and the chloride-insensitive manganese at the oxygen-evolving site. Black arrow indicates the Lys-C cleavage site which prevents reconstitution of the manganese-stabilizing protein and inhibits oxygen evolution. B. CP43. Colored brackets and amino acid residues are as in A, above. A black arrow indicates trypsin cleavage site that prevents rebinding of the manganese-stabilizing protein and inhibits oxygen evolution.

138 Possible involvement of CP47 in chloride sequestration A number of site-directed mutants in the psbB gene exhibit alterations in their phenotypes upon exposure to low chloride concentration environments. The arginyl residue at position 448, which is conserved in all species examined, has also been examined intensively. Six mutants have been produced at this position: R448G, R448S, R448Q, R448K, R448W (Putnam-Evans and Bricker 1994; Wu et al. 1999; Bricker et al. 2001) and R448D (T.M. Bricker, unpublished observations). R448W and R448D are obligate photoheterotrophs, R448K can grow photoautotrophically but exhibits very low rates of oxygen evolution (10% of wild type) while R448G, R448S and R448Q grow well photoautotrophically and exhibit moderate oxygen evolution rates (30–50% of wild type). Interestingly, the phenotypes of R448G and R448S change dramatically when grown under chloride-limiting conditions (<20 µM Cl− ). Under these conditions, no photoautotrophic growth or oxygen evolution is observed and the mutants assemble few functional PS II reaction centers (Putnam-Evans and Bricker 1994). When R448S is grown at normal chloride concentrations (480 µM) and then placed in low-chloride medium, oxygen-evolving capacity is lost at the same rate as wild type (t1/2 16–17 min). Upon addition of saturating amounts of chloride (5 mm), however, R448S recovers its oxygen-evolving capability much slower than does wild type (t1/2 308 s vs. 50 s) (Bricker et al. 2001). This appears to indicate a defect in the low affinity, rapidly exchanging chloride binding site described by Lindberg and Andreasson (1996). It should be noted that the high-affinity, slowly exchanging site appears to act normally in this mutant. Additionally, under normal chloride conditions R448S exhibits a significant increase in the number of oxygen-evolving complexes in the S3 state and a nearly 10-fold increase in the S3 -state lifetime. Under chloride-limiting conditions, the mutant exhibits a six-fold increase in the number of oxygen-evolving complexes in the S2 state with a concomitant 4-fold increase in the S2 state lifetime (Bricker et al. 2001). Since these defects are very similar to those observed in RR384385EE strains and (A373–D380) and (R384–V392) mutants, the binding of the manganese-stabilizing protein was investigated. Unlike the RR384385EE, (A373–D380) and (R384–V392) mutants, however, R448S exhibits manganese-stabilizing protein binding which is comparable to that of wild type (Bricker et al. 2001).

Earlier, we had hypothesized that the manganesestabilizing protein participates in the formation of a sequestered domain for maintenance of chloride in the vicinity of the oxygen-evolving site (Bricker and Frankel 1998). The results which we have obtained for R448S appear to indicate that CP47 also participates in chloride sequestration. Identification of putative chlorophyll-binding ligands The histidyl residues located in the transmembrane αhelices are putative chl ligands and have been the targets of a number of site-directed mutagenesis studies. Shen et al. (1993b) and Shen and Vermaas (1994) have introduced a number of mutations in histidyl residues (9 His, 23 His, 26 His, 100 His, 114 His, 142 His, 157His, 201 His, 202 His, and 216 His; all of these are conserved except for 201 His) which are located in transmembrane helices I, II, III, and IV in the PS I-less/apc− strain of Synechocystis which lacks PS I and assembled phycobilosomes (Shen et al. 1993a). Several mutations produce only modest effects on PS II function when the histidyl target residue is replaced with a tyrosyl residue, which cannot serve as a chl ligand. H9Y, H26Y, H157Y, and H201Y exhibit 60–75% of the oxygen evolution rates of controls. Other mutations produced serious defects in the assembly and function of the photosystem. The H23Y and H114Y mutants do not assemble any detectable PS II reaction centers. H100Y, H202Y, and H216Y exhibit oxygen evolution rates 30–40% of control. When the histidyl residues at many of these sites are replaced by glutamyl or asparginyl residues, which can serve as chl ligands, the resulting mutants H23N, H114Q, H114N, and H216N exhibit less severe phenotypes, evolving substantially more oxygen and assembling more PS II reaction centers. Interestingly, the H23N mutant exhibits a 70% reduction in the intensity of the 695 nm fluorescence emission peak and this fluorescence peak exhibits a 2 nm blue shift. H114N exhibits similar characteristics. The 695 nm fluorescence emission peak is associated with PS II and is believed to arise from a low temperature trap associated with CP47 involving a small number of chlorophylls (van Dorssen et al. 1987). It is possible that the 695 nm emission band arises from chl ligated to 23 His and/or 114 His. The results obtained by these authors for all of the histidyl residues examined are consistent with the hypothesis that most of these residues are serving as chl ligands. The three conserved histidyl residues (455His, 466 His, and 469 His) in transmembrane helix VI have

