The pentatricopeptide repeat protein OTP82 is required for RNA editing of plastid ndhB and ndhG transcripts


  • Kenji Okuda,

    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto 606-8502, Japan
    Search for more papers by this author
    • These authors contributed equally.

  • Kamel Hammani,

    1. ARC Centre of Excellence in Plant Energy Biology, Molecular and Chemical Sciences Building (M316), University of Western Australia, 35 Stirling Highway, Crawley, Perth 6009 WA, Australia
    Search for more papers by this author
    • These authors contributed equally.

  • Sandra K. Tanz,

    1. ARC Centre of Excellence in Plant Energy Biology, Molecular and Chemical Sciences Building (M316), University of Western Australia, 35 Stirling Highway, Crawley, Perth 6009 WA, Australia
    Search for more papers by this author
  • Lianwei Peng,

    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto 606-8502, Japan
    Search for more papers by this author
  • Yoichiro Fukao,

    1. Plant Science Education Unit, Graduate School of Biological Sciences, Nara Institute of Science and Technology, Takayama, Ikoma, Nara 630-0101, Japan
    Search for more papers by this author
  • Fumiyoshi Myouga,

    1. Plant Genome Network Research Team, RIKEN Plant Science Center, 1-7-22 Yokohama 203-0045, Japan
    Search for more papers by this author
  • Reiko Motohashi,

    1. Faculty of Agriculture, University of Shizuoka, Shizuoka 422-8529, Japan
    Search for more papers by this author
  • Kazuo Shinozaki,

    1. Plant Genome Network Research Team, RIKEN Plant Science Center, 1-7-22 Yokohama 203-0045, Japan
    Search for more papers by this author
  • Ian Small,

    1. ARC Centre of Excellence in Plant Energy Biology, Molecular and Chemical Sciences Building (M316), University of Western Australia, 35 Stirling Highway, Crawley, Perth 6009 WA, Australia
    Search for more papers by this author
  • Toshiharu Shikanai

    Corresponding author
    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto 606-8502, Japan
    Search for more papers by this author

For correspondence (fax +81 75 753 4257; e-mail


Several hundred nucleus-encoded factors are required for regulating gene expression in plant organelles. Among them, the most numerous are the members of the pentatricopeptide repeat (PPR) protein family. We found that PPR protein OTP82 is essential for RNA editing of the ndhB-9 and ndhG-1 sites within transcripts encoding subunits of chloroplast NAD(P)H dehydrogenase. Despite the defects in RNA editing, otp82 did not show any phenotypes in NDH activity, stability or interaction with photosystem I, suggesting that the RNA editing events mediated by OTP82 are functionally silent even though they induce amino acid alterations. In agreement with this result, both sites are partially edited even in the wild type, implying the possibility that a single gene produces heterogeneous proteins that are functionally equivalent. Although only five nucleotides separate the ndhB-8 and ndhB-9 sites, the ndhB-8 site is normally edited in otp82 mutants, suggesting that both sites are recognized by different PPR proteins. OTP82 falls into the DYW subclass containing conserved C-terminal E and DYW motifs. As in CRR22 and CRR28, the DYW motif present in OTP82 is not essential for RNA editing in vivo.


Plastids were derived from cyanobacteria-like ancestors via endosymbiosis. Current plastid genomes, containing approximately 120 genes required for photosynthesis and housekeeping functions, were shaped after large-scale transfer of genes to the host genomes (Timmis et al., 2004). Accordingly, plant nuclear genomes encode many proteins with prokaryotic origin that retain the ancestral function in plastids (Barkan and Goldschmidt-Clermont, 2000). Eukaryotic phototrophs evolved mechanisms to coordinately regulate the expression of nuclear and plastid genomes. Several hundred nucleus-encoded proteins of eukaryotic origin are estimated to be involved in the regulation of plastid gene expression, based on the current data (Barkan et al., 2007; Schmitz-Linneweber and Small, 2008; Kroeger et al., 2009). However, the functions of many of them remain to be elucidated.

Among these proteins, much of the focus has concentrated recently on the pentatricopeptide repeat (PPR) protein family. PPR proteins form a huge protein family that is particularly prevalent in land plants: 450 members are encoded in the Arabidopsis thaliana genome (Lurin et al., 2004; O’Toole et al., 2008). The family members are defined by a tandem array of PPR motifs: each is a highly degenerate unit consisting of 35 amino acids, and is expected to be folded into a pair of antiparallel helices (Small and Peeters, 2000). Most PPR proteins are predicted to be localized in plastids or mitochondria (Lurin et al., 2004). A growing mass of information indicates that PPR proteins are involved in almost all stages of gene expression in both plastids and mitochondria (Schmitz-Linneweber and Small, 2008). The most probable explanation for their divergent roles is that they are sequence-specific RNA binding adaptors recruiting effector enzymes to the target RNA (Delannoy et al., 2007). The PPR protein family is divided into the P and PLS subfamilies (Lurin et al., 2004): the latter accounts for roughly half of the members in Arabidopsis, and is specific to land plants (O’Toole et al., 2008). Based on the differences in C-terminal motifs, the PLS subfamily is further classified into the PLS, E and DYW subclasses (O’Toole et al., 2008). The expansion of the PLS subfamily in plants is correlated with the specific characteristics of plant organelles, particularly RNA editing (Salone et al., 2007; O’Toole et al., 2008; Rüdinger et al., 2008).

