Skip to main content
NCBI home page
aBIOTECH logoLink to aBIOTECH
letter
. 2025 Nov 8;7(1):100007. doi: 10.1016/j.abiote.2025.100007

High levels of ADAR overexpression induce abundant and stochastic off-target RNA editing in rice protoplasts

Shuai Yin a,b, Shumin Li a,b, Yang Yuan a,b, Kabin Xie a,b,⁎
PMCID: PMC12973391  PMID: 41940153

Abstract

Adenosine deaminase acting on RNA (ADAR) mediates adenosine-to-inosine (A-to-I) conversions, with its deaminase domain (ADARdd) acting as a core component of targeted RNA editing. However, the off-target effects of ADARdd remain poorly characterized in plants. Here, we demonstrate that high-level overexpression of ADARdd induces widespread off-target editing in rice protoplasts, an effect not observed in stable transgenic lines. From six biological replicates of ADARdd-overexpressing protoplasts, we identified 2794 off-target editing sites (OFTEs). These OFTEs exhibited stochastic distribution across mRNAs and low editing efficiency (20–40 %), contrasting sharply with the high-confidence editing sites in OsDRB1-ADARdd stable lines. We further identify risky endogenous promoters and establish an empirical expression threshold for ADARdd to minimize off-target editing. Our findings highlight that controlled ADARdd expression is essential for reliable RNA editing in plants, providing valuable guidance for enhancing the precision of plant RNA editing experiments.

Keywords: RNA editing, ADAR, Off-target, Protoplast, Rice

Dear Editor,

Adenosine deaminase acting on RNA (ADAR) enables RNA editing through adenosine-to-inosine deamination, which is biochemically interpreted as adenosine-to-guanosine (A-to-G) mutations [1]. The ADAR deaminase domain (ADARdd) has been repurposed for targeted RNA identification by editing (TRIBE), enabling in vivo mapping of RNA-binding protein (RBP) partners [2,3] (Fig. 1A). The TRIBE method has been successfully employed to identify RNA partners for rice (Oryza sativa) OsDRB1 [4], Arabidopsis thaliana AtECT2 [5], and AtUBP1 and OsUBP1 [6]; however, the method's off-target effects have not been fully evaluated. Here, we demonstrate that high levels of ADARdd overexpression generate abundant transient nonspecific A-to-G edits on RNAs in protoplasts and discuss the implications for RNA editing experiments.

Fig. 1.