139 also been examined (Eaton-Rye and Vermaas 1992). All of these are conserved in the species for which sequence information is available. When these were replaced with glutaminyl residues the mutants H455Q and H466Q were similar to wild type, while the H469Q mutant exhibits an 80% reduction in the number of assembled PS II reaction centers and a concomitant decrease in the rate of photoautotrophic growth. Other mutants at these locations (H455T, H455Y, H469Y, and the double mutant F430L,H466R; Wu et al. 1999) exhibit a complete loss of photoautotrophic growth. These results support the hypothesis that 455His and 466 His are probable chl ligands. However, since glutamine was unable to replace histidine at position 469 His, this residue may possess other as yet unidentified roles within the photosystem. Recently, an intergenic suppressor mutant of a number of the obligate photoheterotrophic strains at position 469 His has been isolated (I.Y. Huang, D.P. Lynch and J.J. EatonRye, in press). The presence of the suppressor restores significant rates of oxygen evolution, assembly of PS II reaction centers and photoautotrophic growth to these strains. The gene responsible for the suppression has not yet been identified. Other site-directed mutants in CP47 Mutations at positions 363Phe, 321Lys and 342 Gly have not been examined in detail. 363 Phe lies within the obligately photoheterotrophic mutant (G351–T365) (Eaton-Rye and Vermaas 1991). The mutant F363R exhibits impaired photoautotrophic growth and increased rates of photoinactivation. Additionally, under low chloride growth conditions photoautotrophy is lost. When F363R was placed in a psbO or a psbV background (but not psbU), photoautotrophy was lost even under chloride-sufficient conditions (Clarke and Eaton-Rye 1999). The mutant K321G also exhibits a chloride-sensitive phenotype, exhibiting loss in its ability to grow photoautotrophically, a decrease in oxygen evolution capacity and a decrease in the number of assembled PS II reaction centers when grown under chloride-limiting conditions (PutnamEvans and Bricker 1997). It should be noted, however, that 321 Lys is not absolutely conserved. The mutant G342D was identified in a screen of randomly generated mutants (Wu et al. 1999) and exhibits a decrease in rate of photoautotrophic growth and a 50% loss of oxygen-evolving capability. A few studies have also addressed possible functions of residues located in other domains of the

protein. Loop C is also predicted to be lumenally exposed and is predicted to span residues 162 Trp– 197 Gly in Synechocystis. Clarke and Eaton-Rye (2000) produced a number of short deletions in this domain. (G176–P180) produced a strain that is an obligate photoheterotroph that does not assemble significant quantities of PS II reaction centers. Other deletion strains in loop C have less serious defects. (S169–P171) and (Y172–G176) assemble about 20% of the PS II centers found in wild type while (E184–A188) and (F190–N194) assemble about 50%. (E184–A188) cannot grow photoautotrophically under low chloride conditions. Concomitant genetic removal of the manganese-stabilizing protein, cytochrome c550 or the 12 kDa protein exacerbated the phenotypes of (E184–A188) and (F190–N194). Earlier, point mutations had been produced at position 167 Trp (Elanskaya et al. 1994; Wu et al. 1996) which is conserved in all species examined. W167K cannot grow photoautotrophically, while the mutant W167S exhibits a 75% loss in oxygen-evolving activity, assembles few PS II reaction centers, and exhibits severely retarded photoautotrophic growth. This mutant is also very susceptible to photoinactivation. It should finally be noted that a number of laboratories have introduced polyhistidine tags onto either CP47 (Bricker et al. 1998; Reifler et al. 1998) or CP43 (Sugiura and Inoue 1999) to facilitate isolation of PS II complexes. These modified complexes can be isolated rapidly and in high yield in a single step following detergent solubilization of the cyanobacterial thylakoid membranes. Functionally, such PS II appears to be essentially unmodified, yielding high rates of steady-state oxygen evolution and exhibiting near normal fluorescence, flash oxygen yield (Li et al. 2000), thermoluminescence (Sugiura and Inoue 1999) and EPR characteristics (Boussac et al. 2000). Additionally, the S2 /S1 difference spectra has been shown to be unmodified by the polyhistidine tag (Noguchi and Sugiura 2000; Chu et al. 2001).

CP43 biochemical studies There have been relatively few biochemical studies performed on CP43. Early results indicated that removal of the CP43 protein by chaotropic reagents or detergent treatments (Akabori et al. 1988; Yamaguchi et al. 1988; Ghanotakis et al. 1989) yields a nonoxygen-evolving PS II preparation which additionally exhibits defects in the reduction of QA . These studies

140 were performed on PS II preparations isolated from higher plants. These results led to the suggestion that CP43 was involved in the stabilization of the QA binding site and was required for the formation of the oxygen-evolving complex (Petersen et al. 1990). However, molecular studies in cyanobacteria (see below) indicated that deletion of the psbC gene leads to the assembly of PS II reaction centers (albeit at low yield) with the capacity to reduce QA (Rogner et al. 1991). Thus, either the biochemical treatments used to remove CP43 from the higher plant PS II complex lead to the release of QA from the reaction center or cyanobacterial PS II centers have an intrinsically higher affinity for QA and can maintain the bound quinone in the absence of CP43. The hypothesis that CP43 is required for oxygen evolution has also been recently challenged. Büchel et al. (1999) have reported photoactivation of monomeric and dimeric CP43-less PS II core complexes isolated from spinach. These experiments were carried out under a flashing light regime using a highly sensitive oxygen electrode. These authors concluded that photoactivation in the absence of CP43 required significantly higher concentrations of calcium and that the photoactivated PS II centers were very susceptible to photoinactivation. Of the PS II complexes, 18% were functionally reconstituted, and it was reported that those PS II centers which were reconstituted exhibited a quantum efficiency of 96% of that observed for PS II-enriched membranes. While these results are intriguing, several questions remain. The proper control for studies examining the photoactivation of CP43-less core complexes are intact core complexes (containing CP43) and not the PS II membranes used in this study. Second, these authors did not demonstrate that the photoactivated CP43-less reaction centers exhibited period four oscillations indicative of oxygen evolution from PS II. Finally, it is well documented that damaged PS II reaction centers can generate H2 O2 from either the oxidizing or reducing sides of the photosystem (Fine and Frasch 1990; Schroeder and Aakerlund 1990; Klimov et al. 1993) and that oxygen can be released from H2 O2 by damaged PS II centers (Berg and Seibert 1987; Mano et al. 1987). Consequently, the authors must demonstrate that the oxygen evolution that they are observing is being produced by a functional oxygen-evolving complex and not from damaged centers oxidizing H2 O2 . Nevertheless, these studies do suggest that oxygen-evolving PS II reaction centers may be assembled in the absence of CP43.

