TCA1, a Single Nuclear-Encoded Translational Activator Specific for petA mRNA in Chlamydomonas reinhardtii Chloroplast

Corresponding author: J. Girard-Bascou, UPR/CNRS 1261, Institut de Biologie Physico-Chimique, 13 rue P. et M. Curie, 75005 Paris, France., girard{at}ibpc.fr (E-mail)

Communicating editor: P. J. PUKKILA

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We isolated seven allelic nuclear mutants of Chlamydomonas reinhardtii specifically blocked in the translation of cytochrome f, a major chloroplast-encoded subunit of the photosynthetic electron transport chain encoded by the petA gene. We recovered one chloroplast suppressor in which the coding region of petA was now expressed under the control of a duplicated 5' untranslated region from another open reading frame of presently unknown function. Since we also recovered 14 nuclear intragenic suppressors, we ended up with 21 alleles of a single nuclear gene we called TCA1 for translation of cytochrome b₆f complex petA mRNA. The high number of TCA1 alleles, together with the absence of genetic evidence for other nuclear loci controlling translation of the chloroplast petA gene, strongly suggests that TCA1 is the only trans-acting factor. We studied the assembly-dependent regulation of cytochrome f translation—known as the CES process—in TCA1-mutated contexts. In the presence of a leaky tca1 allele, we observed that the regulation of cytochrome f translation was now exerted within the limits of the restricted translational activation conferred by the altered version of TCA1 as predicted if TCA1 was the ternary effector involved in the CES process.

THE well-developed tools for genetic analysis in Chlamydomonas reinhardtii offer a unique opportunity to study the regulation of chloroplast gene expression in vivo and more specifically the control exerted by the nucleus on post-transcriptional steps such as mRNA maturation, stabilization, and translation. As discussed in WOLLMAN et al. 1999 , translation appears as the main regulatory step in the expression of organellar genes. This is well documented both in yeast mitochondria (reviewed in FOX 1996 ) and in chloroplasts from higher plants (GAMBLE and MULLET 1989 ; BARKAN et al. 1994 ; KIM et al. 1994 ; FISK et al. 1999 ; MCCORMAC and BARKAN 1999 ) or green algae (reviewed in ZERGES 2000 ). In C. reinhardtii, a number of nuclear mutants are specifically altered in the translation of a single organellar gene. Nuclear-encoded factors acting at the translational step were identified for atpA (DRAPIER et al. 1992 ), psaB (STAMPACCHIA et al. 1997 ), psbA (GIRARD-BASCOU et al. 1992 ; YOHN et al. 1998 ), psbC (ROCHAIX et al. 1989 ; ZERGES and ROCHAIX 1994 ; ZERGES et al. 1997 ), and psbD (KUCHKA et al. 1988 ). While the nuclear mutations affecting psbD translation may act at the level of elongation or stabilization of the nascent product (WU and KUCHKA 1995 ; RATTANACHAIKUNSOPON et al. 1999 ), all other nuclear factors are specific activators of translation acting on the 5' untranslated region (UTR) of their target mRNA. In most cases we still do not know whether these factors are merely constitutive of chloroplast gene expression or have a genuine regulatory function.

In C. reinhardtii, the rate of translation of several chloroplast-encoded polypeptides also depends on the presence of those polypeptides with which they ultimately assemble in an oligomeric protein (for reviews see WOLLMAN et al. 1999 ; CHOQUET and VALLON 2000 ). This process was termed "control by epistasy of synthesis" (CES) to account for the experimental observation that the synthesis of a CES subunit is markedly reduced in the absence of its assembly partners, viewed as "dominant" subunits from the same protein complex. To date, cytochrome f, a subunit of the cytochrome b₆f complex encoded by the chloroplast petA gene, is the best-characterized CES subunit (KURAS and WOLLMAN 1994 ; CHOQUET et al. 1998 ). At the molecular level, we have shown that the assembly-mediated control of cytochrome f synthesis is an autoregulation of translation initiation (CHOQUET et al. 1998 ). Still, the nature of the interaction between a regulatory motif in unassembled cytochrome f and the 5' UTR of the petA transcript is not known. Cytochrome f has no reported RNA-binding activity and displays no typical RNA-binding motif. Thus, the interaction is likely to be indirect. It would rely on the competitive binding of a translation activator to unassembled cytochrome f and to the petA-5' UTR. In this model, repression of cytochrome f synthesis, as observed in the absence of subunit IV (KURAS and WOLLMAN 1994 ), should result from a lack of translational activation.

The aim of this study is to dissect genetically the specific nuclear control of petA translation and to explore the links between the CES process and the translation of petA mRNA mediated by trans-acting factors of nuclear origin.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media, culture conditions, and strains:
Wild-type and mutant strains were grown on Tris-acetate-phosphate (TAP) medium, pH 7.2, at 25° under dim light (5–6 µE m^-2 sec^-1), unless otherwise indicated. To assay phototrophic growth, cells were streaked on minimal medium plates and allowed to grow under an illumination of 80 µE m^-2 sec^-1 for 10 days. Antibiotic resistance tests were performed as described in CHOQUET et al. 1998 on TAP plates supplemented with various concentrations of spectinomycin and streptomycin as indicated. Antibiotic concentrations were corrected for the percentage of impurity of the batches.

For genetic crosses and chloroplast transformation we used wild-type strains of C. reinhardtii from our laboratory that are derived from the original 137c strains. The nuclear mutant strains used in this study were mcd1-F16, mt- (DRAGER et al. 1998 ) and mca1-M11 (GIRARD-BASCOU et al. 1995 ; GUMPEL et al. 1995 ). The chloroplast mutants were the deletion strains petD, mt+ and petA, mt+ (KURAS and WOLLMAN 1994 ) and the mt+ chloroplast transformant KF303Q304St, where the first Lysine (K₃₀₃) of the stromal extension of cytochrome f is substituted by a Glutamine and immediately followed by a stop codon that truncates the protein by its last 14 residues.

Genetic analysis:
As a convention, all crosses are indicated with the mt+ parent first, i.e., the strain whose chloroplast genome is transmitted to the whole progeny. For gametogenesis, the cells were grown for 3–4 days on TAP plates containing one-tenth the usual amount of nitrogen. Mating, germination, and tetrad analysis were performed according to HARRIS 1989 . Germination of zygotes was controlled to be >75% unless otherwise indicated. Tetrad progeny were tested for a 2:2 segregation of mating types. For some experiments we pooled the meiotic products from all tetrads even if some were incomplete. Reversion tests and recombination analysis were performed as described in KURAS et al. 1997 , while complementation analysis was done according to GOLDSCHMIDT-CLERMONT et al. (1990).

Isolation of mutant strains:
Eight mutants deficient in cytochrome f synthesis are described in this work. Six of these, tca1-1, tca1-3, tca1-4, tca1-5, tca1-6, and mca1-792, were identified among a population of UV-mutagenized CC125 (mt+) cells (GIRARD-BASCOU et al. 1995 ; XIE et al. 1998 ). Two other mutant strains, tca1-2 and tca1-7, were screened out of a population of FdUrd-mutagenized wild-type cells from our laboratory. Mutagenesis using 5-fluorodeoxyuridine (FdUrd) was achieved at a concentration of 1 mM (WURTZ et al. 1979 ). UV irradiation was performed as in LI et al. 1996 , followed by an enrichment step in the presence of metronidazole according to BENNOUN and DELEPELAIRE 1982 . Cytochrome b₆f mutants were screened for their fluorescence yield as described in ZITO et al. 1997 , using a video-imaging system built in-house (BENNOUN and BEAL 1997 ).

