Plant Molecular Biology 33: 79–86, 1997. 79c 1997 Kluwer Academic Publishers. Printed in Belgium.

Mutations of cytochrome b6 in Chlamydomonas reinhardtii disclose thefunctional significance for a proline to leucine conversion by petB editing inmaize and tobacco

Francesca Zito1, Richard Kuras1, Yves Choquet1, Hans Kossel2;y and F.-A. Wollman1;�

1UPR 9072, Institut de Biologie Physico-Chimique, 13 rue Pierre et Marie Curie, 75005 Paris, France (�authorfor correspondence); 2Institut fur Biologie III der Universitat Freiburg, Schanzlestrasse 1, D-7800, Freiburg,Germany

Received 28 May 1996; accepted in revised form 23 September 1996

Key words: Chlamydomonas, cytochrome b6f complexes, cytochrome b6, RNA editing

Abstract

We have introduced a proline codon in place of a leucine codon at position 204 of the petB gene of Chlamydomonasreinhardtii. This gene modification mimics the presence of proline codons at the same position in the petB genes ofmaize and tobacco, which are subsequently edited to leucine codons at the RNA level. Following transformation,we observed no editing at this position in C. reinhardtii, independent of the type of proline codon we have used:the CCA codon, edited in maize, or a CCT codon. Strains carrying the introduced mutation were non phototrophicand displayed a block in photosynthetic electron transfer, consistent with a lack of cytochrome b6f activity. Thusthe presence of a proline residue at position 204 in cytochrome b6 is detrimental to photosynthesis. We showthat the mutant phenotype arose from a defective assembly of cytochrome b6f complexes and not from alteredelectron transfer properties in the assembled protein complex. Biochemical comparison of the proline-containingtransformants with a cytochrome b6 mutant deficient in heme-attachment indicates that their primary defect is at thelevel of assembly of apocytochrome b6 with the bh heme, thereby preventing assembly of the whole cytochromeb6f complex.

Abbreviations: DCMU, 3-(3,4-dichlorophenyl)-1,1-dimethylurea; PAGE, polyacrylamide gel electrophoresis;TMBZ, 3,30,5,50-tetramethylbenzidine; RFLP, restriction fragment length polymorphism

Introduction

Transcript editing is less extensive in the chloroplaststhan in the mitochondria of a plant cell. Less thantwenty target sites for editing have been identified sofar on plastid transcripts and they all correspond tocytosine-to-uridine conversions.

The functional significance of these editing eventsis easily understood when a C-to-U conversion restoresan initiation codon AUG in place of an ACG codon.This is the case for the editing of rpl2 transcripts in

y Dr Kossel deceased before submission of this manuscript.His co-authors wish to acknowledge his remarkable contribution tothe field of RNA editing and to express their gratitude for havinginitiated this study.

maize [14] or of psbL transcripts in tobacco and spinach[16]. In other instances the editing sites are found with-in the coding sequence of the transcript, well down-stream from initiation codon,at positions where editingrestores the codon for a residue which is highly con-served among homologous proteins from other sources[20, 21, 28]. It has therefore been assumed that theseconserved residues are critical for the function of theprotein, a requirement which would explain the select-ive pressure for keeping, in some species, an edit-ing process at these specific sites. This assumptionhas been tested experimentally in the case of psbF,which encodes the � subunit of cytb559, a core com-ponent of the photosystem II oligomeric protein [2].The psbF transcript is edited at one site in spinach to

80

create a phenylalanine codon which is already presentat the gene level in tobacco. The unedited psbF genefrom tobacco was modified according to the pre-editedform of the spinach transcript, therefore bearing aserine instead of the usual phenylalanine codon [2].The resulting transplastomic plants, which expresseda non-edited transcript still bearing the serine, had ahigh fluorescence phenotype, indicative of some alter-ations in their photosynthetic properties [2]. Whetherthe mutant phenotype was due to an altered biogenesisof the multimeric psbF-containing protein complex orto altered electron transfer properties was not workedout.

