The pentatricopeptide repeat (PPR) is a degenerate 35-amino acid repeating motif that is found in animal, fungal, and plant proteins. The PPR protein family is particularly large in plants, where the majority of family members are predicted to be targeted to mitochondria or chloroplasts. PPR proteins are believed to fall into the larger family of helical repeat proteins, which typically bind macromolecules through a surface formed by the stacking of consecutive helical repeating units. Prior findings implicate several PPR proteins in organellar RNA metabolism, but the biological functions of few PPR proteins have been explored and in no case has a direct substrate been definitively identified. We present a characterization of the maize nuclear gene ppr2, which encodes a chloroplast PPR protein. PPR2 is found in large, heterogeneous protein complexes in the chloroplast stroma, some of which may be associated with RNA. Null ppr2 mutants have albino leaves and lack plastid rRNA and translation products. Plastid rRNAs are absent in both dark- and light-grown leaf tissues, indicating that their absence does not result from photo-oxidative damage. The population of plastid transcripts in ppr2 mutants is similar to that in other maize mutants lacking plastid ribosomes, and no ppr2-specific defects in plastid RNA metabolism have been detected. Taken together, the results suggest that ppr2 functions in the synthesis or assembly of one or more component of the plastid translation machinery.
Protein domains that mediate macromolecular interactions are key determinants of the networks through which distinct biochemical activities are linked in the cell. The recently described pentatricopeptide repeat (PPR) motif (Small and Peeters, 2000) has been proposed to mediate macromolecular interactions because of its resemblance to the tetratricopeptide repeat (TPR), a well-characterized protein interaction motif. The TPR motif consists of degenerate 34-amino acid repeats, often in tandem arrays; each repeating unit forms a pair of antiparallel alpha helices, with the tandem repeats packing to form a broad, concave substrate-binding surface (reviewed by Blatch and Lassle, 1999). The PPR motif is a degenerate 35-amino acid motif that is typically found in tandem arrays of 5–15 copies; a similar pattern of conserved structure-determining residues suggests that PPR arrays adopt a structure that is similar to that of TPR arrays (http://www.evry.inra.fr/public/projects/ppr/ppr.html). TPRs and, presumably, PPRs fall into the larger class of helical hairpin repeats, which also includes the HEAT, the ARM (Huber et al., 1997), and PUF repeats (Edwards et al., 2001). All characterized members of this family mediate macromolecular interactions – usually protein–protein interactions – through structurally similar binding surfaces.
These observations suggest that PPR domains can function as nucleic-acid-binding domains, and most typically as RNA-binding domains. This possibility was initially suggested by Small and Peeters (2000), based on TPR-templated structural models of PPR domains, which revealed a positively charged central groove that could accommodate a single-stranded nucleic acid. However, progress in understanding the biological and biochemical functions of PPR proteins has been hampered by the absence of examples in which an in vitro RNA binding activity has been firmly linked with a genetically defined RNA substrate. To gain insight into the functions of PPR domains, we are seeking to identify a set of PPR genes that are amenable to both genetic and biochemical analyses. Previously, we found that the maize PPR protein CRP1 influences the metabolism and translation of two chloroplast mRNAs (Barkan et al., 1994; Fisk et al., 1999). Although the crp1 mutant phenotype pointed to potential RNA targets of CRP1, biochemical tests of those interactions were hampered by the intractability of recombinant CRP1. This study presents an initial biochemical and genetic analysis of a previously uncharacterized PPR gene in maize, designated ppr2. PPR2 is found in large macromolecular complexes in the chloroplast stroma. Mutant ppr2 alleles were identified through a reverse-genetic screen; their phenotype shows ppr2 to be required for the accumulation of plastid ribosomes. Based on the prior observations implicating PPR proteins in RNA metabolism, we hypothesized that PPR2 functions in the metabolism of translation-related chloroplast RNAs (i.e. rRNA, tRNA, or ribosomal protein mRNA). However, despite an exhaustive search, we were unable to identify defects in the metabolism of these RNAs in ppr2 mutants, aside from pleiotropic defects observed in other mutants lacking plastid ribosomes. We observed that the ppr2 transcript accumulates to increased levels in several maize mutants that are blocked at early stages of chloroplast biogenesis, suggesting that ppr2 expression is negatively regulated in response to signals from developing chloroplasts.
