| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 31, 32474-32482, July 30, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶
From the
Department of Chemistry and Biochemistry and the ||Department of Molecular, Cell, and Developmental Biology, UCLA, Los Angeles, California 90095, the **Division of Biological Sciences, Section of Plant Biology, University of California, Davis, California 95616, and the Genomic Analysis Laboratory, The Salk Institute for Biological Studies, La Jolla, California 92037
Received for publication, April 19, 2004 , and in revised form, May 18, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
In bacteria, c-type cytochromes are assembled in the periplasm via system I or system II (10, 11). In the bacterial periplasm, a disulfide bond formation machinery (Dsb system in Gram-negative and Bdb system in Gram-positive) promotes the oxidative folding of proteins, which is counter to the requirement for apocytochrome cysteinyl thiols in the heme attachment reaction (12, 13). The current thinking is that apocytochromes c upon translocation to the p-side are first a substrate of the Dsb/Bdb system which introduces intramolecular disulfides between the two cysteines of the CXXCH sequence. Subsequently, the disulfide bond undergoes reduction to provide free sulfhydryls that are competent to react with heme.
The requirement for redox chemistry in the context of in vivo cytochrome c assembly was appreciated through the discovery in bacteria of thiol-metabolizing components in systems I and II that constitute a transmembrane thio-reduction pathway (12–14). The components of this pathway include a sulfhydryldisulfide membrane transporter that conveys reducing power from the cytoplasm to the periplasmic space and a dedicated membrane-anchored thioredoxin that reduces the heme binding site of apocytochrome c prior to the heme attachment (13–15).
Plastid c-type cytochromes are assembled through system II (5, 16–21), but the requirement for redox chemistry in cytochrome maturation has remained so far largely unexplored in this organelle. We suspected the operation of a redox subpathway in the plastid cytochrome assembly based on the occurrence of a CcdA homolog encoded in the plastid genome of a red alga, Porphyra purpurea, and in the nuclear genomes of green alga Chlamydomonas reinhardtii and vascular plant Arabidopsis thaliana. CcdA is related to the DsbD/DipZ family of membrane polytopic proteins (22) that functions in transmembrane transfer of thiol-reducing equivalents from the bacterial cytoplasm to the periplasm (12, 13, 23). Loss-of-function of CcdA in bacteria results in a pleiotropic deficiency in c-type cytochromes, presumably because of a block in the reduction of apocytochrome c thiols in the heme binding site (24–28). Interestingly, CcdA is physically linked to a gene encoding a homolog of Ccs1, a system II-specific cytochrome c biogenesis (19) in P. purpurea, Mycobacterium spp., and several cyanobacteria (6). This gene clustering speaks to a role for CcdA in the plastid cytochrome c maturation pathway, but no biochemical or genetic evidence is presently available to assign a function to the plant or cyanobacterial CcdAs.
In this work, we investigated the function of Arabidopsis CCDA in plastid cytochrome c assembly. In vitro import experiments and topological analysis using PhoA and LacZ established that the CCDA polypeptide is a polytopic membrane protein that is targeted to the plastid organelle. We have isolated T-DNA and Ds (dissociation) element insertions in the CCDA gene and shown that loss-of-function alleles resulted in photosynthetic incompetence at the level of cytochrome b6f assembly. We propose that impaired accumulation of cytochrome b6f is a consequence of a defect in holocytochrome f maturation, a c-type cytochrome resident in the plastid lumen. We hypothesize that CCDA is a component of a thylakoid transmembrane thioreduction pathway that operates in the delivery of reducing equivalents to apocytochrome f prior to its conversion to the holoform.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
were used as hosts for recombinant techniques. Strain LMG194 (phoA–) was used to measure PhoA activity of the sandwich constructs, and strain CC118 (phoA– lacZ–) was used to measure PhoA and LacZ activities. The sources of DNAs used in this work are provided in the on-line supplement. Constructs for Chloroplast Import Experiments—The AtCCDA, NAtCCDA-DHFR,3 and N-AtPC-DHFR constructions were designed in the pGEM2 vector. AtCCDA corresponds to the entire ORF of the CCDA gene. N-AtCCDA-DHFR expresses the first 141 amino acids of Arabidopsis CCDA fused to the entire coding sequence of the mouse DHFR. N-AtPC-DHFR was engineered by fusing the sequence encoding the 77 amino acids of the plastocyanin chloroplast-targeting sequence plus the first 5 residues of the mature protein to the DHFR sequence. All gene fusions were under the control of the SP6 promoter that enables in vitro transcription. The details of the molecular cloning are described in the supplemental material.