RNA editing is a post-transcriptional process that alters specific cytidine residues to uridine in the mitochondrial and plastid RNA of higher plants (Shikanai, 2006). Thirty-four sites are edited in Arabidopsis plastids (Chateigner-Boutin and Small, 2007), whereas more than 450 editing sites are edited in Arabidopsis mitochondria (Giege and Brennicke, 1999; Bentolila et al., 2008; Zehrmann et al., 2008). In vivo approaches using plastid transformation and in vitro RNA editing assays clarified that a cis-element consisting of fewer than 30 nucleotides surrounding the editing site is essential for site recognition (Chaudhuri and Maliga, 1996; Hirose and Sugiura, 2001). The case is also similar in mitochondria (Takenaka et al., 2004). Our genetic and biochemical studies revealed that the PPR protein CHLORORESPIRATORY REDUCTION 4 (CRR4) acts as the site-specific factor for recognizing the site 1 of RNA editing (ndhD-1) in the plastid ndhD transcript encoding a subunit of NAD(P)H dehydrogenase (NDH) (Kotera et al., 2005; Okuda et al., 2006). Subsequently, two PPR proteins, CRR21 and CLB19, were identified as site-specific factors required for the RNA editing of ndhD-2 or RNA editing of rpoA and clpP, respectively (Okuda et al., 2007; Chateigner-Boutin et al., 2008). CRR4, CRR21 and CLB19 belong to the E subclass (Kotera et al., 2005; Okuda et al., 2007; Chateigner-Boutin et al., 2008). Subsequently, multiple reports indicated that DYW subclass members are also required for RNA editing. The DYW subclass member YS1 is involved in RNA editing of rpoB-1 (Zhou et al., 2008), whereas both CRR22 and CRR28 are involved in the RNA editing of multiple plastid transcripts (Okuda et al., 2009). Likewise, the DYW subclass members RARE1, LPA66 and ECB2 were found to be involved in RNA editing of plastid transcripts (Cai et al., 2009; Robbins et al., 2009;Yu et al., 2009). Similar DYW subclass proteins, MEF1 and OGR1, were also shown to be required for RNA editing in mitochondria of Arabidopsis and rice, respectively (Kim et al., 2009; Zehrmann et al., 2009). Although the involvement of PPR proteins in RNA editing of plant organelles is already beyond any doubt, questions still remain about the molecular mechanisms by which these proteins recognize editing sites, not to mention the unresolved mystery concerning the identity of the editing enzyme itself. Further examples of site-specific factors will be helpful for progressing our understanding of this process. In this context, the best-studied model system is the Arabidopsis plastid, where 34 sites are edited.

To identify further factors involved in plastid RNA editing, we are performing a comprehensive survey of E and DYW subclass members predicted to target to plastids. Here, we report the identification of a new member of the DYW subclass required for RNA editing of plastid ndh transcripts.


At1g08070 (OTP82) encodes a DYW member required for multiple RNA editing events in plastids

To clarify the genes involved in plastid RNA editing, we focused on the chlorophyll fluorescence changes after actinic light (AL) illumination, reflecting NDH activity in vivo (Shikanai et al., 1998), and identified the mutants defective in RNA editing of plastid ndh genes (Kotera et al., 2005; Okuda et al., 2007, 2009). However, the strategy cannot identify mutants impaired in RNA editing in which the defect does not severely influence NDH activity. To identify mutants defective in plastid RNA editing more systematically, we focused on E and DYW members that are predicted to target to plastids. Among them we selected a DYW member, At1g08070, that we named ORGANELLE TRANSCRIPT PROCESSING 82 (OTP82) after the molecular phenotype of its mutants (see below). OTP82 is disrupted by two independent Ds insertions in the lines otp82-1 and otp82-2 in the Nössen background (Kuromori et al., 2004; Ito et al., 2005) (Figure 1a). In addition, two T-DNA insertion lines, otp82-3 and otp82-4, in the Col-0 background were independently characterized (Figure 1a). The homozygous plants for each mutant allele did not show any obvious visible phenotypes under the standard culture conditions. The OTP82 gene does not contain introns, and encodes a putative DYW member of the PPR family consisting of 741 amino acids (Figure 1a). Putative orthologs of Arabidopsis OTP82 were found in several plants, including rice (Os01g08120), by the program of POGs/PlantRBP (plant RNA-binding Protein Database) ( OTP82 contains 14 PPR or PPR-like (P, L, L2 and S) motifs, as well as the E and DYW motifs that are characteristic of the DYW subclass (Figure 1a). Figure S1 shows the alignment of E and DYW motifs of OTP82 with those of PPR proteins involved in RNA editing, and with that of CRR2 involved in intergenic RNA cleavage (Hashimoto et al., 2003), indicating the high level of sequence similarity among the family members.

