Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

A Domain of the Manganese-stabilizing Protein fromSynechococcus elongatus Involved in Functional Binding to Photosystem II

A Domain of the Manganese-stabilizing Protein fromSynechococcus elongatus Involved in Functional... THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 277, No. 17, Issue of April 26, pp. 14747–14756, 2002 © 2002 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. A Domain of the Manganese-stabilizing Protein from Synechococcus elongatus Involved in Functional Binding to Photosystem II* Received for publication, January 26, 2001, and in revised form, January 22, 2002 Published, JBC Papers in Press, January 23, 2002, DOI 10.1074/jbc.M100766200 Akihiro Motoki‡§, Mina Usui‡, Tsuneo Shimazu‡, Masahiko Hirano‡, and Sakae Katoh¶ From the ‡Biological Sciences Department, Toray Research Center Inc., Kamakura 248-8555, Japan and the ¶Department of Biology, Faculty of Sciences, Toho University, Funabashi 274-8510, Japan Site-directed mutagenesis was performed to investi- cytochrome c , the 12-kDa protein, and a 20-kDa protein in red algae (5). MSP stabilizes the manganese cluster, which gate whether the two protease-sensitive sequences 156 163 184 191 -Gly and Arg -Ser , of the manganese-stabi- Phe catalyzes oxidation of water to molecular oxygen. Dissociation lizing protein (MSP) from a thermophilic cyanobacte- of the protein by washing with 1 M CaCl or MgCl (6) or 2.6 M 2 2 rium, Synechococcus elongatus (Motoki, A., Shimazu, T., urea plus 0.2 M NaCl (7) leads to a gradual release of two of four Hirano, M., and Katoh, S. (1998) Biochim. Biophys. Acta Mn ions present in the cluster. Oxygen evolution is strongly 1365, 492–502), are involved in functional interaction suppressed by the release of MSP, but the lost activity can be with photosystem II (PSII). The ability of MSP to bind to restored by rebinding of the protein (8, 9). Because no cofactor its functional site on the PSII complex and to reactivate is associated with MSP, it is considered that binding of the oxygen evolution was dramatically reduced by the sub- protein itself is responsible for optimal activity of oxygen 152 158 160 162 , Asp , Lys ,orArg with un- stitution of Arg evolution. charged residues, by insertion of a single residue be- Cross-linking experiments with various bifunctional re- 156 157 157 and Leu , or by deletion of Leu . tween Phe agents have shown that MSP is associated with or in close Substitution of each of the four charged residues with proximity to essentially all of the major intrinsic proteins of the an identically charged residue showed that the charges 158 160 PSII complex (10 –15). Various attempts have been made to , and possibly Lys , are important for the elec- at Asp locate binding sites for PSII or identify amino acid residues on trostatic interaction with PSII. The reactivating ability MSP that are involved in binding to its functional site on the to was also strongly affected by the alteration of Phe PSII complex. The N-terminal sequence of spinach MSP was , the only strictly conserved Leu. Replacement of Lys 184 191 suggested to have a binding site to PSII because removal of 16 –Ser sequence, by Gln charged residue in the Arg or 18 amino acid residues from its N terminus by protease had only a marginal effect on the function of MSP. High digestion resulted in total loss of the protein binding (16). affinity binding of MSP to PSII was also affected signif- , which is located in a re- Recently, evidence was presented indicating that the N-termi- icantly by mutation at Arg 148 152 –Arg ) strictly conserved among the 14 se- gion (Val nal sequence is necessary for maintaining the binding ability of quences so far reported. These results imply that the the protein to PSII but might not be involved in the intermo- 148 163 –Gly sequence, which is well conserved among Val lecular binding itself (17). It was also suggested that Asp , the MSPs from cyanobacteria to higher plants, is a domain only conserved, charged residue in the N-terminal 18-amino of MSP for functional interaction with PSII. acid sequence, might engage in both intra- and intermolecular interactions (18). Cross-linking with EDC, which covalently links amino and carboxyl groups in van der Waals contact, Photosynthetic oxygen evolution is mediated by PSII, a indicated that MSP is bound directly to CP47, an intrinsic multiprotein complex carrying chlorophylls, carotenoids, and a chlorophyll-carrying protein of the PSII complex through elec- variety of redox cofactors (for reviews see Refs. 1 and 2). Among trostatic interaction (10 –12). Analysis of EDC-cross-linked the protein components of PSII, the extrinsic 33-kDa protein is products showed that a charge-pair interaction exists between 1 76 also called the manganese-stabilizing protein (MSP) and is a charged residue in the Asp –Lys sequence of the spinach 364 440 associated with the lumenal surface of the PSII complex to- MSP and an oppositely charged residue in the Phe –Asp gether with two or three other extrinsic proteins, i.e. the 23- sequence of CP47 (13). kDa and 17-kDa proteins in higher plants and green algae (3), Chemical modification with specific reagents showed that six cytochrome c and a 12-kDa protein in cyanobacteria (4), and conserved lysyl residues distributed over the entire sequence of spinach MSP are accessible to at least one of the reagents when the protein is free in solution but not when the protein is * This work was performed under the management of the Research associated with PSII membranes (19, 20). This suggests that Association for Biotechnology as a part of the R&D Project of Basic these lysyl residues are located in domains of the protein that Technology for Future Industries, supported by the New Energy and are in contact with the PSII complex and might participate in Industrial Technology Development Organization. The costs of publica- 159 236 tion of this article were defrayed in part by the payment of page intermolecular interaction. Cross-linking of the Lys –Lys charges. This article must therefore be hereby marked “advertisement” domain of spinach MSP to CP47 with a reagent that cross-links in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. lysyl residues within a 1.2-nm radius has been reported (21). § To whom correspondence should be addressed: Biological Sciences Four arginyl residues of the spinach protein may also be in- Dept., Toray Research Center Inc., Kamakura 248-8555, Japan. Tel.: 81-467-32-9963; Fax: 81-467-32-0414; E-mail: akihiro_motoki@trc. volved in the binding to PSII because they were modified with toray.co.jp. a specific reagent only when the protein was free in solution The abbreviations used are: PSII, photosystem II; MSP, extrinsic (20). 33-kDa manganese-stabilizing protein; EDC, 1-ethyl-3-(3-dimeth- Involvement of negatively charged amino acid residues on ylaminopropyl)-carbodiimide; MES, 2-(N-morpholino)ethanesulfonic acid. MSP in binding to PSII was also suggested by EPR studies with This paper is available on line at http://www.jbc.org 14747 This is an Open Access article under the CC BY license. 14748 Binding of MSP to Photosystem II a free radical-relaxing agent dysprosium, which showed that inserting a residue with a side group smaller than that of Met 156 157 157 between Phe the binding site for MSP on the PSII complex is positively and Leu or by deleting Leu , in order to charged (22). Cross-linking experiments in which carboxyl elucidate the structural configuration of MSP around Phe and Leu groups of spinach MSP were activated for cross-linking prior to , which are required for the functional binding of MSP. reconstitution with PSII membranes or vice versa, showed that all carboxyl groups involved in intermolecular cross-linking EXPERIMENTAL PROCEDURES with EDC are located on MSP (23). The binding affinity of Site-directed Mutagenesis of the psbO Gene—Site-directed mutagen- spinach MSP was strongly reduced by chemical modification of esis was performed by the method described previously (28). The pET8c carboxyl groups of aspartyl and glutamyl residues with glycine expression vector, which contains the wild-type psbO gene of S. elon- gatus, was constructed as described (29). The XbaI/BamHI fragment of methyl ester when it was free in solution but not when it was the vector containing the whole psbO gene was cloned into the multiple bound to PSII membrane (24). The reduced binding affinity cloning site of pBluescriptII phagemid vector (Stratagene) and sub- was related to modification of two or three acidic amino acid cloned into SacI/HindIII site of the M13mp18 phagemid vector. The 157 168 212 247 residues located in domains Asp –Asp and Glu –Gln , TM Sculpture in vitro mutagenesis system (Amersham Biosciences) was because these residues were modified only when the protein used to introduce a desired mutagenesis into a M13mp18 template was free in solution. Mutation of Asp to other residues in (30 –32). The synthesized, mutagenic oligonucleotides employed are shown in Table I. The mutation was checked by sequencing the single- Synechocystis MSP affected oxygen evolution significantly (25). stranded DNA of each mutated psbO gene. The mutated psbO genes On the other hand, in another study, chemical modification of were expressed in Escherichia coli cells, and the mutant MSPs were carboxyl groups with glycine methyl ester failed to affect the purified by the same method employed for isolation of the wild-type ability of the spinach protein to reactivate oxygen evolution MSP (29). The yields of purified protein were 5–10 mg of protein/liter of upon reconstitution at a high protein/PSII ratio (20). Site- culture medium. directed mutagenesis of completely or partially conserved as- Preparation of Oxygen-evolving PSII Complexes—Oxygen-evolving PSII complexes were isolated from S. elongatus as described by partyl and glutamyl residues on spinach MSP was performed, Ichimura et al. (33). Thylakoid membranes suspended in solution A (50 but none of the mutations had a dramatic effect (26, 27). All of M MES/NaOH (pH 6.0), 10 mM NaCl, 5 mM MgCl m ) containing 1 M the mutant proteins were able to bind to PSII membranes, and sucrose (1 mg of chlorophyll/ml) were treated with 0.20 – 0.25% sucrose the ability of the protein to reactivate oxygen evolution was monolaurate for1hat20 °C in the dark. After the addition of an equal only marginally affected by alteration of several residues. volume of solution A to reduce the sucrose concentration to 0.5 M,2 volumes of the suspension was placed on 1 volume of solution A con- Thus, amino acid residues on MSP that are involved in func- taining 1 M sucrose and centrifuged at 300,000  g for 1 h. A green band tional interaction with PSII still remain to be identified. that appeared at the interface between 0.5 and 1 M sucrose contained We showed previously that cleavage of MSP from the ther- PSII complexes. The complexes were collected by centrifugation at mophilic cyanobacterium Synechococcus elongatus at a single 370,000  g for 40 min and suspended in solution A containing 25% 156 163 184 191 site between Phe and Gly , or between Agr and Ser , glycerol. Binding of MSP—PSII complexes were treated with 2.6 M urea and by trypsin, chymotrypsin, or lysylendopeptidase resulted in 0.2 M NaCl for 30 min at 0 °C to extract the three extrinsic proteins, total loss of the ability of the protein to bind to PSII and to precipitated by centrifugation, and suspended in solution A containing restore the oxygen-evolving activity (28). Introduction of a nick 25% glycerol. Washed PSII complexes were incubated with each mutant 156 157 between Phe and Leu , a cleavage site specific to chymo- protein in 20 mM MES/NaOH (pH 6.0), 10 mM CaCl , and 10% glycerol trypsin, by site-directed mutagenesis also led to loss of the at 0 °C at an indicated protein/PSII ratio. A stoichiometry of 45 chloro- binding and reactivating ability of MSP. The cause of the phylls/PSII was assumed for the PSII complexes (33). After incubation for 1 h, the PSII complexes were precipitated by centrifugation at inactivation of MSP may, however, not be related to a local 370,000  g for 40 min and then washed once with and resuspended in change at the protease-cleavage site, because the mutation was M MES/NaOH (pH 6.0), 10 mM MgCl 50 m , and 25% glycerol. Wild-type accompanied by a significant change in the secondary confor- and mutant MSPs bound to PSII were analyzed by SDS-PAGE. Sam- mation of the protein. Insertion of a methionyl residue between M dithiothreitol for 30 min, ples were denatured with 5% SDS and 60 m 156 157 Phe and Leu abolished functional binding of MSP and then applied to polyacrylamide gels. The acrylamide concentration was 4.5% for the stacking gel and 12.5% for the resolving gel, and both gels unexpectedly created an appreciable binding affinity for non- contained 0.1% SDS and 6 M urea. After electrophoresis, gels were specific sites of the PSII complex. Because the protein confor- stained with Coomassie Brilliant Blue R-250 and scanned at 560 nm mation was little affected by insertion of the residue, the ob- with an ATTO AE-6900 densitometer. Amounts of MSPs bound to PSII served changes in the protein binding were ascribed to a small were estimated by measuring the peak area of the protein with that of structural change(s) in the mutated polypeptide. Based on CP47 as reference. these observations, we proposed that the two protease-sensi- Determination of Oxygen Evolution—Oxygen evolution was deter- 156 163 184 191 mined with a Clark-type oxygen electrode at 40 °C as described (33). tive regions, Phe –Gly and Arg –Ser , are possible White light of a saturating intensity was provided from a halogen lamp candidates for interaction with the PSII complex. (Watanabe Shoko R&D Corp). The reaction medium contained 50 mM In the present study, site-directed mutagenesis was per- MES/NaOH (pH 6.0), 10 mM NaCl, 5 mM MgCl ,10mM CaCl ,1 M 2 2 formed to identify the amino acid residues involved in effective sucrose, and 0.4 mM 2,6-dichloro-p-benzoquinone. binding of Synechococcus MSP to PSII. Conserved, charged Circular Dichroism Spectrometry—Far-UV and near-UV CD spectra were determined with a JASCO J-820 spectrometer. Data were col- residues in the first protease-sensitive sequences were substi- lected every 0.1 nm with a bandwidth of 1 or 2 nm and a 0.5-s time tuted with an uncharged residue, and the ability of the mutant constant. For determination of the far-UV CD spectra, proteins were proteins to bind to urea/NaCl-washed PSII and to restore the dissolved in 20 mM potassium phosphate (pH 6.5) at a protein concen- oxygen evolution was determined. Charge-preserving substitu- tration of 80 g/ml, which was determined by amino acid analysis. tion with an identically charged residue was also performed to Eight repetitive scans were collected and averaged; the secondary struc- ture elements were analyzed with the program VARSLC (34, 35). examine whether the residues are involved in electrostatic Near-UV CD spectra were determined with proteins dissolved in 20 mM interaction. For comparison, conserved, uncharged residues in 156 163 MES/NaOH (pH 6.5). Measurement was repeated 6 –16 times depend- and near the first protease-sensitive sequence Phe –Gly ing on the protein concentrations (see Fig. 4), and the signals obtained and conserved, charged residues in other domains of the pro- were averaged. tein were altered. The results obtained indicate that Arg , 156 158 160 162 RESULTS Phe , Asp , Lys , and Arg located in or near the first protease-sensitive sequence are essential for binding of MSP to Substitution of Charged Amino Acid Residues—There are its functional site. Mutant proteins were also constructed by several lines of evidence indicating that binding of MSP to its Binding of MSP to Photosystem II 14749 TABLE I Design of oligonucleotides used for the site-directed mutagenesis of MSP Mutants Mutagenetic oligonucleotides K59Q 5-CGGCCTCTTGGCGTTGGTTCTTAGGCTCTTCC-3 R73Q 5-CAAGCTCGTGGTTTCTTGGGTCACCAGTTTGG-3 K123Q 5-TCGCCACCAAGTTTTGCACTGTAAACAGCAG-3 V148T 5-GTGCGGTAGGAAGGAGTGTTGAATTCCCCCTTG-3 P149A 5-GTGCGGTAGGAAGCAACGTTGAATTCCCC-3 Y151F 5-TTGGCGGTGCGGAAGGAAGGAACGTTG R152Q 5-TCGAGGAAGTTGGCGGTTTGGTAGGAAGGAACGTTG-3 R152K 5-GAGGAAGTTGGCGGTTTTGTAGGAAGGAACGTTG-3 F156L 5-TCCTTTGGGATCGAGGAGGTTGGCGGTGCGCTA-3 F156Y 5-GTCCTTTGGGATCGAGGTAGTTGGCGGTGCGGTAG-3 156G157 5-GTCCTTTGGGATCGAGACCGAAGTTGGCGGTGCG-3 156A157 5-GTCCTTTGGGATCGAGGGCGAAGTTGGCGGTGCG-3 156V157 3-GTCCTTTGGGATCGAGAACGAAGTTGGCGGTGCG-3 157L 5-GTCCTTTGGGATCGAAGTTGGCGGTGCG-3 D158N 5-CCCCGTCCTTTGGGATTGAGGAAGTTGGCGGTG-3 D158E 5-GGGCGATCGCCGATTCATAACCGGAGGC-3 P159A 5-CCCCGTCCTTTGGCATCGAGGAAGTTGGC-3 K160Q 5-CCAAGCCCCGTCCTTGGGGATCGAGGAAGTTG-3 K160R 5-GGCCAAGCCCCGTCCGCGGGGATCGAGGAAGTTGG-3 G161Q 5-GGCCAAGCCCCGTTGTTTGGGATCGAGGAAG-3 R162Q 5-CCGGAGGCCAAGCCCTGTCCTTTGGGATCGAG-3 R162K 5-ACCGGAGGCCAAGCCTTTTCCTTTGGGATCGAGG-3 G163Q 5-CATAACCGGAGGCCAATTGCCGTCCTTTGGGATC-3 G167Q 5-CGATCGCCGAGTCATATTGGGAGGCCAAGCCCC-3 Y168F 5-CGATCGCCGAGTCAAAACCGGAGGCCAAGC-3 K188Q 5-GTGAGGGAGAAGCGTTGGACATTGGCGCGAGC-3 K234Q 5-AAGACCCCTTGAATTTGTACTTCGTGAGGTTC-3 The nucleotide sequences are of the complementary strand. The site-directed nucleotide changes are underlined. functional site involves electrostatic interactions between the high-affinity binding of the protein to PSII was appreciably charged amino acid residues on the protein and oppositely reduced by the mutation. The results show that all three of the 156 163 charged residues on intrinsic proteins of PSII (6, 7, 36, 37). conserved, charged residues located in the Phe –Gly se- Single-amino acid substitution experiments were, therefore, quence are required for high affinity binding of the protein to performed for conserved, charged residues located in the two PSII, consistent with the notion that the first protease-sensi- protease-sensitive sequences. The first protease-sensitive se- tive sequence is a domain for interaction with PSII. The high 156 163 quence, Phe –Gly , contains two positively charged amino affinity binding, however, was also severely affected by alter- 160 162 152 ation of Arg acid residues, Lys and Arg , and a negatively charged to Gln, which is located three residues upstream residue, Asp , which are completely conserved among 14 of the first protease-sensitive sequence. This indicates that the 156 163 binding domain is larger than the Phe sequences of MSPs from cyanobacteria, algae, and higher –Gly sequence and 148 152 plants (Fig. 1). Mutant proteins were constructed by replacing possibly involves the Val –Arg sequence, which is strictly 160 162 Lys and Arg with Gln (K160Q,R162Q), and by replacing conserved in all the sequences of MSP shown in Fig. 1. Asp with Asn (D158N). Binding affinities of the mutant The effects of the substitution mutations on the ability of MSP to restore the oxygen-evolving activity were also investi- proteins to PSII were determined by measuring the amount of each mutant protein bound to urea/NaCl-washed PSII at dif- gated. As shown in Table II, oxygen evolution was strongly ferent protein/PSII ratios. Reconstituted PSII complexes were suppressed by urea/NaCl washing, but the activity was par- washed once to remove loosely bound MSP. As shown in Fig. 2, tially restored by the addition of wild-type MSP. This reflects binding of wild-type MSP was saturated at a protein/PSII ratio the normal function of the protein because the small magnitude of two. A further increase in the protein/PSII ratio led to only a of reactivation (about 30% of activity in untreated PSII com- small increase in the amount of the protein bound, reflecting plexes) can be ascribed to the following two reasons. First, the the ability of the protein to bind specifically and stoichiomet- full activity of oxygen evolution of cyanobacterial or red algal rically to its functional site on the PSII complex. On the other PSII requires three or four extrinsic proteins (38 – 40); MSP hand, binding of the protein to PSII complexes was dramati- only partially supports the activity. Second, the smaller reac- cally altered by the mutations introduced. The binding curves tivating effect of MSP may be because a larger proportion of of the K160Q and R162Q mutant proteins were sigmoidal, and PSII preparations had been irreversibly damaged during pro- the amounts of the two mutant proteins bound to PSII at low tein extraction. This, however, will in principle not affect our protein/PSII ratios were much lower than that of the wild-type reconstitution results as far as the comparison of the relative 160 162 protein, indicating that alteration of Lys effects of wild-type and mutant MSPs is concerned. The muta- and Arg leads to a large decrease in the binding affinity for the functional site. tions that strongly affected the high affinity binding of the At protein/PSII ratios above four, the amount of mutant pro- protein had a severe impact on the reactivation ability of MSP. teins bound increased linearly with an increase in protein The levels of oxygen evolution restored by the mutant MSPs concentration; this increase is apparently steeper than that of were only less than 20% of that restored by wild-type protein the wild-type protein. This is a feature of nonspecific binding even at a protein/PSII ratio of five. The results confirm that the that has been described first for mutant proteins with Met four charged residues are essential for functional interaction 156 157 157 inserted between Phe and Leu and with Leu replaced with PSII and that the nonspecific binding to PSII is nonfunc- tional. Table II also shows that not all of the strictly conserved, by Met (28). Nonspecific binding was particularly enhanced upon substitution of Asp with Asn so that the amount of the charged residues are required for the function of MSP. Substi- tution of Lys D158N mutant protein bound to PSII far exceeded that of the , the only strictly conserved, charged residue in 184 191 wild-type protein at a protein/PSII ratio of six. Nevertheless, the second protease-sensitive sequence, Arg –Ser with 14750 Binding of MSP to Photosystem II FIG.1. Alignment of amino acid sequences of MSP from 14 oxygenic photosynthetic organisms. Amino acid residues substituted or deleted by site-directed mutagenesis in the present study are indicated by triangles. Dashes indicate amino acid residues identical to those in S. Binding of MSP to Photosystem II 14751 197–250 nm region as well as Fourier-transform infrared spec- trum (41). Because CD data provide more accurate estimates of the proportions of secondary structure elements when the CD spectrum is extended to shorter wavelengths (35), the CD spec- tra of wild-type and mutant proteins were determined from 181 to 260 nm. The far-UV CD spectrum of wild-type protein showed a strong positive band at 197 nm, a broad negative band between 205 and 225 nm, and a zero-line crossover point 152 158 160 at 192 nm (Fig. 3). Substitution of Arg , Asp , Lys , and Arg had no significant effect on the spectrum (Fig. 3 and data not presented). The relative abundance of secondary structure elements estimated from the far-UV CD data is shown in Table III. The wild-type protein contains -sheet severalfold more abundantly than -helix. The values are sim- ilar to those of spinach MSP (42– 44) but different from the previous estimate from the CD spectrum in the 197–250 nm region, which predicted a higher proportion of -helix and a 152 158 160 FIG.2. Effects of substitution of Arg , Asp , Lys , and lower proportion of -sheet than those shown in Table III (41). Arg with an uncharged residue on the binding affinity of MSP Although not shown here, Fourier-transform infrared spectros- to PSII. Urea/NaCl-washed PSII complexes were reconstituted with copy with an improved precision also indicated a larger abun- each mutant protein at the indicated protein/PSII ratios. The amount of wild-type protein bound to PSII at the protein/PSII ratio of three was dance of -sheets than shown previously (41). The mutations taken as 100%. , wild-type MSP; ‚, D158N; Œ, R152Q; E, K160Q; , that significantly decreased the binding and reactivating abil- R162Q. ities of MSP had no significant effect on the secondary confor- mation of the protein; small variations in the proportion of the TABLE II Effects of substitution of a charged amino-acid residue with an secondary structure elements observed are in the range of uncharged residue on the ability of MSP to restore experimental error. Thus, loss of the functional binding to PSII oxygen-evolving activity is not a result of distortion of the secondary conformation of the Urea/NaCl-washed PSII complexes were reconstituted with each mu- protein. tant protein at a protein/PSII ratio of five. The near-UV CD spectra were determined to examine Oxygen Preparations Additions Reactivation whether tertiary structure of the protein is affected by the evolution substitution mutations. The near-UV CD spectrum of spinach mol of % MSP shows two prominent bands, a tryptophan band at 294 nm O /mg of and a tyrosine band at 285 nm. These bands disappeared when chlorophyll/h the tertiary structure that contributes to the local environment PSII complexes None 1153 Washed PSII complexes None 41 0 of the aromatic amino acids was altered by cleavage of the Washed PSII complexes Wild-type MSP 306 100 sulfhydryl bond (45), acidification (43), or heat treatment (44) Washed PSII complexes D158N 89 18 of the protein. Synechococcus MSP lacks tryptophan, and ac- Washed PSII complexes K160Q 71 11 cordingly no band at 294 nm was observed (Fig. 4). Moreover, Washed PSII complexes R162Q 88 18 Washed PSII complexes R152Q 82 16 and unexpectedly, no conspicuous tyrosine band was observed PSII complexes None 1065 at 285 nm. A simple explanation would be that the 285 nm Washed PSII complexes None 78 0 band originates from one (or more) of the three tyrosine resi- Washed PSII complexes Wild-type MSP 360 100 dues that are present in the spinach protein but not in the Washed PSII complexes K59Q 345 94 Synechococcus protein (see Fig. 1). Fig. 4, however, shows four Washed PSII complexes R73Q 391 111 Washed PSII complexes K123Q 368 103 positive bands at 261, 266, 274, and 282 nm, which can be Washed PSII complexes K188Q 329 89 ascribed to the fine-structure bands of phenylalanine and ty- Washed PSII complexes K234Q 352 97 rosine (46). A more likely explanation, therefore, is that a peak at 290 nm corresponds to a tyrosine band that is overlapped with a sharp negative band at shorter wavelengths. In any Gln, had only a marginal effect on the reactivation ability of case, these spectral features remained essentially unaltered MSP. This casts a doubt on the notion that the second protease- upon substitution of any of the four charged residues, indicat- sensitive sequence is a region for interaction with PSII. The 59 73 123 234 ing that inactivation of the mutant MSPs was not associated alteration of Lys , Arg , Lys , and Lys to Gln had no effect on the function of MSP. Thus, inactivation of MSP is with a significant change in the tertiary structure of the pro- tein, either (Fig. 4 and data not presented). specific to the substitution of a charged residue in the Val – Gly sequence. The present study, however, does not exclude If the four charged residues contribute to the effective bind- ing of MSP through electrostatic interaction with oppositely the existence of a binding site for PSII in other regions of the protein because there are still eight strictly conserved, charged charged residues on the PSII complex, substitution of each residues that remain to be investigated (see Fig. 1). residue with an identically charged residue might have little CD spectroscopy was performed to investigate the effect of impact on the function of the protein. This was found to be the the substitution mutations on the secondary and tertiary con- case with Asp . Alteration of the aspartyl residue to a glu- formations of MSP. The secondary conformation of the protein tamyl residue had no effect on the binding of the protein to PSII has previously been estimated from the UV CD spectrum in the (Fig. 5). The D158E mutant protein was able to reconstitute elongatus; dots indicate vacant position in the sequence; stars indicate residues completely conserved in the 14 sequences. The two protease- sensitive sequences are boxed. 