Photosynthesis Research65: 207–217, 2000.© 2000Kluwer Academic Publishers. Printed in the Netherlands.

207

Regular paper

Identification of Photosystem I components from a glaucocystophyte,Cyanophora paradoxa: The PsaD protein has an N-terminal stretchhomologous to higher plants

Hiroyuki Koike∗, Mari Shibata, Kanako Yasutomi, Yasuhiro Kashino & Kazuhiko SatohDepartment of Life Science, Faculty of Science, Himeji Institute of Technology, Harima Science Garden City,Hyogo 678-1297, Japan;∗Author for correspondence (e-mail: hkoike@sci.himeji-tech.ac.jp; fax: +81-791-58-0185)

Received 18 January 2000; accepted in revised form 7 September 2000

Key words: Cyanophora paradoxa, glaucocystophyte, light harvesting complex, Photosystem I, PsaD, PsaL

Abstract

Thylakoid membranes and Photosystem I (PS I) complexes were isolated from a glaucocystophyte,Cyanophoraparadoxa, which is thought to have the most primitive ‘plastids’, and the proteins related to PS I were ex-amined. The intrinsic light-harvesting chlorophyll protein complexes of PS I (LHC I) were not detected by animmunological method. The PS I complexes consisted of at least eight low-molecular-mass proteins in additionto PS I reaction center proteins. The N-terminal sequence of the PsaD protein has higher homology to that ofChlamydomonas reinhardtiiand land plants, than to that of other algae or cyanobacteria. On the other hand, thePsaL sequence has the highest homology to those of cyanobacteria. Taking into account the other sequences ofPS I components whose genes are encoded in the cyanelle genome, and the fact that LHC I is not detected, it isconcluded that PS I ofC. paradoxahas chimeric characteristics of both ‘green’ lineages and cyanobacteria.

Abbreviations: CAB – chlorophyll a/b light-harvesting complex; Chl – chlorophyll; DCPIP – 2,6-dichlorophenolindophenol; LHC – light-harvesting chlorophyll protein complex; PS – photosystem;SDS–PAGE – SDS–polyacrylamide gel electrophoresis

Introduction

Cyanophora paradoxais a unique oxygenic photo-synthetic eukaryote. This organism contains a specialtype of plastid, cyanelles, which perform oxygenicphotosynthesis. Cyanelles possess phycobilisomes onthe thylakoid membranes and carboxysome-like struc-ture in the stroma. In addition, they have a thinpeptidoglycan wall between outer and inner envelopemembranes (Löffelhardt and Bohnert 1994a, b; Löf-felhardt et al. 1997). In this sense cyanelles resemblecyanobacteria.

Red algae also utilize phycobiliproteins as ac-cessory pigment–protein complexes. However, theirchloroplasts are not surrounded by a peptidoglycanwall nor are carboxysome-like structure present. In

this respect, the chloroplasts of red algae are dis-tinct from those ofC. paradoxa. It has also beenfound that all chloroplasts examined so far havemembrane-intrinsic light-harvesting chlorophyll pro-tein complexes (LHCs) that are immunologically re-lated to higher plant LHCs with a possible exceptionof cryptophytes. The sequence analysis of LHC pro-teins suggests that they belong to a superfamily whichincludes the Chla/bproteins, the Chla/c proteins andthe Chla proteins (Green and Durnford 1996; Durn-ford et al. 1999). However, whether the cyanelles havemembrane-instrinsic LHC(s) has not been investigateduntil now.

The complete DNA sequence of cyanelle genomehas been determined (Stirewalt et al. 1995). The size ofthe DNA is 136 Kbp, which is not much different from

208

those of chloroplast genome of other algae (Hallicket al. 1993; Reith and Munholland 1995; Kowallik etal. 1995; Douglas and Penny 1999) or higher plants(Ohyama 1986; Shinozaki et al. 1986; Hiratsuka et al.1989; Wakasugi et al. 1994,1997; Maier et al. 1995).The cyanelle genome has a set of inverted repeat thatcontains ribosomal RNA genes. Many, but not all,components of PS I and II, cytochromeb6/f complex,and phycobilisomes are encoded in the cyanelle gen-ome. But most enzymes of the Calvin-Benson cycleand other components are not found in the cyanellegenome and are assumed to have been transferred tothe nuclear genome. Martin et al. (1998) performedphylogenetic analysis of concatenated 11 039 aminoacid sequences from 45 proteins encoded in plastidgenomes from nine completely sequenced species,including C. paradoxa, and concluded that chloro-plasts ofC. paradoxabranched first from a commonchloroplast ancestor.

In contrast to many achievements by using mo-lecular biological techniques, biochemical research ofcyanelles is less developed. Because many cyanellegenes have been transferred to the nuclear genome,biochemical studies are needed to identify the com-ponents of PS I and PS II complexes. Burnap andTrench (1989) have isolated PS I and II particles fromthe thylakoid membranes and analyzed their proteincompositions by SDS–PAGE. Significant activities ofelectron transport from DCPIPH2 to methylviologenwas found in PS I particles that consisted of four low-molecular mass subunits of 21, 18, 14 and 11 kDa.The 21 and 18 kDa proteins were identified as PsaDprotein and its degradation product, respectively, byimmunoblotting analysis.

