- © 1999 American Society of Plant Physiologists
Abstract
A recessive mutation in Arabidopsis, named chaos (for chlorophyll a/b binding protein harvesting–organelle specific; designated gene symbol CAO), was isolated by using transposon tagging. Characterization of the phenotype of the chaos mutant revealed a specific reduction of pigment binding antenna proteins in the thylakoid membrane. These nuclear-encoded proteins utilize a chloroplast signal recognition particle (cpSRP) system to reach the thylakoid membrane. Both prokaryotes and eukaryotes possess a cytoplasmic SRP containing a 54-kD protein (SRP54) and an RNA. In chloroplasts, the homolog of SRP54 was found to bind a 43-kD protein (cpSRP43) rather than to an RNA. We cloned the CAO gene, which encodes a protein identified as Arabidopsis cpSRP43. The product of the CAO gene does not resemble any protein in the databases, although it contains motifs that are known to mediate protein–protein interactions. These motifs include ankyrin repeats and chromodomains. Therefore, CAO encodes an SRP component that is unique to plants. Surprisingly, the phenotype of the cpSRP43 mutant (i.e., chaos) differs from that of the Arabidopsis cpSRP54 mutant, suggesting that the functions of the two proteins do not strictly overlap. This difference also suggests that the function of cpSRP43 is most likely restricted to protein targeting into the thylakoid membrane, whereas cpSRP54 may be involved in an additional process(es), such as chloroplast biogenesis, perhaps through chloroplast–ribosomal association with chloroplast ribosomes.
INTRODUCTION
Higher plants require chloroplasts for essential functions ranging from photosynthesis to sugar, lipid, and amino acid biosyntheses. Many of the proteins required for the light reactions of photosynthesis, including the chlorophyll-containing antenna proteins, are localized in the thylakoid membrane. Thylakoid proteins are encoded by both the nuclear and chloroplast genomes. Nuclear-encoded proteins destined for the thylakoid membrane are synthesized with a cleavable N-terminal extension (transit peptide) that targets the protein across the envelope membranes into the stroma (reviewed in Cline and Henry, 1996; Kermode, 1996; Lübeck et al., 1997). Multiple mechanisms exist for targeting proteins from the stroma to the thylakoid membrane (reviewed in Cline and Henry, 1996; Klösgen, 1997). These include spontaneous insertion, a chloroplast Sec r–dependent pathway, a chloroplast signal recognition particle (cpSRP) pathway, and a pathway requiring only an electrochemical potential of ΔpH across the thylakoid membrane. These pathways can be distinguished by specific requirements for stromal proteins and/or by energetic requirements when in vitro reconstitution assays are performed. In vitro, precursor proteins are targeted exclusively via one of the aforementioned pathways.
LHCIIb proteins are the most abundant of the thylakoid membrane proteins. They are the major polypeptides of the light-harvesting chlorophyll a/b binding protein complex of photosystem II (PSII). They constitute approximately one-third of the total thylakoid protein and bind half of the chlorophylls (Yamamoto and Bassi, 1996). They belong to a large family of related polypeptides, which include the minor chlorophyll protein complexes of PSII and the chlorophyll antenna of photosystem I (PSI). The members of this family are collectively called LHCPs in this study. Many of the proteins targeted to the thylakoid require a bipartite transit peptide consisting of an N-terminal envelope transit sequence followed by a thylakoid transfer domain that directs the protein across the thylakoid membrane. The transit peptides of LHCPs, however, comprise only the envelope transit sequences, which target the hydrophobic proteins to the stroma, where they are able to insert into the thylakoid membrane in the absence of their transit peptides (Lamppa, 1988). Thus, the information for LHCP integration into the thylakoid resides in the mature protein (Hand et al., 1989). Attempts at localizing the targeting information have not been successful (Auchincloss et al., 1992; Huang et al., 1992).
Mutant Phenotype.
(A) Seedling stage.
(B) Rosette stage. WT, wild type.
(C) Electron microscopic analysis of chloroplasts. g, grana thylakoids; s, starch grains; v, vesicles present in the mutant only.
Carotenoid and Chlorophyll Content of Leaves
LHCP trafficking requires the intervention of protein factors in the stroma (Cline, 1986; Chitnis et al., 1987; Yuan et al., 1993) and GTP (Hoffman and Franklin, 1994). One of these protein factors is a chaperone that maintains the solubility of LHCPs as they are transported through the stroma (Payan and Cline, 1991). In vitro studies have shown that the subunit of the chaperone that interacts with the LHCPs is cpSRP54 (Franklin and Hoffman, 1993; Li et al., 1995). CpSRP54 is a chloroplast homolog of the 54-kD protein of the mammalian SRP, which is required for the transport of proteins across the endoplasmic reticulum membrane (reviewed in Rapoport, 1992). SRPs are not restricted to eukaryotes: they have been found in a wide range of organisms and may function in protein translocation in every living organism.
When cpSRP54 was removed from stromal extract by immunodepletion, activities required for in vitro trafficking of LHCPs were inhibited. However, the addition of cpSRP54 did not restore activity (Li et al., 1995). This observation suggests the possibility that an additional factor was coimmunoprecipitated along with cpSRP54. Most organisms have a cytoplasmic SRP that contains SRP54 bound to an SRP RNA (reviewed in Lütcke, 1995). However, in plants, no such RNA that binds to cpSRP54 has been detected. Rather, cpSRP54 was found attached to a 43-kD protein named cpSRP43 required for LHCP targeting to the thylakoid membrane (Schuenemann et al., 1998). However, the exact nature and functions of the 43-kD protein have yet to be determined.