1, Synechococcus elongatus; 2, Anacystis nidulans R2; 3, Synechocystis sp. PCC6803; 4, Anabaena sp. PCC7120; 5, Chlamydomonas reinhardtii; 6, Euglena gracilis Z; 7, garden pea; 8, Arabidopsis thaliana; 9, wheat; 10, potato; 11, tomato; 12, tobacco; 13, rice; 14, spinach. 14752 Binding of MSP to Photosystem II FIG.3. Far-UV CD spectra of wild- type and mutant MSPs with Arg replaced by Gln and Asp replaced by Asn. Thick line, wild-type MSP; thin line, R152Q; dotted line, D158N. TABLE III nonspecific sites (not shown). Alteration of Phe to Tyr had no Secondary structural elements of wild-type and mutant MSPs effect on the function of the protein. This indicates that the determined by far-UV CD spectra structure of the side group is crucial for effective interaction Sample -Helix -Sheet Turns Other 161 163 with PSII. Substitution of Gly or Gly with Gln led to a % small decline in the reactivating ability of MSP, whereas alter- Wild type 7 39 21 33 ation of Pro to Ala had no effect. 153 155 164 166 R152Q 8 38 21 33 The Thr –Asn and Leu –Ser sequences, which are K160Q 8 38 21 33 located adjacent to the first protease-sensitive sequence, are R162Q 7 39 21 34 not conserved even in cyanobacterial MSPs. Replacement of a D158N 9 38 21 33 single residue in these two sequences or the entire three-amino acid sequences had no effect on the ability of MSP to restore the oxygen-evolving activity (not shown). Although Arg is re- oxygen evolution activity as effectively as wild-type protein 148 149 quired for the functional binding of MSP, Val , Pro , and (Table IV). Thus, a negative charge at Asp is critically im- 151 148 152 Tyr , located in the conserved Val –Arg sequence, could portant for the functional binding of MSP. In contrast, charge- 152 160 162 be replaced by Thr, Ala, and Phe, respectively, without affect- preserving alteration of Arg to Lys, Lys to Arg, or Arg ing the reactivating ability of MSP. No attempt was made to to Lys resulted in a significant decrease in the high affinity 150 167 168 alter Ser . Gly and Tyr are also completely conserved in binding and an increase in nonspecific binding to PSII (Fig. 5). the 14 sequences of MSP. The reactivating ability of MSP was The amounts of the R152K, K160R, and R162K mutant pro- moderately affected by substitution of Tyr with Phe, teins bound to PSII at low protein/PSII ratios appear to be whereas replacement of Gly by Gln had no influence on the larger than those of the corresponding charge-deleted mutant function of MSP. proteins. The difference should be ascribed to the difference in 156 157 Insertion of a Residue between Phe and Leu and dele- nonspecific binding, because the charge-preserving mutations tion of Leu —We showed previously (28) that MSP becomes affected the ability of the protein to reconstitute oxygen evolu- unable to bind to its functional site and reactivate oxygen tion as severely as the corresponding charge-deleting muta- evolution when a methionyl residue is inserted between Phe tions. Only the K160R mutant protein restored oxygen evolu- and Leu . The mutant protein, however, showed a strong tion somewhat more effectively than the K160Q mutant affinity for nonspecific binding. Extension of this experiment protein, suggesting participation of this residue in electrostatic interaction (Table IV). provided important information on both effective and nonspe- cific interaction of MSP with PSII. Because the protein confor- Substitution of Uncharged Residues—Five uncharged resi- 156 157 159 161 163 dues, Phe mation was little affected by the insertion of Met, loss of the , Leu , Pro , Gly , and Gly , which are 156 163 functional binding of the protein is ascribed to a small struc- located in the Phe –Gly sequence, are strictly conserved tural change in the protease-sensitive sequence. If this inter- among MSPs from cyanobacteria to higher plants, except for Gly , which is conservatively replaced by Ala in two cya- pretation was correct, the inactivation of MSP can be ascribed to an effect of either the side group of the methionyl residue nobacteria (see Fig. 1). Substitution of Phe with Leu resulted in a nearly total loss of the reactivating ability of MSP (Table inserted or a small increase in the polypeptide length between 156 156 157 V). Thus, Phe Phe and Leu . To examine the effect of the side group, is required for effective binding of the protein, although the effects of the mutation on binding of the protein to mutant MSPs were constructed by inserting an amino acid its functional site could not be determined accurately because residue that has a side group smaller than that of Met between 156 157 of an extremely strong affinity the mutant protein exhibited for Phe and Leu . The three mutant MSPs, which were con- Binding of MSP to Photosystem II 14753 FIG.4. Near-UV CD spectra of wild- type and mutant MSPs with Arg or Arg replaced by Gln. Solid line, wild- type MSP (985 g/ml); dashed line, R152Q (827 g/ml); dotted line, R162Q (1760 g/ml). TABLE IV Effects of substitution of a charged amino acid residue with an identically charged residue on the ability of MSP to restore oxygen-evolving activity Experimental conditions are the same as in Table II. Oxygen Preparations Additions Reactivation evolution mol of % O /mg of chlorophyll/h PSII complexes None 1147 Washed PSII complexes None 96 0 Washed PSII complexes Wild-typeMSP 353 100 Washed PSII complexes D158E 382 110 PSII complexes None 1090 Washed PSII complexes None 40 0 Washed PSII complexes Wild-typeMSP 306 100 Washed PSII complexes K160R 136 36 Washed PSII complexes R162K 79 15 Washed PSII complexes R152K 108 26 TABLE V Effects of substitution of an uncharged amino acid residue on the ability of MSP to restore oxygen-evolving activity 152 158 160 FIG.5. Effects of substitution of Arg , Asp , Lys , and Urea/NaCl-washed PSII complexes were reconstituted with each mu- Arg with an identically charged residue on the binding affin- tant protein at the protein/PSII ratios of five except that the ratio of four ity of MSP to PSII. Experimental conditions were the same as in Fig. was used for of G163Q and G167Q. 2. , wild-type MSP; ‚, D158E; Œ, R152K; E, K160R; , R162K. Additions Reactivation structed by inserting Gly, Ala, and Val, are called 156G157, None 0 156A157 and 156V157, respectively. Leu is not es- Wild-type MSP 100 sential for the functional binding of MSP to PSII because the V148T 121 residue could be substituted with Met without affecting the P149A 113 reactivating ability of the protein (28). In the fourth mutant Y151F 94 protein (157L), therefore, Leu was deleted to examine F156Y 98 F156L 4 whether the protein binding is affected by a change in the P159A 104 polypeptide length. As shown in Table VI, the protein became G161Q 73 totally ineffective in reactivation of oxygen evolution when Val, G163Q 71 Ala, or even Gly, which has no side chain, was inserted between G167Q 110 156 157 Y168F 79 Phe and Leu , indicating that inactivation of the protein is independent of the size and structure of side chain of the residue inserted. Deletion of Leu also led to a large decline in the reactivating ability of the protein. None of the four mutant protein to PSII (data not shown). These results indicate that proteins bound to PSII to an appreciable amount at the protein/ binding of MSP to its functional site is extremely sensitive to a PSII ratio of five, indicating that the mutations abolished high small change in the polypeptide length in the region between 156 157 affinity binding without enhancing nonspecific binding of the Phe and Leu . The observation that nonspecific binding 14754 Binding of MSP to Photosystem II TABLE VI 156 157 Effects of insertion of Gly, Ala, or Val between Phe and Leu and deletion of Leu on the ability of MSP to restore the oxygen-evolving activity Experimental conditions are the same as in Table II. Oxygen Preparations Additions Reactivation evolution mol of % O /mg of FIG.6. Mutations that affected the function of MSP. The large chlorophyll/h and small upward arrows indicate mutations that affected the ability of PSII complexes None 889 Synechococcus MSP to reactivate oxygen evolution strongly and weakly, Urea/NaCl-washed PSII None 81 0 respectively. Downward arrows indicate mutations that had no effect complexes on the reactivating ability of the protein. The first protease-sensitive Urea/NaCl-washed PSII Wild-type MSP 337 100 sequence is boxed. Asterisks indicate residues completely conserved in complexes the 14 sequences of MSP. Urea/NaCl-washed PSII 156G157 87 2 complexes Urea/NaCl-washed PSII 156A157 85 2 and cyanobacteria or in the flexibility of proteins between me- complexes sophiles and thermophiles. Urea/NaCl-washed PSII 156V157 89 3 66 76 130 159 Six conserved lysyl residues, Lys , Lys , Lys , Lys , complexes 186 236 Lys , and Lys , of spinach MSP were suggested to be lo- Urea/NaCl-washed PSII 157L 94 5 complexes cated in regions of the protein in contact with PSII because they were modified with amino group-specific reagents when MSP was free in solution but not when the protein was bound appeared upon insertion of Met but not upon insertion of Gly, to PSII membranes (19, 20). The present study shows that not Ala, or Val implies that the structural factor that confers an all of the conserved lysyl residues are required for functional affinity for nonfunctional sites on the protein is the side group 59 123 188 binding to PSII. Substitution of Lys , Lys , Lys , and of Met inserted. This is consistent with the previous observa- 234 66 Lys of Synechococcus MSP (which correspond to Lys , tion that nonfunctional binding became appreciable upon re- 130 190 236 Lys , Lys , and Lys of the spinach protein, respectively) placement of Leu by Met (28). with Gln had no influence on the function of the protein. Al- 160 159 DISCUSSION teration of Lys (which corresponds to Lys of the spinach We have constructed more than 30 variants of Synechococcus protein) to Gln had, however, a severe impact on the ability of MSP by substituting, inserting, or deleting an amino acid res- the protein to effectively bind to PSII. Loss of protein binding is idue to investigate whether the protease-sensitive Phe not related to a change in the secondary or tertiary structure of 163 184 191 Gly and Arg –Ser sequences are regions for functional the protein. The effect of the charge-preserving substitution of Lys with Arg on the reactivating ability of the protein was binding to PSII. The cyanobacterial MSP served as an excellent tool for investigation of the protein binding by means of site- somewhat less than that of the charge-deleting substitution of directed mutagenesis and in vitro reconstitution, allowing us to the residue with Gln. This can be explained by assuming that identify several charged and uncharged amino acid residues Lys participates in the electrostatic interaction essential for that are essential for binding of MSP to its functional site. The the function of the protein and that Arg acts only imperfectly mutations that effectively altered the ability of MSP to reacti- for Lys due to the difference in the side group. The corre- vate oxygen evolution are illustrated in Fig. 6. sponding lysyl residue (Lys ) on spinach MSP was unable to The first protease-sensitive sequence contains a highly con- participate in an intramolecular charge-pair interaction be- served acidic residue, Asp cause the residue was accessible to amino group modifiers . Substitution of this residue with Asn resulted in a large decrease in the ability of MSP to bind to when the protein was free in solution (20). It is highly likely, functional site and to restore the oxygen-evolving activity, therefore, that Lys contributes to binding of MSP to its whereas its charge-preserving replacement by Glu had no in- functional site by electrostatically interacting with a negatively hibitory effect, indicating that the negative charge at Asp charged residue on an intrinsic protein of the PSII complex. is required for the protein binding. Thus, Asp is involved either The contribution of arginyl residues to binding of spinach MSP has been suggested by chemical modification of the gua- in charge-pair interaction with an oppositely charged residue on an intrinsic protein of PSII or in intramolecular charge-pair nidino group of arginyl residues with 2,3-butanedione (20). interaction necessary for maintenance of a functional structure None of the six arginyl residues present in MSP reacted with of MSP. Far- and near-UV CD spectroscopy demonstrated that the reagent when the protein was bound to PSII membranes, inactivation of MSP by mutation of Asp whereas treatment of the protein in solution led to modification to Asn was not accompanied by a significant change in the secondary and of four arginyl residues, which was accompanied by loss of the tertiary conformation of the protein, respectively. Substitution ability of the protein to bind to PSII and to reactivate oxygen of the corresponding aspartyl residue (Asp ) in spinach MSP evolution. Modified residues have not been identified yet be- cause of the instability of the modified products (20). Site- was also shown to have no influence on intramolecular salt bridges formed upon cross-linking with EDC (17). In addition, directed mutagenesis allowed us to successfully identify argi- 157 168 carboxyl groups in the domain Asp nyl residues that are essential for the protein binding and also –Asp of the spinach protein were shown to participate in the binding of spinach provided important information as to a domain of the protein MSP to PSII (24). These observations favor the involvement of for effective binding to PSII. There are three arginyl residues, 158 73 152 162 Asp in an intermolecular charge-pair interaction between Arg , Arg , and Arg , which are completely conserved among MSPs from cyanobacteria, algae, and higher plants (Fig. MSP and PSII. Substitution of the corresponding aspartyl res- idue on MSPs from other photosynthetic organisms had weaker 1). Arg is not essential because the residue could be substi- effects, a 5–15% decline in the reactivating ability of spinach tuted with Gln without affecting the function of the protein. 