Here, we report the identification of PS I compon-ents of PS I-enriched particles by means of N-terminalamino acid sequencing and immunochemical analysis.Our results show that LHC proteins were not im-munologically detectable in the thylakoid membranesof the cyanelle. We further show by the N-terminalsequencing that the PsaD protein whose gene is en-coded in nuclear genome, has higher homology tothat of Chlamydomonas reinhardtiiand land plantsthan toPorphyra purpurea, Odontella sinensisor cy-anobacteria. However, N-terminal sequences of othercomponents are similar to cyanobacterial counter-parts. Thus, we suggest that the PS I of cyanellesshows chimeric characters of both higher plants andcyanobacteria.

Materials and methods

Growth of cells and preparation of PS I-enrichedparticles

C. paradoxaStrain NIES 547 was obtained fromGlobal Environmental Forum, Tsukuba, Japan. Thecells were grown with aeration in the C medium(Ichimura 1971) at 25◦C for 10–14 days illuminatedby an incandescent lamp. Cells were harvested bycentrifugation at 2000× g for 5 min and washed onceby HEMS (50 mM HEPES–NaOH (pH 7.5), 2 mMEGTA, 1 mM MgCl2 and 0.5 M sucrose) buffer andwere suspended in the same buffer. Harvested cellswere disrupted osmotically by dilution of the suspen-sion with HEM buffer (50 mM HEPES–NaOH (pH7.5), 2 mM EGTA and 1 mM MgCl2), and cell frag-ments and cyanelles were collected by centrifugationat 2500× g for 5 min at 4 ◦C. The pellet was sus-pended in the HEMS buffer, and the suspension wasplaced on the same medium containing 40% Percolland centrifuged at 10 000× g for 10 min. The intactcyanelles recovered in the precipitate were suspendedin the HEMS buffer and treated with 0.1% (w/v) lyso-zyme for 15 min at 25◦C. The treated cyanelles werewashed once and were disrupted by passing through aFrench pressure cell twice at 60 MPa. The homogen-ate was centrifuged at 10 000× g for 10 min at 4◦Cto remove unbroken cyanelles. The supernatant wascentrifuged at 300 000× g for 2 h at 4◦C to collect thethylakoid membranes. The precipitated membraneswere suspended in the HEMS buffer and kept frozenat –80◦C until use.

The thylakoid membranes were diluted to 0.5 mgChl/ml and were treated with 1.0% n-heptylthiogluco-side (HTG) or 0.8% dodecyl-β-D-malotoside (DM)for 30 min at 0◦C. The suspension was placed onthe discontinuous sucrose density gradient (5–30%(w/v)) containing the HEM buffer, and centrifuged at300 000× g for 90 min. The PS I-enriched fractionwas obtained as a band at 20% sucrose. The fractionwas diluted with the HEM buffer, and centrifuged at300 000× g for 2 h at 4◦C to collect the PS I-enrichedparticles. There was no difference in the protein com-ponents of PS I-enriched particles solubilized withHTG or DM.

Preparation of thylakoid membranes andPS I-enriched particles of other organisms

Spinach thylakoid membranes were prepared as de-scribed previously (Koike et al. 1998).Cyanidium

209

caldariumthylakoid membranes were kindly providedby Dr Enami, Science University of Tokyo.C. rein-hardtii cell wall-less mutant was disrupted by glasshomogenizer and the thylakoid membrnes were col-lected by centrifugation (7000× g for 10 min). Cellsof Synechocystissp. PCC6803 was disrupted by agit-ation with glass beads and the thylakoid membraneswere collected by 100 000× g 1 h centrifugation.

Thylakoid membranes ofS. 6803 and spinach weresolubilized with 1.0% DM and PS I-enriched particleswere prepared by sucrose density gradient centrifuga-tion as described above.

SDS–PAGE and immunostaining

Samples for SDS–PAGE were delipidated with a mix-ture of methanol and diethylether (1:9, v/v) as de-scribed by Ikeuchi et al. (1992). Samples were thensolubilized with 5% lithium dodecylsulfate, 60 mMdithiothreitol and 60 mM Tris–HCl (pH 6.8). Highresolution SDS–PAGE was performed according toIkeuchi and Inoue (1988) with some modifications.

For immunostaining, the proteins in the gelwere electroblotted to a nitrocellulose membrane in100 mM Tris, 192 mM glycine, 5% methanol and0.02% SDS for 30 min at 2A at room temperature. Theamounts of proteins transferred to the membrane werechecked by staining with amido black 10B. The mem-brane was immunodecorated using antiserum raisedagainst LHC I ofPorphyridium cruentum(Tan et al.1995) orC. reinhardtii(Bass et al. 1992).

Protein sequencing and homology search

For protein sequencing, polypeptides were electroblot-ted to a polyvinylidene difluoride membrane (ProBlot,Applied Biosystems, USA) as described previously(Koike et al. 1989). The membranes were treatedwith 0.6 N HCl overnight (Ikeuchi and Inoue 1988).N-terminal sequences of the PS I components weredetermined by a 473A protein sequencer (AppliedBiosystems, USA) as described previously (Koike etal. 1989).

Homology search was performed using TFASTAprogram equipped with GCG software packageagainst DDBJ/EMBL/GenBank data bases installedon a DEC 3000 workstation.

Other measurements

P700 contents were determined by light-induced ab-sorbance changes at 703 nm by a Hitachi 356

spectrophotometer. Contents of cytochormeb559were determined by dithionite or hydorquinone-reducedminus ferricyanide-oxidized difference ab-sorption spectra using Shimadzu MPS-2000 spectro-photometer. Oxygen evolution or consumption acivit-ies were measured with a Clark-type oxygen electrode(Rank Brothers, UK).

