INTRODUCTION

Studies in the unicellular green alga Chlamydomonas reinhardtii and vascular plants have demonstrated that the proteins required for photosynthesis are encoded by genes that reside in two distinct cellular compartments: the nucleus and the chloroplast (reviewed in Refs. 1-3). The nucleus encodes both structural polypeptides and proteins that play pivotal roles in the regulation of chloroplast gene expression (4-10) and in the maturation and assembly of the photosynthetic apparatus (11, 12). Although biochemical and genetic studies are beginning to lead to a better understanding of the roles that nucleus-encoded regulatory proteins play in these processes (7, 8, 10, 13), the identities of many of these proteins and the mechanisms by which they exert their influence remain largely unknown.

Defects in genes that regulate chloroplast gene expression or that are involved in the maturation or assembly of photosynthetic complexes are likely to cause a nonphotosynthetic phenotype. C. reinhardtii is an attractive organism for studying these genes, because nonphotosynthetic mutants are viable when provided with acetate as a carbon and energy source. With the ultimate goal of understanding the mechanisms by which nucleus-encoded factors regulate chloroplast gene expression and function, we have generated tagged nonphotosynthetic mutants, focusing initially on the isolation and characterization of nuclear genes required for the biogenesis of the cytochrome b₆f complex of C. reinhardtii.

The cytochrome b₆f complex carries out photosynthetic electron transfer between photosystem (PS)¹ II and PS I and is composed of both chloroplast- and nucleus-encoded subunits (14, 15). Its compositional simplicity relative to the PS II and PS I complexes makes it an attractive model for studying the assembly of chloroplast energy-transducing complexes. The chloroplast-encoded subunits of the cytochrome b₆f complex include products of petA (cytochrome f), petB (cytochrome b₆), petD (subunit IV), and two 4-kDa proteins encoded by the petG and petL genes (15-20). The nucleus-encoded components include the Rieske Fe-S protein (product of PetC; Ref. 21) and a 4-kDa protein (product of PetM; Refs. 22 and 23). The binding sites for the hemes and the Fe₂S₂ center are well defined, and each cofactor binding site lies within a single polypeptide. This feature makes the b₆f complex particularly suited for studies of cofactor-polypeptide association and assembly.

Previous studies have identified a number of nuclear genes required for the biogenesis of the cytochrome b₆f complex (9, 11, 14, 24, 25). Among these are a distinct subset of four nuclear loci that are also required for synthesis of cytochrome c₆.² Cytochrome c₆, a functional substitute for plastocyanin, is a soluble heme-containing protein that carries out electron transfer between the cytochrome b₆f complex and PS I; its expression is induced under copper-deficient conditions (11, 26). Cytochrome c₆ and cytochrome f are c-type cytochromes in which the heme prosthetic group is covalently attached to the apoproteins. While cytochrome f is chloroplast-encoded, cytochrome c₆ is encoded by the nuclear Cyc6 gene, synthesized in the cytosol, and imported post-translationally into the chloroplast (26, 27). Thus, there are few biosynthetic steps that are common for cytochrome c₆ and cytochrome f besides translocation across the thylakoid membrane, lumenal processing, and heme attachment.

In previous work, we identified several mutants that display a cytochrome b₆f/cytochrome c₆ phenotype, and biochemical characterization confirmed that the strains were able to translocate and process precursor and intermediate forms of cytochrome f and c₆ but were unable to convert the apocytochromes to their respective holoforms (11, 28). This work defined a specific biochemical phenotype for heme attachment mutants (cytochrome b₆f/cytochrome c₆) and suggested that a common set of proteins is required for holocytochrome formation in the thylakoid lumen. One of these is the product of the chloroplast-encoded ccsA gene, which is required for the biosynthesis and accumulation of the cytochrome b₆f complex and cytochrome c₆ (29). Loss of CcsA function results in a defect at the step of heme attachment to apocytochromes c₆ and f.

In the present study, we describe the isolation and characterization of a nuclear gene, Ccs1 (c-type cyt synthesis), which is required for assembly of the cytochrome b₆f complex and cytochrome c₆. On the basis of a pleiotropic c-type cytochrome deficiency in a strain lacking functional Ccs1, we suggest that its gene product functions in the same pathway as CcsA.