In this study, we have identified chaos (for chlorophyll a/b binding protein harvesting–organelle specific), a tagged mutant of Arabidopsis that is deficient in the production of cpSRP43. By using transposon tagging and by cloning, we were able to identify and characterize CAO, the gene encoding the cpSRP43 polypeptide, and to determine a role for cpSRP in the biogenesis of PSII light-harvesting proteins. The three-letter gene symbol CAO is used for the chaos mutation, which is in keeping with Arabidopsis nomenclature guidelines.
RESULTS
The chaos Mutant Is Depleted of Chlorophyll and Carotenoids
The chaos mutant was identified in a population of 217 independent lines of Arabidopsis ecotype Landsberg erecta (Ler-0) containing a transposed enhancer trap derivative of the maize Dissociation (Ds) transposable element. Self-pollination of a plant from line 348/74/A resulted in progeny exhibiting a recessive chlorotic mutant phenotype (Klimyuk et al., 1995). The mutant was named chaos (Figures 1A and 1B). The pale coloration was observed throughout the vegetative cycle of the plant and uniformly affected all of the aerial tissues.
The chlorotic phenotype suggested that major pigment levels in the mutant had been reduced. Therefore, we measured the concentration of chlorophylls and carotenoids in the leaves of the mutant and the wild type. Total chlorophyll declined by 43%, and the chlorophyll a/b ratio increased from 2.6 in the wild type to 3.2 in the chaos mutant. Members of the LHCP family are the only proteins that bind chlorophyll b. In addition, LHCPs bind carotenoids, such as lutein, neoxanthin, and violaxanthin (Bassi et al., 1993), which were significantly reduced in the mutant, whereas carotenoids that are not present in the LHCPs of PSII, such as β-carotene, exhibited little if any decrease (Table 1). Thus, the mutant phenotype appeared to affect primarily LHCP-associated pigments.
Electron microscopic analysis of chloroplasts revealed no differences between the ultrastructure of the mutant and wild-type organelles (Figure 1C). In both cases, we found starch grains, plastoglobuli (lipids), and typical thylakoid grana. The size, shape, and number of plastids in the mutant chlorotic cells did not vary when compared with those of the wild type (data not shown). The only apparent difference was the presence of numerous vesicles in the chaos stroma (Figure 1C).
Photosynthetic Oxygen Production Measured with a Clark Electrode
The Light-Harvesting Step of Photosynthesis Is Reduced in the Mutant
To investigate the effects of the chaos mutation on photosynthesis, we measured oxygen production by leaves with a Clark-type oxygen electrode under different light conditions (Table 2). The level of oxygen production was always lower in the mutant when compared with the wild type on the basis of leaf area. There was a 16% decrease in oxygen production under high light conditions, whereas the difference increased to 43% under low light. This suggests that the mutation selectively affects the light-harvesting process.
The light distribution between PSI and PSII was evaluated by using the Emerson effect (Canaani and Malkin, 1984). Under state 2 light conditions, a quinone of the cytochrome b6/f complex becomes reduced and activates the kinase system, which in turn phosphorylates the LHCPs (Williams and Allen, 1987). Phosphorylated LHCP complexes dissociate from PSII and associate with PSI. Under state 1 light conditions, the quinone is oxidized, thus inducing the reverse reaction and leading to the return of LHCP to PSII.
For both the wild type and chaos, we observed, as expected, a very low Emerson effect (E) value in state 2, indicating that light energy absorption was well balanced between the two photosystems (Table 3). In state 1, we saw an increase in the E value effect, indicating that LHCP migration from PSII to PSI occurred in both genotypes. However, the E value was significantly higher in the wild type when compared with chaos, and these differences are more significant for state 1 than for state 2. Although both chaos and wild-type chloroplasts are able to regulate light distribution between PSII and PSI, light absorption in PSII is selectively reduced when compared with PSI in the chaos mutant.
Photosynthetic Oxygen Production Measured by Photoacoustic Spectroscopy
Electron Transfer in Both Photosystems Is Not Inhibited
The maximal quantum yield of PSII photochemistry (Φ) was quantified by in vivo measurements of chlorophyll fluorescence. The Φ value obtained for the wild type was 0.80 ± 0.01(mean value of 17 independent experiments ±SE). This is a typical value for healthy plants (Björkman and Demmig, 1987). The chaos mutant had a similar value (0.82 ± 0.01). This clearly demonstrates that the maximal photochemical efficiency of PSII was not affected by the chaos mutation. We also measured the actual quantum yield of PSII photochemistry in white light (100 μmol of photons m−2 sec−1), and no difference was observed between chaos and the wild type (0.74 ± 0.01). These findings indicate that PSII remained photochemically competent in chaos and that the efficiency of photosynthetic electron transfer (PSI plus PSII) was not affected by the mutation.