Chloroplast Import Experiments—In vitro protein import reactions were performed with chloroplasts isolated from pea seedlings (29). Radiolabeled proteins were translated in wheat germ extract (Promega). Reactions were performed in the light for 20 min prior to thermolysin treatment (29).
Construction of CCDA-phoA, CCDA-lacZ, and CCDA-phoA Sandwich Fusions—Plasmid pRGK200 (30) was used to generate seven CCDA-phoA and seven CCDA-lacZ translational fusions (I–VII) by a PCR-based strategy. The phoA sandwich fusion approach consists of the design of fusion proteins where PhoA is inserted into an otherwise intact target protein (31). Two CCDA-phoA sandwich fusions carrying an insertion of phoA into CCDA at position 202 (SWA) or 242 (SWB) of the CCDA protein were constructed in the expression vector pBAD22, a plasmid that harbors an arabinose-inducible phoA gene (32). The details of the molecular cloning are provided in the supplemental material.
Measurement of PhoA and LacZ Activities—PhoA assays were performed as described previously in (33), except that isopropyl-1-thio-
-D-galactopyranoside was not used for induction. LacZ activity was measured as detailed in Ref. 34. PhoA assays of the sandwich fusions were performed on overnight grown cultures supplemented with 100 µM L-arabinose for induction. PhoA activity expressed in Miller units was calculated using Equation 1,
 | (Eq. 1) |
 | (Eq. 2) |
Growth of A. thaliana—Col-0 seeds originated from Lehle Seeds, Round Rock, TX. Plants were propagated on solid minimal media comprising either the commercial soluble fertilizer Peters Professional General Purpose 20-10-20 (35) (Western Farm Services) or 0.5x Murashige & Skoog (MS) minimal medium, solidified with 0.8% (w/v) Select agar (Invitrogen), and containing 0.5 or 5% sucrose, or on a soil mix (Sunshine Professional Basic Mix 2; Sun Gro, Canada). Plants were grown at 24 ± 0.5 °C under white light (80–120 µE/m2/s) with a light/dark cycle of 16/8 h.
Extraction of Plant Genomic DNA and mRNAs—A. thaliana genomic DNA for Southern hybridization was prepared as described in Ref. 36, except that plant material was frozen in liquid nitrogen and ground to a powder prior to extraction. For PCR amplification, genomic DNAs were extracted as in Ref. 37. Total RNA from plants was isolated with an RNeasy Plant Mini Kit (Qiagen) starting from 100 mg of plant tissue according to the manufacturer's procedure. Following the elution step, purified RNAs were treated for 1 h with RNase-free DNase (Stratagene) to remove any contaminating genomic DNA and phenol-chloroform extracted. To confirm the absence of any contaminating DNA, 50 ng of RNA was subjected to a PCR amplification using the same primers that were used in the RT-PCR analysis.
PCR and RT-PCR Amplifications—PCR amplifications performed as part of the University of Wisconsin-Madison screening process were carried out using TAKARA Ex-Taq under the conditions recommended by the manufacturer. The reverse transcription reaction was performed using the RT system (Promega); see the supplemental material.
Screening and Molecular Characterization of ccda Mutant Lines— The PCR-based service provided by the Arabidopsis Knockout Facility at the University of Wisconsin-Madison Biotechnology Center was used to screen 60,480 T-DNA insertion lines, and one insertion in the CCDA gene was identified (ccda-1). The ET5250 line carrying a Ds element insertion (ccda-2), the SALK_025766 line (ccda-3) and SALK_014416 line (ccda-4), both harboring a T-DNA insertion in CCDA, were identified through querying data bases of sequenced insertion sites (Cold Spring Harbor laboratory and Salk Institute Genomic Analysis Laboratory, respectively). The details of the screening and molecular characterization of the mutant lines are available in the supplemental material.
