RNA editing in plant mitochondria: 20 years later
In 1989, three laboratories (in Canada, France and Germany) independently and simultaneously reported the discovery of C-to-U RNA editing in plant mitochondria (1–3). To mark the 20th anniversary of this finding, the leaders of the three research teams have written personal essays describing the events leading up to the discovery in each of their laboratories. These essays are intended not only to capture historical facts but also to illustrate unexpected convergence in the process of scientific discovery, with different groups coming to the same conclusion, often very close together in time, drawing on different types of evidence and via sometimes quite different hypotheses and approaches. Essential background information pertaining to RNA editing in general and RNA editing in plant organelles in particular is provided in this overview. © 2009 IUBMB IUBMB Life, 61: 1101–1104, 2009
The year 2009 marks the 20th anniversary of the discovery of RNA editing in plant mitochondria (1–3). The roots of this tri-lab discovery trace back to the publication by Fox and Leaver in 1981 (4) of the first sequence of a plant mitochondrial protein-coding gene (cox2 in maize, specifying Cox2p, subunit 2 of cytochrome c oxidase). Comparison of the inferred amino acid sequence with Cox2p sequences from non-plant species showed CGG (Arg) codons aligning at positions of conserved Trp residues, in addition to conserved Arg residues. This observation led the authors to suggest “that codon CGG (normally Arg) codes for Trp in maize mitochondria, in addition to the standard Trp codon TGG.” Considering that codon reassignments had previously been inferred for the mitochondrial translation systems of both animals (5) and yeast (6–8), this suggestion did not seem an unreasonable one at the time. Nevertheless, it remained unclear how the same codon could be used to specify two different amino acids, and no plant mitochondrial tRNA that could decode CGG as Trp was ever found. This puzzle was solved by the discovery 8 years later (1–3) that the supposed Trp-encoding CGG codons are, in fact, changed to normal UGG Trp codons by C-to-U editing during maturation of the cox2 mRNA.
RNA editing, one definition of which is “the programmed alteration of the nucleotide sequence of an RNA species, relative to the sequence of the encoding DNA” (9), was first described in 1986 by Benne et al. (10), who coined the term to describe a process that inserts uridine residues into transcripts of protein-coding genes in the mitochondria of kinetoplastid protozoa such as Trypanosoma brucei, the causative agent of African sleeping sickness (this editing process was subsequently found to delete uridine residues, as well). A year later, Powell et al. (11) and Chen et al. (12) reported a single, tissue-specific C-to-U modification resulting in a CAA-to-UAA codon change in the mRNA coding for apolipoprotein B in mammals, resulting in the synthesis of a truncated form (apoB48) of apolipoprotein B in intestinal cells. Trypanosomatid mitochondrial editing and apolipoprotein B editing exemplify two general types of editing: insertion/deletion, in which residues are added to and/or taken away from the gene-specified sequence; and substitution, where the sequences of the edited RNA and its gene are co-linear, nucleotide for nucleotide, but not identical.
The term “RNA editing” now encompasses a range of diverse RNA processing systems that change the nucleotide sequence of an RNA species relative to that of the gene that encodes it. Although RNA editing predominantly occurs within coding regions of mRNA, it also occasionally targets non-coding and intronic regions of protein-coding transcripts, as well as acting on tRNA, rRNA and certain viral RNAs (9). RNA editing systems are mostly a eukaryotic phenomenon; such systems are especially prominent in organelles, particularly mitochondria (13).
In the case of protein-coding transcripts, editing works to create mature, translatable mRNAs, with effects that include: changes in the encoded amino acid sequence; generation of initiation or termination codons; elimination of stop codons (e.g., UGA→UGG); correction of frameshifts; and, in the case of certain trypanosomatid mitochondrial transcripts (e.g., 14), wholesale creation of open reading frames, where upwards of 50% or more of the nucleotides in the final mature mRNA sequence may be contributed by the editing process. Where it has emerged in evolution, RNA editing has become an integral part of the genetic information transfer process.
RNA EDITING IN PLANT ORGANELLES
In angiosperms (flowering plants) and gymnosperms, most of the mtDNA-encoded protein-coding transcripts undergo multiple site-specific C-to-U edits during mRNA maturation. Most of these changes are at first or second positions of codons, and so most of them alter the nature of the encoded amino acid (9). Rarely, U-to-C changes are also seen in advanced land plants, but in some early diverging plants such as ferns and hornworts, these “reverse” edits may be as frequent as C-to-U edits (15). A few primitive plant species, such as the bryophyte Marchantia polymorpha (a liverwort), appear to lack RNA editing altogether, but otherwise the phenomenon is widespread within the diversity of land plants, even in other bryophyte taxa (16).
On the other hand, comparison of mitochondrial genome sequences has uncovered no evidence of either C-to-U or U-to-C RNA editing (or any other form of editing) within the phylogenetic sister group to land plants, the green algae—and, in particular, within the specific green algal relatives of land plants, the charophyte algae (Charophyceae). By whatever pathway this type of editing emerged in evolution (17), it is clearly a derived trait, first appearing at the base of the land plant lineage, after the latter's separation from the charophyte lineage (13).
Two years after the discovery of C-to-U editing in plant mitochondria, the same process was reported in plant chloroplasts (18). In angiosperms, the frequency of chloroplast editing is roughly one-tenth that observed in mitochondria, with only a fraction of chloroplast mRNAs affected by editing. However, in some early diverging plant lineages, a relatively high level of RNA editing in chloroplasts has been observed, with a greatly elevated frequency of U-to-C relative to C-to-U editing. These parallels between RNA editing in the two plant organelles indicate that they likely use a common nucleus-encoded system, components of which are synthesized in the cytosol and differentially targeted to mitochondria and/or chloroplasts.