Isolation of diploid strains:
Vegetative diploids were isolated according to published protocols (HARRIS 1989 ), using complementation between arg2 and arg7 mutations.

Isolation of revertant strains:
Revertants were isolated from tca1, mt+ strains using various mutagenic agents. UV mutagenesis was conducted as described in GIRARD et al. 1980 . EMS mutagenesis was performed with exponentially growing cells on pretreated cells that were grown for 4 days on TAP plates containing 0.1 mM FdUrd or on gametes. Cells were treated with 2% EMS for 30 min, washed three times, allowed to recover for 2 days in TAP liquid medium, and then selected in liquid minimal medium. Viability after mutagenesis, from 25 to 80% depending on batches, was estimated by counting the cells prior to and after treatment. FdUrd mutagenesis was performed on TAP plates containing 1 mM FdUrd (WURTZ et al. 1979 ). Typically, reversion became detectable after 3–5 weeks of culture in minimal medium under high light illumination (80 µE m^-2 sec^-1). Cells were subcloned and only one revertant clone per mutagenesis flask was retained.

Transformation experiments:
Cells were transformed by tungsten particle bombardment as previously described (KURAS and WOLLMAN 1994 ) with a helium particle gun built in-house by D. Béal, according to TAKAHASHI et al. 1991 . Phototrophic transformants were selected on minimum medium under high light (80 µE m^-2 sec^-1). Transformants containing the aadA cassette were selected on TAP-spectinomycin-containing plates (60 µg ml^-1) and subcloned on the same medium under dim light (5–6 µE m^-2 sec^-1) until they reached homoplasmy, as determined by DNA filter hybridization. At least three independent transformants were analyzed for each construct.

Nucleic acid manipulation:
Plasmids pWQ encompassing the wild-type petD gene, petB containing a deletion of cytochrome b₆-coding sequences (KURAS and WOLLMAN 1994 ), and pFKR12 (CHOQUET et al. 1998 ) were described previously. For Northern analysis, total RNAs were extracted from whole cells and analyzed as described in DRAPIER et al. 1998 , using petA and petD DNA probes described in BUSCHLEN et al. 1991 . The atpB probe is the 2.9-kb EcoRI-KpnI fragment of the Ba5 chloroplast DNA fragment (DRAPIER et al. 1992 ). Probe aadA was obtained by a NcoI-HindIII digestion of plasmid pUC-atpX-AAD (GOLDSCHMIDT-CLERMONT 1991 ). For Southern blots, the 2.3-kb HindIII probe was prepared from a HindIII digestion of the piAH1.9 plasmid (KURAS and WOLLMAN 1994 ). The petA-5' UTR probe was prepared by PCR, using plasmid pWF (KURAS and WOLLMAN 1994 ) as a template and oligonucleotides FT7Sac and PETAAUG as primers (probe B). Probe D was obtained by a HindIII-HinfI digestion of a PCR product obtained using oligonucleotides R1cod3 and PETArevA as primers and chloroplast DNA from the Su_C, tca1-2 strain as a template.

Chloroplast DNA manipulation:
Purified chloroplast DNA was isolated as described in CHOQUET et al. 1992 . For Southern blots, digestion products were separated on 0.7% TBE-agarose gels and transferred onto nylon membranes by capillarity. The 2-kb HindIII fragment of the Su_C, tca1-2 strain was recovered from a HindIII bank of the Su_C, tca1-2 chloroplast DNA cloned in the pUN121 vector (NILSSON et al. 1983 ), probed with the 2.3-kb HindIII wild-type probe (see Fig 6A). It was sequenced using primers designed in the pUN121 vector, pUN121codH and pUN121revH. The remaining reorganized DNA was amplified by PCR, using chloroplast DNA of the Su_C, tca1-2 strain as a template and oligonucleotides PETArevA and R1cod3 as primers.

View larger version (67K): [in this window] [in a new window] [Download PPT slide]   Figure 1. tca1 mutants are deficient in cytochrome f translation. (A) Chloroplast translates in the mutant strains tca1-1 and tca1-5 compared to those of the wild-type strain and of the deletion strain petA. Other tca1 mutants are indistinguishable from those two. (B) Accumulation of petA mRNA in wild-type and tca1 mutant strains. atpB mRNA accumulation is presented as a loading control. (C) Accumulation of cytochrome f and subunit IV polypeptides from cytochrome b6f complex in tca1 mutants and in wild type, detected using specific antibodies. OEE2 accumulation is presented as a loading control. tca1-6 and -7 had a similar phenotype as the tca1-1 and -2 representative stringent mutant strains.

View larger version (62K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Cytochrome f synthesized under the control of atpA-5' UTR no longer requires the wild-type TCA1 factor. (A) Maps of the petA gene in wild-type and AFFF strains (CHOQUET et al. 1998 ). Relevant restriction sites (B, BglII; N, NcoI) are indicated. The heavily hatched box in the 5' region of atpA denotes the sequence encoding the first 25 amino acids from the -subunit of the ATP synthase complex. (B) Top, newly synthesized cytochrome f detected by pulse-labeling experiments in wild-type parental strains AFFF and tca1-1 and in progeny of a representative tetrad from the cross AFFF, mt+ x tca1-1, mt-. Bottom, accumulation of cytochrome f, subunit IV, and OEE2 as a loading control, in the same strains, detected with specific antibodies.

View larger version (82K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Expression of the FKR12 reporter gene driven by the petA-5' UTR is dependent upon the wild-type TCA1 factor. (A) Maps of the R12 fragment in wild-type and FKR12 strains (CHOQUET et al. 1998 ). Positions of known genes and relevant restriction sites are indicated (S, StuI; R, EcoRI). (B) Identification of the tca1 members lacking cytochrome f accumulation in the progeny of a representative tetrad from the cross FKR12, mt+ x tca1-1, mt-. (C) Analysis of antibiotic resistance conducted on the same tetrad (Sp, spectinomycin; St, streptomycin). (D) Accumulation of the chimeric FKR12 mRNA in the four progeny of the tetrad determined with an aadA probe.

View larger version (54K): [in this window] [in a new window] [Download PPT slide]   Figure 4. Cytochrome f accumulation in revertant strains. Cytochrome f and subunit IV accumulation in the various revertant strains obtained either from tca1-1 or tca1-2 mutant strains was detected using specific antibodies. OEE2 accumulation is presented as a loading control. Estimated percentage of cytochrome f accumulation in each strain is presented at the bottom.

View larger version (46K): [in this window] [in a new window] [Download PPT slide]   Figure 5. The SuC chloroplast suppressor allows a TCA1-independent translation of petA. (A) Cytochrome f accumulation in the progeny of one representative tetrad from the cross SuC, tca1-2, mt+ x TCA1, mt- and in the parental strains. OEE2 is presented as a loading control. (B) TCA1 genotype was determined from the recovery or not of spectinomycinresistant transformants (+ for TCA1, - for tca1) after biolistic transformation from each meiotic product by the FKR12 chimeric gene. (C) petA mRNA accumulation in the same tetrad progeny and in the parental strains. petD mRNA accumulation is presented as a loading control.