Fluorescence Measurements—The fluorescence imaging system, Fluorcam 700 MF (Photon Systems Instruments Ltd.), was used to record chlorophyll fluorescence induction and decay kinetics of Arabidopsis plants that had been dark-adapted for at least 5 min. Plants were illuminated for 20 s under actinic light of 50 µE/m2/s. Emitted fluorescence was captured by the camera, and the collected data were converted into a graph using Excel software (Microsoft).
Protein Preparation and Analysis—To extract proteins, plant tissue was frozen in liquid nitrogen, ground to a powder, and ground further in ice-cold extraction buffer (50 mM Tris-HCl, pH 8.0, 5 mM EDTA). The homogenates were centrifuged at 13,000 x g for 15 min at 4 °C, and the pellets, enriched in membrane proteins, were washed twice and resuspended in extraction buffer (0.5–2.0 ml/g of fresh tissue). Protein concentration was determined using the bicinchoninic acid assay (38). In-gel heme stain of holocytochome f was performed using TMBZ (39). Immunoblotting of proteins separated by SDS-PAGE was performed as in Ref. 40, except that samples were heated at 37 °C for 30 min. Antibody dilutions were 1:1,000 for anti-CF1 (spinach), anti-Cox1, anti-cyt f, anti-subunit IV/PetD (C. reinhardtii), and anti-HCF164 (A. thaliana); 1:2,000 for anti-Sec12 (A. thaliana) and anti-cyt b6 (C. reinhardtii); and 1:3,000 for anti-Rieske FeS (maize). A 1:3,000 dilution of alkaline phosphatase-conjugated goat anti-rabbit IgG was used as the secondary antibody. Sample loading was adjusted so that each sample contained the same amount of the subunit of CF1. Quantitation of the bands was performed using a MultiImage system running ChemImager 4400 software (Alpha Innotec).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Arabidopsis CCDA Is Targeted to the Plastid—The direct approach of immunolocalizing CCDA polypeptide after cell fractionation was not successful because the antisera we raised reacted very weakly if at all with native CCDA, perhaps because it is a low abundance hydrophobic protein (data not shown). Therefore, we chose to localize the in vitro synthesized radiolabeled protein after in vitro import into pea chloroplasts. We certainly expected CCDA to localize to chloroplasts based on the occurrence of the gene in the less derived red algal plastid genome of P. purpurea (6). Indeed, AtCCDA could be imported into isolated chloroplasts and was processed to the mature form, suggesting that the N-terminal extension functions as a chloroplast-targeting presequence in vitro (Fig. 1A). We further solidified this model by showing that the 141-amino acid N-terminal extension on CCDA could target a passenger protein like the widely used reporter, DHFR, to the chloroplast in vitro (Fig. 1B). As a positive control for import, we used the 77-amino acid targeting sequence of plastocyanin, a known thylakoid lumen resident protein. As expected, the targeting sequence of plastocyanin is able to drive import of DHFR into the chloroplast (Fig. 1C). We concluded that CCDA is targeted to the chloroplast via its N-terminal extension and that its plastid location is compatible with a function in chloroplast cytochrome c biogenesis.
|
and 
fragments of LacZ in the cytoplasm, and therefore an n-side location of the fusion site can be deduced. Location of the third extramembrane loop to the n-side or the p-side of the membrane could not be ascertained because both CCDA-PhoA and CCDA-LacZ fusions at position 236 of the polypeptide display high enzymatic activity (Table I, see fusion IV). We have ruled out that this is the result of a position effect at this particular site because fusions at positions 243, 249, and 254 in CCDA, all predicted to be located on the n-side, also led to high PhoA activity (data not shown). To overcome the ambiguity concerning the location of the third loop, we exploited the so-called sandwich approach, which offers greater reliability for examining membrane protein topology (31). In this method, PhoA is inserted in an otherwise intact CCDA protein (see arrows marked SW in Fig. 2). The measurement of PhoA activity of sandwich constructs SWA and SWB yielded a definitive placement of the third loop to the stromal n-side (Table I). Using an anti-PhoA antibody, we were able to detect the CCDA-PhoA sandwich protein in the membrane fraction of bacterial cells expressing the SWA construct (data not show). We could not detect the CCDA-PhoA sandwich protein expressed from the SWB construct, suggesting that the protein is not stable and undergoes degradation. This has been observed for fusion proteins where the cytoplasmic form of the alkaline phosphatase often fails to fold into a stable structure (26, 41). The experimentally deduced model is in agreement with the topological prediction and consists of a six-transmembrane protein with N and C termini on the lumen side.