MECHANISM OF PLANT RNA EDITING: CATALYTIC ACTIVITY AND SPECIFICITY DETERMINANTS
Whether the same or two different enzyme systems effect C-to-U and U-to-C changes in plant organelles is still unknown, as the catalytic activity(ies) remain(s) elusive, some 20 years after the initial discovery of RNA editing in plant mitochondria. Data indicate that C-to-U edits occur by base modification rather than by base or nucleotide replacement (19–21), pointing to a cytidine deaminase as the responsible activity (the available evidence does not, however, exclude a transaminase activity, which could theoretically account for both C-to-U and U-to-C edits). Reproducible mitochondrial or chloroplast extracts capable of supporting efficient editing in vitro have proven very difficult to generate, so that analysis of the mechanism of C-to-U editing has to date been limited, with no biochemical investigation at all being reported in the case of U-to-C editing.
In contrast, considerable progress has been made recently in defining the specificity determinants of C-to-U editing: i.e., those elements/factors that determine which C residues will be selected for editing. A number of studies have identified essential sequence motifs in the vicinity of editing sites in both chloroplasts and mitochondria, with these cis-acting elements interacting with trans-acting factors to select C residues for editing (see 22). The cis-acting elements seem to be specific for one or at most a few editing sites (e.g., 23). By analogy with the small guide RNAs that identify the sites of U insertion and deletion during trypanosomatid mitochondrial RNA editing (24), it was long anticipated that trans-acting factors in plant organellar RNA editing would likely be small guide-type RNAs; however, no such RNAs have been identified in either plant mitochondria or chloroplasts. Instead, in this case the trans-acting specificity factors turn out to be proteins, so that interactions with the cis-acting specificity elements are RNA-protein rather than RNA-RNA.
Considering the very large number of editing sites typically found in plant organelles—e.g., 441 mitochondrial (25) and 34 chloroplast (26) in Arabidopsis thaliana—a correspondingly large number of proteins would have to be elaborated for such a process, even accepting that some editing factors might be able to direct editing at more than one site (see, e.g., 27). Increasingly, it appears that this role is fulfilled by a family of organellar-targeted RNA-binding proteins, uniquely encoded by land plants and characterized by tandem arrays of a degenerate 35-amino-acid repeat, termed the pentatricopeptide repeat, or PPR (28). Genetic and biochemical studies have implicated PPR proteins in a wide range of RNA metabolic processes in plant organelles (29), including not only editing but also transcript splicing, stability and translation (30). The current working hypothesis is that the PPR proteins interact directly with a specific site in the target transcript, and that the PPR protein in turn recruits the generic enzymes responsible for RNA maturation, such as C deaminase (editing) or RNA endonuclease (processing) (22).
The first specific editing factors have recently been identified in both land plant chloroplasts (31–37) and land plant mitochondria (27, 38). All of these specificity factors have turned out to be PPR proteins; moreover, all are members of a specific class, PLS, characterized by triplet repeats of P, L (“long,” generally 36 amino acids) and S (“short,” generally 31 amino acids) motifs and an extended C-terminal domain. PLS proteins having the foregoing structure are members of the “E” subclass, whereas a second subclass (“DYW”) possesses an additional distinct C-terminal extension, the DYW domain, named for a defining Asp-Tyr-Trp triad.
Although PPR genes comprising only tandem PPR repeats (P class) are phylogenetically widely distributed, genes for PLS-type PPR proteins have so far been found only in land plants (29). Moreover, PPR proteins (i.e., P + PLS) are uniquely present in exceptionally high numbers only in land plants, ranging from ∼100 in the moss, Physcomitella patens, to ∼600 in rice, Oryza sativa (30), a distribution that parallels the known occurrence of the plant-type of substitution editing. As noted above, no such organellar editing is evident within the green algae, including within the Charophyceae, the green algal lineage that is sister to land plants. No charophyte nuclear genome sequences have yet been determined, so the number of PPR proteins encoded by charophyte genomes is unknown; however, the genome of Chlamydomonas reinhardtii, a more distantly related green alga, encodes only 12 PPR genes, on a par with other eukaryotes (29).
Intriguingly, a correlation seems to be emerging between the total number of editing sites and the number of PLS-type PPR proteins encoded by a given plant genome. Of ∼450 PPR genes identified in the genome of A. thaliana, 200 are PLS class. In contrast, in P. patens, where only 11 mitochondrial and two chloroplast edits have been documented (39), only 10 of 103 identified PPR genes are PLS (all DYW type). These results have focused attention on the PLS class of PPR proteins as key specificity determinants for RNA editing in plant organelles.
The discovery of RNA editing in plant organelles ushered in a new era of research in plant organellar biology. One immediate practical consequence was that workers in the field now had to infer protein sequences from full-length cDNA (edited mRNA) sequences rather than from the corresponding (unedited) gene sequences encoded by plant mitochondrial or chloroplast DNA. The continued development of more efficient in vitro editing systems (e.g., 40) will undoubtedly provide further insights into both the contribution of cis-acting elements to the editing process and the nature of the catalytic activity, the current “Holy Grail” of the field. A combination of biochemical, genetic and bioinformatic approaches has led to the exciting discovery that PPR proteins are RNA editing factors in plant organelles, a finding that has greatly stimulated research in the field. Increasingly, the critical role of RNA editing in plant development is being recognized and explored (e.g., 37). The authors of the accompanying essays hope that their reflections will provide a flavour of the early work that led, in the ensuing 20 years, to what has become a key area of research in plant organellar biology.