Size-separation of PS I-enriched particles wasperformed by Superdex-200 column chromatography(SMART System, Pharmacia Biotech) as described byShen (1999).

Results and discussion

Immunological detection of LHC I polypeptides

C. paradoxautilizes phycobilisomes as light-harvestingpigment–protein complexes. Cyanobacteria and redalgae also posses phycobilisomes as antennae for PSII. In addition to them, red algae have another typeof membrane-intrinsic Chla-binding light-harvesingcomplexes associated with PS I (Wolfe et al. 1994).Amino acid sequence analyses deduced from cDNAclones ofP. cruentumrevealed that they belong toCAB superfamily (Tan et al. 1997). It is of interestto investigate whetherC. paradoxacarries this typeof complexes or not because all the chloroplasts sofar examined have LHC proteins belonging to CABsuperfamily (Green and Durnford 1996).

Figure 1A shows immunostaining analysis withantiserum raised against LHC I ofP. cruentum(Tanet al. 1995) in thylakoid membranes of spinach,C.caldarium (a red alga) andC. paradoxa. Two poly-peptides with apparent molocular masses of 28 and21 kDa were strongly immunodecorated in spinachmembranes (Figure 1A, lane d). Some additionalcross-reacting polypeptides were also present. InC.caldarium thylakoids, a polypeptide with an appar-ent molecular mass of 18 kDa also cross reacted(Figure 1A, lane e). Contrary to spinach andC. cal-darium thylakoids, there was no detectable bandscross reacted inC. paradoxathylakoid membranes(Figure 1A, lane f).

In order to further check the presence of LHC I inC. paradoxa, immunostaining analysis with antiserumraised against LHC I proteins ofC. reinhardtii wasperformed because anti-LHC I ofP. cruentummighthve not recognized the LHC I ofC. paradoxa. An-tiserum raised against p17.2 ofC. reinhardtii (Basset al. 1992) strongly immunodecorated the 29 and 24

210

Figure 1. Immunoblot analysis of thylakoid membrane components with an antiserum toP. cruentumLHC I (A) or to p17.2 ofC. reinhardtii(B). (A) SDS–PAGE profile and immunoblot analysis of spinach (a and d),C. caldarium(b and e), andC. paradoxa(b and e), respectively. (B)SDS–PAGE profile and immunoblot analysis of spinach (a and d),C. reinhardtii (b and e), andC. paradoxa(c and f), respectively.

kDa proteins ofC. reinhardtiimembranes (Figure 1B,lane e). The antiserum also cross reacted strongly withspinach membranes (Figure 1B, lane d). However,no cross reacting bands were observed forC. para-doxamembranes (Figure 1B, lane f). When anotherantiserum raised against p18.1 ofC. reinhardtii waschallenged toC. paradoxamembranes, cross reactingbands were not detected as well (data not shown). Itis thus indicated thatC. paradoxamembranes did notcross react with antisera raised against LHC I fromeitherP. cruentum, or C. reinhardtii.

There are two possibilities to account for the non-reactivity: (1) C. paradoxacarries LHC I proteinswhich bind only Chla like red algae, but the anti-bodies tested in the present study did not detect theLHC I, or (2) it is, in principle, the same as cyanobac-terial thylakoid membranes and does not have LHCI. Although there still remains a possibility thatC.paradoxahas LHC I which is very much different

from those ofP. cruemtumor C. reinhardtii, we preferthe possibility (2) by the following two reasons. (1)We did not identify PS I components other than thoseconstituting PS I core even after milder detergent treat-ment of thylakoid membranes (data not shown). (2)Chl to P700 and cytochromeb559ratios in thylakoidmembranes were 177 and 84, respectively. If we takeit that the two copies of cytochromeb559are presentin the PS II reaction center (De Paula et al. 1985), Chl– PS II ratio is calculated to be 168. Judging from theChl – P700 ratio of 110–120 in PS I-enriched particlesand the reported number of 50 Chl molecules whichbelong to the PS II core (Durnford and Green 1996),respectively, the total number of Chl molecules will be170 assuming that PS I – PS II ratio is 1 – 1. This valueis very close to the ratios of Chl against PS I and PS IIobserved in thylakoid membranes. This indicates thatall the Chl molecules are present in the cores of PS Iand PS II. There would be no Chl for LHC(s) to carry.

211

Figure 2. Polypeptide compositions of cyanelles (a), thylakoidmembranes (b), and Photosystem I-enriched particles (c) fromC.paradoxacyanelles. SDS–PAGE was performed on 16–22% acryl-amide gradient gel. The bands were stained with Coomassie brilliantblue R-250. Assignments of the proteins based on Figures 3 and 5are indicated on the right.

If this is the case, the LHC-related genes are aquiredafter glautocystophyte was branched off from a line toother eukaryotic algae.

SDS–PAGE profile

Resolution of protein components of cyanelles,thylakoids and the PS I-enriched fraction by SDS–PAGE are depicted in Figure 2. The cyanelles showintense bands of phycobiliproteins (14–17 kDa), aswell as distinct bands of other phycobilisome com-ponents (anchor (100 kDa) and linker (30–31 kDa)polypeptides, Figure 2, lane a). Most of the phycobili-proteins and anchor and linker polypeptides wereremoved in the thylakoid membrane fraction (Fig-ure 2, lane b). The membranes did not show an

oxygen-evolving activity, but still retained activities ofP700 photo-oxidation and PS I electron transport fromDCPIPH2 to methylviologen.