|
|
|
|
The progeny arising from self-fertilization of the heterozygous ccda-1 plant exhibited a kanr:kans ratio of 322:17, or 
15:1, suggesting the presence of two segregating T-DNA insertions/genome. This was confirmed by Southern hybridization (Fig. 2A in the supplemental data), and heterozygous plants carrying only the ccda-inserted T-DNA were identified (Fig. 2A in the supplemental data). These were grown to maturity and allowed to self-pollinate to generate the ccda-1 homozygous line.
The progeny of self-fertilized CCDA/ccda-4 plants exhibited a kanr:kans ratio of 72:26, or 3:1, suggesting the presence of a single segregating T-DNA/genome. This was confirmed by Southern hybridization (Fig. 2B in the supplemental data). The progeny of self-fertilized CCDA/ccda-2 and CCDA/ccda-3 plants exhibited kanr:kans ratios of 3:1 and 15:1, indicative of the segregation of one single Ds element and two T-DNA elements, respectively. Preliminary characterization revealed that plants homozygous for ccda-2 and ccda-3 alleles could be grown to maturity on soil (see below), and further clean-up of the ccda-3 allele carrying line was omitted.
The molecular arrangement of each insertion in the CCDA gene was analyzed by PCR using T-DNA/Ds-specific and CCDA-specific primers (see supplemental material). This analysis indicated that the T-DNA insertions in ccda-1, ccda-3, and ccda-4 are all tandem or complex insertions carrying two left borders at their extremities (45). In the case of the ccda-2 allele, the simplest interpretation is the integration of a single Ds element (Fig. 3 and Fig. 3 in the supplemental data).
Growth Characteristics and Fluorescence Transients of the ccda Mutant Lines—Because loss of CCDA function is anticipated to block assembly of plastid c-type cytochrome, holocytochrome f, and consequently impair photosynthesis, homozygous ccda-interrupted plants were examined at the level of growth and photosynthetic electron transfer by recording chlorophyll fluorescence induction and decay kinetics. For fluorescence measurements, plants were propagated as segregating populations on MS medium containing 0.5% sucrose, and each plant was subsequently genotyped by PCR to identify ccda homozygotes. Plants homozygous for either the ccda-2 or the ccda-3 allele were phenotypically indistinguishable from wild type lines (data not shown). The plants grew on soil, flowered, set seeds normally, and behaved as their corresponding wild type with respect to photosynthetic electron transfer (data not shown). Homozygotes for the ccda-1 allele grew very slowly on soil and remained pale green (Fig. 4 and not shown), but after 60 days the plants, although quite small, produced a few flowers and set seeds. Consistent with the growth defect, mutant plants displayed a leaky hcf (high chlorophyll fluorescence) phenotype (Fig. 4) typical of photosynthetic mutants that are not completely blocked in the function of the b6f complex (46). Of 77 plants analyzed in the segregating population, 63 exhibited wild type fluorescence and 14 a hcf phenotype, making a 3:1 ratio (
2 = 1.06). That the hcf phenotype cosegregated with the ccda-1 genotype suggested that the T-DNA integration in CCDA was responsible for the lesion we observed. Mutants homozygous for ccda-4 were seedling-lethal (not shown) and could only grow on a minimal synthetic medium supplemented with 0.5% sucrose, as expected for photosynthesisminus Arabidopsis plants (46). The ccda-4 homozygous plantlets obtained on 0.5% sucrose-containing medium were about half the size of the wild type and heterozygous segregants grown in the same conditions (not shown). They displayed a tight hcf phenotype consistent with a complete loss of the cytochrome b6f complex and hence photosynthetic growth (Fig. 4). Of 167 plants examined among the segregants, 131 exhibited wild type fluorescence and growth and 36 a hcf and reduced growth phenotype. This 3:1 distribution (2 = 0.554) and the fact that the hcf phenotype cosegregated with the ccda-4 genotype suggested that the observed phenotype is attributable to the T-DNA integration. Interestingly, growth in the presence of 5% sucrose resulted in ccda-4 homozygous plants that were similar in size to the corresponding wild-type and heterozygous segregants but which produced only a few small flowers that did not fully open (data not shown).