Figure 1.

 Editing defects in otp82 mutants.
(a) Predicted motif structure of OTP82. pentatricopeptide repeat (PPR) (or PPR-like), E and DYW motifs are depicted as boxes with letters. The designation of the P, L and S corresponds to the PPR motif, PPR-like S (for short) motif, and PPR-like L (for long) motif, respectively, proposed by Lurin et al. (2004). Sites of Ds and T-DNA insertions in mutant alleles are indicated.
(b) Poisoned primer extension assays were conducted on the editing sites ndhG-1 (118 858) and ndhB-9 (95 644). The editing sites are specified relative to the nucleotide sequence of the complete Arabidopsis chloroplast genome (Genbank accession number AP000423). RT-PCR products were obtained with primers surrounding the editing sites, and serve as templates for the extension reaction from a labeled 6-carboxyfluorescein primer that anneals next to the target editing site (a forward poisoned primer extension primer was used for all sites). The extension is stopped by the incorporation of ddGTP at the location of the editing site for unedited molecules, producing a short unedited product. The extension is stopped at the next G/C for the edited molecules, producing a longer edited product.
(c) RNA editing efficiency; pU, plasmid not edited; pE, plasmid edited.

To test the possibility that the otp82 disruptants are defective in RNA editing, we systematically examined the editing status of chloroplast transcripts using a new high-resolution melting screen (Chateigner-Boutin and Small, 2007). Among the 34 RNA editing sites present in Arabidopsis plastids (Chateigner-Boutin and Small, 2007), we identified defects in the RNA editing of ndhB-9 and ndhG-1 in otp82 (Figure S2) These defects were confirmed by more sensitive poisoned primer extension assays (Figure 1b). The sites at ndhB-9 and ndhG-1 were partially edited in the wild type at efficiencies estimated to be 71.1 and 67.2% in the Nössen background, respectively, and 78.9 and 86.3% in the Col-0 background, respectively (Figure 1c). RNA editing of these two sites was completely impaired in otp82 mutants, except that otp82-1 showed residual editing at the ndhB-9 and ndhG-1 sites (efficiency of 0.8 and 2.4%, respectively) (Figure 1c). We also confirmed that the introduction of the wild-type genomic sequence of OTP82 fully restored the editing of ndhB-9 and ndhG-1 (Figure 2), confirming that the defect in RNA editing results from the disruption of OTP82. Although only five nucleotides separate the ndhB-8 and ndhB-9 sites, the ndhB-8 site is normally edited in otp82 mutants (Figures 2 and S2), suggesting that these two adjacent RNA editing sites are recognized by different site-specificity factors. All other known sites were also edited correctly in otp82 mutants (Figure S2).

Figure 2.

 Effect of the deletion of the DYW motif in OTP82.
Nucleotide sequences including the RNA editing sites of ndhB-9 and ndhG-1 are shown as sequence chromatograms. Editing sites of ndhB-8, ndhB-9 and ndhG-1 are indicated by arrows pointing to the corresponding peaks; otp82-1 OTP82ΔDYW, otp82-1 transformed with OTP82 lacking the DYW motif; otp82-1 OTP82, otp82-1 complemented by the introduction of the wild-type genomic OTP82.

Defects in RNA editing may be secondarily caused by aberrant RNA processing. To test this possibility, the levels and patterns of transcripts were analyzed by RNA gel blot. Figure 3 shows that there are no obvious alterations in ndhB and ndhG transcripts in otp82. We conclude that otp82 is primarily defective in multiple RNA editing events in ndhB and ndhG transcripts.

Figure 3.

 Transcript profiles of genes with the editing defects in otp82.
A 10 μg portion of RNA from 15-day-old seedlings was loaded onto formaldehyde agarose gels and transferred onto a membrane. Hybridizations were performed under high-stringency conditions using antisense RNA probes for the genes specified for each blot. The arrows point at the mature transcript. The sizes of RNA markers are shown in kilobases. Ribosomal RNA stained with methylene blue on the membranes is shown as a loading control (membrane).

OTP82 is localized to plastids

OTP82 is predicted to be targeted to chloroplasts by TargetP (Emanuelsson et al., 2000) and Predotar (Small et al., 2004). To confirm the localization experimentally, we constructed a chimeric gene encoding a fusion protein consisting of the N-terminal 100 amino acids of OTP82 and green-fluorescent protein (GFP) under the control of the Cauliflower mosaic virus 35S promoter. As the cleavage site of the transit peptide was unknown, we used the N-terminal 100 amino acids including the first PPR motif to ensure that the complete targeting information was included. The plasmid containing the chimeric gene was introduced into wild-type Arabidopsis cells by bombardment. Analysis of GFP fluorescence in transformed cells revealed that the fluorescence co-localized with a chloroplast marker: the fusion of red-fluorescent protein (RFP) with the small subunit of Arabidopsis ribulose biphosphate carboxylase (SSU) (Carrie et al., 2007) (Figure 4). We conclude that OTP82 is localized to plastids, consistent with the otp82 phenotype specifically defective in plastid RNA editing.