Here we address the functional significance ofanother chloroplast editing process, involving the petBtranscript for cytochrome b6, one of the major sub-units of the cytb6f complex. This transcript is edited inmaize [9] and tobacco [13] in order to convert a prolinecodon into a leucine codon which is already present inall unedited transcripts for cytochrome b6 from othersources, including that from the green alga C. rein-hardtii [4]. The target leucine residue (L204) belongsto the D helix span which is involved in heme bind-ing through two histidines in position 187 and 202 [25,Kuras et al., unpublished data]. Therefore the chemicalnature of the residue at position 204 may be essentialfor the function of this helical domain within the cytb6f complex.

Editing of organellar transcripts is thought to berestricted to vascular plants, including both angio-sperms and gymnosperms [11]. It has not beenobserved in either bryophytes or chlorophytes [24].Thus we decided to replace the Leu-204 codon of thepetB gene from the chlorophyte C. reinhardtii with aproline codon. This can be easily achieved nowadaysthat the chloroplast genome of C. reinhardtii is trans-formable by homologous recombination. In addition,the functional consequences of this substitution couldbe rigorously assessed with this alga which has beenextensively used for biochemical and functional stud-ies of cytochrome b6f complexes [17, 18, 19, 22].To replace the resident TTA leucine codon, we chosetwo distinct proline codons: the CCA codon, whichis edited in maize, and the most frequently used CCTcodon in order (1) to assess that there was no latentediting activity at this site in green alga and (2) tounderstand the functional significance of the editing inmaize by a detailed analysis of a C. reinhardtii strainexpressing a mutated cytochrome b6 bearing a prolineinstead of a leucine in position 204.

Material and methods

Cell growth conditions

Wild type (mt+) derived from strain 137C and trans-formants were grown on Tris-acetate phosphate (TAP)medium, pH 7.2 at 25 �C under 300 lux continuousillumination.

Oligonucleotides, mutagenesis and plasmids

Plasmid pdWB, encompassing the cytochrome b6 cod-ing region has been previously described [17]. PlasmidpdWBA results from the introduction of the 1.9 kbEcoRV-SmaI fragment from plasmid pUC-atpX-AADcarrying the aadA cassette [10], into the unique NsiIsite of pdWB, that has been previously blunt-endedwith T4 polymerase. In that plasmid, the aadA cassetteis located 227 bp after the petB gene stop codon and istranscribed in the same direction as petB as illustratedin Fig. 1. The putative editing sites were introducedin the pdWB plasmid by site directed mutagenesis,according to the method of Kunkel et al. [15], usingthe following synthetic oligonucleotides:

Ed CCA: 50-CCTGTTTACGAATCATTGGGAA-ATGCATTAACATGAA-30

Ed CCT: 50-CCTGTTTACGAATCATAGGGAA-ATGCATTAACATGAA-30

where the nucleotides differing from the wild-typesequence are written in bold and a new restriction site(NsiI) generated by silent mutation is underlined. Thislead to plasmids pdB204P(cca) and pdB204P(cct).From these, the new plasmids pdBA204P(cca) andpdBA204P(cct) were obtained by ligation of the 3.6 kbBalI-AlwnI fragment from pdWBA with the 3.9 kbfragment derived from pdB204Ps upon digestion withthe same enzymes.

Chloroplast transformation in C. reinhardtii

Wild type and �petB strains [17] were transformedby tungsten particle bombardment according to Boyn-ton et al. [3] with a particle gun, operating undervacuum, built in the laboratory by P. Bennoun andD. Beal. Phototrophic transformants were selected onminimum medium at 3000 lux. Transformants con-taining the aadA cassette were selected for growthon TAP-spectinomycin-containingplates as previouslydescribed [17].

81

Figure 1. Schematic representation of the C. reinhardtii chloroplastgenome region encompassing the petB gene, cloned in the pdWBplasmid. The genes in that region are shown with an arrow indicatingthe direction of transcription. Also indicated on the figure are theNsiI site, where the aadA cassette has been inserted to produce thepdWBA plasmid, the position of the mutations we introduced in thepetB gene and the sequence location of the oligonucleotides used inthis work (? rev3; � cod0; � cod1).

