ORIGINAL PAPER

Loreta Gudynaite-Savitch Æ Michael Gretes

Rachael M. Morgan-Kiss Æ Leonid V. Savitch

John Simmonds Æ Susanne E. Kohalmi

Norman P. A. Huner

Cytochrome f from the Antarctic psychrophile, Chlamydomonas raudensisUWO 241: structure, sequence, and complementation in the mesophile,Chlamydomonas reinhardtii

Received: 4 October 2005 / Accepted: 16 December 2005 / Published online: 20 January 2006� Springer-Verlag 2006

Abstract Although cytochrome f from the Antarcticpsychrophile, Chlamydomonas raudensis UWO 241,exhibits a lower apparent molecular mass (34 kD) thanthat of the mesophile C. reinhardtii (41 kD) based onSDS-PAGE, both proteins are comparable in calculatedmolecular mass and show 79% identity in amino acidsequence. The difference in apparent molecular masswas maintained after expression of petA from bothChlamydomonas species in either E. coli or a C. rein-hardtii DpetA mutant and after substitution of a uniquethird cysteine-292 to phenylalanine in the psychrophiliccytochrome f. Moreover, the heme of the psychrophilicform of cytochrome f was less stable upon heating thanthat of the mesophile. In contrast to C. raudensis, a

C. reinhardtii DpetA mutant transformed with petA fromC. raudensis exhibited the ability to undergo state tran-sitions and a capacity for intersystem electron transportcomparable to that of C. reinhardtii wild type. However,the C. reinhardtii petA transformants accumulated lowerlevels of cytochrome b6/f complexes and exhibited lowerlight saturated rates of O2 evolution than C. reinhardtiiwild type. We show that the presence of an altered formof cytochrome f in C. raudensis does not account for itsinability to undergo state transitions or its impairedcapacity for intersystem electron transport as previouslysuggested. A combined survey of the apparent molecularmass, thermal stability and amino acid sequences ofcytochrome f from a broad range of mesophilic speciesshows unequivocally that the observed differences incytochrome f structure are not related to psychrophilly.Thus, caution must be exercised in relating differences inamino acid sequence and thermal stability to adaptationto cold environments.

Keywords Chlamydomonas raudensis Æ Chlamydomonasreinhardtii Æ Cytochrome f Æ Thermostability ÆState transitions Æ Electron transport

Abbreviations DA820: Change in absorbance at820 nm Æ BBM: Bold’s Basal medium Æ SD: Shine–Dalgarnosequence Æ CES: Control by epistaticsynthesis Æ cpDNA: Chloroplast DNA Æ CP43:Photosystem II core antenna polypeptide Æ CTAB:Cetyltrimethyl-ammonium bromide Æ cyt f, cyt b6/f:Cytochrome f and b6/f, respectively Æ D1: PhotosystemII reaction centre polypeptide Æ DTT:Dithiothreitol Æ F682 and F710: 77 K Chl a fluorescenceemission maximum at 682 nm and at 710 nm,respectively Æ FR: Far red light Æ HS: High saltmedium Æ IPTG: Isopropylthio-b-D-galactoside Æ LHCI,LHCII: Light harvesting complex I and II,respectively Æ MT and ST: Multiple- and single-turnoverflash, respectively Æ P700 and P700+: Reduced and

Electronic Supplementary Material Supplementary material isavailable for this article at http://dx.doi.org/10.1007/s00438-005-0094-4 and is accessible for authorized users.

Communicated by S. Hohmann

L. Gudynaite-Savitch Æ M. Gretes Æ S. E. KohalmiN. P. A. Huner (&)Department of Biology and The Biotron,University of Western Ontario, London, Canada, N6A 5B7E-mail: nhuner@uwo.caTel.: +1-519-6612111Fax: +1-519-6613935

R. M. Morgan-KissDepartment of Microbiology, University of Illinois,Champagne-Urbana, 61801, USA

L. V. Savitch Æ J. SimmondsECORC, Agriculture and Agri-Food Canada,Ottawa, Canada, K1A 0C6

Present address: L. Gudynaite-SavitchDepartment of Biology, Plant Molecular Biology,University of Ottawa, Ottawa, Canada, K1N 6N5

Present address: M. GretesDepartment of Botany, University of British Columbia,Vancouver, Canada

Mol Gen Genomics (2006) 275: 387–398DOI 10.1007/s00438-005-0094-4

oxidized photosystem I reaction centre,respectively Æ PC: Plastocyanin Æ petA: Cyt f codinggene Æ PQ: Plastoquinone Æ PQH2: Plastoquinol Æ PsaA/B: Photosystem I reaction centre polypeptides Æ PsaE,PsaH, PsaL: Photosystem I polypeptides Æ PSI, PSII:Photosystem I and II, respectively Æ RApetA: petA genefrom C. raudensis; RA1-5 – C. reinhardtii RApetAtransformant lines Æ rbcL: Rubisco Large subunit codinggene Æ REpetA: petA gene from C. reinhardtii Æ TAP:Tris–acetate–phosphate medium Æ TMBZ: 3,3¢,5,5¢-tetramethylbenzidine

Introduction

Cytochrome b6/f is an obligatory electron transfer andproton-translocating complex in all oxygenic photosyn-thetic organisms. This complex couples the oxidation ofplastoquinol (PQH2) with the reduction of plastocyanin(PC), the electron donor to P700+ (Hauska et al. 1996).Proton translocation across the photosynthetic mem-brane, driven by the Q cycle associated with the cyt b6/fcomplex contributes to the formation of the protongradient required for ATP synthesis through either lin-ear or PSI cyclic electron transport (Hauska et al. 1996).

The unprocessed form of cyt f consists of 317 aminoacids in C. reinhardtii and is the largest polypeptide inthe cyt b6/f complex. The signal peptide of 31 residuesmediates translocation of cyt f across the thylakoidmembrane and is cleaved from the mature protein by alumenal processing peptidase. This signal sequence is themost variable region of cyt f and is characterized by ahydrophobic region flanked by a charged residue on itsamino terminus (Bailet and Kohorn 1996).

The processed form of cyt f from C. reinhardtii con-sists of 286 amino acids and its apparent molecular massis 41 kD as estimated by SDS-PAGE. It is anchored tothe thylakoid membrane by a single trans-membranehelix, consisting of amino acids 252–271 of the maturecyt f whereas the conserved, C-terminal 15 amino acidsextend into the stroma. The soluble part of cyt f resides inthe thylakoid lumen and consists of large and small do-mains. Mutations within the large domain (Tyr-1 to Gly-171) of cyt f inhibit electron transfer (Ponamarev andCramer 1998), proton translocation (Ponamarev andCramer 1998), and maturation of cyt f. The heme-Fe iscoordinated by the side chain of His25 and Tyr1 and theheme-binding motif, 21-Cys-Ala-Asn-Cys-His-25, is lo-cated in the longest conserved stretch of amino acids incyt f (Gray 1992; Ponamarev et al. 2000). Two conservedcysteine residues (Cys 21 and Cys 24) provide the thioe-ther linkages for heme-binding in cyt f (Gray 1992).