The PS I-enriched fraction consists of eight low-molecular-mass components in addition to an intensebut fussy PsaA/B band at the molecular mass of 61kDa and two contaminating proteins in between. Themolecular masses of the low-molecular-mass compon-ents, whose N-terminal sequences were determined,were 19, 18, 11.5, 9, 8.6, 4.2, 4.1 and 3.7 kDa. In addi-tion to them, 7.2 and 5.4 kDa bands were reproduciblyobserved.

Components whose genes are encoded in nucleargenome

Forty four amino acid residues were determined forthe 19 kDa protein (Figure 3A). The protein wasidentified as the PsaD protein by homology search.Surprisingly, it has the highest homology to the PsaDprotein ofC. reinhardtii. It showed 61% identity and68% homology in the determined 44 residues. The se-quence was also homologous to those of tomato andspinach, showing 55 and 50% identity, respectively.However, the homology is decreased to 32% or loweragainstP. purpurea, O. sinensis, Guillardia thetaorcyanobacteria.

The PsaD protein posseses the following featuresin the determined sequence: (1) The protein ofC.paradoxa, as well asC. reinhardtii and land plants,has an N-terminal stretch of about 20 amino acidscompared with that of other algae or cyanobacteria.ThepsaDgene of algae represented in Figure 3A (P.purpurea, O. sinensisand G. theta) is encoded inchloroplast genome, while that ofC. reinhardtii andhigher plants is encoded in nuclear genome. (2) The N-terminal stretch is highly homologous with each other.It is characterized by two clusters; one is enriched inGlu and Ala residues in the N-terminal side of thestretch and the other is enriched in Lys and Ala in theinner part of the stretch. Clusters enriched in Glu/Lysand Ala in the stretch seem to be a prominent feature ofPsaD protein whose gene is transferred to the nucleargenome.

The N-terminal stretch is very important for stablebinding of PsaC protein to PS I reaction center coreof barley (Naver et al. 1995). They assumed thatN-terminal stretch may express its function throughinteraction with a plant-specific subunit, PsaH. In thepresent study, PsaH protein was not detected inC.

212

Figure 3. N-terminal amino acid sequences of the PsaD (A) and PsaL (B) proteins and their alignments with corresponding sequences. TheN-terminal amino acid sequences determined in the present study are indicated in the respective top row. Only amino acid residues differentfrom those ofC. paradoxaare indicated in the Figure for other sources: the same residue is shown by a dot, a hyphen denotes a deletion, and Xindicates a residue which was not determined. Accession Nos: [1]: S05281 [2]: A32124 [3]: P51279 [4]: P49481 [5]: O78502 [6]: Q39615 [7]:P12372 [8]: P12353 [9]: S22209 [10]: P37277 [11]: F42799 [12]: P51222 [13]: P49486 [14]: O78469 [15]: Q41285 [16]: D83700 [17]: P2399.

paradoxaPS I-enriched particles. Based on their idea,the subunit interaction of PS I particles ofC. paradoxamay not be strong as that of higher plants.

We can conclude that the PsaD protein is moresimilar to that of chlorophytes and higher plants ratherthan cyanobacteria or other algae. In this respect, thePsaD protein has a characteristic of ‘green’ lineages.However, it should also be emphasized that the innerpart other than the N-terminal stretch showed higherhomology to those of cyanobacteria.

The N-terminus of the 11.5 kDa protein was notblocked and 39 amino acid residues were determined(Figure 3B). The reading frame of the correspondingprotein was not found in the cyanelle genome, and theprotein was identified as the PsaL protein by homo-logy search. Opposed to the case of the PsaD protein,the sequence of the protein has the highest homologyto that ofSynechococcus elongatus(Mühlenhoff et al.1993), a cyanobacterium and lowest homology to hi-ger plants. Although its gene is encoded in nucleargenome, the PsaL protein does not have an N-terminalstretch unlike the PsaD protein. Judging from the PsaLsequences, a stretch found in the PsaD protein is notthe prerequisite for PS I components whose geneswere transferred to nuclear genome.

Figure 4. Elution profiles from a Superdex 200 gel-filtrationcolumn of PS I particles ofS. 6803 (a),C. paradoxa(b) and spinach(c). Absorbance was monitored at 435 nm. Base lines of each elutioncurves were shifted for clarity.

The PsaL protein of cyanobacteria is supposed tobe necessary for trimerization of PS I complexes and

213

to span thylakoid membrane twice from its deducedamino acid sequence (Krauß et al. 1996). It is ofinterest whether the PS I-enriched particles exist astrimers or not because the determined PsaL sequenceshows highest homology to that ofS. elongatus. Thesize of the PS I particles were determined by gel per-meation chromatography and compared with those ofcyanobacterium and spinach (Figure 4). When PS Iparticles ofS. 6803 were applied to the column, anaggregate which eluted at void volume (0.92 ml) wasfollowed by trimeric form eluted at 1.00 ml. A verysmall peak of monomeric form was also eluted at1.15 ml (Figure 4a). The particles were aggregatedprobably due to concentration of the particles by ultra-centrifugation and by resuspension in the mediumwith no detergents. Apart from this, it was clearly seenthat trimeric form of PS I particles is present inS.6803. However, when PS I-enriched particles fromC.paradoxawere applied to the column, it gave a singlepeak eluted at 1.17 ml with an apparent molecularmass of 420 kDa, which corresponded to the molecu-lar mass of monomeric form of PS I (Figure 4b). ThePS I particles prepared from spinach also eluted at thesame position as that ofC. paradoxa(Figure 4c). It isalso of note that the elution volume of PS I-enrichedparticles ofC. paradoxacoincided with that of mono-meric form ofS. 6803 (see Figure 4a). This indicatesthat the PS I-enriched particles ofC. paradoxaexist asa monomeric form.