EXPERIMENTAL PROCEDURES

C. reinhardtii strain nit1-305 cw15, which is the wild type for photosynthesis, was used as a recipient for nuclear transformations. C. reinhardtii strain CC-125 (wild type) was obtained from the Chlamydomonas Genetics Center (Duke University, Durham, NC). Strain B6, isolated by Gal et al. (30), does not accumulate cytochrome c₆ or the cytochrome b₆f complex due to a frameshift mutation in the chloroplast ccsA gene (11, 29, 31). CC-125 and B6 were used as positive and negative control strains, respectively, for the analyses of cytochrome c₆ and cytochrome f synthesis. E. coli strains XL1-Blue and DH5 were used as hosts for recombinant DNA techniques (32).

Cells were grown in Sager-Granick medium (33) supplemented with ammonium nitrate (SGII-NH₄) under continuous light (50 microeinsteins/m² s) and transformed by the glass bead method (34). 0.5 µg of plasmid pNIT1 DNA (35) linearized with EcoRI was used per transformation, after which the cells were plated on SGII agar plates supplemented with potassium nitrate (SGII-NO₃) and grown under dim light (5 microeinsteins/m² s) until colonies were visible (~3 weeks). The colonies were then picked onto plates to screen for nonphotosynthetic mutants by growth under the following conditions: 1) acetate-containing SGII-NO₃ plates in dim light (permissive condition), and 2) acetate-deficient SGII-NO₃ plates in bright light (100 microeinsteins/m² s; restrictive condition). Transformation experiments to demonstrate complementation of the nonphotosynthetic phenotype were performed with 5 µg of  DNA, 1 µg of plasmid DNA, or no DNA, as described. Photosynthetic colonies were selected on acetate-free SGII-NO₃ agar plates.

For analyses of cellular proteins, cells were grown in high salt acetate, Tris-acetate phosphate (33) or copper-free Tris-acetate phosphate (36) media to exponential to early stationary phase (2-6 × 10⁶ cells/ml). For immunoblots of total cell extracts, cells were harvested by centrifugation and boiled for 5 min in loading buffer (0.0625 M Tris-HCl, pH 6.8, 10% glycerol, 2% SDS, 1% -mercaptoethanol, 0.00005% bromphenol blue). After centrifugation, 25 µl of the supernatant was loaded into a 12 or 15% SDS-polyacrylamide gel, and the proteins were separated by electrophoresis as described previously (27). Proteins were then transferred onto nitrocellulose paper using a Bio-Rad Transblot semidry apparatus (Bio-Rad). Antibodies against cytochrome f and cytochrome b₆ were obtained from Dr. R. Malkin (University of California, Berkeley, CA). The antibody against subunit IV (SuIV) has been described previously (37). The antibody against the D2 protein of PS II was obtained from Dr. M. Kuchka (Lehigh University, Bethlehem, PA). These primary antibodies were used at a dilution of 1:1000. The ECL chemiluminescent system (Amersham Life Science, Inc.) was used to detect immunoreactive proteins.

For detection of cytochromes, cells were collected by centrifugation (3,000 × g, 5 min), washed in 10 mM sodium phosphate (pH 7.0), and resuspended to a chlorophyll concentration of 1 mg/ml in the same buffer. Cells were lysed by two cycles of slow freezing to 80 °C followed by thawing to room temperature. The soluble cell extract was separated from the insoluble membrane fraction by centrifugation (16,000 × g) at 4 °C for 10 min, and both fractions were returned to the original sample volume with buffer. The proteins were separated by nondenaturing or denaturing gel electrophoresis as described previously (11), except that the denaturation buffer contained 50 mM dithiothreitol instead of -mercaptoethanol, and the samples were denatured on ice to preserve the integrity of the heme group of the cytochromes. Proteins were then transferred to Immobilon P polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA) at 4 °C, and heme-containing proteins were identified by a heme staining procedure using chemiluminescent reagents (38). Membranes were immersed in the Supersignal Chemiluminescent Substrate CL-HRP (Pierce) for 30 s and exposed to film (Reflection, NEN Life Science Products). Following exposure to film, membranes were rinsed with TBS buffer (40 mM Tris, pH 7.5, 150 mM NaCl) and subjected to immunoblot analysis using antibodies against plastocyanin (1:5000), cytochrome c₆ (1:1000), and cytochrome f (1:5000) as a primary antibody. Bound primary antibody was detected with an alkaline phosphatase-conjugated secondary antibody for cytochrome immunoblots and a horseradish peroxidase-conjugated secondary antibody for plastocyanin immunoblots.

