Analysis of the psbU Gene Encoding the 12-kDa Extrinsic Protein of Photosystem II and Studies on Its Role by Deletion Mutagenesis in Synechocystis sp. PCC 6803*

The gene encoding the 12-kDa extrinsic protein of photosystem II from Synechocystis sp. PCC 6803 was cloned based on N-terminal sequence of the mature protein. This gene, namedpsbU, encodes a polypeptide of 131 residues, the first 36 residues of which were absent in the mature protein and thus served as a transit peptide required for its transport into the thylakoid lumen. A psbU gene deletion mutant grew photoautotrophically in normal BG11 medium at almost the same rate as that of the wild type strain. This mutant, however, grew apparently slower than the wild type did upon depletion of Ca2+ or Cl− from the growth medium. Photosystem II oxygen evolution decreased to 81% in the mutant as compared with that in the wild type, and the thermoluminescence B- and Q-bands shifted to higher temperatures accompanied by an increase in the Q-band intensity. These results indicate that the 12-kDa protein is not essential for oxygen evolution but may play a role in optimizing the ion (Ca2+ and Cl−) environment and maintaining a functional structure of the cyanobacterial oxygen-evolving complex. In addition, a double deletion mutant lacking cytochrome c-550 and the 12-kDa protein grew photoautotrophically with a phenotype identical to that of the single deletion mutant of cytochrome c-550. This supports our previous biochemical results that the 12-kDa protein cannot bind to photosystem II in the absence of cytochromec-550 (Shen, J.-R., and Inoue, Y. (1993) Biochemistry32, 1825–1832).

The oxygen-evolving system of cyanobacteria contains three extrinsic proteins, namely, a 33-kDa protein, cytochrome c-550, and a 12-kDa protein. The genes coding for the 33-kDa protein and cytochrome c-550 are psbO and psbV genes, respectively (for reviews see Refs. 1 and 2), whereas the gene for the 12-kDa protein has been tentatively named psbU (3,4). The 33-kDa protein is commonly found in higher plant and cyanobacterial PSII, 1 and its function has been studied extensively by both in vitro biochemical approaches and in vivo mutagenesis studies. Results from these studies suggested that the 33-kDa protein plays an important role in stabilizing the tetramanganese cluster, which directly catalyzes the water-splitting reaction. Loss of this protein by biochemical removal from isolated PSII (5)(6)(7) or genetic deletion (8,9) from algal cells leads to a significant loss of the oxygen-evolving activity and in some conditions loss of manganese atoms from the tetramanganese cluster. The other two proteins, cytochrome c-550 and the 12-kDa protein, however, are present only in algal-type PSII but absent in higher plant PSII. Our previous in vitro biochemical (10) and in vivo genetic studies (11) have indicated that cytochrome c-550 is required for maintaining both the oxygen evolution and PSII stability in vivo. This cytochrome can bind to PSII essentially independent of the other extrinsic proteins (10). In accordance with this, a double deletion mutant of Synechocystis sp. PCC 6803 lacking both the 33-kDa protein and cytochrome c-550 showed a complete loss of photoautotrophic growth, which is caused by deactivation of oxygen evolution and destabilization of PSII in vivo (12). In contrast, both the single deletion mutant of the 33-kDa protein (9) or cytochrome c-550 (11) were able to grow autotrophically, albeit with reduced rates. These studies suggested that cytochrome c-550 binds to and functions in cyanobacterial PSII independent of the 33-kDa protein in maintaining the oxygen-evolving activity and PSII stability (12).
The 12-kDa protein was first found as a 9-kDa protein in PSII purified from a thermophilic Phormidium laminosum (13,14). Dissociation of this protein from the isolated PSII caused a decrease in oxygen evolution that was partially restored upon rebinding of the protein (14,15). The P. laminosum gene encoding the protein has been cloned and includes a leader sequence required for its transport into the thylakoid lumen (16). We have confirmed the presence of a homologous 12-kDa protein in PSII purified from another thermophilic cyanobacterium, Synechococcus vulcanus (17). Lack of the 12-kDa protein also led to a decrease in oxygen evolution of the Synechococcus PSII that was restored by rebinding of the protein (10). We further showed that binding of the 12-kDa protein in the isolated Synechococcus PSII requires presence of both the 33-kDa protein and cytochrome c-550; in the absence of either of the two proteins, the 12-kDa protein cannot bind to PSII efficiently (10). These results suggested that the 12-kDa protein is in close contact with both the 33-kDa protein and cytochrome c-550 at the lumenal side of cyanobacterial PSII; this has been confirmed by in vitro cross-linking studies showing that all of the three proteins were cross-linked together by 1-ethyl-3-(3-dimethylaminopropyl) carbodimide hydrochloride (18). These studies were carried out in vitro with PSII purified from thermophilic cyanobacteria. In the present work, we demonstrated the presence of the 12-kDa protein in thylakoids of the mesophilic cyanobacterium, Synechocystis sp. PCC 6803 by N-terminal amino acid sequencing and cloned the gene coding for the 12-kDa protein based on its N-terminal sequence. The role of the 12-kDa protein was then studied by construction and characterization of the psbU gene deletion mutants.