It is reported that when PS I – PS II ratio becomescloser to 1, the absorption band specific to the trimeris reduced, suggesting that the population of the tri-meric form of PS I in thylakoid membranes is reduced(Westerman et al. 1999). Judging from the fact thatthe PS I – PS II ratio is close to 1 in theC. para-doxa thylakoid membranes as discussed above, PS Iparticles may exist as a monomer.

There would be other possibilities to account forthe existence of the monomeric form in the isolatedPS I-enriched particles. Interaction within the trimer ofPS I particles ofC. paradoxamight not be so strong asin the case of cyanobacteria: the concentration of DM(0.8%) used in the present experiment totally solubil-ized the thylakoid membranes ofC. paradoxa, whilethe ratio of the solubilized Chl proteins inS. 6803thylakoid membranes was only 70%. Excess amountsof detergent might have dissociated trimeric form ofPS I particles to monomeric form.

An internal segment specific to higher plants ispresent in the PsaL sequence. Although the determ-ined N-terminal sequence shows higher affinity to

those of cyanobacteria, this internal segment mightalso be present inC. paradoxaPsaL protein, whichdoes not give rise to trimerization in thylakoid mem-branes. However, the conclusion should await fordetermination of the full length amino acid sequenceof the protein.

As to nuclear encoded PS I components, the PsaDsequence is rather similar to that of a chlorophyte andland plants than those of other eukaryotic algae. As faras the PsaD protein is concerned,C. paradoxashouldbe positioned closer to ‘green’ lineages. For phylo-genetic analysis of plastids, one should also take intoaccount the proteins which were transferred to nucleargenome.

Components whose genes are encoded in cyanellegenome

Entire sequence of the cyanelle genome has been re-ported (Stirewalt et al. 1995), however, determinationof N-termini of the PS I components of the matureform is not possible from the DNA sequences. Ac-cordingly, it is also very important to determine theN-terminus of each components.

N-terminal sequencing of the 61 kDa diffuse bandyielded a single sequence which completely matchedwith the deduced PsaB sequence ofC. paradoxawhilethe first Met residue was missing (Figure 5A). Thisindicates that the PsaB protein has an open N-terminuswhile the first Met residue was cleaved off. On theother hand, the PsaA protein sequence was not ob-tained, indicating that the protein has a blocked N-terminus. It has been reported that the N-terminus ofthe PsaB protein is not blocked inAnabaena29413(Nyhus et al. 1992) andSynechococcus vulcanus(Ikeuchi pers comm), but those of the PsaA and PsaBproteins in higher plants are both blocked (Hiyama1996). Thus, the characteristic of PsaB protein thathas an open N-terminus is not limited to cyanobac-teria but is also found inC. paradoxa. It is of notethat the N-terminus of the PsaB protein ofProchloronsp. is also open (Kashino and Maruyama, personalcommunication). Based on these results, lineages of‘cyanobacteria’ may have an open N-terminus of thePsaB protein. Sequencing of the psaB protein of redalgae and diatoms is of special interest to answerthis question; whether blockage of N-terminus of theprotein is limited in ‘green’ lineages or not.

The N-terminal sequence of the 18 kDa proteincompletely matched with that of internal amino acidsequence of the PsaF protein deduced from the DNA

214

Figure 5. N-terminal amino acid sequences and their alignments with corresponding sequences deduced from cyanelle genome. The N-terminalamino acid sequences determined in the present study are indicated in the respective higher row. Residues different from those deduced fromDNA sequence (Stirewalt et al. 1995) are indicated in the lower row: the same residue is shown by a dot, and X indicates a residue which wasnot determined. The first Met residue underlined in (C) indicates that the residue is missing in a large fraction of the protein.

sequence except a Cys residue (Figure 5D). The ap-parent molecular mass of the protein coincided wellwith the calculated molecular mass (17.8 kDa) of themature protein in which a lumen-targeted presequenceof 25 amino acid residues has been cleaved off. Thedetermined N-terminus matched with that of matureprotein estimated by Löffelhardt et al. (1997). Whenthe whole PsaF sequence is compared with those ofcyanobacteria, eukaryotic algae and higher plants,C.paradoxaposseses an internal stretch which is com-mon to chloroplast bearing eukaryotes. The over allsequence was relatively similar to those ofP. pur-purea, O. sinensisandG. theta, which are consideredas ‘red’ lineages.

The 9 kDa protein band gave two sequences (Fig-ure 5C). Both the sequence with a higher signal intens-ity and that with a intensity of about one third of thestronger signal corresponded to the PsaE protein. Themajor component started from the second Val residue,but the minor one started from the first Met residue.Judging from the signal intensity mentioned above, theresults indicate that the first Met residue is remaining

in about 25% of the PsaE protein. This kind of fea-ture has also been observed inS. vulcanus(Koike etal. 1989) but not inA. variabilis 29413 (Nyhus et al.1992). It is not clear whether the minor one has justescaped from processing of its N-terminus or it hasdifferent function(s) from the major one.