Fig. 1

Open in a new tab

High levels of ​ADARdd ​overexpression induce thousands of off-target RNA edits in rice protoplasts. A Schematic of targeted editing and transient editing by ADARdd. In targeted RNA editing, ADARdd is fused with an RBP to mark the RNA partners with A-to-I editing, which is interpreted as A-to-G mutations. In transient off-target RNA editing, excess ADARdd transiently deaminates adenosines in nearby RNAs. B Experimental workflow for assessing the off-target effects of ADARdd. ADARdd-OE rice leaves (ADARdd-OE-leaf), protoplasts extracted from ADARdd-OE rice plants transfected with empty vector (EV; ADARdd-OE ​+ ​EV) or UBI10p::ADARdd (ADARdd-OE ​+ ​ADARdd) plasmids, protoplasts extracted from wild-type rice plants transfected with UBI10p::ADARdd (WT ​+ ​ADARdd) or UBI10p::MCP-ADARdd (WT ​+ ​MCP-ADARdd) plasmids, and OsDRB1-OE leaves (OsDRB1-OE-leaf) were subjected to RNA-seq to identify RNA‒DNA variants (RDVs); the number of biological replicates is shown. OsDRB1-OE-leaf samples served as negative controls. 2 ​× ​MS2, MS2 RNA; U3p, snoRNA U3 promoter; UBI10p, UBIQUITIN 10 promoter; MCP, MS2 coat protein. NLS, nuclear localization sequence. C Radar chart comparing the average counts of 12 types of RDVs in the OsDRB1-OE-leaf, ADARdd-OE-leaf, ADARdd-OE ​+ ​EV, ADARdd-OE ​+ ​ADARdd, WT ​+ ​ADARdd, and WT ​+ ​MCP-ADARdd samples. D Comparisons of the counts of RNA-edited RDVs and non-RNA-edited RDVs across different samples. Data are presented as means ​± ​SDs (n ​= ​3 replicates, except for ADARdd-OE ​+ ​EV and OsDRB1-OE-leaf [n ​= ​2 replicates]). Different lowercase letters denote statistically significant differences (Kruskal–Wallis test followed by Wilcoxon rank-sum tests, α ​= ​0.05). EADARdd expression levels in different samples. TPM, transcripts per kilobase million. Data represent means ​± ​SDs. Different lowercase letters denote statistically significant differences (one-way ANOVA followed by post hoc Tukey's HSD test, α ​= ​0.05). F Correlation between ADARdd expression level (TPM) and the number of A-to-G/T-to-C RDVs. r, Pearson's correlation coefficient. G Bioinformatics pipeline for RDV filtering and off-target edit (OFTE) identification. H Counts and percentages of filtered RDVs. For the WT ​+ ​ADARdd and WT ​+ ​MCP-ADARdd samples, potential RNA editing (A ​> ​G and T ​> ​C) and other RDVs are displayed separately as separate categories in the bar graph. I-J Frequencies of OFTEs in WT ​+ ​ADARdd and WT ​+ ​MCP-ADARdd samples. Bar graph (I) shows the occurrence frequency of OFTEs across three combined WT ​+ ​ADARdd datasets and three WT ​+ ​MCP-ADARdd datasets (total replicates, n ​= ​6). Pie chart (J) displays the percentages of OFTEs detected in 1, 2, or ​> ​2 replicates. K-L Distribution of RNA edits across mRNA transcripts. UTR, untranslated region; CDS, coding sequence. M Density plots of the editing percentages of OFTEs and HiCEs. The dashed lines denote the mean editing efficiency for each dataset. N Venn diagram identifying 197 overlapping editing sites (within 75 genes) shared between OFTEs and OsDRB1-HiCEs. O Correlation analysis between the expression level and edit count for 75 overlapping genes. r, Pearson's correlation coefficient (for HiCE, n ​= ​4 replicates; for OFTE, n ​= ​6 replicates). P-Q Expression levels of 75 genes that overlap between the OFTE and OsDRB1-HiCE datasets. The genes were categorized based on their off-target editing frequencies in six protoplast samples, which were combined from the WT ​+ ​ADARdd and WT ​+ ​MCP-ADARdd datasets (Q). Also shown is a comparison of the average TPM values from the protoplast and leaf datasets (P). Numbers above boxplots denote P values (Wilcoxon signed-rank test, paired samples, one-tailed). R–S RNA-EMSA validation of the binding between OsDRB1 and the 3′ UTR of Os04g0380300. (R) Predicted stem‒loop structure of the 3′ UTR sequence of Os04g0380300. The adenosine in red is the editing site. (S) RNA-EMSA with recombinant OsDRB1 and a Cy5-labeled 3′ UTR probe. Unlabeled probes (100-fold excess) served as competitors.

Although global off-target effects were not previously detected in stable ADARdd-overexpressing (ADARdd-OE) rice [4], we unexpectedly observed numerous RNA-to-DNA variants (RDVs), particularly A-to-G and T-to-C variants, in RNA-seq data from rice protoplasts transfected with plasmids overexpressing ADARdd. To investigate this phenomenon, we simultaneously examined the RDVs in leaf samples and ADARdd-transfected protoplasts from the same rice seedlings (Fig. S1). In leaf samples, RNA-seq revealed low levels (169–352) and comparable frequencies of all 12 types of RDVs (Fig. S1A and S1B). In contrast, protoplasts transfected with plasmids overexpressing ADARdd driven by the rice UBIQUITIN10 promoter (UBI10p::ADARdd; p33-ADARdd) contained 16,005 and 15,983 A-to-G and T-to-C RDVs, respectively. These RDVs were approximately 30–55 times more abundant than the other 10 types of RDVs (Fig. S1C and S1D). In our previous study, leaf samples from stable transgenic lines overexpressing ADARdd generated from the same plasmid (p33-ADARdd) presented RDV spectra similar to those of wild-type rice (Fig. S1E–G) [4]. This discrepancy raises concerns that ADARdd has pronounced off-target effects, at least in the protoplast transient expression system.