MATERIALS AND METHODS
The mesophilic, glucose-tolerant, transformable cyanobacterium, Synechocystis sp. PCC 6803, was grown in BG11 medium at 30°C at a light intensity of 25-30 einsteins m Ϫ2 s Ϫ1 . Thylakoid membranes were isolated from the Synechocystis cells by glass beads according to Ref. 19, and the isolated thylakoids were treated by sonication to release extrinsic proteins as described in Ref. 20. The obtained sonication extract was separated by a Mono Q column using a fast protein liquid chromatography system (Pharmacia Biotech Inc.) (11). The fractions from column chromatography were collected and electrophoresed in a gel containing 16% polyacrylamide and 6.0 M urea, and protein bands around the 12-kDa molecular mass region were transferred to a polyvinylidene difluoride membrane followed by N-terminal amino acid sequencing. The 12-kDa protein was identified by comparing the obtained sequences with the 9-kDa protein sequence from P. laminosum (16) and the N-terminal sequence of the 12-kDa protein from S. vulcanus (17).
Based on the N-terminal sequence, the gene coding for the 12-kDa protein was cloned from genomic DNA of Synechocystis PCC 6803 by a two-step, reversed PCR method (11). This gene was named psbU. DNA sequencing was performed with an ALFexpress DNA sequencer (Pharmacia) using Cy5 fluorescent primers, based on the Sanger dideoxysequencing method. PCR reaction was carried out using an Expand High Fidelity PCR kit (Boehringer Mannheim), and each PCR product was sequenced from three independent clones to eliminate any possible inconsistencies. To obtain the clone containing the complete psbU gene, another PCR reaction was carried out with genomic DNA as template and primers flanking the whole gene. The resulted PCR product was cloned into pUC119 and used for subsequent mutagenesis studies.
To delete the 12-kDa protein from Synechocystis PCC 6803, a plasmid was constructed in which a 0.45-kb EcoT14 I-XbaI fragment containing the entire coding region of the gene was replaced by a kanamycin-resistant cassette. This plasmid was used to transform the glucosetolerant Synechocystis PCC 6803 strain. Cyanobacterial transformation and selection of the kanamycin-resistant transformant were carried out according to published procedures (21). As a control, another plasmid was constructed in which the kanamycin-resistant cassette was inserted into an XbaI site downstream of the coding region of the 12-kDa protein gene, and this plasmid was also transformed to Synechocystis PCC 6803.
A double deletion mutant lacking cytochrome c-550 and the 12-kDa protein was constructed by introducing the psbU gene deletion plasmid into a host cell in which the psbV gene coding for cytochrome c-550 has been replaced by an erythromycin-resistant cassette as described previously (11). The transformants were selected and propagated in BG11 plates containing 25 M atrazine, 5 mM glucose, 10 g kanamycin/ml, and 20 g/ml erythromycin. Growth curves of the wild type and mutant strains in liquid BG11 were recorded by measuring the light scattering of cells at 730 nm. For depletion of Ca 2ϩ or Cl Ϫ from the BG11 medium, 0.24 mM CaCl 2 in the original medium was replaced by either 0.48 mM NaCl or 0.24 mM Ca(NO 3 ) 2 . Synechocystis cells grown in their mid-logarithmic phase were harvested, washed twice with distilled water, and then transferred to the growth medium lacking Ca 2ϩ or Cl Ϫ . For electrophoresis, harvested cells were broken by glass beads (100 -150 m), solubilized, and then applied to a 16 -22% polyacrylamide gradient gel containing 6.0 M urea. Cytochromes of c-type were detected by monitoring peroxidase activity of the hemes on the gel with 3,3Ј,5,5Ј-tetramethylbenzidine and H 2 O 2 as described in Refs. 20 and 22. Oxygen evolution of the Synechocystis cells was measured with a Clark-type oxygen electrode under continuous, saturating yellow light at 30°C in BG11 with 0.6 mM 2,6-dichlorobenzoquinone and 1 mM potassium ferricyanide as acceptors. For thermoluminescence measurement, harvested Synechocystis cells were suspended in BG11 in the absence or the presence of 20 M DCMU at 100 g chlorophyll/ml, adapted in the dark for 5 min at room temperature, and then illuminated with single turn-over flashes at 0°C for samples without DCMU or with continuous light for 30 s at Ϫ5°C for samples supplemented with DCMU. Thermoluminescence glow curves were recorded as described in Ref. 23, with a heating rate of 1.0°C/s.