152 162 MSP (27) and a 35– 40% reduction in the rate of oxygen evolu- Mutations at either Arg or Arg affected the function of tion in Synechocystis cells (25). These differences may be re- MSP significantly. Because these mutations did not induce any lated to differences in protein binding between higher plants significant changes in the protein conformation, the results Binding of MSP to Photosystem II 14755 indicate that the two arginyl residues directly participate in arranged spatially so as to become in van der Waals contact effective interaction with PSII. No direct evidence was obtained with respective partner residues on the PSII complex. As such, 152 162 indicating that Arg and Arg are involved in charge-pair the protein binding will be weakened or abolished by insertion interaction; the reactivating ability of MSP was nearly equally or deletion of a residue, which alters the location of the charged 148 163 residues in the Val –Gly domain. suppressed by charge-deleting and charge-preserving substitu- 148 163 tions of each residue. A possibility remains, however, that the The Val –Gly domain becomes a region for nonfunc- two residues participate in charge-pair interaction but that a tional binding to PSII upon modification. Replacement of a single essential residue in the domain conferred on the protein positive charge each residue carries becomes incapable of in- teracting with a negative charge on the partner residue when the ability to nonspecifically bind to the PSII complex. A com- Arg is altered to Lys. Because Arg is located three residues parison of the insertion mutants constructed in the present and apart from the first protease-sensitive sequence, the result also previous experiments (28) showed that the structure of the side shows that the binding domain of MSP is larger than suggested chain of an inserted residue is critical for the nonspecific bind- ing of the protein. It is concluded, therefore, that the conserved by the proteolytic experiments (28) and may involve the well 148 163 148 152 conserved Val –Arg sequence. amino acid residues in the Val –Gly domain are also im- Substitution of uncharged residue in the first protease-sen- portant for the protein to minimize nonspecific interactions with PSII. Several mutant proteins exhibited an extremely sitive sequence also affected the function of MSP. Substitution of Phe with Leu resulted in an almost complete loss of the strong affinity for nonspecific binding. This implies caution in using binding of MSP to PSII as an index of the functional ability of the protein to restore the oxygen-evolving activity, integrity of the protein when reconstitution is performed at a whereas replacement of the same residue by Tyr had no effect on the function of MSP. Thus, the aromatic side group of Phe low protein/PSII protein ratio. In contrast to the first protease-sensitive sequence where all is required for the functional binding of the protein. The reac- the eight residues are completely conserved except for Gly , tivating ability of MSP was moderately affected by substitution 161 163 which is conservatively replaced, the second protease-sensitive of either Gly or Gly with Gln. This suggests that the side 184 191 Arg –Ser sequence has only two strictly conserved resi- chain of Gln, which replaced Gly sterically, interferes with the 186 188 dues, Asn and Lys . No evidence was obtained to support protein-protein interaction. Alternatively, because Gly has no the proposition that a binding site for PSII is located in the side chain, it may confer a greater conformational flexibility, 184 191 188 Arg -Ser sequence. Substitution of Lys , the sole strictly possibly required in the first protease-sensitive region for the effective association of the protein with its functional site. The conserved charged residue in the sequence, with Gln had only a marginal effect on the reactivating ability of MSP. This is protein binding, however, was not appreciably affected by sub- consistent with the results with spinach MSP, where the cor- stitution of Pro with Ala; this may give another constraint responding lysyl residue (Lys ) was accessible to the water- on the polypeptide conformation. soluble reagents even when the protein was bound to PSII Conserved, uncharged residues located near the Phe – 163 152 membranes (19, 20). There is another charged residue, Arg , Gly sequence were also examined. As stated above, Arg is in the second protease-sensitive sequence, which is conserved located in the completely conserved five-amino acid sequence. in two other cyanobacteria and conservatively replaced by Although substitution of three uncharged residues, Val , 149 151 Lys in other photosynthetic organisms. Because the lysyl resi- Pro , and Tyr , in the sequence failed to affect the function due (Lys ) of spinach MSP was modified with specific re- of MSP, the result should be interpreted with caution because 148 151 agents only when the protein was free in solution (19, 20), a Val and Tyr were replaced by residues with relatively possibility remains that Arg participates in the functional small differences in the structure of the side chain. In partic- binding of Synechococcus MSP. ular, because Tyr was substituted with Phe, a possibility remains that an aromatic group at position 151 plays a role in Acknowledgments—We are grateful to Dr. T. Takakuwa, Nihon the functional binding of the protein, which will be addressed Bunko Co., for measurement and analysis of CD spectra. We thank Dr. in future studies. Furthermore, amino acid residues in the Jian-Ren Shen of the Institute of Physical and Chemical Research 167 169 (Riken) for reading the manuscript. Gly –Asp sequence, which is located three residues after the first protease-sensitive sequence, are either strictly con- REFERENCES served or conservatively replaced. Possible contribution of this 1. Seidler, A. (1996) Biochim. Biophys. Acta 1277, 35– 60 sequence to the protein binding cannot be excluded because 2. Bricker, T. M., and Frankel, L. K. (1998) Photosynth. Res. 56, 157–173 substitution of Tyr with Phe weakly affected the function of 3. Murata, N., and Miyao, M. (1985) Trends Biochem. Sci. 10, 122–124 4. Shen, J.-R., Ikeuchi, M., and Inoue, Y. (1992) FEBS Lett. 301, 145–149 MSP, and single amino acid substitution has not been per- 5. Enami, I., Murayama, H., Ohta, H., Kamo, M., Nakazato, K., and Shen, J.-R. formed yet for Asp . (1995) Biochim. Biophys. Acta 1232, 208 –216 Extension of the previous experiments with the mutant pro- 6. Ono, T., and Inoue, Y. (1984) FEBS Lett. 166, 381–384 156 157 7. Miyao, M., and Murata, N. (1984) FEBS Lett. 170, 350 –354 tein having Met inserted (28) between Phe and Leu 8. Kuwabara, T., Miyao, M., Murata, T., and Murata, N. (1985) Biochim. Biophys. showed that binding of MSP to its functional site is extremely Acta 806, 283–289 9. Ono, T., and Inoue, Y. (1986) Biochim. Biophys. Acta 850, 380 –389 sensitive to a small change in the polypeptide length caused by 10. Enami, I., Satoh, K., and Katoh, S. (1987) FEBS Lett. 226, 161–165 156 157 insertion of an amino acid residue between Phe and Leu 11. Bricker, T. M., Odam, W. R., and Queirolo, C. B. (1988) FEBS Lett. 231, or deletion of Leu . This is consistent with the above de- 111–117 12. Enami, I., Kaneko, M., Kitamura, N., Koike, H., Sonoike, K., Inoue, Y., and scribed results that residues in or near the first protease- Katoh, S. (1991) Biochim. Biophys. Acta 1060, 224 –232 sensitive region participate in the interaction of MSP with PSII 13. Odom, W. R., and Bricker, T. M. (1992) Biochemistry 31, 5616 –5620 14. Mei, R., Green, J. P., Sayer, R. T., and Frash, W. D. (1989) Biochemistry 28, and further suggests that the relative location of these residues 5560 –5567 is critical for effective binding to PSII. We suggest that binding 15. Enami, I., Ohta, S., Mitsuishi, S., Takahashi, S., Ikeuchi, M., and Katoh, S. of MSP to its functional site involves multiple charged residues (1992) Plant Cell Physiol. 33, 291–297 148 163 16. Eaton-Rye, J. J., and Murata, N. (1989) Biochim. Biophys. Acta 977, 219 –226 in the Val –Gly region, which interact electrostatically 17. Seidler, A., Rutherford, A. W and Michel, H. (1996) Plant Mol. Biol. 31, with oppositely charged residues on intrinsic proteins of the 183–188 18. Qian, M., Al-Khaldi, S. F., Putnam-Evans, C., Bricker, T. M., and Burnap, PSII complex, although contribution of other interactions can- R. L. (1997) Biochemistry 36, 15244 –15252 not be excluded. Such a protein-complex interaction is possible 19. Frankel, L. K., and Bricker, T. M. (1995) Biochemistry 34, 7492–7497 only when the charged residues on the surface of MSP are 20. Miura, T., Shen, J.-R., Takahashi, S., Kamo, M., Nakamura, E., Ohta, H., 14756 Binding of MSP to Photosystem II Kamei, A., Inoue, Y., Domae, N., Takio, K., Nakazato, K., Inoue, Y., and 33. Ichimura, T., Miyairi, S., Satoh, K., and Katoh, S. (1992) Plat Cell Physiol. 33, Enami, I. (1997) J. Biol. Chem. 272, 3788 –3798 299 –305 21. Queirola, C. B. (1992) Doctoral dissertation, Louisiana State University, Baton 34. Manavalan, P., and Johnson, W. C., Jr. (1987) Anal. Biochem. 167, 76 – 85 Rouge 35. Johnson, W. C., Jr. (1990) Proteins Struct. Funct. Genet. 7, 205–214 22. Isogai, Y., Nishimura, M., and Itoh, S. (1987) Plant Cell Physiol. 28, 1493–1499 36. Enami, I., Tomo, T., Kitamura, M., and Katoh, S. (1994) Biochim. Biophys. 23. Seidler, A. (1996) Eur. J. Biochem. 242, 485– 490 Acta 1185, 75– 80 24. Frankel, L. K., Cruz, J. A., and Bricker, T. M. (1999) Biochemistry 38, 37. Enami, I., Kitamura, M., Tomo, T., Isokawa, Y., Ohta, H., and Katoh, S. (1994) 14271–14278 Biochim. Biophys. Acta 1186, 52–58 25. Burnap, R. L., Qian, M., Shen, J.-R., Inoue, Y., and Sherman, L. A. (1994) 38. Shen, J.-R., and Inoe, Y. (1993) Biochemistry 32, 1825–1832 Biochemistry 33, 13712–13718 39. Enami, I., Kikuchi, S., Fukuda, T., Ohta, H., and Shen, J.-R. (1998) Biochem- 26. Seidler, A., Ro ¨ ll, K., and Michel, H. (1992) in Current Research in Photosyn- istry 37, 2787–2793 thesis (Murata, N., ed) Vol. II, pp. 409 – 412, Kluwer Academic Publishers, 40. Enami, I., Yoshihara, S., Tori, A, Okamura, A., Ohta, H., and Shen, J.-R. Dordrecht (2000) Plant cell Physiol. 41, 1354 –1364 27. Seidler, A., Michel, H., and Rutherford, A. W. (1995) in Photosynthesis: From 41. Sonoyama, M., Motoki, M., Okamoto, G., Hirano, M., Ishida, H., and Katoh, S. Light to Biosphere (Mathis, P., ed) Vol. II, pp. 259 –262, Kluwer Academic (1996) Biochim. Biophys. Acta 1297, 167–170 Publishers, Dordrecht 42. Xu, Q., Nelson, J., and Bricker, T. M. (1994) Biochim. Biophys. Acta 1188, 28. Motoki, A., Shimazu, T., Hirano, M., and Katoh, S. (1998) Biochim. Biophys. 427– 431 Acta 1365, 492–502 43. Shutova, T., Irrgang, K.-D., Shubin, V., Klimov, V., and Renger, G. (1997) 29. Motoki, A., Shimazu, T., Hirano, M., and Katoh, S. (1995) in Photosynthesis: Biochemistry 36, 6350 – 6358 from Light to Biosphere (Mathis, P., ed) Vol. II, pp. 487– 490, Kluwer 44. Lydakis-Simantris, N., Hutchinson, R. S., Betts H. S., Barry, B. A., and Academic Publishers, Dordrecht Yocum, C. F. (1999) Biochemistry 38, 404 – 414 30. Taylor, J. W., Ott, J., and Eckstein, F. (1985) Nucleic Acids Res. 13, 8765– 8785 45. Tanaka, S., and Wada, K. (1988) Photosynth. Res. 17, 255–266 31. Nakamae, K. L., and Eckstein, F. (1986) Nucleic Acids Res. 14, 9679 –9698 46. Strickland E. H. (1976) in Handbook of Biochemistry and Molecular Biology 32. Sayers, J. R., Krekel, C., and Eckstein, F. (1992) BioTechniques 13, 592–596 (Fasman, G. D., ed) Vol. III, pp. 141–147, CRC Press, Cleveland, OH http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

A Domain of the Manganese-stabilizing Protein fromSynechococcus elongatus Involved in Functional Binding to Photosystem II

Motoki, Akihiro; Usui, Mina; Shimazu, Tsuneo; Hirano, Masahiko; Katoh, Sakae
Journal of Biological Chemistry – Apr 1, 2002
Download PDF
10 pages

Loading next page...
