Analysis of the Nucleus-Encoded and Chloroplast-Targeted Rieske Protein by Classic and Site-Directed Mutagenesis of Chlamydomonas

Correspondence to: Catherine de Vitry, devitry{at}ibpc.fr (E-mail), 331-1-40-46-83-31 (fax)

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Three mutants of the alga Chlamydomonas reinhardtii affected in the nuclear PETC gene encoding the Rieske iron–sulfur protein 2Fe–2S subunit of the chloroplast cytochrome b₆f complex have been characterized. One has a stable deletion that eliminates the protein; two others carry substitutions Y87D and W163R that result in low accumulation of the protein. Attenuated expression of the stromal protease ClpP increases accumulation and assembly into b₆f complexes of the Y87D and W163R mutant Rieske proteins in quantities sufficient for analysis. Electron-transfer kinetics of these complexes were 10- to 20-fold slower than those for the wild type. The deletion mutant was used as a recipient for site-directed mutant petC alleles. Six glycine residues were replaced by alanine residues (6G6A) in the flexible hinge that is critical for domain movement; substitutions were created near the 2Fe–2S cluster (S128 and W163); and seven C-terminal residues were deleted (G171och). Although the 6G6A and G171och mutations affect highly conserved segments in the chloroplast Rieske protein, photosynthesis in the mutants was similar to that of the wild type. These results establish the basis for mutational analysis of the nuclear-encoded and chloroplast-targeted Rieske protein of photosynthesis.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Photosynthesis, organelle biogenesis, and many other processes in plant cells involve the coordinated activity of nuclear and plastid-encoded proteins. Elucidation of these processes will require multiple approaches, including nuclear transformation with modified genes. Genes have been widely introduced into plant nuclei; however, site-directed mutagenesis of plant nuclear genes has not been reported (Bevan et al. 1999 ). Homologous insertion and replacement events are extremely rare in both plant and algal nuclear genomes.

DNA introduced into the green alga Chlamydomonas reinhardtii cells by particle bombardment or other methods integrates largely randomly into the nuclear genome and permits the inactivation and tagging of selected genes (reviewed in Kindle 1998 ). Recently, site-directed mutagenesis and nuclear transformation have been used to introduce modified genes for the subunit of coupling factor 1 (Ross et al. 1995 ) and the PsaF subunit of photosystem I (PSI; Hippler et al. 1998 ) into Chlamydomonas mutants carrying null mutations of these genes. Chlamydomonas cells possess the photosynthetic apparatus and fundamental features of plant cells but are facultative phototrophs, synthesizing the photosynthetic apparatus under heterotrophic as well as phototrophic growth conditions. They serve as amenable and unique models for mutational studies of plant cell functions involving nuclear genes. We exploited this potential for studies of the nuclear-encoded Rieske iron–sulfur protein of the chloroplast cytochrome b₆f complex, a transmembrane proton pump of the photosynthetic electron transfer chain.

The high midpoint potential (E_m of ~+300 mV) Rieske 2Fe–2S protein plays essential roles in photosynthetic electron transport, redox sensing, and signal transduction in chloroplasts (reviewed in Keren and Ohad 1998 ; Wollman 1998 ). The cytochrome b₆f complex transfers electrons from photosystem II (PSII) to PSI by plastoquinol oxidation and reduction of plastocyanin and contributes to translocation of protons across the thylakoid membrane. The Rieske protein catalyzes the oxidation of plastoquinol at the Q_o site close to the lumenal membrane face, leading to bifurcated electron transfer along both a high-potential chain (via the Rieske protein and cytochrome f to plastocyanin) and a low-potential chain (via the hemes b_l and b_h, allowing electron transfer across the membrane to a quinone at the Q_i site near the stromal face of the thylakoid membrane). Related cytochrome bc complexes are part of mitochondrial and bacterial respiratory chains. High-resolution structures have been determined for the soluble domains of the bovine mitochondrial (Iwata et al. 1996 ) and spinach chloroplast (Carrell et al. 1997 ) Rieske proteins and show generally similar organization. Recent x-ray crystal structures of these complexes have shown that the Rieske protein comprises a soluble domain tethered by a flexible region to a transmembrane anchor (Xia et al. 1997 ; Kim et al. 1998 ; Zhang et al. 1998 ). The soluble domain occupied different positions in bc₁ complexes that were cocrystallized with inhibitors, which suggests a unique domain movement during catalysis (Kim et al. 1998 ; Zhang et al. 1998 ).

The chloroplast Rieske protein is encoded by the nuclear PETC gene and is imported into the chloroplast via a cleavable presequence; the soluble domain is translocated across the thylakoid membrane and into the chloroplast lumen via residues within the N-terminal hydrophobic domain of the mature protein (de Vitry 1994 ; Madueno et al. 1994 ). The assembly of the 2Fe–2S cluster and of the Rieske protein into the b₆f complex, as well as the role of the protein in electron transfer and signal transduction, are the subjects of ongoing research. The Chlamydomonas PETC gene has been cloned (de Vitry 1994 ). This unicellular alga is readily amenable to pulse labeling, has a rapid growth rate (generation time ~8 hr), and is a facultative phototroph; its haplobiontic cell cycle allows a variety of genetic studies; and transformation of the nucleus and chloroplast genome is easily performed (Rochaix et al. 1998 ). The cytochrome b₆f complex has been purified (Pierre et al. 1995 ), and measurements of rapid electron transfer reactions in vivo (reviewed in Joliot et al. 1998 ) are possible. A Chlamydomonas Rieske-deficient mutant, ac21, has long been available (Levine and Smillie 1962 ) and assembles the remaining subunits of the b₆f complex (Lemaire et al. 1986 ); however, this mutant reverts. Additional, more stable Rieske-deficient mutants 788 (petC-1) and 794 are described in the present work. The use of a Chlamydomonas mutant deficient in ClpP protease (W. Majeran, F.-A. Wollman, and O. Vallon, manuscript in preparation) allowed the overexpression of the defective ac21 (petC-W163R) and 794 (petC-Y87D) Rieske proteins in quantities that permitted biophysical analysis.