RESULTS
Cloning and Sequence Analysis of the psbU Gene Coding for the 12-kDa Protein-The 12-kDa protein was identified by Nterminal sequencing of several protein bands around 12-kDa molecular mass region obtained by column chromatography of sonication extracts from thylakoid membranes of Synechocystis PCC 6803. The sequence obtained is ELNAVDAKLTTDF-GQKIDLNNSDIXDFXGLRGFYPXLAXXIIKN; this sequence is homologous to the 9-kDa protein of P. laminosum and the 12-kDa protein of S. vulcanus.
Based on the N-terminal sequence, the psbU gene encoding the 12-kDa protein was amplified and cloned by a two-step PCR method (11). The obtained DNA sequence, together with its 5Јand 3Ј-flanking regions, is shown in Fig. 1. The start codon for this gene was assigned at the ATG codon of nucleotide number 253 in the DNA fragment shown in Fig. 1, because this codon is preceded by a typical ribosome-binding sequence (Shine-Dalgarno sequence, AGGAG, underlined in Fig. 1) 7 base pairs upstream, as well as a stop codon 39 base pairs upstream in the same reading frame (indicated by an asterisk in Fig. 1). According to this assignment, the gene encodes a polypeptide of 131 amino acid residues with a total molecular mass of 14231 Da. Hydropathy analysis indicated that the derived amino acid sequence is mostly hydrophilic, except the first 36 residues, which form a major hydrophobic loop (not shown). This is in agreement with the previous results that the 12-kDa protein is a hydrophilic one that is associated with lumenal surface of the thylakoid membranes (10,(13)(14)(15)(16)(17). As upon PCR amplification of the psbU gene, we detected only one band of the gene, we consider that there is only one copy of the gene in Synechocystis PCC 6803. This is consistent with the results from computer analysis of the complete genome sequence of Synechocystis PCC 6803 that was recently determined (24).
The N-terminal sequence determined for the mature 12-kDa protein corresponds to the sequence starting from residue number 37 of the gene-derived amino acid sequence. This indicates the cleavage of the first 36 residues after synthesis of the protein, which gave rise to a mature 12-kDa protein of 95 residues with a calculated molecular mass of 10490 Da. Thus, the first 36 residues served as a transit peptide for the protein to be transferred across membranes. This is consistent with the previous results that the 12-kDa protein is associated with the lumenal side of thylakoid membranes (10). In fact, a similar transit peptide has been found for the homologous 9-kDa protein from P. laminosum (16), and recently it has been reported that the same gene from Synechococcus sp. PCC 7002 also carried a transit peptide (3). Although the three transit peptides for the 12-kDa protein have virtually no similarity in their primary sequences, they share some common features of bacterial-type transit peptides (25), e.g. there is a positive charge in the N terminus and a hydrophobic cluster in the middle, and the residue immediately before the cleavage site is Ala in all the three sequences (Fig. 2). Fig. 2 compared the gene-derived 12-kDa protein sequence from Synechocystis PCC 6803 with those from Synechococcus sp. PCC 7002 (3) and P. laminosum (16) and also with the N-terminal partial sequence from S. vulcanus (17) and a red alga, Cyanidium caldarium (26). Among the three cyanobacteria for which the psbU gene has been cloned, an overall identity of 45.5-56.8% can be seen for the mature part of the sequence. When this comparison was made for the first 65 residues for which the sequence from S. vulcanus is known, the total ho-mology lies between 49.2-57.1% among the four species of cyanobacteria compared. The sequence homology in the same N-terminal region, however, are only 30.3-36.5% between cyanobacteria and the red alga C. caldarium; this is significantly lower than those observed among different species of cyanobacteria. This indicates a remarkably higher divergence in the 12-kDa protein sequence between cyanobacteria and red algae than those found among cyanobacteria.