View larger version (120K): [in this window] [in a new window] [Download PPT slide]   Figure 6. Molecular characterization of the suppression event in SuC, tca1-2. (A) Map of the WT Pst9 restriction fragment (HARRIS 1989 ). Relevant restriction sites HindIII (H), PstI (P), BglII (B), and features such as the petA gene, the Wendy DNA element, and ORF469 are indicated. BglII restriction fragments are numbered according to HARRIS 1989 . The petA region is widened to indicate the petA 5' UTR, signal sequence (stippled box), coding region, and 3' UTR. Probes A–C used for the Southern blots are indicated, as well as the wild-type hybridizing fragments (, ). (B) Southern blots on WT and SuC, tca1-2 chloroplast DNA. Purified chloroplast DNA was digested by HindIII or BglII, separated by agarose gel electrophoresis, and hybridized with probes A–D. Probes A–C are depicted in A and correspond, respectively, to the intragenic wild-type HindIII-AccI petA restriction fragment, to the wild-type petA-5' UTR, and to the upstream 2.3-kb HindIII fragment (BUSCHLEN et al. 1991 ). The main hybridizing fragments in the wild-type strain are indicated by open symbols (, ), whereas fragments specific to the SuC, tca1-2 strain are depicted with solid symbols (•, , ). The Bg11 and Bg13 fragments hybridizing to probe D are indicated. (C) Proposed mechanism for the reorganization of the Pst9 fragment in the SuC, tca1-2 strain. Relevant restriction sites are indicated: HindIII (H), XhoI (X), PstI (P), and BglII (B). Direction of transcription of petA gene and ORF469 is indicated with arrows. A region comprising part of Bg11 and -13 fragments was duplicated and inserted instead of the lost petA-5' UTR, as shown by the larger arrows. The enlargement shows the sequence breakpoints (||) of the petA-5' UTR. (D) Map of the resulting reorganized petA region in the SuC strain: The inserted region contains a tRNAVal, the 5' UTR (lightly hatched box), and the first 93 residues (darkly hatched box) of ORF469 fused to the last 19 amino acids of the petA transit peptide with the downstream coding sequence. The amino acid sequence of the fusion region is shown at the bottom with amino acids from ORF469 written in boldface type. The hybridizing fragments specific to the SuC strain (•, , ) and the HindIII-HinfI fragment corresponding to probe D used in Southern blots are indicated.

Oligonucleotides used for sequencing and/or PCR:

• pUN121codH: GGTTGAAGGTAATTCCATGACCG

• pUN121revH: GCTCAACAGCCTGCTCAGGGTCAA

• R1cod1: CGTTACAGGCATGAGCTAGTA

• R1cod2: GAATTTTAGTGGCAGTTGCCTCCT

• R1cod3: GTTACTACTTCCTCGAGACAGAACC

• R1rev1: CTGGTTAGAGTCTATCGAGCA

• FT7Sac: CGCGAGCTCAGATATAATATATGTTGAGAAG

• AAAAAAAATAAAATTTAAATAGT

• PETArevA: ACAGCTTGTGGTACTTCGATTTCAACTGCT

• PETAAUG: GCGGATCCATGGACATAATTTTATTAA

• TCTTAAAACG

Protein isolation, separation, and analysis:
Pulse-labeling experiments, protein isolation, separation, and analysis were carried out on cells grown to a density of 2 x 10⁶ cells ml^-1, according to KURAS and WOLLMAN 1994 .

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of nuclear mutants deficient in cytochrome f synthesis:
More than 80 mutants lacking cytochrome b₆f activity have been generated by either UV mutagenesis (XIE et al. 1998 ) or FdUrd treatment. They were identified as being deficient both in immunoreactive forms of cytochrome f and in cytochrome b₆f activity on the basis of their typical fluorescence induction kinetics (Cyt b₆f^- phenotype; BENNOUN and DELEPELAIRE 1982 ; ZITO et al. 1997 ). Twelve mutants were specifically affected in cytochrome f synthesis as illustrated in Fig 1A by tca1-1 and tca1-5 strains that showed no detectable cytochrome f in 5-min pulse-labeling experiments with [¹⁴C]acetate. Seven mutants, referred to as class I mutants and named tca1-1 to -7, still displayed petA mRNA accumulation, albeit at reduced levels, from 15 to 30% of the wild-type amount (Fig 1B). mca1-M11, a mutant strain previously described (GUMPEL et al. 1995 ), was totally deficient in petA mRNA accumulation as were four other mutants that we refer to as class II mutants. Cytochrome f accumulation was assayed in class I mutants by immunoblotting experiments (Fig 1C). Strains tca1-3, -4, and -5 accumulated 0.1, 1.6, and 0.2% of wild-type accumulation of cytochrome f, respectively [whereas strains tca1-1, -2, -6, and -7 showed no or quite undetectable amounts of cytochrome f (see tca1-1 and tca1-2 in Fig 1C)]. The other subunits of the cytochrome b₆f complex, such as subunit IV (Fig 1C) or cytochrome b₆ (not shown), accumulated only in trace amounts in the tca1 mutants, in agreement with previous reports (LEMAIRE et al. 1986 ; KURAS and WOLLMAN 1994 ).

Genetic analysis of tca1 mutants:
The seven class I mutants (Cyt b₆f^- phenotype) affected in cytochrome f synthesis were backcrossed with the wild-type strain (Cyt b₆f⁺ phenotype) to determine the inheritance of the mutant phenotype. All tetrads showed a 2:2 segregation for the Cyt b₆f^-:Cyt b₆f⁺ phenotypes, indicating that the Cyt b₆f^- phenotype was due to a single nuclear mutation. We tested the recessivity of the tca1-1 mutation by generating heterozygous vegetative diploids that exhibited the same fluorescence induction kinetics and phototrophic growth as did wild-type vegetative diploid cells. The seven tca1 mutants were analyzed in recombination and complementation tests (Table 1). We used a rapid recombination test to detect mutations at the same locus. The progeny of individual zygotes derived from pairwise crosses between all mutants, except tca1-5, were tested for phototrophy on minimal medium. None gave rise to a recombinant progeny capable of phototrophic growth. Thus, all genetic distances were below 2.4 cM, indicative of a tight linkage. By contrast, no linkage was detected in crosses with the five class II mutant strains since phototrophic recombinants were recovered at high frequency (see Table 1, where mca1-792 is taken as representative of class II mutants). Percentages of zygotes giving rise to wild-type progeny, which corresponds to the frequency of tetratype and nonparental ditype tetrads, were high indicating that the mutations probably affect two independent genes. In C. reinhardtii, complementation tests between photosynthetic mutants can be performed by testing the fluorescence induction kinetics of layers of young zygotes obtained from pairwise crosses (GOLDSCHMIDT-CLERMONT et al. 1990 ). In crosses between the seven tca1 mutants with the five class II mutants, young zygotes showed wild-type fluorescence induction kinetics, thus demonstrating altogether genetic complementation between the two mutant classes and recessivity of all mutations. In contrast, pairwise crosses between all class I mutants yielded young zygotes that had retained fluorescence induction kinetics of cytochrome b₆f^- mutants, indicating an absence of complementation. Thus, the seven class I nuclear mutations corresponded to recessive alleles of a single nuclear gene, which we called TCA1 (translation of cytochrome b₆f petA mRNA).