In addition to regulating intersystem electron trans-port, cyt b6/f complex is involved in regulating state I tostate II transitions (Zito et al. 1999; Finazzi et al. 2001).State transitions involve the reversible phosphorylationof polypeptides associated with the mobile, peripherallight harvesting complex (LHCII) of PSII. This causes adissociation of LHCII from PSII with its concomitant

re-association with PSI (Allen 1995; Wollman 2001). Itappears that the redox state of the PQ pool acts as thesensor for state transitions (Wollman and Lamaire1988). The binding of PQH2 to the Qo site of cyt b6induces the rotation of the Rieske Fe–S protein as wellthe dimerization of the cyt b6/f complex (Zito et al. 1999;Finazzi et al. 2001; Wollman 2001). The dimerization ofthe cyt b6/f complex activates a thylakoid LHCII-kinasewhich phosphorylates LHCII polypeptides (Wollman2001). This kinase was recently identified as a Stt7chloroplast protein kinase which is required for theregulation of state transitions and LHCII phosphoryla-tion in Chlamydomonas (Depege et al. 2003). In addi-tion, Arabidopsis thaliana deletion mutants of the PSIpolypeptides, PsaE, PsaH and PsaL, exhibit an impairedability to phosphorylate the LHCII polypeptides andwere unable to undergo state transitions (Lunde et al.2000; Pesaresi et al. 2002).

State I to state II transitions in C. reinhardtii havebeen monitored in vivo by measuring 77 K chlorophyll afluorescence emission spectra (Lebedev et al. 1988; Linand Knox 1991; Morgan-Kiss et al. 2002a; 2002b).Emission maxima in the regions of 680–682 nm (F682),686 nm (F686) and 696 nm (F696) reflect dissociatedperipheral LHCII, proximal light harvesting antenna ofPSII, and the PSII core complex respectively. Fluores-cence emission maxima associated with the light har-vesting complex I and the core proteins of PSI appear at707 nm (F707) and 710–717 nm (F710), respectively. StateI to state II transitions can be assessed by the decrease inthe F682 relative to the concomitant increase in the F710

emission (Lebedev et al. 1988; Lin and Knox 1991;Morgan-Kiss et al. 2002a).

In contrast to C. reinhardtii, 77 K Chl a fluorescenceemission spectra of C. raudensis indicated that the psy-chrophile is unable to modulate energy distribution be-tween PSII and PSI through state transitions (Morgan-Kiss et al. 2002a, b). Since C. raudensis appears to belocked in state I, it was suggested that this psychrophileis a natural state transition mutant (Morgan-Kiss et al.2002a). The absence of the major PSI-associated Chlfluorescence emission band in C. raudensis even underconditions favoring state II is not associated withphosphorylation of LHCII but rather phosphorylationof a unique set of high molecular mass thylakoid poly-peptides (Morgan-Kiss et al. 2002a, 2005). Structurally,this characteristic is associated with reduced levels of PsaA/B polypeptides relative to D1 and either the absenceor a significant decrease in specific LHCI polypeptides(Morgan et al. 1998). In addition, the cyt b6/f complex ofC. raudensis exhibits a form of cyt f protein which isapproximately 7 kD smaller in apparent molecular massthan that of C. reinhardtii (Morgan-Kiss et al. 2002a).

To test the hypothesis that the lower apparentmolecular mass reported for cyt f from the Antarcticpsychrophile, C. raudensis, and its deficiency in state I tostate II transitions reflect alterations in the chloroplast-encoded petA gene (Morgan-Kiss et al. 2002a), wecloned and sequenced petA from C. raudensis and

388

compared it with the petA sequence from the modelmesophile, C. reinhardtii. Through chloroplast trans-formation of a DpetA mutant of C. reinhardtii, we have,for the first time to our knowledge, complemented amesophilic with the psychrophilic form of cyt f. Thesetransformants were tested for their ability to undergostate transitions and their capacity for intersystem elec-tron transport.

Materials and methods

Growth conditions

Chlamydomonas reinhardtii UTEX 89 (ChlamydomonasGenetics Centre) was grown in Bold’s Basal Medium(BBM) (Nichols and Bold 1965) whereas C. reinhardtiiWT.11 and the RApetA and REpetA transformantswere grown in Sueoka High-Salt Media (HS) (Harris1989). The C. reinhardtii cyt f deletion mutant, DpetA,was grown at 25�C under ambient CO2 and continuouslight (20 lmol photons/m2s) in TAP medium (Harris1989) containing 100 lg/ml of spectinomycin. C. raud-ensis was grown under ambient CO2 conditions andcontinuous light (100–150 lmol photons/m2s) in BBMsupplemented with 0.7 M NaCl at either 8�C or 16�Caccording to (Morgan et al. 1998).

DNA isolation and cloning

Total algal DNA was isolated according to Rochaixet al. (1988) and purified from polysaccharides withCTAB (cetyltrimethyl-ammonium bromide) as describedby Murray and Thompson (1980). Chloroplast DNA(cpDNA) was purified using a NaI gradient as describedby Harris (1989). HaeIII digestion was used to generatea �2.9 kb, fragment containing the entire C. raudensispetA gene. Digested cpDNA was separated on a 1% (w/v) agarose gel and the �2.3 to 3.5 kb HaeIII fragmentswere isolated from the agarose gel and cloned into theSmaI site of pUC19 (Gibco BRL). A plasmid (pUC19-RApetA) containing the C. raudensis petA gene wasdetected by Southern hybridization with a petA probefrom C. reinhardtii.

The gene coding CP43 (psbC) was amplified by PCRwith a High Fidelity PCR system (Roche MolecularBiochemical) using the cpDNA from C. raudensis and C.reinhardtii as templates. The forward primer was PsbC-CR-Start (5¢ ATGGAAACACTTTTTAATGGAAC 3¢)and the reverse primer was PsbC-CR-Stop (5¢ TTAGTCTAAAGGACGCATTGAAAG 3¢).

To clone the C. raudensis Rubisco large subunit gene,rbcL, purified cpDNA was digested with HaeIII. The�3.75 kb fragment detected by Southern hybridizationwith an rbcL probe from C. reinhardtii was isolated fromagarose and cloned into the SmaI site of pUC19.

Nucleotide sequences for petA, psbC, and rbcL weredetermined by sequencing both strands of the entire

cloned or PCR-amplified fragments (GenBank accessionnumbers: petA, A4039799; psbC, DQ196176; rbcL,DQ196177). Corresponding amino acid sequences werepredicted withDNAman program using universal geneticcode and aligned with corresponding C. reinhardtii se-quences (GenBank accession number BK000554) usingDNAman fast alignment (K-tuple 3, gap penalty 5).

SDS-PAGE and immunoblotting

Thylakoid membrane polypeptides were isolated fromexponentially growing algal cultures as described byMorgan et al. (1998). Samples were loaded on an equalchlorophyll basis (5–10 lg per lane) and electrophoret-ically separated in a 15% (w/v) polyacrylamide gelcontaining 6 M urea according to Laemmli (1970).Polypeptides separated by SDS-PAGE were transferredelectrophoretically to 0.2 lm supported nitrocellulosemembranes (Bio-Rad) and immunobloted with specificantibodies following the manufacturer’s recommenda-tions. Probes used included polyclonal antibodies raisedagainst the 15 stromal C-terminal amino acids (cyt fCter) and the 21 amino acids at the N-terminus of maturecyt f (cyt f Nter) from spinach (Fig. 1S) as well as anti-bodies against isolated cyt f, cyt b6 and Rieske Fe–Sprotein and PsaA from C. reinhardtii. Antibodies to D1,CP43, Lhcb2, Lhcb1 and b-ATPase were obtained fromAgriSera (Umea, Sweden). Phosphorylated polypeptideswere detected using phosphothreonine-specific antibod-ies (Zymed Lab. Inc.).