Mutant Constructions-A deletion mutant ⌬psbU, was constructed by replacing a 0.45-kb EcoT14I-XbaI fragment containing the whole psbU gene with a 1.3-kb kanamycin-resistant cassette. This replacement was confirmed by PCR amplification of the DNA region containing the psbU gene. As shown in Fig. 3, although PCR amplification with two primers whose sequences correspond to the 5Ј-and 3Ј-sequences of the DNA fragment shown in Fig. 1 yielded a fragment of 0.92 kb from the wild type cells, the same amplification yielded a fragment of 1.8 kb from the targeted 12-kDa protein deletion mutant, which is exactly the same as would be expected if the 0.45-kb psbU gene-containing fragment was replaced by the 1.3-kb kanamycin cassette. The original 0.92-kb fragment disappeared completely, indicating a successful deletion of the only psbU gene in the mutant. Fig. 3 also shows the successful construction of a double deletion mutant, ⌬psbU/⌬psbV, which is depleted of both the 12-kDa protein and cytochrome c-550. In this mutant, the psbU gene was replaced by the kanamycin cassette (Fig. 3,  lane 4 in panel A) and a large part of the psbV gene was replaced by an erythromycin-resistant cassette (11) (Fig. 3,  lane 4 in panel B). As a control, a Synechocystis strain designated psbU/Km was constructed by inserting the 1.3-kb kanamycin cassette into the XbaI site downstream of the psbU gene. This insertion does not inactivate the psbU gene and gives rise to a PCR fragment of 2.2 kb as shown in lane 2 of panel A in Fig.  3, which thus indicates a successful construction of the psbU/Km strain.
Characterization of the Mutants- Fig. 4 shows the photoautotrophic growth curves and oxygen evolution of wild type and mutant strains. Both the deletion mutant ⌬psbU and the control strain psbU/Km showed a growth rate close to that of the wild type, indicating that deletion of the 12-kDa protein had very little, if any, effect on growth of the Synechocystis cells. The PSII oxygen evolution decreased to 81% in the ⌬psbU mutant as compared with that of the wild type strain. In contrast, the psbU/Km strain had the same oxygen-evolving activity as that of the wild type. These results indicate a slight decrease in the oxygen evolution upon deletion of the psbU FIG. 2. Comparison of the 12-kDa protein sequence among four species of cyanobacteria and one species of red alga currently known. The complete sequence of the 12-kDa protein from cyanobacteria Synechocystis sp. PCC 6803 (this study), Synechococcus sp. PCC 7002 (3), P. laminosum (16), the N-terminal partial sequence of the mature 12-kDa protein from S. vulcanus (17), and a red alga C. caldarium (26) were compared. A downward arrowhead indicates the cleavage site; an asterisk indicates identical residues among all the sequences compared; and a slash (/) indicates an incomplete sequence. The table below the sequence shows the ratio of identical residues out of the total residues compared, and the numbers in parentheses are the same ratio when the N-terminal 65 residues for which the sequence from S. vulcanus was known were compared.
gene. Apparently, this decrease in the PSII activity was not large enough to give rise to a remarkable change in the growth rate of the mutant strain, presumably because that the growth rate of the Synechocystis cells is determined primarily by some photosynthetic steps other than the PSII oxygen-evolving reaction. In fact, most of the mutants leading to an impairment in PSII activity showed a smaller decrease in their growth rates or overall photosynthetic electron transport activities than the decrease observed in the PSII oxygen-evolving activity (see for example Ref. 9).
The growth rate of the double deletion mutant ⌬psbU/⌬psbV decreased to half of the wild type strain (Fig. 4); this is comparable with that of the single deletion mutant ⌬psbV reported previously (11). The activity of the double deletion mutant, ⌬psbU/⌬psbV, is 41% of the wild type strain (Fig. 4), which is also similar to the activity observed for the single deletion mutant, ⌬psbV (11). These results are in contrast to the results observed with the single deletion mutant ⌬psbU and indicates that further deletion of the psbU gene from the ⌬psbV mutant strain had no effect on its growth and oxygen evolution. This in turn suggests that the 12-kDa protein did not function in oxygen evolution of the ⌬psbV mutant, in agreement with the previous in vitro results that the 12-kDa protein cannot bind to and function in cyanobacterial PSII in the absence of both cytochrome c-550 and the 33-kDa protein (10).