The maize ppr2 gene encodes a PPR protein that resides in the chloroplast stroma
Predicted plastid-localized PPR proteins in Arabidopsis were used to identify potential maize orthologs by querying sequence databases. Despite the abundance of ppr genes in plant genomes, putative orthologs stand out as having particularly high levels of sequence identity both within and outside their PPR motifs. A maize expressed sequence tag (EST; TC69373) predicted to be orthologous to Arabidopsis PPR gene At3g06430 was used as a starting point to construct a full-length cDNA via PCR amplification of 5′ and 3′ sequences out of a maize cDNA library. A corresponding genomic clone was also isolated, and its nucleotide sequence was determined. The gene, designated ppr2, is diagrammed in Figure 1(a).
Northern hybridizations (Figure 1b) and cDNA sequences revealed two mRNA forms that differ in their polyadenylation site but that have identical open-reading frames (Figure 1a, transcripts ‘a’ and ‘b’). EST sequences derived from tassel primordia cDNA libraries revealed a third mRNA form in which an additional intron is spliced out (Figure 1a, transcript ‘c’). This mRNA is predicted to encode a truncated protein product. However, we have been unable to detect this transcript by RT-PCR of RNA purified from seedling leaf, root tip, endosperm, immature ear, immature tassel, or tassel primordia (data not shown). The ppr2 mRNAs accumulate to fairly similar levels in a variety of tissues (tassle, ear, stem, green leaf, and etiolated leaf; Figure 1c).
An alignment between PPR2, its putative Arabidopsis ortholog (At3g06430), and a second closely related Arabidopsis PPR protein (At5g48730) is shown in Figure 2(a). PPR2 contains 11 tandem PPR repeats; these are aligned with one another and compared with the PPR consensus in Figure 2(b). The targetp (Emanuelsson et al., 2000) and predotar (http://www.inra.fr/predotar/) algorithms predict that PPR2 and its putative Arabidopsis ortholog are targeted to the chloroplast. To test whether PPR2 is a chloroplast protein, antibody raised to recombinant PPR2 was used to probe immunoblots of different subcellular fractions. The molecular weight predicted for PPR2 after transit peptide removal is 58 kDa. The antibody detected a protein of approximately 60 kDa in leaf and chloroplast extract (Figure 3a,b). This protein is absent in ppr2 mutant leaf extract but is present in iojap mutant leaves, which resemble ppr2 mutants in having ivory leaves and a severe plastid ribosome deficiency (ppr2 mutants described below). These results provide strong evidence that the protein detected by the anti-PPR2 antibody is PPR2.
As anticipated by the targeting prediction algorithms, the PPR2 protein is enriched in isolated chloroplasts, with respect to its concentration in total leaf extract (Figure 3a,b). When isolated chloroplasts were themselves fractionated (Figure 3c), PPR2 co-fractionated with Cpn60, a stromal marker, and did not co-fractionate with proteins associated with the stromal face of the thylakoid membrane (ATPβ), the thylakoid lumen (OE33), or the envelope membranes (IM35). Therefore, PPR2 is localized to the chloroplast stroma.
To determine whether PPR2 is bound in a stable fashion to other macromolecules in vivo, stromal extracts were fractionated by sedimentation through sucrose gradients (Figure 4). The position of PPR2 in the gradients was determined by immunoblot analysis of gradient fractions. PPR2 was broadly distributed, but was found primarily in gradient fractions containing particles in the 300–600-kDa range. A small fraction of the PPR2 co-sedimented with ribosomes. When the extract was treated with RNase A prior to sedimentation, a PPR2 peak at approximately 550 kDa became more distinct and material in the ribosome region of the gradient may have diminished (Figure 4). These results indicate that PPR2 is found in heterogeneous protein complexes in vivo and that a small fraction may be associated with RNA and/or with ribosomes.