Heme thermostability

Thylakoid membrane proteins were isolated as describedpreviously and suspended into solubilization buffer (Mor-gan et al. 1998)with a final SDS concentrationof 4%(w/v).After 1 min incubation at 95�C, aliquots were incubated at75�C and sampled at 5 min intervals for a total of 40 min.Polypeptides were separated by SDS-PAGE and trans-ferred to nitrocellulose membrane as described above.Heme binding to cyt f was visualized by incubation withECL chemiluminescence detection reagents (PharmaciaAmersham) and developed onX-ray film (Kodak) (Vargaset al. 1993). The intensity of the heme bands after heattreatment was compared to the intensity of bands in sam-ples prior to heating and quantified by densitometry usingScion Image (Scion Corporation).

Cyt f expression and site-directed mutagenesis

Cyt f coding sequences from C. raudensis and C. rein-hardtii were amplified by PCR introducing a HindIIIrestriction site after the stop codon at the 3¢ end (primerRApetA-H-Cter: 5¢ GCATAAGCTTAGAAGT TCATTTCTGCT 3¢) as well as an EcoRI restriction site, aShine–Dalgarno sequence and a start codon at the 5¢ end

389

(primers RApetA-E-Nter: 5¢ GGCGGCATTCTG-GAGGCTTTTACATATGTACCCAA TTTTTG CTCAAC3¢ and REpetA-E-Nter: 5¢GGCGGAATTCTGGAGGA TTT TACA TAT GTA CCC TGT ATT TGC-ACAA 3¢ for C. raudensis and C. reinhardtii, respec-tively). PCR products from both C. raudensis and C.reinhardtii were directionally cloned into EcoRI andHindIII sites of pSD80 placing the petA sequences underthe control of a Lac promoter (Smith et al. 1996) gen-erating pSD80-RApetA and pSD80-REpetA, respec-tively. The pSD80-RApetA plasmid was altered by site-directed mutagenesis using a Quick Change site-directedmutagenesis kit (Stratagene) and the primers: C292F-F(5¢ATTCTTCGCTTTTATTtTATTAGC3¢), C292F-R(5¢GCTAATAaAATAAAAGCGAAGAATG3¢) tointroduce a single nucleotide substitution into petA fromC. raudensis resulting in the C292F substitution. Allplasmids were transformed into competent DH5a E. colicells (Invitrogen) according to the manufacturer’s direc-tions. To induce petA expression and cyt f biosynthesis,IPTG was added during exponential growth, and cellswere collected after 4 h of incubation at 30�C. Crudeprotein extracts were prepared for SDS-PAGE by sus-pending cells in solubilization buffer containing 60 mMTris (pH 7.8), 1 mM EDTA, 12% (w/v) sucrose, 2% (w/v) SDS and incubating at 95�C for 5 min.

Chloroplast transformation

A partial petA sequence from C. raudensis was amplifiedby PCR using pUC19-RApetA as template. A 5¢ HindIIIrestriction site was introduced at the mature cyt f N-terminus coding sequence using the RApetA-Tr-H pri-mer (5¢ GCGAAGCTTACCCAATTTTTGCTC 3¢). A3¢AflII restriction site was added at the C-terminus of thetransmembrane helix coding sequence using the RA-petA-Tr-A primer (5¢ GCGCTTAAGAACTAATAAAATTTGAGC 3¢). The amplified PCR product wasdirectionally cloned into the HindIII and AflII restric-tion sites of pWF replacing the internal petA sequencefor C. reinhardtii and generating the plasmid pWF-RApetA. Thus, the resulting chimeric petA gene had thepromoter, terminator and N-terminal signal peptidesequences from C. reinhardtii but the mature cyt f codingsequence from C. raudensis (Fig. 1a). Due to the intro-duction of an AflII restriction site, the last amino acidfrom the transmembrane helix of C. raudensis wasaltered from Val to Leu. A C292F was introduced bysite-directed mutagenesis at G875T in pWF-RApetA ofthe cyt f coding region as described above.

C. reinhardtii DpetA cells used for transformationwere grown to mid-log phase in TAP medium containing100 lg/ml of spectinomycin (Kuras and Wollman 1994).Gold particles (1.0 lm diameter, Bio-Rad) were pre-pared for particle bombardment (Bio-Rad, model PDS-1000/He) according to Boynton and Gillham (1993).The plasmids pWF-RApetA, pWF-RApetA-C292F andpWF were used to transform DpetA cells and the gen-

eration of RApetA, C292F and REpetA transformantsrespectively. After bombardment, cells were incubatedovernight on TAP media without spectinomycin at roomtemperature at 20 lmol photons m�2 s�1. The next daycells were washed from the plate and spread on a Petridish containing HS medium. After 3–4 weeks of incu-bation at 20�C and an irradiance of 20 lmol photonsm�2 s�1, colonies were detected. After 2–3 transfers tofresh HS plates, transformants were tested for specti-nomycin resistance. Since all transformants were spec-tinomycin sensitive, we considered them to behomoplastic.

77 K Chl a fluorescence emission

Low temperature (77 K) Chl a fluorescence emissionspectra of wholeChlamydomonas cells weremeasured on aPTI Fluorometer (Model LS-1, Photon TechnologyInternational Inc.) as described by Morgan-Kiss et al.(2002a). To induce state I, mid-log phase cells were incu-bated in the dark for 20 min at their growth temperature.The transition from state I to state II was induced byincubating cells in the dark under anaerobic conditions for30 min to inhibit chlororespiration (Bennoun 1982).Anaerobiosis was induced by bubbling samples with N2

according to Wollman and Delepalaire (1984).

P700 measurements

The relative redox state of P700 in Chlamydomonas cellscollected from mid-log phase cultures was determined invivo under ambient CO2 conditions using a PAM-101modulated fluorometer (Walz, Germany) equipped withED-800T emitter-detector and PAM-102 units (Ivanovet al. 1998). Far red light (FR) (lmax=715 nm, 10 W/m,Schott filter RG 715) was provided by the FL-101 lightsource. Multiple turnover (MT, 50 ms) and singleturnover (ST, half peak width 14 ls) saturating lightflashes were applied with a XMT-103 and XST-103power control units (Walz, Germany), respectively, via amultibranched fiber optic system connected to theemitter-detector unit and the sample holder. The extentand kinetics of P700 oxidation were estimated as DA820/A820 and recorded using an oscilloscope card (PC-ScopeT6420, Version 2.43x; Intelligent Messtechnik GmbH).The complementary areas between the oxidation curveof P700 after ST and MT excitation, and the steady-statelevel of P700+ under FR were used to calculate the ratioof MT/ST which provides an estimate of the intersystemelectron donor pool size (e/P700) (Asada et al. 1992;Ivanov et al. 2001).

