Cyanobacterial Small Chlorophyll-binding Protein ScpD (HliB) Is Located on the Periphery of Photosystem II in the Vicinity of PsbH and CP47 Subunits*

Cyanobacteria contain several genes coding for small one-helix proteins called SCPs or HLIPs with significant sequence similarity to chlorophyll a/b-binding proteins. To localize one of these proteins, ScpD, in the cells of the cyanobacterium Synechocystis sp. PCC 6803, we constructed several mutants in which ScpD was expressed as a His-tagged protein (ScpDHis). Using two-dimensional native-SDS electrophoresis of thylakoid membranes or isolated Photosystem II (PSII), we determined that after high-light treatment most of the ScpDHis protein in a cell is associated with PSII. The ScpDHis protein was present in both monomeric and dimeric PSII core complexes and also in the core subcomplex lacking CP43. However, the association with PSII was abolished in the mutant lacking the PSII subunit PsbH. In a PSII mutant lacking cytochrome b559, which does not accumulate PSII, ScpDHis is associated with CP47. The interaction of ScpDHis with PsbH and CP47 was further confirmed by electron microscopy of PSII labeled with Ni-NTA Nanogold. Single particle image analysis identified the location of the labeled ScpDHis at the periphery of the PSII core complex in the vicinity of the PsbH and CP47. Because of the fact that ScpDHis did not form any large structures bound to PSII and because of its accumulation in PSII subcomplexes containing CP47 and PsbH we suggest that ScpD is involved in a process of PSII assembly/repair during the turnover of pigment-binding proteins, particularly CP47.

microarray data confirm that scp genes are expressed under various stress conditions, such as: HL, low temperature, hyperosmotic stress, salt stress, or the presence of inhibitors of photosynthetic electron transport (18 -22). Recent studies suggest that a sensory histidine kinase NblS of Synechococcus sp. PCC 7942 may regulate the expression of scp genes (23). Moreover, the NblS homologue in Synechocystis 6803, Hik33 (also called DspA), also controls the expression of scp genes in response to low temperature and osmotic stress (20,24).
Interestingly, the expression of scp genes is strikingly similar to that of Elip genes (13,15,25) with transient mRNA accumulation during exposure of a plant to a variety of stress conditions that result in the absorption of excess excitation energy (26 -28). The ELIPs have been indicated to function in photoprotection (26), most likely by functioning as transient pigment carriers during light stress-induced turnover of Chl-binding proteins (29). Like the ELIPs, the SCPs are localized in the thylakoid membranes and are important for photoacclimation during HL exposure (15,30). By analogy to ELIPs, it has been proposed that SCPs may function directly or indirectly in the dissipation of excess light energy (25,30,31). They could also serve as transient carriers of Chl (12) and modulate tetrapyrrole biosynthesis (32,33). All light stress-induced LHC relatives from plants studied so far have been found to be associated with photosystems (8, 34 -36).
Here we demonstrate that the ScpD protein tagged by the His 6 epitope is specifically associated with PSII. Immunogold labeling, followed by single particle analysis, identified the location of the ScpD protein to be at the periphery of the PSII complex. With the help of various PSII mutants, we have shown that ScpD protein interacts with the PSII proteins PsbH and CP47. These results are discussed in terms of the ScpD role in PSII.

EXPERIMENTAL PROCEDURES
Growth Conditions-Wild type (WT) and mutants (as described in Table 1) of Synechocystis 6803 were grown in liquid BG-11 medium (37) supplemented by 10 mM glucose at 30°C at 30 mol photon m Ϫ2 s Ϫ1 (in PSI containing strain) or 3 mol photon m Ϫ2 s Ϫ1 (in PSI Ϫ strain) on a rotary shaker. For HL treatment, cells were incubated at 500 mol photon m Ϫ2 s Ϫ1 .