Rieske proteins of bc₁ complexes have been studied extensively by using mutagenesis (reviewed in Graham et al. 1993 ; Kallas 1994 ) and suppressor analyses. Residues with roles in cluster binding, midpoint potential determination (Gatti et al. 1989 ; Liebl et al. 1997 ; Denke et al. 1998 ; Schroter et al. 1998 ), and domain interactions within the bc₁ complex Rieske proteins and with neighboring subunits have been defined (Brasseur et al. 1997 ; Saribas et al. 1998 ). Site-directed mutagenesis of b₆f subunits of cyanobacteria is also possible (Wu et al. 1995 ); however, only nonlethal mutants can be propagated because the b₆f complex is required for both photosynthetic and heterotrophic growth (reviewed in Kallas 1994 ). Mutational investigations of chloroplast Rieske proteins have not been attempted. For initial study (de Vitry et al. 1998 ), we chose substitutions of the conserved S128 residue, which is involved in cluster assembly and redox potential determination, and of conserved aromatic residue W163, which may play a role in stability and electron transport. We also created mutants in the flexible neck region implicated in the Rieske protein movement and in the unique conserved 12- to 14-residue C-terminal extension of b₆f Rieske proteins.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Mutants Deficient in the Rieske Protein Accumulate the Other Cytochrome b₆f Complex Subunits, with an Apparently Modified Cytochrome b₆
Mutant ac21, obtained by UV mutagenesis (Levine and Smillie 1962 ), was identified as being deficient in the Rieske protein (Lemaire et al. 1986 ). Mutants 788 and 794 also were obtained by UV mutagenesis and were identified in a collection of b₆f mutants kindly provided by S. Merchant (University of California, Los Angeles). Heme staining of the cells shown in Figure 1A demonstrates that ac21, 788, and 794 accumulated cytochromes f and b₆ as heme binding proteins. The results of the immunodetection analysis shown in Figure 1B indicate that the b₆f subunits, except the Rieske protein, accumulated to ~60% of wild-type amounts. Mutants deficient in the Rieske protein had a cytochrome b₆ of higher apparent molecular mass than did the wild type when the proteins were electrophoresed (Figure 1A and Figure 1B).

View larger version (33K): [in this window] [in a new window]   Figure 1. Subunits of the Cytochrome b6f Complex in Rieske Protein–Deficient Mutants. (A) Heme staining of Chlamydomonas wild-type and mutant cell extracts separated by SDS-PAGE. (B) Immunodetection (cell extracts). (C) The Rieske protein of the wild type and ac21 mutant seen after 7 min of 14C acetate pulse-labeling in the presence of chloramphenicol (membrane extracts). (D) Immunodetection (cell extracts) of ClpP protease and Rieske protein in the single mutant 794 (named petC-Y87D after molecular characterization) and the double mutant petC-Y87D clpP-AUU (794 mutant attenuated in stromal protease ClpP expression). Cyt, cytochrome; IV, subunit IV; WT, wild type.

Immunodetection of cell extracts revealed the complete absence of the Rieske protein in the 788 mutant and variable but low accumulation in the 794 and ac21 mutants, typically at ~1 and 4% of the wild-type amount, respectively (in early exponential-phase cultures of ~10⁶ cells/mL). The apparent molecular mass of the Rieske protein when electrophoresed was higher in the 794 mutant and lower in the ac21 mutant than in the wild type. Analysis of polypeptides synthesized on cytoplasmic ribosomes and inserted into thylakoid membranes during 7 min of pulse labeling showed (1) insertion of the Rieske protein into thylakoids in the wild type and the ac21 mutant, (2) a lower apparent molecular mass of the Rieske protein in the ac21 mutant (as shown in Figure 1B), and (3) less intense labeling of the Rieske protein in the ac21 mutant compared with that in the wild-type cells (Figure 1C).

Localization of the Mutations in the Nuclear PETC Gene Encoding the Chloroplast Rieske Protein
Genetic analyses are detailed in Table 1. The absence of recombination and complementation indicated that the 788, 794, and ac21 mutants were affected in the same gene.

PETC transcript analysis is shown in Figure 2. The PETC transcript was undetectable in the 788 mutant, equaled the wild-type quantity in the 794 mutant, and overaccumulated in the ac21 mutant. Complementation with a 4.5-kb genomic PETC DNA fragment restored phototrophy and the wild-type quantity of the PETC transcript and indicated that the 788 and 794 mutants are directly affected in the PETC gene (de Vitry et al. 1998 ). PETC complementation was not performed with ac21 mutant because of a high frequency of reversion events. Revertants of 788 were not detected, whereas reversion occurred in approximately one per 10⁷ ac21 cells.

View larger version (39K): [in this window] [in a new window]   Figure 2. petC Transcript Analysis of Rieske Protein–Deficient Mutants. T788 and T794 are the 788 and 794 mutants transformed with the wild-type PETC gene encoding the Rieske protein. The atpA transcript was used as a control. WT, wild type.

Sequencing of polymerase chain reaction (PCR) products identified lesions in the PETC gene that cause (1) a frameshift (TG deletion in codon 7 of the mature protein) leading to a truncated Rieske protein of 52 residues in the 788 mutant (petC-1) with a second, probably neutral mutation in intron 3 (G784A; numbered relative to the initiator ATG codon of the wild-type preprotein); (2) the substitution Y87D (TACGAC) in the mature Rieske protein of mutant 794; and (3) the substitution W163R (TGGCGG) in the mature Rieske protein of mutant ac21.

Dimeric Assembly of Cytochrome b₆f Complex in Mutants Deficient in the Rieske Protein
Cytochrome b₆f complexes were purified from exponentially growing cells of the Rieske-deficient mutant and wild type by a modification of the Hecameg solubilization procedure described in Pierre et al. 1995 . Figure 3 shows immunoblot analysis of the distribution of the cytochrome f and Rieske proteins upon sucrose gradient centrifugation. The distribution of these proteins peaked in fractions of similar sucrose density, at a position known to be that of the dimeric form of the wild-type b₆f complex (Breyton et al. 1997 ). In the ac21 (petC-W163R) mutant, some of the Rieske protein assembled into b₆f complexes (Figure 3). The cytochrome f, cytochrome b₆, subunit IV, PetG, PetL, and PetM subunits comigrated in the gradients from the 788 (petC-1) and ac21 (petC-W163R) mutants, indicative of their assembly into complexes in these mutants (data not shown).

View larger version (65K): [in this window] [in a new window]   Figure 3. Dimeric State of the Cytochrome b6f Complex in Rieske Protein–Deficient Mutants. The cytochrome b6f complex was purified from thylakoid membranes by using the Hecameg solubilization technique (see Results). The supernatant was centrifuged on a 10 to 30% sucrose gradient. Fractions from the gradient were analyzed by SDS-PAGE, and cytochrome f and the Rieske protein were immunodetected. The open and filled arrowheads indicate the sucrose concentration at the peak of cytochrome b6f complex in the gradient with and without Rieske protein, respectively. The sucrose concentration of the fraction is given above the arrowhead as a percentage. Cyt, cytochrome; WT, wild type; (+), with; (±), with or without; (-), without.

Stabilization of Mutant Rieske Proteins by Attenuated Expression of the Stromal Protease ClpP
When a mutation affects the properties of a protein complex without compromising its stability, this provides valuable material for functional analysis. The other subunits of the Chlamydomonas b₆f complex assemble even in the absence of the Rieske protein. However, some mutations in the Rieske protein, such as petC-Y87D and petC-W163R, allowed only very low accumulation of the Rieske protein. A chloroplast mutant that attenuates the expression of the stromal protease ClpP has been created by site-directed mutagenesis of the CLPP gene, resulting in a fourfold decrease in the chloroplast content of ClpP (W. Majeran, F.-A. Wollman, and O. Vallon, manuscript in preparation). Because nuclear mutations have biparental heredity, whereas chloroplast mutations are transmitted uniparentally by the mating type (+), we attempted to improve the stability of the mutant Rieske protein by crossing the nuclear mutants petC-Y87D and petC-W163R of mating type (-) with the chloroplast clpP-AUU mutant of mating type (+). In the resulting double mutants, petC-Y87D clpP-AUU and petC-W163R clpP-AUU, the accumulation of the Rieske protein could be increased in early exponential-phase cultures to ~8% (Figure 1D) and 40% (W. Majeran, F.-A. Wollman, and O. Vallon, manuscript in preparation) of the wild-type amount, respectively. We observed an increase in the growth rate of the double mutant petC-W163R clpP-AUU but no growth of the petC-Y87D or petC-Y87D clpP-AUU mutants under phototrophic conditions.

