SEARCH

SEARCH BY CITATION

Keywords:

  • chloroplast development;
  • PPR proteins;
  • chloroplast RNA editing;
  • RNA polymerase;
  • Arabidopsis

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

RNA editing changes the sequence of many transcripts in plant organelles, but little is known about the molecular mechanisms determining the specificity of the process. In this study, we have characterized CLB19 (also known as PDE247), a gene that is required for editing of two distinct chloroplast transcripts, rpoA and clpP. Loss-of-function clb19 mutants present a yellow phenotype with impaired chloroplast development and early seedling lethality under greenhouse conditions. Transcript patterns are profoundly affected in the mutant plants, with a pattern entirely consistent with a defect in activity of the plastid-encoded RNA polymerase. CLB19 encodes a pentatricopeptide repeat protein similar to the editing specificity factors CRR4 and CRR21, but, unlike them, is implicated in editing of two target sites.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The biogenesis of chloroplasts is a complex process that occurs during early differentiation of mesophyll cells, when undifferentiated proplastids coordinately mature into chloroplasts (Lopez-Juez and Pyke, 2005). Chloroplast differentiation is modulated by environmental cues, and requires the expression of a specific set of genes, a minimal set of which are encoded within the plastids themselves. The majority of genes required for chloroplast functions are encoded in the nucleus, and their proteins are transported post-translationally into the organelle (Barkan and Goldschmidt-Clermont, 2000). Nucleus-encoded chloroplast proteins are also largely responsible for expression and regulation of organelle genes. In recent years, the major role that post-transcriptional and translational regulation plays in plastid gene expression has become apparent (Herrin and Nickelsen, 2004; Stern et al., 2004).

Characterization of various chloroplast biogenesis mutants has shown that a significant proportion affect proteins that have roles in chloroplast RNA-processing events, including endo- and exonucleolytic cleavage, splicing and RNA editing (reviewed by Herrin and Nickelsen, 2004; Nakamura et al., 2004; Shikanai, 2006). Several of these regulatory proteins are members of the pentatricopeptide repeat (PPR) family, characterized by tandem repeats of a degenerate 35-amino-acid structural motif (Andres et al., 2007; Saha et al., 2007; Small and Peeters, 2000).

The PPR family is one of the largest protein families in plants, with approximately 450 members in Arabidopsis (Lurin et al., 2004). Computational analysis has predicted that more than 70% of the plant PPR proteins are targeted to either mitochondria or plastids (Small et al., 2004). Structural analysis has suggested that PPRs might function as RNA-binding proteins (Small and Peeters, 2000; Williams and Barkan, 2003). This binding activity has been corroborated in vitro for some family members (Lahmy et al., 2000; Lurin et al., 2004; Mancebo et al., 2001; Meierhoff et al., 2003; Nakamura et al., 2003; Okuda et al., 2006). More recently, in vivo interaction between these proteins and particular RNAs has been shown through direct co-immunoprecipitation of the specific targets with antibodies against a particular PPR protein (Schmitz-Linneweber et al., 2005, 2006). These results have led to the generally accepted view that PPR proteins act as sequence-specific adaptors, bringing catalytic proteins to the correct site on the correct transcript at the correct time (Delannoy et al., 2007; Lurin et al., 2004). Despite the high number of genes in the PPR family and the sequence similarity between some of them, mutations in specific members often result in specific and severe phenotypes (Ding et al., 2006; Gothandam et al., 2005; Gutierrez-Marcos et al., 2007; Lurin et al., 2004). The molecular functions of only a few PPR proteins have been characterized in detail, and many questions remain unanswered, such as how their specificity of action is achieved and which transcripts are targets for each particular PPR protein.

In this study, we report the characterization of clb19, a pale-yellow Arabidopsis mutant lacking expression of a PPR gene (CLB19) that is essential for correct chloroplast development. Our results demonstrate that this protein is required for the correct processing of various chloroplast transcripts and for RNA editing of the chloroplast clpP and rpoA transcripts. In the context of research on PPR protein specificity, the molecular phenotype of this mutant is particularly interesting as it implies that the CLB19 protein can recognize two different RNA sequences.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Phenotypic characterization of the clb19-1 mutant

In a genetic screen for mutants that affect chloroplast biogenesis (clb), we selected several plants that accumulate low levels of photosynthetic pigments (Gutiérrez-Nava et al., 2004). One of them, clb19-1, develops pale-yellow cotyledons and seedling leaves when grown under standard light conditions and dies shortly afterwards (Figure 1a,b). When supplemented by sucrose as a carbon source, clb19-1 continues development, but with low pigment levels (Table 1), producing numerous leaves (Figure 1f), an inflorescence stem (Figure 1g), flowers (Figure 1g,h), siliques (Figure 1g), viable pollen (Figure 1i) and even viable seeds (data not shown). Development takes considerably longer than for wild-type plants. Neither viability nor pigmentation of this mutant are restored under low-light growing conditions (Figure 1c and Table 1) that reduce photo-oxidative damage (Oelmüller, 1989). Therefore, this phenotype is a direct consequence of the clb19-1 mutation and is not caused by photo-oxidative stress.

image

Figure 1.  Phenotype of clb19 mutants. (a–c) Morphology of 15-day-old wild-type (a, c) and clb19-1 (b, c) seedlings grown under light conditions of 120 μE m−2 sec−1 (a, b) or 5 μE m−2 sec−1 (c). (d, e) Transmission electron micrographs of plastids from 15-day-old leaves of wild-type (d) and clb19-1 (e). Scale bar = 1 μm. The arrow in (e) indicates the internal membrane structures observed in the mutant plant. (f–h) Morphology of a 60-day-old clb19-1 mutant grown in germination medium, showing pale-green leaves (f), siliques (g) and flowers (h). The arrow in (g) indicates the siliques. (i) Acetocarmine-stained pollen grains from wild-type and clb19-1. (j) Molecular complementation of a representative clb19-1 kanamycin/hygromycin-resistant 15-day-old transgenic seedling containing wild-type CLB19 cDNA.