A diffuse band was separated in the 8.6 kDa region(see Figure 2), which is a characteristic feature of thePsaC protein (Koike et al. 1989). Its N-terminal se-quence was determined up to 19 amino acid residues(Figure 5B). The sequence completely matched withthat of the PsaC protein except for the first Metand internal Cys residues (Rhiel et al. 1992). ThePsaC protein so far sequenced has lost its N-terminalMet residue posttranslationally (Hiyama 1996), whichseems to be a common feature of the PsaC protein.

Identification of three proteins ranging 4.2–3.7kDa has been reported elesewhere (Shibata et al.1999). In summary, 4.2, 4.1 and 3.7 kDa proteins wereidentified as PsaJ, PsaI and PsaM proteins, respect-ively. The first Met residue was cleaved off in PsaJ,while PsaI and PsaM proteins started with formyl-

215

Table 1. Properties of PS I subunits ofC. paradoxa

Protein Location of Molecular mass Similarityb to Functions

the genea Deduced Determined S. 6803 spinach

PsaA C 84 61 87 87 Primary reaction/

light harvesting

PsaB C 82 61 89 88 Primary reaction/

light harvesting

PsaC C 8.7 8.6 94 95 Reduction of Fd

PsaD N – 19 23 55 Fd-docking

PsaE C 8.0 9 64 66 Interact with Fd

PsaF C 18 18 53 57 PC-docking

PsaI C 3.8 4.1 66 74c Unknown

PsaJ C 4.4 4.2 62 69c Interact with PsaF

PsaK ND – – – – Unknown

PsaL N – 11.5 54 49 Trimerization of PS I

PsaM C 3.3 3.7 67 60c Unknown

a C – cyanelle genome, N – nuclear genome, ND – not detected.b Similarities are expressed as percent for the whole sequences for plastid encoded components, whilefor PsaD and PsaL, they were caluculated in the range determined in the present study.cThe sequences of PsaI, PsaJ and PsaM ofMarchintia polymorphaare used because data of spinachare unavailable. Fd – ferredoxin, PC – plastocyanin.

Met. The PsaJ and PsaI proteins are found in all theplastid genomes whose entire DNA sequence has beendetermined, while the PsaM protein is not found in an-giosperm chloroplasts (Ohyama et al. 1986; Shinozakiet al. 1986; Hiratsuka et al. 1989; Hallick et al.1993; Wakasugi et al. 1994, 1997; Reith and Mun-holand 1995; Kowallick et al. 1995; Maier et al. 1995;Douglas and Penny 1999).

We did not identify the PsaK protein in PS I-enriched particles. The PsaK protein has been iden-tified from cyanobacteria to higher plants althoughhomology is not high between ‘green’ lineages andcyanobacteria andP. purpurea(Hiyama 1996). Mo-lecular mass of the protein is 6–7 kDa depending onspecies. The 7.2 kDa protein of the particles is closeto the expected molecular mass of the PsaK protein.Sequencing of this band gave no significant signalssuggesting the blockage of its N-terminus. It is pos-sible that the PsaK protein does exist in the particles,but if its N-terminus is blocked, the protein was notidentifiable.

No significant signals were obtained as well for the5.4 kDa protein presumably due to blockage of its N-terminus even after HCl treatment. The identificationof the 5.4 kDa protein is also remaining to be clarified.

Table 1 summarizes the properties of PS I com-ponents investigated in the present study (Golbeck1992). The genes of the components are encoded in

the cyanelle genome exceptpsaD and psaL. All thePS I components are encoded in chloroplast genome in‘red’ lineages (P. purpurea, O. sinensisandG. theta)while psaE, psaF, psaKandpsaMgenes are located inthe nuclear genome together withpsaDandpsaLin theangiosperms (Shinozaki et al. 1986; Hiratsuka et al.1989; Wakasugi et al. 1994; Maier et al. 1995). As faras the location of the genes is concerned,C. paradoxais situated between ‘red’ lineages and ‘green’ lineages.

The molecular masses determined in the presentstudy coincided well with those deduced from DNAsequences except PsaA and PsaB. These two poly-peptides migrate faster in a SDS–PAGE gel (Koikeet al. 1989; Nyhus et al. 1992) probably because thepolypeptides are not completely unfolded even in thepresence of SDS due to their strong hydrophobicity.

The sequences of the PsaA, PsaB and PsaC ofC.paradoxaare highly homologous to eitherS. 6803and spinach; the similarities are almost 90% or higher(Table 1). The similarity of other components is lowercompared with the three components. As indicated,the determined N-terminal sequence of the PsaD issimilar to that of spinach rather thanS. 6803.1 On theother hand, PsaL shows higher similarity to that ofS. 6803 in the range determined in the present study.Comparison of the PsaI and PsaJ in a whole sequenceshows higher similarity to those ofM. polymorpha,while the PsaM is more similar to that ofS. 6803. This

216

also supports the idea thatC. paradoxaPS I has a chi-meric characteristic between cyanobacteria and higherplants. Further investigation is of course necessary tocharacterize PS I components.

Acknowledgements

Authors wish to thank Dr E. Gantt, University ofMaryland, and Dr Y. Takahashi, Okayama University,for the immunoblotting analysis. We thank Dr J. R.Shen, RIKEN, for determination of molecular size ofPS I particles. We express our thanks to Dr E. Ganttfor critcal reading of our manuscript. We also wishto thank Dr H. Hirata and Dr M. Kanamori, HimejiInstitute of Technology, for helpful instructions of pro-tein sequencing. This work was supported in part by aGrant-in-Aid for the scientific research from Ministryof Education, Science, Sports and Culture to H.K. (No.08640836).