The 5' UTR of petA mRNA is the target of TCA1:
We investigated the putative role of the petA-5' UTR as a target for the translational control mediated by the TCA1 nuclear gene product. To this end, we analyzed the expression of two chloroplast chimeric genes, AFFF and FKR12, in TCA1 or tca1 nuclear contexts. AFFF is a chimeric gene where the regular 5' UTR of the petA gene has been substituted by the atpA-5' UTR that preserves the phototrophic property of the AFFF strain (CHOQUET et al. 1998 ; Fig 2A). FKR12 is a reporter gene driven by the petA-5' UTR that confers resistance to spectinomycin and streptomycin in a wild-type nuclear context (CHOQUET et al. 1998 ; Fig 3A). We crossed mating-type plus (mt+) strains bearing FKR12 or AFFF genes with a mating-type minus (mt-) tca1-1 mutant strain. In the two crosses, the chloroplast chimeric genes should be mainly transmitted uniparentally in tetrads while the nuclear mutation tca1 should be transmitted to only one-half of the progeny.

In AFFF, mt+ x tca1-1, mt- crosses, the whole progeny from 12 tetrads was phototrophic. As shown for one representative tetrad in Fig 2B, the rate of cytochrome f synthesis was similar in the four daughter cells. Accordingly, cytochrome f accumulated to the same extent in the four members of the tetrad (Fig 2B, bottom). No changes in petA mRNA levels were observed among the tetrad progeny (data not shown). Thus, translation of cytochrome f driven by the atpA-5' UTR is no longer dependent upon the wild-type allele of the TCA1 gene. This observation demonstrates that the mRNA target for TCA1 includes elements from the petA-5' UTR.

In FKR12, mt+ x tca1-1, mt- crosses, the two tca1 meiotic products were detected by their Cyt b₆f^- fluorescence phenotype that was confirmed by their deficiency in cytochrome f and their inability to grow on minimum medium. The other two TCA1 members of the tetrads showed a Cyt b₆f⁺ phenotype, accumulated cytochrome f, and were phototrophic (Fig 3B). In five representative tetrads, none of the tca1-1 progeny grew on antibiotic-supplemented medium (50 µg ml^-1 spectinomycin plus 4 µg ml^-1 streptomycin), in contrast to the TCA1 progeny that were antibiotic resistant (Fig 3C). We took the loss of antibiotic resistance in the tca1-1 context as indicative of a lack of translation of the FKR12 chimeric mRNA. The 5' UTR of the petA mRNA was thus sufficient to confer TCA1 sensitivity to a reporter gene. This observation demonstrates that the mRNA target for TCA1 is located within the petA-5' UTR.

Reversion strategy of tca1 mutants:
The molecular identification of TCA1 partners in the control of petA translation should be tractable by generating extragenic suppressor mutations either in the chloroplast or in the nuclear genome of a primary tca1 mutation. Therefore we undertook a search for chloroplast revertants, aimed at the characterization of cis-acting elements controlling translation within the petA-5' UTR, and a search for nuclear revertants to identify other possible nuclear gene products interacting with the TCA1 factor.

All tca1 mutants showed weak frequencies of spontaneous reversion, ranging from <10^-9 to 3.10^-7, as indicated in the diagonal of Table 1. A search for phototrophic revertants induced by mutagenesis was conducted most extensively with tca1-1, mt+ and tca1-2, mt+ strains. We used EMS and UV, which cause preferentially nuclear mutations, and FdUrd, which is known to reduce chloroplast polyploidy and induce chloroplast mutations preferentially (WURTZ et al. 1977 , WURTZ et al. 1979 ). From 49 treatments, 21 independent revertants were isolated on the basis of their restored growth in phototrophic conditions under an 80 µE m^-2 sec^-1 illumination (Table 2). As illustrated in Fig 4, revertant strains recovered from 11 to 89% of the wild-type level of cytochrome f and accumulated subunit IV accordingly. True reversions could be excluded since none of the revertants recovered wild-type levels of cytochrome f. Consequently the suppressed phenotype could be most often distinguished from a wild-type phenotype by fluorescence tests in further crosses. We noted during our mutagenesis experiments that suppressor mutations from the UV-induced tca1-1 mutant were most often found after UV treatments than after FdUrd treatments, while they were most often obtained after FdUrd treatments than after UV treatments from the FdUrd-induced tca1-2 mutant (Table 2). Thus, the molecular mechanisms at work on the nuclear genome in FdUrd mutagenesis are different from those of UV.

Genetic analysis of revertant strains:
To determine the genetic origin of the suppression events, each mt+ revertant strain was backcrossed with a mt- strain bearing the original tca1 mutation. The suppressed and Cyt b₆f^- phenotypes should segregate 2:2 if the suppression event is due to a single nuclear mutation whereas the suppressed phenotype should be transmitted to the whole progeny if it is due to a chloroplast mutation (Table 2). For 16 revertants, a 2:2 segregation of the suppressed and Cyt b₆f^- phenotypes was observed either in tetrads or statistically on batches of meiotic products. This was indicative of a single nuclear mutational event at the origin of the reversion. However, for some of these revertant strains, we noted that the level of cytochrome f varied over time as well as among the progeny after a cross. For example, cytochrome f accumulated to 50% of the wild-type level in the original tca1-1 r3, mt+ strain but dropped down to 15% in a mt- progeny from a tca1-1 r3, mt+ x tca1-1, mt- cross. From a further backcross of this tca1-1 r3, mt- strain with the original tca1-1, mt+ strain, six meiotic products with a suppressed phenotype showed variable cytochrome f accumulation ranging from 15 to 30% of the wild-type amount. These variations point to the possible instability of the TCA1 mutated factor whose steady-state concentration may then depend on the genetic background in each strain.

Fourteen of these 16 nuclear revertants due to monogenic nuclear suppression were crossed with the wild type to test the linkage of the suppressor and tca1 mutations. No meiotic progeny yielded recombinant clones of Cyt b₆f^- phenotype, indicating low genetic distances between forward and reverse mutations (Table 2). Thus, all these 14 nuclear suppressor mutations were tightly linked to the original tca1 mutations as one would expect for intragenic suppressors. We thus most likely obtained 14 new alleles of TCA1 but got no genetic evidence for additional TCA nuclear genes.

Four suppressors were not analyzed further because they failed to show either predicted monogenic Mendelian inheritance or chloroplast uniparental inheritance or they showed poor spore viability after meiosis.

In a backcross of the FdUrd-induced r1 revertant from tca1-2, mt+ strain with a tca1-2, mt- strain, the whole progeny from 10 tetrads exhibited the suppressed phenotype. We then used a mt- clone from the progeny to perform a symmetric backcross with a tca1-2, mt+ mutant strain. This cross yielded only clones of Cyt b₆f^- phenotype among the progeny from 6 tetrads. Thus, the suppressed phenotype was transmitted only by the mt+ parental strain, indicative of a chloroplast suppressor mutation.