Electron Transfer Is Blocked in the Deletion Mutant petC-1 and Considerably Slowed in the Substitution Mutants petC-Y87D and petC-W163R
Figure 4 shows the patterns of fluorescence induction at room temperature of dark-adapted wild-type and mutant cells when illuminated continuously. The addition of 3-(3',4'-dichlorophenyl)-1,1-dimethylurea (DCMU) prevents reoxidation of the primary PSII quinone acceptor and leads to maximal fluorescence yield F_max. In the absence of DCMU, the initial fluorescence value defines F₀, and the steady state fluorescence value defines F_stat. The ratio (F_max-F_stat)/(F_max-F₀) is a relative measure of the linear electron flow (LEF): the smaller the difference, the lower the reoxidation rate of the PSII acceptor by the downstream electron transport chain. Table 2 summarizes the LEF results. The petC-1 mutant clearly was completely blocked in PSII reoxidation and therefore in photosynthetic LEF. The petC-Y87D (data not shown) and petC-Y87D clpP-AUU (Table 2) mutants also seemed to be blocked in linear electron transfer under these conditions. The fluorescence pattern of petC-W163R (Figure 4) and petC-W163R clpP-AUU (Table 2) showed little difference and significant difference, respectively, between F_max and F_stat, which is indicative of increased efficiency of linear LEF when ClpP is attenuated.

View larger version (20K): [in this window] [in a new window]   Figure 4. Fluorescence Induction Transients of Dark-Adapted Cells from WT, petC-1, petC-W163R, and Site-Directed Rieske Mutants in the Presence or Absence of 10 µM DCMU. Maximal levels of fluorescence (Fmax) were normalized to 100%.

Table 2 summarizes b₆f kinetics in vivo. The half-time of the slow rise of the electrochromic shift of carotenoids at 515 nm was induced by changes in the transmembrane potential during charge transfer through the b₆f complex. The permanent electrochemical potential of the living algae is abolished by addition of the uncoupler carbonylcyanide-p-trifluoromethoxyphenylhydrazone (FCCP). In agreement with previous reports (Finazzi et al. 1997 ), the wild-type cells showed a half-time of 1.2 to 2.4 msec under these conditions. The half-time of cytochrome f reduction was 1.8 to 2.7 msec for the wild type. In contrast, in the petC-1 mutant, the slow rise of the carotenoid band shift was absent. Cytochrome f became oxidized by plastocyanin, but its rereduction was 10⁴ times slower than in the wild type (Finazzi et al. 1997 ). This value is similar to that observed in a mutant that accumulates a soluble cytochrome f and lacks the remainder of the b₆f complex (Kuras et al. 1995 ), which indicates that this reduction is not mediated by the Q_o site.

The quantity of Rieske protein–containing b₆f complexes in the mutants petC-Y87D and petC-W163R was far below the wild-type amount. Charge separation in PSI induced by an actinic flash therefore will induce oxidation of more plastocyanin molecules than functional b₆f complexes. This induces multiple turnover of the few active complexes and results in increased half-times of the slow rise of the carotenoid band shift and of cytochrome f reduction. To obtain accurate half-time values for single turnovers of the b₆f complex, we used the double mutants petC-Y87D clpP-AUU and petC-W163R clpP-AUU. These accumulated more Rieske protein than did the single mutants bearing the petC mutations alone. Furthermore, the intensity of the actinic light was lowered until further attenuation did not lead to additional acceleration of the reaction kinetics. Time-resolved spectroscopic analysis of the petC-W163R clpP-AUU mutant is shown in Figure 5A and Figure 5B. Real half-time values for electron transfer in b₆f complexes of petC-W163R and petC-Y87D were 10- to 20-fold lower than wild-type values (Table 2). Kinetics of the electrochromic shift and of cytochrome f reduction were proportionally affected in both mutants. These data indicate a perturbation in the concerted electron transfer from the quinol at the Q_o site to the Rieske protein.

View larger version (13K): [in this window] [in a new window]   Figure 5. Time-Resolved Spectroscopic Analysis of the Double Mutant petC-W163R clpP-AUU. petC-W163R was named ac21 before molecular characterization. clpP-AUU mutants are attenuated in stromal protease ClpP expression. (A) Kinetics of the slow electrochromic phase at 515 nm indicative of charge transfer across the membrane with nonsaturating actinic flashes (10% saturation). AU, arbitrary units. (B) Time-resolved redox changes of cytochrome f. -I/I is the difference between the light transmitted through the measuring and the reference cuvettes divided by the light transmitted through the reference cuvette (Joliot et al. 1998 ).

Strategy for Site-Directed Mutagenesis of the Nuclear-Encoded Rieske Protein
Mutations created in the chloroplast Rieske protein are described in Table 3 (see Methods) and shown in a model of the protein in Figure 6. These replace glycine residues in the flexible hinge region implicated in domain movement (6G6A), create substitutions of S128 (S128A, S128C, and S128T) or of the aromatic residue W163 (W163A) near the 2Fe–2S cluster, or delete residues from the C terminus (G171och).

View larger version (59K): [in this window] [in a new window]   Figure 6. Mutation Sites in the Chlamydomonas Chloroplast Rieske Protein. Cysteine and histidine ligands for the 2Fe–2S cluster are represented by heavily outlined squares. C110 and C125 form a disulfide bond. The box TM represents a likely transmembrane helix. The helix 1 and ß sheet strands are indicated, as is the location of the N terminus of the spinach Rieske fragment (sp ISF). Locations corresponding to the C terminus and an extra loop of mitochondrial Rieske proteins are marked bc1 C-ter and bc1 loop, respectively. Boldface residues are conserved in the bc and bf Rieske proteins. Black circles represent sites at which mutations were detected or created. In this work, the model is based on the structures of mitochondrial (Iwata et al. 1996 ) and spinach chloroplast (Carrell et al. 1997 ) Rieske fragments and of Rieske subunits of bc1 complexes (Xia et al. 1997 ; Kim et al. 1998 ; Zhang et al. 1998 ).

Mutant 788 (petC-1) was used as the recipient for site-directed mutations. Mutant petC alleles contained a BamHI restriction site tag in intron 1 and a second, specific restriction tag associated with each directed mutation (Table 3). Mutant i1Bam is a control containing only the BamHI restriction site tag in intron 1. For our initial study, we focused on Rieske protein mutations that could support photosynthesis, selecting these by their ability to grow phototrophically on minimal medium. Phototrophic transformants were recovered at a frequency of ~1 to 10 per 10⁸ recipient cells after transformation with plasmids carrying the i1Bam, 6G6A, S128T, and G171och petC alleles.