Mutant Construction-The Synechocystis 6803 ScpDHis strain with ScpD protein tagged with the His 6 epitope (His tag) on its N terminus, expressed under psbA2 promoter, was constructed in a procedure analogous to that described in Ref. 38. The scpD gene was amplified by PCR using a mix of Taq and Pfu DNA polymerases and gene-specific primers with artificially generated restriction sites for NdeI and BamHI and containing 6 histidine codons (CAT) in the forward primer. After restric-tion, the PCR fragment was cloned into the NdeI and BamHI sites of the pSBA plasmid (39) containing the upstream and downstream regions of the Synechocystis 6803 psbA2 gene. The ligation mix was amplified by PCR using pSBA primers amplifying the scpDHis gene flanked by 0.5-kbp upstream and downstream regions of psbA2. This 1.2 kbp PCR product was directly used to transform the Synechocystis 6803 psbAII-KS strain, where the psbA2 gene was replaced by a kanamycin-resistance/ sacB cartridge (39). The sacB gene codes for levan sucrase, leading to sucrose sensitivity of this strain. After transformation, Synechocystis cells were grown on BG-11 plates for 4 days. The transformants were then transferred to plates with 5% sucrose and sucrose-resistant colonies were checked for kanamycin sensitivity. The resulting strain expressing both the wild-type and His-tagged forms of ScpD protein was transformed with chromosomal DNA from the ⌬ScpD strain (32). Proper insertion of the scpDHis gene was confirmed by DNA sequencing, and deletion of the WT copy of the scpD gene was confirmed by PCR.
The ScpDHis/⌬ScpD strain was transformed with chromosomal DNA from a PSI Ϫ cell to obtain the ScpDHis/⌬ScpD/ PSI Ϫ strain. The ScpDHis/⌬ScpD/PSI Ϫ strain was further transformed with chromosomal DNA from ⌬Cyt, ⌬CP47, and ⌬H cells (see Table 1). The deletions were confirmed by PCR.
Radiolabeling of the Cells, Membrane Preparation, and Protein Analysis Using Two-dimensional Blue Native/SDS, Two-dimensional Native Deriphat/SDS Electrophoresis, and Immunoblotting-Radioactive labeling of cells using a mixture of L-[ 35 S]methionine and L-[ 35 S]cysteine (Ͼ1000 Ci mmol Ϫ1 , Trans-label, ICN, final activity 400 Ci ml Ϫ1 ) and isolation of membranes was performed as described in Ref. 40. Isolated membranes were solubilized with n-dodecyl-␤-maltoside (DM) (DM/Chl ϭ 20 and 100 (w/w) in PSI-containing and PSI Ϫ strain, respectively), and obtained complexes were separated in the first dimension by either blue-native electrophoresis at 4°C in 5-14% polyacrylamide gel according to (41) or native Deriphat electrophoresis as described in (42). Subunit composition of the complexes was assessed by electrophoresis in a denaturing 12-20% linear gradient polyacrylamide gel containing 7 M urea (in the second dimension) (40). Proteins separated in the gel were transferred onto PVDF membrane. The membrane was incubated with specific primary antibodies followed by incubation with a secondary antibody conjugated with horseradish peroxidase (Sigma) and chemiluminescence Lumilight substrate (Roche Applied Science). The primary antibodies used in the study were raised in rabbits against: (i) residues 58 -86 of the spinach D1 polypeptide, (ii) residues 380 -394 of barley CP47 (40), and (iii) monoclonal antipolyhistidine (Sigma). For autoradiography, the gel or the membrane with labeled proteins was exposed to x-ray film at laboratory temperature for 2-3 days.
Isolation of PSII and Protein Analysis-The isolation of thylakoid membranes was performed as described in Ref. 42. The thylakoid membranes were washed until the supernatant became colorless with a buffer containing 25 mM MES, pH 6.5, 5 mM MgCl 2 and 5 mM CaCl 2 . The washed thylakoid membranes (0.25 mg Chl ml Ϫ1 ) were solubilized with DM (1% final concentration) for 10 min on ice. Nonsolubilized material was Determination of Chl-Chl was extracted from cell pellets (50 ml, OD 730 ϳ0.5) by 100% methanol. Chl content was measured spectrophotometrically on a Spectronic Unicam UV500 spectrophotometer (43).