The Mutation Affects the Level of Nuclear-Encoded Antenna Proteins
To identify the possible changes in the composition of photosystems in the chaos mutant, we solubilized the thylakoid pigment protein complexes and analyzed them by partially denaturing gel electrophoresis (Figure 2A). A clear difference between the wild type and the mutant could be observed due to a strong reduction of the trimeric and monomeric forms of the LHCPs. Figure 2B shows that the chaos mutant lost a significant number of major chloroplast proteins between 24 and 29 kD. To identify the proteins affected, we performed immunoblot analyses with representative proteins of PSI (PSA D, E, F, and L) and PSII (D1, D2, CP47, CP43, CP24, and LHCIIb). The results given in Figure 2C can be considered as quantitative (the amounts of protein required to avoid saturation problems were investigated in preliminary experiments). As already shown (Figure 2A), the levels of the major LHCP (i.e., LHCIIb) were lower in the mutant. Densitometric analysis revealed a decrease of ~50% for the LHCIIb proteins (the result was similar when the thylakoid-pigmented complexes of Figure 2A were analyzed). This result was also observed with CP24, another nuclear-encoded antenna protein of PSII, whereas all other components of PSI and PSII remained unchanged (Figure 2C). Immunoblotting was also conducted with components of the Cyt b6/f complex and ATP synthase. These did not reveal any differences between the wild type and the mutant (data not shown).
The CAO Locus Is Tagged by Using the Ds Transposon
Genetic analysis performed with selfed progeny of a plant from line 348/74/A demonstrated linkage of the chaos allele to the Ds insertion (Klimyuk et al., 1995). To confirm that the chaos mutation was caused by the Ds insertion, we selected revertants from among the progeny of a plant homozygous for Ds insertion and carrying the Activator (Ac) transposase. Three revertants were isolated from the ~15,000 plantlets that were screened. Reversion of the mutation to the wild-type phenotype was correlated with the excision of the Ds element (data not shown). Analysis of the excision footprints in the three revertants revealed complete restoration of the open reading frame and a substitution of the first two bases (G and A) flanking the 3′ end of the Ds element (G and A being converted to C and A or G and C, depending on the revertant). The probability of such events is extremely low, which explains the low frequency of reversions that were observed. Nevertheless, this analysis demonstrated that the chaos mutation was due to the Ds insertion.
Gel Electrophoretic and Immunoblot Analyses of the Photosynthetic Apparatus.
(A) Separation of pigment–protein complexes solubilized from thylakoids of the wild type (WT) and the chaos mutant. The pigment–protein complexes were solubilized with a mixture of SDS and octylglucoside and separated by gel electrophoresis. The green bands were visible without staining. CCI, core complex I; CCII, core complex II; FP, free pigment; LHCII, light-harvesting complex II; (LHCII)3, trimeric form of LHCII.
(B) Total proteins and thylakoid membrane polypeptides of the chaos mutant and the wild type (WT) after electrophoresis in a 15% polyacrylamide–SDS gel and silver staining. Each lane contains 8 μg of protein. Molecular masses (given in kilodaltons at the left) were estimated by co-electrophoresis with commercially available size standards.
(C) Immunoblot analysis of chaos mutant and wild-type (WT) proteins. Wild-type and mutant proteins of thylakoids were electrophoresed in a 12% polyacrylamide–SDS gel and immunodetected with antisera, as described by Meurer et al. (1996). Thylakoid proteins used in the immunoblot analysis are as follows: PSI; PSII; PSA D, E, F, and L (subunits of PSI); CP47 and CP43 (chlorophyll a apoproteins of the inner antennae of PSII); D1 and D2 (PSII reaction center proteins); CP24 (one of the minor light-harvesting complexes of PSII); and LHCIIb (major light-harvesting complex of PSII).
Cloning of the CAO Gene
A 300-bp DNA sequence flanking the right border of the Ds element was amplified by inverse polymerase chain reaction, cloned, and sequenced. This probe was used to isolate four cDNA clones of 1.4 kb (pLMC C6, pLMC C8, pLMC F1, and pLMC F2) from an Arabidopsis seedling library. A genomic cosmid DNA library was screened, and a 6-kb BamHI fragment hybridizing with the probe was subcloned into the pBluescript KS+ vector. The sequence of the entire genomic region of the CAO gene (GenBank accession number AF013115) was determined (Figure 3). No introns were detected.
The four cDNA clones differed only at their 3′ ends. Two alternative polyadenylation sites were found at positions 1149 and 1328 of the genomic sequence (Figure 3). A near upstream element (a sequence that can control the utilization of a polyadenylation site), AAUGAA (Wu et al., 1994), was found at positions 1119 and 1123. These positions, 16 and 20 bp before the first polyadenylation site, are characteristic of such signals (Li and Hunt, 1995). A putative TATA box, TATAAA, was located 158 bp upstream of the ATG codon.
Genomic Sequence of the CAO Gene and the Deduced Amino Acid Sequence.
The amino acids comprising the potential chloroplast transit peptide are underlined with dots, and an arrowhead marks the most likely position for the cleavage. The ankyrin repeats are boxed, and the chromodomain is underlined. The two positions observed for the poly(A) tail on CAO transcripts are indicated by dots. Insertion of the transposable element is indicated with an asterisk in the coding sequence.
The Ds insertion was found just after position 825 of the genomic sequence. The typical 8-bp target site duplication usually flanking the insertion of Ds elements (Fedoroff, 1989) was absent. In its place was one extra base (a cytidine) preceeding the 5′ end of the Ds element.
The CAO Gene Exists as a Single Copy on Chromosome 2 in the Arabidopsis Genome
Gel blot analyses of genomic DNA from three Arabidopsis ecotypes (Columbia, Ler, and Wassilewskija) digested with different restriction enzymes were performed using the cloned cDNA as a probe. The gene was found to exist as a single copy in the Arabidopsis genome (data not shown). The CAO gene was mapped to the bottom of chromosome 2. It shares residues identical to those of the open reading frame encoding an unknown protein and is positioned on bacterial artificial chromosome T30B22, which is part of yeast artificial chromosome CIC06C03 and most likely located upstream of the recombinant inbred marker mi79a.