|
Analysis of CCDA Transcripts in ccda Mutant Lines—We monitored the level of expression of the CCDA gene in the different mutant lines by RT-PCR analysis. As shown in Fig. 5A, both ccda-2 and ccda-3 homozygous lines that display no visible phenotype still express the CCDA mRNA, suggesting that the insertion of T-DNA or Ds elements in the 5'-untranslated region did not abolish the transcriptional activity of the CCDA gene. By contrast, plants carrying the ccda-1 and ccda-4 mutations that result in impaired photosynthesis do not accumulate the CCDA mRNA, and consequently, those mutations correspond to loss-of-function type of alleles. However, the difference in the severity of the photosynthetic phenotypes caused by the ccda-1 or ccda-4 alleles suggests that those mutations do not both cause a complete lack of function of the CCDA protein. By RT-PCR, we were able to show that both mutant lines expressed a truncated form of the CCDA mRNA (Fig. 5B). One possible explanation for the phenotypic differences is that in ccda-4, the mRNA does not encode a functional protein, whereas in ccda-1 the truncated protein still exhibits some activity.
|
Mutations in CCDA Result in a Defect in Cytochrome b6f Complex Accumulation—We assessed the abundance of holocytochrome f in the mutant lines by heme staining and immunoblotting and also that of Rieske FeS, cytochrome b6, and subunit IV, which are subunits of the cytochrome b6f complex. As expected from the lack of phenotype in the ccda-2 and ccda-3 lines, holocytochrome f accumulation is not affected by these mutations, and the molecule assembles to essentially wild type level (Fig. 6). By contrast, the ccda-1 mutation causes a severe reduction in the level of holocytochrome f. Only 20–25% of wild type levels of holocytochrome f is detected in the ccda-1 line. Concomitant with the diminution in holocytochrome f abundance, we observed a decrease in the accumulation of the subunits of the cytochrome b6f complex (Fig. 6). The ccda-4 mutation that leads to a more drastic phenotype than the ccda-1 mutation has a more pronounced effect on holocytochrome f accumulation. Only 10% of wild type level of cytochrome f is assembled in this mutant line, and the abundance of cytochrome b6f subunits is also likewise severely reduced (Fig. 6). We concluded that loss-of-function ccda alleles lead to a defect in plastid biogenesis that is the result of impaired accumulation of the cytochrome b6f complex in the thylakoid membranes.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Arabidopsis CCDA Is a Chloroplast Polytopic Membrane Protein—All vascular plant CCDA homologs that we identified exhibit an additional stretch of 100–115 amino acids in the N-terminal part of the polypeptide that is missing in the rhodophyte homolog encoded in the plastid genome (see Fig. 1 in the supplemental data). Chlamydomonas CCDA displays the shortest N-terminal extension (61 amino acids), a feature that is common to algal transit peptides that are 32 amino acids shorter, on average, than those from vascular plants (47). The N-terminal extensions present typical features of bipartite transit sequences that direct proteins to the thylakoid membrane via the so-called "spontaneous import pathway" (48, 49). The first domain directs the protein across the envelop into the stroma, whereas the second domain carries targeting information to the thylakoid. Both domains in the N-terminal extensions exhibit the expected characteristics: the former is hydrophilic and rich in hydroxylated residues, and the latter contains a positively charged N terminus followed by a hydrophobic segment, which is a crucial determinant for thylakoid translocation and ends with a processing site (AXA