A few studies have provided evidence that CP43 interacts with components of the oxygen-evolving complex. In an early study, it was shown that trypsin treatment of PS II membranes from which the manganese-stabilizing protein had been removed led to a rather specific cleavage of CP43 (Isogai et al. 1985). In a significantly more detailed study, it was demonstrated that tryptic cleavage occurred at 357 Arg in the large extrinsic loop E of CP43 in the absence but not in the presence of the manganese-stabilizing protein (Enami et al. 1997). The α-subunit of cytochrome b559 was also cleaved under these conditions. Reconstitution experiments indicated that rebinding of the manganese-stabilizing protein was lost in parallel to the cleavage at 357 Arg and/or cytochrome b559. This pattern of cleavage is remarkably similar to that observed to occur at 389 Lys in CP47 (Hayashi et al. 1993). Cleavage was also documented in the stromally-exposed C-terminus of CP43, specifically at residue 457 Lys, which, did not occur in NaCl-washed membranes. This result indicates that the removal of the manganese-stabilizing protein induced a conformational change in the stromally exposed C-terminal domain of CP43. Several studies have indicated that under conditions of photoinactivation CP43 can become covalently crosslinked to the D1 protein (Mori and Yamamoto 1992; Mori et al. 1995). Reconstitution studies indicated that the presence of the manganesestabilizing protein appears to inhibit the formation of the crosslinked product (Yamamoto and Akasaka 1995). These studies documented the close association of CP43 with the D1 protein and led to the assignment of CP43 to a position adjacent to the D1 protein in the current crystal structure (Zouni et al. 2001). Early studies indicated that a serine-type protease was responsible for D1 degradation during photoinhibition (Virgin et al. 1991, 1992). A number of reports have suggested that CP43 may act as the protease which cleaves the D1 protein during photoinactivation (Giacometti et al. 1992; Salter et al. 1992). Salter et al. (1992) demonstrated that CP43 could be labeled with the serine protease/esterase inhibitor [14 C]diisopropyl fluorophosphate and that this labeling inhibited the cleavage of D1 during photoinactivation at a stromally exposed cleavage site in the DE loop. These studies were carried out in higher plant oxygen-evolving PS II core preparations. The authors utilized a variety of different methods to assure that the labeled protein was indeed CP43. This is difficult, however, since even trace amounts of a

141 contaminating protease which comigrates with CP43 could yield the effects observed. Nevertheless, the data are reasonably convincing that the labeled component is indeed CP43. Unfortunately, the labeled serine was not identified in these studies. Since cleavage occurs on the stromal side of the membrane, it is expected that, if CP43 is acting as a protease, the labeled seryl residue would be located on stromally exposed domains of CP43, the N- and C-termini of the protein, or extrinsic loops B and D. Only a few seryl residues are possible candidates. In spinach, 33 Ser (N-terminal domain), 130 Ser (extrinsic loop B), 254Ser (extrinsic loop D), and 455 Ser (C-terminal domain) are predicted to be stromally exposed. Only 33 Ser is conserved in all species for which sequence information is available.

CP43 site-directed mutagenesis studies Deletion of the psbC gene leads to a loss of photoautotrophic growth and oxygen-evolving capacity (Vermaas et al. 1988; Rochaix et al. 1989). Nevertheless, these mutants do accumulate detectable quantities of PS II core components D1, D2 and CP47. Synechocystis mutants lacking CP43 due to the deletion of the psbC gene were observed to accumulate CP43-less PS II core complexes at low yield (about 10% of wild type) (Rogner et al. 1991). These investigators were able to isolate and characterize the CP43-less core complex and determined that it exhibited the ability to carry out primary charge separation and the reduction of QA , and could form Y+ Z . Neither the CP43-less core complex nor the CP43-containing core complex isolated from wild type contained any detectable QB . Additionally, these investigators examined the formation of Q− A under different intensities of flash illumination. From these light saturation studies they estimated that CP43 contributed about 30% of the light-harvesting capacity of the core complex. The introduction of short deletions in the large extrinsic loop E of CP43 has a profound effect on the assembly of functional PS II core complexes (Kuhn and Vermaas 1993) comparable to that observed upon deletion of the psbC gene. Eight short deletions (3– 12 residues) were introduced in the region between 298 Gln and 386 Ala, which lies roughly in the middle of this extrinsic loop domain. In contrast to the deletion studies performed in extrinsic loop E of CP47 (see above) in which a variety of phenotypes were observed ranging from mildly affected to obligate photoheterotrophy, all of the deletions produced in CP43 were

obligate photoheterotrophs. No oxygen evolution or electron transfer from Z to QA (as measured by electron transfer from diphenylcarbazide to dichlorophenolinolephenol) or assembly of PS II centers capable of binding significant quantities of [14C]diuron was observed. In most of the mutants, CP43, D2, D1 and CP47 could be detected immunologically, but it was unclear if these components were assembled into PS II reaction centers. It is clear from these studies that domains located in the large extrinsic loop E of CP43 are required for the functional assembly of oxygenevolving PS II reaction centers in vivo and that these domains appear to be more critical than the majority of similar domains located in the extrinsic loop E of CP47. Given the apparent intimate association of CP43 with the D1 protein (Mori and Yamamoto 1992; Mori et al. 1995; Zouni et al. 2001) we hypothesize that the interaction of the extrinsic loop E of CP43 with lumenally exposed domains of D1 may be required for the efficient assembly and stability of the oxygen-evolving complex. Only a few point mutations have been introduced into CP43. This is surprising given the extreme nature of the deletion mutations noted above. C. PutnamEvans’ laboratory has introduced a number of mutations in the large extrinsic loop E of CP43 (Knoepfle et al. 1999; Rosenberg et al. 1999). The alteration of arginyl residues at positions 305 Arg and 342Arg yielded quite interesting phenotypes (Knoepfle et al. 1999). These residues are conserved in all species which have been examined. Both the R305S and R342S mutants contain near normal amounts of the CP47, CP43, and D1 proteins as determined immunologically. These proteins appear to assemble into functional PS II reaction centers at moderate yields. Fluorescence measurements indicated that both mutants assemble 60–70% of the PS II centers capable of transferring electrons from hydroxylamine to QA . Nevertheless, R342S can not grow photoautotrophically, exhibits only 10% of the steady-state oxygen evolution rate of wild type when grown photoheterotrophically (<3% when grown mixotrophically), with this low residual oxygen evolution capacity being very susceptible to photoinactivation (t1/2 4 min vs. >12 min for wild type). R305S exhibits a considerably more robust phenotype, growing photoautotrophically and exhibiting oxygen evolution rates of about 70% of wild type. This mutant also exhibits a higher rate of photoinactivation than wild type (t1/2 7 min). Interestingly, this mutant exhibits a markedly more severe phenotype when grown under chloride-limiting conditions