To monitor cytochrome f synthesis, cells were radiolabeled as described previously (39), with the following modifications. The cells were pulse-labeled for 10 min with Na₂³⁵SO₄ followed by a 40-min chase in the presence of unlabeled sulfate and chloramphenicol (250 µg/ml). The initial immunoprecipitation of cytochrome f is essentially quantitative in the case of extracts from mutant strains, while for wild-type extracts, approximately 50% of total cytochrome f is removed in the first immunoprecipitate.

Isolation of plasmid DNA was performed either by the boiling method (40) or by a modified alkaline lysis procedure (41). C. reinhardtii DNA was isolated as described previously (42). C. reinhardtii total RNA was prepared either by the method of Shepherd et al. (43) or as described by Merchant et al. (27), while poly(A)⁺ RNA was isolated as described previously (44). Southern blotting and hybridization of C. reinhardtii DNA to radioactive probes was performed as described previously (35). RNA blot hybridization was performed as described by Hill et al. (45). petA mRNAs were identified by hybridization to a 600-bp HindIII/AccI radiolabeled fragment from plasmid pHA0.6 (16), and mRNAs encoding cytochrome c₆ (Cyc6) were identified by hybridization to a 710-bp EcoRI radiolabeled fragment from plasmid pGEM1Crc552 7A-4 (46). The RbcS2-specific probe consisted of the 0.8-kb EcoRI cDNA insert of the plasmid p149A, which was obtained from the Chlamydomonas Genetics Center. 5-rapid amplification of cDNA ends experiments were performed using poly (A)⁺ RNA from C. reinhardtii strain nit1-305 cw15, Ccs1-specific primers (5-AAGCTCGAGAACTGGGC-3 and 5-ATGAAGGTGCCCAGGCC-3), and the Superscript Preamplification System (Life Technologies, Inc.) according to the manufacturer's instructions.

Genomic DNA was digested with SalI, ligated, and transformed by electroporation into E. coli (47). A fragment from the rescued plasmid was used as a probe to screen a C. reinhardtii genomic library comprising strain 21gr DNA in FIX II (Stratagene, La Jolla, CA); this library was obtained from Dr. R. Schnell and Dr. P. Lefebvre (University of Minnesota, St. Paul, MN; Ref. 48).

A C. reinhardtii cDNA library made from RNA isolated from vegetative cells of strain CC-621 in ZAP II (Stratagene) was obtained from Dr. J. Woessner and Dr. U. Goodenough (Washington University, St. Louis, MO; Ref. 49). The library (8 × 10⁴ independent plaques) was screened with a mixture of probes made from the 0.45-kb SphI and 1.5-kb SphI/KpnI fragments (probes 2 and 3, Fig. 4) of pCcs1-1. Since we were unable to convert the cDNA insert from the single positive plaque into a plasmid by superinfection with a helper, the cDNA insert was polymerase chain reaction-amplified (30 cycles of the following: 92 °C for 30 s; 45 °C for 1 min; 72 °C for 2 min) using T3 and T7 primers. The fragment was digested with EcoRI/ScaI and XhoI/ScaI to produce 0.5-kb ScaI/EcoRI and 2.0-kb XhoI/ScaI fragments, which were then subcloned into pBluescript KSII (the ScaI site is located within the cDNA, while the terminal EcoRI and XhoI sites are derived from the ZAP II vector).