To confirm that phototrophic colonies arose from integration of modified petC alleles and not from reversion, recombination, or gene conversion events that might restore a wild-type PETC gene, we performed allele-specific PCR and restriction tests (Figure 7). Under stringent conditions, the primers 788TG and C5aFS amplified only the defective 788 petC allele and not the wild type or introduced petC alleles (Figure 7A). These primers generated the expected 1242-bp PCR products from mutant 788 and from 788 transformed with the 6G6A, S128T, and G171och petC alleles but not from wild-type Chlamydomonas (Figure 7C). Mutant 788 and these transformants therefore retained copies of the defective 788 petC allele. Primers WT107TG and C5aFS amplify petC alleles that are wild type in exon 1 (Figure 7B). Accordingly, these primers generated the expected 1243-bp PCR product from wild-type Chlamydomonas and from mutant 788 transformed with the mutant petC alleles but did not amplify DNA from the original 788 mutant (Figure 7D). To confirm that these transformants contained the mutant petC genes, PCR products from the WT107TG/C5aFS primer set were digested with restriction enzymes specific for the mutant 6G6A, S128T, and G171och petC alleles. The nearly complete digestion of these products and generation of expected restriction fragments (Figure 7E) confirm that restoration of phototrophy to the 788 (petC-1) mutant resulted from transformation with specific mutant petC alleles.

View larger version (47K): [in this window] [in a new window]   Figure 7. Allele-Specific PCR and Restriction Analysis of Chlamydomonas Transformants. (A) Map of the petC gene of the Rieske deletion mutant 788. The location of the deletion mutation in mutant 788 is indicated, as are the binding sites for the primers 788TG and C5aFS. The 788TG primer is specific for the 788 petC allele and, along with primer C5aFS, generates a 1242-bp product by PCR. (B) Map of the wild-type and site-mutated petC genes. Each modified petC allele carries a BamHI site (i1Bam) in intron 1. The location of each site-directed mutation and its accompanying restriction tag is shown. The WT107TG primer is specific for the wild-type PETC sequence of exon 1 and, along with primer C5aFS, generates a 1243-bp product by PCR. (C) Stained gel showing PCRs with primers 788TG and C5aFS. DNA extracts were from the wild type (WT) and Rieske deletion (788) mutant, and from mutant 788 transformed with the 6G6A, S128T, and G171och petC alleles (fragment sizes in base pairs are shown at right). The left lanes show size standards (1-kb DNA ladder from BRL). (D) PCRs with primers WT107TG and C5aFS. (E) Restriction digestion of PCR products obtained with the WT107TG and C5aFS primer set on DNAs from the 788 transformants and the wild type.

Subunits of the Cytochrome b₆f Complex and Restoration of the Expected Full-Length or Truncated Rieske Proteins in Transformants of Mutant 788 ( petC-1)
Comparable amounts of cytochromes f and b₆ were found in the wild type and transformants (Figure 8A). Furthermore, as shown in Figure 1A and in lane 1 of Figure 8A, the wild type typically displays two diffuse 3,3',5,5'-tetramethylbenzidine–stained bands for cytochrome b₆. The lower band was undetectable in the Rieske deletion mutant 788 but was restored in the transformants. Figure 8B shows restoration of the Rieske protein to mutant 788 (petC-1) transformed with mutant petC alleles. The G171och Rieske protein migrated at a lower apparent molecular mass, as expected for deletion of seven residues from the C terminus. The 6G6A Rieske protein also migrated at a slightly lower position.

View larger version (49K): [in this window] [in a new window]   Figure 8. Subunits of the Cytochrome b6f Complex in Cell Extracts of Site-Directed Mutants. (A) Heme staining of Chlamydomonas wild-type and mutant cell extracts separated by SDS-PAGE. (B) Immunodetection of the Rieske protein and cytochrome f. Cyt, cytochrome; WT, wild type.

Mutant Rieske Proteins Bearing a Less Flexible Hinge Region or a Deletion of the C Terminus Show Electron Transport Activities Similar to Those of the Wild Type
6G6A, G171och (Figure 4), and the control mutant i1Bam (data not shown) had wild-type fluorescence characteristics. The S128T mutant showed a fluorescence pattern somewhat different from wild type, but nevertheless its LEF approximated that of the wild type, indicating functional b₆f complexes. These data (summarized in Table 2) are consistent with restoration of photosynthetic electron transport and phototrophic growth by nuclear transformation with the mutant petC alleles. The functional properties of these mutant Rieske proteins were analyzed further by flash-induced, time-resolved spectroscopic measurements in vivo. As shown in Table 2, values for both phase b of the 515-nm electrochromic band shift and the rereduction kinetics of cytochrome f fell largely within the range of values expected for wild-type Chlamydomonas cells (Finazzi et al. 1997 ). These data indicate that both of the major reactions of the cytochrome b₆f complex, the transfer of an electron through the high-potential chain (Rieske 2Fe–2Scytochrome f plastocyanin or cytochrome c₆PSI), and the transfer of a charge across the membrane operated essentially with wild-type kinetics in the petC mutants.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Substitutions Y87D and W163R Impede Rieske Protein Assembly and Function
Chlamydomonas is the only genetically tractable eukaryote in which photosynthesis and carbon assimilation mutations are conditional rather than lethal. We have used a mutant lacking the Rieske protein (petC-1) to develop site-directed mutagenesis of the nuclear-encoded, chloroplast-targeted Rieske protein in Chlamydomonas. One previous mutant, ac21, was identified as lacking the Rieske protein and shown to accumulate the remaining subunits of the b₆f complex (Lemaire et al. 1986 ). We have shown here that ac21 is indeed a Rieske mutant carrying the substitution W163R. However, the petC-W163R mutant protein can accumulate under suitable conditions, integrate into the thylakoid membrane, and transfer electrons, although 10 times more slowly than does the wild-type Rieske protein.

Rieske mutant 794 also carried an aromatic substitution (Y87D) and accumulated only small amounts of the Rieske protein assembled into the b₆f complex. Figure 9A shows the locations of the Chlamydomonas Y87D and W163R substitutions in the structure of the Rieske protein soluble domain. In the spinach protein, Y89 (the counterpart of Chlamydomonas Y87) occupies the gap between the cluster binding fold and the larger subdomain of the soluble fragment. Y89 may form hydrogen bonds with either V114 or F116 (Chlamydomonas V112 or F114) of the cluster binding fold (Figure 9B). The negatively charged aspartic acid substitution at this position is predicted to leave the gap unoccupied, to increase its polarity, and to prevent hydrogen bonds to the cluster fold (Figure 9C). This detachment of the cluster binding fold from the base fold may resemble that recently described in the intermediate Rieske protein state in bc₁ crystals (Iwata et al. 1998 ). The Chlamydomonas Y87D substitution had a marked effect, presumably because it destabilized the soluble domain fold and allowed accessibility to proteases, resulting in degradation. The W163R substitution of mutant ac21 similarly replaced an aromatic residue with a charged residue. Its phenotype was like that of petC-Y87D but less severe in that more of the protein assembled into a b₆f complex. The W163R substitution has a more peripheral location. In spinach, W164 (Chlamydomonas W163) forms hydrogen bonds with the tail region, including a hydrogen bond between its backbone and W177 (Chlamydomonas W176). However, we report here that deletion of residues G171 to A177 had no functional impact in Chlamydomonas.