Characterization of the chloroplast suppressor event in the r1 revertant:
The r1 revertant strain of tca1-2 strain contained both the original tca1-2 nuclear mutation and a chloroplast suppressor mutation that we called Su_C. It was crossed to the wild type to recover progeny carrying the suppressor mutation in a wild-type nuclear context. From two tetrads, the two TCA1 progeny were distinguished from the tca1-2 members by transformation with the FKR12 chimeric gene (Fig 5B): the chimeric construct conferred spectinomycin resistance to the TCA1 strains, while tca1-2 mutation prevented its expression (see Fig 3). However, cytochrome f accumulated to about the same amount in the whole tetrads (Fig 5A) and in the parental r1 revertant mt+ strain. In this experiment cytochrome f accumulated to 20% of the wild-type level while it accumulated to 50% in a first experiment (Fig 4; a similar situation was discussed above with tca1-1 r3 mutant strains). Thus, the chloroplast mutation Su_C allowed only a limited accumulation of cytochrome f, even in a wild-type nuclear context. Interestingly, the petA mRNA accumulating in the r1 revertant strain and in the four members of the tetrad was of larger size than the wild-type petA mRNA and its accumulation was reduced (Fig 5C).

To determine the specificity of the chloroplast suppressor Su_C, the Su_C, tca1-2, mt+ revertant strain was crossed with tca1-1, tca1-3, and mca1-792 mutant strains. The whole tetrad progeny from these crosses (10, 3, and 6 tetrads tested, respectively) displayed the suppressed phenotype in fluorescence tests (data not shown). Furthermore, cytochrome f accumulated to the same level as in the parental Su_C, tca1-2 strain in two tetrads analyzed (data not shown). Thus, the chloroplast Su_C mutation suppressed other tca1 or mca1 mutations. We then investigated whether the CES process mentioned in the Introduction still occurred in this chloroplast revertant strain. We introduced a deletion of the petB gene encoding cytochrome b₆ by biolistic transformation of the chloroplast genome from r1 revertant strain with plasmid petB (see MATERIALS AND METHODS). While cytochrome b₆ deletion mutants exhibit a 10-fold reduction of cytochrome f expression (KURAS and WOLLMAN 1994 ), the strain Su_C, petB, tca1-2 displayed no reduction of cytochrome f expression compared to the original Su_C, tca1-2 strain (data not shown).

Thus, the chloroplast suppression turned out to confer a TCA1-, MCA1-, and CES-independent expression of the petA gene. Since the 5' UTR of petA mRNA is the target of both factors and of the CES process, this suggested, together with the larger size of the petA mRNA accumulated in the Su_C strains, that the 5' UTR region was deeply altered in these strains.

Molecular characterization of the chloroplast suppression event:
To characterize the putative rearrangement in the petA region in Su_C, tca1-2 strains, HindIII and BglII digests of purified chloroplast DNA isolated from the mutant and wild-type strains were compared by Southern blot analysis. The 3.5-kb HindIII fragment (, Fig 6A), containing the coding sequence for mature cytochrome f and the downstream regions detected with probe A, remained unaffected, as expected from the accumulation of a functional cytochrome f (Fig 6B, lane A). Conversely, probe B, corresponding to the petA-5' UTR, clearly showed the absence of any hybridizing fragment in the Su_C, tca1-2 strain (Fig 6B, lane B). Thus, the petA-5' UTR, the target of TCA1 and MCA1 factors and of the CES process, has been lost as a result of the mutation that leads to the suppressed phenotype. This was confirmed by hybridization with probe C, corresponding to the HindIII fragment upstream of the cytochrome f coding sequence (Fig 6A). This 2.3-kb fragment () was not detected any longer in the Su_C, tca1-2 strain, which contained instead two new hybridizing fragments of 2.0 () and 0.5 kb (; Fig 6B, lane C). The 2.0-kb HindIII fragment was subcloned into pUN121 and sequenced. The characterization of the rearrangement was completed by sequencing a PCR product amplified using oligonucleotides primers derived from the sequence of the cloned fragment and from the petA coding region (see MATERIALS AND METHODS). In the Su_C, tca1-2 strain, the deleted region of the petA-5' UTR is replaced by 1.1 kb of unidentified sequences. To address the origin of this sequence, Southern blots of chloroplast BglII digestion products were probed with a HindIII-HinfI fragment internal to the inserted region (probe D, see Fig 6D). As depicted in Fig 6B, lane D, this probe hybridized to the BglII fragments Bg13 and Bg11 in the wild-type and Su_C, tca1-2 strains. This assignment was consistent with other features of the sequence of the insert such as the presence of a tRNA_val and of XhoI and BglII restriction sites (HARRIS 1989 ). A final confirmation of the identity between the sequence of this fragment and that of the Bg13-Bg11 region came from the sequence data kindly communicated by D. Stern and J. Maul in the frame of the Chlamydomonas chloroplast sequencing project. In the Su_C, tca1-2 strain, probe D strongly hybridized to a new band of 2.6 kb (•). The origin of this new band is shown diagrammatically in Fig 6C. Two other faint hybridizing fragments of 2- and 1-kb size are of unknown origin. As confirmed by Southern blots using other restriction enzymes (data not shown), the Bg11 and Bg13 fragments remained unaffected in the Su_C, tca1-2 strain compared to the wild-type strain. Thus, the rearrangement results from a duplication of this region.

This rearrangement leads to a chimeric gene encoding a new protein made of the cytochrome f sequence, including 19 of the 31 residues from its transit peptide fused in frame with 93 amino acids corresponding to the N-terminal part of a 469-amino-acid open reading frame (ORF469) present in the Bg13-Bg11 region (Fig 6C). The fusion preserves the 7-amino-acid hydrophobic core of the cytochrome f transit peptide described by SMITH and KOHORN 1994 and the AQA target sequence for the lumenal peptidase (BUSCHLEN et al. 1991 ; KURAS et al. 1995A ; BAYMANN et al. 1999 ; Fig 6D). Thus, the mature product from the chimeric gene is identical to wild-type cytochrome f (Fig 5A). This chimeric petA gene is now under the control of cis-acting sequences normally required for the expression of ORF469, suggesting that this putative ORF is expressed. The modification of the 5' UTR of the petA gene explains the accumulation to 5% of the wild-type level of a transcript of higher molecular weight as depicted in Fig 5C.

Study of the CES process in strains with partially restored TCA1 activity:
As mentioned in the Introduction, the CES process for cytochrome f is likely to operate through a ternary effector that would interact with the petA-5' UTR. This effector should modulate the translational efficiency of petA mRNA, depending on its binding to unassembled cytochrome f. The above experiments indicated that the TCA1 factor and the CES process both involved as a target the petA-5' UTR. We wondered whether the partially restored translation of cytochrome f in a nuclear tca1-suppressed strain, which should express a mutated version of the TCA1 factor, still showed sensitivity to the CES process. To this end we studied the CES process in a revertant strain carrying a leaky TCA1 allele, tca1-1 r3, mt-, which accumulated 15% of wild-type cytochrome f (Fig 7A and Fig 8A). We investigated the expression of the petA gene in tetrads obtained from crosses between tca1-1 r3, mt- and two types of chloroplast transformants in which cytochrome f translation was either repressed or enhanced.