) for the thylakoid processing peptidase (see Fig. 1 in the supplemental data). Our in vitro experiments have established that this N-terminal extension indeed serves as a chloroplast-targeting sequence and is cleaved off upon import as expected for nuclearly encoded proteins that function in the plastid organelle. The sublocalization of CCDA could not be examined by conventional cell fractionation techniques because we were unable to generate an immunoreactive probe for the polypeptide. However, based on the presence of the bipartite chloroplast-targeting sequence and the polytopic model of CCDA that we have generated from PhoA-LacZ fusion analysis, we can deduce indirectly that CCDA is localized to the thylakoid membrane. Our topogical model is compatible with the experimentally deduced topology of bacterial Rhodobacter capsulatus CcdA (26), and it is noteworthy that in both plastid and bacterial CcdAs, the critical cysteine residues are assigned a within-membrane location in the first and the fourth transmembrane segments. The exact position of these cysteines, whether embedded or in proximity to the stromal or luminal sides, cannot be resolved by our topological experiments. The function of the within-membrane cysteines in CCDA is not really known, but functional dissection of the related bacterial DsbD/DipZ transporter has demonstrated that the redox active cysteines face the cytoplasmic side of the membrane (50). These cysteines have the potential to form a disulfide bond, an indication of their physical vicinity, and are key residues in the relay of reducing equivalents originating from the cytoplasm (50). It is likely that conserved cysteines in CCDA are also facing the stromal side of the thylakoid membrane and operate in a similar manner.
Mutations in CCDA Indicate a Function in the Biogenesis of the Cytochrome b6f Complex—The severity of the photosynthetic phenotype in the ccda lines correlates nicely with the impact of the interrupting DNA elements on the expression of the CCDA gene. The two insertions in ccda-2 and ccda-3 lines which lie in the 5'-untranslated region of the CCDA gene did not result in any observable phenotype under our conditions, either at the level of growth on soil or when examined for the accumulation of the cytochrome b6f complex. Neither insertion prevented the transcription of the CCDA gene, a rather surprising finding considering that the interrupting elements are only 15 and 23 bp away from the initiation codon of the CCDA gene. One likely explanation is that both the T-DNA and the Ds element harbor some promoter-like activity that enables transcription of CCDA. The Ds element is known to carry a promoter-like sequence at its 3'-end that can be activated upon integration (51). Indeed, we could detect a hybrid RNA in the ccda-2 line using a Ds element-specific primer and a CCDA-specific primer (Fig. 5B), confirming that the element does not block expression of CCDA. Despite several attempts, a transcript originating from the T-DNA in the ccda-3 line could not be detected by RT-PCR (Fig. 5B). However, based on the observation that the T-DNA in the ccda-4 line allows transcription of a truncated CCDA mRNA (Fig. 5B), we assume that the CCDA mRNA produced in the ccda-3 line also originates from the T-DNA. Lack of phenotype because of insertion of interrupting element in proximity to the initiation codon has also been reported previously in Arabidopsis (52).
The ccda-1 and ccda-4 lines do not express the full-length CCDA mRNA and result in photosynthetic incompetence that is correlated to decreased level of assembled cytochrome b6f complex. The ccda-1 and ccda-4 alleles cause distinct phenotypes: the former leads to a partial loss of function with 25% of assembled holocytochrome f that enables photosynthetic growth, and the latter results in complete loss of function with only 10% of accumulated holocytochrome f, which prevents photosynthetic growth. It is interesting to note that 25% of the level of holocytochrome f is sufficient to sustain germination and photosynthetic growth, whereas only 10% of the level of cytochrome f is insufficient for photosynthetic growth.
Unexpectedly, truncation of the first three exons in the ccda-1 allele does not produce a complete knock-out of the CCDA gene. Somehow, despite the lack of the first 90 amino acids that carry the targeting information, the truncated protein is still able to function in the assembly of the cytochrome b6f complex. It is possible that targeting information lies also in the hydrophobic core of the polypeptide or that the N-terminal sequence created by the fusion between the T-DNA and the CCDA gene is compatible with import into the thylakoid membrane.