Nucleic acids analyses

Cellular DNA was purified according to Roffey et al.[23] and total RNA was extracted using TRI reagent(Molecular Research Center, MRC). Total DNA wasamplified using the specific oligonucleotides CodO(50-AGCCTAATGGTCATGTCACA-30) and Rev3 (50-TCAATGGCCAACGCCTTGGA-30). Routine ampli-fication was performed with 40 cycles, using a cycleof denaturation for 1 min at 94 �C, annealing for 1 minat 55 �C and an extention for 2 min at 72 �C. Ampli-fication products were sequenced according to Changet al. [5] using the Sequenase kit (USB) and oligonuc-leotide Cod1 (50-ACCAAGTTGGTTACTGGGCG-30) as primer. Reverse transcription of total RNA wasperformed using the oligonucleotide Rev3 and cDNAwas then amplified and sequenced as described above.Specific probes for petA and petB mRNA, used innorthern experiments, were as described in Kuras andWollman [17].

Protein isolation, separation and analyses

Proteins from whole cells, grown at a density of 2�106

cells/ml, were resuspended in 100 mM DTT/Na2CO3

and solubilized in the presence of 2%SDS at 100 �Cfor 1 min. Polypeptides were separated by denaturingSDS-PAGE (8 M urea, 12–18% polyacrylamide gels).Protein analyses were performed according to Kurasand Wollman [17].

Pulse labelling experiment

Whole cells (2 � 106 cells/ml) were pulse-labeledaccording to Delepelaire [6] in the presence of cyc-

loheximide (8 �g/ml) to inhibit the cytoplasmic trans-lation. 14C acetate was added at the concentration of0.1 mM (5�Ci/ml) for 5 min and the pulse was stoppedby adding 10 vol of 50 mM ice-cold sodium acet-ate. Labeled polypeptides were separated as describedabove and detected by autoradiography of dried gels.

Fluorescence measurements

Fluorescence induction kinetics of dark adaptedalgae (4 � 106 cells/ml) were recorded on an inhouse built fluorimeter. Continuous illumination at605 nm was provided with electroluminescent diodesHLMA/CLOO. Fluorescence was measured at rightangles with photodiodes UV444B.

Results

Introduction of a proline codon at position 204 intothe chloroplast petB gene of C. reinhardtii

To allow direct selection of petB transformants andavoid the rather difficult recovery of homoplasmic,non phototrophic petB mutants in co-transformationexperiments, we decided to insert the aadA cassetteconferring spectinomycin resistance [10] at a neutralsite, within the pdWB plasmid, which encompasses thepetB-coding region. To this end we inserted the 1.9 kbEcoRI-SmaI fragment from plasmid pUC-atpX-AADinto the unic NsiI site of pdWB, 227 bp downstream ofthe petB stop codon, with the aadA being transcribedin the same direction as petB (Fig. 1). The resultingplasmid pdWBA was introduced in the chloroplast ofC. reinhardtii and conferred spectinomycin resistanceto the transformants. Their biochemical characteriza-tion showed that they had a wild-type phenotype and anunaltered expression of the pet genes (data not shown).

We converted the TTA codon for Leu-204 into twoproline codons, either CCA or CCT, the former beingthe edited codon in maize and the latter correspondingto the higher codon usage for proline in C. reinhardtiichloroplast DNA [12]. These sequence modificationswere introduced by site-directed mutagenesis in plas-mid pdWB [17], yielding plasmids pdB204P(CCA)and pdB204P(CCT). In order to follow easily the intro-duction of the mutations in the petB gene,a silent muta-tion was also introduced at position 202, generating anew NsiI restriction site used as a screenable RFLPmarker. The aadA selectable marker, was introducedin the two plasmids containing the mutation in pos-

82

ition 204, resulting in two new constructions namedpdBA204P(CCA) and pdBA204P(CCT) (see materialand methods).