142 (A. Young, T.M. Bricker and C. Putnam-Evans, in preparation). Mutations have also been introduced at positions 293 Glu, 339 Glu and 352 Glu in extrinsic loop E (Rosenberg et al. 1999). These residues are also conserved. E352Q cannot grow photoautotrophically, cannot evolve oxygen and cannot assemble reaction centers capable of transferring electrons from hydroxylamine to QA . This strain does accumulate significant quantities of CP47 and D1 but contains no immunologically detectable CP43. Overall, the phenotype of E352Q appears very similar to that of psbC strains. 352Glu appears to be required for the assembly and/or stability of the PS II reaction center. The mutant E339Q exhibits barely perceptible photoautotrophic growth and evolves oxygen at about 20% of wild type rates. This mutant appears to accumulate near normal amounts of the proteins CP43, CP47 and D1, and these assemble into functional PS II reaction centers, with nearly 100% of the reaction centers capable of hydroxylamine to QA electron transfer. Interestingly, less than half of these centers can use water as an electron donor to QA and the mutant exhibits a relative quantum yield for oxygen evolution of 12%. This mutant clearly exhibits a defect at the oxygen-evolving site and it is tempting to speculate that 339 Glu might be required for the stability and/or assembly of the manganese cluster. The mutant E293Q exhibits a relatively mild phenotype. It exhibits near normal photoautotrophic growth, evolves oxygen at 60% the rate observed in wild type and exhibits a relative quantum yield for oxygen evolution of 71%. Additionally, E293Q accumulates near normal amounts of CP43, CP47 and D1. It is interesting to note that the mutants R342S, E339Q and E352Q exhibit significantly more profound defects in PS II function and assembly than do any point mutations introduced into the large extrinsic loop of CP47. This result parallels the results obtained for the deletion strains in extrinsic loop E of CP43 (Kuhn and Vermaas 1993). We hypothesize that, while CP47 provides the major binding domain for the manganese-stabilizing protein, CP43 interacts more directly with the oxygen-evolving site of the photosystem. A number of the histidyl residues located in the transmembrane helices of CP43 have also been probed by site-directed mutagenesis (Manna and Vermaas 1997). 40 His, 105 His and 119His were changed to glutaminyl and tyrosyl residues. These three residues are located in transmembrane helices I and II and conserved in all CP43 sequences available. The mutants

H105Q, H119Q and H119Y are photoautotrophic while H40Q, H40Y and H105Y are obligate photoheterotrophs. The results obtained for 105 His are consistent with this residue serving as a chlorophyll ligand. The results are unclear for 119His since replacement with a tyrosyl residue has no major effect other than a decrease in the growth rate under photoautotrophic conditions. Mutations at 40 His yielded interesting phenotypes. H40Q, while exhibiting relatively high rates of oxygen evolution, is an obligate photoheterotroph. This may be due to significant alterations in fluorescence decay kinetics observed both in the presence and absence of dichloromethylurea, indicating defects both on the oxidizing and reducing sides of the photosystem. H40Y fails to assemble significant quantities of PS II reaction centers. It is, however, unclear if 40 His serves as a chlorophyll ligand. It should be noted that there is some confusion with respect to the numbering of the amino acid residues in CP43. This arises from several factors. First, there was an initial ambiguity in identifying the start codon for psbC translation (see Bricker 1990 for discussion). Some authors (Enami et al., 1997) still use the original numbering system (Alt et al. 1984) presented for spinach, rather than the correct translational start 12 codons upstream (Chisolm and Williams 1988; Golden and Stearns 1988; Carpenter et al. 1990). Synechocystis also exhibits a one amino acid deletion at position 7 vis-à-vis other species. Taking this into account has yielded the numbering system used in (Eaton-Rye and Vermaas 1991) and in subsequent papers from the Vermaas and Eaton-Rye laboratories. Finally, CP43 is N-terminally processed with 3 Thr being acetylated and phosphorylated. Some authors have identified this as 1 Thr in their numbering system (Knoepfle et al. 1999). In this communication, we have cited residue numbers as reported in the original papers. In Figure 2B, however, we have used the numbering system of (Eaton-Rye and Vermaas 1991). CP43 CP43 is the product of the isiA gene which was identified as one of the principal proteins induced under low iron stress conditions (Guikema and Sherman 1983; Laudenbach and Straus 1988; Burnap et al. 1993). The principal difference between CP43 and CP43 is the absence of a large extrinsic loop E in the latter protein where it is replaced by a much smaller extrinsic loop

143 domain. Recently, CP43 was found to form an alternative chlorophyll antenna for PS I under low iron stress conditions (Bibby et al. 2001; Boekema et al. 2001). Single particle image analysis indicates that eighteen CP43 molecules form a ring which surrounds the PS I trimer. Under low iron stress conditions, the synthesis of both phycobilosomes and PS I is reduced (Straus 1994). The CP43 ring antenna increases the photosynthetic efficiency of the PS I that is synthesized. The PS I-isiA complex exhibits a 44% increase in optical cross section over PS I alone (Boekema et al. 2001). Future directions Clearly our understanding of the structure and function of CP47 and CP43 has matured significantly since our original 1990 review. These proteins are intimately associated with the photochemical core of the photosystem, with the manganese-stabilizing protein, and probably the other extrinsic components of the oxygen-evolving complex. They may also interact more directly with the oxygen-evolving site of the photosystem. Further progress in the investigation of the structure and function of CP43 and CP47 will be greatly enhanced with the acquisition of a high resolution structure of the photosystem. This will allow the identification of putative protein interaction domains which can be probed by intelligent site-directed mutagenesis protocols. Additionally, possible interactions of these proteins with calcium, chloride and the manganese cluster would be defined. It is also clear, however, that some questions will not be directly addressable by structural studies. The modification of the D1–CP43 interaction during photoinactivation, the possible functioning of quinocofactors during electron transport, and the mechanism(s) which govern PS II complex stability will all be probed by insightful (hopefully!) biochemical investigations. It is hoped that in the next few years an even more detailed understanding of the roles of these important PS II components will be forthcoming. Acknowledgements Supported by grants from the National Science Foundation and the Department of Energy to T.M.B and L.K.F