Oxygen evolution

Gross, light-dependent, CO2 saturated O2 evolution ofexponentially grown cells (25�C, 100 lmol photons/

390

m2 s) was measured at 25�C using a Clark-type oxygenelectrode (Hansatech Ltd.,). Actinic light from a fiber-optic halogen light source (Fiber-Lite, Dolan-JennerIndustries) was modulated with neutral density filtersand measured from within the cuvette by a customizedminiature photodetector (Hamamatsu G1125-02, Ham-amatsu Photonics Sys.) attached to a Licor LI-189radiometer. Apparent quantum yield for O2 evolutionand maximum photosynthetic capacity were calculatedfrom light response curves (Falk and Samuelsson 1992).

Chlorophyll and protein determinations

The chlorophyll concentration of thylakoid membranesamples was measured in the 90% (v/v) acetoneaccording to Harris (1989). Protein concentration ofthylakoid membrane samples was measured by solubi-lizing membrane aliquots in 2% (w/v) SDS, and using

the DC Protein Assay kit (Bio-Rad) as described by themanufacturer.

Results

Structure and stability of cyt f

The apparent molecular mass of cyt f from C. raudensis(34 kD) was 7 kD less than cyt f from C. reinhardtii(41 kD) (Fig. 2a) regardless of whether antibodies to theC-terminus (Fig. 2a, cyt f Cter), the N-terminus (Fig. 2a,cyt f Nter), the whole polypeptide (Fig. 2a, cyt f W) orheme staining with TMBZ (Fig. 2a, cyt fH) were used todetect cyt f. However, variation in the relative abun-dance of cyt f as assessed by the band intensity afterdetection by immunoblotting with either cyt f Cter, Nter,or W antibodies indicated a differential affinity of theseantibodies for the two forms of cyt f. The differences in

Fig. 1 a Scheme of chimericpetA gene in RApetAtransformants. b Amino acidsequence alignment of cyt fproteins from C. raudensis andC. reinhardtii. The proteinsequence was predicted fromthe DNA sequence and alignedusing the DNAman program.Black boxes indicate identicalamino acids; N-ter N terminusof mature cyt f protein; greyunderline heme binding motif;black underline transmembranehelix; black block arrowsposition of two conservedcysteine and third cysteineresidue found in cyt f fromC. raudensis; Asterisk lysineresidues involved inplastocyanin docking; � aminoacid residues involved in waterchain formation

391

affinity of antibodies raised against the Cter and the Nter

of cyt f is most likely due to the fact that the Cter se-quence was 100% identical whereas the Nter sequenceexhibited a single amino acid difference between petAfrom C. raudensis as compared to C. reinhardtii(Fig. 1S).

Neither the presence of urea nor reducing reagentssuch as DTT and b-mercaptoethanol affected the dif-ference in apparent molecular mass of cyt f from C.raudensis and C. reinhardtii (data not shown). The dif-ference in apparent molecular mass of cyt f was not dueto differential proteolytic cleavage since the combinationof three freshly prepared proteolytic inhibitors, benz-amidine, e-amino-N-caproic acid and phenymethyl-sul-fonyl fluoride, was used in all buffers used for thylakoidmembrane isolation. In addition, the recognition of bothproteins by the antibodies raised either against the first21 amino acids of N-terminus of mature cyt f or the last15 amino acids of the stromal C-terminus of cyt f,indicates the absence of proteolytic cleavage at eitherterminus. The differences in apparent molecular massof cyt f were also observed in the purified cyt b6/f

complexes from the two Chlamydomonas species (datanot shown).

In contrast, cyt b6 and the Rieske Fe–S proteins ofthe cyt b6/f complex did not show any differences inapparent molecular mass when both algal species werecompared (Fig. 2a). Thus, the differences in apparentmolecular mass appeared to be specific for cyt f of thecyt b6/f complex. However, as shown in Fig. 2b, thelower apparent molecular mass of the psychrophilic cyt frelative to the mesophilic form was not unique to thisthylakoid polypeptide. The apparent molecular mass ofthe large subunit of the soluble, stromal enzyme, Rubi-sco, of CP43, the core PSII antenna complex, as well asthe b-subunit of the chloroplast ATP synthase fromC. raudensis were consistently lower than those from C.reinhardtii (Fig. 2b).

To test whether the difference in apparent molecularmass is associated with psychrophily, we surveyed theapparent molecular mass of cyt f from 15 different spe-cies (Fig. 2c). Cyt f from Arabidopsis thaliana, Brassicanapus, and Psilotum sp. exhibited an apparent molecularmass comparable to that of C. raudensis whereas the

Fig. 2 a Immunoblots ofthylakoid membranepolypeptides from C. reinhardtii(C. reinh.) and C. raudensis(C. raud.) with specificantibodies against C-terminus(Cter), N-terminus (Nter), wholeprotein (W) of cyt f and wholeproteins of cyt b6 and RieskeFe–S (Fe–S) as well as the c-typeheme staining of cyt f using ECL(H). The calculated molecularweight for whole mature cyt fproteins is indicated undercorresponding protein lanes.b Immunoblots of total proteinextracts from C. reinhardtii(C. reinh.) and C. raudensis(C. raud.) with specificantibodies against Rubiscolarge subunit (Rbc LS), CP43protein and b-ATPase. Thecalculated molecular weight forwhole CP43 and partial Rbc LS(1-407 a.a.) proteins is indicatedunder immunoblot ofcorresponding protein.Molecular mass markers in kDshown on the left. All sampleswere loaded on SDS-PAGE onequal chlorophyll basis (5 lgper lane). c Immunoblots ofthylakoid membranepolypeptides from differentalgal, plant and cyanobacterialspecies with specific antibodiesagainst C-terminus of cyt f. Themolecular marker sizes (kD) areindicated on both sides of theblot. Samples were loaded onequal chlorophyll basis—10 lgof chlorophyll per lane

392

other 12 species tested exhibited apparent molecularmasses comparable to C. reinhardtii (Fig. 2c). SinceArabidopsis, Brassica and Psilotum are mesophilicplants, there is no correlation between the apparentmolecular mass of cyt f and psychrophily.

To determine if we could detect any correlation be-tween the amino acid sequence and the apparentmolecular mass of cyt f, we compared the amino acidsequence alignment of cyt f from C. raudensis with eightother known sequences from eukaryotic and prokaryoticphotosynthetic organisms (Fig. 2S). We were unable todetect any differences in amino acid sequence that werecorrelated with the lower apparent molecular mass ob-served for cyt f from C. raudensis, Arabidopsis thalianaand Brassica napus compared to that observed forC. reinhardtii, Chlorella vulgaris, spinach, wheat, pineand Synechococcus sp (Fig. 2S).