Figure 4.

 Analysis of subcellular localization of the OTP82 protein.
A chimeric protein consisting of the first 100 amino acids of At1g08070 fused to GFP was targeted to chloroplasts in transformed Arabidopsis cells. Fluorescence of the green fluorescent protein (100 aa OTP82-GFP), red fluorescence of the red fluorescent protein in fusion with the targeting sequence of the small subunit of the ribulose biphosphate carboxylase (SSU-RFP), and the overlay of the two fluorescence images are shown.

RNA editing mediated by OTP82 is not required for NDH activity, stability and the supercomplex formation with PSI

The RNA editing of ndhB-9 and ndhG converts Ser279 of NdhB (uCa) to leucine (uUa) and Ser17 of NdhG (uCc) to phenylalanine (uUc), respectively (Figure 5d). The alterations between amino acids with different characters might be expected to have a strong impact on the encoded protein. However, both the ndhB-9 and ndhG-1 sites are only partially edited, even in the wild type (Figure 1c), implying that RNA editing might be dispensable for the protein function. First, we analyzed the level of NDH complex in otp82. As the NDH complex is unstable without the membrane subunits, NdhA, NdhB, NdhC, NdhD, NdhE, NdhF and NdhG, as reported for mutants defective in NdhB, NdhD or NdhF (Peng et al., 2008), antibodies against NdhH and NdhL can be used to monitor the accumulation of the NDH complex, and consequently that of NdhB and also probably that of NdhG. In otp82, neither NdhH nor NdhL levels were affected (Figure 5a). As editing at the ndhB-9 and ndhG-1 site was below the detection limit in otp82-2 (Figure 1c), all the NdhB and NdhG accumulating in the mutants was probably translated from unedited RNA. These results suggest that Leu279 in NdhB and Phe17 in NdhG are unlikely to be essential for stabilizing the NDH complex.

Figure 5.

 Analyses of the NAD(P)H dehydrogenase (NDH) complex in otp82.
(a) Immunodetection of NDH subunits NdhH and NdhL, and a subunit of the cytochrome (Cyt) b6f complex, Cytf. The proteins were extracted from thylakoid membrane fractions. Lanes were loaded with protein samples corresponding to 0.5 μg chlorophyll for Cytf, 1 μg chlorophyll for NdhL and 5.0 μg chlorophyll for NdhH (100%), and the series of dilutions indicated.
(b) BN-PAGE analysis of thylakoid protein complex isolated from the wild type and from otp82. Thylakoid membrane complexes isolated from the wild type, crr2-2, crr4-3 and otp82-3 were solubilized and separated by BN-PAGE, and then stained with Coomassie Briliant Blue, and are shown on the left panel. The top part of the gels is closed up on the right panel. Positions of band I are indicated by arrows.
(c) Monitoring of NDH activity by using chlorophyll fluorescence analysis. The curve shows a typical trace of chlorophyll fluorescence in the wild type (WT). Leaves were exposed to actinic light (AL) (50 μmol of photons m−2 s−1) for 5 min. AL was turned off and the subsequent change in chlorophyll fluorescence level was monitored. The transient increase in chlorophyll fluorescence is caused by the plastquinone reduction based on NDH activity. Insets are magnified traces from the boxed area. The fluorescence levels were normalized by the maximum fluorescence at closed PSII centers in the dark (Fm) levels; ML, measuring light; SP, a saturating pulse of white light.
(d) Partial sequence alignments of NdhB and NdhG around the amino acids affected by RNA editing. Arabidopsis NdhB and NdhG proteins were aligned with their homologs from other species. The alignment was performed using the ClustalW program (version 1.87, DNA Data Bank of Japan). Amino acids that are fully or semiconserved are shaded black and gray, respectively. Numbers indicate amino acid positions in the protein. The arrows above the sequences indicate the positions of edited codons.

Chloroplast NDH interacts with the photosystem I (PSI) complex to form a supercomplex (NDH-PSI) in Arabidopsis (Peng et al., 2008). It is possible that Leu279 of NdhB and Phe17 of NdhG are required for the supercomplex formation, although they are not essential for stabilizing NDH. To test this possibility, we analyzed the level of the NDH-PSI supercomplex in otp82 by Blue Native (BN)-PAGE. The high molecular weight green band (band I) corresponds to the NDH-PSI supercomplex in the wild type, and is missing in the crr4-3 mutant defective in ndhD expression, and is greatly reduced in the crr2-2 mutant defective in ndhB expression (Figure 5b), consistent with previous results (Peng et al., 2008). The BN-PAGE detected the same level of band I in otp82 as in the wild type (Figure 5b). The result indicates that Leu279 of NdhB and Phe17 of NdhG are not essential for the interaction between NDH and the PSI complex.