Sequence Analysis of the Protein Encoded by CAO
The nucleotide sequence of the CAO cDNA contained one open reading frame with the potential to encode a polypeptide of 376 amino acid residues, with a predicted molecular mass of 41.5 kD and an isoelectric point of 4.3 (Figure 3). The deduced amino acid sequence of the protein showed several interesting features. It contains a domain similar to those of typical transit peptides of chloroplast proteins (Kermode, 1996). The first 60 amino acids are rich in the hydroxylated amino acid serine (25%) and in hydrophobic amino acids (valine [11%] and alanine [6%]) but contain very few acidic amino acids (5%). Based on the cleavage site consensus sequence (V/I)X(A/C)/A defined by Gavel and von Heijne (1990), the cleavage site is predicted to occur after A-60. The processing site of the pea protein has been determined by N-terminal sequencing and suggests that processing occurs at either A-60 or C-59, depending on how the peptide is aligned with the Arabidopsis sequence.
Database searches using the Blast program (Altschul et al., 1990) identified two segments of the protein encoded by CAO with similarity to motifs mediating protein–protein interactions. The first one corresponds to four tandem ankyrin repeats located between residues 130 and 254. These four ankyrin repeats (Figure 4A) exhibit a conservation of the consensus motif (defined by Zhang et al., 1992), ranging from 62 to 85%. This is in agreement with the 54 to 100% identity observed among hundreds of ankyrin repeats (Bork, 1993).
Domains of Significant Similarity to Known Proteins.
(A) Similarity to ankyrin consensus motifs (defined by Zhang et al., 1992).
(B) Similarity to chromodomains (Aasland and Stewart, 1995).
CHD1, helicase-domain; Dm, Drosophila melanogaster; HET, heterochromatin protein; HP1, heterochromatin protein 1; Mo, mouse; Pc, Planococcus citri; PC, polycomb; Sp, Schizosaccharomyces pombe; SWI6, repressor of mating-type loci. A secondary structure prediction generated with the PREDICT (Rost and Sander, 1994) and HCA (Lemesle-Varloot et al., 1990) programs is shown. H indicates a prediction for an α helix and E for a β sheet. The numbering of the sequences refers to their position in the protein. Amino acids strictly conserved or closely related in many of the sequences are underlined in the CHAOS sequences. Dashes were introduced to optimize alignment.
The second domain has similarity to the chromatin binding domain (chromodomain) originally identified in Polycomb and heterochromatin protein 1, which are two Drosophila proteins (Paro and Hogness, 1991). In these proteins, chromodomains have been implicated in the regulation of chromatin structure through protein–protein interactions. Chromoboxes are conserved motifs in both plants and animals, as demonstrated by gel blot analyses (Singh et al., 1991). They vary in length from 30 to 70 amino acids (Aasland and Stewart, 1995; Ball et al., 1997) and typically have three blocks of conserved sequences. The protein encoded by CAO has two chromodomains in tandem spanning residues 271 to 368 (see alignment in Figure 4B).
The CAO Transcript Is Expressed in Aerial Tissues
The CAO gene was tagged by using an enhancer trap Ds element (Klimyuk et al., 1995). The β-glucuronidase (GUS) gene, which was used as an enhancer trap, was activated in the chaos mutant, with GUS being very weakly expressed in the aerial tissues of the plant (Figure 5A). To confirm these observations, we performed RNA gel blot analyses with the cloned CAO cDNA as a probe. Poly(A) mRNA purified from 100 μg of total RNA for each sample was used. This step was necessary due to the very low abundance of the CAO mRNA (no signal was obtained with 10 μg of total RNA). The transcript was detected in leaves and not in roots (Figure 5B). Thus, consistent with the chlorotic phenotype, expression of CAO appears to be confined to chlorophyll-containing tissues. Circadian fluctuations of the CAO mRNA levels were investigated. The amplitude of the oscillation was not very marked and did not exceed 55 to 65% (data not shown). The level of mRNA accumulation was maximal during the later part of the night period and decreased throughout the day.
Identification of chaos as a Mutant in CpSRP43
The presence of a transit peptide suggested that the protein encoded by CAO was localized in the chloroplast. To confirm this, we probed blots of proteins prepared from wild-type and chaos leaves with an antiserum raised against the recombinant CAO-encoded protein. Immunoblot analysis revealed that the antiserum detected a protein in wild-type leaves that was absent in the chaos mutant (Figure 5C). Furthermore, a unique polypeptide in the stromal fraction of purified Arabidopsis (Figure 5D) and pea (Figure 6B) chloroplasts, respectively, cross-reacted with the antiserum, confirming the chloroplast localization of the CAO gene product. The use of silver-stained gel and protein mass markers indicates that the sizes of the detected polypeptides were, respectively, 42 kD for Arabidopsis and 43 kD for pea (data not shown). We established that the stromal and thylakoid membrane fractions were basically free of contamination by using antisera raised against stromal heat shock protein Hsp70 and the LHCIIb. More than 90% of cpSRP43 is estimated to be in the stroma, because the protein was not detected in 10 μg of thylakoid proteins but was clearly detected in 2 μg of stromal proteins.