View larger version (86K): [in this window] [in a new window] [Download PPT slide]   Figure 7. Cytochrome f accumulation in the progeny of a representative tetrad from the cross petD, mt+ x tca1-1 r3, mt-. (A) Cytochrome f and subunit IV accumulations, detected using specific antibodies, are compared in the wild-type and parental strains and in a representative tetrad progeny from the cross petD, mt+ x tca1-1 r3, mt-. OEE2 accumulation is presented as a loading control. (B) Accumulation of cytochrome f and subunit IV was similarly monitored in strains obtained by biolistic transformation of the strains depicted in A with the wild-type petD gene. OEE2 accumulation is presented as a loading control.

View larger version (43K): [in this window] [in a new window] [Download PPT slide]   Figure 8. Cytochrome f accumulation in the progeny of a representative tetrad from the cross KF303Q304St, mt+ x tca1-1 r3, mt-. Accumulation of cytochrome f and OEE2 (as a loading control) is detected by specific antibodies in the wild type, in the parental strains KF303Q304St and tca1-1 r3, and in the progeny of the cross KF303Q304St, mt+ x tca1-1 r3, mt-.

The first type of cross involved the petD, mt+ strain and consequently displays a cytochrome f synthesis reduced by 90% as illustrated in Fig 7A. We analyzed four tetrads from a petD, mt+ x tca1-1 r3, mt- cross. The whole progeny inherited the petD deletion and lacked subunit IV, as illustrated in Fig 7A. We observed a Mendelian segregation of cytochrome f accumulation: two members of the tetrads accumulated 15% of wild-type cytochrome f, while this accumulation was reduced to 3% in the other two members of the tetrad (see Fig 7A). We observed changes in the petA mRNA levels among these various strains (data not shown). However, as previously reported (SAKAMOTO et al. 1994 ; CHOQUET et al. 1998 ) we found no direct quantitative relation between petA mRNA levels and the rates of cytochrome f synthesis. To determine the genotype of that tetrad progeny, the wild-type petD gene was reintroduced by biolistic transformation using plasmid pWQ in each member of the representative tetrad presented in Fig 7A. Accumulation of cytochrome f in the resulting transformant strains is shown in Fig 7B. The third and fourth members of the tetrad had the wild-type TCA1 allele since they exhibited wild-type levels of cytochrome f after transformation, while the first and second members had inherited the partially active tca1-1 r3 allele since their transformants accumulated cytochrome f to the same level as the parental strain tca1-1 r3, mt-. Comparison of the state of cytochrome f accumulation in the first and second members of the tetrad before and after reintroduction of the petD gene expressing subunit IV shows that the CES process still occurred in cells bearing the leaky tca1-1 r3 allele, with a repressed expression of cytochrome f in the absence of subunit IV.

For the second type of cross, we used the KF303-Q304St, mt+ strain that expresses a mutated version of cytochrome f deleted for the last 14 residues of the C-terminal domain. As shown in Fig 8, this strain behaves similarly to the FK₂₈₃St strain lacking the whole C-terminal domain (KURAS et al. 1995B ): the mutated cytochrome f is overexpressed threefold, irrespective of the presence or absence of its assembly partners. Two tetrads were analyzed from the cross KF303Q304St, mt+ x tca1-1 r3, mt-. All progeny displayed the mutated version of cytochrome f as evidenced by its faster electrophoretic migration pattern upon SDS-PAGE. Two members of the tetrads overexpressed cytochrome f to the same extent as the parental KF303Q304St, mt+ strain (Fig 8, members 2 and 3 of a representative tetrad). The other two members of the progeny accumulated the mutated version of cytochrome f to the same level as did wild-type cytochrome f in the parental tca1-1 r3, mt- strain (Fig 8, members 1 and 4). Thus, clones 2 and 3 had a wild-type version of TCA1 while clones 1 and 4 bore the mutated TCA1 factor. According to the molecular model we have proposed for the CES process (CHOQUET et al. 1998 ), TCA1 (either in its wild-type or mutated form) is no longer trapped by unassembled cytochrome f since the regulatory motif carried by the C-terminal domain involved in this mechanism is deleted in the KF303Q304St, mt+ strain. The wild-type or mutated versions of TCA1 are therefore entirely available to promote petA mRNA translation, but the reduced activity of the mutated TCA1 factor turned out to be limiting in the expression of cytochrome f.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

TCA1 is a nuclear-encoded activator required for the initiation of translation of petA mRNA:
We performed a genetic identification of a nuclear factor specifically required for the expression of the chloroplast petA gene encoding cytochrome f, a major subunit of cytochrome b₆f complexes from the thylakoid membranes. Mutations in the TCA1 gene still allow a reduced but significant accumulation of the petA mRNA but impair the synthesis of cytochrome f. The lower stability of the untranslated petA mRNA in a tca1 context correlates with a block at the stage of translation initiation since we showed, using chimeric genes, that the target for TCA1 is located entirely in the 5' UTR region of the petA transcript. As the seven TCA1 mutations presented in this study were recessive, we conclude that the TCA1 product acts as a specific translational activator and not as a translational repressor. The reason for the destabilization of petA mRNA in tca1 mutants is unclear. In a tca1 mutant nuclear context, the accumulation of the petA messenger is reduced (Fig 1B) but not that of the chimeric FKR transcript (Fig 3D). A comparative analysis of several other studies shows that there is no simple relationship between translation and stability of chloroplast mRNAs in Chlamydomonas. The tda1-F54 nuclear mutant strain, impaired in the translation of the -subunit of the ATP synthase complex, showed a threefold increase of the level of atpA mRNA (DRAPIER et al. 1992 ), while the tbc1-F34 and tbc2-F64 mutants, lacking translation initiation of psbC mRNA, showed no changes at the transcript level (ROCHAIX et al. 1989 ). From the analysis of translational defects for the psaB chloroplast mRNA it is tempting to suggest that a translational block after initiation may protect mRNAs from degradation whereas impaired initiation would compromise mRNA steady-state accumulation: indeed, there was a fivefold reduction in the tab1-F15 mutant specifically defective in translation initiation (STAMPACCHIA et al. 1997 ). In contrast premature arrest of translation of the same psaB mRNA, because of a frameshift or because of addition of lincomycin, caused an increase in the transcript level (XU et al. 1993 ). However, in the absence of initiation of translation, the lifetime of a chloroplast mRNA certainly depends on determinants downstream of the 5' UTR since, in contrast to the resident psaB mRNA, a chimeric 5'psaB-aadA-3'rbcL mRNA that contains the target sequence of the TAB1 factor accumulated to similar levels in tab1 and TAB1 genetic nuclear contexts (STAMPACCHIA et al. 1997 ).