The structural stability of cyt f was assessed by mea-suring the thermal sensitivity of Fe-binding to the c-typeheme by staining cyt f with ECL in combination withimmunoblotting with the cyt f Cter antibody to determinethe relative abundance of the cyt f polypeptide (Fig. 3a,b). The results show that the heme in cyt f from C.raudensis was less stable to heating than that from C.reinhardtii (Fig. 3a, b). The calculated t1/2 for in vitro cytf heme-stability at 75�C was 6.7 and 15.7 min for C.raudensis and C. reinhardtii, respectively. Although thet1/2 was temperature dependent, C. raudensis cyt f heme-stability was consistently twofold to threefold lower thanthat of C. reinhardtii when measured at either 55 or 95�C(data not shown). Incubation temperature had minimaleffects on the relative abundance of cyt f indicating thatthe effect of incubation temperature was specific forheme-stability rather than polypeptide degradation overthe course of the experiment (Fig. 3a). To test of whetherthere is a correlation between the apparent molecularmass of cyt f and the thermostability of heme, we ana-lyzed the thermostability of the heme in cyt f from Ara-bidopsis thaliana, Brassica napus, Spinacia oleracea,Triticum aestivum and Zea mays. Although the thermo-stability of the heme in cyt f from Arabidopsis (34 kD)and spinach (41 kD) followed the same pattern as that ofC. raudensis (34 kD) and C. reinhardtii cyt f (41 kD),respectively, the thermostability of the heme of cyt f fromZea mays (41 kD) was comparable to that ofC. raudensiscyt f. The thermostability of the heme of cyt f fromBrassica napus and Triticum aestivum was intermediaterelative to that of the thermostability of cyt f from twoChlamydomonas species (Fig. 3c). Thus, it appears thatthere is no correlation between the apparent molecularmass of cyt f and the thermostability of heme.

Sequence analysis of petA

To test whether the difference in apparent molecularmass between cyt f from the psychrophile versus themesophile was due to possible differences in genestructure, petA from C. raudensis was cloned, sequenced

Fig. 3 Thermostability of c-type heme in cyt f at 75�C. aThylakoid membrane polypeptides from C. reinhardtii (C. reinh.)and C. raudensis (C. raud.) incubated at 75�C for 35 min with5 min intervals, were stained for c-type heme (lanes Heme) andimmunoblotted against C-terminus of cyt f (lanes cyt f). Sampleswere loaded on equal chlorophyll basis—15 lg of chlorophyll perlane. bKinetics for the heat-induced decrease in peroxidase activityof the c-type heme associated with cyt f proteins relative tounheated controls (n=4); black bars C. reinhardtii, grey bars C.raudensis. c Kinetics for the heat-induced decrease in peroxidaseactivity of the c-type heme associated with cyt f proteins fromdifferent plant and algal species. Open symbols correspond tospecies containing cyt f with apparent molecular mass comparableto C. raudensis, closed symbols correspond to species containing cytf with apparent molecular mass comparable to C. reinhardtii. Theperoxidase activity of heme bound to cyt f was measured at 5 minintervals during the incubation at 75�C (n=3)

393

and compared to that of C. reinhardtii. Analysis of theopen reading frame corresponding to the complete petAfrom C. raudensis revealed that the petA coding sequencewas identical in length to that of C. reinhardtii (954nucleotides) and encodes 317 amino acids (Fig. 1b). Thegreatest variability in amino acid sequence between cyt ffrom C. raudensis and C. reinhardtii (23% identity) wasin the N-terminal 31 amino acids of the thylakoid signalsequence (Fig. 1b). The 286 amino acids of the matureform of cyt f from C. raudensis and C. reinhardtii were79% identical with a predicted molecular mass of31.2 kD. In addition, the predicted protein charge andpI for both cyt f proteins are comparable (data notshown). One of the differences in amino acid sequencebetween the cyt f of C. raudensis and C. reinhardtii wasfound at position 292 in the transmembrane helix of cyt fwhich was occupied by a Cys in the psychrophile but aPhe residue in the mesophile (Fig. 1b). Thus, the psy-chrophile exhibited a third cysteine (Cys-292) near theC-terminus in addition to Cys-52 and Cys-55 residuesfound in the heme-binding motif near the N-terminus ofcyt f from C. raudensis and C. reinhardtii (Fig. 1b). Acomparison of the complete cyt f amino acid sequencesavailable in the database indicates that Cys-292 is absentnot only in C. reinhardtii but also in all cyt f proteinsinvestigated to date [Chlorella vulgaris (NC001865),Spinacia oleracea (AJ400848), Arabidopsis thaliana(AP000423), Brassica rapa (X77011), Nicotiana tabacum(NC001879), Vicia faba (X04920), Lotus japonicus(NC002694), Triticum aestivum (NC002762), Zea mays(NC001666), Oryza sativa (X15901), Picea abies(AJ001023), Pinus thunbergii (NC001631), Marchantiapolymorpha (NC001319), Cyanophora paradoxa cyanelle(NC001675), Synechocystis sp. (X58532), Synechococcussp. (NC005070), Porphyra purpurea (NC000925), Guil-lardia theta (NC000926)] (data not shown).

Expression in E. coli and site directed mutagensisof C. raudensis petA

As indicated above, the presence of Cys-292 in thetransmembrane helix of cyt f appeared to be unique toC. raudensis. To test whether this unique difference af-fects the apparent molecular mass of the psychrophilicform of cyt f, Cys-292 was converted to Phe-292 by site-directed mutagenesis of petA from C. raudensis and thealtered sequences transformed into E. coli as well as a C.reinhardtii DpetA mutant. The results illustrated inFig. 4a (lanes 2 and 3) indicate that the difference inapparent molecular mass of cyt f between C. raudensisand C. reinhardtii was still evident when petA was ex-pressed in E. coli. Furthermore, transformation of eitherE. coli (Fig. 4a, lanes 3 and 4) or C. reinhardtii (Fig. 4b,lanes 3 and 4) with C292F had no effect on the apparentmolecular mass of the psychrophilic form of cyt f. Thus,the presence of Cys-292 in cyt f from C. raudensis doesnot appear to be critical in determining its apparentmolecular mass.

Complementation of petA in C. reinhardtiiby chloroplast transformation

We hypothesized that the lower apparent molecularmass of cyt f from C. raudensis compared to C. rein-hardtii might be a consequence of species-specific, post-translational modification(s). To test this possibility, weintroduced C. raudensis petA sequences into C. rein-hardtii by chloroplast transformation to express cyt ffrom C. raudensis in a DpetA mutant of C. reinhardtii.To ensure proper chloroplast gene expression and pro-tein transport specificity to the thylakoid membranes,the C. reinhardtii cyt f promoter, terminator and signalsequence peptide was used in conjunction with thecoding sequence for cyt f from C. raudensis (Fig. 1a).Thus, only the mature C. raudensis cyt f protein waspresent in the transformed C. reinhardtii DpetA mutant,RApetA. All C. reinhardtii RApetA transformants ex-pressed a cyt f that was 7 kD smaller than that of C.reinhardtii WT and similar to the expected apparentmolecular mass of cyt f from C. raudensis (Fig. 5b). As acontrol, DpetA mutant was also transformed with thepetA from C. reinhardtii generating REpetA transfor-mants. All control lines exhibited an abundance, as as-sessed by the relative band intensity, and an apparentmolecular mass for cyt f that was similar to that of WTcells (Fig. 5a). In contrast, all five RApetA transfor-mants analyzed generally showed lower abundance ofcyt f than C. reinhardtii WT. Transformant line RA3exhibited the lowest level of cyt f, transformant linesRA4 and RA5 the highest abundance and transformantlines RA1 and RA2 exhibited intermediate abundance ofcyt f as assessed by the relative band intensity. A parallelpattern of abundance in the 5 RApetA transformant cell