Download figure to PowerPoint

Table 1.   Chlorophyll and carotenoid content of clb19 under various light conditions
Light conditionsPlantChlorophyll aChlorophyll bTotal chlorophyllTotal carotenoids
  1. Pigments were extracted from 18-day-old seedlings grown under standard or low-light conditions. Values are expressed as microgram pigment per gram fresh weight of seedling tissue. Pigments were extracted for 5 h at room temperature using a 2:1 v/v methanol:dicloromethane mixture in glass vials protected from light. Pigment concentration was estimated in a Beckman (http://www.beckmancoulter.com) DUR650 spectrophotometer at 645 nm (chlorophyll a) or 663 nm (chlorophyll b). Values are means ± SEM.

StandardCol-0 wild-type 70.4 ± 2031.58 ± 7.3104.9 ± 29.534.4 ± 9.9
clb1922.2 ± 7.5515.01 ± 4.637.2 ± 12.112.6 ± 4
Low lightCol-0 wild-type31.8 ± 4.911.29 ± 1.942.95 ± 6.812.9 ± 2
clb194 ± 1.85.94 ± 3.39.9 ± 58.1 ± 2.6

The leaf plastids of clb19-1 are irregularly shaped (Figure 1e) and smaller than wild-type chloroplasts (Figure 1d). Internal short linear appressed membranes are present, but no membrane stacking is observed. These features indicate that organelle maturation is arrested at an early stage of chloroplast biogenesis (Vothknecht and Westhoff, 2001).

The CLB19 gene encodes a PPR protein

The yellow phenotype of the clb19-1 mutant is inherited as a single recessive allele, through several generations. This mutant was isolated from a T-DNA insertion collection (Gutiérrez-Nava et al., 2004), and analysis of T3 families showed 100% co-segregation between the mutant phenotype and the kanamycin resistance marker. Sequences flanking the T-DNA insertion site were isolated from a partial clb19-1 genomic library screened with the left border of the T-DNA. The 125 bp genomic sequence adjacent to the T-DNA was identical to the central region of gene At1g05750 (henceforth named CLB19) in the Arabidopsis genome. This makes clb19-1 almost certainly allelic to pigment defective 247, of which two alleles are described as presenting pale-yellow embryos in the SeedGenes database (Tzafrir et al., 2003).

The At1g05750 locus encodes a predicted ORF of 500 amino acids that lacks introns. In the clb19-1 mutant, the T-DNA was inserted at codon 230 (Figure 2a). Two additional mutant alleles (SALK_104250 and SALK_123752) were identified from the SALK insertion mutant collection (Alonso et al., 2003), and designated clb19-2 and clb19-3. Phenotypic characterization of these lines showed seedlings with a pale-yellow phenotype, indistinguishable from the phenotype of clb19-1 (data not shown), and also segregating with a 3:1 ratio. The insertion sites for clb19-2 and clb19-3 (verified by PCR and sequencing) are indicated in Figure 2(a,b). Final confirmation that the yellow phenotype in the mutants was due to a defect in expression of CLB19 was obtained from molecular complementation. A full-length CLB19 cDNA was introduced into the clb19-1 mutant via Agrobacterium tumefaciens-mediated transformation (Bechtold et al., 1993). A representative transgenic T2 plant derived from the complementation test is shown in Figure 1(j), displaying a normal green phenotype. PCR analysis corroborated the genotype of the complemented lines using primers for the CLB19 gene, the kanamycin resistance marker (inserted in the clb19-1 locus) and the hygromycin resistance marker present in the 35S::CLB19 transgene (data not shown). No yellow plants segregated in successive generations amongst progeny of hygromycin- and kanamycin-resistant individuals. The CLB19 sequence contains 10 PPR or PPR-like motifs (P, L, L2 or S) present in tandem (Lurin et al., 2004). These motifs are present in a triple motif arrangement (P-L-S) (Figure 2b and Figure S1). In addition to the PPR motifs, the C-terminus of CLB19 contains two extra motifs, E and E+. Thus, based on this structure, CLB19 belongs to the E/E+ subclass as proposed by Lurin et al. (2004).

image

Figure 2. CLB19 gene structure. (a) Amino acid sequence of the CLB19 gene. The putative plastid transit peptide sequence (54 amino acids) identified by TargetP (http://www.cbs.dtu.dk/services/TargetP) is underlined; the T-DNA insertion sites in the clb19-1, clb19-2 and clb19-3 alleles are indicated by open triangles. (b) Diagrammatic representation of the PPR motifs within the CLB19 protein. The designations of the P, L, L2, S, and C-terminal motifs (E and E+) correspond to those proposed by Lurin et al. (2004). cTP indicates the plastid transit peptide. (c) Transcript expression profile of clb19-1 and the complemented transgenic line (TC). Northern blot analysis was performed using 10 μg of total RNA from 15-day-old seedlings of Col-0 (wild-type), clb19-1 mutant and the complemented transgenic line. The complete CLB19 cDNA was used as a probe. Hybridization to 25S rRNA is shown as a loading control. (d) Accumulation of the CLB19 protein in wild-type, clb19-1 and clb19-2 plants and in a representative transgenic line (TC). Western blot analysis was performed using 10 μg of total protein from 15-day-old seedlings and CLB19 polyclonal antibodies. Gels stained with Coomassie brillant blue (BBG) are shown as a loading control. Membranes are representative of at least two biologically independent experiments.