View larger version (59K): [in this window] [in a new window]   Figure 9. Locations of Mutation Sites in the Soluble Domain of the Chloroplast Rieske Protein. (A) The structure is that of the soluble fragment (ISF) of the spinach chloroplast Rieske protein (Carrell et al. 1997 ). The view was generated by the molecular graphics program Rasmol, version 2.6 (R. Sayle, Glaxo, Greenford, UK). Residue numbering is that of the Chlamydomonas protein (de Vitry 1994 ). Strands represent the peptide backbone. The 2Fe–2S cluster is shown by van der Waals spheres. Stick structures show Y87 (replaced by aspartic acid residue in the 794 mutant) and W163 (replaced by arginine and alanine residues in ac21 and petC-W163A mutants, respectively). Sticks also represent the mutated S128 and the two tryptophan residues of the C terminus contained within the deletion (shown as a dark ribbon) created by the G171och mutation. (B) Detail of the Y87 region (equivalent to spinach Y89). Increasing polarity of amino aci d residues is represented from light gray to dark gray and possible hydrogen bonds as dashed lines. (C) Predicted detail of the region shown in (B) after substitution with Y87D. The molecular graphics program Swiss-PdbViewer, version 3.0 (N. Guex and C. Peitsh, Geneva Biomedical Resarch Institute, Geneva, Switzerland), was used to visualize amino acid side chain modifications.

The Rieske protein is synthesized in the cytosol with an N-terminal signal peptide that directs its import into the stroma of the chloroplast and then is inserted into the thylakoid membrane. The Chlamydomonas chloroplast Rieske protein is easily extracted from the membrane (Breyton et al. 1994 ) but is most probably anchored in the membrane by an N-terminal helix with the majority of the protein in the lumen by analogy to the bc₁ complex structure (Kim et al. 1998 ; Zhang et al. 1998 ). Stromal proteases have access to the Rieske protein during its translocation and also can cleave a short stromal-exposed segment of the assembled Rieske protein in spinach (Karnauchov et al. 1997 ). The mutant W163R Rieske protein accumulated less label than did the wild-type protein during short pulse-labeling experiments (Figure 1C). It is improbable that synthesis of the mutant protein was diminished; more probably, however, translocation or stability was affected.

A chloroplast mutant attenuating the expression of the stromal protease ClpP has been constructed (W. Majeran, F.-A. Wollman, and O. Vallon, unpublished results). The double mutants petC-W163R clpP-AUU and petC-Y87D clpP-AUU reported here had significantly increased quantities of the mutant Rieske proteins compared with the single mutants petC-W163R and petC-Y87D. Lower labeling and ClpP sensitivity of the mutant proteins are compatible with translocation of the lumenal domain of the Rieske protein across the thylakoid membrane by a pH-dependent mechanism (Madueno et al. 1994 ) in a prefolded, cofactor binding conformation as suggested by Berks 1996 . Less efficient translocation of more open structures of the W163R and Y87D Rieske proteins could allow stromal degradation of a portion of these molecules. Some Y87D and W163R proteins escaped degradation and assembled into b₆f complexes that showed charge translocation (phase b) and cytochrome f reduction kinetics ~10- to 20-fold slower than the wild type (Table 2). Cpn60 and heat shock Hsp70 chloroplast-localized chaperones have been reported to interact with the Rieske protein in the stroma (Madueno et al. 1993 ). Overexpression of Hsp70 (Schroda et al. 1999 ) may represent a further strategy for increased accumulation of mutant proteins.

Functional Organization of the Thylakoid Membrane in a Stable PetC Deletion Mutant
Mutant 788 showed a two-base deletion in PETC leading to a frameshift in codon 7 and termination at codon 53 (86 bases before the second intron) that entirely eliminated the Rieske protein. The PETC transcript was undetectable in mutant 788. In eukaryotic cells, mechanisms exist for eliminating mRNAs in which translation terminates prematurely. It has been proposed that any intron located more than 50 to 55 nucleotides downstream of a termination codon can mediate a reduction in mRNA abundance (Nagy and Maquat 1998 ). This rule applies to 788 and may explain the absence of PETC mRNA from this mutant. In contrast, the PETC transcript overaccumulated in the petC-W163R mutant and accumulated to wild-type quantities in the petC-Y87D mutant.

Purified b₆f complexes from 788 (petC-1) mutant retained a dimer structure, indicating that the Rieske protein does not control the oligomerization of the extractable complex. This is in agreement with experiments on Chlamydomonas mutant fud2 (Finazzi et al. 1997 ), which has a duplication of 12 amino acids in the cytochrome b₆ loop between transmembrane helices C and D of the Q_o binding site. During purification, the Rieske protein detaches from the FUD2 b₆f complex without leading to monomerization. These observations are of note because loss of the Rieske protein and monomerization often occur simultaneously, which therefore raises the question of their causal relationship (Breyton et al. 1997 ). The state of oligomerization of nonextractable complexes of the petC-1 mutant is not known. Extraction of b₆f complexes from thylakoid membranes is less efficient in Rieske protein–deficient mutants (~50% in petC-1 or petC-W163R mutants compared with ~80% in the wild type). This may reflect a modified functional organization of the thylakoid membrane in these mutants.

The "state transition" phenomenon refers to redistribution of antenna proteins between the two photosystems in response to cellular requirements for ATP or reducing equivalents. These transitions do not occur in the absence of the Rieske protein in the 788 (petC-1) mutant (R. Delosme, J. Olive, and F.-A. Wollman, data not shown).

The Rieske protein is implicated in a modification of cytochrome b₆. The cytochrome b₆ protein appears upon electrophoresis as two diffuse bands in early exponential-phase cultures (10⁶ cells/mL). The lower band is undetectable in the petC-1 mutant and is restored after transformation with PetC alleles. Several distinct immunoreactive bands of cytochrome b₆ have been detected in chloroplast mutants bearing substitutions in one or the other of the histidine ligands for heme b_h (two bands) or b_l (one band) or in nuclear ccb mutants defective in holocytochrome b₆ formation (two bands) (Kuras et al. 1997 ). The absence of the lower band in the petC-1 mutant suggests that there is a modification of cytochrome b₆ requiring the presence of the Rieske protein.

Structural and Functional Implications of the Site-Directed Mutants
Because of inefficient homologous recombination, directed mutagenesis of the chloroplast Rieske protein depended on a stable, null mutant that could be used as the recipient for modified petC alleles. Mutant 788 (petC-1), characterized in this work, meets this criterion. As demonstrated by allele-specific PCR and restriction analysis (Figure 7), Chlamydomonas 788 (petC-1) mutants transformed with mutant petC alleles retained the defective 788 petC allele and carried, in addition, introduced mutant petC alleles. In this study, we have characterized petC mutations that restored phototrophy to the 788 (petC-1) mutant. Recovery of nonphototrophic transformants should be possible by cotransformation of a 788 (petC-1) arg^- mutant with ARG7 (Debuchy et al. 1989 ) and petC alleles.