 
/lp/unpaywall/a-domain-of-the-manganese-stabilizing-protein-fromsynechococcus-67bSFJxDYD

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.m100766200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 277, No. 17, Issue of April 26, pp. 14747–14756, 2002 © 2002 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. A Domain of the Manganese-stabilizing Protein from Synechococcus elongatus Involved in Functional Binding to Photosystem II* Received for publication, January 26, 2001, and in revised form, January 22, 2002 Published, JBC Papers in Press, January 23, 2002, DOI 10.1074/jbc.M100766200 Akihiro Motoki‡§, Mina Usui‡, Tsuneo Shimazu‡, Masahiko Hirano‡, and Sakae Katoh¶ From the ‡Biological Sciences Department, Toray Research Center Inc., Kamakura 248-8555, Japan and the ¶Department of Biology, Faculty of Sciences, Toho University, Funabashi 274-8510, Japan Site-directed mutagenesis was performed to investi- cytochrome c , the 12-kDa protein, and a 20-kDa protein in red algae (5). MSP stabilizes the manganese cluster, which gate whether the two protease-sensitive sequences 156 163 184 191 -Gly and Arg -Ser , of the manganese-stabi- Phe catalyzes oxidation of water to molecular oxygen. Dissociation lizing protein (MSP) from a thermophilic cyanobacte- of the protein by washing with 1 M CaCl or MgCl (6) or 2.6 M 2 2 rium, Synechococcus elongatus (Motoki, A., Shimazu, T., urea plus 0.2 M NaCl (7) leads to a gradual release of two of four Hirano, M., and Katoh, S. (1998) Biochim. Biophys. Acta Mn ions present in the cluster. Oxygen evolution is strongly 1365, 492–502), are involved in functional interaction suppressed by the release of MSP, but the lost activity can be with photosystem II (PSII). The ability of MSP to bind to restored by rebinding of the protein (8, 9). Because no cofactor its functional site on the PSII complex and to reactivate is associated with MSP, it is considered that binding of the oxygen evolution was dramatically reduced by the sub- protein itself is responsible for optimal activity of oxygen 152 158 160 162 , Asp , Lys ,orArg with un- stitution of Arg evolution. charged residues, by insertion of a single residue be- Cross-linking experiments with various bifunctional re- 156 157 157 and Leu , or by deletion of Leu . tween Phe agents have shown that MSP is associated with or in close Substitution of each of the four charged residues with proximity to essentially all of the major intrinsic proteins of the an identically charged residue showed that the charges 158 160 PSII complex (10 –15). Various attempts have been made to , and possibly Lys , are important for the elec- at Asp locate binding sites for PSII or identify amino acid residues on trostatic interaction with PSII. The reactivating ability MSP that are involved in binding to its functional site on the to was also strongly affected by the alteration of Phe PSII complex. The N-terminal sequence of spinach MSP was , the only strictly conserved Leu. Replacement of Lys 184 191 suggested to have a binding site to PSII because removal of 16 –Ser sequence, by Gln charged residue in the Arg or 18 amino acid residues from its N terminus by protease had only a marginal effect on the function of MSP. High digestion resulted in total loss of the protein binding (16). affinity binding of MSP to PSII was also affected signif- , which is located in a re- Recently, evidence was presented indicating that the N-termi- icantly by mutation at Arg 148 152 –Arg ) strictly conserved among the 14 se- gion (Val nal sequence is necessary for maintaining the binding ability of quences so far reported. These results imply that the the protein to PSII but might not be involved in the intermo- 148 163 –Gly sequence, which is well conserved among Val lecular binding itself (17). It was also suggested that Asp , the MSPs from cyanobacteria to higher plants, is a domain only conserved, charged residue in the N-terminal 18-amino of MSP for functional interaction with PSII. acid sequence, might engage in both intra- and intermolecular interactions (18). Cross-linking with EDC, which covalently links amino and carboxyl groups in van der Waals contact, Photosynthetic oxygen evolution is mediated by PSII, a indicated that MSP is bound directly to CP47, an intrinsic multiprotein complex carrying chlorophylls, carotenoids, and a chlorophyll-carrying protein of the PSII complex through elec- variety of redox cofactors (for reviews see Refs. 1 and 2). Among trostatic interaction (10 –12). Analysis of EDC-cross-linked the protein components of PSII, the extrinsic 33-kDa protein is products showed that a charge-pair interaction exists between 1 76 also called the manganese-stabilizing protein (MSP) and is a charged residue in the Asp –Lys sequence of the spinach 364 440 associated with the lumenal surface of the PSII complex to- MSP and an oppositely charged residue in the Phe –Asp gether with two or three other extrinsic proteins, i.e. the 23- sequence of CP47 (13). kDa and 17-kDa proteins in higher plants and green algae (3), Chemical modification with specific reagents showed that six cytochrome c and a 12-kDa protein in cyanobacteria (4), and conserved lysyl residues distributed over the entire sequence of spinach MSP are accessible to at least one of the reagents when the protein is free in solution but not when the protein is * This work was performed under the management of the Research associated with PSII membranes (19, 20). This suggests that Association for Biotechnology as a part of the R&D Project of Basic these lysyl residues are located in domains of the protein that Technology for Future Industries, supported by the New Energy and are in contact with the PSII complex and might participate in Industrial Technology Development Organization. The costs of publica- 159 236 tion of this article were defrayed in part by the payment of page intermolecular interaction. Cross-linking of the Lys –Lys charges. This article must therefore be hereby marked “advertisement” domain of spinach MSP to CP47 with a reagent that cross-links in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. lysyl residues within a 1.2-nm radius has been reported (21). § To whom correspondence should be addressed: Biological Sciences Four arginyl residues of the spinach protein may also be in- Dept., Toray Research Center Inc., Kamakura 248-8555, Japan. Tel.: 81-467-32-9963; Fax: 81-467-32-0414; E-mail: akihiro_motoki@trc. volved in the binding to PSII because they were modified with toray.co.jp. a specific reagent only when the protein was free in solution The abbreviations used are: PSII, photosystem II; MSP, extrinsic (20). 33-kDa manganese-stabilizing protein; EDC, 1-ethyl-3-(3-dimeth- Involvement of negatively charged amino acid residues on ylaminopropyl)-carbodiimide; MES, 2-(N-morpholino)ethanesulfonic acid. MSP in binding to PSII was also suggested by EPR studies with This paper is available on line at http://www.jbc.org 14747 This is an Open Access article under the CC BY license. 14748 Binding of MSP to Photosystem II a free radical-relaxing agent dysprosium, which showed that inserting a residue with a side group smaller than that of Met 156 157 157 between Phe the binding site for MSP on the PSII complex is positively and Leu or by deleting Leu , in order to charged (22). Cross-linking experiments in which carboxyl elucidate the structural configuration of MSP around Phe and Leu groups of spinach MSP were activated for cross-linking prior to , which are required for the functional binding of MSP. reconstitution with PSII membranes or vice versa, showed that all carboxyl groups involved in intermolecular cross-linking EXPERIMENTAL PROCEDURES with EDC are located on MSP (23). The binding affinity of Site-directed Mutagenesis of the psbO Gene—Site-directed mutagen- spinach MSP was strongly reduced by chemical modification of esis was performed by the method described previously (28). The pET8c carboxyl groups of aspartyl and glutamyl residues with glycine expression vector, which contains the wild-type psbO gene of S. elon- gatus, was constructed as described (29). The XbaI/BamHI fragment of methyl ester when it was free in solution but not when it was the vector containing the whole psbO gene was cloned into the multiple bound to PSII membrane (24). The reduced binding affinity cloning site of pBluescriptII phagemid vector (Stratagene) and sub- was related to modification of two or three acidic amino acid cloned into SacI/HindIII site of the M13mp18 phagemid vector. The 157 168 212 247 residues located in domains Asp –Asp and Glu –Gln , TM Sculpture in vitro mutagenesis system (Amersham Biosciences) was because these residues were modified only when the protein used to introduce a desired mutagenesis into a M13mp18 template was free in solution. Mutation of Asp to other residues in (30 –32). The synthesized, mutagenic oligonucleotides employed are shown in Table I. The mutation was checked by sequencing the single- Synechocystis MSP affected oxygen evolution significantly (25). stranded DNA of each mutated psbO gene. The mutated psbO genes On the other hand, in another study, chemical modification of were expressed in Escherichia coli cells, and the mutant MSPs were carboxyl groups with glycine methyl ester failed to affect the purified by the same method employed for isolation of the wild-type ability of the spinach protein to reactivate oxygen evolution MSP (29). The yields of purified protein were 5–10 mg of protein/liter of upon reconstitution at a high protein/PSII ratio (20). Site- culture medium. directed mutagenesis of completely or partially conserved as- Preparation of Oxygen-evolving PSII Complexes—Oxygen-evolving PSII complexes were isolated from S. elongatus as described by partyl and glutamyl residues on spinach MSP was performed, Ichimura et al. (33). Thylakoid membranes suspended in solution A (50 but none of the mutations had a dramatic effect (26, 27). All of M MES/NaOH (pH 6.0), 10 mM NaCl, 5 mM MgCl m ) containing 1 M the mutant proteins were able to bind to PSII membranes, and sucrose (1 mg of chlorophyll/ml) were treated with 0.20 – 0.25% sucrose the ability of the protein to reactivate oxygen evolution was monolaurate for1hat20 °C in the dark. After the addition of an equal only marginally affected by alteration of several residues. volume of solution A to reduce the sucrose concentration to 0.5 M,2 volumes of the suspension was placed on 1 volume of solution A con- Thus, amino acid residues on MSP that are involved in func- taining 1 M sucrose and centrifuged at 300,000  g for 1 h. A green band tional interaction with PSII still remain to be identified. that appeared at the interface between 0.5 and 1 M sucrose contained We showed previously that cleavage of MSP from the ther- PSII complexes. The complexes were collected by centrifugation at mophilic cyanobacterium Synechococcus elongatus at a single 370,000  g for 40 min and suspended in solution A containing 25% 156 163 184 191 site between Phe and Gly , or between Agr and Ser , glycerol. Binding of MSP—PSII complexes were treated with 2.6 M urea and by trypsin, chymotrypsin, or lysylendopeptidase resulted in 0.2 M NaCl for 30 min at 0 °C to extract the three extrinsic proteins, total loss of the ability of the protein to bind to PSII and to precipitated by centrifugation, and suspended in solution A containing restore the oxygen-evolving activity (28). Introduction of a nick 25% glycerol. Washed PSII complexes were incubated with each mutant 156 157 between Phe and Leu , a cleavage site specific to chymo- protein in 20 mM MES/NaOH (pH 6.0), 10 mM CaCl , and 10% glycerol trypsin, by site-directed mutagenesis also led to loss of the at 0 °C at an indicated protein/PSII ratio. A stoichiometry of 45 chloro- binding and reactivating ability of MSP. The cause of the phylls/PSII was assumed for the PSII complexes (33). After incubation for 1 h, the PSII complexes were precipitated by centrifugation at inactivation of MSP may, however, not be related to a local 370,000  g for 40 min and then washed once with and resuspended in change at the protease-cleavage site, because the mutation was M MES/NaOH (pH 6.0), 10 mM MgCl 50 m , and 25% glycerol. Wild-type accompanied by a significant change in the secondary confor- and mutant MSPs bound to PSII were analyzed by SDS-PAGE. Sam- mation of the protein. Insertion of a methionyl residue between M dithiothreitol for 30 min, ples were denatured with 5% SDS and 60 m 156 157 Phe and Leu abolished functional binding of MSP and then applied to polyacrylamide gels. The acrylamide concentration was 4.5% for the stacking gel and 12.5% for the resolving gel, and both gels unexpectedly created an appreciable binding affinity for non- contained 0.1% SDS and 6 M urea. After electrophoresis, gels were specific sites of the PSII complex. Because the protein confor- stained with Coomassie Brilliant Blue R-250 and scanned at 560 nm mation was little affected by insertion of the residue, the ob- with an ATTO AE-6900 densitometer. Amounts of MSPs bound to PSII served changes in the protein binding were ascribed to a small were estimated by measuring the peak area of the protein with that of structural change(s) in the mutated polypeptide. Based on CP47 as reference. these observations, we proposed that the two protease-sensi- Determination of Oxygen Evolution—Oxygen evolution was deter- 156 163 184 191 mined with a Clark-type oxygen electrode at 40 °C as described (33). tive regions, Phe –Gly and Arg –Ser , are possible White light of a saturating intensity was provided from a halogen lamp candidates for interaction with the PSII complex. (Watanabe Shoko R&D Corp). The reaction medium contained 50 mM In the present study, site-directed mutagenesis was per- MES/NaOH (pH 6.0), 10 mM NaCl, 5 mM MgCl ,10mM CaCl ,1 M 2 2 formed to identify the amino acid residues involved in effective sucrose, and 0.4 mM 2,6-dichloro-p-benzoquinone. binding of Synechococcus MSP to PSII. Conserved, charged Circular Dichroism Spectrometry—Far-UV and near-UV CD spectra were determined with a JASCO J-820 spectrometer. Data were col- residues in the first protease-sensitive sequences were substi- lected every 0.1 nm with a bandwidth of 1 or 2 nm and a 0.5-s time tuted with an uncharged residue, and the ability of the mutant constant. For determination of the far-UV CD spectra, proteins were proteins to bind to urea/NaCl-washed PSII and to restore the dissolved in 20 mM potassium phosphate (pH 6.5) at a protein concen- oxygen evolution was determined. Charge-preserving substitu- tration of 80 g/ml, which was determined by amino acid analysis. tion with an identically charged residue was also performed to Eight repetitive scans were collected and averaged; the secondary struc- ture elements were analyzed with the program VARSLC (34, 35). examine whether the residues are involved in electrostatic Near-UV CD spectra were determined with proteins dissolved in 20 mM interaction. For comparison, conserved, uncharged residues in 156 163 MES/NaOH (pH 6.5). Measurement was repeated 6 –16 times depend- and near the first protease-sensitive sequence Phe –Gly ing on the protein concentrations (see Fig. 4), and the signals obtained and conserved, charged residues in other domains of the pro- were averaged. tein were altered. The results obtained indicate that Arg , 156 158 160 162 RESULTS Phe , Asp , Lys , and Arg located in or near the first protease-sensitive sequence are essential for binding of MSP to Substitution of Charged Amino Acid Residues—There are its functional site. Mutant proteins were also constructed by several lines of evidence indicating that binding of MSP to its Binding of MSP to Photosystem II 14749 TABLE I Design of oligonucleotides used for the site-directed mutagenesis of MSP Mutants Mutagenetic oligonucleotides K59Q 5-CGGCCTCTTGGCGTTGGTTCTTAGGCTCTTCC-3 R73Q 5-CAAGCTCGTGGTTTCTTGGGTCACCAGTTTGG-3 K123Q 5-TCGCCACCAAGTTTTGCACTGTAAACAGCAG-3 V148T 5-GTGCGGTAGGAAGGAGTGTTGAATTCCCCCTTG-3 P149A 5-GTGCGGTAGGAAGCAACGTTGAATTCCCC-3 Y151F 5-TTGGCGGTGCGGAAGGAAGGAACGTTG R152Q 5-TCGAGGAAGTTGGCGGTTTGGTAGGAAGGAACGTTG-3 R152K 5-GAGGAAGTTGGCGGTTTTGTAGGAAGGAACGTTG-3 F156L 5-TCCTTTGGGATCGAGGAGGTTGGCGGTGCGCTA-3 F156Y 5-GTCCTTTGGGATCGAGGTAGTTGGCGGTGCGGTAG-3 156G157 5-GTCCTTTGGGATCGAGACCGAAGTTGGCGGTGCG-3 156A157 5-GTCCTTTGGGATCGAGGGCGAAGTTGGCGGTGCG-3 156V157 3-GTCCTTTGGGATCGAGAACGAAGTTGGCGGTGCG-3 157L 5-GTCCTTTGGGATCGAAGTTGGCGGTGCG-3 D158N 5-CCCCGTCCTTTGGGATTGAGGAAGTTGGCGGTG-3 D158E 5-GGGCGATCGCCGATTCATAACCGGAGGC-3 P159A 5-CCCCGTCCTTTGGCATCGAGGAAGTTGGC-3 K160Q 5-CCAAGCCCCGTCCTTGGGGATCGAGGAAGTTG-3 K160R 5-GGCCAAGCCCCGTCCGCGGGGATCGAGGAAGTTGG-3 G161Q 5-GGCCAAGCCCCGTTGTTTGGGATCGAGGAAG-3 R162Q 5-CCGGAGGCCAAGCCCTGTCCTTTGGGATCGAG-3 R162K 5-ACCGGAGGCCAAGCCTTTTCCTTTGGGATCGAGG-3 G163Q 5-CATAACCGGAGGCCAATTGCCGTCCTTTGGGATC-3 G167Q 5-CGATCGCCGAGTCATATTGGGAGGCCAAGCCCC-3 Y168F 5-CGATCGCCGAGTCAAAACCGGAGGCCAAGC-3 K188Q 5-GTGAGGGAGAAGCGTTGGACATTGGCGCGAGC-3 K234Q 5-AAGACCCCTTGAATTTGTACTTCGTGAGGTTC-3 The nucleotide sequences are of the complementary strand. The site-directed nucleotide changes are underlined. functional site involves electrostatic interactions between the high-affinity binding of the protein to PSII was appreciably charged amino acid residues on the protein and oppositely reduced by the mutation. The results show that all three of the 156 163 charged residues on intrinsic proteins of PSII (6, 7, 36, 37). conserved, charged residues located in the Phe –Gly se- Single-amino acid substitution experiments were, therefore, quence are required for high affinity binding of the protein to performed for conserved, charged residues located in the two PSII, consistent with the notion that the first protease-sensi- protease-sensitive sequences. The first protease-sensitive se- tive sequence is a domain for interaction with PSII. The high 156 163 quence, Phe –Gly , contains two positively charged amino affinity binding, however, was also severely affected by alter- 160 162 152 ation of Arg acid residues, Lys and Arg , and a negatively charged to Gln, which is located three residues upstream residue, Asp , which are completely conserved among 14 of the first protease-sensitive sequence. This indicates that the 156 163 binding domain is larger than the Phe sequences of MSPs from cyanobacteria, algae, and higher –Gly sequence and 148 152 plants (Fig. 1). Mutant proteins were constructed by replacing possibly involves the Val –Arg sequence, which is strictly 160 162 Lys and Arg with Gln (K160Q,R162Q), and by replacing conserved in all the sequences of MSP shown in Fig. 1. Asp with Asn (D158N). Binding affinities of the mutant The effects of the substitution mutations on the ability of MSP to restore the oxygen-evolving activity were also investi- proteins to PSII were determined by measuring the amount of each mutant protein bound to urea/NaCl-washed PSII at dif- gated. As shown in Table II, oxygen evolution was strongly ferent protein/PSII ratios. Reconstituted PSII complexes were suppressed by urea/NaCl washing, but the activity was par- washed once to remove loosely bound MSP. As shown in Fig. 2, tially restored by the addition of wild-type MSP. This reflects binding of wild-type MSP was saturated at a protein/PSII ratio the normal function of the protein because the small magnitude of two. A further increase in the protein/PSII ratio led to only a of reactivation (about 30% of activity in untreated PSII com- small increase in the amount of the protein bound, reflecting plexes) can be ascribed to the following two reasons. First, the the ability of the protein to bind specifically and stoichiomet- full activity of oxygen evolution of cyanobacterial or red algal rically to its functional site on the PSII complex. On the other PSII requires three or four extrinsic proteins (38 – 40); MSP hand, binding of the protein to PSII complexes was dramati- only partially supports the activity. Second, the smaller reac- cally altered by the mutations introduced. The binding curves tivating effect of MSP may be because a larger proportion of of the K160Q and R162Q mutant proteins were sigmoidal, and PSII preparations had been irreversibly damaged during pro- the amounts of the two mutant proteins bound to PSII at low tein extraction. This, however, will in principle not affect our protein/PSII ratios were much lower than that of the wild-type reconstitution results as far as the comparison of the relative 160 162 protein, indicating that alteration of Lys effects of wild-type and mutant MSPs is concerned. The muta- and Arg leads to a large decrease in the binding affinity for the functional site. tions that strongly affected the high affinity binding of the At protein/PSII ratios above four, the amount of mutant pro- protein had a severe impact on the reactivation ability of MSP. teins bound increased linearly with an increase in protein The levels of oxygen evolution restored by the mutant MSPs concentration; this increase is apparently steeper than that of were only less than 20% of that restored by wild-type protein the wild-type protein. This is a feature of nonspecific binding even at a protein/PSII ratio of five. The results confirm that the that has been described first for mutant proteins with Met four charged residues are essential for functional interaction 156 157 157 inserted between Phe and Leu and with Leu replaced with PSII and that the nonspecific binding to PSII is nonfunc- tional. Table II also shows that not all of the strictly conserved, by Met (28). Nonspecific binding was particularly enhanced upon substitution of Asp with Asn so that the amount of the charged residues are required for the function of MSP. Substi- tution of Lys D158N mutant protein bound to PSII far exceeded that of the , the only strictly conserved, charged residue in 184 191 wild-type protein at a protein/PSII ratio of six. Nevertheless, the second protease-sensitive sequence, Arg –Ser with 14750 Binding of MSP to Photosystem II FIG.1. Alignment of amino acid sequences of MSP from 14 oxygenic photosynthetic organisms. Amino acid residues substituted or deleted by site-directed mutagenesis in the present study are indicated by triangles. Dashes indicate amino acid residues identical to those in S. Binding of MSP to Photosystem II 14751 197–250 nm region as well as Fourier-transform infrared spec- trum (41). Because CD data provide more accurate estimates of the proportions of secondary structure elements when the CD spectrum is extended to shorter wavelengths (35), the CD spec- tra of wild-type and mutant proteins were determined from 181 to 260 nm. The far-UV CD spectrum of wild-type protein showed a strong positive band at 197 nm, a broad negative band between 205 and 225 nm, and a zero-line crossover point 152 158 160 at 192 nm (Fig. 3). Substitution of Arg , Asp , Lys , and Arg had no significant effect on the spectrum (Fig. 3 and data not presented). The relative abundance of secondary structure elements estimated from the far-UV CD data is shown in Table III. The wild-type protein contains -sheet severalfold more abundantly than -helix. The values are sim- ilar to those of spinach MSP (42– 44) but different from the previous estimate from the CD spectrum in the 197–250 nm region, which predicted a higher proportion of -helix and a 152 158 160 FIG.2. Effects of substitution of Arg , Asp , Lys , and lower proportion of -sheet than those shown in Table III (41). Arg with an uncharged residue on the binding affinity of MSP Although not shown here, Fourier-transform infrared spectros- to PSII. Urea/NaCl-washed PSII complexes were reconstituted with copy with an improved precision also indicated a larger abun- each mutant protein at the indicated protein/PSII ratios. The amount of wild-type protein bound to PSII at the protein/PSII ratio of three was dance of -sheets than shown previously (41). The mutations taken as 100%. , wild-type MSP; ‚, D158N; Œ, R152Q; E, K160Q; , that significantly decreased the binding and reactivating abil- R162Q. ities of MSP had no significant effect on the secondary confor- mation of the protein; small variations in the proportion of the TABLE II Effects of substitution of a charged amino-acid residue with an secondary structure elements observed are in the range of uncharged residue on the ability of MSP to restore experimental error. Thus, loss of the functional binding to PSII oxygen-evolving activity is not a result of distortion of the secondary conformation of the Urea/NaCl-washed PSII complexes were reconstituted with each mu- protein. tant protein at a protein/PSII ratio of five. The near-UV CD spectra were determined to examine Oxygen Preparations Additions Reactivation whether tertiary structure of the protein is affected by the evolution substitution mutations. The near-UV CD spectrum of spinach mol of % MSP shows two prominent bands, a tryptophan band at 294 nm O /mg of and a tyrosine band at 285 nm. These bands disappeared when chlorophyll/h the tertiary structure that contributes to the local environment PSII complexes None 1153 Washed PSII complexes None 41 0 of the aromatic amino acids was altered by cleavage of the Washed PSII complexes Wild-type MSP 306 100 sulfhydryl bond (45), acidification (43), or heat treatment (44) Washed PSII complexes D158N 89 18 of the protein. Synechococcus MSP lacks tryptophan, and ac- Washed PSII complexes K160Q 71 11 cordingly no band at 294 nm was observed (Fig. 4). Moreover, Washed PSII complexes R162Q 88 18 Washed PSII complexes R152Q 82 16 and unexpectedly, no conspicuous tyrosine band was observed PSII complexes None 1065 at 285 nm. A simple explanation would be that the 285 nm Washed PSII complexes None 78 0 band originates from one (or more) of the three tyrosine resi- Washed PSII complexes Wild-type MSP 360 100 dues that are present in the spinach protein but not in the Washed PSII complexes K59Q 345 94 Synechococcus protein (see Fig. 1). Fig. 4, however, shows four Washed PSII complexes R73Q 391 111 Washed PSII complexes K123Q 368 103 positive bands at 261, 266, 274, and 282 nm, which can be Washed PSII complexes K188Q 329 89 ascribed to the fine-structure bands of phenylalanine and ty- Washed PSII complexes K234Q 352 97 rosine (46). A more likely explanation, therefore, is that a peak at 290 nm corresponds to a tyrosine band that is overlapped with a sharp negative band at shorter wavelengths. In any Gln, had only a marginal effect on the reactivation ability of case, these spectral features remained essentially unaltered MSP. This casts a doubt on the notion that the second protease- upon substitution of any of the four charged residues, indicat- sensitive sequence is a region for interaction with PSII. The 59 73 123 234 ing that inactivation of the mutant MSPs was not associated alteration of Lys , Arg , Lys , and Lys to Gln had no effect on the function of MSP. Thus, inactivation of MSP is with a significant change in the tertiary structure of the pro- tein, either (Fig. 4 and data not presented). specific to the substitution of a charged residue in the Val – Gly sequence. The present study, however, does not exclude If the four charged residues contribute to the effective bind- ing of MSP through electrostatic interaction with oppositely the existence of a binding site for PSII in other regions of the protein because there are still eight strictly conserved, charged charged residues on the PSII complex, substitution of each residues that remain to be investigated (see Fig. 1). residue with an identically charged residue might have little CD spectroscopy was performed to investigate the effect of impact on the function of the protein. This was found to be the the substitution mutations on the secondary and tertiary con- case with Asp . Alteration of the aspartyl residue to a glu- formations of MSP. The secondary conformation of the protein tamyl residue had no effect on the binding of the protein to PSII has previously been estimated from the UV CD spectrum in the (Fig. 5). The D158E mutant protein was able to reconstitute elongatus; dots indicate vacant position in the sequence; stars indicate residues completely conserved in the 14 sequences. The two protease- sensitive sequences are boxed. 