Gold Labeling of the His-tagged PSII Particles-To visualize the position of the ScpDHis subunit, Ni-NTA Nanogold (Nanoprobes Inc.) was used. This label is a gold cluster with attached Ni 2ϩ enabling specific binding of the gold particle to the Histagged proteins. Well-washed thylakoid membranes were suspended in a buffer containing 20 mM Tris (pH 7.4), 100 mM NaCl, and 5 mM CaCl 2 at a Chl concentration of 100 g ml Ϫ1 and mixed with an equal volume of Ni-NTA Nanogold solution (30 nM). The solution was incubated for 30 min at 4°C and the free Ni 2ϩ -NTA groups on gold labels were saturated by 1 mM L-histidine for a further 10 min on ice. Thylakoid membranes were then solubilized with 1% DM for 15 min and the unsolubilized material was removed by centrifugation at 60,000 ϫ g for 30 min. The supernatant was loaded onto a freshly prepared 0.1-1.2 M continuous sucrose density gradient prepared by freezing and thawing the centrifuge tubes filled with a buffer containing 20 mM Tris (pH 7.4), 0.6 M sucrose, 10 mM NaCl, 5 mM CaCl 2 , and 0.05% DM (44), and the PSII complexes were separated by centrifugation at 150,000 ϫ g (Hitachi, P56SW rotor) for 14 h at 4°C. The lower green band containing the PSII dimers was harvested with a syringe and loaded onto a Sephadex G-25 (Amersham Biosciences) desalting column equilibrated with 20 mM Tris (pH 7.4) containing 0.05% DM. Nonlabeled PSII particles were prepared as described above, but the Ni-NTA Nanogold labeling step was omitted during the procedure.
Electron Microscopy and Image Analysis-Freshly prepared labeled PSII complexes eluted from the desalting column were placed on glow-discharged carbon-coated copper grids and negatively stained with 0.75% uranyl acetate. Electron micros-copy was performed with a Philips TEM 420 electron microscope using 80 kV at ϫ60,000 magnification. Micrographs free from astigmatism and drift were scanned with a pixel size corresponding to 4.5 Å at the specimen level. Image analyses were carried out using SPIDER software (45). From 63 micrographs of the PSII cores, about 2900 top-view projections of unlabeled particles and 472 side-view projections of labeled particles were selected for analysis. Both separate data sets were rotationally and translationally aligned, and subjected to multivariate statistical analysis in combination with classification (46,47). Classes from each of the subsets were used for refinement of alignments and subsequent classifications. For the final sum, the best of the class members were summed using a cross-correlation coefficient of the alignment procedure as a quality parameter. The resolution of the images was calculated using the Fourier ring correlation method (48). For molecular modeling, the coordinates were taken from the Protein Data Bank under the code 1S5L for PSII structure at 3.5-Å resolution (49). The overlay cartoon was generated by the freeware program Accelrys ViewerLite 4.2

ScpDHis Protein Is Localized in the Thylakoid Membrane-
To localize the ScpD protein in the cells of Synechocystis 6803, we constructed mutants in which the scpD gene had been tagged by the His 6 epitope (ScpDHis strains) as described under "Experimental Procedures." As higher accumulations of ScpC/D proteins have been detected in PSI Ϫ strains (12), the ScpDHis/⌬ScpD/PSI Ϫ strain was used to determine the localization of the ScpDHis protein. Using the anti-His monoclonal antibody, a single band was detected in the thylakoids of the ScpDHis/⌬ScpD/PSI Ϫ strain (Fig. 1). No such band was detected in the fraction of soluble proteins from this strain and in the thylakoids of PSI Ϫ (not shown) and ScpDHis/⌬ScpD strains grown at 30 mol photon m Ϫ2 s Ϫ1 (Fig. 1). However, the ScpDHis band was detectable in the ScpDHis/⌬ScpD strain grown under HL (Fig. 1).
