Unparalleled Variation in RNA Editing among Selaginella Plastomes^,

Abstract

The number and position of C-to-U RNA editing sites in Selaginella plastomes can be extremely variable, to a degree that is currently unparalleled in any other photosynthetic genus.

Dear Editor,

One of the most extreme documented examples of chloroplast RNA editing comes from the seedless vascular plant Selaginella uncinata (Lycopodiophyta), for which an astonishing 3494 cytosine-to-uracil editing events have been discovered (Oldenkott et al., 2014). Is posttranscriptional chloroplast editing as rampant in other Selaginella species? Here, I examine plastome-wide RNA editing profiles for Selaginella kraussiana and Selaginella lepidophylla and report that the number and position of edited sites can be extremely variable among Selaginella plastomes, to a degree that is currently unparalleled in any other photosynthetic genus.

RNA editing sites were identified by mapping publicly available Illumina RNA sequencing (RNA-seq) reads from S. kraussiana (GenBank accessions SRR2045379–82) and S. lepidophylla (SRR6345606–15) onto the respective chloroplast genome sequences of these two lycophytes (Supplemental Materials and Methods; Mower et al., 2019). For each species, the RNA and plastome sequencing data came from the same cultivar (and laboratory; Ge et al., 2016; VanBuren et al., 2018), greatly reducing the potential of mistaking polymorphisms between specimens as editing events. Mapping of the RNA-seq reads gave near-complete coverage (98%) of the reference chloroplast genomes, including all genes. Mean coverage of the plastomes exceeded 500×, providing robust alignments for identifying edited sites, which were only characterized in regions with ≥ 5× coverage and ≥ 25% read support (Supplemental Materials and Methods); thus, keep in mind that sites with low editing efficiency (< 25%) were not recorded in this study.

A total of 1353 and 720 C-to-U changes, respectively, were identified in the S. kraussiana and S. lepidophylla chloroplast transcriptomes (Table 1; Supplemental Tables S1 and S2), making them the most heavily RNA-edited plastomes from the Viridiplantae (Ichinose and Sugita, 2016), outdone only by that of Selaginella uncinata. Approximately 80% of the observed edits from the two plastomes occurred in protein-coding regions and included synonymous and nonsynonymous changes as well as many instances in which start and/or stop codons were restored (Table 1; Supplemental Tables S3 and S4). The remainder of the edits (∼20%) were restricted to intergenic and intronic segments (Table 1), meaning not a single change was recorded in ribosomal RNAs (rRNAs) or transfer RNAs (tRNAs), and no U-to-C changes were found, which parallels the editing data from S. uncinata (Oldenkott et al., 2014).

Table 1. Cytosine-to-uracil RNA editing in the plastomes of Selaginella lepidophylla (Sl), S. kraussiana (Sk), and S. uncinata (Su).

Genomic Feature	Sl	Sk	Sub
Total no. of editing sitesa	720	1353	>3494
Editing sites in protein-coding regions	581	1,104	3,415
Nonsynonymous substitutions	530	972	2987
Synonymous substitutions	51	132	428
Start codon restoration	22	18	52
Stop codon restoration	9	12	31
Most edited gene (no. sites)	rpoB (65)	ccsA (66)	rpoB (214)
Editing sites in intergenic regions	128	236	>5c
Editing sites in intronic regions	11	13	74c
Plastome size (kb)	114.69	129.97	144.17
Plastome GC content	51.9	52.3	54.8

^aRNA editing sites in the large inverted (or direct) repeat region were counted only once.

^bData from Oldenkott et al. (2014).

^cOnly 1139 nt of intergenic chloroplast RNA and four introns were surveyed for editing.

The huge amount of RNA editing in Selaginella plastomes is striking, but equally remarkable is the variation in the number of edited sites among species. Indeed, The S. uncinata plastome has ∼2150 and ∼2775 more C-to-U alterations than its S. kraussiana and S. lepidophylla counterparts. In other words, there is a 2- to 5-fold difference in plastome editing across these three taxa—and that is likely an underestimate, as only 1139 nucleotides (nt) of intergenic chloroplast RNA from S. uncinata have been surveyed for editing (Oldenkott et al., 2014). To the best of my knowledge, this is the largest reported difference in chloroplast editing for any genus studied to date—but see Klinger et al. (2018) for other extreme examples.