The �petB strain, deleted for the petB gene [17],and the wild type were bombarded respectively withpdB204P and pdBA204P constructs. These two trans-formation strategies were used in parallel since theformer allows one to recover phototrophic transform-ants when the mutation preserves the function of cytb6

whereas the latter is used to recover non-phototrophictransformants based on their resistance to spectinomy-cin.

No phototrophic clones were recovered upon�petB transformation with pdB204P(CCA) orpdB204P(CCT). This strongly supported the viewthat substitution of a proline for Leu-204 in cytb6,was detrimental to photosynthesis. Consistent withthis assumption, the spectinomycin-resistant trans-formants, obtained from the wild type as a recipi-ent strain, displayed fluorescence induction kineticsafter dark adaptation which were typical of mutantswith impaired cyt b6/f activity [1]. These are shown inFig. 2 (top and bottom) in comparison with the wildtype induction curve (middle). The latter displays acomplex pattern of rise and decay phases before reach-ing a low stationary fluorescence level (Fstat) higherthan the initial fluorescence yield F0. Duysens andSweers [8] have shown that the fluorescence yield iscontrolled by the redox state of the primary quinoneacceptor of PSII: it is low in ‘open centres’, whenthis acceptor is in an oxidized state, because excitonsreaching the reaction centres are efficiently used forphotochemistry. In contrast, it is high in closed centres,when the primary quinone acceptor is reduced,becauseexcitons reaching the centres can no longer be used forphotochemistry and have an increased probability tobe re-emitted by fluorescence. Thus, in the presenceof DCMU, which inhibits electron transfer after theprimary quinone acceptor, the fluorescence yield risesrapidly to the Fmax level where all PSII centres are in aclosed state with a reduced primary acceptor. In CCAand CCT transformants (top and bottom of Fig. 2),the fluorescence induction curves without DCMU rosecontinuously up to a fluorescence level similar to theFmax level, reached in the presence of DCMU. There-fore electron transfer is interrupted in the transformantssince their stationary level of fluorescence is close toFmax. However, the shape of the fluorescence induc-tion curves with and without DCMU are widely differ-ent in the CCA and CCT transformations. This is dueto the larger reservoir of PSII electron acceptors avail-

Figure 2. Fluorescence induction curves of dark adapted cells fromwild-type or CCA and CCT transformant strains. In the presenceof 10 �M DCMU, added 60 s before illumination, the fluorescenceyield rises to the Fmax level, typical of closed PSII centres. !1, areabound by the fluorescence induction curve +DCMU and its Fmax

asymptote, corresponding to reduction of one acceptor per PSII; !2, additional area corresponding to reduction of plastoquinone pool.F0: initial fluorescence yield due to open PSII centres (oxidizedprimary quinone acceptor; Fmax, maximal fluorescence yield dueto closed PSII centres (reduced primary quinone acceptor); Fstat,stationary level of fluorescence, representative of the steady-statelevel of reduction of the primary quinone acceptor upon continuousillumination. Note that Fstat and Fmax are similar in CCA andCCT transformants, which is indicative of a block in phosyntheticelectron transfer.

able in the absence of the inhibitor. The area bound bythe fluorescence induction curve and the Fmax asymp-tote (defined on the top panel of Fig. 2) is a measureof the size of the acceptor pool available after PSII [7].Comparison of these area with DCMU (indicated by!1 in the top panel of Fig. 2) and without DCMU(indicated by !2 in the top panel of Fig. 2) showsthat it is more than ten times larger in the latter casethen in the former case where only the primary quinoneacceptor is available. Thus, in the CCA and CCT trans-formants, the fluorescence induction curves recordedwithout DCMU are indicative of a block in electrontransfer at a site located after the plastoquinone pool(about fourteen electron acceptors available), corres-ponding to the position of cytochrome b6f complexesin the electron transfer chain.