[View Larger Version of this Image (25K GIF file)]

Plasmid pCcs1-1 was digested with SacI and religated to make the plasmid pCcs1-4 in which the 5-region of genomic DNA in pCcs1-1 was deleted. pCcs1-4 was used as a template for genomic sequencing. Nested deletions of plasmids containing cDNA fragments were made using the Erase-a-Base kit (Promega Corp., Madison, WI). Plasmids were sequenced using the Silver Sequence DNA sequencing system (Promega) or by dye terminator cycle sequencing using 3 dye-labeled dideoxynucleotide triphosphates and run on an ABI PRISM 377 DNA sequencer (Perkin-Elmer). The DNA sequence was derived from sequencing the entire length of both strands of the cDNA and genomic clones.

RESULTS

Nuclear transformation was used to generate insertion mutations in genes required for biogenesis of the cytochrome b₆f complex of C. reinhardtii. Approximately 2500 independent transformants were grown on acetate-supplemented plates and subsequently screened for mutants that had lost the ability to grow phototrophically. Eight nonphotosynthetic transformants were identified and screened further for defects in accumulation of the cytochrome b₆f complex by immunoblot analysis using an antibody against subunit IV (SuIV) to probe total cell extracts. Since strains that carry mutations affecting the accumulation of cytochrome f, cytochrome b₆, or SuIV also fail to accumulate the other subunits of the complex (14, 37, 50), the abundance of SuIV is diagnostic of the entire cytochrome b₆f complex. Four of the eight strains accumulated drastically reduced amounts of SuIV (strains 3, 9, 30, and 34; Fig. 1). These strains were designated as abf (accumulation of the b₆f complex) mutants. The abf mutants did not accumulate cytochrome b₆ or cytochrome f, as expected (Fig. 1). All abf mutants exhibited wild-type levels of the D2 protein of PS II, which suggests that these mutations do not affect the accumulation of cytochrome b₅₅₉, which is an integral component of PS II.

[View Larger Version of this Image (33K GIF file)]

We next performed a Southern blot analysis to determine whether the mutants exhibited an RFLP that might indicate an insertion in either PetC and PetM, the two known nuclear genes that encode subunits of the complex. Since no PetC or PetM RFLPs were observed (data not shown), we concluded that the lesions in the abf mutants most likely affected 1) a previously unidentified nucleus-encoded subunit of the complex, 2) a regulatory gene required for the expression of structural genes of the complex, or 3) a gene required for the maturation or assembly of the complex.

Nuclear mutations have defined four loci, CCS1-CCS4, involved specifically in the biogenesis of chloroplast c-type cytochromes.² These strains exhibit a pleiotropic c-type cytochrome deficiency that is characteristic of a defect at the post-translational step of heme attachment. To determine whether any of the abf mutants harbored lesions of this type, the mutants were screened for cytochrome c₆ accumulation by performing immunoblots, which assess polypeptide abundance, and heme staining, which assesses holocytochrome abundance. Cell extracts were isolated from cells grown in copper-deficient medium to induce transcription of Cyc6, which is repressed by copper (27, 51). The B6 mutant of C. reinhardtii, which contains a mutation in the chloroplast ccsA gene and does not accumulate cytochrome f or cytochrome c₆ (29), was included as a control. As shown in Fig. 2, A and B, holocytochromes c₆ and f do not accumulate in either abf3 or B6. To confirm that the cells were copper-deficient and hence under conditions allowing expression of the Cyc6 gene, the extracts were tested for plastocyanin content, since cells accumulate plastocyanin only when the medium contains copper. Since little plastocyanin was detected under copper-deficient conditions (Fig. 2B, bottom; compare lanes 2, 4, and 6 with lanes 1, 3, and 5), the cells were indeed copper-deficient. Moreover, Cyc6 transcripts accumulated under these conditions (Fig. 2C). Hence, the cytochrome c₆ deficiency could not result simply from lack of Cyc6 expression due to copper contamination in the medium. The mutation did not affect accumulation of petA mRNA (Fig. 2C), and pulse-chase labeling experiments demonstrated that abf3, like B6 and previously described heme attachment mutants (11), synthesizes but is unable to accumulate cytochrome f (Fig. 2D). The presence of plastocyanin and the PS II-associated Oee1 protein (data not shown) in extracts of cells grown in the presence of copper confirms that the synthesis of other lumenal proteins is not affected in these strains. Thus, it is unlikely that abf3 is defective in some component of the chloroplast-SecA thylakoid transport pathway (52, 53). Together, these data suggest that abf3 most likely harbors a mutation(s) in a gene(s) required for the attachment of heme to chloroplast c-type cytochromes.