Figure 9 shows the locations of the site-directed mutations on the soluble domain structure of the spinach chloroplast Rieske protein. S163 in the bovine mitochondrial Rieske protein forms a hydrogen bond with S-1 of the 2Fe–2S cluster and participates in a hydrogen bonding network that helps stabilize the reduced form of the cluster (Iwata et al. 1996 ). We constructed the mutations S128A, S128C, and S128T at the corresponding position in the chloroplast Rieske protein (Figure 6). S128T is a conservative substitution designed to test the mutagenesis strategy. The mutations S128A and S128C as well as W163A failed to restore phototrophic growth.

Structural analysis of bc₁ complexes (Kim et al. 1998 ; Zhang et al. 1998 ) suggests a catalytic domain movement of the Rieske protein that depends on flexibility of the region joining the membrane anchor to the soluble domain. We constructed the mutation 6G6A, which should make the flexible hinge or "neck" region (residues 40 to 50 in Figure 6) less flexible by replacing six glycine residues with six alanine residues. Cyanobacterial chloroplast Rieske proteins contain four to six conserved glycine residues in this region. Our 6G6A mutation had no effect on charge translocation or electron transfer (Table 2). Apparently several various structures of the neck region may provide adequate flexibility. The bc₁ Rieske proteins contain no glycines in this region but do carry three noncontiguous alanines. Recent mutational studies in Rhodobacter sphaeroides are consistent with a requirement for flexibility in the neck region (Tian et al. 1998 , Tian et al. 1999 ).

The unique and highly conserved 14- to 17-residue C-terminal extension of cyanobacterial chloroplast Rieske proteins contains several aromatic residues that may play roles in protein stability and perhaps electron transport (Gray and Winkler 1996 ; Carrell et al. 1997 ). To test the role of the b₆f Rieske tail region, we introduced a stop codon at position G171 that deletes seven residues from the C terminus (mutation G171och). Deletions of seven but not 14 residues from the C terminus of an overproduced Nostoc Rieske protein still allowed cluster integration in vitro (X. Wu, A. Tsapin, and T. Kallas, unpublished data). Our in vivo seven-residue deletion in the petC-G171och mutant showed no detectable functional change from the wild type. Therefore, these unique residues, including the aromatic W176 and W177, play no apparent role in cluster insertion—or in protein stability, electron transport, or subunit interactions in vivo. Interestingly, the site-directed aromatic mutation W163A fell on the same site as mutation W163R in ac21 mutant. Both petC-W163R and petC-W163A mutants showed little or no phototrophy. Analysis of the petC-W163R and petC-Y87D mutants indicates that certain aromatic residues are necessary for Rieske protein stability and assembly, as are specific aromatic residues of the yeast bc₁ protein (Snyder et al. 1999 ). Furthermore, we find that perhaps as a consequence of structural changes, electron transfer is considerably slowed in these mutants that assemble a portion of the mutant Rieske proteins into only partially functional b₆f complexes. The lack of a more pronounced phenotype in the petC-6G6A and petC-G171och mutants was surprising. Presumably there are selective reasons for conservation of residues in the flexible hinge and C-terminal extensions of chloroplast Rieske proteins, but these were not apparent from our analyses.

Data presented here establish mutational analysis of the nuclear-encoded and chloroplast-targeted Rieske protein and illustrate the advantages of Chlamydomonas as a model photosynthetic cell.

	METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Chlamydomonas reinhardtii Mutants and Genetic and RNA Analyses
Nuclear mutants deficient in the b₆f complex were isolated by using UV mutagenesis of wild-type strain CC125 (mt+; Kuras et al. 1997 ). The ac21 mutant has been isolated previously (Levine and Smillie 1962 ). Strains were grown on Tris-acetate-phosphate medium, pH 7.2, at 25°C under dim light (5 to 6 µE m^-2 sec^-1). Crosses, complementation analyses between nuclear mutants, and reversion and recombination tests were performed according to Kuras et al. 1997 . Total RNA analysis was performed according to Drapier et al. 1992 .

Plasmid Construction, Site-Directed Mutagenesis, and Nuclear Transformation
Molecular cloning techniques were performed as in Sambrook et al. 1989 . Escherichia coli DH5, DH5MCR (Gibco BRL, Gaithersburg, MD), and ES1301 (Promega, Madison, WI) were grown in Luria–Bertani medium supplemented as needed with 150 µg/mL ampicillin or 15 µg/mL tetracycline. Plasmids were introduced into E. coli by electroporation (Gene Pulser apparatus; Bio-Rad) at 12 kV/cm, 25 µF, and 200 settings. Plasmid pACR4.5 was constructed by joining a 4.5-kb BamHI DNA fragment containing the Chlamydomonas PETC gene to BamHI-digested, dephosphorylated plasmid pALTER-1 (Promega). Plasmid pACRi1Bam was created by site-directed mutagenesis (altered-sites procedure; Promega) of pACR4.5 with the oligonucleotide i1Bam (Table 3). The resulting ampicillin-resistant and tetracycline-sensitive pACRi1Bam plasmid carries a neutral BamHI restriction site tag in intron 1 of the PETC gene; it was used as the template for subsequent mutagenesis with the oligonucleotides described in Table 3. Mutant plasmids, which could be detected by restoration of tetracycline resistance, were subsequently screened for the specific restriction site associated with each directed mutation in PETC. Oligonucleotides were purchased in desalted form from Oligo Express (Paris, France). Derivatives of plasmid pACRi1Bam carrying mutations in PETC were purified from E. coli and introduced into the Rieske deletion 788 mutant (petC-1) by helium-driven particle bombardment (Kuras et al. 1997 ). Phototrophic transformants were selected on minimal medium at light intensities of 50 and 200 µE m^-2 sec^-1. Transformant colonies were visible after 1 to 2 weeks.