83

DNA and RNA analyses of the transformants

We sequenced the petB gene region around codon 204in two homoplasmic clones resulting from transform-ation of wild type C. reinhardtii with either the CCA-or CCT-containing plasmid, hereafter referred to asCCA and CCT transformants. This was achieved usingsuitable primers for PCR amplification of a 800 bpinternal petB fragment (see Fig. 1). It confirmed thepresence of the transforming proline codon at the petBgene level (results not shown). In order to check theabsence of editing at this site in C. reinhardtii, totalRNA, after being reverse-transcribed, was also sub-mitted to PCR amplification, using the same set ofprimers as for the genomic DNA amplification. TheRNA sequences were identical to the DNA sequences(not shown) therefore ruling out the possibility that alatent editing would have been activated by introducingthe CCA codon, which is edited in maize, in the petBgene of C. reinhardtii chloroplasts.

Figure 3 shows an autoradiogram of RNA filtershybridized with specific probes for petB and, as acontrol, petA genes. The CCA and CCT transform-ants (lanes 4 and 5) were compared with three controlstrains: the wild type (lane 1), a strain lacking thepetB gene (�petB, lane 2) and a cytb6 mutant strain(TB202Q, lane 3) bearing a His-202!Gln substitution,two residues upstream of the Leu-204!Pro substitu-tion we introduced in the CCA and CCT transformants.This latter strain fails to accumulate cytochrome b6

because it lacks one of the coordinating histidines forheme bh (Kuras et al. unpublished data). The 0.9 kbpetB transcript accumulated in all strains but in theone deleted for the petB gene (lane 2). Therefore ourstrategy to construct the CCA and CCT transformantsdid not lead to a decrease of petB gene expression atthe pre-translational steps.

Accumulation of the cytochrome b6f subunits

Figure 4A shows a urea/SDS-PAGE of whole-cell pro-tein extracts followed by heme staining with TMBZfor the same strains as in Fig. 3. In the wild type(lane 1), cytochrome b6 is seen as a broad band withan electrophoretic migration much higher than that ofcytochrome f which appears as a doublet. In strain�petB (lane 2), which lacks cytochrome b6 becauseof the deletion of the petB gene, the stained band cor-responding to cytochrome f drops to about 10% of thewild-type level, owing to an epistatic control of its rateof synthesis [27]. In the CCA and CCT transformants

Figure 3. Analysis of accumulation of petB and petA RNA in themutants of the cytochrome b6f complex. 10 �g of total RNA fromwild type and mutant strains were separated on a formaldeide agarosegel and hybridized with DNA probes of the chloroplast genes petBand petA encoding respectively cytb6 and cytf.�petB, strain deletedfor the petB gene; TB202Q, strain lacking the bh heme; CCA andCCT, proline-containing strains.

Figure 4. Immunoblots of whole cell protein extracts probed withantibodies against cytochrome f and b6 (A). A 1/5000 dilution wasused for detection with ECL method (Amersham). Heme-associatedperoxidase activity of denaturated subunits of cytochrome b6f com-plex in wild-type and mutants cells (B). Loads for each samplecorresponded to 20 �g chlorophyll. �petB, strain deleted for thepetB gene; TB202Q, strain lacking the bh heme; CCA and CCT,proline-containing strains.

(lanes 4 and 5), cytochrome b6 was not detectable byheme staining, as was the case of the heme attach-ment mutant TB202Q (lane 3). In addition these threemutants displayed the same 90% reduced staining ofcytochrome f as in the�petB strain.

The absence of a TMBZ-stained band in positionof cytochrome b6 in the CCA and CCT transformantscould be indicative of either an absence of cytochromeb6 polypeptide or of a block in heme binding to apo-cytochrome b6. We then probed these strains for thepresence of apocytochrome b6 using a specific anti-body (Fig. 4B). Whereas a broad band in position ofcytochrome b6 was immunodetected in the wild type(lane 1), a doublet was distinctly observed in the CCAand CCT transformants (lanes 4 and 5). Its labellingintensity remained far below that of the wild-type band.It should be noted that when less loaded by a factor of10, cytochrome b6 was still detected as a broad band in