Download figure to PowerPoint

Neither transcripts nor protein expressed from the CLB19 gene are detectable in the clb19-1 mutant in contrast to wild-type plants and the complemented transgenic plants (Figure 2c,d). These data demonstrate that the clb19-1 and clb19-2 mutants are null alleles. The CLB19 antibody does not show significant cross-reactivity with other proteins in the mutant.

The CLB19 protein is targeted to plastids

The prediction programs TargetP and Predotar (Emanuelsson et al., 1999; Small et al., 2004) predict that the 54 amino acids at the N-terminus of CLB19 could function as a putative plastid transit peptide (Figure 2a). The N-terminal 92 codons of CLB19 were fused to a GFP gene. This fusion (CLB19–GFP) was introduced into Arabidopsis cells by DNA bombardment (Figure 3a). A second construct that includes an additional 131 bp upstream of the ATG codon of the CLB19 ORF was also generated (5UTRCLB19–GFP), as recent reports indicate that 5′ UTR sequences can affect targeting (Christensen et al., 2005; Millar et al., 2006). Analysis by confocal imaging of Arabidopsis leaves after cell bombardment with either CLB19–GFP (Figure 3b,d) or 5UTRCLB19–GFP (Figure 3e,g), showed fluorescence exclusively in chloroplasts. The GFP co-localizes with chlorophyll autofluorescence (Figure 3c,d,f,g).

image

Figure 3.  Subcellular localization of CLB19. (a) Schematic representation of the two GFP fusion constructs used in subcellular localization assays. The CLB19–GFP construct contains the putative chloroplast transit peptide (cTP) and a part of the mature CLB19 (CR). The 5′UTRCLB19–GFP construct contains an additional 121 bp of the 5′ UTR immediately upstream of the CLB19 ATG codon. (b–g) Localization of GFP in Arabidopsis 15-day-old leaf cells bombarded with CLB19–GFP (b–d) or 5′UTRCLB19–GFP (e–g) constructs using confocal microscopy (b, e) and chlorophyll autoflorescence (c, f). Merged images for each construct are shown in (d) and (g). (h) CLB19 accumulates in the chloroplast-enriched protein fraction of Arabidopsis leaves. Aliquots (5 μg) of proteins from an enriched chloroplast fraction (C), the depleted supernatant (S), or total protein extracts (T) were resolved by SDS–PAGE. Immunodetection was performed using polyclonal antibodies to CLB19 or the beta subunit of ATP synthase (AtpB). Gels stained with Coomassie brilliant blue are shown as a loading control.

Download figure to PowerPoint

The chloroplast localization of CLB19 was further corroborated by immunoblot analysis after chloroplast enrichment. As shown in Figure 3h, CLB19 is detected in total leaf protein extracts (T) and enriched in the chloroplast protein fractions (C). This pattern is similar to that for other chloroplast proteins such as AtpB (Figure 3h). CLB19 is undetectable in the non-chloroplast fraction (S).

CLB19 is required for efficient expression of various chloroplast transcripts

To address the role of CLB19 protein in chloroplast development, the transcript profiles of various chloroplast genes were compared between clb19 and wild-type plants (Figure 4a). This analysis was performed under low light to reduce photo-oxidative damage (Oelmüller, 1989). Three different behaviors were observed among the genes analyzed: some genes, including psbA and rrn16, showed decreased levels but displayed a pattern basically indistinguishable from that of the wild-type plant, which suggest that either their transcription or their transcript stability might be affected. The transcripts for other genes (psbB, petD, psaB, rrn23 and rbcL) also showed a substantial decrease in accumulation of the typical mature transcripts, but, in addition, novel transcripts for these genes were observed. Finally, transcripts such as atpB, atpE and accD also showed alterations in their transcript pattern, but, in contrast to the previous examples, a significant over-accumulation of transcripts was observed (Figure 4a). These transcript changes are accompanied by dramatic alterations in the levels of many chloroplast proteins, notably a massive decrease in the amounts of RbcL and components of the photosystem complexes, while others, such as AtpB, are less affected (Figure S2).

image

Figure 4.  Expression analysis of chloroplast-endoded genes in the clb19 and ptac2 mutants. (a) Aliquots (5 μg) of total RNA from 15-day-old wild-type (Wt) and clb19 (19) plants grown under low-light conditions were fractionated on agarose gels. Hybridization was performed with the genes specified in each lane. Black bars below the gels indicate that the gene is transcribed by PEP, shaded bars indicate transcription by NEP, and black/shaded bars indicate transcription by both PEP and NEP. The nuclear-encoded cytosolic 25S rRNA is shown as a loading control. (b) The transcript abundance of all protein-encoding and ribosomal genes of the chloroplastic genome in the mutants (clb19-1 and ptac2) and complemented transgenic line (TC) displaying a wild-type phenotype was measured by quantitative RT-PCR. Values are log2 of the induction factor (IF), where IF is the ratio of the mutant to the wild-type transcript abundance. The genes are sorted according to their physical location on the chloroplastic chromosome. accD and rbcL transcripts are indicated by arrows, and the full data set with the list of genes is given in Table S1.

Download figure to PowerPoint

All the transcripts that display higher levels in clb19 are at least in part transcribed by the nuclear-encoded RNA polymerase (NEP) (Legen et al., 2002; Serino and Maliga, 1998; Swiatecka-Hagenbruch et al., 2007), whereas those with low levels have been reported to be transcribed by the plastid-encoded RNA polymerase (PEP) (Legen et al., 2002). Moreover, these transcriptional patterns resemble those observed in tobacco PEP mutants (Allison et al., 1996; Krause et al., 2000). Although no Arabidopsis PEP mutants have been described, a set of mutants (named ptac) that are affected in genes encoding proteins associated with transcriptionally active plastid DNA have similar phenotypes (Pfalz et al., 2006). To verify whether the alterations in transcript patterns in clb19 were consistent with alterations in PEP activity, a systematic comparison of transcript levels was carried out between clb19 and ptac2. PTAC2 encodes a PPR protein that is unrelated to CLB19 but has been shown to be associated with plastid transcription complexes by proteomics analysis (Pfalz et al., 2006). The two mutants display very similar and characteristic changes (Figure 4b and Table S1). This was also corroborated by Northern analysis for the rbcL and accD genes (data not shown).