References Akabori K, Tsukamoto H, Tsukihara J, Nagatsuka T, Motokawa O and Toyoshima Y (1988) Disintegration and reconstitution of Photosystem II reaction center core complex. I. Preparation and characterization of three different types of subcomplexes. Biochim Biophys Acta 932: 345–357 Alt J, Morris J, Westhoff P and Herrmann R (1984) Nucleotide sequence for the clustered genes for the 44 kD chlorophyll a apoprotein and the ‘32 kD-like’ protein of the Photosystem II reaction center in the spinach plastid chromosome. Curr Genet 8: 597–606 Anderson LB, Ouellette AJA and Barry BA (2000) Probing the structure of Photosystem II with amines and phenylhydrazine. J Biol Chem 275: 4920–4927 Aro E-M, Virgin I and Andersson B (1993) Photoinhibition of Photosystem II. Inactivation, protein damage and turnover. Biochim Biophys Acta 1143: 113–134 Barbarto R, Race HL, Friso G and Barber J (1991) Chlorophyll levels in the pigment binding proteins of Photosystem II. A study based on chlorophyll to cytochrome ratio in different Photosystem II preparations. FEBS Lett 286: 86–90 Berg SP and Seibert M (1987) Is functional manganese involved in hydrogen peroxide stimulated anomalous oxygen evolution in calcium chloride-washed Photosystem II membranes. Photosynth Res 13: 3–17 Bibby TS, Nield J and Barber J (2001) Iron deficiency induces the formation of an antenna ring around trimeric Photosystem I in cyanobacteria. Nature 412: 743–745 Boekema EJ, Hifney A, Yakushevska AE, Piotrowski M, Keegstra W, Berry S, Michel K-P, Pistorius EK and Kruip J (2001) A giant chlorophyll–protein complex induced by iron deficiency in cyanobacteria. Nature 412: 745–748 Boussac A, Sugiura M, Inoue Y and Rutherford AW (2000) EPR study of theoOxygen evolving complex in his-tagged Photosystem II from the cyanobacterium Synechococcus elongatus. Biochemistry 39: 13788–13799 Bricker TM (1990) The structure and function of CPa-1 and CPa-2 in Photosystem II. Photosynth Res 24: 1–13 Bricker TM (1992) Oxygen evolution in the absence of the 33 kDa manganese-stabilizing protein. Biochemistry 31: 4623–4628 Bricker TM and Frankel LK (1987) Use of a monoclonal antibody in structural investigations of the 49 kDa polypeptide of Photosystem II. Arch Biochem Biophys 256: 295–301 Bricker TM and Frankel LK (1998) The structure and function of the 33 kDa extrinsic protein of Photosystem II. A critical review. Photosynth Res 56: 157–173 Bricker TM and Ghanotakis DF (1996) Introduction to oxygen evolution and the oxygen-evolving complex In: Ort DR and Yocum CF (eds) Oxygenic Photosynthesis: The Light Reactions, Vol 4. Advances in Photosynthesis, pp 113–136. Kluwer Academic Publishers, Dordrecht, The Netherlands Bricker TM, Odom WR and Queirolo CB (1988) Close association of the 33 kDa extrinsic protein with the apoprotein of CPa-1 in Photosystem II. FEBS Lett 231: 111–117 Bricker TM, Morvant J, Masri N, Sutton H and Frankel LK (1998) Isolation of a highly active Photosystem II preparation from Synechocystis 6803 using a histidine-tagged mutant of CP 47. Biochim Biophys Acta 1409: 50–57 Bricker TM, Lowrance J, Sutton H and Frankel LK (2001) Alterations of the oxygen-evolving apparatus in a 448 Arg→448 S mutant in the CP 47 protein of Photosystem II under normal and low chloride conditions. Biochemistry 40: 11483–11489