CCDA May Correspond to an Eighth Component in the System II Pathway in Plastid—Our data support the implication of CCDA in a process required for the biogenesis of cytochrome b6f complex but do not elucidate whether heme attachment to apocytochrome f is compromised by ccda mutations. The defect in cytochrome b6f biogenesis in the absence of CCDA is most likely the result of compromised assembly of the complex, consistent with the observation that holocytochrome f is required for the assembly of the b6f complex in Chlamydomonas (53). A definite placement of CCDA in the plastid c-type cytochrome assembly pathway also relies on the examination of the assembly of cytochrome cX, the other c-type cytochrome occurring in the thylakoid lumen in Arabidopsis (54). This was not feasible because of the extraordinarily low abundance of this molecule in the plastid, whose holoform can only be detected in plants engineered to overexpress the protein (55). The green alga C. reinhardtii remains the ideal system to study plastid c-type cytochrome assembly because of the availability of ccs mutants (cytochrome c synthesis) that are blocked at the level of heme attachment to the apoforms of cytochrome f and cytochrome c6, the two c-type cytochromes resident in the lumen (56–59). Seven loci, plastid ccsA and nuclear CCS1 to CCS6, are required for the accumulation of both plastid holocytochrome f and c6 (17, 20).6 The presence of a CCDA homolog in Chlamydomonas prompted us to examine whether unidentified CCS2 to CCS6 loci could correspond to the CCDA gene. Two cosmids (pARG7-CCDA) carrying the ARG7 selection marker and containing genomic DNA7 corresponding to Chlamydomonas CCDA plus 5'- and 3'-flanking DNA were identified. The pARG7-CCDA cosmids were tested directly for their ability to rescue representative mutants corresponding to each of the CCS2 through CCS6 loci or indirectly for restoration of photosynthetic growth in arginine prototrophs resulting from transformation of ccs arg7 mutants. For most strains, we also analyzed the CCDA locus physically by blot hybridization and sequence analysis of the entire ORF. Based on the lack of complementation and absence of physical change in the CCDA gene in the ccs mutants, we concluded that if CcdA is involved in cytochrome c biogenesis, then it must correspond to a component that is not yet genetically defined.
There is the formal possibility that plastid CCDA is necessary for other processes controlling the biogenesis of cytochrome b6f complex, such as, for instance, maturation of the recently discovered covalent heme on holocytochrome b6 (60, 61). However, a function of CCDA in holocytochrome b6 assembly seems unlikely based on the fact that the covalent heme in holocytochrome b6 is on the stromal side (60, 61) and that covalent binding of heme to bacterial holocytochrome b is not dependent upon CcdA (25). That CCDA is recruited to complete plastid cytochrome c maturation seems in our sense the most compelling hypothesis, and we propose that it corresponds to an eighth assembly factor in the system II plastid pathway.
Emergence of a Thioreduction Pathway in Plastid c-Type Cytochrome Biogenesis—The existence of a thio-reduction pathway in the chloroplast was also suggested recently with the discovery of HCF164, a thylakoid membrane-anchored thioredoxin-like protein with disulfide reductase activity which is required for cytochrome b6f biogenesis in Arabidopsis (62). HCF164 displays similarity to CcsX/ResA, a thioredoxin-like protein involved in system II bacterial cytochrome c assembly pathway (63, 64). Remarkably, CcsX, ResA, and HCF164 proteins share the same membrane topology with the thioredoxin-like domain facing the p-side of the membrane, suggesting that their relevant targets of action are also located in this compartment (62–64). The definitive placement of HCF164 in plastid cytochrome c maturation is still uncertain because the defect in cytochrome b6f assembly could not be localized specifically to a block in the maturation of a particular subunit (i.e. conversion of apocytochrome f to its holoform) (62). A nucleus-encoded HCF164-like protein is also present in Chlamydomonas, but functional assignment in chloroplast cytochrome c biogenesis could not be provided in the algal system because none of the ccs2 to ccs6 mutations appears to lie in the HCF164-encoding gene.8