The mutants crr22 and crr28, both defective in the RNA editing of ndhB transcripts, showed altered NDH activity, despite the stable accumulation of the NDH complex (Okuda et al., 2009). It is possible that Leu279 of NdhB and Phe17 of NdhG are essential for NDH activity, rather than for its stability. To examine this possibility, we analyzed NDH activity in otp82. The chloroplast NDH complex catalyzes electron donation to plastoquinone from the stromal electron pool, and its activity can be monitored as a transient increase in chlorophyll fluorescence, reflecting plastoquinone reduction after turning off AL (Shikanai et al., 1998). Figure 5c shows a typical trace of the chlorophyll fluorescence level in the wild type. In otp82, the post-illumination increase of chlorophyll fluorescence was not suppressed (Figure 5c), indicating that NDH activity was not affected in otp82. These results suggest that Leu279 of NdhB and Phe17 of NdhG are not required for NDH activity. This result is consistent with the fact that the mutants were not identified in our chlorophyll fluorescence screen looking for mutants affected in NDH activity (Okuda et al., 2007).

Taken together, the Ser279 → Leu279 conversion of NdhB and Ser17 → Phe17 conversion of NdhG are unlikely to be essential for the activity or stability of the NDH complex, or even for NDH-PSI supercomplex formation, although the editing of ndhB-9 and ndhG-1 results in the restoration of codons for amino acids conserved in other land plants (Figure 5d) (Lutz and Maliga, 2001; Tsudzuki et al., 2001), except that ndhG-1 editing in pea converts serine to leucine (Inada et al., 2004). The results are consistent with the fact that the sites are only partially edited even in the wild type (Figure 1c).

Mass spectrometric analysis of the PSI-NDH supercomplex in the wild type and in otp82

The observations above suggest that the proteins translated from edited and unedited mRNA are functionally equivalent, and that both proteins are incorporated into the NDH complex. To examine this possibility, we analyzed the NDH-PSI supercomplex by mass spectrometry (MS). Consistent with the previous MS analysis (Peng et al., 2009), we failed to detect NdhG in both the wild type and the otp82 mutant, probably because of the hydrophobic nature of NdhG. In the wild type, the MS detected only NdhB derived from the ndhB transcript in which the ndhB-9 site was edited (Table 1). The ndhB-8 site is separated by only five nucleotides and the site was also edited, resulting in a Ser → Leu substitution in the detected peptide (Table 1). We also analyzed the NDH-PSI supercomplex in otp82-3, where the ndhB-9 site was not edited at all. The MS identified oligopeptides common to NdhB translated from both edited and unedited RNAs, but did not identify the peptide reflecting the RNA editing status of ndhB-9 (Table 1). The oligopeptide derived from unedited RNA (VAALASATR) may not be detected by MS for technical reasons, although the corresponding peptide translated from edited RNA (VAALALATR) was detected in the wild type. Thus, it is still possible that two versions of NdhB translated from edited and unedited mRNA are incorporated into the NDH complex.

Table 1.   Identified polypeptides from linear ion-trap triple quadrupole (LTQ)-Orbitrap mass analysis of the NDH-PSI supercomplex against NdhB amino acid sequences
  1. aSpecific peptide derived from NdhB that is translated from the ndhB transcript edited at sites 8 and 9.


The DYW motif of OTP82 is not essential for RNA editing in vivo

The DYW motif was proposed as the catalytic site of organelle RNA editing, based on the apparent similarity to the active site of cytidine deaminases, including the human RNA editing enzyme APOBEC1, and the phylogenetic correlation between the occurrence of RNA editing and the presence of DYW motifs (Salone et al., 2007; Rüdinger et al., 2008). However, the DYW motifs of CRR22 and CRR28 were shown to be dispensable for RNA editing in vivo, even though that of CRR2 was essential for RNA cleavage in vivo (Okuda et al., 2009). To examine whether this finding can be extended to other DYW motifs present in the PPR proteins required for RNA editing, a truncated version of OTP82 lacking its DYW motif was introduced into otp82. Consistent with our previous result, the mutant version of OTP82 restored the RNA editing activity of ndhB-9 and ndhG-1 to the wild-type level (Figure 2, otp82-1 OTP82ΔDYW). This result indicates that the DYW motif present in OTP82 is not essential for RNA editing in vivo.