144 Buchel C, Barber J, Ananyev G, Eshaghi S, Watt R and Dismukes C (1999) Photoassembly of the manganese cluster and oxygen evolution from monomeric and dimeric CP47 reaction center Photosystem II complexes. Proc Natl Acad Sci USA 96: 14288–14293 Burnap RL, Shen J-R, Jursinic PA, Inoue Y and Sherman LA (1992) Oxygen yield and thermoluminescence characteristics of a cyanobacterium lacking the manganese-stabilizing protein of Photosystem II. Biochemistry 31: 7404–7410 Burnap R, Troyan T and Sherman LA (1993) The highly abundant chlorophyll–protein complex of iron-deficient Synechococcus sp. PCC7942 is encoded by the isiA gene. Plant Physiol 103: 893– 902 Burnap RL, Qian M and Pierce C (1996) The manganese-stabilizing protein of Photosystem II modifies the in vivo deactivation and photoactivation kinetics of the H2 O oxidation complex in Synechocystis sp. PCC6803. Biochemistry 35: 874–882 Chisholm D and Williams JGK (1988) Nucleotide sequence of psbC, the gene encoding CP-43 chlorophyll a-binding protein of Photosystem II, in the cyanobacterium Synechocystis 6803. Plant Mol Biol 10: 293–201 Chu HA, Debus RJ and Babcock GT (2001) D1-Asp170 is structurally coupled to the oxygen evolving complex in Photosystem II as revealed by light-induced Fourier transform infrared difference spectroscopy Biochemistry 40: 2312–2316 Clarke SM and Eaton-Rye JJ (1999) Mutation of Phe-363 in the Photosystem II protein CP 47 impairs photoautotrophic growth, alters the chloride requirement, and prevents photosynthesis in the absence of either PS II-O or PS II-V in Synechocystis sp. PCC 6803. Biochemistry 38: 2707–2715 Clarke SM and Eaton-Rye JJ (2000) Amino acid deletions in loop C of the chlorophyll a-binding protein CP47 alter the chloride requirement and/or prevent the assembly of Photosystem II. Plant Mol Biol 44: 591–601 de Vitry C, Wollmann F-A and Delepelaire P (1984) Function of the polypeptides of the Photosystem II reaction center in Chlamydomonas reinhardtii. Biochim Biophys Acta 767: 415–422 Eaton-Rye JJ and Vermaas WFJ (1991) Oligonucleotide-directed mutagenesis of psbB, the gene encoding CP 47, employing a deletion strain of the cyanobacterium Synechocystis sp PCC 6803. Plant Mol Biol 17: 1165–1177 Eaton-Rye J and Vermaas W (1992) Characterization of a histidine to glutamine substitution at residue 469 in CP47 of Photosystem II In: Murata N (ed) Research in Photosynthesis, Vol I, pp 239– 242. Kluwer Academic Publishers, Dordrecht, The Netherlands Elanskaya IV, Allakhverdiev SI, Boichenko VA, Klimov V, Demeter S, Timofeev KN and Shestakov SV (1994) Photochemical characterization of cyanobacterium Synechocystis sp. PCC 6803 mutants with impaired Photosystem II proteins. Biochemistry (Moscow) 59: 929–934 Enami I, Satoh K and Katoh S (1987) Crosslinking between the 33 kDa extrinsic protein and the 47 kDa chlorophyll-carrying protein of the PS II reaction center core complex. FEBS Lett 226: 161–165 Enami I, Kaneko M, Kitamura N, Koike H, Sonoike K, Inoue Y and Katoh S (1991) Total immobilization of the extrinsic 33 kDa protein in spinach Photosystem II membrane preparations. Protein stoichiometry and stabilization of oxygen evolution. Biochim Biophys Acta 1060: 224–232 Enami I, Tohri A, Kamo M, Ohta H and Shen J-R (1997) Identification of domains on the 43 kDa chlorophyll-carrying protein (CP 43) that are shielded from tryptic attack by binding of the extrinsic 33 kDa protein with Photosystem II complex. Biochim Biophys Acta 1320: 17–26

Fine PL and Frasch WD (1990) The mechanism of hydrogen peroxide production by the S2 state of the oxygen-evolving complex In: Baltscheffsky M (ed) Current Research in Photosynthesis, Vol 1, pp 905–908. Kluwer Academic Publishers, Dordrecht, The Netherlands Frankel LK and Bricker TM (1989) Epitope mapping of the monoclonal antibody FAC2 on the apoprotein of CPa-1 in Photosystem II. FEBS Lett 257: 279–282 Frankel LK and Bricker TM (1990) Mapping of NHS-biotinylation sites and the epitope of the monoclonal antibody FAC2 on the apoprotein of CPa-1. In: Batcheffskey M (ed) Current Research in Photosynthesis, Vol I, pp 825–828. Kluwer Academic Publishers, Dordrecht, The Netherlands. Frankel LK and Bricker TM (1992) Interaction of CPa-1 with the manganese-stabilizing protein of Photosystem II: identification of domains on CPa-1 which are shielded from Nhydroxysuccinimide biotinylation by the manganese-stabilizing protein. Biochemistry 31: 11059–11063 Ghanotakis DF, de Paula JC, Demetriou DM, Bowlby NR, Peterson J, Babcock GT and Yocum CF (1989) Isolation and characterization of the 47 kDa protein and the D1–D2-cytochrome b559 complex. Biochim Biophys Acta 974: 44–53 Giacometti GM, Barbato R, Friso G, Frizzo A and Rigoni F (1992) Photosystem II degradation pathways after photoinhibition of isolated thylakoids. In: Murata N (ed) Research in Photosynthesis, Vol 4, pp 505–508. Kluwer Academic Publishers, Dordrecht, The Netherlands Gleiter HM, Haag E, Shen J-RS, Eaton-Rye JJ, Inoue Y and Vermaas WFJ (1994) Functional characterization of mutant strains of the cyanobacterium Synechocystis PCC 6803 lacking short domains within the large, lumen-exposed loop of the chlorophyll protein CP47 in Photosystem II. Biochemistry 33: 12063–12071 Gleiter HM, Haag E, Shen J-R, Eaton-Rye JJ, Seeliger AG, Inoue Y, Vermaas WFJ and Renger G (1995) Involvement of the CP47 protein in stabilization and photoactivation of a functional wateroxidizing complex in the cyanobacterium Synechocystis sp. PCC 6803. Biochemistry 34: 6847–6856 Golden SS and Stearns GW (1988) Nucleotide sequence and transcript analysis of three Photosystem II genes from the cyanobacterium Synechococcus sp. PCC7942. Gene 67: 85–96 Guikema J and Sherman LA (1983) Organization and function of chlorophyll in membranes of cyanobacteria during iron starvation. Plant Physiol 73: 250–256 Haag E, Eaton-Rye JJ, Renger G and Vermaas WFJ (1993) Functionally important domains of the large hydrophilic loop of CP 47 as probed by oligonucleotide-directed mutagenesis in Synechocystis sp. PCC 6803. Biochemistry 32: 4444–4454 Hackett CS and Strittmatter P (1984) Covalent crosslinking of the active sites of vesicle-bound cytochrome b5 and NADH cytochrome b5 reductase. J Biol Chem 259: 3275–3282 Hayashi H, Fujimura Y, Mohanty PS and Murata N (1993) The role of CP 47 in the evolution of oxygen and the binding of the extrinsic 33-kDa protein to the core complex of Photosystem II as determined by limited proteolysis. Photosynth Res 36: 35–42 Isogai Y, Yamamoto Y and Nishimura M (1985) Association of the 33 kDa polypeptide with the 43 kDa component in Photosystem II particles. FEBS Lett 187: 240–244 Klimov V, Ananyev G, Zastryzhnaya O and Wydrznski T (1993) Photoproduction of hydrogen peroxide in Photosystem II membrane fragments. Photosynth Res 38: 409–416 Knoepfle N, Bricker TM and Putnam-Evans C (1999) Site-directed mutagenesis of basic arginine residues 305 and 342 in the CP 43 protein of Photosystem II affects oxygen-evolving activity in Synechocystis 6803. Biochemistry 38: 1582–1588