The ScpDHis Protein Is Associated with Photosystem II-To further localize the ScpDHis protein in the thylakoid membranes, we performed two-dimensional analysis of thylakoid proteins. Mildly solubilized thylakoid protein complexes from the ScpDHis/⌬ScpD/PSI Ϫ strain were first separated by nondenaturing blue native (BN) or Deriphat PAGE and their subunit composition was determined by SDS-PAGE in combination with Western blotting in the second dimension. Both methods yielded similar results with the majority of the Scp-DHis being present in two complexes at about 300 and 220 kDa  1, 2, and 3, respectively). OCTOBER 27, 2006 • VOLUME 281 • NUMBER 43 ( Fig. 2A). These complexes had been previously identified as a monomeric reaction center core complex (RCC1) and a smaller reaction center core subcomplex (RC47), which contains CP47 but no CP43 (40) (Fig. 3). BN-PAGE provided a better separation of the complexes than Deriphat-PAGE (not shown) and was therefore used subsequently. In the ⌬CP47 background where no PSII core is assembled, the ScpDHis/⌬ScpD/PSI Ϫ / ⌬CP47 mutant accumulated a significant amount of ScpDHis; however this protein was localized in a region of the small protein complexes and free proteins (Fig. 2B). Thus, ScpDHis comigrated with PSII in two different buffer systems and ScpDHis association with large thylakoid complexes was dependent on PSII accumulation, indicating that ScpDHis is associated with PSII.

Localization of the CAB-like Protein
To confirm that the ScpDHis protein is indeed associated with PSII we used a combination of several independent methods for isolation of PSII. Monomeric and dimeric reaction center core complexes were isolated from the ScpDHis/⌬ScpD/ PSI Ϫ strain by sucrose density gradient centrifugation or by a combination of affinity and ion exchange chromatography. The complexes were further purified by BN-PAGE. Green PSII bands were cut from the native gel and subjected to SDS-PAGE and immunodetected with anti-His antibody. ScpDHis protein was present in all PSII preparations (Fig. 4).
The His tag is widely used to isolate tagged proteins under native conditions by nickel affinity chromatography allowing co-purification of proteins associated with the tagged protein.
However, native isolation of the ScpDHis protein was not successful and no ScpDHis protein was found in the fraction eluted from the nickel column (not shown).
ScpDHis Protein Expression-As the ScpDHis protein had been expressed under light-inducible psbA2 promoter, we tried to follow its expression after HL treatment. Protein expression in WT and ScpDHis/⌬ScpD cells was compared using autoradiography after two-dimensional PAGE (BN/SDS-PAGE). Cells were pulse-labeled with a mixture of [ 35 S]Met and Cys under HL conditions. Electrophoretic analysis of labeled proteins and autoradiography showed the expression of a new 7-kDa protein in the ScpDHis/⌬ScpD strain that was absent in WT. Again, this protein was associated with RCC1 and RC47 (Fig. 3). The position of this 7 kDa protein on the membrane was identical to that of the ScpDHis protein detected by immu-  Thylakoid membrane proteins from both strains were pulse radiolabeled for 30 min, separated in the first dimension by BN-PAGE and in the second dimension by denaturing SDS-PAGE, blotted onto PVDF membrane, and exposed to x-ray film. For the ScpDHis/⌬ScpD autoradiogram, same membrane was used as in Fig. 2A.  1 and 2) or sucrose density gradient centrifugation (dimeric and monomeric PSII; lanes 3 and 4) followed by separation by BN-PAGE. PSII from the ScpDHis/⌬ScpD/PSI Ϫ strain was isolated by ion exchange chromatography or sucrose density gradient centrifugation as described under "Experimental Procedures." The isolated PSII protein complexes were further purified by BN-PAGE. The separated proteins were stained with Coomassie Blue (lane 1) or blotted onto PVDF membrane and immunodetected using anti-His and anti-D1 antibodies (lanes 2, 3, and 4). noblotting ( Fig. 2A) demonstrating that the expressed protein is ScpDHis.
The ScpDHis Protein Is Associated with CP47 and PsbH-Previous results showed that the ScpDHis protein is associated with PSII. To localize this protein more precisely, we followed the association of ScpDHis with other PSII proteins in several mutants lacking individual PSII subunits. Interestingly, in the ⌬H background (strain ScpDHis/⌬ScpD/PSI Ϫ /⌬H containing ScpDHis but lacking PSI and the PsbH subunit of PSII), the ScpDHis protein was not associated with PSII but was present in a region of small protein complexes or free proteins (Fig. 5). Note that the ⌬H mutant grown in the presence of glucose exhibits activity in oxygen evolution similar to that of WT (50) and also contains a similar set of PSII complexes (51). This indicates that the PsbH protein is necessary for ScpDHis association with PSII.