The effect of deletion of the 12-kDa protein on growth was also examined in the absence of Ca 2ϩ or Cl Ϫ in the growth medium. As shown in Fig. 5, the wild type strain was able to grow in the medium depleted of either Ca 2ϩ or Cl Ϫ with a slightly reduced rate in agreement with the results reported previously (27,28). A similar growth was observed for the strain psbU/Km. The deletion mutant ⌬psbU, however, showed an apparent decrease in the growth rate in the absence of either Ca 2ϩ or Cl Ϫ . This is different from the growth in the normal BG11 medium, where no effect was seen upon deletion of the gene (Fig. 4). This suggests a role of the 12-kDa protein in maintaining the optimum ion (Ca 2ϩ and Cl Ϫ ) environment required for cyanobacterial oxygen evolution. Presumably, deletion of the 12-kDa protein decreased the affinity of PSII for Ca 2ϩ and Cl Ϫ , which then leads to a reduction of the growth in the absence of either one of these two ions. In contrast, the double deletion mutant ⌬psbU/⌬psbV cannot grow at all in the absence of either Ca 2ϩ or Cl Ϫ . This is caused primarily by deletion of the psbV gene, because the single deletion mutant ⌬psbV is already unable to grow in the absence of Ca 2ϩ or Cl Ϫ . 2 The oxygen-evolving system in the mutant strains was further investigated by thermoluminescence from cells excited with single turn-over flashes. The thermoluminescence B-band, which is obtained in the absence of DCMU and arises principally from charge recombination between S 2 and Q B Ϫ (29), had a peak temperature of 24°C for the wild type strain (Fig. 6). Its intensity becomes larger following two flashes of illumination than after one flash of illumination, a typical feature observed for thermoluminescence from whole cyanobacterial cells (30). The peak temperature and intensity of the B-band from the control strain psbU/Km is very similar to those of the wild type strain. The ⌬psbU strain also showed a similar intensity of the B-band as that of the wild type or psbU/Km strains following either one, two, or three flashes. Its peak temperature, however, shifted to 28°C; this is apparently higher than the peak temperature of the B-band from wild type or psbU/Km strains. This suggests a possible modification of the S 2 -state upon deletion of the psbU gene. The peak temperature of B-band from the double deletion mutant ⌬psbU/⌬psbV is around 30°C, which is close to the single deletion mutant ⌬psbU. The intensity of the B-band from the double deletion mutant ⌬psbU/⌬psbV, however, decreased significantly than that of the wild type or psbU/Km strains. This implies that the shift of the B-band peak temperature is caused by deletion of the 12-kDa protein, whereas deletion of cytochrome c-550 mainly affected the overall activity of PSII oxygen evolution but had essentially no effects on properties of any single S-states.
The thermoluminescence Q-band, which is obtained in the presence of DCMU and arises from charge recombination of S 2 Q A Ϫ , showed a peak temperature of 18°C for the wild type and psbU/Km strains. This temperature is slightly higher than that reported previously for cyanobacterial cells (30), presumably because of the difference in the growth conditions between the present and previous studies. The intensity of the Q-band is similar to that of the B-band from the same strains. The Qband from the ⌬psbU mutant strain showed a peak temperature of 22°C, which is higher than the Q-band from the wild type or psbU/Km strains. This resembles the situation observed for the B-band, thus confirming the modification of the S 2 -state upon removal of the 12-kDa protein. The intensity of the Q-band in the ⌬psbU mutant strain, however, was significantly higher than the intensity of the B-band from the same strain. The cause for this increase is not clear at present but may be due to a further modification on Q A Ϫ upon removal of the 12-kDa protein. In the double deletion mutant ⌬psbU/ ⌬psbV, peak temperature of the Q-band is around 24°C, which is close to that of the Q-band observed for the ⌬psbU mutant. Its intensity also increased to be slightly higher than that of the B-band from the same strain, although this increase is not so large as that observed in the ⌬psbU mutant. These features are also observed in the single deletion mutant ⌬psbV, 2 suggesting that the 12-kDa protein already lost its function in the ⌬psbV mutant. The smaller increase in the intensity of the Q-band in the ⌬psbU/⌬psbV mutant is apparently due to the fact that the number of functional PSII reaction centers is decreased in the double deletion mutant, because deletion of psbV has been shown to destabilize the PSII complex (11). DISCUSSION Features of the psbU Gene-The psbU gene cloned in the present study represents the first gene from mesophilic cyanobacteria coding for the 12-kDa extrinsic protein of algal PSII. As in the thermophilic cyanobacteria (16), the gene obtained here had a leader sequence typical of a prokaryotic type. This suggests the transport of the protein after its synthesis, in support of the previous biochemical evidence that this protein is located in the lumenal side of the thylakoid membrane (10).