Allele-Specific Polymerase Chain Reaction and Restriction Analysis of Transformants
DNA templates were purified from 1.0 mL of late-log phase cultures or from colonies (~10 µL of cell volume). Briefly, cells were pelleted (colonies were suspended in 1.0 mL of water and then pelleted), resuspended in 100 µL of sterile water, and lysed by the addition of 300 µL of freshly prepared lysis buffer (100 mM Tris-HCl, pH 8, 0.4 M NaCl, 40 mM EDTA, 100 µg/mL RNase, 2%SDS) and incubated at 50°C for 15 min. This suspension was extracted with phenol:chloroform:isoamyl alcohol (25:24:1 mixture, v/v/v). DNA from the aqueous phase was precipitated with isopropanol, washed with 70% ethanol, and resuspended in 50 µL of 10 mM Tris-HCl, pH 7.5, and 0.1 mM EDTA buffer. Five-microliter samples were used in 50-µL reactions performed in a Crocodile III thermal cycler (Appligene, Illkirch, France) in a polymerase chain reaction (PCR) buffer containing 20% Qiagen (Chatsworth, CA) Q-solution and Taq polymerase. Amplification products were purified by the Qiaquick method of Qiagen. Allele-specific PCR primers were designed for amplification of the defective petC, 788 allele (primer 788TG) 5'-CGCTGCCTCGTCGGAGGC-3' of mutant 788, or of the corresponding PetC alleles that are wild type in exon 1 (primer WT107TG) 5'-GCTGCCTCGTCGGAGGTG-3'. The same reverse primer (C5aFS) 5'-GAGAGACAATCAACACGCGG-3' was used with each forward primer. At stringent annealing temperatures of 67 to 68°C, these primers allowed specific amplification and detection of the 788 petC allele and of wild type or introduced mutant petC alleles, both of which are wild type in exon 1.

Fluorescence Induction Measurements and Spectroscopic Analysis of Electron Transfer Kinetics in Vivo
Fluorescence measurements were performed with a home-built fluorometer by using exponential-phase algal cultures maintained in aerobic conditions by vigorous agitation in darkness before experiments. Fluorescence was triggered by continuous actinic light at 590 nm (intensity 125 µE m^-2 sec^-1). F_max values were determined in the presence of 10^-5 M DCMU.

For spectroscopic analysis, algae were harvested in the early exponential phase of growth (2 x 10⁵ to 2 x 10⁶ cells/mL) and resuspended in 50 mM Mops buffer, pH 7 (or 20 mM Hepes buffer, pH 7.2), containing 10% Ficoll, 10 µM DCMU, and 10 mM hydroxylamine to inhibit PSII activity (Joliot et al. 1998 ). Algae in the cuvette were maintained in anaerobic conditions under a stream of argon. FCCP was added at 1 µM to collapse the permanent transmembrane electrochemical potential. Electron transfer reactions were triggered by an actinic flash from a xenon lamp that was filtered by a dark-red filter (RG695). The intensity of the actinic flash was adjusted to excite less than one PSI per b₆f complex to obtain single turnover of these complexes. The interval between actinic flashes was 8 sec. Kinetics at 515 nm were measured to follow the electrochromic carotenoid band shift, induced by transmembrane charge-transfer reactions. Charge separation in PSI is reflected by a fast rise of the signal that was used to normalize kinetics. The slow phase of the carotenoid band shift was corrected for its decline by a linear approximation of the latter and fitted to an exponential curve to determine the kinetics of b₆f turnover. Cytochrome f kinetics were measured at 554 nm and were corrected by subtraction of 3% of the signal at 515 nm for the spectral contribution from the carotenoid band shift at 554 nm. The baseline was assumed to be linear between 545 and 573 nm.

Protein Isolation, Separation, Analysis, and Pulse-Labeling
Biochemical analyses were performed on cells grown to a density of ~2 x 10⁶ cells/mL. For polypeptide analysis, samples were resuspended in 100 mM 1,4-DTT and 100 mM Na₂CO₃ and solubilized in the presence of 2% SDS at 100°C for 50 sec. Polypeptides were separated on 12 to 18% SDS–polyacrylamide gels in the presence of 8 M urea, and heme staining was detected by peroxidase activity of heme binding subunits by using 3,3',5,5'-tetramethylbenzidine as described by Lemaire et al. 1986 . For immunodetection, proteins were electroblotted onto nitrocellulose sheets in a semidry system. Detection by using antisera raised against cytochrome b₆, cytochrome f, subunit IV, and ¹²⁵I-labeled protein A was conducted as given in Breyton et al. 1994 and by using the anti-ClpP antiserum (W. Majeran, F.-A. Wollman, and O. Vallon, manuscript in preparation). Anti-Rieske antiserum prepared against the entire polypeptide was kindly provided by Y. Pierre (Institut de Biologie Physico-Chimique). Thylakoid membrane proteins were purified as in Atteia et al. 1992 . Cytochrome complexes were extracted by Hecameg solubilization (at 30 rather than 25 mM) of thylakoid membranes (Pierre et al. 1995 ).

Whole cells (2 x 10⁶ cells/mL) were pulse-radiolabeled for 7 min as in Lemaire et al. 1986 . ¹⁴C-Acetate was used in the presence of 100 µg/mL chloramphenicol (an inhibitor of chloroplast protein synthesis) at 10^-4 M (5 µCi/mL). Labeling was terminated by dilution of the isotope with 10 volumes of chilled sodium acetate (50 mM).

	ACKNOWLEDGMENTS

We thank Sabeeha Merchant for the b₆f mutants; Yves Choquet, Jacqueline Girard-Bascou, Richard Kuras, and David Reardon for help with characterization of Rieske-deficient mutants; Yves Pierre for the anti-Rieske antiserum; Wojciech Majeran and Olivier Vallon for the ClpP-attenuated mutant and anti-ClpP antiserum, crossing petC-W163R and clpP-AUU mutants, and collaboration in petC-W163R clpP-AUU double mutant biochemical analysis; William Cramer for chloroplast Rieske protein coordinates; and Francis-André Wollman for help with fluorescence and ongoing input and support. This work was supported by the Centre National de la Recherche Scientifique (UPR 1261). F.B. was supported by the European Community Fellowship Grant No. ERBFMBICT960707. T.K. acknowledges the U.S. Department of Agriculture (Grant No. 9701960), University of Wisconsin sabbatical, and University of Wisconsin–Oshkosh faculty development grants, and thanks Francis-André Wollman and numerous others at the Institut de Biologie Physico-Chimique for their hospitality. C.d.V. and T.K. acknowledge a Vander Putten international exchange grant.

Received May 14, 1999; accepted August 3, 1999.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Atteia, A., de Vitry, C., Pierre, Y., and Popot, J.-L. (1992) Identification of mitochondrial proteins in membrane preparations from Chlamydomonas reinhardtii.. J. Biol. Chem. 267:226-234[Abstract/Free Full Text].

Berks, B.C. (1996) A common export pathway for proteins binding complex redox cofactors? Mol. Microbiol. 22:393-404[CrossRef][ISI][Medline].

Bevan, M., Bancroft, I., Mewes, H.W., Martienssen, R., and McCombie, R. (1999) Clearing a path through the jungle: Progress in Arabidopsis genomics. Bioessays 21:110-120[CrossRef][ISI][Medline].

Brasseur, G., Sled, V., Liebl, U., Ohnishi, T., and Daldal, F. (1997) The amino-terminal portion of the Rieske iron-sulfur protein contributes to the ubihydroquinone oxidation site catalysis of the Rhodobacter capsulatus bc₁ complex. Biochemistry 36:11685-11696[CrossRef][Medline].

Breyton, C., de Vitry, C., and Popot, J.-L. (1994) Membrane association of cytochrome b₆f subunits; the Rieske protein of Chlamydomonas reinhardtii is an extrinsic protein. J. Biol. Chem. 269:7597-7602[Abstract/Free Full Text].