Unlike the ⌬H strain, the mutant ⌬Cyt lacking the psbEFLJ operon does not accumulate any stable PSII subcomplexes, however it does contain a small amount of unassembled CP47 that is resolved in the BN gels as a double band (40) containing not only CP47 but also PsbH (51). In the ScpDHis/⌬ScpD/ PSI Ϫ /⌬Cyt strain, most of the ScpDHis protein was present in a region of small protein complexes or free proteins (Fig. 5). However, a significant amount of ScpDHis was associated with one of the two bands of unassembled CP47. The observation that ScpD is indeed part of this CP47⅐PsbH complex and that co-migration with CP47 on a two-dimensional gel (Fig. 5) is not just accidental, was confirmed by isolation of this complex from the ⌬Cyt strain expressing the His-tagged PsbH (not shown). This indicates that ScpDHis is associated with CP47 and PsbH under conditions when PSII assembly is blocked at the early step of formation of reaction center (RC) complexes because of the absence of cytochrome b 559 and D2 (40).
Localization of ScpDHis Using Electron Microscopy-To further confirm the location of the His-tagged ScpD subunit within PSII, the dimeric PSII complexes were labeled with Ni-NTA Nanogold and visualized in an electron transmission microscope. As the labeling procedure previously used for the localization of the PsbH protein (52) yielded no labeling, an alternative procedure had to be developed that included direct labeling of thylakoid membranes prior to their solubilization and isolation of PSII complexes.
The electron micrographs revealed that the preparation contained dispersed particles with uniform size and shape and almost free of contaminants. The images contained dimeric PSII particles, mostly in their top-view projections (i.e. perpendicular to the original membrane plane). Side-view projections were less common and usually occurred as an aggregation of two single PSII complexes attached by their stromal surfaces (Fig. 6). Fig. 6 also shows that not all PSII particles carried gold labels. To localize the gold label within PSII dimers, both labeled and unlabeled particles were extracted from the micrographs, aligned, subjected to multivariate statistical analysis and classified. The most representative class averages of both labeled and unlabeled particles are depicted in Fig. 7. The topview projections showed diamond-shaped particles with 2-fold rotational symmetry around the center of the complex. The class averages closely resembled the PSII dimers without the ScpDHis protein (44,53,54). All the projections had the same type of handedness and no mirror images were detected, thus indicating a preferred orientation of the PSII dimers with their stromal side to the carbon support film. One part of the labeled top views had a gold label only on one side of the complex, whereas another part exhibited gold labels on both sides of the complex (Fig. 7, D and E). This indicates the ScpDHis protein as being located on both sides of the complex, reflecting the 2-fold rotational symmetry of the dimeric PSII complex.
The side-view projections showed two single PSII dimers aggregated with their stromal surface (Fig. 7C). Protrusions on the luminal sides of the single PSII dimer corresponded to the proteins of the oxygen-evolving complex (44). In side views the labels were located between the two single PSII particles in the paired structures (Fig. 7F). This clearly identifies the His tag, and therefore, the N terminus of the ScpD protein on the stromal side of the PSII complex.

DISCUSSION
We have expressed ScpD as a His-tagged protein on its N terminus to enable its isolation and localization within the thylakoid membranes. The ScpDHis protein was expressed under psbA2 promoter. Unlike the native scpD promoter, the psbA2 promoter belongs to those promoters with a high expression level even in low light (55) that is further increased under HL conditions. However, no accumulation of ScpDHis protein was observed in the cells grown in normal light, similar to what was observed for the native ScpD protein (12). This shows that different levels of the scpDHis and the scpD transcript did not influence the accumulation of both proteins in the cell, indicating that accumulation of ScpD is controlled mostly by other factors. Accordingly, in the PSI Ϫ background, ScpDHis accumulated even under low-light conditions, similar to that observed for the native ScpD (12).