The 12-kDa protein, together with cytochrome c-550, are two PSII extrinsic proteins first found in prokaryotic algae cyanobacteria (17), and they recently have been confirmed in a eukaryotic red alga C. caldarium (26). The psbV gene for cytochrome c-550 has been found in the plastid genomes of eukaryotic algae Cyanophora paradoxa (31), Porphyra purpurea (32) and a diatom Odontella sinensis (33) for which the complete plastid genome sequences have been determined. The psbV genes from all the three types of algae had a leader sequence in their N terminus, consistent with a thylakoid lumenal location of this cytochrome. In the plastid genomes from these three species of eukaryotic algae, however, the psbU gene for the 12-kDa protein was not found. Although the presence of the 12-kDa protein in Cyanophora and diatom has not been confirmed, this protein has been found in the red alga C. caldarium (26), and therefore, the psbU gene must be present in nuclear genome of the red algae. In fact, we have recently cloned the psbU gene from the red alga C. caldarium, which shows features of a typical nuclear gene coding for thylakoid lumenal proteins, e.g. the presence of a bipartite transit peptide in its N terminus that is required for transport of the protein across both the envelope and thylakoid membranes of chloroplasts. 3 The overall homology of the gene-derived amino acid sequence was 45.5-56.8% among three species of cyanobacteria P. laminosum, Synechocystis 7002, and Synechocystis 6803. The homology in the N-terminal region of the protein was 49.2-57.1% among four cyanobacteria including S. vulcanus. On the other hand, the homology in the same N-terminal region was 30.3-36.5% between the four species of cyanobacteria and the red alga C. caldarium (26). This is significantly lower than those observed among different species of cyanobacteria and therefore suggests an evolutionary divergence in the oxygen-evolving complex between prokaryotic cyanobacteria and eukaryotic red algae.
Function of the 12-kDa Protein-Deletion of the psbU gene slightly decreased the PSII oxygen evolution. This effect was apparently caused by loss of the 12-kDa protein, because insertion of the kanamycin cassette downstream of the psbU gene had no detectable effects. This agrees with the previous in vitro biochemical results that reconstitution with the 12-kDa protein to purified S. vulcanus PSII resulted in a small increase in the oxygen evolution (10). Similar results have been observed with PSII purified from P. laminosum, although in this case, the decrease observed after release of the 12-kDa protein was rather large presumably partly due to a concomitant loss of cytochrome c-550 from the PSII particles (14,15). The decrease in oxygen evolution caused by loss of the 12-kDa protein, however, is rather small compared with that caused by loss of the other two extrinsic proteins, the 33-kDa protein and cytochrome c-550; loss of either of these two components resulted in a significant loss of oxygen evolution accompanied by an apparent reduction in the growth rate (9,11,27,34,35). In contrast, the 12-kDa protein deletion mutant grew photoautotrophically at a rate very similar to that of the wild type strain.
The growth rate of the ⌬psbU mutant, however, was apparently lower than that of the wild type strain in the absence of Ca 2ϩ or Cl Ϫ . This suggests a role of the 12-kDa protein in maintaining the optimum ion (Ca 2ϩ and Cl Ϫ ) environment required for cyanobacterial oxygen evolution. In addition, a shift to higher temperature in the thermoluminescence B-and Q-bands, as well as an increase in the intensity of the Q-band, were observed upon deletion of the psbU gene. These results may suggest a modification of the S-state, in particularly the S 2 -state, upon depletion of the 12-kDa protein, which thus suggests a role of the 12-kDa protein in maintaining a functional structure of the algal PSII complex.
The double deletion mutant, ⌬psbU/⌬psbV, showed a similar phenotype to that of the single deletion mutant ⌬psbV in terms of its oxygen evolution, growth in the presence or the absence of Ca 2ϩ or Cl Ϫ , and thermoluminescence properties. 2 This indicates that deletion of the psbU gene from the ⌬psbV mutant had essentially no further effect on the PSII oxygen-evolving complex, thus suggesting that the 12-kDa protein did not function in the single psbV deletion mutant lacking cytochrome c-550. This agrees with the previous results that the 12-kDa protein cannot bind to and function in isolated PSII in the absence of either cytochrome c-550 or the 33-kDa protein. This double deletion mutant ⌬psbU/⌬psbV is thus very much different from another double deletion mutant, ⌬psbO/⌬psbV, which cannot grow photoautotrophically in spite of the fact that either of the single deletion mutants ⌬psbO or ⌬psbV can grow autotrophically (12).