1, Synechococcus elongatus; 2, Anacystis nidulans R2; 3, Synechocystis sp. PCC6803; 4, Anabaena sp. PCC7120; 5, Chlamydomonas reinhardtii; 6, Euglena gracilis Z; 7, garden pea; 8, Arabidopsis thaliana; 9, wheat; 10, potato; 11, tomato; 12, tobacco; 13, rice; 14, spinach. 14752 Binding of MSP to Photosystem II FIG.3. Far-UV CD spectra of wild- type and mutant MSPs with Arg replaced by Gln and Asp replaced by Asn. Thick line, wild-type MSP; thin line, R152Q; dotted line, D158N. TABLE III nonspecific sites (not shown). Alteration of Phe to Tyr had no Secondary structural elements of wild-type and mutant MSPs effect on the function of the protein. This indicates that the determined by far-UV CD spectra structure of the side group is crucial for effective interaction Sample -Helix -Sheet Turns Other 161 163 with PSII. Substitution of Gly or Gly with Gln led to a % small decline in the reactivating ability of MSP, whereas alter- Wild type 7 39 21 33 ation of Pro to Ala had no effect. 153 155 164 166 R152Q 8 38 21 33 The Thr –Asn and Leu –Ser sequences, which are K160Q 8 38 21 33 located adjacent to the first protease-sensitive sequence, are R162Q 7 39 21 34 not conserved even in cyanobacterial MSPs. Replacement of a D158N 9 38 21 33 single residue in these two sequences or the entire three-amino acid sequences had no effect on the ability of MSP to restore the oxygen-evolving activity (not shown). Although Arg is re- oxygen evolution activity as effectively as wild-type protein 148 149 quired for the functional binding of MSP, Val , Pro , and (Table IV). Thus, a negative charge at Asp is critically im- 151 148 152 Tyr , located in the conserved Val –Arg sequence, could portant for the functional binding of MSP. In contrast, charge- 152 160 162 be replaced by Thr, Ala, and Phe, respectively, without affect- preserving alteration of Arg to Lys, Lys to Arg, or Arg ing the reactivating ability of MSP. No attempt was made to to Lys resulted in a significant decrease in the high affinity 150 167 168 alter Ser . Gly and Tyr are also completely conserved in binding and an increase in nonspecific binding to PSII (Fig. 5). the 14 sequences of MSP. The reactivating ability of MSP was The amounts of the R152K, K160R, and R162K mutant pro- moderately affected by substitution of Tyr with Phe, teins bound to PSII at low protein/PSII ratios appear to be whereas replacement of Gly by Gln had no influence on the larger than those of the corresponding charge-deleted mutant function of MSP. proteins. The difference should be ascribed to the difference in 156 157 Insertion of a Residue between Phe and Leu and dele- nonspecific binding, because the charge-preserving mutations tion of Leu —We showed previously (28) that MSP becomes affected the ability of the protein to reconstitute oxygen evolu- unable to bind to its functional site and reactivate oxygen tion as severely as the corresponding charge-deleting muta- evolution when a methionyl residue is inserted between Phe tions. Only the K160R mutant protein restored oxygen evolu- and Leu . The mutant protein, however, showed a strong tion somewhat more effectively than the K160Q mutant affinity for nonspecific binding. Extension of this experiment protein, suggesting participation of this residue in electrostatic interaction (Table IV). provided important information on both effective and nonspe- cific interaction of MSP with PSII. Because the protein confor- Substitution of Uncharged Residues—Five uncharged resi- 156 157 159 161 163 dues, Phe mation was little affected by the insertion of Met, loss of the , Leu , Pro , Gly , and Gly , which are 156 163 functional binding of the protein is ascribed to a small struc- located in the Phe –Gly sequence, are strictly conserved tural change in the protease-sensitive sequence. If this inter- among MSPs from cyanobacteria to higher plants, except for Gly , which is conservatively replaced by Ala in two cya- pretation was correct, the inactivation of MSP can be ascribed to an effect of either the side group of the methionyl residue nobacteria (see Fig. 1). Substitution of Phe with Leu resulted in a nearly total loss of the reactivating ability of MSP (Table inserted or a small increase in the polypeptide length between 156 156 157 V). Thus, Phe Phe and Leu . To examine the effect of the side group, is required for effective binding of the protein, although the effects of the mutation on binding of the protein to mutant MSPs were constructed by inserting an amino acid its functional site could not be determined accurately because residue that has a side group smaller than that of Met between 156 157 of an extremely strong affinity the mutant protein exhibited for Phe and Leu . The three mutant MSPs, which were con- Binding of MSP to Photosystem II 14753 FIG.4. Near-UV CD spectra of wild- type and mutant MSPs with Arg or Arg replaced by Gln. Solid line, wild- type MSP (985 g/ml); dashed line, R152Q (827 g/ml); dotted line, R162Q (1760 g/ml). TABLE IV Effects of substitution of a charged amino acid residue with an identically charged residue on the ability of MSP to restore oxygen-evolving activity Experimental conditions are the same as in Table II. Oxygen Preparations Additions Reactivation evolution mol of % O /mg of chlorophyll/h PSII complexes None 1147 Washed PSII complexes None 96 0 Washed PSII complexes Wild-typeMSP 353 100 Washed PSII complexes D158E 382 110 PSII complexes None 1090 Washed PSII complexes None 40 0 Washed PSII complexes Wild-typeMSP 306 100 Washed PSII complexes K160R 136 36 Washed PSII complexes R162K 79 15 Washed PSII complexes R152K 108 26 TABLE V Effects of substitution of an uncharged amino acid residue on the ability of MSP to restore oxygen-evolving activity 152 158 160 FIG.5. Effects of substitution of Arg , Asp , Lys , and Urea/NaCl-washed PSII complexes were reconstituted with each mu- Arg with an identically charged residue on the binding affin- tant protein at the protein/PSII ratios of five except that the ratio of four ity of MSP to PSII. Experimental conditions were the same as in Fig. was used for of G163Q and G167Q. 2. , wild-type MSP; ‚, D158E; Œ, R152K; E, K160R; , R162K. Additions Reactivation structed by inserting Gly, Ala, and Val, are called 156G157, None 0 156A157 and 156V157, respectively. Leu is not es- Wild-type MSP 100 sential for the functional binding of MSP to PSII because the V148T 121 residue could be substituted with Met without affecting the P149A 113 reactivating ability of the protein (28). In the fourth mutant Y151F 94 protein (157L), therefore, Leu was deleted to examine F156Y 98 F156L 4 whether the protein binding is affected by a change in the P159A 104 polypeptide length. As shown in Table VI, the protein became G161Q 73 totally ineffective in reactivation of oxygen evolution when Val, G163Q 71 Ala, or even Gly, which has no side chain, was inserted between G167Q 110 156 157 Y168F 79 Phe and Leu , indicating that inactivation of the protein is independent of the size and structure of side chain of the residue inserted. Deletion of Leu also led to a large decline in the reactivating ability of the protein. None of the four mutant protein to PSII (data not shown). These results indicate that proteins bound to PSII to an appreciable amount at the protein/ binding of MSP to its functional site is extremely sensitive to a PSII ratio of five, indicating that the mutations abolished high small change in the polypeptide length in the region between 156 157 affinity binding without enhancing nonspecific binding of the Phe and Leu . The observation that nonspecific binding 14754 Binding of MSP to Photosystem II TABLE VI 156 157 Effects of insertion of Gly, Ala, or Val between Phe and Leu and deletion of Leu on the ability of MSP to restore the oxygen-evolving activity Experimental conditions are the same as in Table II. Oxygen Preparations Additions Reactivation evolution mol of % O /mg of FIG.6. Mutations that affected the function of MSP. The large chlorophyll/h and small upward arrows indicate mutations that affected the ability of PSII complexes None 889 Synechococcus MSP to reactivate oxygen evolution strongly and weakly, Urea/NaCl-washed PSII None 81 0 respectively. Downward arrows indicate mutations that had no effect complexes on the reactivating ability of the protein. The first protease-sensitive Urea/NaCl-washed PSII Wild-type MSP 337 100 sequence is boxed. Asterisks indicate residues completely conserved in complexes the 14 sequences of MSP. Urea/NaCl-washed PSII 156G157 87 2 complexes Urea/NaCl-washed PSII 156A157 85 2 and cyanobacteria or in the flexibility of proteins between me- complexes sophiles and thermophiles. Urea/NaCl-washed PSII 156V157 89 3 66 76 130 159 Six conserved lysyl residues, Lys , Lys , Lys , Lys , complexes 186 236 Lys , and Lys , of spinach MSP were suggested to be lo- Urea/NaCl-washed PSII 157L 94 5 complexes cated in regions of the protein in contact with PSII because they were modified with amino group-specific reagents when MSP was free in solution but not when the protein was bound appeared upon insertion of Met but not upon insertion of Gly, to PSII membranes (19, 20). The present study shows that not Ala, or Val implies that the structural factor that confers an all of the conserved lysyl residues are required for functional affinity for nonfunctional sites on the protein is the side group 59 123 188 binding to PSII. Substitution of Lys , Lys , Lys , and of Met inserted. This is consistent with the previous observa- 234 66 Lys of Synechococcus MSP (which correspond to Lys , tion that nonfunctional binding became appreciable upon re- 130 190 236 Lys , Lys , and Lys of the spinach protein, respectively) placement of Leu by Met (28). with Gln had no influence on the function of the protein. Al- 160 159 DISCUSSION teration of Lys (which corresponds to Lys of the spinach We have constructed more than 30 variants of Synechococcus protein) to Gln had, however, a severe impact on the ability of MSP by substituting, inserting, or deleting an amino acid res- the protein to effectively bind to PSII. Loss of protein binding is idue to investigate whether the protease-sensitive Phe not related to a change in the secondary or tertiary structure of 163 184 191 Gly and Arg –Ser sequences are regions for functional the protein. The effect of the charge-preserving substitution of Lys with Arg on the reactivating ability of the protein was binding to PSII. The cyanobacterial MSP served as an excellent tool for investigation of the protein binding by means of site- somewhat less than that of the charge-deleting substitution of directed mutagenesis and in vitro reconstitution, allowing us to the residue with Gln. This can be explained by assuming that identify several charged and uncharged amino acid residues Lys participates in the electrostatic interaction essential for that are essential for binding of MSP to its functional site. The the function of the protein and that Arg acts only imperfectly mutations that effectively altered the ability of MSP to reacti- for Lys due to the difference in the side group. The corre- vate oxygen evolution are illustrated in Fig. 6. sponding lysyl residue (Lys ) on spinach MSP was unable to The first protease-sensitive sequence contains a highly con- participate in an intramolecular charge-pair interaction be- served acidic residue, Asp cause the residue was accessible to amino group modifiers . Substitution of this residue with Asn resulted in a large decrease in the ability of MSP to bind to when the protein was free in solution (20). It is highly likely, functional site and to restore the oxygen-evolving activity, therefore, that Lys contributes to binding of MSP to its whereas its charge-preserving replacement by Glu had no in- functional site by electrostatically interacting with a negatively hibitory effect, indicating that the negative charge at Asp charged residue on an intrinsic protein of the PSII complex. is required for the protein binding. Thus, Asp is involved either The contribution of arginyl residues to binding of spinach MSP has been suggested by chemical modification of the gua- in charge-pair interaction with an oppositely charged residue on an intrinsic protein of PSII or in intramolecular charge-pair nidino group of arginyl residues with 2,3-butanedione (20). interaction necessary for maintenance of a functional structure None of the six arginyl residues present in MSP reacted with of MSP. Far- and near-UV CD spectroscopy demonstrated that the reagent when the protein was bound to PSII membranes, inactivation of MSP by mutation of Asp whereas treatment of the protein in solution led to modification to Asn was not accompanied by a significant change in the secondary and of four arginyl residues, which was accompanied by loss of the tertiary conformation of the protein, respectively. Substitution ability of the protein to bind to PSII and to reactivate oxygen of the corresponding aspartyl residue (Asp ) in spinach MSP evolution. Modified residues have not been identified yet be- cause of the instability of the modified products (20). Site- was also shown to have no influence on intramolecular salt bridges formed upon cross-linking with EDC (17). In addition, directed mutagenesis allowed us to successfully identify argi- 157 168 carboxyl groups in the domain Asp nyl residues that are essential for the protein binding and also –Asp of the spinach protein were shown to participate in the binding of spinach provided important information as to a domain of the protein MSP to PSII (24). These observations favor the involvement of for effective binding to PSII. There are three arginyl residues, 158 73 152 162 Asp in an intermolecular charge-pair interaction between Arg , Arg , and Arg , which are completely conserved among MSPs from cyanobacteria, algae, and higher plants (Fig. MSP and PSII. Substitution of the corresponding aspartyl res- idue on MSPs from other photosynthetic organisms had weaker 1). Arg is not essential because the residue could be substi- effects, a 5–15% decline in the reactivating ability of spinach tuted with Gln without affecting the function of the protein. 