We hypothesized that the high degree of off-target RNA edits in ADARdd-transfected protoplasts may arise from (1) excessively elevated ADARdd expression, (2) altered ADARdd activity under physiological conditions of protoplasts, and/or (3) the lack of an RNA-binding domain. To investigate the effect of protoplast physiology on ADARdd off-target editing, we prepared protoplasts from ADARdd-OE seedlings and transfected them with either empty vector (ADARdd-OE ​+ ​EV) or UBI10p::ADARdd-bearing plasmids (ADARdd-OE ​+ ​ADARdd) (Fig. 1B). In the ADARdd-OE ​+ ​ADARdd experiment, we reduced the amount of plasmid DNA used for transfection by 75 ​% relative to the other protoplast transfections (See Materials and Methods). We also prepared protoplasts from wild-type rice seedlings and transfected them with either the UBI10p::ADARdd (WT ​+ ​ADARdd) or UBI10p::MCP-ADARdd (WT ​+ ​MCP-ADARdd) plasmid (Fig. 1B; see Materials and Methods). The UBI10p::MCP-ADARdd plasmid expresses MCP-fused ADARdd, which contains an N-terminal nuclear localization signal (NLS) driven by the rice UBIQUITIN10 promoter, with an MS2-RNA ligand driven by the rice snoRNA U3 promoter (Fig. 1B). This construct was designed to assess whether the RBP-ADARdd fusion protein or the NLS would improve ADARdd specificity, as observed in animal systems [7]. We obtained RNA-seq data from three biological replicates for all constructs except ADARdd-OE ​+ ​ADARdd, which had two replicates.

ADARdd expression levels were 2–4 times greater in protoplasts than in leaf samples, as determined by RT-qPCR (Fig. S2A and S2B; Table S1 and S2). In addition to protoplasts, we included RNA-seq data from the leaves of ADARdd-OE (ADARdd-OE-leaf) seedlings and the negative control: OsDRB1-overexpressing seedlings (OsDRB1-OE-leaf) [4]. The global gene expression profiles were highly correlated within the sample groups (protoplasts or leaves; Pearson's correlation coefficient, r ​= ​0.95–0.98; Fig. S2C), whereas the gene expression patterns differed slightly between the protoplast and leaf samples (Pearson's correlation coefficient, r ​= ​0.77–0.81; Fig. S2C).

Analysis of RDV spectra indicated that all protoplast samples transfected with ADARdd or MCP-ADARdd had elevated levels of A-to-G and T-to-C RDVs, whereas the ten other RDV types remained at levels comparable to those of the control (Fig. 1D, Table S3). Moreover, ADARdd-OE ​+ ​ADARdd protoplasts contained fewer of the 10 non-RNA-editing RDV types than the control. Pearson's correlation coefficient analysis indicated that these two RNA-seq datasets differed slightly from other protoplast samples (Fig. S2), suggesting that the modest reduction in RDVs may be attributable to batch effects during RNA-seq. Consistent with this, all 12 types of RDVs showed low abundance in ADARdd ​+ ​EV protoplasts (Fig. 1C and D). These data rule out the possibility that the physiological conditions of protoplasts induced off-target editing by ADARdd. Furthermore, both the WT ​+ ​ADARdd and WT ​+ ​MCP-ADARdd protoplast samples contained 2.9–3.1 times more A-to-G and T-to-C RDVs than the other variants (Fig. 1C and D), indicating that neither the RNA-binding domain nor the NLS reduces off-target activity. Notably, off-target editing activity increased significantly when ADARdd expression (measured as log2transformed transcripts per million, log2[TPM]) exceeded 13 (Fig. 1E and F). Furthermore, the number of A-to-G and T-to-C RDVs strongly correlated with the expression levels of ADARdd and MCP-ADARdd (Pearson's correlation coefficient, r ​= ​0.92, P ​< ​0.001; Fig. 1F), suggesting that high levels ADARdd overexpression causes global off-target RNA editing.