Are there other TCA factors?
The number of nuclear genes required for the achievement of a given step in the expression of one chloroplast gene in C. reinhardtii is probably higher than estimated from our current knowledge. In most cases, these nuclear genes were identified by a single mutant allele, indicating that we are far from genetic saturation and that other genes remain to be discovered. The situation is widely different for cytochrome f. Although we used several mutagenic agents to carry out a large-scale screening of cytochrome b₆f-deficient mutants, we identified only a single nuclear gene, TCA1, controlling cytochrome f translation, out of seven independent mutants deficient in cytochrome f translation. The five other nuclear mutations responsible for a deficiency of cytochrome f synthesis lacked mature petA mRNA and were all located in another unlinked gene called MCA1, identified for the first time by the M11 mutation (GUMPEL et al. 1995 ; J. GIRARD-BASCOU and Y. CHOQUET, unpublished results). Thus, the 12 selected nuclear mutants deficient in cytochrome f synthesis presented in this study were mutated either in TCA1 or MCA1 genes. Moreover, the 14 nuclear tca1 suppressors that we could further analyze in crosses were most likely intragenic suppressors exhibiting a broad range of cytochrome f accumulation. This extensive search for genes involved in cytochrome f synthesis suggests that there are indeed only two genes, one controlling specifically the stability of petA mRNA (MCA1 gene) and the other its translation (TCA1 gene).

In yeast mitochondria, where extensive screens have been used, a limited number of specific translational activators have been characterized for each mitochondrial gene (for a review, see FOX 1996 ). For instance, translation of COX3 mRNA requires three nuclear genes (PET54, PET122, and PET494) whose products form a complex. Only a single nuclear gene (PET111) was found to be required for COX2 mRNA translation, the products of two nuclear genes (CBS1 and CBS2) specifically activate translation of the COB mRNA, and one gene (PET309) is involved in COX1 mRNA translation.

In maize, a nuclear gene, Crp1, has also been shown to affect cytochrome f translation (BARKAN et al. 1994 ; FISK et al. 1999 ). However, the phenotypes of the crp1 mutant and the tca1 mutant differ in their specificity. Besides its role in cytochrome f translation, Crp1 is also involved in the processing of the dicistronic petB-petD transcript and translation of petD mRNA. The Crp1 gene, cloned by transposon tagging, encodes a large soluble protein not associated with ribosomes, which is a component of a multisubunit complex in the chloroplast stroma (FISK et al. 1999 ). Crp1 homologs have been found in Neurospora crassa (cya5), Saccharomyces cerevisiae (PET309; FISK et al. 1999 ), and recently in C. reinhardtii (Tbc2; for a review, see ZERGES 2000 ). Tbc2 is specifically required for translation of the chloroplast psbC mRNA, which encodes a subunit of PSII complex. Thus Crp1 and TCA1 genes are probably not related to one another.

Chloroplast suppression:
Only one chloroplast suppression event was obtained after several FdUrd treatments of tca1 mutant strains (Table 2). Furthermore, while chloroplast point mutations induced with FdUdr have been found on the 5' UTR of psbC or psaB (STAMPACCHIA et al. 1997 ; ZERGES et al. 1997 ), the suppression event obtained in our study corresponds to an extensive chloroplast DNA rearrangement that led to the replacement of the whole 5' UTR of petA by that of another gene, so that the petA gene expression in this strain is now independent of TCA1 and MCA1 and of the CES process. The rearrangement does not result from a reciprocal event as in a chloroplast revertant of a C. reinhardtii strain carrying a deletion in the 5' UTR of the petD gene (STURM et al. 1994 ; HIGGS et al. 1998 ). Here, the petA-5' UTR has been deleted from the chloroplast genome of the revertant strain, and a duplication of a 1.1-kb sequence located on the other side of the wendy transposon has been inserted immediately upstream of the petA coding region. The Wendy DNA element (see Fig 6A) is bordered by inverted repeats and several additional degenerate copies of repeated sequences in direct or inverted orientation (FAN et al. 1995 ), which may have played a role in the mutation event. However, no sequence homology has been detected at the breakpoints of the rearrangement, but illegitimate transpositional recombination without duplication of the element has been previously reported (FAN et al. 1995 ).

The rearrangement led to the disruption of the 31-amino-acid transit peptide of the apocytochrome f by deleting the first 12 amino acids but preserved the hydrophobic core required for the translocation of the protein (SMITH and KOHORN 1994 ). Even though the molecular characterization of the suppression event gave no further insights about the target of TCA1 within the petA 5' UTR, it led to the identification of a novel ORF of 469 amino acids, which is likely expressed since the chimeric gene resulting from the fusion between this ORF and the 5'-truncated petA gene is translated. This ORF being unaltered in the Su_C strain, we obtained no indication about its possible function. However, the C-terminal half of this ORF shares homologies (35–45% identity, 40–60% similarity) with the first 200 amino acids of chloroplast rpoC1 or bacterial rpoC gene products. This could suggest that this ORF is essential for cell viability and explain why the suppressing event involved a duplication rather than a reciprocal event. Further characterization of this gene product is now in progress.

TCA1, a candidate to act as the ternary effector involved in the CES process that controls cytochrome f translation:
The CES process that controls cytochrome f synthesis is an assembly-dependent autoregulation of translation that involves a regulatory motif carried by the C-terminal domain of the unassembled protein and the 5' UTR of the petA mRNA (CHOQUET et al. 1998 ). The CES process appears as a general regulation mechanism in the biogenesis of photosynthetic proteins in C. reinhardtii chloroplast, with at least one CES subunit present in each oligomeric protein of the thylakoid membrane: the -subunit of the ATP synthase complex (DRAPIER et al. 1992 ), cytochrome f of the cytochrome b₆f complex (KURAS and WOLLMAN 1994 ; CHOQUET et al. 1998 ), the PsaA reaction center subunit of PSI (GIRARD-BASCOU et al. 1987 ; STAMPACCHIA et al. 1997 ), the D1 and CP47 subunits of PSII (BENNOUN et al. 1986 ; ERICKSON et al. 1986 ; JENSEN et al. 1986 ; DE VITRY et al. 1989 ), the large subunit of Rubisco (KHREBTUKOVA and SPREITZER 1996 ; reviewed in CHOQUET et al. 1998 ; WOLLMAN et al. 1999 ; CHOQUET and VALLON 2000 ). It is highly unlikely that all these CES subunits would have evolved specific RNA-binding domains, able to interact specifically with their own encoding mRNAs. Thus, a competitive binding of a ternary effector to the unassembled CES subunit and its mRNA is the most likely mechanism underlying the CES process. As discussed above, there is overwhelming evidence for the presence of chloroplast gene-specific translational activators of nuclear origin in C. reinhardtii (see BARKAN and GOLDSCHMIDT-CLERMONT 2000 ; ZERGES 2000 for reviews). Some of these nuclear factors could have been recruited for the CES process during evolution.

The present analysis of cytochrome f expression in a strain carrying a leaky allele of TCA1 showed that the mutated TCA1 factor became limiting in the CES process, as expected from a ternary CES effector. First, cytochrome f expression could no longer be stimulated in that leaky tca1 strain, even in the absence of the C-terminal regulatory motif of cytochrome f: The rate of synthesis of C-ter truncated mutated cytochrome f and wild-type cytochrome f remained similar in the tca1 leaky strain, although these two protein versions showed a threefold difference in their expression levels in a wild-type nuclear context. Second, due to the absence of subunit IV, cytochrome f accumulation in the tca1 leaky strain dropped from 15 to 3% of the wild-type level. The 5-fold repression observed in the presence of this tca1 leaky allele compared to the 10-fold repression observed in a wild-type TCA1 context can be attributed to the tca1 leaky allele. This tca1 leaky allele could correspond either to a drop in the concentration of a fully functional TCA1 factor or to a mutated TCA1 factor with altered function. Thus, the characteristics of the CES-mediated up- and downregulation of cytochrome f synthesis in the presence or absence of the tca1 leaky allele are consistent with TCA1 being the CES ternary effector. The molecular identification of TCA1 should open the way to a refined characterization of the mechanism underlying the regulation of cytochrome f synthesis by the CES process.