The present study shows that OTP82 is a site-specific factor required for two RNA editing events in plastids. Like all other editing factors identified so far, OTP82 belongs to the PLS family of PPR proteins that is specific to land plants. Of 34 editing sites known in Arabidopsis plastids, three E and seven DYW subclass members required for 14 sites have been determined (Table S1). In total in Arabidopsis, 20 E and 24 DYW proteins are predicted to be targeted to plastids (Lurin et al., 2004). Even though some of these proteins (such as CRR2) may have other roles (Hashimoto et al., 2003; Nakamura and Sugita, 2008; Okuda et al., 2009), the number of E and DYW proteins appears more than sufficient to deal with all editing sites present in plastids, especially considering that a single PPR protein can recognize multiple sites (Chateigner-Boutin et al., 2008; Okuda et al., 2009; this work). This case appears to be similar in mitochondria, where the DYW members MEF1 and OGR1 were identified as editing factors (Zehrmann et al., 2009;Kim et al., 2009). The number of the PLS subfamily members predicted to localize to mitochondria correlates well with the number of editing sites in mitochondria (Zehrmann et al., 2009). The E and DYW domains are highly conserved in the PPR proteins involved in RNA editing, even between plastid and mitochondrial proteins (Figure S1). If these C-terminal motifs, especially the E motif, are the binding sites for the still-unidentified editing enzyme (Okuda et al., 2007), similar or identical editing enzymes are likely to be recruited in both organelles.

In vitro RNA editing assays have suggested that the cis-elements recognized by the same trans-factor show high sequence identity (a 60% identity in nucleotide sequence) (Kobayashi et al., 2007). Consistent with this idea, thirteen nucleotides are conserved within the 35 nucleotides surrounding the ndhB-9 and ndhG-1 sites that are recognized by OTP82 (Figure 6). The ndhB-8 editing site that is 6-bp upstream from the ndhB-9 site increases the number of identical nucleotides from 13 to 14 (Figure 6). Similarly, the nucleotides surrounding the two editing sites recognized by CRR28 are highly conserved (Okuda et al., 2009). However, it is not simple to interpret the recognition of multiple cis-elements by a single PPR protein. The sets of putative cis-elements recognized by CLB19, CRR22, MEF1 and OGR1 show no obvious similarity (Chateigner-Boutin et al., 2008; Kim et al., 2009; Okuda et al., 2009; Zehrmann et al., 2009). Different sets of PPR motifs within a single protein may independently recognize unrelated cis-sequences.

Figure 6.

 Comparison of the nucleotide sequences in the region surrounding the editing sites affected in otp82.
Sequence alignment of the region surrounding the ndhB-9 and ndhG-1 editing sites. The alignment includes the sequence from −30 to +10 around the edited C (asterisk), with identical nucleotides shown in boxes. The ndhB-8 site is indicated by the asterisk.

The ndhB-8 and ndhB-9 sites are separated by only five nucleotides in Arabidopsis (Figure 6). Are the closely adjacent sites recognized by distinct proteins, or by a common factor? In the tobacco ndhB gene a pair of editing sites, IV and V, are separated by only eight nucleotides. Both sites were independently edited in partially edited transcripts, implying the existence of two independent site-specificity factors (Bock et al., 1997). Both ndhB-8 and ndhB-9 sites are edited in Nicotiana tabaccum and Nicotiana sylvestris, whereas the ndhB-8 site is not edited in Nicotiana tomentosiformis, implying that distinct site-specificity factors are involved in the two editing events (Sasaki et al., 2003). Consistent with this idea, otp82 lacks only the editing activity of ndhB-9, and the ndhB-8 site is edited as in the wild type (Figure 2). Our result provides evidence that the ndhB-8 and ndhB-9 sites are recognized by distinct proteins, despite their close proximity.

otp82 is defective in RNA editing of two sites, ndhB-9 and ndhG-1, leading to amino acid alterations. However, the amino acid alterations induced by these two RNA editing events are functionally silent, at least under the culture conditions used in this study. Even leaky defects in NDH activity cause a severe growth phenotype in the pgr5 mutant background impaired in the main pathway of PSI cyclic electron transport (Munekage et al., 2004; Peng et al., 2009). To test the possibility that NDH activity is subtly affected in otp82, we generated the double mutant otp82-2 pgr5. The phenotype of otp82-2 pgr5 was identical to that of pgr5, supporting our idea that NDH activity was not affected in otp82 (data not shown).

The ndhB-9 and ndhG-1 sites are not edited completely, even in wild-type plants (Figure 1c), suggesting the possibility that proteins translated from edited and unedited mRNA are incorporated into the NDH complex. However, our MS analysis preferentially detected NdhB that originated from edited RNA in the wild type (Table 1). We cannot conclude that the NDH-PSI supercomplex solely consists of NdhB originating from edited mRNA, despite the high mass accuracy, high resolution and high sensitivity of linear ion-trap triple quadrupole (LTQ)-Orbitrap MS.

RNA editing at ndhB-9 and ndhG-1 may be physiologically significant under certain conditions, because the amino acids restored by editing are highly conserved in plants, presumably by natural selection. We reported that CRR22 is involved in RNA editing of rpoB-3, and that the resulting amino acid alteration was also functionally silent (Okuda et al., 2009). At this site a trace level of unedited transcript was also detected even in the wild type (Okuda et al., 2009). We also observed that editing of the ndhD-1 site that creates the translational initiation site of ndhD is only partial (Kotera et al., 2005), and that the efficiency of this RNA editing event can be determined by CRR4 activity (Okuda et al., 2008). Partial editing seems to be restricted to cases where the editing event is not essential for protein expression or activity. As yet we cannot conclude whether this is simply a result of lower selection pressure for efficient editing, or whether there is active selection for partial editing to generate genetic diversity.