Isolation of loss-of-function alleles of ppr2
Mutant ppr2 alleles were identified in a PCR-based reverse-genetic screen of our collection of transposon-induced non-photosynthetic maize mutants (http://chloroplast.uoregon.edu/). Two mutant alleles were identified (ppr2-1 and ppr2-4), both with a Mu insertion disrupting the open-reading frame early in the PPR-encoding region (Figures 1, 2, and 5). The insertion in ppr2-1 is likely to be the Mu1 member of the Mu transposon family (Lisch, 2002) based on the sequence of its terminal inverted repeats, and on its size (approximately 1.3 kbp) as deduced from genomic Southern blot analysis (Figure 5b). The insertion in ppr2-4 is likely to be the Mu2 member of the family, based upon its size (approximately 2 kbp, Figure 5b), and the sequence of its terminal inverted repeats.
Homozygous ppr2-1 and ppr2-4 mutant seedlings are albino, with an ‘ivory’ leaf phenotype (Figure 5c). Crosses between ppr2-1/+ and ppr2-4/+ plants yielded approximately 25% albino seedlings, indicating that the ivory leaf phenotype results from the mutations in ppr2. Ivory leaves are typical of maize mutants that lack chloroplast ribosomes; this pigmentation differs subtly from the ‘paper white’ phenotype resulting from mutations that disrupt carotenoid synthesis, in which albinism results from the photo-oxidation of chlorophyll (Han et al., 1995). Ivory pigmentation can result from treatment with antibiotics that inhibit plastid translation (Zubko and Day, 2002), from mutations that disrupt the synthesis of any component of the plastid translation machinery, or from mutations that block chloroplast development at an early stage. Ivory pigmentation is accompanied by characteristic changes in plastid RNA metabolism (Han et al., 1993; Hess et al., 1993; Silhavy and Maliga, 1998; Zubko and Day, 2002; our unpublished data), which presumably result from the combined effects of the absence of the plastid-encoded RNA polymerase, the absence of plastid ribosomes, and the changes in nuclear gene expression that result from these plastid defects. Two well-characterized ivory maize mutants are crs2-1 (Jenkins et al., 1997) and iojap (Walbot and Coe, 1979): crs2-1 mutants lack chloroplast ribosomes because of the failure to splice several chloroplast tRNAs and ribosomal protein mRNAs; iojap mutants also lack chloroplast ribosomes although the basis for this loss is unknown.
RNA gel blot hybridization showed that ppr2 mRNA accumulation is reduced in both mutant ppr2 alleles (Figure 5d). Strikingly, ppr2 mRNA accumulates to increased levels in two other ivory mutants (iojap and w1), suggesting that the ppr2 gene might be negatively regulated in response to a signal from developing chloroplasts. However, this increase was not observed in ivory crs2-1 mutants, indicating that this regulation is complex. A microarray study involving the Arabidopsis prpl11 mutant, which has a chloroplast translation defect (albeit a more subtle one than that in the ppr2 mutants, as described below), revealed increased mRNA accumulation from a variety of nuclear genes encoding chloroplast proteins (Kurth et al., 2002; Richly et al., 2003).
Absence of plastid ribosomes in ppr2 mutants
Pentatricopeptide repeat proteins with known functions influence post-transcriptional aspects of gene expression. To address whether PPR2 might function in chloroplast gene expression, we characterized chloroplast gene expression in the ppr2 mutants. The photosynthetic enzyme complexes Rubisco, ATP synthase, photosystem II, photosystem I, and the cytochrome b6f complex each includes chloroplast-encoded subunits; a defect in the synthesis of any single subunit results in reduced accumulation of other closely associated subunits (Barkan, 1998; Barkan et al., 1995). Immunoblot analysis showed that representative subunits of these complexes did not accumulate to detectable levels in the ppr2 mutants (Figure 6). Similar results were obtained with the ivory maize mutants w1 and iojap (Figure 6), which have severe plastid ribosome deficiencies. These results suggested that ppr2 mutants have a tight and global defect in plastid gene expression.
To determine whether these protein deficiencies result from a defect in the metabolism of chloroplast RNAs, transcripts from a wide variety of plastid genes were analyzed by RNA gel blot hybridization (Figure 7 and data not shown). Emphasis was placed on analyzing RNAs involved in translation, as the failure to synthesize any essential component of the translation machinery (e.g. a tRNA, rRNA, or ribosomal protein mRNA) is anticipated to result in the ppr2 mutant protein and pigmentation phenotype.