To comprehensively characterize ADARdd-mediated off-target editing in protoplasts, we used a published bioinformatics pipeline [4] to filter out genomic single-nucleotide polymorphisms (SNPs) and identify off-target edits from the WT ​+ ​ADARdd and WT ​+ ​MCP-ADARdd protoplast samples, with other samples used as controls (Fig. 1G). All RDVs in ADARdd-OE-leaf and ADARdd-OE ​+ ​EV protoplast samples were classified as either genomic SNPs or low-confidence RDVs (depth <10). For the WT ​+ ​ADARdd protoplast and WT ​+ ​MCP-ADARdd protoplast datasets, this pipeline filtered approximately 97.9 ​% of non-RNA-edited RDVs and 62.9 ​% of A-to-G/T-to-C RDVs as genomic SNPs and low-confidence variants. Conversely, 37.1 ​% of A-to-G/T-to-C RDVs, corresponding to 1747 and 1391 off-target RNA edits (OFTEs), respectively, remained (Fig. 1H; Table S3 and S4). The integration of these two datasets yielded 2794 OFTEs, which exhibited distinct recurrence patterns: 65 ​% were detected in only one replicate, while 18 ​% were detected in two replicates. Notably, 1 ​% (35 sites) of the OFTEs were consistently detected across all six replicates (Fig. 1I and J; Fig. S3). Comparative analysis with high-confidence editing sites (HiCEs) from the OsDRB1-ADARdd datasets [4] revealed distinctions between off-target editing and on-target editing events: OFTEs exhibited a uniform mRNA distribution with modest editing percentages (20–40 ​%), whereas OsDRB1-HiCEs were enriched with untranslated regions (UTRs) and showed higher editing percentages (40–100 ​%) (Fig. 1K-M). These findings suggest that ADARdd-mediated off-target editing produces stochastic, low-efficiency modifications that are distinct from targeted editing outcomes.

We investigated potential false positives among previously reported OsDRB1-HiCEs derived from leaf samples of stable transgenic lines expressing OsDRB1-ADARdd. Within the OFTE dataset, 197 sites, corresponding to 75 genes, overlapped with the OsDRB1-HiCEs (Fig. 1N; Table S5). These overlapping genes showed modest correlations between editing frequency and gene expression levels in protoplast samples but not in OsDRB1-ADARdd-overexpressing leaf samples (Fig. 1O). Notably, 19, 13, and 7 genes were consistently edited by ADARdd in 4, 5, and 6 biological replicates, respectively, in the OFTE datasets, and these genes were expressed at higher levels in protoplasts than in leaves (Fig. 1P-Q; Fig. S4). Given that these RDVs observed in more than three replicates resisted bioinformatic filtering, we experimentally validated their interactions with OsDRB1. Among the 19 genes edited in more than three replicates, Os04g0380300 contained the most OFTE sites, which were further confirmed by Sanger sequencing of RT-PCR products (Fig. S4). The 3′ UTR of the Os04g0380300 transcript contains a stem‒loop structure and is edited by both ADARdd and OsDRB1-ADARdd (Fig. 1R). An RNA electrophoresis mobility shift assay (RNA-EMSA) using recombinant OsDRB1 protein confirmed the specific binding between OsDRB1 and the 3′ UTR of Os04g0380300 (Fig. 1S), indicating that this gene is the genuine RNA partner of OsDRB1. These data demonstrate that the off-target edits triggered by high levels of ADARdd overexpression in protoplasts did not compromise the results from stable OsDRB1-ADARdd transformants.

Finally, we assessed the potential risk of off-target editing caused by the abnormal high levels of ADARdd overexpression in rice. Based on our observations, we established that ADARdd expression beyond a critical threshold (log2[TPM] ​> ​13, Fig. 1F) is associated with a significant risk of global off-target editing. Among the 23 ADARdd-OE rice plants, the levels of ADARdd mRNA (maximum log2[TPM] ​= ​11.75) were 2–5-fold lower than those observed in protoplasts (Fig. S2 and S5), suggesting that TRIBE vectors employing the rice UBIQUITIN10 promoter yield minimal off-target RNA editing in stable transgenic lines. To evaluate whether other endogenous promoters might trigger high levels of overexpression, we analyzed RNA-seq data from our previous study [4]. We identified at least two rice genes with log2[TPM] values exceeding 13 in leaves (Fig. S6): OsGR-RBP4 [8] and OsLHCB2.1 [9]. Notably, stress-inducible genes, such as OsSalT (Os01g0348900) [10] and OCPI2 (Os01g0615100) [11], also surpassed this threshold in protoplasts, implying that implementing TRIBE using these promoters under stress conditions requires careful evaluation. We therefore conclude that ADARdd expression driven by the UBIQUITIN10 promoter, or potentially other rice promoters, results in negligible off-target editing events and that the few such off-targeting edits can be filtered out via a stringent bioinformatic pipeline using multiple biological replicates.