Experimental procedures

Plant materials

Arabidopsis thaliana ecotypes Nössen and Col-0 were used in this study. otp82-1 (11-6902-1, insertion site +465 bp, Nössen) and otp82-2 (54-2824-1, insertion site +793 bp, Nössen) carry Ds transposon insertions (Kuromori et al., 2004; Ito et al., 2005). otp82-3 (SALK_027812, Col-0) and otp82-4 (SAIL_851_G04, Col-0) are T-DNA insertion mutants obtained from the ABRC Stock Center (

Chlorophyll fluorescence analysis

Chlorophyll fluorescence was measured using a MINI-PAM portable chlorophyll fluorometer (Waltz, The transient increase in chlorophyll fluorescence after turning off AL was monitored as previously described (Shikanai et al., 1998).

Construction of GFP fusion proteins to analyze targeting

The first 300 bp of the coding sequence of At1g08070 was amplified using Phusion DNA polymerase (Finnzymes, with primers z0797 and z0798 containing the attB sites for Gateway® cloning according to the manufacturer’s instructions (Invitrogen, The GFP vector used was the same as that previously described (Carrie et al., 2009). The chloroplast targeting marker consists of RFP fused to SSU (Carrie et al., 2009). The primers are listed in Table S2.

Subcellular localization

Biolistic transformations of GFP and RFP constructs were performed on Arabidopsis cell culture, as previously reported (Carrie et al., 2007). The GFP construct and the chloroplast RFP marker (5 μg each) were co-precipitated onto gold particles and transformed using the biolistic PDS-1000/He system (Bio-Rad, Particles were bombarded onto 2 ml of Arabidopsis cell suspension resting on filter paper on osmoticum plates. After bombardment, the cells were placed in the dark at 22°C for 24 h. Observation of transient GFP and RFP expression was performed using an Olympus BX61 fluorescence microscope (Olympus, with excitation wavelengths of 460/480 nm (GFP) and 535/555 nm (RFP), and with emission wavelengths of 495–540 nm (GFP) and 570–625 nm (RFP). Subsequent images were captured using Cell® imaging software, as previously described (Carrie et al., 2007; Murcha et al., 2007).

Analysis of RNA editing

A high-resolution melting analysis of amplicons was performed as previously described (Chateigner-Boutin and Small, 2007), except that the primers used for the PCR were designed to give shorter amplicons than in the previous study. Poisoned primer extension of reverse transcription (RT)-PCR products was performed as previously described (Chateigner-Boutin and Small, 2007). RNA editing was also analyzed by directly sequencing the RT-PCR products, including editing sites, as previously described (Okuda et al., 2009). The primers are listed in Table S2.

Plant transformation

For complementation of the otp82 mutation, the wild-type genomic sequence surrounded by 5′-GTACAAGATCGGAAGAGC-3′ and 5′-CTACCAGTAGTCATTGCAG-3′ was cloned in pGEW-NB1 vector. For the expression of OTP82 truncated in the DYW motif, the wild-type genomic sequence surrounded by 5′-GTACAAGATCGGAAGAGC-3′ and 5′-CTACTCTAGTAACACCTCCATT-3′ was cloned in the pGEW-NB1 vector. The resultant plasmids were introduced into otp82-1 via Agrobacterim tumefaciens strain ASE.

RNA preparation and RNA gel blot analysis

Total RNA from leaves of 15-day-old plantlets was isolated using TRIzol reagent (Invitrogen), as recommended in the manufacturer’s instructions. Fifteen micrograms of RNA was fractionated on 1.2% (w/v) formaldehyde agarose gels, and then transferred onto Hybond N+ nylon membranes (GE Healthcare, RNA integrity, loading and transfer were checked by staining the membrane with methylene blue. RNA probes were internally labeled with biotinylated cytidine by transcription of PCR products cloned in pGEM-T Easy vector (Promega, The primers used for the PCR were ndhB.AT.rev2 and AndhB, corresponding to the second exon of ndhB, and ndhG.AT.for and ndhG.AT.rev for ndhG. Clones with inserts in antisense orientation were amplified by PCR using the forward primer and M13/pUC reverse primer. The PCR products served as a template for in vitro transcription with SP6 polymerase, following the manufacturer’s instructions (Maxiscript; Ambion, Prehybridization of the membrane was carried out for 1 h in hybridization buffer containing 5 x SSC, 50% (v/v) formamide, 0.5% SDS and 100 μg ml−1 heparin, at 68°C. Hybridization with RNA probes was carried out in the same buffer overnight at 68°C, followed by three 15-min washes at 25°C in 1 x SSC/0.5% SDS, and two washes at 60°C in 0.1 x SSC/0.1% SDS for, respectively, 20 min and 1 h. Signal detection was performed using the Chemiluminescent Nucleic Acid Detection Module (Pierce,, read in an ImageQuant-RT ECL (GE Healthcare). The primers are listed in Table S2.