The abundance of plastid rRNAs reflects plastid ribosome abundance as rRNA does not detectably accumulate out of the context of the ribosome in normal chloroplasts (unpublished observations). Plastid rRNA content in the ppr2 mutants was compared with that of iojap and crs2-1 mutants, both of which lack plastid ribosomes (Han et al., 1993; Jenkins et al., 1997; Walbot and Coe, 1979). Mature 16S rRNA, 5S rRNA, 4.5S rRNA, and the 5′ fragment of cleaved 23S rRNA were not detected in the ppr2 mutants, while only trace amounts of the 3′ fragment of cleaved 23S rRNA were detected (Figure 7a). This indicates that ribosomes are virtually absent in ppr2 mutant plastids. The magnitude of this rRNA deficiency was similar in dark-grown ppr2 mutant leaf tissue (Figure 7b), showing that the ribosome loss in ppr2 mutants is not the result of photo-oxidative damage.
The absence of plastid ribosomes is associated with numerous pleiotropic changes in plastid RNA metabolism (Han et al., 1993, 1995; Jenkins et al., 1997). Therefore, plastid mRNAs and tRNAs from ppr2 mutants were compared to those from crs2-1 and iojap mutants. Transcript patterns unique to the ppr2 mutants would be informative with regard to ppr2 function, whereas transcript defects observed in all three mutants would reflect their early block in chloroplast biogenesis or plastid ribosome deficiency. All plastid ribosomal protein mRNAs were examined, but no differences were detected in transcript patterns between the ppr2 and control ivory mutants (Figure 7c and data not shown). Twelve tRNAs were examined, but the mature tRNAs were undetectable, or nearly so, in ppr2 and in the control ivory mutants (data not shown). Numerous mRNAs encoding components of the photosynthetic apparatus were also examined (see legend to Figure 7), and again, the RNA patterns in the ppr2 mutants were identical to those in the control ivory mutants (data not shown). Thus, all of the RNA metabolism defects detected in the ppr2 mutants can be accounted for as pleiotropic effects of their plastid ribosome deficiency or their early block in chloroplast biogenesis. Although we failed to detect any defects in RNA metabolism that were unique to ppr2 mutants, we cannot eliminate the possibility that such defects exist.
We present a genetic and biochemical characterization of a PPR protein in maize, designated PPR2. PPR2 contains 11 tandem PPR repeats, preceded at its N-terminus by a chloroplast transit peptide and approximately 100 amino acids lacking homology to known proteins or motifs. We have shown that PPR2 is localized to the chloroplast stroma, where it is found in heterogeneous complexes that are largely resistant to ribonuclease treatment. These complexes are larger and more heterogeneous than those described previously for PPR proteins (CRP1: approximately 300 kDa, Fisk et al., 1999; HCF152: approximately 180 kDa, Meierhoff et al., 2003; TBC2: approximately 400 kDa, Auchincloss et al., 2002); however, the PPR2 complexes resemble previously characterized PPR complexes in that they are largely resistant to RNase treatment.
PPR2 plays an essential role early in chloroplast biogenesis, its absence preventing the accumulation of plastid ribosomes in both light- and dark-grown leaf tissue. The plastid ribosome deficiency in ppr2 mutants is much more severe than that in tobacco plastome mutants in which plastid rpo genes have been deleted (Allison et al., 1996; Santis-Maciossek et al., 1998), so the ppr2 mutant phenotype is unlikely to result solely from a defect in the expression of the plastid RNA polymerase. Furthermore, ppr2 seems unlikely to function in chloroplast RNA editing: the editing of 13 sites was examined in the ppr2 mutants (including all known edited sites in mRNAs that function in plastid gene expression), but no editing defects were detected (N. Peeters and M. Hanson, personal communication). The results are consistent with the possibility that PPR2 functions in the translation or processing of plastid ribosomal protein mRNAs, the processing of plastid tRNAs, or the assembly or activity of plastid ribosomes.