Using several independent methods, we have determined that after HL treatment most of the ScpDHis protein in a cell is associated with PSII. The ScpDHis protein was present in both monomeric and dimeric PSII reaction center core complexes and also in the RC47 subcomplex lacking CP43, indicating that the CP43 subunit is not necessary for the association of Scp-DHis with PSII. The specific interaction of ScpDHis with PSII was further confirmed by the observation that in the absence of PSII, the ScpDHis protein was present in the form of small protein complexes. This also indicated that ScpDHis does not interact with other large thylakoid membrane complexes, particularly PSI, as was shown recently for the plant SCP homologue OHP (36). We also succeeded in a more detailed localization of ScpDHis within PSII as we detected the protein in the complex of CP47 and PsbH that accumulates at low levels in the strain unable to assemble PSII complexes because of the missing cytochrome b 559 (Fig. 5). Thus, ScpDHis seems to bind to PSII in proximity to the CP47-PsbH proteins. This corresponds well with the observation that in the strain ⌬H lacking the small PsbH protein, ScpDHis is no longer detected within PSII, indicating that the PsbH protein is important for the proper association of ScpDHis with PSII.
The His-tagged ScpD protein, this time with the tag on the C terminus, has been used previously to follow ScpD accumulation and potential association with other proteins (15). Interestingly, in this study the ScpD protein was found in the ϳ100 kDa fraction after gel filtration, apparently not associated with PSII. This raises the question as to whether the observed association of ScpDHis with PSII is not an artifact caused by the His tag. This is probably not the case as the native forms of ScpD and ScpC seem to be also associated with PSII (see below). We also attempted to isolate complexes containing ScpDHis under native conditions by nickel affinity chromatography in a similar procedure that we have used recently for the His-tagged PsbH protein (52). However, no ScpDHis protein was found in the fraction eluted from the column. We assume that this may be caused by the low accessibility of the His tag on ScpD bound to the PSII complex.
Similar problems were encountered during labeling of Scp-DHis for single particle analysis by Ni-NTA Nanogold, when no label was bound to PSII. Therefore, a different labeling procedure than the one used recently for the His-tagged PsbH protein (52) had to be developed for a successful labeling. The combination of Ni-NTA Nanogold labeling of His-tagged protein and single particle analysis provides an excellent tool to localize protein subunits within a protein complex (52,56). The main advantage to conventional immunogold labeling procedures is the close proximity of the gold particle to the His tag, enabling a more accurate localization of the target protein.
Using this procedure we localized the ScpDHis protein on the periphery of PSII dimers. To identify PSII proteins in the vicinity of the gold label we have overlaid a model of the transmembrane helix organization of PSII into a top-view projection map of the labeled PSII particle. Fig. 8 shows that the Ni-NTA gold label is found to be close to the transmembrane helices of the CP47 and PsbH proteins. This is in full agreement with other data indicating the proximity of ScpDHis to the CP47 and PsbH proteins. Moreover, the labeling confirmed that the Ni-NTA Nanogold and therefore, also, the N terminus of the ScpD protein is localized on the stromal side of the thylakoid membrane. Electron microscopy and single particle analysis did not reveal any differences between PSII dimers isolated from normal or HL-treated cells, indicating that SCPs do not form large complexes at or around PSII comparable to IsiA rings formed around the PSI trimers in iron-depleted conditions (57). When compared with the location of the PSII subunit PsbH using a similar approach (52), localization of the gold label on the Scp-DHis protein is less distinct (Fig. 7, D and E). This may indicate that the position of ScpDHis in PSII is more ambiguous. Moreover, several gold labels were distinguishable on each side of the PSII dimer, indicating that more than one copy of ScpDHis may be bound per PSII monomer.
We believe that the native ScpD protein (without the His tag) is also associated with PSII. On two-dimensional autoradiograms, ScpDHis is identified as a strong 7-kDa band (Fig. 3), the size of the native ScpD being expected to be about 6 kDa. Interestingly, a prominent 6-kDa band is present in WT, with a similar pattern of association with both PSII core complexes as we have shown for ScpDHis (Fig. 3). Moreover, this 6-kDa protein did not associate with PSII in the ⌬H background similarly as in the case of Scp-DHis. 3 We think that this 6-kDa band in WT is formed at least in part by native ScpD. However, the 6-kDa band is still present in the ScpDHis/⌬ScpD mutant lacking the native ScpD protein (Fig. 3).