In conclusion, unlike the global off-target effect reported for ADARdd in animal cells [7,12,13], this study, together with our previous report [4], demonstrates that promiscuous off-target editing by ADARdd requires abnormally high expression levels, a condition uniquely observed in the protoplast transient expression system. This elevated expression likely results from the high plasmid copy number introduced during transfection. Since off-target editing depends on the hyperaccumulation of ADARdd, the precise regulation of RBP-ADARdd expression in stable transgenic plants effectively minimizes unintended editing during targeted RNA editing. Our findings further emphasize that protoplast-based assessments of proximity-directed RNA editors require rigorous validation.

1. Materials and methods

1.1. Plasmid construction

Plasmids expressing ADARdd and MCP-ADARdd in rice protoplasts were constructed as previously described [4].

1.2. Protoplast transfection and rice transformation

Rice protoplast preparation and transfection were performed as previously reported [14,15]. Briefly, 20 μg of plasmid DNA was used to transfect approximately 1 × 107 rice protoplasts, except for ADARdd-OE protoplasts, which were transfected with 5 μg of plasmid DNA. The transfected protoplasts were incubated in the dark for 24–36 h before harvesting. Rice (Oryza sativa L. ssp. japonica) cultivar Nipponbare was used in this study.

1.3. Genomic DNA extraction, RNA purification, and RT-PCR

Genomic DNA was isolated from rice protoplasts and leaves via the cetyltrimethylammonium bromide method [16]. RNA was extracted using TRIzol reagent (Life Technologies) following the manufacturer's instructions.

For RT-PCR, 1 μg of total RNA was treated with DNase I (New England Biolabs) and reverse transcribed using M-MLV (Takara Bio) with oligo (dT)18 primers or sequence-specific primers. PCR was performed using Q5 High-Fidelity DNA Polymerase (New England Biolabs, Ipswich, MA, USA) for Sanger sequencing (see Table S6 for primer sequences).

1.4. Quantification of editing efficiency via Sanger sequencing

After the protoplasts were transfected with plasmids, the RNA targets were amplified using RT-PCR, and the PCR products were analyzed using Sanger sequencing. A-to-I RNA editing was calculated based on the T/C peak heights in the sequencing chromatogram as described previously [17].

1.5. Electrophoretic mobility shift assay

The RNA-EMSA experiment was performed as we described previously [18]. For recombinant protein expression, the coding sequence of OsDRB1 was amplified via PCR and inserted into the pDEST15 vector using a ClonExpress II One Step Cloning Kit (Vazyme). GST-tagged OsDRB1 was expressed in the E. coli strain BL21 (DE3) pLysS and purified using glutathione Sepharose 4B (Amersham Biosciences) according to the manufacturer's instructions. The oligonucleotide sequences used in this study are listed in Table S6. The 3 UTR of Os04g0380300 RNA was amplified using primers with the T7 promoter sequence, followed by in vitro transcription using a HiScribe Quick T7 High Yield RNA Synthesis Kit (New England Biolabs, USA) according to the manufacturer's instructions. Briefly, 1 μg of the 3 UTR of Os04g0380300 template DNA, NTP buffer mix, and T7 RNA polymerase were combined and incubated overnight at 37 °C. The synthesized RNA was purified using an RNA Clean & Concentrator Kit (Zymo Research, USA). To generate the Cy5-labeled 3 UTR of Os04g0380300, 2 μg of Os04g0380300 3 UTR RNA was preheated at 85 °C for 5 min and then placed immediately on ice. The RNA was mixed with 1 × T4 RNA ligase buffer, 40 units of T4 RNA Ligase 1 (NEB), 40 units of RNase inhibitor, 1 mM ATP, 10 % DMSO, 33.3 μM pCp-Cy5 (Jera Bioscience), and 15 % PEG8000 and incubated overnight at 16 °C. The labeled RNA was purified and eluted with 20 μL of water, heated at 85 °C for 5 min, and annealed by gradual cooling to 25 °C before use. The Cy5-labeled and unlabeled 3 UTRs were incubated with GST-OsDRB1 protein in 20 μL of binding buffer (20 mM HEPES [pH 7.5], 10 mM ZnCl2, 20 mM KCl, 1 mM MgCl2, 5 mM DTT, 10 % glycerol, 0.1 % NP-40) at room temperature for 1 h in the dark. The resulting protein‒substrate complexes were resolved on 6 % nondenaturing polyacrylamide gels at 60 V for 2 h in 0.5 × TBE buffer. After electrophoresis, the Cy5 signals were detected using a FUJIFILM FLA-9000 imager.