Note

1 After submission of the manuscript, it turned out that cDNA se-quence ofpsaD was disclosed (Accession No.: AJ132477). Thedetermined N-terminal sequence except one residue which was notdetermined, completely matched with the deduced amino acid se-quence started at the position of 54. It was thus confirmed that thecyanelle targeting transit peptide is cleaved between 53rd Ala reidueand 54th Glu residue. Comparison of entire amino acid sequence ofthe mature protein still shows higher homology to spinach (64%)rather thanS. 6803 (60%), although the degree of similarity toS.6803 increased because internal part of the sequence shows higherhomology toS. 6803. It is of note thatC. paradoxaPsaD pro-tein was placed to higher plants andChlamydomonasbranch ratherthan cyanobacteria or other algae by phylogenetic analysis usingCLUSTAL W program.

References

Bass R, Soen SY, Frank G, Zuber H and Rochaix J-D (1992) Char-acterization of Chlorophylla/b proteins of Photosystem I formChlamydomonas reinhardtii. J Biol Chem 267: 25714–25721

Burnap RL and Trench RK (1989) The biogenesis of the cyanel-lae ofCyanophora paradoxa. I. Polypeptide composition of thecyanellae. Proc R Soc London B 238: 53–72

Douglas SE and Penny SL (1999) The plastid genome of the crypto-phyte alga,Guillardia theta: Complete sequence and conservedsynteny groups confirm its common ancestry with red algae. JMol Evol 48: 236–244

Durnford DG, Deane JA, Tan S, McFadden GI, Gantt E andGreen BR (1999) A phylogenetic assessment of the eukaryoticlight-harvesting antenna proteins, with implications for plastidevolution. J Mol Evol 48: 59–68

Green BR and Durnford DG (1996) The Chlorophyll-carotenoidproteins of oxygenic photosynthesis. Annu Rev Plant PhysiolPlant Mol Biol 47: 685–714

Golbeck JH (1992) Structure and function of Photosystem I. AnnRev Plant Physiol Plant Mol Biol 43: 293–324

Hall WT and Claus G (1963) Ultrastructural studies on the blue-green algal symbiont inCyanophora paradoxaKorschikoff. JCell Biol 19: 551–563

Hallick RB, Hong L, Drager RG, Favreau MR, Montfort AB, OrsatA, Spielmann A and Stutz E (1993) Complete sequence of theEuglena gracilisChloroplast DNA. Nucleic Acids Res 21: 3537–3544

Hiratsuka J, Shimada H, Whittier R, Ishibashi T, Sakamoto M, MoriM, Kondo C, Honji Y, Sun C-R, Meng B-Y, Li Y-Q, Kanno A,Nishizawa Y, Hirai A, Shinozaki K and Sugiura M (1989) Thecomplete sequence of the rice (Oryza sativa) chloroplast gen-ome: Intermolecular recombination between ditinct tRNA genesaccount for a major plastid inversion during the evolution ofcereals. Mol Gen Gen 217: 185–194

Hiyama T (1996) Photosystem I: Structures and functions. In: Pess-arakli M (ed) Handbook of Photosynthesis, pp 195–217. MarcelDekker, New York

Ichimura T (1971) Sexual cell division and conjugation-papillaformation in sexual reproduction ofChlosterium strigosum. In:Proc. 7th Internatl. Seaweed Symposium pp 208–214. Universityof Tokyo Press, Tokyo

Ikeuchi M and Inoue Y (1988) A new Photosystem II reaction centercomponent (4.8 kDa protein) encoded by chloroplast genome.FEBS Lett 241: 99–104

Ikeuchi M, Sonoike K, Koike H, Pakrasi HB and Inoue Y (1992) Anovel 3.5 kDa protein component of cyanobacterial PhotosystemI complexes. Plant Cell Physiol 33: 1057–1063

Kaneko T, Sato S, Kotani H, Tanaka A, Asamizu E, NakamuraY, Miyajima N, Hirosawa M, Sugiura M, Sasamoto S, KimuraT, Hosouchi T, Matsuno A, Muraki A, Nakazaki N, Naruo K,Okumura S, Shimpo S, Takeuchi C, Wada T, Watanabe A, Ya-mada M, Yasuda M and Tabata S (1996) Sequence analysis ofthe genome of the unicellular cyanobacteriumSynechocystissp.Strain PCC 6803. II. Sequence determination of the entire gen-ome and assignment of potential protein-coding regions. DNARes 3: 109–136

Koike H, Ikeuchi M, Hiyama T and Inoue Y (1989) Identificationof Photosystem I components from the cyanobacterium,Syne-chococcus vulcanusby N-terminal sequencing. FEBS Lett 253:257–263

Kowallik KV, Stoebe B, Schaffran I, Kroth-Pancic P and Freier U(1995) The complete genome of a chlorophylla+c-containingalga,Odontella sinensis.Plant Mol Biol Reptr 13: 336–342

Krauß N, Schubert W-D, Klukas O, Fromme P, Witt HT and SängerW (1996) Photosystem I at 4̊A resolution represents the firststructural model of a joint photosystem reaction centre and coreantenna system. Nature Struct Biol 3: 965–973

Löffelhardt W and Bohnert HJ (1994a) Molecular biology of cyan-elles In: Bryant DA (ed) The Molecular Biology of Cyanobac-teria, pp 65–89. Kluwer Academic Publishers, Dordrecht, TheNetherlands

Löffelhardt W and Bohnert HJ (1994b) Structure and function ofcyanelle genome. Internatl Rev Cytol 151: 29–65