[View Larger Version of this Image (28K GIF file)]

A Southern blot analysis of DNA isolated from abf3 that was hybridized with a Nit1-specific probe revealed that the mutant harbored a single insertion of pNIT1 (data not shown). Since the mutation in abf3 was presumably a consequence of this plasmid DNA insertion, we attempted plasmid rescue to isolate the DNA that flanked the insertion site in abf3 (47). The recovered plasmid contained a 1.3-kb C. reinhardtii DNA fragment. When a genomic DNA blot was probed with this fragment, an RFLP was detected between wild type and abf3 (Fig. 3A, lanes 1 and 2). If the nonphotosynthetic phenotype was indeed due to the insertion, the RFLP should cosegregate with the mutant phenotype in genetic crosses. However, crosses between abf3 and an appropriate marker strain did not yield viable progeny. Therefore, a molecular genetic approach was pursued. To isolate a wild-type genomic sequence corresponding to the rescued DNA fragment, we screened a bacteriophage  genomic library with the rescued fragment. Of the six positive clones, two (III-4 and III-7) contained DNA that complemented the photosynthetic defect in abf3 (Fig. 3B). Southern blot analysis of DNA isolated from wild type cells, abf3, and a transformant rescued by one of the  clones (abf3::III-4) provided molecular evidence of complementation; abf3::III-4 contains both the introduced wild-type gene as well as the RFLP due to the insertion (Fig. 3A, lane 3). Subsequent subcloning and transformation experiments delimited the complementing region to a 6.5-kb DNA fragment (pCcs1-2 in Fig. 3B). Control transformation experiments with no DNA did not yield spontaneous photosynthetic revertants of abf3. The phototrophic transformants accumulated cytochrome f, as assessed by immunoblots and heme staining (Fig. 3C), and as expected, the complemented strains also accumulated wild-type levels of both cytochrome b₆ and SuIV (data not shown). Since selection for phototrophic growth relies only on restoration of cytochrome f function, complemented transformants were also tested for cytochrome c₆ accumulation by immunoblot and heme stain analyses. As shown (Fig. 3D), abf3 strains that had been complemented with either III-4 or plasmid pCcs1-1 also accumulated wild-type levels of cytochrome c₆. These results demonstrated that a single nuclear mutation resulted in the deficiencies of the cytochrome b₆f complex and cytochrome c₆ in abf3. We conclude that the product of this nuclear gene is essential for the biogenesis of both chloroplast c-type cytochromes. The gene has been designated Ccs1 (c-type cytochrome synthesis).

[View Larger Version of this Image (21K GIF file)]

The primary sequence of the protein potentially encoded by the Ccs1 gene was determined by sequencing a cDNA clone isolated from a  ZAP library. We screened 8 × 10⁴ plaques and identified a single positive plaque, which was significantly smaller than the other plaques. Although the  ZAP vector system was designed to permit the direct recovery of plasmids containing cDNA inserts, we were unable to accomplish this with Ccs1. Therefore, the cDNA insert was obtained by amplification of the phage DNA using vector-specific primers and polymerase chain reaction. Attempts to clone the amplified insert into a plasmid vector also proved unsuccessful. The small size of the positive plaque and our inability to recover a cDNA-containing plasmid suggested that the cDNA might be expressed and toxic to E. coli. These problems were circumvented by subcloning the polymerase chain reaction-amplified cDNA insert as two separate fragments. The resultant plasmids were used as templates to determine the nucleotide sequence of the cDNA. The cDNA sequence was confirmed by determining the sequence of the genomic DNA in pCcs1-3.