The biochemical basis for the ribosome deficiency in ppr2 mutant plastids could not be discerned: despite an exhaustive analysis, no defects in chloroplast RNA metabolism were detected in ppr2 mutants that differed from the characteristic defects in other mutants that lack plastid ribosomes. The absence of plastid ribosomes in iojap and other ivory maize mutants, barley albostrians, and in streptomycin-treated rice and maize seedlings are all accompanied by similar patterns of aberrant chloroplast transcript accumulation (Han et al., 1993; Hess et al., 1993; Silhavy and Maliga, 1998; Zubko and Day, 2002): some plastid mRNAs encoding components of the gene expression machinery accumulate to increased levels; some transcripts are reduced in abundance; and characteristic aberrant transcript forms are found. These changes presumably result from the combined effects of the absence of the plastid-encoded RNA polymerase, the absence of plastid ribosomes, and the changes in nuclear gene expression that result from these plastid defects. In one maize mutant lacking plastid ribosomes, crs2-1, a plastid splicing defect was superimposed on the characteristic RNA defects associated with plastid ribosome loss (Jenkins and Barkan, 2001; Jenkins et al., 1997; see Figure 7c, for an example); this lead to the conclusion that CRS2 functions in chloroplast splicing. However, the population of plastid transcripts in ppr2 mutants was indistinguishable from that in iojap mutants and other mutants lacking plastid ribosomes, and so these RNA analyses did not provide clues about the specific biochemical role of PPR2.
Despite their apparent lack of plastid ribosomes, ppr2 mutant seedlings are normal in morphology and grow at near normal rates for approximately 2 weeks in soil, until seed reserves are depleted. This adds to the evidence that the loss of plastid ribosomes, whether induced by mutation or by antibiotics, does not impact leaf cell viability or morphology in the grasses (Han et al., 1993, 1995; Hess et al., 1993; Jenkins et al., 1997; Walbot and Coe, 1979; Zubko and Day, 2002). In contrast, there are several examples in dicots in which early blocks in chloroplast biogenesis are accompanied by defects in palisade cell morphogenesis (Chatterjee et al., 1996; Keddie et al., 1996; Reiter et al., 1994; Wang et al., 2000). These observations suggest the possibility that dicot plastid genomes provide a cellular house-keeping function that is provided by a nuclear gene(s) in the grasses.
To explore whether PPR2 might interact with RNA, we modeled its putative substrate binding surface (Figure 8). Each PPR repeating unit has been proposed to consist of a pair of alpha helices, helix A and helix B, separated by a sharp turn; consecutive helical hairpins are thought to stack upon one another to form a superhelix with a concave substrate binding surface (Blatch and Lassle, 1999; Small and Peeters, 2000). The ‘A’ helices are expected to contribute to the putative substrate binding surface, whereas the ‘B’ helices are predicted to play a structural role. In PPR2 and CRP1, the predicted ‘A’ helices are more highly conserved with their putative Arabidopsis orthologs than are the ‘B’ helices, consistent with the idea that the ‘A’ helices bind substrate (data not shown).
Figure 8 shows the amino acid residues on the predicted solvent-exposed surface formed by the stacked ‘A’ helices in PPR2, corresponding to the putative substrate binding surface. Shown for comparison are the predicted Arabidopsis PPR2 ortholog, and CRP1, a chloroplast-localized PPR protein in maize that impacts the metabolism of the petA, petB, and petD mRNAs (Barkan et al., 1994; Fisk et al., 1999). The residues predicted to lie on this surface are highly conserved between maize and Arabidopsis PPR2 proteins, whereas they are quite divergent in CRP1; this is consistent with the idea that PPR2 and CRP1 bind different substrates. However, certain features are similar between the three proteins, suggesting similar modes of substrate recognition. In all three proteins, the ‘upper’ edge of this surface (corresponding to residue 12 in each repeating unit) is highly charged. In all three proteins, the ‘lower’ edge of the surface (corresponding to residue 4 in each repeating unit) is composed almost exclusively of uncharged, polar amino acids. Particularly striking is the abundance of asparagine residues on the ‘lower’ edge; indeed, an asparagine at position 4 forms part of the PPR consensus (Small and Peeters, 2000; see Figure 2b). Interestingly, asparagine is highly represented at the RNA/protein interface in structurally characterized RNA/protein complexes, whereas it is less characteristic of protein–protein interaction surfaces (Treger and Westhof, 2001). Among the remaining amino acids, arginine, serine, and lysine are the most frequently found at RNA/protein interfaces (Treger and Westhof, 2001); these are also highly represented on the putative substrate binding surfaces of PPR2 and CRP1 (Figure 8). An appealing possibility is that the arginines and lysines along the upper edge interact with phosphate moieties in the RNA backbone, presenting the nucleotide bases for hydrogen bonding with the asparagines, serines, and threonines along the lower edge.