The phenotype of the chaos mutant suggested that the protein encoded by CAO might play a role in LHCP biogenesis. LHCP trafficking requires a complex of the proteins cpSRP54 and cpSRP43 (Schuenemann et al., 1998). CpSRP54 has been characterized (Franklin and Hoffman, 1993; Li et al., 1995). However, the nature of cpSRP43 has not yet been elucidated. Three lines of evidence demonstrate that the protein encoded by CAO corresponds to cpSRP43. First, peptide sequence data from pea cpSRP43 are similar to the predicted sequence from the Arabidopsis cDNA (Figure 6A). Second, antibodies raised against the protein encoded by CAO recognize, respectively, the pea 43-kD (Figure 6B) and the Arabidopsis 42-kD (data not shown) bands coimmunoprecipitated by antibodies raised against cpSRP54. As shown in Figure 6B, the 43-kD polypeptide in the pea stroma recognized by the CAO antiserum was efficiently removed from the stroma by the anti-cpSRP54 antibody and was recovered in the pellet fraction, whereas none of the 43-kD polypeptide was precipitated by an unrelated antiserum. Thus, the anti-cpSRP54 antibody immunodepletes the stroma of both cpSRP54 and the CAO-encoded polypeptide. Third, we have shown that in pea stroma immunodepleted of the SRP transit complex (composed of cpSRP54 and cpSRP43), LHCP integration in thylakoids is reduced by >90%. This integration could only be resaturated when bacterially expressed cpSRP54 and CAO-encoded proteins were added to the immunodepleted pea stroma (Schuenemann et al., 1998). However, the same experiments performed with Arabidopsis chloroplasts failed, presumably because it is difficult to isolate chloroplasts competent for import in this species. These data unambiguously demonstrate that the CAO-encoded protein is the pea cpSRP43 homolog.
Analysis of CAO Transcript and Products.
(A) Pattern of GUS enzyme activity in the chaos mutant.
(B) RNA gel blot analysis of the CAO transcript. Poly(A)+ RNAs corresponding to 100 μg of total RNA for each sample were analyzed using the CAO cDNA as probe. The blot was reprobed with a cDNA of a nuclear gene encoding the β subunit of mitochondrial ATPase to control the loading. WT, wild type.
(C) Protein gel blot analysis of the indicated amounts of total soluble protein extracts from wild-type (WT) and chaos plants. The blot was probed with antibodies directed against the CAO-encoded protein.
(D) Protein gel blot analysis of stromal and thylakoid proteins from the wild type. The blot at the top was probed with antibodies directed against the CAO-encoded protein. Controls were performed with antibodies raised against light-harvesting complex II purified from pea thylakoids (bottom) and with antibodies raised against chloroplast Hsp70 purified from pea stroma (middle; Yuan et al., 1993).
DISCUSSION
In this study, we have described a leaf pigmentation mutant, chaos, that resulted from the insertion of an enhancer trap Ds element at the CAO locus of Arabidopsis. Sequencing of the interrupted gene revealed one open reading frame containing two types of motifs involved in protein–protein interactions. Consistent with the phenotype and based on biochemical studies, we have determined that CAO encodes cpSRP43, an SRP protein that appears to be found only in the plastid. Indeed, the CAO-encoded Arabidopsis homolog of cpSRP43 has no apparent homology to any other known SRP subunits.
What is the function of cpSRP43 and why is it unique to cpSRP? Because normal levels of cpSRP54 accumulate in chaos (data not shown), we can exclude the possibility that cpSRP43 is required to stabilize cpSRP54. The presence of an RNA component has been described for all cytoplasmic SRPs studied, including those from animals, plants, and prokaryotes (reviewed in Rapoport, 1992). Recently, putative SRP RNAs have been identified in the plastid genome of a red alga and a diatom, but no such candidates have been detected in higher plants (Packer and Howe, 1998). One possibility is that cpSRP43, an acidic protein with a calculated pI of 4.3, might be a functional equivalent of the SRP RNA. However, we note that cpSRP is not only structurally distinct from cytoplasmic SRP but possesses the unusual feature of being able to interact post-translationally with its substrate, LHCP. Chemical cross-linking revealed a direct interaction between LHCPs and cpSRP54 when the stroma was incubated with LHCPs, whereas cpSRP54 alone was inactive (Li et al., 1995). These results are consistent with the idea that cpSRP43 configures cpSRP54 into a form capable of binding LHCPs post-translationally.
The CAO-Encoded Protein Is an Arabidopsis Homolog of the Pea CpSRP43.
(A) Comparison of peptide sequence data from pea cpSRP43 and the CAO-encoded protein (CHAOS). Amino acids conserved or closely related are underlined. The numbering of the sequence refers to their positions in the protein. The pea peptide listed at the top of each pair of sequences was obtained by sequencing the N-terminal end of the pea cpSRP43.
(B) Antisera against the CAO-encoded protein recognizes cpSRP43. CpSRP was immunoprecipitated with antibodies raised against cpSRP54 (α-54) from pea stroma (input, 90 μg of chlorophyll), as described in Methods. Immunoprecipitation was also performed with an antiserum raised against a nucleotide binding protein (α-NBP) from Synechococcus sp strain PCC7942. Proteins remaining in the supernatant (spnt) were concentrated by trichloroacetic acid precipitation. The CAO-encoded protein in the supernatant and coimmunoprecipitates (co-IP) was detected as a 43-kD band by immunoblot analysis after SDS-PAGE and electrotransfer to nitrocellulose membranes.