To characterize more precisely the type of alterations caused by the absence of CLB19, we focused on a chloroplast region that contains two of the mono-cistronic units whose transcriptional pattern was strongly affected in clb19 plants, the rbcL and accD genes. RNA hybridization experiments were conducted using sequential probes that cover the coding sequences as well as the 5′ and the 3′ untranslated regions of these genes (Figure 5a). The mature rbcL transcript of around 1.8 kb that normally accumulates in wild-type plants is practically non-existent in young clb19 seedlings. However, a longer transcript of approximately of 4 kb is detected, albeit at lower levels (Figure 5b). Sequencing of 5′ RACE products (Figure 5c) indicates that the major, highly abundant rbcL transcript in Col-0 wild-type starts at −177 with respect to the AUG start codon. In clb19, the rbcL transcript starts from almost 2 kb upstream and includes part of the complementary sequence of the atpB gene. This is entirely consistent with the low-resolution mapping data from both the Northern hybridizations (Figure 5b) and with a defect in transcript initiation by PEP. This analysis also demonstrates that these processing defects are not a consequence of overall chloroplast arrest, as they are not observed in the unrelated albino clb6 mutant (Figure 5b) that also has undifferentiated plastids (Gutiérrez-Nava et al., 2004).

image

Figure 5.  Expression analysis of the atpE, atpB, rbcL and accD regions in the clb19 mutant. (a) Schematic representation of the chloroplast region studied. The bars indicate the regions (1–7) used as probes in (b). The black arrows indicate atpB and accD transcripts present in both wild-type and clb19 plants. The light-gray arrow indicates the major rbcL transcript in wild-type that is virtually absent in clb19. The dark-gray arrows indicate the long transcripts that are only detectable in clb19. The diagram is not to scale. (b) Northern analysis. Each lane contains 10 μg of total RNA from 15-day-old wild-type (Wt), clb19 (19) or clb6 (6) seedlings grown under low-light conditions. The probes used in each case, as indicated in (a), are: atpB ORF (1), rbcL 5′ UTR (2), rbcL ORF (3), rbcL 3′ UTR (4), accD 5′ UTR (5), accD ORF (6) and accD 3′ UTR (7). RNA molecular weight markers are shown (MW) and a methylene blue-stained membrane (MB) is shown as a loading control. (c) 5′ RACE analysis of wild-type (Wt) and clb19 mutant plants (19). The bands are nested PCR products obtained using rbcL-specific primers and SMART kit primers designed to amplify the 5′ extremities of cDNAs. Molecular weight markers are shown on the right (MW). Sequencing of the products indicated that the shorter product (in the wild-type) derives from transcripts starting at −177 with respect to the initiation codon.

Download figure to PowerPoint

Editing of two sites is abolished in clb19 mutants

In addition to checking obvious transcript differences, we also systematically examined the editing of transcripts in clb19 plants using a new high-resolution melting screen (Chateigner-Boutin and Small, 2007). Of the 34 sites known to exist in chloroplast transcripts (Chateigner-Boutin and Small, 2007; Tillich et al., 2005), 32 are edited correctly in clb19, as in the case of the ndhG site shown in Figure 6(c). The remaining two sites in the rpoA and clpP transcripts are not edited at all in clb19 mutant plants. Editing of codon 67 in wild-type rpoA transcripts encoding the alpha subunit of PEP results in a serine to phenylalanine coding change that was not detected in two independent clb19 alleles (Figure 6a). Similarly, editing of codon 187 in wild-type clpP1 transcripts encoding a catalytic subunit of the ATP-dependent ClpP serine protease (Yu and Houry, 2007) that changes a histidine into tyrosine is also not detectable in clb19 plants (Figure 6b). These two sites are edited in wild-type plants and in complemented clb19 plants grown under the same conditions, and in the non-photosynthetic otp51 mutant that is physiologically similar to clb19. Neither of these two sites are particularly variable sites under different growth conditions, i.e. they are edited to a high degree even in etiolated plants (Chateigner-Boutin and Small, 2007). Both editing events result in the restoration of codons for amino acids that are conserved at the same positions in other organisms (Figure S3), and thus the non-editing of these sites in clb19 mutants might be expected to deleteriously affect the function of RpoA and ClpP1.

image

Figure 6.  Chloroplast editing defects in clb19 mutants. (a–c) Poisoned primer extension assays were conducted on editing sites clpP (69942) (a) and rpoA (78691) (b), with ndhG (118858) (c) as a control. RT-PCR products were obtained using primers surrounding the editing sites, and served as templates for the extension reaction from a 5′ 6-carboxyfluorescein (FAM)-PPE primer that anneals close to the target editing site (a forward PPE primer was used for clpP and rpoA and a reverse primer for ndhG). The extension was stopped by incorporation of ddCTP for clpP and rpoA or ddGTP for ndhG at the location of the editing site (producing a short product if the template is unedited) or at the subsequent C/G (producing a longer product if the template is edited). Data are shown for 15-day-old seedlings of the alleles clb19-1 and clb19-3, for complemented clb19-1 (TC), and compared to seedlings of Col-0 (wild-type) and from another mutant showing a severe chloroplast biogenesis defect (otp51) grown under the same conditions. (d) Sequence alignment of the region surrounding the rpoA (78691) and clpP (69942) editing sites. The alignment includes the sequence from −30 to +30 around the edited C (upper case), with identical nucleotides shown in bold.

