TL;DR
This study comprehensively maps plastid RNA editing sites in the medicinal plant Glycyrrhiza uralensis, identifying 38 high-confidence sites, many in ndh genes, suggesting roles in stress adaptation. Crucially, it demonstrates that total RNA-seq is the most reliable method for de novo RNA editing discovery, while repurposed mRNA-seq data can also be effective with stringent filtering.
What Problem Does This Paper Address?
RNA editing is a critical post-transcriptional modification in plant organelles (plastids and mitochondria), primarily involving C-to-U conversions. These edits can restore conserved codons, correct genomic mutations, and influence RNA stability, playing a vital role in gene regulation and function, particularly for photosynthetic components. While RNA editing is widespread, it remains underexplored in many non-model plant species, including medicinal ones like Glycyrrhiza uralensis. Furthermore, the choice of RNA sequencing (RNA-Seq) library preparation strategy can significantly impact the accuracy and completeness of RNA editing detection. Previous studies have often focused on nuclear transcripts, leaving a gap in understanding how different RNA-Seq library types (total RNA, rRNA-depleted, and mRNA-selected) perform for organellar RNA editing detection, especially in species with economic or medicinal importance. This paper aims to fill this gap by characterizing the plastid RNA editing landscape in G. uralensis and systematically comparing the effectiveness of three common RNA-Seq library strategies for this purpose.
Methods
The researchers employed a multi-faceted approach:
- Plastid Genome Assembly and Annotation: They used whole-genome resequencing data to assemble complete circular plastid genomes for three G. uralensis individuals using the GetOrganelle pipeline. These genomes were then annotated using PGA to identify genes and their boundaries.
- RNA-Seq Library Preparation and Sequencing: Three distinct RNA-Seq library preparation strategies were used:
- Total RNA-seq: Direct RNA extraction without prior enrichment or depletion.
- rRNA-depleted RNA-seq: Removal of ribosomal RNA using the Epicentre Ribo-Zero kit.
-
mRNA-seq: Enrichment of polyadenylated mRNA using oligo(dT) magnetic beads. All libraries were prepared using a strand-specific protocol (dUTP method) to preserve strand information. Sequencing was performed on an Illumina NovaSeq 6000 platform.
-
Bioinformatic Analysis:
- Read Mapping: Cleaned RNA-Seq reads were mapped to the corresponding sample's plastid genome using HISAT2, optimized for strand-specific mapping.
- Variant Calling: Putative RNA editing sites (single nucleotide variants) were identified using REDItools 2.0.
- Stringent Filtering: A rigorous filtering process was applied to identify high-confidence editing sites. This included criteria such as a minimum number of uniquely strand-specific supporting reads, a minimum variant allele fraction, visual inspection in IGV, cross-library reproducibility checks, and BLAST searches against nuclear and mitochondrial genomes to rule out paralogs. Special attention was paid to filtering technical artifacts, including those arising from rRNA mapping, noncanonical substitutions, and antisense signals from dUTP-based library preparation.
Datasets
- Genomic Data: Whole-genome resequencing data from three wild-sampled Glycyrrhiza uralensis individuals.
- RNA-Seq Data: Strand-specific RNA-Seq data generated from the same three G. uralensis individuals using three different library preparation strategies (total RNA, rRNA-depleted, and mRNA-selected).
- Reference Genomes: The annotated G. uralensis reference plastome (Accession number: MZ329070.1), its mitochondrial genome (Accession number: NC_053919.1, MZ066515.1), and its nuclear genome (GCA_027886165.1) were downloaded from NCBI for artifact filtering.
Key Insights and Ideas
- Medicinal Plant RNA Editing Landscape: This study provides the first comprehensive characterization of plastid RNA editing in G. uralensis, a medicinally important plant.
- Adaptive Significance of ndh Editing: The identification of numerous editing sites in ndh genes (involved in cyclic electron flow and photoprotection) suggests a potential adaptive role in coping with the stressful environmental conditions (arid, saline-alkali soils) characteristic of its native habitat.
- Total RNA-Seq Superiority: Total RNA-seq emerges as the most robust and reliable method for de novo plastid RNA editing discovery due to its high sensitivity, excellent reproducibility, and significantly fewer rRNA-derived artifacts.
- mRNA-Seq Repurposing Potential: Even with inherently low plastid mapping rates, poly(A)-selected mRNA-seq, when coupled with stringent strand-specific bioinformatic filtering, can effectively recover a substantial portion of genuine plastid RNA editing sites. This highlights the potential to leverage vast public mRNA-seq datasets for organellar RNA editing research.
- Artifact Identification and Filtering: The study meticulously details and filters various technical artifacts, including those related to rRNA mapping, noncanonical substitutions, and strand-specific biases from dUTP-based library preparation, emphasizing the importance of rigorous bioinformatic pipelines.
Key Results
- 38 High-Confidence Editing Sites: A total of 38 high-confidence plastid RNA editing sites were identified across 19 genes in G. uralensis.
- Library Strategy Performance:
- Total RNA-seq: Consistently detected 37 out of 38 high-confidence sites across all replicates.
- rRNA-depleted RNA-seq: Recovered 35 editing sites but produced a higher number of rRNA-linked artifacts.
- mRNA-seq: Despite low mapping rates, recovered a significant fraction of bona fide sites when analyzed with stringent filtering.
- Editing Site Distribution: Editing sites were predominantly found in protein-coding regions, with ndhB harboring the most sites (8), followed by rpoB (4) and ndhD (3).
- Codon and Amino Acid Changes: Serine (18 codons) was the most frequently edited amino acid, followed by proline (10) and histidine (4). After editing, leucine became the most frequent amino acid encoded.
- Strand Specificity: All bona fide plastid RNA editing events were confirmed to be strand-specific and occur exclusively on the coding strand.
Limitations and Future Work
- Artifacts in rRNA Regions: While the study identified numerous variants in rRNA regions, these were largely dismissed as technical artifacts due to noncanonical substitutions and low reproducibility. The authors acknowledge that dynamic rRNA modifications are a biologically significant phenomenon, distinct from canonical plastid RNA editing, and may contribute to ribosomal heterogeneity.
- mRNA-Seq Limitations: While mRNA-seq can be repurposed, its low mapping efficiency for plastid transcripts makes it suboptimal for comprehensive, artifact-resistant genome-wide profiling compared to total RNA-seq.
- Future Directions: The authors suggest further studies linking editing efficiency at specific sites to environmental factors and conducting expression profiling of nuclear-encoded editing factors to better understand the roles of genetic divergence and adaptive plasticity in shaping RNA editing landscapes. They also highlight the vast potential of publicly available mRNA-seq data for large-scale exploratory analyses of RNA editing.
Summary updated