Fig. 4 a Immunoblots of total protein extracts from E. coliexpressing mature wild type cyt f from C. reinhardtii (lane 2) andC. raudensis (lane 3), cyt f from C. raudensis with C292Fsubstitution (lane 4) and thylakoid membrane polypeptides fromC. reinhardtii and C. raudensis (lanes 5 and 6, respectively) withspecific antibodies against C-terminus of cyt f. Lane 1 E. colicontaining empty expression vector pSD80. b Immunoblots ofthylakoid membrane polypeptides from wild type C. raudensis (lane1) and C. reinhardtii (lane 2) cells and C. reinhardtii RApetAtransformants expressing wild type cyt f from C. raudensis (lane 3)and cyt f containing C292F substitution (lane 4) with specificantibodies against C-terminus of cyt f. Molecular weight markersare indicated on the left

394

lines analyzed was observed for cyt b6 and Rieske-FeS ofthe cyt b6/f complex as well as the PSI reaction centrepolypeptide, PsaA. The abundance of the PSII reactioncentre polypeptide, D1, the b-subunit of the chloroplastATPase as well as polypeptides associated with the lightharvesting complex of PSII (Lhcb2) and PSI (Lhca1)exhibited minimal variation between the 5 RApetAtransformant lines tested (Fig. 5b). These data are con-sistent with the fact that the RApetA transformant celllines exhibited minimal differences in total Chl contentper cell and Chl a/b ratios compared to C. reinhardtiiWT (Table 1S). However, the pseudo-first orderexponential growth rates of RApetA transformant lines,measured either as changes in A750 or as changes intotal Chl ml�1 of culture, were 23–46% lower than thatof C. reinhardtii WT (lmax=0.0834±0.0048 h�1). Light

saturated rates of gross O2 evolution were generally also30–40% lower in RApetA transformant lines comparedto C. reinhardtii WT (183±13 lmol O2 evolved mgChl�1 h�1) with minimal differences in the apparentquantum yield of O2 evolution (data not shown).

P700 and intersystem electron transport

The effect of the cyt f from C. raudensis on intersystemelectron transport in the C. reinhardtii RApetA trans-formants was assessed by DA820/A820 which is anestimate of the extent of P700 photo-oxidation (Asadaet al. 1993). Typical traces for wild type C. raudensis, C.reinhardtii and RApetA transformant (line RA1) areshown in Fig. 6a. When cells were illuminated with farred light, a rapid increase in DA820/A820 was observed,which indicates the oxidation of P700 to P700+. Thiswas followed by a slower re-reduction of P700+ to a newsteady-state level (Fig. 6a). The steady-state DA820/A820 in C. raudensis cells was significantly lower thanthat of C. reinhardtii (Fig. 6a, Table 1). Although ourmeasurements were performed at the optimal growthtemperatures for each species, the difference in thesteady-state DA820/A820 was observed regardless of thegrowth temperature for both species, which is consistentwith previous results (Morgan-Kiss et al. 2002a). Al-though the DA820/A820 in RApetA transformants was15–35% lower than that observed for C. reinhardtii WT,it was higher than that observed for C. raudensis WT(Fig. 6a, Table 1). This most likely reflects the fact thatRApetA transformants exhibited lower levels of PsaAthan C. reinhardtii WT (Fig. 5b).

Imposition of multiple turnover (MT) and singleturnover (ST) saturating flashes of white light induce atransient reduction of P700+ by PSII followed by the re-oxidation of P700 to P700+ by the background FR light.The MT flash induced a minimal, transient reduction ofP700+ and the effect of the ST flash was undetectable inC. raudensis cells (Fig. 6a). As expected, the MT flashinduced a complete, transient reduction of P700+ and asmaller, transient reduction of P700+ by the ST flash inC. reinhardtii WT.11 cells (Fig. 6a). Thus, intersystemelectron transport appeared to be impaired in C. raud-ensis compared to C. reinhardtii WT.11 (Fig. 6a). How-ever, the kinetics for the transient reduction of P700+

induced by the MT and ST flashes in RApetA transfor-mants resembled that of C. reinhardtiiWT.11 rather thanthat ofC. raudensisWT (Fig. 6a). We calculated the ratiobetween the areas of MT and ST which estimates thenumber of electrons in the intersystem electron transportchain per P700 (Asada et al. 1992). All three RApetAtransformants analyzed possessed lower MT/ST ratiosthan the wild type C. reinhardtii cells (Table 1), whichprobably reflects the lower abundance of cyt b6/f complex(Fig. 5b). Thus, the impairment of photosynthetic in-tersystem electron transport in C. raudensis can not be aconsequence of the presence of an altered form of cyt f aspreviously suggested (Morgan-Kiss et al. 2002a).

Fig. 5 a Immunoblots of thylakoid membrane polypeptides fromwild type C. reinhardtii cells (WT) and five independent REpetAtransformants (lanes 1–5) with specific antibodies against C-terminus of cyt f. b Immunoblots of thylakoid membranepolypeptides from wild type C. reinhardtii cells (WT) and fiveindependent RApetA transformants (lanes 1–5) with specificantibodies against cyt f, cyt b6, Rieske Fe–S protein, D1, PsaA,b-ATPase, Lhcb2 and Lhcb1 polypeptides. The thylakoid mem-brane samples were loaded on an equal chlorophyll basis (5 lg perlane)

395

State transitions

Previously, we reported that C. raudensis was locked instate I and was unable to regulate energy distributionbetween PSI and PSII (Morgan-Kiss et al. 2002a). Giventhe central role of the cyt b6/f complex in the regulationof state transitions and the phosphorylation of theLHCII polypeptides (Zito et al. 1999), we exploited theuse of the RApetA transformants to assess whether thepresence of an altered cyt f in C. raudensis could accountfor the inability of the psychrophile to undergo statetransitions as was suggested previously using 77 K flu-orescence emission spectroscopy (Morgan-Kiss et al.2002a). As shown in Fig. 6b, incubation of C. raudensiscells in the dark under anaerobic conditions causedminimal changes in the F682/F710 reflecting an inabilityto undergo a state I to state II transition. As expected,incubation of C. reinhardtii WT.11 cells under identical

conditions induced a state I to state II transition asindicated by a 25% decrease in the F682/F710 ratio from1.02 to 0.746 (Fig. 6b). As illustrated for transformantline RA1, three independent RApetA transformant linestested were able to undergo state I to state II transitionssimilar to that observed C. reinhardtii WT.11 (Fig. 6b).

The transition from state I to state II is regulated bythe phosphorylation of the LHCII polypeptides (Zitoet al. 1999). As expected, an increase in the phosphor-ylation of Lhcb2 was observed in C. reinhardtii WT cellsincubated in the dark under anaerobic conditions(Fig. 6c, C. reinh.). All three RApetA transformantstested also exhibited increased levels of Lhcb2 phos-phorylation (Fig. 6c, RApetA transformants). In con-trast, C. raudensis showed minimal phosphorylation ofLHCII polypeptides under conditions which induceeither state I or state II (Fig. 6c, C. raud.). Thus, theresults for state transitions and Lhcb2 phosphorylation

Fig. 6 a In vivo measurementsof the redox state of P700 inwild type C. raudensis,C. reinhardtii WT.11 cells andtransformant line RA1 usingthe change in the absorbance at820 nm. FR far red light; STsingle turnover; MT multipleturnover. b Low temperature(77 K) chlorophyll afluorescence of C. raudensis,C. reinhardtii WT.11 andtransformant line RA1. Solidlines correspond to the emissionspectra of dark adapted cells(state I); dotted lines correspondto cells incubated in the darkanaerobic conditions (state II).c Phosphorylation of thylakoidmembrane polypeptides duringstate transitions in wild typeC. reinhardtii (C. reinh.),C. raudensis (C. raud.) andthree RApetA transformants(1, 2, 3). The immunoblots ofthylakoid membranepolypeptides (10 lg chlorophyllper lane) from the cellsincubated in the dark (D) anddark anaerobic conditions (-O2)using antibodies againstPhosphothreonine (Zymed).Molecular mass standards (kD)are indicated on the right

396

in the RApetA transformants indicate that the presenceof an altered form of cyt f cannot account for theinhibition of state transitions in C. raudensis.