Download figure to PowerPoint

The RpoA protein fails to accumulate normally in clb19 mutants

The molecular phenotype of the clb19 mutant is consistent with a defect in PEP function. Thus expression of the four plastid-encoded subunits of PEP was analyzed. As shown in Figure 7(a), no visible alterations in the rpoA, rpoB, rpoC1 and rpoC2 transcript patterns were observed in the clb19 mutant compared to wild-type. Thus, neither the defect in PEP function nor the defect in rpoA editing can be attributed to a problem in transcription or processing of rpo transcripts. However, the level of immunodetectable RpoA protein is lower in clb19 seedlings (Figure 7b). This decrease cannot be attributed to a decrease in rpoA mRNA, and suggests that the lack of editing of rpoA transcripts has a critical effect on the translation or stability of the RpoA protein.

image

Figure 7.  Analyses of the chloroplastic-encoded PEP subunits in wild-type and in clb19 mutant. (a) For analysis of the chloroplast-encoded rpo transcripts, 5 μg of total RNA from 15-day-old wild-type (Wt) and clb19 (19) plants grown under low-light conditions were fractionated on agarose gels. Hybridization was performed using the rpoA, rpoB, rpoC1 and rpoC2 genes encoding the four subunits of the plastid-encoded RNA polymerase (PEP). The cytosolic 25S rRNA is shown as loading control, and the size of the ribosomal markers is shown. The asterisks correspond to the transcript size reported for each rpo transcript. (b) For protein analysis, total protein extracts were obtained from 15-day-old seedlings grown under low-light conditions. Immunoblots were performed using antibodies against the RpoA polypeptide. Each lane contains 10 μg of total protein extract. Gels stained with Coomassie brilliant blue (BBG) are shown as loading controls.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

In this paper, we present the identification and characterization of CLB19, a novel PPR protein that plays an essential role in proper chloroplast development in Arabidopsis thaliana. The lethal phenotype caused by the absence of CLB19 demonstrates its unique role during early stages of seedling development. CLB19 is targeted to chloroplasts, and its absence results in a complex phenotype including alterations to transcripts for many chloroplast genes. As it is known that transcription, processing, mRNA stability and translation are often coupled in plastids (Barkan et al., 1995), CLB19 could participate in one or more of these mechanisms. Phenotypes similar to those of clb19 have been found in PPR mutants that affect general chloroplast translation in maize (Schmitz-Linneweber et al., 2006; Williams and Barkan, 2003). However, our data show that general translation is not totally impaired in clb19, as, although drastic decreases are observed in proteins such as several photosystem proteins and RbcL, others such as RpoA and AtpB accumulate in this mutant at near-normal levels compared to wild-type plants (Figure S2). Our data also show that it is unlikely that the processing defects could be attributed to a role for CLB19 in general editing or splicing, because many of the transcripts affected in the clb19 mutant do not contain introns and are edited normally. Hence we believe that the transcript differences observed are secondary effects due to a primary malfunction in a basic gene expression step within the chloroplasts. The best candidates for primary defects are the two specific and total editing defects in clb19 mutants. Editing of two cytidines in the rpoA and clpP transcripts, which are highly edited in wild-type plants, is abolished in clb19 mutants.

The rpoA gene encodes the alpha subunit of PEP, which is believed to predominantly transcribe photosynthesis genes (Hajdukiewicz et al., 1997). The expression pattern of the clb19 mutant showed decreases in transcript accumulation of several genes but also over-accumulation of some transcripts. This differential pattern of transcript abundance in clb19 is very similar to those in mutants that are impaired in the function of plastid-encoded RNA polymerase. They include mutants that directly affect the RNA polymerase subunits (rpo genes) (Krause et al., 2000; Legen et al., 2002; Serino and Maliga, 1998) or those that alter the general transcription components (Pfalz et al., 2006). Similar to the case in these mutants, the transcripts of genes expressed by PEP (e.g. rbcL, psbA, psaB) are less abundant in clb19. In contrast, transcripts from genes such as atpB, accD, atpE and clpP, whose expression depends partially or completely on NEP, are more abundant in clb19 than in wild-type plants. These results exemplify the complexity underlying chloroplast transcription, and are in agreement with the finding that, although some genes may be transcribed from a single promoter, the majority of them are transcribed from multiple NEP and PEP promoters (Swiatecka-Hagenbruch et al., 2007). Similar to the case in PEP-deficient mutants (Legen et al., 2002), we observed very long transcripts in clb19 (Figures 4 and 5) that are probably initiated from cryptic NEP promoters as suggested for ptac mutants (Pfalz et al., 2006).

Editing of rpoA leads to insertion of a hydrophobic phenylalanine instead of a hydrophilic serine residue. Phenylalanine is genomically encoded at this position in several plant chloroplasts such as rice, maize and bryophytes (Figure S3). In bacteria, there is a similar hydrophobic isoleucine at the corresponding position. In Escherichia coli, the C-terminal part of the alpha subunit is involved in binding to the promoter, whereas the N-terminal part (residues 1–235) enables the assembly of the polymerase complex (Kimura et al., 1994). If the bacterial model can be applied to PEP, we can speculate that the change in the nature of the amino acid at position 67 in the N-terminal part of RpoA may prevent assembly of PEP.

The second editing defect in clb19 is in the clpP transcript encoding the catalytic subunit of the plastid ClpPRS protease complex. It is possible that a defect in ClpPRS protease activity contributes to the gross physiological phenotype of clb19 plants, because previous work in Arabidopsis has shown that a lack of the regulatory subunit (clpR2) led to a yellow phenotype associated with a reduced accumulation of clpPRS protease complex (Rudella et al., 2006). However, the defects in plastid gene expression observed are unlikely to be linked to defects in Clp protease activity, as clpR2 plants contain a normal quantity of RbcL (Rudella et al., 2006), whereas this protein is hardly detectable in clb19 seedlings.