The DNA sequence of cpSRP43 revealed two types of motifs that can mediate protein–protein interactions: four ankyrin repeats and two closely spaced chromodomains situated at the C terminus. This makes cpSRP43 the second plant chromoprotein candidate to be characterized (Henikoff and Comai, 1998). Although we expect plants to contain chromoproteins that bind chromatin, as described in other eukaryotes, cpSRP43 is noteworthy because it is a novel example of a nonnuclear chromoprotein. Chromoproteins are not known in prokaryotes, suggesting that cpSRP43 evolved from eukaryotic genes relatively recently. Chromodomains frequently are found in pairs that function independently (Platero et al., 1995). The two cpSRP43 chromodomains are more closely spaced than usual (Aasland and Stewart, 1995), possibly creating a single but more stable binding target. Because chromodomains are known to be involved in dimerization (Cowell and Austin, 1997), one possibility is that this motif mediates the self-association of cpSRP43, which was recently found to be a dimer (D. Schuenemann and N.E. Hoffman, unpublished results). Ankyrin was first described as an animal protein that couples integral membrane proteins to spectrin. The occurrence of ankyrin motifs in functionally diverse proteins from animals, yeasts, plants, and even prokaryotes is inconsistent with a highly specialized function. Indeed, ankyrin repeats have been implicated in protein–protein interactions between heterologous sequences (Thompson et al., 1991). As such, we speculate that the ankyrin-repeat domain of cpSRP43 is the region that binds cpSRP54.
The salient characteristic of the phenotype of the chaos mutant is the specific effect on the light-harvesting apparatus. The observed chlorosis was due to a partial lack of photosynthetic pigments. Pigments most affected were those associated with LHCPs. The mutation reduced light-harvesting efficiency but did not alter the photochemical abilities of PSI and PSII; neither did it reduce the steady state levels of core complex and PSII internal antenna proteins (CP47 and CP43). In contrast, at least two LHC proteins were found at lesser amounts in the mutant. Other LHCPs may also be affected because the reduction of the trimer proteins and CP24 appears to be insufficient to account for the observed decreases in pigment levels (Table 1; Yamamoto and Bassi, 1996). In addition, in the expected size range for LHCPs, several thylakoid membrane polypeptides appear to be depleted in chaos. Earlier biochemical studies have indicated that cpSRP54 is required for LHCP biogenesis in vitro (Li et al., 1995). More recent in vitro studies have also demonstrated a requirement for both cpSRP54 and cpSRP43 (Schuenemann et al., 1998). Because of the abundance of LHCPs in the thylakoid membrane, efficient targeting is important. The reduced level of LHCP in the chaos mutant now provides genetic confirmation that cpSRP is utilized for LHCP trafficking.
In vitro, LHCPs require the cpSRP pathway for integration into the thylakoid membrane (Li et al., 1995). However, a substantial amount of LHCP is integrated into the thylakoid membranes of the chaos mutant. This may be due to partial activity of cpSRP54 in the absence of cpSRP43 or to an alternative targeting pathway that compensates for the defect in cpSRP. Such an alternative pathway (reviewed in Cline and Henry, 1996; Klösgen, 1997) may require intact chloroplast to be functional. In this case, it would not have been detected with the in vitro experiments performed. In these experiments, chloroplast extracts were used to investigate the role of the cpSRP pathway for LHCP integration into the thylakoid membrane (Li et al., 1995).
Plants containing reduced levels of cpSRP54 have been obtained by introducing a transgenic copy of cpSRP54 with a dominant negative mutation (Pilgrim et al., 1998). Surprisingly, these mutants have a phenotype different from that of chaos. The cpSRP54 mutants produce yellow first true leaves that gradually become green. The chlorophyll a/b ratio is identical to that of the wild type in the first true leaves that are yellow, indicating that the mutation affects proteins besides LHCPs. These first true leaves were not observed in the chaos mutant, and the chlorophyll a/b ratio was significantly increased. The recent isolation of true null cpSRP54 mutants exhibiting the same phenotype (D. Sy, P. Amin, M. Pilgrim, and N.E. Hoffman, unpublished data) has confirmed that the phenotype is due to cpSRP54 deficiency.
The fact that cpSRP54 and cpSRP43 mutants have different phenotypes suggests that the two proteins do not always function as a complex. Although most of cpSRP43 is bound to cpSRP54 (Figure 6), none binds to ribosomes, which contain approximately half of the plastid cpSRP54 (Franklin and Hoffman, 1993; Schuenemann et al., 1998). Thus, cpSRP54 may have a function on the ribosome that is independent of cpSRP43. One idea currently being tested is that ribosome-bound cpSRP54 mediates the cotranslational targeting of certain chloroplast-encoded thylakoid proteins, whereas the complex containing cpSRP43 functions primarily in LHCP biogenesis.
METHODS
Plant Material and Growth Conditions
Transgenic plants carrying an enhancer trap Dissociation (Ds) element were described previously (Klimyuk et al., 1995). Arabidopsis thaliana ecotype Landsberg erecta (Ler-0) self-progeny of the line 348/74/A segregated a chlorotic mutant named chaos. After in vitro germination of seeds at 23 and 17°C with an 8-hr photoperiod, young plantlets were transferred to soil in a growth chamber at 23 and 17°C with an 8-hr photoperiod and 60 to 85% relative humidity. The effect of light on the mutant was investigated by using four different light intensities (100, 300, 600, and 1000 μmol m−2 sec−1) for growing the plants. All molecular and biochemical analyses were performed with leaves harvested at the rosette stage.