145 Kuhn MG and Vermaas WFJ (1993) Deletion mutations in a long hydrophilic loop in the Photosystem II chlorophyll-binding protein CP43 in the cyanobacterium Synechocystis sp. PCC 6803. Plant Mol Biol 23: 123–133 Laudenbach D and Straus NA (1988) Characterization of a cyanobacterial iron stress-induced gene similar to psbC. J Bacteriol 170: 5018–5026 Leuschner C and Bricker TM (1996) Interaction of the 33 kDa extrinsic protein with Photosystem II: rebinding of the 33 kDa extrinsic protein to Photosystem II membranes which contain four, two, or zero manganese per Photosystem II reaction center. Biochemistry 35: 4551–4557 Li Z-L, Bricker TM and Burnap R (2000) Kinetic characterization of His-tagged CP47 Photosystem II in Synechocystis sp. PCC6803. Biochim Biophys Acta 1503: 350–356 Lindberg K and Andreasson L-E (1996) A one-site, two-state model for the binding of anions in Photosystem II. Biochemistry 35: 14259–14267 Manna P and Vermaas W (1997) Mutational studies on conserved histidine residues in the chlorophyll-binding protein CP43 of Photosystem II. Eur J Biochem 247: 666–672 Mano J, Takahashi M and Asada K (1987) Oxygen evolution from hydrogen peroxide in Photosystem II: flash induced catalytic activity of water-oxidizing Photosystem II membranes. Biochemistry 26: 2495–2501 Morgan T, Shand J, Clarke S and Eaton-Rye J (1998) Specific requirements for cytochrome c-550 and the manganese-stabilizing protein in photoautotrophic strains of Synechocystis sp. PCC 6803 with mutations in the domain 351 Gly to 436 Thr of the chlorophyll binding protein CP47. Biochemistry 37: 14437– 14449 Mori H and Yamamoto Y (1992) Deletion of antenna chlorophyll-abinding proteins CP43 and CP47 by tris-treatment of PS II membranes in weak light – evidence for a photo-degradative effect on the PSII components other than the reaction center-binding proteins. Biochim Biophys Acta 1100: 293–298 Mori H, Yamashita YTA and Yamamoto Y (1995) Further characterization of the loss of antenna chlorophyll-binding protein CP43 from Photosystem-II during donor-side photoinhibition. Biochim Biophys Acta 1228: 37–42 Noguchi T and Sugiura M (2000) Structure of an active water molecule in the water-oxidizing complex of Photosystem II as studied by FTIR spectroscopy. Biochemistry 39: 10943–10949 Odom WR and Bricker TM (1992) Interaction of CPa1 with the manganese-stabilizing protein of Photosystem II: identification of domains crosslinked by 1-ethyl-3-[3(dimethylamino)propyl]carbodiimide. Biochemistry 31: 5616– 5620 Ohta H, Yoshida N, Sano M, Hirano M, Nakazato K and Enami I (1995) Evidence for electrostatic interaction of the loop A on CP 47 with the extrinsic 33 kDa protein. In: Mathis P (ed) Photosynthesis: From Light to Biosphere, Vol II, pp 361–364. Kluwer Academic Publishers, Dordrecht, The Netherlands Ouellette AJA, Anderson LB and Barry BA (1998) Amine binding and oxidation at the catalytic site for photosynthetic water oxidation. Proc Natl Acad Sci USA 95: 2204–2209 Petersen J, Dekker JP, Bowlby NR, Ghanotakis DF, Yocum CF and Babcock GT (1990) EPR characterization of the CP47-D1–D2cytochrome b559 complex of Photosystem II. Biochemistry 29: 3226–3231 Philbrick JB, Diner BA and Zilinskas BA (1991) Construction and characterization of cyanobacterial mutants lacking the manganese-stabilizing protein of Photosystem II. J Biol Chem 266: 13370–13376

Prasil O, Adir N and Ohad I (1992) Dynamics of photosystem II: Mechanism of photoinhibition and recovery process. In: Barber J (ed) The Photosystems: Structure, Function and Molecular Biology, pp 295–348. Elsevier, Amsterdam Putnam-Evans C and Bricker TM (1992) Site-directed mutagenesis of the CPa-1 protein of Photosystem II: alteration of the basic residue pair 384,385 R to 384,385 G leads to a defect associated with the oxygen-evolving complex. Biochemistry 31: 11482–11488 Putnam-Evans C and Bricker TM (1994) Site-directed mutagenesis of the CP 47 protein of Photosystem II: alteration of the basic residue 448 R to 448 G prevents the assembly of functional Photosystem II centers under chloride-limiting conditions. Biochemistry 33: 10770–10776 Putnam-Evans C and Bricker TM (1997) Site directed mutagenesis of the basic residues 321 K to 321 G in the CP 47 protein of Photosystem II alters the chloride requirement for growth and oxygen-evolving activity in Synechocystis 6803. Plant Mol Biol 34: 455–463 Putnam-Evans C, Wu J, Burnap R, Whitmarsh J and Bricker TM (1996) Site-directed mutagenesis of the CP 47 protein of Photosystem II: alteration of conserved charged residues in the domain 364 E–444 R. Biochemistry 35: 4046–4053 Qian M, Al-Khaldi S, Putnam-Evans C, Bricker TM and Burnap RL (1997) Photoassembly of the Photosystem II (Mn)4 cluster in site-directed mutants impaired in the binding of the manganesestabilizing protein. Biochemistry 36: 15244–15252 Queirolo C (1992) Assemblage of spinach Photosystem II proteins: CPa-1 and MSP interactions. Dissertation, Louisiana State University, Baton Rouge, Louisiana Reifler MJ, Chisholm DA, Wang J, Diner BA and Brudvig GW (1998) Engineering and rapid purification of histidine-tagged Photosystem II from Synechocystis PCC 6803. In: Garab G (ed) Proceedings of the 11th International Congress on Photosynthesis, Vol 2, pp 1189–1192. Kluwer Academic Publishers, Dordrecht, The Netherlands Rhee K-H, Morris EP, Barber J and Kuhlbrandt W (1998) Threedimensional structure of Photosystem II reaction center at 8 Å resolution. Nature 396: 283–286 Rochaix JD, Kuchka M, Mayfield S, Schirmer-Rahire M, GirardBascou J and Bennoun P (1989) Nuclear and chloroplast mutations affect the synthesis or stability of the chloroplast psbC gene product in Chlamydomonas reinhardtii. EMBO J 8: 1013–1021 Rogner M, Chisholm DA and Diner B (1991) Site-directed mutagenesis of the psbC gene of Photosystem II: isolation and functional characterization of CP43-less Photosystem II core complexes. Biochemistry 30: 5387–5395 Rosenberg C, Christian J, Bricker T M and Putnam-Evans C (1999) Site-directed mutagenesis of glutamate residues in the large extrinsic loop of the Photosystem II protein CP 43 affects oxygen-evolving activity and PS II assembly. Biochemistry 38: 15994–16000 Salter AH, Virgin I, Hagman A and Andersson B (1992) On the molecular mechanism of light-induced D1 protein degradation in Photosystem II core particles. Biochemistry 31: 3990–3998 Sayre RT and Wrobel-Boerner EA (1994) Molecular topology of the Photosystem II chlorophyll a binding protein, CP43: Topology of a thylakoid membrane protein. Photosynth Res 40: 11–19 Schroeder WP and Aakerlund HE (1990) Hydrogen peroxide production in Photosystem II preparations. In: Baltscheffsky M (ed) Current Research in Photosynthesis, Vol 1, pp 901–904. Kluwer Academic Publishers, Dordrecht, The Netherlands