84

Figure 5. Autoradiogram of chloroplast translates from pulse-labeled cells (one week exposure). Wild type (lane 5) and transform-ants (lanes 1, 2, 3 and 4) were pulse-labelled for 5 min with 14C-acetate in presence of cycloheximide to inhibit cytoplasmic trans-lation. Loads for each sample corresponded to 20 �g chlorophyll.�petB, strain deleted for the petB gene; TB202Q, strain lacking thebh heme; CCA and CCT, proline-containing strains.

the wild-type sample (results not shown). Interestingly,mutant TB202Q (lane 3) displayed a doublet similarto that in the proline-containing strains. In agreementwith the TMBZ staining, cytochrome f was immuno-detected at about 10% of its wild-type level, in thesecytb6 mutant strains.

That both cytochrome f and apocytochrome b6 werein reduced contents in the CCA and CCT transformantscould be due to either lower stability in the thykaloidmembranes or to their decreased rate of synthesis. Therate of synthesis of the chloroplast-encoded subunitscan be studied in vivo in C. reinhardtii by pulse labelingexperiments. The various strains were labeled with14C-acetate for 5 min in the presence of an inhibitorof cytoplasmic translation. Figure 5 shows an auto-radiogram after urea/SDS-PAGE of these whole-cellchloroplast translates. Whereas the rate of cytochromef synthesis dropped in the four cytochrome b6 mutants(lanes 1–4), that of cytochrome b6 remained in theCCA, CCT and TB202Q transformants (lanes 2–4). Itappears as a doublet, indicated by � in Fig. 4, consist-ing of a heavily labelled lower band and a faint upperband. Because the labelled cytochrome b6 bands weresharpened in these transformants (lanes 2–4), they evenlooked more highly labelled than in the wild type (lane5) in which cytochrome b6 displays the same charac-teristics as on immunoblots: it appears as a broad andfuzzy band, poorly resolved on the autoradiogram ofFig. 4. Here again the behavior of apocytochrome b6

in the CCA and CCT transformants appears strikinglysimilar to that in the heme attachment mutant TB202Q.

Discussion

Chloroplast and cyanobacterial cytochrome b6 arehomologous to the N-terminal domain of mitochon-drial and bacterial cytochrome b. This protein domainconsists of four transmembrane �-helices. Helices Band D contribute, each, two histidines whose pair-ing allows coordination of two b-hemes with differentredox potential [26]. Heme bh is coordinated by his-tidines 100 and 202, close to the stromal end of helicesB and D respectively. A leucine codon is consistentlyfound in position 204, two residues downstream ofHis-202, in all cytochrome b6 gene sequences exceptfor a few cases where a proline codon was found. Thisfeature was particularly intriguing because hydropathyplots predicted position 204 to be included in helix Dwhereas prolines are known to break the periodicity of�-helices. It was later shown that the proline codon,found in cytochrome b6 sequences from maize andtobacco, was edited at the RNA level to a codon for theubiquitous leucine residue [9, 13].

We chose to address experimentally the physiolo-gical need for editing at this site when a proline codonis improperly specified in the gene sequence, by usingchloroplast transformation of the chlorophyte C. rein-hardtii whose cytochrome b6 contains an appropriateTTA leucine codon at the gene level. We replaced thiscodon by a proline codon by site-directed mutagenesis.We used both the most commonly used proline codonin C. reinhardtii chloroplast, CCT, and the edited codonfrom maize, CCA, in case the latter would becomea target for editing when placed in C. reinhardtii.Although no editing event had been reported in bry-ophytes or chlorophytes, it occurred to us that theirediting capacities would have been obscured in casea highly active reverse transcriptase would be presentin their organelles. However, both types of proline-containing transformants carried the proline codon wehad introduced at the DNA as well as at the RNAlevels. Thus no latent editing could be detected by thisapproach in C. reinhardtii.