1.6. RNA-seq

Total RNA extracted from leaves and protoplasts was used for RNA sequencing. Three biological replicates were performed for each experimental group. RNA-seq libraries were prepared using a NEBNext Ultra RNA Library Prep Kit for Illumina, and 2 × 150 bp paired-end sequencing was performed on a NovaSeq 6000 sequencer (Illumina, San Diego, CA, USA).

1.7. RNA-seq data analysis

The rice reference genome sequence and annotation (IRGSP 1.0) were downloaded from Ensembl Plants (http://plants.ensembl.org/).

The RNA-seq data were analyzed using a bioinformatic pipeline as previously described with some modifications [4]. In brief, adaptor sequences were trimmed, and low-quality reads were removed via TrimGalore (v0.6.10) with the following parameters: j 4 -q 30 --phred33 --stringency 3 --length 150 -e 0.1 --paired. The clean reads were aligned to the rice reference genome using STAR (v2.5.3a) [19]with the following settings: -outFilterMismatchNoverLmax 0.07 --outSAMstrandField intronMotif --outFilterMultimapNmax 1 --outSAMtype BAM SortedByCoordinate. The PCR duplicates were removed using Picard (v2.23.1). Gene expression profiles were generated using featureCounts (v2.0.1) [20] and converted to transcripts per million (TPM) values. The RDVs were extracted from the alignment results using SAMtools (v1.7.1) and BCFtools (v1.15.1). To identify off-target editing sites, the RDVs were filtered using in-house scripts (Fig. 1G) [1]: Genomic SNPs detected in the negative controls (ADARdd-OE-leaf and ADARdd-OE ​+ ​EV samples) were removed [2]; low-depth RDVs (reads <10) were removed. All remaining A-to-G and T-to-C RDVs were regarded as off-target edits (OFTEs). The percentage of each editing site was determined using the formula ReadG/(ReadA ​+ ​ReadG), where ReadA and ReadG represent the number of RNA-seq reads containing A and G that cover the editing site, respectively. All computer code used in the RNA-seq analysis is available on GitHub (https://github.com/YS-HZAU/protoplast-TRIBE).

CRediT authorship contribution statement

Shuai Yin: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Shumin Li: Validation, Resources. Yang Yuan: Resources, Funding acquisition. Kabin Xie: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This study was supported by the Science and Technology Innovation 2030-Major Projects (2023ZD0407403), the National Natural Science Foundation of China (32293243), the Fundamental Research Funds for the Central Universities (2662025ZKPY008), and the Earmarked Fund for the China Agriculture Research System (CARS-01).

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.abiote.2025.100007.

Appendix A. Supplementary data

The following is/are the supplementary data to this article:

Multimedia component 1
mmc1.xlsx (6.8MB, xlsx)
Multimedia component 2
mmc2.docx (1.2MB, docx)

Data availability

The RNA-seq data described in this study have been deposited in the BIG Data Center (https://bigd.big.ac.cn/) under accession numbers CRA023046for the protoplast data and CRA005307 for the leaf data (under BioProject PRJCA007082). The computer code used for data analysis is available on GitHub (https://github.com/YS-HZAU/protoplast-TRIBE). All other data underlying this study are included in the article and the online supplementary materials.