Electron Microscopy
Leaves were fixed as described by Lascève et al. (1987). The micrographs were taken with a JEM 1200 EXII electron microscope (JEOL, Tokyo, Japan).
Pigments Analysis
Leaf discs 0.8 cm in diameter were frozen in liquid nitrogen before pigment analysis was performed according Havaux and Tardy (1996).
Chlorophyll Fluorescence and Photosynthetic Oxygen Evolution Measurements
Photosystem II Efficiency
The chlorophyll fluorescence measurements were conducted with intact leaves by using a PAM-2000 fluorometer (Walz, Effeltrich, Germany), as previously described (Havaux and Davaud, 1994). The maximal quantum yield of photosystem II (PSII) photochemistry (Φ) was estimated in dark-adapted leaves (30 min) with an Fv/Fmax ratio. Fv (variable chlorophyll fluorescence) = (Fmax − Fo), where Fo is the initial level of fluorescence under red light (655 nm) modulated at 600 Hz, and Fmax is the maximal level of chlorophyll fluorescence induced by a 800-msec pulse of strong white light (photon flux density >4000 μmol m−2 sec−1). The actual quantum yield of PSII in illuminated leaves was measured as (Fs − Fmax)/Fmax, where Fs is the actual fluorescence level. This parameter is a good indicator of the quantum yield for whole-chain electron transport (involving both PSI and PSII).
Measurement of the Emerson Effect
The quantum yield of photosynthetic oxygen evolution was measured on leaf discs 0.8 cm in diameter by the photoacoustic technique, as previously described by Havaux and Davaud (1994). The samples were illuminated with 26 μmol m−2 sec−1 blue-green light (400 to 500 nm) modulated at 20 Hz. The Emerson effect (E) was measured in samples adapted to far-red light (730 nm; 45 W m−2) for 10 min (state 1) or to blue-green light (state 2). E = O2 (+far-red light)/O2 (−far-red light). The absolute rate of oxygen evolution in white light was measured with a Clark-type oxygen electrode (LD2/2; Hansatech Instruments Ltd., King's Lynn, UK), as described by Havaux and Davaud (1994).
Chloroplast Isolation and Pigment Protein Electrophoresis
The experimental procedures for pigment protein electrophoresis (chloroplast isolation, pigment–protein complex solubilization, and separation by partially denaturing 11% PAGE) were conducted as described by Dörmann et al. (1995).
For assaying the CAO-encoded protein in the stroma and thylakoids, we first isolated Arabidopsis or pea chloroplasts as described by Spector et al. (1998). Chloroplasts were resuspended in 20 mM Hepes-KOH, pH 8.0, and 1 mM phenylmethylsulfonyl fluoride to a final chlorophyll concentration of 1.5 mg/mL. The lysed chloroplasts were centrifuged at 50,000 rpm for 5 min. The clear supernatant was used as stroma. Thylakoids were washed two additional times in 0.33 M sorbitol and 50 mM Hepes-KOH, pH 8.0, and resuspended in a solution of 20 mM Hepes-KOH, pH 8.0, 1% SDS, and 1 mM phenylmethylsulfonyl fluoride. Protein was measured using the bicinchoninic acid reagent (Smith et al., 1985).
PAGE and Immunoblotting
PAGE and immunoblotting were performed as described by Meurer et al. (1996).
Isolation of Ds Flanking Sequences and Genomic and cDNA Clones of CAO
One sequence flanking the 5′ end of Ds was isolated by inverse polymerase chain reaction (PCR). chaos DNA was digested by Sau3AI and self-ligated. Two PCRs (94°C for 20 sec, 58°C for 20 sec, and 72°C for 60 sec for 35 cycles) were performed, the first one with oligonucleotides B34 (5′-ACGGTCGGTACGGGATTTTCCCA-3′) and B35 (5′-TATCGTATAACCGATTTTGTTAGTT-3′) and the second one with nested primers D73 (5′-TTTCCCATCCTACTTTCATCCCTG-3′) and D74 (5′-CCGATTTTGTTAGTTTTATCCCGC-3′). The 300-bp product generated was cloned into pBluescript KS+ (Stratagene, La Jolla, CA). This probe was used to screen an Arabidopsis genomic library (kindly provided by C. Lister and C. Dean, John Innes Centre). One positive clone was recovered, and its 6-kb BamHI fragment hybridizing with the inverse PCR probe was sequenced. A 1.3-kb PstI-SalI subfragment of this genomic clone was used as a probe to screen a λ ZAPII cDNA library (Stratagene) constructed from poly(A)+ RNA from 20-day-old Ler-0 plants (Parker et al., 1997). Four clones (pLMC C6, pLMC C8, pLMC F1, and pLMC F2) were isolated and sequenced. The GenBank accession number for CAO is AF013115.
mRNA Isolation and Analysis
Total RNA was isolated from 0.5 g of Arabidopsis tissue, as described by Verwoerd et al. (1989). The purification of mRNA was performed with a Dynabeads mRNA purification kit (Dynal, Oslo, Norway), using 100 μg of total RNA for each sample. RNA gel blot analyses were performed under stringent conditions (Sambrook et al., 1989) by using the 1.3-kb PstI-SalI fragment of the CAO gene and the 1.6-kb EcoRI-SalI cDNA fragment of the nuclear-encoded β subunit of the mitochondrial ATPase from Nicotiana plumbaginifolia (Boutry and Chua, 1985). Quantification of the signals from the hybridized filter was performed using a Storm 840 PhosphorImager (Molecular Dynamics Inc., Sunnyvale, CA).