	ACKNOWLEDGMENTS

We are grateful to D. Culler and S. Merchant for providing us with some of the mutant strains used in this work. We thank S. Bujaldon for technical assistance and D. Drapier and R. Kuras for critical reading of the manuscript and stimulating discussion during the course of this work. We also thank J. Maul and D. Stern for providing unpublished sequences from the Chloroplast genome sequencing project. This work was supported by CNRS/UPR1261. K. Wostrikoff was supported by a fellowship from the French Ministère de l'Education Nationale, de la Recherche et de la Technologie.

Manuscript received March 20, 2001; Accepted for publication June 22, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BARKAN, A. and M. GOLDSCHMIDT-CLERMONT, 2000  Participation of nuclear gene in chloroplast gene expression. Biochimie 82:559-572[Medline].

BARKAN, A., M. WALKER, M. NOLASCO, and D. JOHNSON, 1994  A nuclear mutation in maize blocks the processing and translation of several chloroplast mRNAs and provides evidence for the differential translation of alternative mRNA forms. EMBO J. 13:3170-3181[Medline].

BAYMANN, F., F. ZITO, R. KURAS, L. MINAI, and W. NITSCHKE et al., 1999  Functional characterization of Chlamydomonas mutants defective in cytochrome f maturation. J. Biol. Chem. 274:22957-22967[Abstract/Free Full Text].

BENNOUN, P. and D. BEAL, 1997  Screening algal mutant colonies with altered thylakoid electrochemical gradient through fluorescence and delayed luminescence digital imaging. Photosynth. Res. 51:161-165.

BENNOUN, P., and P. DELEPELAIRE, 1982 Isolation of photosynthesis mutants in Chlamydomonas, pp. 25–38 in Methods in Chloroplast Molecular Biology, edited by M. EDELMAN, N.-H. CHUA AND R. B. HALLICK. Elsevier Biomedical Press, Amsterdam/New York.

BENNOUN, P., M. SPIERER-HERZ, J. ERICKSON, J. GIRARD-BASCOU, and Y. PIERRE et al., 1986  Characterization of photosystem II mutants of Chlamydomonas reinhardii lacking the psbA gene. Plant Mol. Biol. 6:151-160.

BUSCHLEN, S., Y. CHOQUET, R. KURAS, and F. A. WOLLMAN, 1991  Nucleotide sequences of the continuous and separated petA, petB and petD chloroplast genes in Chlamydomonas reinhardtii. FEBS Lett. 284:257-262[Medline].

CHOQUET, Y. and O. VALLON, 2000  Synthesis, assembly and degradation of thylakoid membrane proteins. Biochimie 82:615-634[Medline].

CHOQUET, Y., M. RAHIRE, J. GIRARD-BASCOU, J. ERICKSON, and J. D. ROCHAIX, 1992  A chloroplast gene is required for the light-independent accumulation of chlorophyll in Chlamydomonas reinhardtii. EMBO J. 11:1697-1704[Medline].

CHOQUET, Y., D. B. STERN, K. WOSTRIKOFF, R. KURAS, and J. GIRARD-BASCOU et al., 1998  Translation of cytochrome f is autoregulated through the 5' untranslated region of petA mRNA in Chlamydomonas chloroplasts. Proc. Natl. Acad. Sci. USA 95:4380-4385[Abstract/Free Full Text].

DE VITRY, C., J. OLIVE, D. DRAPIER, M. RECOUVREUR, and F. A. WOLLMAN, 1989  Posttranslational events leading to the assembly of photosystem II protein complex: a study using photosynthesis mutants from Chlamydomonas reinhardtii. J. Cell Biol. 109:991-1006[Abstract/Free Full Text].

DRAGER, R. G., J. GIRARD-BASCOU, Y. CHOQUET, K. L. KINDLE, and D. B. STERN, 1998  In vivo evidence for 5'3' exoribonuclease degradation of an unstable chloroplast mRNA. Plant J. 13:85-96[Medline].

DRAPIER, D., J. GIRARD-BASCOU, and F.-A. WOLLMAN, 1992  Evidence for nuclear control of the expression of the atpA and atpB chloroplast genes in Chlamydomonas. Plant Cell 4:283-295[Abstract/Free Full Text].

DRAPIER, D., H. SUZUKI, H. LEVY, B. RIMBAULT, and K. L. KINDLE et al., 1998  The chloroplast atpA gene cluster in Chlamydomonas reinhardtii. Functional analysis of a polycistronic transcription unit. Plant Physiol. 117:629-641[Abstract/Free Full Text].

ERICKSON, J. M., M. RAHIRE, P. MALNOE, J. GIRARD-BASCOU, and Y. PIERRE et al., 1986  Lack of the D2 protein in a Chlamydomonas reinhardtii psbD mutant affects photosystem II stability and D1 expression. EMBO J. 5:1745-1754[Medline].

FAN, W. H., M. A. WOELFLE, and G. MOSIG, 1995  Two copies of a DNA element, 'Wendy', in the chloroplast chromosome of Chlamydomonas reinhardtii between rearranged gene clusters. Plant Mol. Biol. 29:63-80[Medline].

FISK, D. G., M. B. WALKER, and A. BARKAN, 1999  Molecular cloning of the maize gene crp1 reveals similarity between regulators of mitochondrial and chloroplast gene expression. EMBO J. 18:2621-2630[Medline].

FOX, T. D., 1996 Genetics of mitochondrial translation, pp. 733–758 in Translational Control, edited by J. W. B. HERSHEY, M. B. MATHEWS and N. SONENBERG. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.

GAMBLE, P. E. and J. E. MULLET, 1989  Translation and stability of proteins encoded by the plastid psbA and psbB genes are regulated by a nuclear gene during light-induced chloroplast development in barley. J. Biol. Chem. 264:7236-7243[Abstract/Free Full Text].

GIRARD, J., N.-H. CHUA, P. BENNOUN, G. H. SCHMID, and M. DELOSME, 1980  Studies of mutants deficient in the photosystem I reaction centers in Chlamydomonas reinhardtii. Curr. Genet. 2:215-221.

GIRARD-BASCOU, J., Y. CHOQUET, M. SCHNEIDER, M. DELOSME, and M. DRON, 1987  Characterization of a chloroplast mutation in the psaA2 gene of Chlamydomonas reinhardtii. Curr. Genet. 12:489-495[Medline].

GIRARD-BASCOU, J., Y. PIERRE, and D. DRAPIER, 1992  A nuclear mutation affects the synthesis of the chloroplast psbA gene production Chlamydomonas reinhardtii. Curr. Genet. 22:47-52[Medline].