Alignment of the genomic and cDNA nucleotide sequences revealed that Ccs1 consists of 10 exons and 9 introns (Fig. 4A). Southern blot analysis revealed that Ccs1 is a single copy gene in C. reinhardtii (data not shown). We also determined the nucleotide sequence of the region spanning the site of the pNIT1 insertion in abf3 and found that the plasmid had inserted into the 10th exon of the gene, 70 bp from the termination codon (shown by an arrowhead in Fig. 4A). Southern blots revealed that the insertion in abf3 was accompanied by a deletion of approximately 200 bp of C. reinhardtii DNA upstream of the insertion site (data not shown). These alterations prevented the accumulation of Ccs1 transcripts (Fig. 4B).

The Ccs1 cDNA contains an ORF of 1839 nucleotides capable of encoding a 613-amino acid protein with a calculated molecular mass of 64.9 kDa (Fig. 5). The initiation codon is preceded by a short A-rich region (underlined sequence in Fig. 5) that is typical of many nuclear genes in C. reinhardtii (54-57) and vascular plants (58). 5-Rapid amplification of cDNA ends of poly (A)⁺ RNA isolated from wild-type cells showed that the 5-end of the mRNA contained approximately 120 additional nucleotides upstream of the 5-end of the cDNA. However, an examination of the nucleotide sequence of this region did not reveal alternative initiation codons. A consensus polyadenylation addition site (TGTAA) was not found in the 3-untranslated region, despite the fact that the cDNA is intact as evidenced by the presence of a poly(A) tract at the 3-end.

[View Larger Version of this Image (61K GIF file)]

A search of the protein data bases revealed five proteins of unknown function with significant similarity to Ccs1 (Fig. 6). Three of the proteins are ORFs encoded by ycf44 (hypothetical chloroplast frame) genes in oxygen-evolving photosynthetic organisms including the cyanobacterium Synechocystis strain PCC 6803 genome (33% identity; Ref. 59), the chloroplast genomes of the red alga, Porphyra purpurea (26% identity; Ref. 60), and the brown alga Odontella sinensis (25% identity; Ref. 61). The fourth and fifth proteins, the resB gene product of Bacillus subtilis (62) and an ORF in the genome of Mycobacterium leprae,³ exhibit much less similarity to Ccs1 (11 and 12% similarities, respectively). Alignment of these proteins revealed two domains that are unique to Ccs1: an N-terminal extension, and an alanine plus serine-rich region in the central portion of the polypeptide (positions 385-428). Overall, the hydropathy profiles of Ccs1, ResB, and the Ycf44 proteins display remarkable similarity (Fig. 7). The N-terminal part of the extension resembles the stroma-targeting domains of chloroplast transit peptides, since it has a net positive charge, and the sequence VRC at positions 36-38 corresponds to the cleavage site ((V/I)X(A/C)) of proteins that are imported into chloroplast (63). However, there is an acidic residue and a tyrosine in this region, which are unusual in transit peptides. The remainder of the N-terminal extension has numerous charged residues and no extended hydrophobic region that might serve as a lumen-targeting domain, so it is unclear whether it plays any role in intrachloroplast localization.

[View Larger Version of this Image (110K GIF file)]

[View Larger Version of this Image (23K GIF file)]

DISCUSSION

We have isolated a C. reinhardtii mutant, abf3, that is deficient in both the cytochrome b₆f complex and cytochrome c₆. Since an intact copy of the Ccs1 gene, which is disrupted in abf3, complements both deficiencies, we conclude that Ccs1 is required for the biogenesis of both the cytochrome b₆f complex and cytochrome c₆. abf3 belongs to a class of previously described C. reinhardtii nuclear mutants defective in one of the final steps of chloroplast c-type cytochrome biosynthesis, e.g. heme transport across the thylakoid membrane, maintenance of protoheme and the apocytochrome heme-binding site in a reduced state, or covalent attachment of the heme prosthetic group to the apocytochrome (11). Mutations in these genes prevent formation of the holocytochromes and result in the accelerated turnover of the nonfunctional apocytochromes. Neither the integrity of PS II (assessed by the accumulation of D2; Fig. 1) nor its photochemical activity (assessed by the rate of oxygen evolution),⁴ is significantly affected in the ccs1 mutant. Since PS II requires the heme-containing cytochrome b₅₅₉, it is unlikely that heme is disrupted. It is also unlikely that the mutation affects a general process like protein import, processing, thylakoid membrane insertion, or lumen translocation by the SecA pathway. For these reasons, we believe that the Ccs1 gene encodes a polypeptide required for covalent attachment of heme to chloroplast c-type cytochromes.