Similarities between known editing factors

CLB19 is the third PPR protein of the E/E+ subgroup found to be essential for RNA editing of specific sites. Previously, PPR proteins CRR4 and CRR21 had been identified as specificity factors required for editing sites 1 and 2, respectively, in the ndhD transcript (Kotera et al., 2005; Okuda et al., 2006). So far, no other group of PPR proteins has been shown genetically or biochemically to be directly implicated in editing, although, based on evolutionary considerations, the related DYW sub-family of PPR proteins has been proposed to carry the catalytic function for organelle RNA editing (Salone et al., 2007). The CRR4 protein has been shown to bind just upstream of its target editing site (Okuda et al., 2006), and it seems likely that CRR21 and CLB19 function in the same way. Exchange of domains between CRR4 and CRR21 has strongly implicated the PPR motifs in determining RNA binding specificity, and the E/E+ domain is required for the editing reaction (Okuda et al., 2007). Alignment of these domains between CLB19 and the CRR4/CRR21 proteins (Figure 8) showed similarity, particularly within the conserved 15 amino acid domain noted by Okuda et al. (2006). No other molecular function has been found so far for any E/E+ PPR proteins, raising the possibility that all 100 or so members of this sub-group are editing specificity factors.

image

Figure 8.  Comparison of the predicted motif structure and C-terminal alignment of the PPR editing proteins CLB19, CRR4 and CRR21. PPR motifs P, L, L2, S, E and E+ are defined as described by Lurin et al. (2004), and the consensus sequences of the E and E+ domains are shown below. Amino acids that are conserved between the three PPR proteins are shown in bold. The sequence underlined is a motif that is highly conserved between CRR4 and CRR21, as noted by Okuda et al. (2007).

Download figure to PowerPoint

It is important to note that, although the absence of many different organelle RNA-binding proteins can lead to decreases in RNA editing of one or even many sites (Wang et al., 2006), presumably more or less indirectly through changes in RNA abundance or conformation, the crr4, crr21 and clb19 mutants differ in that the affected sites are no longer detectably edited at all, implying a much more direct requirement for these proteins.

CLB19 is involved in editing two different sites

In clb19 plants, neither the site in rpoA nor the site in clpP are edited. As editing of other sites is not noticeably affected, and neither the rpoA nor clpP transcripts show obvious quantitative or qualitative alterations apart from the absence of editing, it seems likely that both editing defects are a direct result of the absence of CLB19 protein. Mutants in either CRR4 or CRR21 lose editing of only a single site, and PPR proteins are believed to show considerable site specificity in vivo (Delannoy et al., 2007), so the participation of CLB19 in editing of two independent sites was unexpected. However, other circumstantial evidence strongly suggests that some editing factors may act at two or more sites. Previous work in transgenic chloroplasts of tobacco identified groups of editing sites, editing of which was decreased by overexpression of an additional copy of the sequence surrounding one site in either rpoB or ndhF (Chateigner-Boutin and Hanson, 2002). The finding that the sequence upstream of these affected sites contained similar nucleotides in equivalent positions led to the hypothesis that the same trans factor could recognize several editing sites that share cis elements. Subsequent in vitro analysis of the conserved nucleotides in rpoB confirmed that mutations at these locations decrease the editing efficiency (Hayes et al., 2006). The participation of one common factor at multiple editing sites has been recently demonstrated by an in vitro RNA editing system in tobacco (Kobayashi et al., 2007).

Preliminary analyses using in vitro binding and immunoprecipitation of protein–RNA complexes (Christian Schmitz-Linneweber, Humboldt University, personal communications, and M. R-V, C.A., unpublished results) did not provide conclusive evidence that CLB19 binds either or both the clpP and rpoA transcripts, but the similarity of CLB19 with CRR4 suggests that it does act in this way at both sites. These studies are technically challenging, and an absence of evidence should not be taken as proof of an absence of direct interaction. Previous studies in vivo and in vitro have demonstrated the presence of cis-acting elements around the editing site that are essential for editing (Shikanai, 2006). However, the sequences surrounding the editing sites in rpoA and clpP transcripts show no obvious contiguous conserved cis elements (Figure 6d). A similar situation was found in the tobacco rpoB cluster of editing sites (Chateigner-Boutin and Hanson, 2002). An editing factor such as CLB19 that may naturally bind two unrelated sites may be of great help in understanding how these proteins specifically recognize their target sites.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant material and growth conditions

Arabidopsis thaliana (L.) Heyhn. ecotype Columbia (Col-0) was used in this study. Seeds were germinated under sterile conditions on germination medium containing 1× Murashige and Skoog basal salts (Gibco BRL, http://www.invitrogen.com), 1% w/v sucrose, 1× Gamborg’s B5 vitamin solution (Sigma-Aldrich Inc., http://www.sigmaaldrich.com/), 0.05% w/v MES, solidified with 0.8% w/v phytoagar (Caisson Laboratories, Inc., http://www.caissonlabs.com), and adult plants were grown on Metro-Mix 200 (Sun Gro Horticulture, Inc., http://www.sungro.com). Seedlings were grown under a 16 h light/8 h dark cycle as standard conditions (120 μE m−2 sec−1) or at 5 μE m−2 sec−1 for low-light conditions, at 22°C in growth chambers, unless otherwise indicated. To break dormancy, seeds were vernalized at 4°C for 5 days. The T-DNA pools from which the clb19-1 mutant was selected were generously provided by Chris Somerville (Carnegie Institute of Washington, Stanford, CA, USA) as described previously (Gutiérrez-Nava et al., 2004). clb19-2 (SALK_104250) and clb19-3 (SALK_123752) were obtained from the ABRC Stock Center. Seeds of ptac2 (SALK_075736) mutants were kindly provided by Jeannette Pfalz (Friedrich Schiller University) and grown as described above. The positions of the T-DNA in the clb19-2 and clb19-3 alleles were confirmed by amplifying the genomic and T-DNA borders using specific primers for the T-DNA (LBb1 and LBa1) and a CLB19-specific primer (clb19-5′UTR) (Table S2).