152 162 MSP (27) and a 35– 40% reduction in the rate of oxygen evolu- Mutations at either Arg or Arg affected the function of tion in Synechocystis cells (25). These differences may be re- MSP significantly. Because these mutations did not induce any lated to differences in protein binding between higher plants significant changes in the protein conformation, the results Binding of MSP to Photosystem II 14755 indicate that the two arginyl residues directly participate in arranged spatially so as to become in van der Waals contact effective interaction with PSII. No direct evidence was obtained with respective partner residues on the PSII complex. As such, 152 162 indicating that Arg and Arg are involved in charge-pair the protein binding will be weakened or abolished by insertion interaction; the reactivating ability of MSP was nearly equally or deletion of a residue, which alters the location of the charged 148 163 residues in the Val –Gly domain. suppressed by charge-deleting and charge-preserving substitu- 148 163 tions of each residue. A possibility remains, however, that the The Val –Gly domain becomes a region for nonfunc- two residues participate in charge-pair interaction but that a tional binding to PSII upon modification. Replacement of a single essential residue in the domain conferred on the protein positive charge each residue carries becomes incapable of in- teracting with a negative charge on the partner residue when the ability to nonspecifically bind to the PSII complex. A com- Arg is altered to Lys. Because Arg is located three residues parison of the insertion mutants constructed in the present and apart from the first protease-sensitive sequence, the result also previous experiments (28) showed that the structure of the side shows that the binding domain of MSP is larger than suggested chain of an inserted residue is critical for the nonspecific bind- ing of the protein. It is concluded, therefore, that the conserved by the proteolytic experiments (28) and may involve the well 148 163 148 152 conserved Val –Arg sequence. amino acid residues in the Val –Gly domain are also im- Substitution of uncharged residue in the first protease-sen- portant for the protein to minimize nonspecific interactions with PSII. Several mutant proteins exhibited an extremely sitive sequence also affected the function of MSP. Substitution of Phe with Leu resulted in an almost complete loss of the strong affinity for nonspecific binding. This implies caution in using binding of MSP to PSII as an index of the functional ability of the protein to restore the oxygen-evolving activity, integrity of the protein when reconstitution is performed at a whereas replacement of the same residue by Tyr had no effect on the function of MSP. Thus, the aromatic side group of Phe low protein/PSII protein ratio. In contrast to the first protease-sensitive sequence where all is required for the functional binding of the protein. The reac- the eight residues are completely conserved except for Gly , tivating ability of MSP was moderately affected by substitution 161 163 which is conservatively replaced, the second protease-sensitive of either Gly or Gly with Gln. This suggests that the side 184 191 Arg –Ser sequence has only two strictly conserved resi- chain of Gln, which replaced Gly sterically, interferes with the 186 188 dues, Asn and Lys . No evidence was obtained to support protein-protein interaction. Alternatively, because Gly has no the proposition that a binding site for PSII is located in the side chain, it may confer a greater conformational flexibility, 184 191 188 Arg -Ser sequence. Substitution of Lys , the sole strictly possibly required in the first protease-sensitive region for the effective association of the protein with its functional site. The conserved charged residue in the sequence, with Gln had only a marginal effect on the reactivating ability of MSP. This is protein binding, however, was not appreciably affected by sub- consistent with the results with spinach MSP, where the cor- stitution of Pro with Ala; this may give another constraint responding lysyl residue (Lys ) was accessible to the water- on the polypeptide conformation. soluble reagents even when the protein was bound to PSII Conserved, uncharged residues located near the Phe – 163 152 membranes (19, 20). There is another charged residue, Arg , Gly sequence were also examined. As stated above, Arg is in the second protease-sensitive sequence, which is conserved located in the completely conserved five-amino acid sequence. in two other cyanobacteria and conservatively replaced by Although substitution of three uncharged residues, Val , 149 151 Lys in other photosynthetic organisms. Because the lysyl resi- Pro , and Tyr , in the sequence failed to affect the function due (Lys ) of spinach MSP was modified with specific re- of MSP, the result should be interpreted with caution because 148 151 agents only when the protein was free in solution (19, 20), a Val and Tyr were replaced by residues with relatively possibility remains that Arg participates in the functional small differences in the structure of the side chain. In partic- binding of Synechococcus MSP. ular, because Tyr was substituted with Phe, a possibility remains that an aromatic group at position 151 plays a role in Acknowledgments—We are grateful to Dr. T. Takakuwa, Nihon the functional binding of the protein, which will be addressed Bunko Co., for measurement and analysis of CD spectra. We thank Dr. in future studies. Furthermore, amino acid residues in the Jian-Ren Shen of the Institute of Physical and Chemical Research 167 169 (Riken) for reading the manuscript. Gly –Asp sequence, which is located three residues after the first protease-sensitive sequence, are either strictly con- REFERENCES served or conservatively replaced. Possible contribution of this 1. Seidler, A. (1996) Biochim. Biophys. Acta 1277, 35– 60 sequence to the protein binding cannot be excluded because 2. Bricker, T. M., and Frankel, L. K. (1998) Photosynth. Res. 56, 157–173 substitution of Tyr with Phe weakly affected the function of 3. Murata, N., and Miyao, M. (1985) Trends Biochem. Sci. 10, 122–124 4. Shen, J.-R., Ikeuchi, M., and Inoue, Y. (1992) FEBS Lett. 301, 145–149 MSP, and single amino acid substitution has not been per- 5. Enami, I., Murayama, H., Ohta, H., Kamo, M., Nakazato, K., and Shen, J.-R. formed yet for Asp . (1995) Biochim. Biophys. Acta 1232, 208 –216 Extension of the previous experiments with the mutant pro- 6. Ono, T., and Inoue, Y. (1984) FEBS Lett. 166, 381–384 156 157 7. Miyao, M., and Murata, N. (1984) FEBS Lett. 170, 350 –354 tein having Met inserted (28) between Phe and Leu 8. Kuwabara, T., Miyao, M., Murata, T., and Murata, N. (1985) Biochim. Biophys. showed that binding of MSP to its functional site is extremely Acta 806, 283–289 9. Ono, T., and Inoue, Y. (1986) Biochim. Biophys. Acta 850, 380 –389 sensitive to a small change in the polypeptide length caused by 10. Enami, I., Satoh, K., and Katoh, S. (1987) FEBS Lett. 226, 161–165 156 157 insertion of an amino acid residue between Phe and Leu 11. Bricker, T. M., Odam, W. R., and Queirolo, C. B. (1988) FEBS Lett. 231, or deletion of Leu . This is consistent with the above de- 111–117 12. Enami, I., Kaneko, M., Kitamura, N., Koike, H., Sonoike, K., Inoue, Y., and scribed results that residues in or near the first protease- Katoh, S. (1991) Biochim. Biophys. Acta 1060, 224 –232 sensitive region participate in the interaction of MSP with PSII 13. Odom, W. R., and Bricker, T. M. (1992) Biochemistry 31, 5616 –5620 14. Mei, R., Green, J. P., Sayer, R. T., and Frash, W. D. (1989) Biochemistry 28, and further suggests that the relative location of these residues 5560 –5567 is critical for effective binding to PSII. We suggest that binding 15. Enami, I., Ohta, S., Mitsuishi, S., Takahashi, S., Ikeuchi, M., and Katoh, S. of MSP to its functional site involves multiple charged residues (1992) Plant Cell Physiol. 33, 291–297 148 163 16. Eaton-Rye, J. J., and Murata, N. (1989) Biochim. Biophys. Acta 977, 219 –226 in the Val –Gly region, which interact electrostatically 17. Seidler, A., Rutherford, A. W and Michel, H. (1996) Plant Mol. Biol. 31, with oppositely charged residues on intrinsic proteins of the 183–188 18. Qian, M., Al-Khaldi, S. F., Putnam-Evans, C., Bricker, T. M., and Burnap, PSII complex, although contribution of other interactions can- R. L. (1997) Biochemistry 36, 15244 –15252 not be excluded. Such a protein-complex interaction is possible 19. Frankel, L. K., and Bricker, T. M. (1995) Biochemistry 34, 7492–7497 only when the charged residues on the surface of MSP are 20. Miura, T., Shen, J.-R., Takahashi, S., Kamo, M., Nakamura, E., Ohta, H., 14756 Binding of MSP to Photosystem II Kamei, A., Inoue, Y., Domae, N., Takio, K., Nakazato, K., Inoue, Y., and 33. Ichimura, T., Miyairi, S., Satoh, K., and Katoh, S. (1992) Plat Cell Physiol. 33, Enami, I. (1997) J. Biol. Chem. 272, 3788 –3798 299 –305 21. Queirola, C. B. (1992) Doctoral dissertation, Louisiana State University, Baton 34. Manavalan, P., and Johnson, W. C., Jr. (1987) Anal. Biochem. 167, 76 – 85 Rouge 35. Johnson, W. C., Jr. (1990) Proteins Struct. Funct. Genet. 7, 205–214 22. Isogai, Y., Nishimura, M., and Itoh, S. (1987) Plant Cell Physiol. 28, 1493–1499 36. Enami, I., Tomo, T., Kitamura, M., and Katoh, S. (1994) Biochim. Biophys. 23. Seidler, A. (1996) Eur. J. Biochem. 242, 485– 490 Acta 1185, 75– 80 24. Frankel, L. K., Cruz, J. A., and Bricker, T. M. (1999) Biochemistry 38, 37. Enami, I., Kitamura, M., Tomo, T., Isokawa, Y., Ohta, H., and Katoh, S. (1994) 14271–14278 Biochim. Biophys. Acta 1186, 52–58 25. Burnap, R. L., Qian, M., Shen, J.-R., Inoue, Y., and Sherman, L. A. (1994) 38. Shen, J.-R., and Inoe, Y. (1993) Biochemistry 32, 1825–1832 Biochemistry 33, 13712–13718 39. Enami, I., Kikuchi, S., Fukuda, T., Ohta, H., and Shen, J.-R. (1998) Biochem- 26. Seidler, A., Ro ¨ ll, K., and Michel, H. (1992) in Current Research in Photosyn- istry 37, 2787–2793 thesis (Murata, N., ed) Vol. II, pp. 409 – 412, Kluwer Academic Publishers, 40. Enami, I., Yoshihara, S., Tori, A, Okamura, A., Ohta, H., and Shen, J.-R. Dordrecht (2000) Plant cell Physiol. 41, 1354 –1364 27. Seidler, A., Michel, H., and Rutherford, A. W. (1995) in Photosynthesis: From 41. Sonoyama, M., Motoki, M., Okamoto, G., Hirano, M., Ishida, H., and Katoh, S. Light to Biosphere (Mathis, P., ed) Vol. II, pp. 259 –262, Kluwer Academic (1996) Biochim. Biophys. Acta 1297, 167–170 Publishers, Dordrecht 42. Xu, Q., Nelson, J., and Bricker, T. M. (1994) Biochim. Biophys. Acta 1188, 28. Motoki, A., Shimazu, T., Hirano, M., and Katoh, S. (1998) Biochim. Biophys. 427– 431 Acta 1365, 492–502 43. Shutova, T., Irrgang, K.-D., Shubin, V., Klimov, V., and Renger, G. (1997) 29. Motoki, A., Shimazu, T., Hirano, M., and Katoh, S. (1995) in Photosynthesis: Biochemistry 36, 6350 – 6358 from Light to Biosphere (Mathis, P., ed) Vol. II, pp. 487– 490, Kluwer 44. Lydakis-Simantris, N., Hutchinson, R. S., Betts H. S., Barry, B. A., and Academic Publishers, Dordrecht Yocum, C. F. (1999) Biochemistry 38, 404 – 414 30. Taylor, J. W., Ott, J., and Eckstein, F. (1985) Nucleic Acids Res. 13, 8765– 8785 45. Tanaka, S., and Wada, K. (1988) Photosynth. Res. 17, 255–266 31. Nakamae, K. L., and Eckstein, F. (1986) Nucleic Acids Res. 14, 9679 –9698 46. Strickland E. H. (1976) in Handbook of Biochemistry and Molecular Biology 32. Sayers, J. R., Krekel, C., and Eckstein, F. (1992) BioTechniques 13, 592–596 (Fasman, G. D., ed) Vol. III, pp. 141–147, CRC Press, Cleveland, OH

Journal

Journal of Biological Chemistry – Unpaywall

Published: Apr 1, 2002

There are no references for this article.