The CCA and CCT transformants were both nonphototrophic, thereby confirming the need for an edit-ing at this site in maize or tobacco in order to preserveactive photosynthesis. According to their fluorescenceproperties, the transformants behaved as bona fidecytochrome b6f mutants. Further biochemical analysisshowed that the Leu-to-Pro substitution led to assemblydefective mutants rather than electron transfer defect-ive mutants. This conclusion could be drawn from thepleiotropic decrease in the various cytochrome b6f sub-

85

units which we have previously observed in strainswhich lack one of the major subunit of the complex[16]. In particular, the CCA and CCT transformantsboth showed decreased rates of synthesis of cyto-chrome f, a phenomenon that we have attributed toits impaired assembly with the rest of the protein com-plex [27].

The proline-containing strains were strikingly sim-ilar to the TB202Q mutant which has impaired bhheme binding to cytochrome b6 due to the substitu-tion of a glutamine for one of the heme-coordinatinghistidine. The three strains had a low amount of cyto-chrome b6 polypeptides which could not be detectedby heme staining. Upon electrophoresis, the neosyn-thesized polypeptides migrated as sharp bands, yield-ing doublets upon immunodetection whereas wild typeneosynthesized cytochrome b6 migrates as a broad anddiffuse band due to its early association with the b-hemes. These common features allowed us to traceback the defect in assembly of the cytochrome b6fcomplex in the CCA and CCT transformants to a lackof assembly of apocytochrome b6 with one of its non-covalently bound heme.

Thus, the presence of a proline which is a helixbreaker, near the bh heme binding site, prevents suit-able folding of the apoprotein as had been suggestedinitially [9].

Acknowledgements

This work was supported by the CNRS/UPR9072 andby the European Union through contracts ERBCHRX-CT920045 and BIO2CT-930076. These included fel-lowship support to F.Z. The authors thank DominiqueDrapier for discussion and assistance during comple-tion of this work.

References

1. Bennoun P, Delepelaire P: Isolation of photosynthesis mutantsin Chlamydomonas. In: Edelman M, Hallick R, Chua N-H(eds) Methods in Chloroplast Molecular Biology, pp. 25–38.Elsevier, Amsterdam (1982).

2. Bock R, Kossel H, Maliga P: Introduction of a heterologousediting site into the tobacco plastid genome: the lack of RNAediting leads to a mutant phenotype. EMBO J 13: 4623–4628(1994).

3. Boynton JE, Gillham NW, Harris EH, Hosler JP, JohnsonAM, Jones AR, Randolph-Anderson BL, Robertson D, KleinTM, Shark KB, Sanford JC: Chloroplast transformation in

Chlamydomonas with high velocity microprojectiles. Science240: 1534–1538 (1988).

4. Buschlen S, Choquet Y, Kuras R, Wollman FA: Nucleotidesequences of the continuous and separated petA, petB and petDchloroplast genes in Chlamydomonas reinhardtii. FEBS Lett284: 257–262 (1991).

5. Chang C, Kwok SF, Bleecker AB, Meyerowitz EM: Arabidop-sis ethylene-response gene ETR1: similarity of product to two-component regulators. Science 262: 539 (1993).

6. Delepelaire P: Characterisation of additional thylakoid mem-brane polypeptides synthesized inside the chloroplast inChlamydomonas reinhardtii. Photochem Photobiophys 6:279–291 (1983).

7. Delosme R, Joliot P, Lavorel J: Sur la complementarite dela fluorescence et de l’emission d’oxigene pendant la perioded’induction de la photosynthese CRA Acad Sci Paris 249:1409–1412 (1959).

8. Duysens LNM, Sweers HE: Mechanism of two photochemicalreactions in algae as studies by means of fluorescence. In:Japanese Society of Plant Physiologists (ed.) Micro algae andPhotosynthesis Bacteria, pp. 352–372. University of TokyoPress, Tokyo (1963).

9. Freyer R, Hoch B, Neckermann K, Maier RM, Kossel H: RNAediting in maize chloroplasts is a processing step independentof splicing and cleavage to monocistronic mRNAs. Plant J 4:621–629 (1993).