DNA Isolation and Analysis
Total DNA was isolated from the plant tissue and digested with different restriction enzymes, as described previously (Klimyuk et al., 1995). DNA gel blots were performed according to Sambrook et al., 1989) by using the 1.3-kb PstI-SalI fragment of the CAO genomic clone as probe.
Expression of the CAO-Encoded Protein in Escherichia coli
The cloned CAO cDNA was used as the template for PCR with primer L5 (5′-AAGATCATTGGATCCCGAACGGCGGGGGAAGGAGC-3′), which contains a BamHI site and hybridizes 270 bp after the ATG codon, and primer L4 (5 ′-GTAATCATCGGATTCAATCATTCATTCATT-3′), which introduced an EcoRI site after the stop codon. The PCR product was then cloned into the BamHI and the EcoRI sites of the pGEX4T3 vector (Pharmacia, Uppsala, Sweden), and the resulting plasmid was named pLMC11. The recombinant protein was produced in E. coli, purified with the bulk glutathione S-transferase purification kit (Pharmacia), and named ΔCHAOS.
Immunodetection of the Protein Encoded by CAO
Two batches of polyclonal IgY raised against ΔCHAOS (antibodies anti-CHAOS 30 and 39) were obtained from the eggs of two chickens injected with the purified recombinant protein. The protein encoded by CAO was detected using antibodies raised to chicken IgY coupled with horseradish peroxidase (303-035-003; Jackson ImmunoResearch Laboratories, West Grove, PA). Light-harvesting complex II was detected using antibodies to rabbit IgG coupled to alkaline phosphatase (170-6518; Bio-Rad).
Immunoprecipitation of the Signal Recognition Particle Complex
Protein A–Sepharose (Sigma) was swollen, washed in 1 × TBS (10 mM Tris-Cl, pH 7.5, and 150 mM NaCl), resuspended in TBS to a final volume of 10% (w/v), and rotated end over end overnight at 4°C with the indicated antisera (4.8 mg of IgG and 3 mg of protein A–Sepharose). The beads were washed three times in 1 × TBS and twice in IP buffer (20 mM Hepes-KOH, pH 8.0, 5 mM MgCl2, and 150 mM KCl). The beads were resuspended in 375 μL of IP buffer. Chloroplasts were lysed to 3.6 mg of chlorophyll per mL in lysis buffer (20 mM Hepes-KOH, pH 8.0, 5 mM MgCl2, 1 mM DTT, and 1 mM phenylmethylsulfonyl fluoride) and centrifuged to pellet thylakoids at 135,000g for 10 min. Ribosomes were removed by centrifuging the stroma (2.5 mL) over a 1 M sucrose cushion (0.5 mL) in lysis buffer at 346,000g for 1 hr. Ribosome-free stroma (25 μL) was added to the treated beads, and the solution was rotated end over end for 4 hr at 4°C. The supernatant was removed, and the beads were washed three times in IP buffer. The bound protein was eluted by the addition of 200 μL of 2 × sample buffer and heated for 3 min at 95°C. Corresponding amounts of the diluted ribosome-free stroma, the supernatant, and the immunoprecipitate were separated by SDS-PAGE in 12% polyacrylamide gels, blotted onto nitrocellulose, and probed with the anti-chaos antiserum (1:5000), as described previously (Pilgrim et al., 1998).
Peptide Sequencing
CpSRP43 was purified as described previously (Schuenemann et al., 1998). N-terminal sequence data were obtained using automated Edman degradation chemistry at the Rockefeller University Protein/DNA Technology Center (New York, NY). Internal sequence data were obtained at the Stanford University Protein and Nucleic Acid facility (Stanford, CA), as described previously (Hwang et al., 1996).
ACKNOWLEDGMENTS
We thank Mark Coleman for the gift of the cDNA library, Ken Cline for antibodies raised against cp hsp70, Christoph Benning and Peter Doermann for their help with the green gels, and Jocelyne Jappé for advice on the electron microscopic analyses. We are very grateful to Caroline Dean and Richard Macknight for their helpful discussions. We also thank Albert-Jean Dorne for help offered during analyses of the lipid content of the mutant and Philippe Lessard, Rob Martienssen, Christophe Robaglia, Catherine Sarrobert, and Marie-Christine Thibault for critical reading of this manuscript. Protein sequence analysis was provided by Richard Winant of the Protein andNucleic Acid Facility, Beckman Center, Stanford University and the Rockefeller University Protein/DNA Technology Center, which is supported in part by the U.S. National Institutes of Health shared instrumentation grants and by funds provided by the U.S. Army and Navy for purchase of equipment. D.S. and N.E.H. were supported by grants from Deutsche Forschungsgemeinschaft, the National Science Foundation (Grant No. MCB-9507745), and the U.S. Department of Agriculture (Grant No. 96-35304-3703). Work at the Sainsbury Laboratory (V.I.K. and J.D.G.J.) is supported by the Gatsby Charitable Foundation.
Footnotes
-
↵1 These authors contributed equally to this work.
- Received August 12, 1998.
- Accepted October 23, 1998.
- Published January 1, 1999.
REFERENCES