The recent structural characterization of a Puf domain (Edwards et al., 2001) provides precedent for the idea that the concave surface presented by a helical repeat protein can be adapted for RNA binding. The body of data concerning PPR proteins raises the possibility that these may bind RNA in a similar manner. Further genetic, biochemical, and structural studies of PPR proteins will be required to deduce the biological and biochemical roles of this large and diverse protein family.
Genetic stocks and plant growth
The mutant ppr2 alleles were identified in a reverse-genetic screen involving our collection of approximately 2000 Mu-induced non-photosynthetic maize mutants (see http://chloroplast.uoregon.edu/). Pooled DNA samples (20 mutants each) were screened by PCR, using a ppr2-specific primer (5′-TTGTACATGGCGTCGGTGAGGT) in conjunction with a Mu TIR primer (5′-AGAGAAGCCAACGCCAWCGCCTCYATTTCGTC). The reactions (50 µl) contained 50 mm KCl, 10 mm Tris–HCl, pH 9.0, 0.1% Triton X-100, 1.5 mm MgCl2, 2.5 mm dNTPs, 10% DMSO, Taq polymerase, 50 ng genomic DNA, and 50 µm of each primer. Amplification conditions were 94°C for 4 min, followed by 35 cycles of 94°C for 45 sec, 62°C for 1 min, 72°C for 2 min, and a final extension of 72°C for 5 min. Positive pools were identified by Southern hybridization of the PCR products, using a ppr2-specific probe. Positive individuals within the positive pools were identified in the analogous fashion. The ppr2 sequence on the other side of each Mu insertion was subsequently amplified by PCR with the Mu TIR primer and a second ppr2-specific primer (5′-TCAGCTCCAAGAACTGGCAAGA), corresponding to sequences on the other side of the insertions. Plants were genotyped by genomic Southern blot and/or PCR analyses. Allelism between ppr2-1 and ppr2-4 was tested by intercrossing heterozygotes. Each of six such crosses, involving six different pairs of plants, yielded approximately 25% ivory seedlings, confirming that the ivory phenotype was the result of the insertions in the ppr2 gene.
Other mutants with similar pigmentation were used as controls in these experiments: crs2-1 mutants lack plastid ribosomes as a consequence of a defect in the splicing of chloroplast introns (Jenkins et al., 1997). Fully albino iojap mutants, which also lack plastid ribosomes (Walbot and Coe, 1979), were recovered as ‘maternal exceptions’ (Han et al., 1993) in the progeny of a cross between ij/ij ears and +/+ pollen. The ivory w1 mutant segregates in a standard Mendelian fashion and does not result from a primary defect in carotenoid biosynthesis (Han et al., 1993). Experiments involving extracts of wild-type plants used the inbred line B73 (Pioneer HiBred, Des Moines, IA, USA). Unless otherwise indicated, seedlings were grown in soil in a growth chamber under a 16-h light/8-h dark cycle at 26°C, and harvested between 7 and 10 days after planting.
Nucleic acid extraction and analysis
DNA for PCR amplification was isolated using plant DNAzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. DNA for genomic Southern analysis was prepared using a urea-based extraction protocol (Voelker et al., 1997). DNA samples (10 µg each) were digested with EcoRI, separated in a 0.8% agarose gel in 1× TBE, and blotted according to standard protocols (Sambrook et al., 1989). The ppr2 probe was generated by PCR, using the two ppr2 gene primers described above, and radiolabeled by random hexamer priming. Blots were probed at 65°C overnight in 7% SDS, 0.5 m sodium phosphate (pH 7; Church and Gilbert, 1984) and washed at the same temperature in 0.5× SSC, 0.5% SDS.