10. Goldschmidt-Clermont M: Transgenic expression of amino-glycoside adenine transferase in the chloroplast: a selectablemarker of site-directed transformation of Chlamydomonas.Nucl Acids Res 19: 4083–4089 (1991).

11. Gray MW, Covello PS: RNA editing in plant mitochondria andchloroplasts. FASEB J 7: 64–71 (1993).

12. Harris EH: The Chlamydomonas Sourcebook: A ComprensiveGuide to Biology and Laboratory Use. Academic Press, NewYork (1989).

13. Hirose T, Wakasugi T, Sugiura M, Kossel H: RNA editing oftobacco petB mRNAs occurs both in chloroplasts and non- pho-tosynthetic proplastids. Plant Mol Biol 26: 509–513 (1994).

14. Hoch B, Maier RM, Appel K, Igloi GL, Kossel H: Editing ofchloroplast mRNA by creation of an initiation codon. Nature353: 178–180 (1991).

15. Kunkel TA, Roberts JD, Zakour RA: Rapid and efficientsite-specific mutagenesis without phenotypic selection. MethEnzymol 154: 367–382 (1987).

16. Kudla J, Igloi GL, Metzlaff M, Hagemann R, Kossel H: RNAediting in tobacco chloroplasts leads to the formation of atranslatable psbL mRNA by a C to U substitution within theinitiation codon. EMBO J 11: 1099–1103 (1992).

17. Kuras R, Wollman FA: The assembly of cytochrome b6/f com-plexes: an approach using genetic transformation of the greenalga Chlamydomonas reinhardtii. EMBO J 13: 1019–1027(1994).

18. Kuras R, Buschlen S, Wollman FA: Maturation of pre- apo-cytochrome f in vivo. A site-directed mutagenesis study inChlamydomonas reinhardtii. J Biol Chem 270: 27797–27803(1995).

19. Kuras R, Wollman FA, Joliot P: Conversion of cytochromef to a soluble form in vivo in Chlamydomonas reinhardtii.Biochemistry 34: 7468–7475 (1995).

20. Maier RM, Neckermann K, Hoch B, Akhmedov NB, KosselH: Identification of editing positions in the ndhB transcriptfrom maize chloroplasts reveals sequence similarities betweenediting sites of chloroplasts and plant mitochondria. Nucl AcidsRes 20: 6189–6194 (1992).

86

21. Maier RM, Hoch B, Zeltz P, Kossel H: Internal editing ofthe maize chloroplast ndhA transcript restores codons for con-served amino acids. Plant Cell 4: 609–616 (1992).

22. Pierre Y, Breyton C, Kramer D, Popot JL: Purification and char-acterization of the cytochrome b6 f complex from Chlamydo-monas reinhardtii. J Biol Chem 270: 29342–29349 (1995).

23. Roffey RA, Golbeck JH, Hille CR, Sayre RT: Photosyntheticelectron transport in genetically altered photosystem II reactioncenters of chloroplasts. Proc Natl Acad Sci USA 88: 9122–9126 (1991).

24. Scott J: A place in the world for RNA editing. Cell 81: 833–836(1995).

25. Szczepaniak A, Cramer WA: Topography of the chloroplastcytochrome b6 Curr Res Photosyn 3: 255–258 (1990).

26. Widger WR, Cramer WA, Herrmann RG, Trebst A: Sequencehomology and structural similarity between cytochrome b ofmitochondrial complex III and the chloroplast b6-f complex:position of the cytochrome b hemes in the membrane. ProcNatl Acad Sci USA 81: 674–678 (1984).

27. Wollman F-A, Kuras R, Choquet Y: Epistatic effects inthykaloid protein synthesis: the exemple of cytochrome f. In:Mathis P (ed) Photosynthesis From Light to Biosphere, vol. 3,pp. 737–742. Kluwer Academic Publishers, Dordrecht (1995).

28. Zeltz P, Hess WR, Neckermann K, Borner T, Kossel H: Editingof the chloroplast rpoB transcript is independent of chloroplasttranslation and shows different patterns in barley and maize.EMBO J 12: 4291–4296 (1993).