Discussion

We show for the first time, to our knowledge, that thepsychrophilic form of cyt f, an integral thylakoidmembrane protein, can complement cyt f in a mesophilicbackground using C. reinhardtii chloroplast transfor-mation. However, the RApetA transformants grewmore slowly and exhibited lower light saturated rates ofO2 evolution than C. reinhardtii WT. This is most likelydue to the fact that chloroplast transformation of C.reinhardtii with C. raudensis cyt f did cause a decrease inthe apparent abundance of cyt f, cyt b6 and the RieskeFe–S protein of the cyt b6/f complex, Lhca1 and PsaAassociated with PSI as well as the b-subunit of thechloroplast ATPase but not the D1 and Lhcb2 poly-peptides associated with PSII relative to C. reinhardtiiWT (Fig. 5b). Moreover, the lower abundance of the cytb6/f complex rather than the functional difference in cyt fresults in the observed differences in the redox state ofP700 in RApetA transformants (Fig. 6a, Table 1). Lowtemperature (77 K) fluorescence emission spectra andLHCII phosphorylation showed that the RApetAtransformants exhibited a capacity for state I to state IItransitions that was comparable to C. reinhardtii WT(Fig. 6b, c). This shows unequivocally that neither theimpairment of intersystem electron transport observedin C. raudensis (Fig. 6a) nor the inability to undergostate transitions are due to the difference in structure ofcyt f as previously suggested (Morgan-Kiss et al. 2002a).We suggest that the impairment of electron transport inC. raudensis may reflect a difference in the overallorganization of the photosynthetic electron transportchain as reflected, in part, by the low PSI/PSII ratio andlow Chl a/b ratios which limits the capacity of the psy-chrophile to acclimate to red light relative to C. rein-hardtii (Morgan-Kiss et al. 2005). The inability toundergo state transitions and phosphorylate LHCII maybe due to either the lack of or the repression of specificgenes required for state transitions (Lunde et al. 2000;Depege et al. 2003).

The sequencing of petA from C. raudensis revealedthat the unprocessed apocytochrome f consists of 317amino acids, which is exactly the same length as thatfrom C. reinhardtii. Thus, the lower apparent molecularmass of C. raudensis cyt f is not due to truncation ofpetA. Since the predicted N- and C-terminal amino acidsequences (Fig. 1S) exhibited 100% identity with theN- and C-terminal protein sequences of petA expressedin E. coli as determined by in-gel protein digestioncoupled with mass spectroscopy and MALDI-TOF (L.Gudynaite-Savitch, G. Lajoie, N. Huner, unpublished),the 7 kD difference in apparent molecular mass betweencyt f from C. raudensis and C. reinhardtii is unlikely to bethe result of differential post-translational protein

cleavage. Furthermore, the difference in the molecularmass of cyt f between C. raudensis and C. reinhardtii wasstill evident after expression in E. coli and in C. rein-hardtii. In fact, the apparent molecular mass of cyt ffrom C. raudensis (34 kD) is closer to the calculatedmolecular mass of 31 kD than that of C. reinhardtii(41 kD). A feature that makes the primary sequence ofcyt f from C. raudensis unique compared to all otherspecies examined in this study is the presence of oneadditional cysteine residue, Cys-292, in the transmem-brane helix near the C-terminus (Fig. 1b). Conversion ofCys-292 to Phe-292, in the C292F mutant generated bysite directed mutagenesis of C. raudensis petA failed toalter the apparent molecular mass of cyt f from C.raudensis. Thus, based on all of these data, we concludethat it is unlikely that the 7 kD difference in molecularmass is due to differences in either post-transcriptional,post-translational regulation or alterations in possi-ble lipid-protein interactions associated with specificdifferences thylakoid lipid and fatty acid composition(Morgan-Kiss et al. 2002b).

The primary sequence associated with heme-bindingis highly conserved in C. raudensis and C. reinhardtii(Fig. 1b). Despite this fact, the thermal stability of theheme in cyt f from the psychrophile was 2–3 times lowerthan that of cyt f from the mesophile (Fig. 3). Thissupports our contention that the difference in apparentmolecular mass between C. raudensis and C. reinhardtiiis due to some subtle but significant difference(s) inprotein conformation that must be associated with the21% difference in amino acid sequence. Lower apparentmolecular masses were also observed for a major solublechloroplastic protein (Rubisco), other integral thylakoidmembrane proteins such as CP43 and the b-subunit ofthe chloroplast ATP synthase (Fig. 2b) as well as themajor polypeptides of LHCII (Morgan et al. 1998;Morgan-Kiss et al. 2002b). This could lead one to sug-gest that such a trend might reflect the psychrophilicnature of this Antarctic alga characterized by increasedflexibility of cold-adapted proteins (Russell 2000).However, this is unlikely since our combined survey ofpetA sequences, thermal stability of the heme as well asthe apparent molecular masses of cyt f from a broadrange of mesophilic species shows unequivocally that theobserved differences in protein conformation are not

Table 1 In vivo measurements of redox state of P700 in C. rein-hardtii WT.11 and RApetA transformant (RA1, RA2, RA3) cellsgrown at 25�C/100 lmol photons/m2 s and C. raudensis 8�C/20 lmol photons/m2 s. e�/P700 was calculated as the area undermultiple turnover flash/area under single turnover flash

Cells DA820/A820 e�/P700

C. raudensis 25.45±2.16 NAC. reinhardtii 64.52±0.89 51.23±3.58RA1 41.58±0.67 13.61±0.62RA2 39.28±0.99 14.93±0.69RA3 41.64±0.51 19.94±0.76

NA not applicable (n=5)

397

related to psychrophily. Thus, caution must be exercisedin relating differences in protein sequence and thermalstability to adaptation to cold environments.

Acknowledgements We are grateful to Dr. F.-A. Wollman (Institutde Biologie Physico-Chimique, Paris, France) for the generous giftof C. reinhardtii DpetA mutant, the pWF plasmid and antibodiesagainst cyt b6 and Rieske proteins, Dr. J. E. Thompson (Dept. ofBiology, Univ. of Waterloo, Waterloo, ON, Canada) for antibodiesagainst C-terminus of cyt f, Dr. K. Redding (Univ. of Alabama,Tuscaloosa, USA) for antibodies to PsaA and Dr. Y. Choquet(Institut de Biologie Physico-Chimique, Paris, France) for advice inthe transformation strategy. LG-S was supported by and NSERCPostgraduate Fellowship. This work was supported by NSERCDiscovery grants to NPAH and SK.