RNA was isolated using Tri Reagent (Molecular Research Center Inc., Cincinnati, OH, USA). RNA gel blot hybridizations were performed as described previously by Barkan et al. (1994). Blots probed for nucleus-encoded mRNAs had 20 µg of total RNA per lane. Blots probed for plastid RNAs had 5 µg of total RNA per lane. Hybridization and wash conditions were similar to those used for Southern analysis, except that the temperature was 68°C, and blots were washed in 0.2× SSC, 0.2% SDS. The following DNA fragments were used to probe for chloroplast RNAs in the data shown here (residue numbers from GenBank Accession number X86563): rrn16, 1276-bp HincII/BamHI fragment (residues 96835–95564); 5′ rrn23/trnA, 1454-bp HindIII/BamHI fragment (residues 122946–124405); 3′ rrn23, 513-bp PCR fragment (residues 121098–121611); rrn4.5/5, 614-bp PvuII/XmaI fragment (residues 102101–102715); rpoC1, 500-bp PCR product (residues 25831–26331); rps3/rpl22/rps19, 749-bp PCR product (residues 81710–82459); rps8/rpl14, 2383-bp BamHI/SalI fragment (residues 77280–79663); and rpl33/rps18/rpl20, 2184-bp KpnI fragment (residues 67061–69245).
The genomic clone of the ppr2 locus was isolated from a size-selected library of DNA extracted from the inbred line B73. A total of 100 µg of B73 DNA was digested with HindIII and fractionated in an agarose gel. DNA in the 5–6-kbp range was isolated from the gel, purified with QIAquick Gel Extraction kit (Qiagen, Valencia, CA, USA), and ligated into Bluescript SK+ (Stratagene, La Jolla, CA, USA) that had been digested with HindIII and treated with calf-intestine phosphatase. The ligation was transformed into XL1-blue MRF′ cells (Stratagene), and colonies with the ppr2 insert were detected by colony hybridization. The genomic sequence was characterized using genscan (http://genes.mit.edu/GENSCAN.html) to identify the putative open-reading frame. A possible promoter was identified 422 bp upstream of the putative start codon. The genomic sequence and deduced protein sequence of PPR2 are deposited in GenBank under Accession number # AY278988.
ppr2 cDNAs were obtained by PCR amplification from a green seedling leaf cDNA library (inbred line B73; Ostheimer et al., 2003), using the primers (5′-TGAATTCGCGTGCACCGTCACCGAGGC) and (5′-AACTGCAGATGGAGTTGAAGAATGGCCGC). Two ppr2 mRNA isoforms that differed in their site of polyadenylation were identified as ESTs in The Institute for Genomic Research database (http://www.tigr.org/), and we recovered examples of both isoforms. The intron/exon junction corresponded with the splice sites predicted by genscan.
Recombinant PPR2 was generated by expressing amino acids 181–571 into pMAL-C2 (New England Biolabs, Beverly, MA, USA), to generate a fusion protein with an amino-terminal maltose binding protein (MBP) tag. MBP::PPR2 was purified on amylose affinity resin (New England Biolabs) and used to generate polyclonal antisera from rabbits. The same PPR2 fragment was sub-cloned into the pet28 A(+) (Novagen, Madison, WI, USA) to generate PPR2 with an amino-terminal 6× histidine tag. HIS::PPR2 was purified on a Ni-NTA agarose column (Qiagen) and was used to affinity-purify the antisera.
Chloroplast fractionation and protein analyses
Proteins were extracted from leaf tissue, size-separated using SDS–PAGE, and immunoblotted as described by Barkan (1998). The antisera to AtpA, D1, PsaD, AtpB, PetA, and OE33 were generated in our laboratory and described previously by McCormac and Barkan (1999) and Voelker and Barkan (1995). The αIM35 serum, αMDH and αCpn60 sera were generously provided by Danny Schnell (Rutgers University), Kathy Newton (University of Missouri), and Tony Gatenby (DuPont), respectively. The PPR2 antiserum was generated in rabbits using MBP::PPR2 recombinant protein and affinity-purified using HIS::PPR2, as described above.
We would like to thank Susan Belcher and Roz Williams-Carrier for technical assistance, Nick Herschberger for help with RNA gel blot experiments, and Dianna Fisk for help with modeling the putative substrate binding faces of CRP1 and PPR2. We are grateful to Danny Schnell, Kathy Newton, and Tony Gatenby for providing antibodies. This work was supported by grants from the Department of Energy (DE-FG03-99ER20343) and the National Science Foundation (DBI 0077756).
Accession number: AY278988: ppr2 genomic sequence and deduced protein sequence.