Löffelhardt W, Bohnert HJ and Bryant DA (1997) The cyanelles ofCyanophora paradoxa. Critical Rev Plant Sci 16: 393–413

Maier RM, Neckermann K, Igloi GL and Koessel H (1995) Com-plete sequence of the maize chloroplast genome: Gene content,hotspots of divergence and fine tuning of genetic information bytranscript editing. J Mol Biol 251: 614–628

217

Martin W, Bettina S, Goremykin V, Hansmann S, Hasegawa Mand Kowallik KV (1998) Gene transfer to the nucleus and theevolution of chloroplasts. Nature 393: 162–165

Mühlenhoff U, Haehnel W, Witt HT and Herrmann RG (1993)Genes encoding 11 subunits of Photosystem I from the thermo-philic cyanobacterium Synechococcus sp. Gene 127: 71–78

Naver H, Scott P, Anderson B, Møller BL and Scheller V (1995) Re-constitution of barley Photosystem I reveals that the N-terminusof the PS I-D subunit is essential for tight binding of PS I-C.Physiol Plant 95: 19–26

Nyhus K, Ikeuchi M, Inoue Y, Whitmarsh J and Pakrasi HB (1992)Purification and characterization of the Photosystem I com-plex from the filamentous cyanobacterium Anabaena variabilisATCC 29413. J Biol Chem 25: 12489–12495

Ohyama K, Fukuzawa H, Kohchi T, Shirai H, Sano T, Sano S,Umesono K, Shiki Y, Takeuchi M, Chang Z, Aota S, InokuchiH and Ozeki H (1986) Chloroplast gene organization deducedfrom complete sequence of liverwort Marchantia polymorpha.Nature 322: 572–574

Paula De JC, Innes JB and Brudvig GW (1985) Electron transferin Photosystem II at cryogenic temperatures. Biochemistry 24:8114–8120

Reith J and Munholland J (1995) Complete nucleotide sequence ofthePorphyrapurpurea chloroplast genome. Plant Mol Biol Reptr13: 333–335

Riehl E, Stirewalt VL, Gasparich GE and Bryant DA (1992) ThepsaC genes of Synechococcus sp. PCC7002 and Cyanophoraparadoxa: Cloning and sequence analysis. Gene112: 123–128

Shen JR (1999) Possible functional differences between dimer andmonomer of Photosystem II complex. In: Garab G (ed) Photo-synthesis: Mechanisms and Effects, Vol II , pp 941–944. KluwerAcademic Publishers, Dordrecht, TheNetherlands

Shibata M, Kashino Y, Satoh K and KoikeH (1999) Identification ofprotein componentsof photosystemsof cyanelle-thylakoids fromCyanophora paradoxa. In: Garab G (ed) Photosynthesis: Mech-anisms and Effects, Vol IV, pp 3107–3110. Kluwer AcademicPublishers, Dordrecht, TheNetherlands

Shinozaki K, Ohme M, Tanaka M, Wakasugi T, Hayashida N,Matsubayashi T, Zaita N, ChunwongseJ, Obokata J, Yamaguchi-Shinozaki K, Ohto C, Torazawa K, Meng BY, Sugita M, DenoH, Kamogashira T, Yamada K, Kusuda J, Takaiwa F, Kato A,Tohdoh N, Simada H and Sugiura M (1986) The complete nuc-leotide sequence of the tobacco chloroplast genome: Its geneorganization and expression. EMBO J 5: 2043–2049

Stirewalt VL, Michalowski CB, Loeffelhardt W, Bohnert HJ andBryant DA (1995) Nucleotide seuence of the cyanelle genomefrom Cyanophora paradoxa. Plant Mol Biol Reptr 13: 327–332

Tan S, Wolfe GR, Cunningham FX Jr and Gantt E (1995) Decreaseof polypeptides in the PS I antenna complex with increas-ing growth irradiance in the red algaPorphyridium cruentum.Photosynth Res 45: 1–10

Tan S, Ducret A, Aebersold R and Gantt E (1997) Red algal LHCI genes have similarities with both Chl a/b- and a/c-binding pro-teins: A 21 kDa polypeptide encoded by LhcaR2 isoneof thesixLHC I polypeptides. Photosynth Res 53: 129–140

Wakasugi T, Tsudsuki J, Ito S, Nakashima K, Tsudzuki T and Sug-iura M (1994) Lossof all ndh genesasdetermined by sequencingtheentire chloroplast genomeof theblack pinePinus thunbergii.Proc Natl Acad Sci USA 91: 9794–9798

Wakasugi T, Nagai T, Kapoor M, Sugita M, Ito M, Ito S, TsudzukiJ, Nakashima K, Tsudzuki T, Suzuki Y, Hamada A, Ohta T,Inamura A, Yoshinaga K and Sugiura M (1997) Complete nuc-leotide sequence of the chloroplast genome from the green algaChlorella vulgaris: The existence of genes possibly involved inchloroplast division. Proc Natl Acad Sci USA 94: 5967–5972

Westermann M, Neuschaefer-Rube O, Mörschel E and WehrmeyerW (1999) Trimeric Photosystem I complexes exist in vivo inthylakoid membranes of the Synechocystis strain B9201 and dif-fer in absorption characteristics from monomeric Photosystem Icomplexes. J Plant Physiol 155: 24–33

Wolfe GR, Cunningham FX, Dunhord D, Green BR and Gantt E(1994) Evidence for a common origin of chloroplasts with light-harvesting complexes of different pigmentation. Nature 367:566–568