Breyton, C., Tribet, C., Olive, J., Dubacq, J.-P., and Popot, J.-L. (1997) Dimer to monomer conversion of the cytochrome b₆f complex: Causes and consequences. J. Biol. Chem. 272:21892-21900[Abstract/Free Full Text].

Carrell, C.J., Zhang, H., Cramer, W.A., and Smith, J.L. (1997) Biological identity and diversity in photosynthesis and respiration: Structure of the lumen-side domain of the chloroplast Rieske protein. Structure 5:1613-1625[Medline].

Debuchy, R., Purton, S., and Rochaix, J.-D. (1989) The argininosuccinate lyase gene of Chlamydomonas reinhardtii: An important tool for nuclear transformation and for correlating the genetic and molecular maps of the ARG7 locus. EMBO J. 8:2803-2809[ISI][Medline].

Denke, E., Merbitz-Zahradnik, T., Hatzfeld, O.M., Snyder, C.H., Link, T.A., and Trumpower, B.L. (1998) Alteration of the midpoint potential and catalytic activity of the Rieske iron-sulfur protein by changes of amino acids forming hydrogen bonds to the iron-sulfur cluster. J. Biol. Chem. 273:9085-9093[Abstract/Free Full Text].

de Vitry, C. (1994) Characterization of the gene of the chloroplast Rieske iron-sulfur protein in Chlamydomonas reinhardtii: Indications for an uncleaved lumen targeting sequence. J. Biol. Chem. 269:7603-7609[Abstract/Free Full Text].

de Vitry, C., Choquet, Y., Baymann, F., Finazzi, G., and Kallas, T. (1998) Deficient mutants and site-directed mutagenesis of the nuclear-encoded chloroplast Rieske 2Fe–2S protein. In Photosynthesis: Mechanisms and Effects, Vol. 4, G. Garab, ed. Dordrecht, The Netherlands, Kluwer Academic Publishers, pp. 3151–3154.

Drapier, D., Girard-Bascou, J., and Wollman, F.-A. (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].

Finazzi, G., Buschlen, S., de Vitry, C., Rappaport, F., Joliot, P., and Wollman, F.-A. (1997) Function-directed mutagenesis of the cytochrome b₆f complex in Chlamydomonas reinhardtii: Involvement of the cd loop of cytochrome b₆ in quinol binding to the Q_o site. Biochemistry 39:2867-2874.

Gatti, D.L., Meinhardt, S.W., Ohnishi, T., and Tzagoloff, A. (1989) Structure and function of the mitochondrial bc₁ complex. A mutational analysis of the yeast Rieske iron-sulfur protein. J. Mol. Biol. 205:421-435[CrossRef][Medline].

Graham, L.A., Brandt, U., Sargent, J.S., and Trumpower, B.L. (1993) Mutational analysis of assembly and function of the iron-sulfur protein of the cytochrome bc₁ complex in Saccharomyces cerevisiae.. J. Bioenerg. Biomembr. 25:245-257[CrossRef][Medline].

Gray, H.B., and Winkler, J.R. (1996) Electron transfer in proteins. Annu. Rev. Biochem. 65:537-561[CrossRef][ISI][Medline].

Hippler, M., Drepper, F., Haehnel, W., and Rochaix, J.-D. (1998) The N-terminal domain of PsaF: Precise recognition site for binding and fast electron transfer from cytochrome c₆ and plastocyanin to photosystem I of Chlamydomonas reinhardtii.. Proc. Natl. Acad. Sci. USA 95:7339-7344[Abstract/Free Full Text].

Iwata, S., Saynovits, M., Link, T.A., and Michel, H. (1996) Structure of a water soluble fragment of the Rieske iron-sulfur protein of the bovine heart mitochondrial cytochrome bc₁ complex determined by MAD phasing at 1.5 Å resolution. Structure 4:567-579[Abstract/Free Full Text].

Iwata, S., Lee, J.W., Okada, K., Lee, J.K., Iwata, M., Rasmussen, B., Link, T.A., Ramaswamy, S., and Jap, B.K. (1998) Complete structure of the 11-subunit bovine mitochondrial cytochrome bc₁ complex. Science 281:64-71[Abstract/Free Full Text].

Joliot, P., Béal, D., and Delosme, R. (1998) In vivo measurements of photosynthetic activity: Methods. In Rochaix J.-D., Goldschmidt-Clermont M., Merchant S., eds. The Molecular Biology of Chloroplasts and Mitochondria in Chlamydomonas. Dordrecht, The Netherlands, Kluwer Academic Publishers. 433–449.pp.

Kallas, T. (1994) The cytochrome b₆f complex. In Bryant D.A., ed. The Molecular Biology of Cyanobacteria. Dordrecht, The Netherlands, Kluwer Academic Publishers. 259–317.pp.

Karnauchov, I., Herrmann, R.G., and Klosgen, R.B. (1997) Transmembrane topology of the Rieske Fe/S protein of the cytochrome b₆f complex from spinach chloroplasts. FEBS Lett. 408:206-210[CrossRef][ISI][Medline].

Keren, N., and Ohad, I. (1998) State transition and photoinhibition. In Rochaix J.-D., Goldschmidt-Clermont M., Merchant S., eds. The Molecular Biology of Chloroplasts and Mitochondria in Chlamydomonas. Dordrecht, The Netherlands, Kluwer Academic Publishers. 569–596.pp.

Kim, H., Xia, D., Yu, C.A., Xia, J.Z., Kachurin, A.M., Zhang, L., Yu, L., and Deisenhofer, J. (1998) Inhibitor binding changes domain mobility in the iron-sulfur protein of the mitochondrial bc₁ complex from bovine heart. Proc. Natl. Acad. Sci. USA 95:8026-8033[Abstract/Free Full Text].

Kindle, K.L. (1998) Nuclear transformation: Technology and applications. In Rochaix J.-D., Goldschmidt-Clermont M., Merchant S., eds. The Molecular Biology of Chloroplasts and Mitochondria in Chlamydomonas. Dordrecht, The Netherlands, Kluwer Academic Publishers. 41–61.pp.

Kuras, R., Wollman, F.-A., and Joliot, P. (1995) Conversion of cytochrome f to a soluble form in vivo in Chlamydomonas reinhardtii.. Biochemistry 34:7468-7475[CrossRef][Medline].

Kuras, R., de Vitry, C., Choquet, Y., Girard-Bascou, J., Culler, D., Buschlen, S., Merchant, S., and Wollman, F.-A. (1997) Molecular genetic identification of a pathway for heme binding to cytochrome b₆. J. Biol. Chem. 272:32427-32435[Abstract/Free Full Text].

Lemaire, C., Girard-Bascou, J., Wollman, F.-A., and Bennoun, P. (1986) Studies on the cytochrome b₆f complex. I. Characterization of the complex subunits in Chlamydomonas reinhardtii. Biochim. Biophys. Acta 851:229-238[CrossRef].

Levine, R.P., and Smillie, R.M. (1962) The pathway of triphosphopyridine nucleotide photoreduction in Chlamydomonas reinhardtii.. Biochemistry 48:417-421