The variation in chloroplast editing is also reflected in the location of C-to-U changes within the Selaginella plastomes, as well as in which protein-coding transcripts are (or are not) edited and in the relative number of editing sites in those genes (Supplemental Tables S1–S4). Alignments of the protein-coding chloroplast DNA and RNA from the three Selaginella species (Supplemental Materials and Methods) showed that more than 40% of the edits identified in S. kraussiana and S. lepidophylla are unique—i.e. the C-to-U change was found in only one of the species. Thus, the total pool of uniquely edited sites currently identified within the Selaginella genus easily exceeds 2500, not including edits within the intergenic regions, which could not be aligned.

Some similarities in the editing patterns were also observed (Table 1; Supplemental Tables S3 and S4). For example, in S. uncinata, S. kraussiana, and S. lepidophylla, the gene encoding the D1 protein of PSII (psbA), which is over 1000 nt long, has no detectable editing sites. Likewise, for both S. uncinata and S. lepidophylla, the chloroplast gene for the beta subunit of RNA polymerase (rpoB) gene has the largest number of editing sites (but not so for S. kraussiana). For all three species, many of the editing events are clustered close together (Supplemental Tables S1 and S2; Oldenkott et al., 2014).

Together, these data suggest that throughout the evolution and diversification of Selaginella, there has been the gain and/or loss of thousands of chloroplast RNA-editing sites and that this process is still ongoing. What’s more, the same is probably true for the mitochondrial genome of this genus, which undergoes equal (or even greater) amounts of C-to-U RNA editing than the plastome (Smith, 2009; Hecht et al., 2011; Oldenkott et al., 2014). I did try to mine mitochondrial transcripts from the S. kraussiana and S. lepidophylla RNA-seq datasets but was unsuccessful.

As more Selaginella species are investigated, the breadth of the variation in RNA editing is sure to grow. Preliminary analyses of the Selaginella moellendorffii chloroplast genome suggest that it has at least 1800 C-to-U modifications (Oldenkott et al., 2014), more than those of S. kraussiana and S. lepidophylla. Not surprisingly, there appears to be a positive relationship between Selaginella plastome GC content, which is among the highest of any lineage, and the number of C-to-U editing sites (Table 1; Smith, 2009). Thus, to capture the complete range of chloroplast RNA editing, it might be useful to target species that are predicted to have very high plastome GC contents, such as Selaginella fragilis, as well as those with much lower predicted GC compositions, like S. sinensis (Smith, 2009).

No matter how large the variation turns out to be, the question remains: why do Selaginella plastomes (and mitogenomes) undergo such extensive C-to-U editing? The evolutionary origins of RNA editing in organelle systems can be eloquently explained by the concept of constructive neutral evolution, which “posits that the biochemical elements of an RNA editing system must be in place before there is an actual need for editing” (Gray, 2012). Among the key players in plant organelle RNA editing are nuclear-encoded, aspartic acid-tyrosine-tryptophan (DYW)-domain-containing pentatricopeptide repeat (PPR) proteins, some of which are known to be site recognition factors for editing events (Ichinose and Sugita, 2018). In land plants, the size and diversity of DYW-type PPR gene families appears to be positively associated with the abundance of organelle RNA editing (Rüdinger et al., 2012). And, as one might expect, the S. moellendorffii genome (Banks et al., 2011) encodes an expanded DYW-type PPR protein family: ∼312 members (Cheng et al., 2016). As more data become available, it will be particularly interesting to compare variation in the number of PPR proteins from Selaginella species with chloroplast RNA editing abundance and to take advantage of bioinformatics programs that use the PPR-RNA binding code to predict binding events (Harrison et al., 2016; Yan et al., 2019). I anticipate that species with the largest number of RNA editing sites will have the most expanded PPR protein gene families and predicted PPR binding events, and vice versa.

Plant organelle RNA editing is a burgeoning field and was recently implicated in chloroplast-to-nucleus communication (Zhao et al., 2019), opening new research avenues. Given its status as a model lineage and its unrivaled number and diversity of chloroplast C-to-U editing sites, Selaginella is well positioned to become a leading system for studying posttranscriptional organelle editing. I look forward to seeing what future work will uncover.

Accession Numbers

All accession numbers used in this study are listed in the Supplemental Materials and Methods.

Supplemental Data

The following supplemental materials are available.

• Supplemental Materials and Methods. Detailed information on experimental procedures.

• Supplemental Table S1. Chloroplast RNA editing in the plastome of S. kraussiana.

• Supplemental Table S2. Chloroplast RNA editing in the plastome of S. lepidophylla.

• Supplemental Table S3. Chloroplast RNA editing in the protein-coding transcripts of S. kraussiana.

• Supplemental Table S4. Chloroplast RNA editing in the protein-coding transcripts of S. lepidophylla.