References

Allen JF (1995) Thylakoid protein phosphorylation, state1-state 2transitions, and photosystem stoichiometry adjustments: redoxcontrol at multiple levels of gene expression. Physiol Plant93:196–205

Asada K, Heber U, Schreiber U (1992) Pool size of electrons thatcan be donated to P700+, as determined in intact leaves:Donation to P700+ from stromal components via the inter-system chain. Plant Cell Physiol 33:927–932

Asada K, Heber U, Schreiber U (1993) Electron flow to the in-tersystem chain from stromal components and cyclic electronflow in maize chloroplasts, as detected in intact leaves bymonitoring redox change of P700 and chlorophyll fluorescence.Plant Cell Physiol 34:39–50

Bailet B, Kohorn BD (1996) Hydrophobic core but not amino-terminal charged residues are required for translocation of anintegral thylakoid membrane protein in vivo. J Biol Chem271:18375–18378

Bennoun P (1982) Evidence for a respiratory chain in the chloro-plast. Proc Natl Acad Sci USA 79:4352–4356

Boynton JE, Gillham NW (1993) Chloroplast transformation inChlamydomonas. Methods in Enzymology 217:510–536

Depege N, Bellafiore S, Rochaix J-D (2003) Role of chloroplastprotein kinase Stt7 in LHCII phosphorylation and state tran-sition in Chlamydomonas. Science 299:1572–1575

Falk S, Sammuelsson G (1992) Recovery of photosynthesis andphotosystem II fluorescence in Chlamydomonas reinhardtii afterexposure to three levels of high light. Physiol Plant 85:61–68

Finazzi G, Zito F, Barbagallo RP, Wollman F-A (2001) Contrastedeffects of inhibitors of cyt b6/f complex on state transitions inChlamydomonas reinhardtii: the role of Qo site occupancy inLHCII kinase activation. J Biol Chem 276:9770–9774

Gray JC (1992) Cytochrome f: structure, function and biosynthesis.Photosynth Res 34:359–374

Harris EH (1989) The Chlamydomonas sourcebook. A compre-hensive guide to biology and laboratory use. Academic, SanDiego

Hauska G, Schlutz M, Butter M (1996) The cytochrome b6f com-plex-composition, structure and function. In: Ort DR, YocumCF (eds) Oxygenic photosynthesis: the light reactions. Kluwer,Dordrecht, pp 377–398

Ivanov AG, Morgan RM, Gray GR, Velitchkova MY, HunerNPA (1998) Temperature/light dependent development ofselective resistance to photoinhibition of photosystem I. FEBSLett 430:288–292

Ivanov AG, Sane PV, Zeinalov Y, Malmberg G, Gardestrom P,Huner NPA, Oquist G (2001) Photosynthetic electron transportadjustments in overwintering Scots pine (Pinus sylvestris L.).Planta 213:575–585

Laemmli UK (1970) Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227:680–685

Lebedev NN, Khatypov RA, Ladygin VG, Krasnovskii AA (1988)Fluorescence excitation spectra and decay kinetics of light-harvesting complexes in Chlamydomonas reinhardtii mutants.Photosynth Res 22:364–370

Lin S, Knox RS (1991) studies of excitation energy transfer withinthe green alga Chlamydomonas reinhardtii and its mutants at77 K. Photosynth Res 27:157–168

Lunde C, Jensen PE, Haldrup A, Knoetzel J, Scheller HV (2000)The PSI-H subunit of photosystem I is essential for statetransitions in plant photosynthesis. Nature 408:613–615

Morgan RM, Ivanov AG, Priscu JC, Maxwell DP, Huner NPA(1998) Structure and composition of the photochemical appa-ratus of the Antarctic green alga, Chlamydomonas subcaudata.Photosynth Res 56:303–314

Morgan-Kiss RM, Ivanov AG, Huner NPA (2002a) The Antarcticpsychrophile, Chlamydomonas subcaudata, is deficient in stateI–state II transitions. Planta 214:435–445

Morgan-Kiss RM, Ivanov AG, Williams J, Khan M, Huner NPA(2002b) Differential thermal effects on the energy distributionbetween photosystem II and photosystem I in thylakoid mem-branes of a psychrophilic and mesophilic alga. Biophys BiochimActa 1561:251–265

Morgan-Kiss RM, Ivanov IG, Pocock T, Krol M, Gudynaite-Savitch L, Huner NPA (2005) The Antarctic psychrophile,Chlamydomonas raudensis Ettl (UWO241) exhibits a limitedcapacity to photoacclimate to red light. J Phycol 41:791–800

Murray MG, Thompson WF (1980) Rapid isolation of highmolecular weight plant DNA. Nucleic Acids Res 8:4321–4325

Nichols HW, Bold HC (1965) Trichosarcina polymorpha. GenEt SpNov J Phycol 1:34–38

Pesaresi P, Lunde C, Jahns P, Tarantino D, Meurer J, Varotto C,Hirtz R-D, Soave C, Scheller HV, Salamini F, Leister D (2002)A stable LHCII-PSI aggregate and suppression of photosyn-thetic state transitions in the psae1-1 mutant of Arabidopsisthaliana. Planta 215:940–948

Ponamarev MV, Cramer WA (1998) Perturbation of the internalwater chain in cytochrome f of oxygenic photosynthesis: loss ofthe concerted reduction of cytochromes f and b6. Biochemistry37:17199–17208

Ponamarev MV, Schlarb BG, Howe CJ, Carrell CJ, Smith JL,Bendall DS, Cramer WA (2000) Tryptophan-Heme p-electro-static interactions in cytochrome f of oxygenic photosynthesis.Biochemistry 39:5971–5976

Rochaix J-D, Mayfield S, Goldschmidt-Clermont M, Erikson J(1988) Molecular biology of Chlamydomonas. In: Shaw CH(eds) Plant molecular biology. A practical approach. IRL Press,Oxford, pp 253–275

Russell NJ (2000) Towards a molecular understanding of coldactivity of enzymes from psychrophiles. Extermophiles 4:83–90

Smith SP, Barber KR, Dunn SD, Shaw GS (1996) Structuralinfluence of cation binding to recombinant human brain S100b:evidence fro calcium-induced exposure of a hydrophobic sur-face. Biochemistry 35:8805–8814

Vargas C, McEwan AG, Downie JA (1993) Detection of c-typecytochromes using enhanced chemiluminescence. Anal Biochem209:323–326

Wollman F-A, Delepalaire P (1984) Correlation between changesin light energy distribution and changes in thylakoid membranepolypeptide phosphorylation in Chlamydomonas reinhardtii. JCell Biol 98:1–7

Wollman F-A, Lamaire C (1988) Studies on kinase controlled statetransition in PSII and b6/f mutants from Chlamydomonasreinhardtii which lack quinine-binding proteins. Biochim Bio-phys Acta 933:85–94

Wollman F-A (2001) State transitions reveal the dynamics andflexibility of the photosynthetic apparatus. EMBO J 20:3623–3630

Zito F, Finazzi G, Delosme R, Nitschke W, Picot D, Wollman F-A(1999) The Qo site of cytochrome b6/f complexes controls theactivation of the LHCII kinase. EMBO J 18:2961–2969

398