The lack of similarity between Ccs1 and gene products with known activities has precluded homology-based assignment of a specific biochemical function in the heme attachment pathway. However, the similarity of Ccs1 to the resB gene product of B. subtilis is consistent with our conclusion that Ccs1 does indeed encode a product involved in the biogenesis of c-type cytochromes. In B. subtilis, the resA, resB, and resC genes belong to the same transcriptional unit, suggesting that they function in a common pathway. The resA gene product is similar to protein disulfide isomerases and thioredoxin-like proteins, the latter of which have been shown to be required for the biogenesis of c-type cytochromes in Rhodobacter capsulatus and Bradyrhizobium japonicum (the helX and the tlpB gene, respectively; Refs. 64-66). Similarly, the resC gene product is related to CcsA and, like CcsA, contains a sequence motif that is found in a family of proteins that are also required for the biosynthesis of c-type cytochromes in bacteria (Ccl1/CycK in R. capsulatus/B. japonicum; Refs. 65 and 66). The presence of the resB gene within this operon is a strong indication that resB and its homologues, Ccs1 and ycf44, also encode proteins involved in the biogenesis of c-type cytochromes, and the phenotype of the abf3 strain confirms the function of one member of this family. Although the putative homologues of CcsA and Ccs1 in M. leprae have not been studied, they are organized in an operon as in B. subtilis.

It is very likely that the Ccs1 gene product is chloroplast-localized, because 1) it has an N-terminal extension relative to Ycf44 that might serve as a chloroplast import signal; 2) the biosynthesis of holocytochromes c₆ and f share common steps only within the chloroplast; and 3) the homologous ycf44 gene products are encoded in the plastid genomes of the brown and red algae O. sinensis and P. purpurea, respectively (60, 61). The difference in the location of these related genes suggests that they were translocated from the plastid to the nuclear genome in C. reinhardtii. Based on their hydrophobic nature (Fig. 7), it is tempting to speculate that these genes encode membrane-associated proteins. We are currently generating antibodies against Ccs1 to permit the localization of the protein by subcellular fractionation and immunolocalization.

Although very little is known about the mechanisms involved in the biosynthesis of c-type cytochromes within the chloroplast, there are clearly similarities with the bacterial pathway. For instance, a common set of gene products participates in the ligation of heme to both membrane-bound and soluble c-type cytochromes in chloroplasts and bacteria (26, 64-66). This is in contrast to the situation in mitochondria, where two distinct cytochrome c heme lyases are required for the attachment of heme to the soluble cytochrome c and membrane-associated cytochrome c₁ (reviewed in Ref. 67). Furthermore, while there appears to be no relationship between mitochondrial cytochrome c heme lyases and the products of cytochrome biogenesis genes in chloroplasts and bacteria, there are clear relationships between at least two sets of genes involved in the biogenesis of c-type cytochromes in chloroplasts and bacteria, specifically, between ccsA and cycK/ccl1 (29, 64) and between Ccs1/ycf44 and resB (this work).

In C. reinhardtii, there is at least one chloroplast locus (ccsA) and four nuclear loci (CCS1-CCS4) involved in the biogenesis of chloroplast c-type cytochromes.² Clearly, interactions between these nucleus- and chloroplast-encoded gene products are involved in the process of c-type cytochrome biogenesis. Ultimately, we would like to understand the roles that these nucleus- and chloroplast-encoded proteins play in the biogenesis of chloroplast c-type cytochromes. The genetic approaches outlined in this paper, combined with biochemical analyses of the proteins and pathways, hold the promise of supplying answers to these questions.