We speculate that this band belongs to the ScpC protein known to have similar electrophoretic mobility as ScpD (15). Indirect evidence indicates that ScpD may associate with the highly homologous ScpC in thylakoid membranes, based on their co-purification during gel filtration of solubilized thylakoids and on their accumulation kinetics following exposure to HL (15). ScpC and ScpD have been also proposed to functionally complement each other (33). SCPs are generally expected to form complexes in the thylakoid membranes based on the presumption that they are Chlbinding proteins and that in LHCII, interaction of two homologous helices is necessary for pigment binding (58). Indeed, SCPs have been detected in high molecular weight complexes/ aggregates by gel filtration (15) or electrophoresis (12).
ScpD is considered to be a general stress protein that accumulates under various stress conditions. Extensive DNA microarray data show that ScpD is induced by almost all stresses tested including HL, nutrient deprivation or low temperature (18,20,59). In our experiments, ScpD protein became one of the most expressed proteins in the thylakoid membranes after 30 min of HL treatment (Fig. 3). However, scp genes are also expressed under standard growth conditions (18 -22). In cells maintained in low light and under optimal growth conditions, low levels of SCPs have been detected after partial purification (15).
SCPs and their larger relatives, ELIPs containing three transmembrane helices, share more than just the Chl binding motif. Both groups exhibit a similar pattern of expression under various stress conditions and both represent relatively short-lived proteins (11,15). Similarly to the association of ScpDHis with PSII, ELIP in pea has been found in the margins of the grana where PSII is assembled, and cross-linking experiments indicated the proximity of the pea ELIP to PSII (34). Based on these similarities, analogous functions have been proposed for both SCPs and ELIPs, ranging from a transient pigment carrier function (33,60) to their role in the dissipation of excess light energy (30,31). Indirect evidence is growing that ELIPs may be involved in the binding of Chl released during turnover of PSII and in protein stabilization during PSII assembly (29). ELIPs are expressed not only under typical stress conditions represented by HL stress or in the process of thylakoid biogenesis during greening, but also during the reverse process of thylakoid disassembly represented by leaf senescence (61) or by chloroplast-3 J. Komenda, unpublished data. to-chromoplast transitions during fruit ripening (62). This emphasizes the ubiquity of ELIPs during normal developmental processes when there is an increased pigment/protein turnover. Also SCPs are implicated in several processes involving Chl biosynthesis and/or accumulation. Similar to ELIPs, SCPs are necessary to ensure a high rate of Chl accumulation in cells during greening (33,63). Also Chl stability in the dark was negatively influenced by the absence of SCPs (33). This may be explained by the destabilization of PSII pigment-protein complexes and/or by inefficient Chl recycling (33). This is in line with our observation that ScpD significantly accumulates after several days of incubation in the dark, 4 suggesting that SCPs are physically present there to fulfill the above functions and that accumulation of SCPs can be also triggered by factors other than external stress. Interestingly, CP47, the binding partner of ScpDHis, is the most stable Chl protein of PSII and during PSII assembly and repair it is the most frequently reused large PSII subunit, as demonstrated by significantly lower levels of its radioactive labeling in comparison with D1, D2, and CP43 (40). Thus, it is tempting to speculate that SCPs contribute to this stability of CP47.
In conclusion, we have shown that the His-tagged ScpD protein, an abundant member of the SCP family in Synechocystis 6803 is located on the periphery of PSII in the proximity of the PSII subunits CP47 and PsbH. This localization is in accordance with most of the functions suggested for SCPs. However, because of the observed stoichiometry of only one to several copies of ScpD per PSII and because of the ScpD presence in PSII subcomplexes containing CP47 and PsbH we favor the idea of ScpD being involved in the process of stabilization of pigment-binding proteins, particularly CP47, during PSII assembly and repair.