Immunoblot analysis

Chloroplasts were isolated from the leaves of 4-week-old plants as previously described (Okuda et al., 2007). Samples were normalized by measuring chlorophyll concentration. The protein samples were separated by 12.5% SDS–PAGE. After electrophoresis, the proteins were transferred onto a Hybond-P membrane (GE Healthcare) and incubated with specific antibodies. The signals were detected using an ECL Advance Western Blotting Detection Kit (for NdhH) (GE Healthecare) or an ECL Plus Western Blotting Detection Kit (for the others) (GE Healthcare), and were visualized by an LAS1000 chemiluminescence analyzer (Fuji Film,

Thylakoid membrane preparation and BN-PAGE

Chloroplasts were isolated as previously described (Okuda et al., 2007) and osmotically ruptured in buffer containing 20 mm HEPES/KOH (pH 7.6), 5 mm MgCl2 and 2.5 mm EDTA. Thylakoid membranes were pelleted by centrifugation (7700 g for 3 min) and resuspended in the same buffer.

BN-PAGE was performed as previously described (Peng et al., 2006) with some minor modifications. The freshly isolated thylakoid membranes were gently washed twice with buffer containing 25 mm BisTris–HCl (pH 7.0), 20% glycerol, and solubilized in 25 mm BisTris–HCl (pH 7.0), 20% glycerol, 1%N-dodecyl-β-d-maltoside, at final chlorophyll concentration of 1 mg ml−1. After incubation on ice for 10 min, the supernatants were supplemented with a 1/10 volume of BN sample buffer [100 mm BisTris–HCl, pH 7.0, 5% Serva blue G, 0.5 m 6-amino-N-caproic acid, 30% sucrose (w/v)]. Thylakoid protein complexes were separated by 5–12% gradient BN-PAGE in 0.75-mm-thick gels connected to a circulating cooler.

Peptide preparation for MS/MS analysis

Thylakoid membrane complexes isolated from wild-type and otp82 mutant plants were solubilized and separated by BN-PAGE. Band I (described in Peng et al., 2008) was excised from the gel. The excised band was treated twice with 25 mm ammonium bicarbonate in 30% (v/v) acetonitrile for 10 min and 100% (v/v) acetonitrile for 15 min, and then dried in a vacuum concentrator. The dried gel pieces were treated with 0.01 mg ml−1 trypsin (sequence grade; Promega)/50 mm ammonium bicarbonate, and then incubated at 37°C for 16 h. The digested peptides in the gel pieces were recovered twice with 20 μl 5% (v/v) formic acid/50% (v/v) acetonitrile. The extracted peptides were combined and then dried in a vacuum concentrator.

Mass spectrometric analysis and database searching

LC-MS/MS analyses were performed on an LTQ-Orbitrap XL-HTC-PAL system. Trypsin-digested peptides were loaded on the column (ø75 μm, 15 cm; L-Column; CERI, by using a Paradigm MS4 HPLC pump (Michrom BioResources, and an HTC-PAL autosampler (CTC Analytics), and were eluted by a gradient of 5–45% (v/v) acetonitrile in 0.1% (v/v) formic acid over 70 min. The eluted peptides were introduced directly into the LTQ-Orbitrap XL MS at a flow rate of 300 nl min−1, and with a spray voltage of 2.0 kV. The range of MS scan was m/z 450–1500, and the top three peaks were analyzed by MS/MS analysis. MS/MS spectra were compared by mascot 2.2 against NdhB amino acid sequences (Figure S3) with the following search parameters: set-off threshold at 0.05 in the ion score cut-off; peptide tolerance, 10 ppm; MS/MS tolerance, ±0.8 Da; peptide charge, 2+ or 3+; trypsin as enzyme allowing up to one missed cleavage; carboxymethylation on cysteines as a fixed modification and oxidation on methionine as a variable modification.


We thank Asako Tahara (Kyoto University, Kyoto, Japan) for skilled technical support. We also thank Tsuyoshi Endo (Kyoto University, Kyoto, Japan) for giving us antibodies. We are also grateful to Tsuyoshi Nakagawa (Shimane University, Matsue, Japan) for giving us the binary vector, pGWB-NB1. This work was supported by: Grant-in-Aid 16085206 for Scientific Research on Priority Areas; Grant 17GS0316 for Creative Science Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan; a grant from the Ministry of Agriculture, Forestry and Fisheries of Japan (Genomics for Agricultural Innovation, GPN0008); a grant from the Australian Research Council (CE0561495); an International Science Linkages grant (CG120098) from the Australian Government Department of Innovation, Industry, Science and Research; and scholarships from the Region of Alsace and the University of Western Australia. KH is the holder of a Lavoisier Fellowship and IS is a West Australian State Premier’s Fellow.