Gene identification

The genomic sequence adjacent to the T-DNA insertion in CLB19 was obtained from a partial genomic library from the clb19-1 mutant. Total genomic DNA digested with HindIII was size-fractionated, and the region containing fragments that included left border T-DNA was cloned into the Lambda ZAP Express vector (Strategene, http://www.stratagene.com/). Clones carrying T-DNA sequences were identified through hybridization, and the region adjacent to T-DNA was analyzed by sequencing.

Mutant complementation

The complete CLB19 open reading frame was amplified by PCR from total DNA, using oligonucleotides 19F and 19R (Table S2) and used to transform heterozygous clb19-1 mutant plants (Bechtold et al., 1993). Hygromycin B-resistant (50 μg ml−1) transgenic lines were selected. Homozygous independent transgenic lines for both transgenes were identified by 100% segregation of hygromycin B and kanamycin resistance, and verified by PCR using primers for the CaMV 35S promoter and for the CLB19 coding sequence (primer 19in).

Analysis of RNA

Total RNA was isolated from plants grown under low light unless otherwise indicated using Trizol (Invitrogen, http://www.invitrogen.com/). For RNA gel-blot analysis, the RNA was fractionated by electrophoresis and transferred onto Hybond N+ nylon membranes (Amersham Biosciences, http://www5.amershambiosciences.com/). Hybridizations were performed under high-stringency conditions according to standard procedures (Church and Gilbert, 1984) using 32P-radiolabeled probes with a Megaprime DNA labeling system (Amersham Biosciences). The fragments used as probes were obtained by PCR using the following specific oligonucleotides, whose sequence is given in Table S2: primers F and R for CLB19 (1.5 kb), accDF and accDR for the accD coding region (589 bp), accD5′F and accD5′R for the accD 5′ UTR (354 bp), accD3′F and accD3′R for the accD 3′ untranslated region (413 bp), rrn23SF and rrn23SR for rrn23S (400 bp), rrn16SF and rrn16SR for rrn16S, atpBF and atpBR for the atpB probe (455 bp), rbcL3′UTRF and rbcL3′UTRR for the rbcL 3′ untranslated sequence (307 bp), rbcL5′UTRF and rbcL5′UTRR for the rbcL 5′ untranslated leader (321 bp), AtpEFw and AtpERv for atpE (330 bp), rpoAF and rpoAR for rpoA (433 bp), rpoBF and rpoBR for rpoB (697 bp), rpoC1F and rpoC1R for rpoC1 (600 bp), and rpoC2F and rpoC2R for rpoC2 (592 bp). Specific fragments were used for the Arabidopsis PC (At1g76100), CAB1 (At1g29930), RBCS (At1g67090), RPL21 (At1g33680), psaB (GenBank accession number CAB88725), psbB (GenBank accession number CAB88753), petD (GenBank accession number CAB88758) and psbA (GenBank accession number CAB88705) genes, and for the pea rbcL gene (GenBank accession number X03850). RACE analysis was performed using the SMART RACE cDNA amplification kit (Clontech-Takara Bio Europe, http://www.clontech.com) according to the manufacturer’s instructions. The first round of PCR used the rbcL-specific primer rbcl5′, and the second nested PCR used the primer rbcL5UTR2.

Plastid transcript abundance

Total RNA was extracted from 4-week-old plants using an RNeasy kit (Qiagen Inc., http://www.qiagen.com/), and genomic DNA was removed using DNA-free DNAse (Ambion, http://www.ambion.com). The absence of DNA was confirmed by PCR prior to reverse transcription with random hexanucleotide primers. Quantitative PCR was performed using a LightCycler 480 real-time PCR system (Roche, http://www.roche.com) using LightCycler 480 SYBR Green 1 master mix and the primer set detailed in Table S3. Each data set was normalized by setting the median value to 0.

Determination of the extent of editing

Poisoned primer extension (PPE) of RT-PCR products and determination of editing efficiency were performed as previously described (Chateigner-Boutin and Small, 2007). PCR templates were generated using primers clpP.AT. for and clpP.AT2.rev for clpP, chloro159_for and chloro159_rev for rpoA and ndhG.AT.for and ndhG.AT.rev for ndhG. The PPE primers and corresponding ddNTPs used were clpP_PPE_C, rpoA_PPE_C and ndhG_PPE_G (Table S2).

Antibody preparation

Polyclonal antibodies were generated against a 6×His-tagged CLB19 fusion protein containing the entire CLB19 coding sequence. The Isopropyl β-D-1-thiogalactopyranoside (IPTG)-induced protein was purified from E. coli crude extracts by affinity chromatography using Ni-NTA agarose (Qiagen Inc.) under native conditions according to the manufacturer’s recommendations. For antibody generation, 50 μg of purified 6xHis–CLB19 fusion protein in PBS (140 mm NaCl, 2.8 mm NaH2PO4, 7.2 mm Na2HPO4 pH 7.4) and complete Freund’s adjuvant (Sigma-Aldrich Inc.) were subcutaneously injected as a 1:1 v/v emulsion into 8-week-old female New Zealand rabbits. Four additional injections (50 μg each) with incomplete Freund’s adjuvant were administered at 10-day intervals starting 15 days after the initial injection. The serum obtained by total bleeding of the rabbit was collected 3 days after the last injection, and the CLB19 antibody titer was determined.

