Evidence that RNA editing in ferns performs as a gene-regulating mechanism

RNA editing occurs only in the organellar genomes of chloroplasts and mitochondria of land plants, where it defies a central dogma of biology: DNA->RNA ->protein. Through post-transcriptional pyrimidine exchange it alters the sequence of the mRNA transcripts to differ in sequence from their DNA templates. C-to-U edits prevail across land plants and U-to-C changes are restricted to ferns, hornworts, and some lycophytes. C-to-U edits typically restore evolutionarily conserved amino acid sequences and can generate AUG start codons, whereas U-to-C edits typically replace DNA-encoded internal stop codons with sense codons. Here we combine genomic and transcriptomic data with bioinformatic and comparative phylogenetic methods to investigate the evolution of RNA editing in ferns. We find that C-to-U and U-to-C editing sites evolve at different rates, with U-to-C edits generally being more evolutionarily conserved. We also find C-to-U edits that generate AUG start codons are maintained. Most nonsynonymous RNA editing sites are edited with high efficiency, whereas RNA edits that create start codons or that remove internal stop codons are edited with variable efficiency, implying they may be subject to selection pressures. Taken together, we provide strong evidence to suggest that these less-efficient RNA edits act as a gene expression regulation mechanism that is selectively advantageous.