146 Shen G, Boussiba S and Vermaas WFJ (1993a) Synechocystis sp PCC 6803 strains lacking Photosystem I and phycobilisome function. Plant Cell Physiol 5: 1853–1863 Shen G, Eaton-Rye JJ and Vermaas WFJ (1993b) Mutation of histidine residues in CP47 leads to destabilization of the Photosystem II complex and to impairment of light energy transfer. Biochemistry 32: 5109–5115 Shen G and Vermaas W (1994) Mutation of chlorophyll ligands in the chlorophyll-binding CP47 protein as studied in a Synechocystis sp. PCC 6803 Photosystem I-less background. Biochemistry 33: 7379–7388 Straus NA (1994) Iron deprivation: physiology and gene regulation. In: Bryant DA (ed) Advances in Photosynthesis: Molecular Biology of the Cyanobacteria, Vol 1, pp 731–750 Kluwer Academic Publishers, Dordrecht, The Netherlands Sugiura M and Inoue Y (1999) Highly purified thermo-stable oxygen-evolving Photosystem II core complex from the thermophilic cyanobacterium Synechococcus elongatus having histagged CP43. Plant Cell Physiol 40: 1219–1231 Tang X-S and Satoh K (1984) Characterization of a 47-kilodalton chlorophyll-binding polypeptide isolated from a Photosystem II core complex. Plant Cell Physiol 25: 935–945 Tichy M and Vermaas W (1998) Functional analysis of combinational mutants altered in a conserved region in loop E of the CP47 protein in Synechocystis sp. PCC 6803. Biochemistry 37: 1523–1531 Tronrud DE, Schmidt MF and Matthews BW (1986) Structure and X-ray amino acid sequence of a bacteriochlorophyll-a protein from Prosthecochloris aestuarii refined at 19 Å resolution. J Mol Biol 188: 443–454 van Dorssen RJ, Breton J, Plijter JJ, Satoh K and van Gorkom H (1987) Spectroscopic properties of the reaction center and of the 47 kDa protein of Photosystem II. Biochim Biophys Acta 893: 267–274 Vermaas WFJ, Ikeuchi M and Inoue Y (1988) Protein composition of the Photosystem II core complex in genetically engin-

eered mutants of the cyanobacterium Synechocystis PCC 6803. Photosynth Res 17: 97–113 Vermaas WFJ, Williams JGK and Arntzen CJ (1987) Sequencing and modification of psbB, the gene encoding the CP 47 protein of Photosystem II in the cyanobacterium Synechocystis 6803. Plant Mol Biol 8: 317–326 Virgin I, Salter AH, Ghanotakis D F and Andersson B (1991) Lightinduced D1 protein degradation is catalyzed by a serine-type protease. FEBS Lett 287: 125–128 Virgin I, Salter AH, Hagman A, Vass I, Styring S and Andersson B (1992) Molecular mechanisms behind light-induced inhibition of Photosystem II electron transport and degradation of reaction center polypeptides. Biochim Biophys Acta 1101: 139–142 Wu J, Masri N, Lee W, Frankel LK and Bricker TM (1999) Random mutagenesis in the large extrinsic loop E and transmembrane αhelix VI of the CP 47 protein of Photosystem II. Plant Mol Biol 39: 381–386 Wu J, Putnam-Evans C and Bricker TM (1996) Site-directed mutagenesis of the CP 47 protein of Photosystem II: 167 W in the lumenally exposed loop C is required for Photosystem II assembly and stability. Plant Mol Biol 32: 537–542 Yamaguchi N, Takahashi Y and Satoh K (1988) Isolation and characterization of a Photosystem II core complex depleted in the 43 kDa chlorophyll binding subunit. Plant Cell Physiol 29: 123–129 Yamamoto Y (2001) Quality control of Photosystem II Plant Cell Physiol 42: 121–128 Yamamoto Y and Akasaka T (1995) Role of an extrinsic 33 kDa protein of Photosystem II in the turnover of the reaction centerbinding protein D1 during photoinhibition. Biochemistry 43: 9038–9045 Zouni A, Witt H-T, Kern J, Fromme P, Krauss N, Saenger W and Orth P (2001) Crystal structure of Photosystem II from Synechococcus elongatus at 38 Å resolution. Nature 409: 739–743 Zuber H, Brunisholz R and Sidler W (1987) Structure and function of light-harvesting pigment-protein complexes In: Amesz J (ed) Photosynthesis, pp 233–271. Elsevier, Amsterdam