Protein gel-blot analysis

Protein analyses were performed on total protein extracts or isolated chloroplast proteins for subcellular detection. Intact chloroplasts were purified from 3-week-old Arabidopsis leaves by two-step Percoll (Sigma-Aldrich Inc.,) gradients as previously described (Voelker and Barkan, 1995). Protein concentration in samples or intact chloroplasts was determined using Bradford reagent (Bio-Rad, http://www.bio-rad.com/). Proteins were separated by SDS–PAGE and transferred onto nitrocellulose Hybond C membranes (Amersham Pharmacia Biotech, http://www5.amershambiosciences.com/). To verify equal protein loading, a parallel gel was run and stained with Coomassie brilliant blue R-250. Immunodetection was performed using the following dilutions of polyclonal antibodies: 1:1000 for CLB19, 1:100 000 for RbcL, 1:500 for PsaD, 1:1000 for PetD, 1:1000 for PetA, 1:10 000 for AtpB, and 1:1000 for RpoA. Detection of the primary antibodies was performed using alkaline phosphatase-conjugated goat anti-rabbit IgG using a BCIP/NBT substrate kit (Zymed Laboratories, http://www.zymed.com).

Subcellular localization

For subcellular localization, two different CLB19–GFP translational fusions were constructed using forward primers CLB19GFPF or 5′UTRCLB19GFP F and reverse primer CLB19GFPR for both (Table S2). CLB19–GFP contained the 92 amino acids from the N-terminus of the CLB19 protein, and 5′UTRCLB19–GFP contained an additional 121 bp from the 5′ UTR. Transient transformation of A. thaliana Col-0 was performed by bombardment of 2-week-old plantlets (Ye et al., 1990) with DNA-coated tungsten particles (Tungsten M17, Bio-Rad). GFP expression was monitored in leaf tissue by confocal microscopy 1 week after bombardment.

Confocal, light, and electron microscopy

Confocal images were obtained using a Carl Zeiss LSM510 META laser scanning microscope (http://www.zeiss.com/), equipped for excitation with an argon (Ar2) 488 nm laser. GFP and chlorophyll emissions were isolated using BP500–530 nm and BP650–710 nm filters, respectively. Transmission electron micrographs were obtained exactly as described by Mandel et al. (1996). All images were processed using Adobe Photoshop (Adobe, http://www.adobe.com).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We are grateful to Dr Alice Barkan (University of Oregon) for providing the antibodies for AtpB, PetA, PsaD and PetD proteins, Drs Kensuke Kusumi (Kyushu University) and Yuzuru Tozawa (Ehime University) for antibodies against RpoA, and Dr Herminia Loza for the antibody against RbcL. We thank Dr Patricia Dupree for help with pollen staining. We also thank Carolina San Roman and Martha Trujillo for technical support, Arturo Ocadíz for computer support and Andrés Saralegi for confocal microscopy. We acknowledge the Salk T-DNA insertion project for the T-DNA insertion lines and the Arabidopsis Biological Resource Center for providing seeds. Seeds of ptac2 mutants were kindly provided by Jeannette Pfalz (Friedrich Schiller University). This work was supported by grants from the Consejo Nacional de Ciencia y Tecnología (31791-N), the Dirección General de Asuntos para el Personal Académico-UNAM (IN218007), the French Ministry of Education and Research, the French Australian S&T (FAST) Programme (FR060030), the Australian Research Council (CE0561495), the Western Australian State Government, and the Howard Hughes Medical Institute.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Figure S1.Comparison of the predicted PPR motifs of the CLB19 protein against the consensus sequences defined in Lurin et  al. (2004).

Figure S2. Protein analysis of chloroplast photosynthetic proteins in Wt and in clb19 mutant. Total protein extracts were obtained from 15 d-old seedlings grown in low light conditions. Immunoblots were performed using antibodies against to the subunit B of ATP synthase (AtpB), the large subunit of RuBISCo (RbcL), a subunit of the cytochrome b6f complex (PetA), a subunit of photosystem I (PsaD) and a subunit of the cytocrome b/f (PetD). (a) 10 μg of total protein extract of wild-type (Wt) or clb19 were used in each lane. (b) For a more quantitative estimation of the difference in the AtpB levels, increasing amounts of total protein extract as indicated in each lane (μg) from Wt and clb19 were used. The Coomasie-stained gels (Coo.) are shown as a loading controls.

Figure S3. Alignment of RpoA and ClpP sequences. The amino acid sequences predicted by translation of representative bacterial or plastid rpoA and clpP genes were aligned using CLC Free Workbench 3.1 software (http//:www.clcbio.com). The amino acids altered by editing corroborated in Arabidopsis and Phalaenopsis (Zeng et  al., 2007, Plant Cell Physiol48: 362) are indicated by an arrow. Editing restores amino acid conservation among plants for RpoA and with bacterial sequences in the case of ClpP.

Table S1. Comparison of chloroplastic transcripts abundance for clb19 and ptac2 mutants. Transcript abundance of all protein-encoding and ribosomal genes of the chloroplastic genome from the mutants (clb19-1, clb19-3 and ptac2) and wild type was measured by quantitative RT-PCR. The genes are listed in the same order as from right to left in Figure 4(b). The standard deviation was calculated from 3 technical replicates. The primers used are given in Supplementary Table 2.

Table S2. Primers used for genotyping plant material, preparing probes and for RACE and RNA editing analyses. All primers are given by the name used in the manuscript.

Table S3. Primers used for the quantitative RT-PCR experiments described in Figure 4(b) and in Supplementary Table 1.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
TPJ_3634_sm_figS1.jpg494KSupporting info item
TPJ_3634_sm_figS2.jpg155KSupporting info item
TPJ_3634_sm_figS3.jpg558KSupporting info item
TPJ_3634_sm_tableS1.xls30KSupporting info item
TPJ_3634_sm_tableS2.doc31KSupporting info item
TPJ_3634_sm_tableS3.doc36KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.