Erldensson and co-workers (28, 64) have proposed that ResA and CcdA are part of a redox relay required for the reduction of the cysteinyls in the heme binding site of apocytochrome c. Interestingly, we have noted that in cyanobacteria Gloeobacter violaceus and Geobacter sulfurreducens, CCDA and HCF164 homologs are encoded by adjacent genes (65, 66), giving support to the hypothesis that they might act together in a same redox pathway in the plastid. Immunoblotting experiments showed that HCF164 accumulation is not affected by loss-of-function alleles in CCDA (Fig. 7). This result, however, does not necessarily rule out a model in which CCDA and HCF164 are interacting partners involved in the same pathway, and we hypothesize that HCF164 is the ninth component involved in plastid cytochrome c formation. We propose that the transfer of electrons from the stroma to the lumen for maintenance of reduced CXXCH motif in apocytochrome c proceeds via sulfide/disulfide transporter CCDA and HCF164 thioredoxin (Fig. 8). We postulate that the provision of reducing equivalents in the plastid system is dependent upon other redox components, like for instance, a membrane-anchored thioredoxin-like protein on the n-side (Fig. 8). These redox components might correspond to the gene products of CCS4 and CCS5 loci in Chlamydomonas (17, 20). Indeed, a preliminary functional assessment of these genes in redox chemistry is based on the observation that ccs4 and ccs5 mutants, similar to ccda and resA/ccsx mutants in bacteria (26, 28, 63, 64), can be rescued for cytochrome c deficiency by exogenous reduced thiol compounds.9 It is conceivable that, because of the compartmentalization of the thylakoid lumen, the redox pathway evolved in plastids relies on several players (at least CCDA, HCF164, Ccs4, and Ccs5) and is therefore more complex than its bacterial counterpart.
|
|
|
FOOTNOTES |
|---|
The on-line version of this article (available at http://www.jbc.org) contains an additional Table I and Figs. 1–3 and supplemental experimental procedures.
These authors contributed equally to this work.
¶ Supported in part by American Heart Association Fellowship 0120100Y and the Muscular Dystrophy Association Fellowship 3618.
To whom correspondence should be addressed: Dept. of Chemistry and Biochemistry, UCLA, Box 951569, Los Angeles, CA 90095-1569. Tel.: 310-825-8300; Fax: 310-206-1035; E-mail: merchant{at}chem.ucla.edu.
1 The p-side (for positive) corresponds to the plastid lumen, mitochondrial intermembrane space, or bacterial periplasm. The n-side (for negative) is the plastid stroma, mitochondrial matrix, or the bacterial cytoplasm.
2 The single letter amino acid code is used where X is any amino acid except cysteine.
3 The abbreviations used are: DHFR, dihydrofolate reductase; kanr, kanamycin-resistant; kans, kanamycin-sensitive; MS medium, Murashige & Skoog medium; ORF, open reading frame; RT, reverse transcription.
4 Descurainia sophia, Glycine max, Hordeum vulgare, Lotus japonicus, and Medicago truncatula (TBLASTN scores ranged from 10–17 to 10–55).
5 GenBankTM accession numbers: C. reinhardtii, AAL84598
[GenBank]
A. thaliana, AAF35369
[GenBank]
D. sophia, AAQ95739
[GenBank]
G. max, AAP81162
[GenBank]
H. vulgare, AAO16019
[GenBank]
L. japonicus, AAO16018
[GenBank]
and M. truncatula, AAO16020
[GenBank]
6 P. Hamel and S. Merchant, unpublished data.
7 We determined the genomic organization of the CCDA locus by sequencing.
8 S. T. Gabilly, P. P. Hamel, and S. Merchant, unpublished data.
9 P. P. Hamel and S. Merchant, unpublished data.
|
ACKNOWLEDGMENTS |
|---|
-Rieske FeS), D. Krogmann (spinach -cyt f), K. Meierhoff (Arabidopsis -HCF164), J. L. Popot (Chlamydomonas -Cox1), N. Raikhel (Arabidopsis -Sec12), and F.-A. Wollman (Chlamydomonas -cyt b6 and -subunit IV/PetD) for the kind donation of antibodies. We also thank Dr. J. Gober for the gift of plasmid placZ290, Drs. J. Beckwith for the gift of plasmid pBAD22 and M. Ehrmann for the gift of strain LMG194. |
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
indicates the position of the integrated DNA, and 
(bp) shows the size of the deletion of A. thaliana genomic DNA upon integration of the interrupting element.
right border for the T-DNA and 5' 