References

  • 1.Pfeiffer L.S., Stafforst T. Precision RNA base editing with engineered and endogenous effectors. Nat Biotechnol. 2023;41(11):1526–1542. doi: 10.1038/s41587-023-01927-0. [DOI] [PubMed] [Google Scholar]
  • 2.McMahon A.C., Rahman R., Jin H., Shen J.L., Fieldsend A., Luo W., et al. TRIBE: hijacking an RNA-editing enzyme to identify cell-specific targets of RNA-binding proteins. Cell. 2016;165(3):742–753. doi: 10.1016/j.cell.2016.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Sekar V., Mármol-Sánchez E., Kalogeropoulos P., Stanicek L., Sagredo E.A., Widmark A., et al. Detection of transcriptome-wide microRNA-target interactions in single cells with agoTRIBE. Nat Biotechnol. 2024;42(8):1296–1302. doi: 10.1038/s41587-023-01951-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Yin S., Chen Y., Chen Y., Xiong L., Xie K. Genome-wide profiling of rice double-stranded RNA-binding Protein 1-associated RNAs by targeted RNA editing. Plant Physiol. 2023;192(2):805–820. doi: 10.1093/plphys/kiad158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Arribas-Hernandez L., Rennie S., Koster T., Porcelli C., Lewinski M., Staiger D., et al. Principles of mRNA targeting via the Arabidopsis m6A-binding protein ECT2. eLife. 2021;10 doi: 10.7554/eLife.72375. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zhou G., Niu R., Zhou Y., Luo M., Peng Y., Wang H., et al. Proximity editing to identify RNAs in phase-separated RNA binding protein condensates. Cell Discov. 2021;7(1):72. doi: 10.1038/s41421-021-00288-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Vallecillo-Viejo I.C., Liscovitch-Brauer N., Montiel-Gonzalez M.F., Eisenberg E., Rosenthal J.J.C. Abundant off-target edits from site-directed RNA editing can be reduced by nuclear localization of the editing enzyme. RNA Biol. 2018;15(1):104–114. doi: 10.1080/15476286.2017.1387711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Sahi C., Agarwal M., Singh A., Grover A. Molecular characterization of a novel isoform of rice (Oryza sativa L.) glycine rich-RNA binding protein and evidence for its involvement in high temperature stress response. Plant Sci. 2007;173(2):144–155. [Google Scholar]
  • 9.Umate P. Genome-wide analysis of the family of light-harvesting chlorophyll a/b-binding proteins in Arabidopsis and rice. Plant Signal Behav. 2010;5(12):1537–1542. doi: 10.4161/psb.5.12.13410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Claes B., Dekeyser R., Villarroel R., Van den Bulcke M., Bauw G., Van Montagu M., et al. Characterization of a rice gene showing organ-specific expression in response to salt stress and drought. Plant Cell. 1990;2(1):19–27. doi: 10.1105/tpc.2.1.19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Singh A., Sahi C., Grover A. Chymotrypsin protease inhibitor gene family in rice: genomic organization and evidence for the presence of a bidirectional promoter shared between two chymotrypsin protease inhibitor genes. Gene. 2009;428(1–2):9–19. doi: 10.1016/j.gene.2008.09.028. [DOI] [PubMed] [Google Scholar]
  • 12.Buchumenski I., Roth S.H., Kopel E., Katsman E., Feiglin A., Levanon E.Y., et al. Global quantification exposes abundant low-level off-target activity by base editors. Genome Res. 2021;31(12):2354–2361. doi: 10.1101/gr.275770.121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Biswas J., Rahman R., Gupta V., Rosbash M., Singer R.H. MS2-TRIBE evaluates both protein-RNA interactions and nuclear organization of transcription by RNA editing. iScience. 2020;23(7) doi: 10.1016/j.isci.2020.101318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Xie K., Yang Y. RNA-guided genome editing in plants using a CRISPR-Cas system. Mol Plant. 2013;6(6):1975–1983. doi: 10.1093/mp/sst119. [DOI] [PubMed] [Google Scholar]
  • 15.Panda D., Karmakar S., Dash M., Tripathy S.K., Das P., Banerjee S., et al. Optimized protoplast isolation and transfection with a breakpoint: accelerating Cas9/sgRNA cleavage efficiency validation in monocot and dicot. aBIOTECH. 2024;5(2):151–168. doi: 10.1007/s42994-024-00139-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Zhang Y., Cui M.M., Ke R.N., Chen Y.D., Xie K. C-terminal frameshift mutations generate viable knockout mutants with developmental defects for three essential protein kinases. aBIOTECH. 2024;5(2):219–224. doi: 10.1007/s42994-024-00165-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Rinkevich F.D., Schweitzer P.A., Scott J.G. Antisense sequencing improves the accuracy and precision of A-to-I editing measurements using the peak height ratio method. BMC Res Notes. 2012;5:63. doi: 10.1186/1756-0500-5-63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Chen Y., Yin S., Yang Y., Xie K. The dsRNA-binding protein OsDRB1.4 is phosphorylated by OsMPK5 and negatively regulates rice defense against Magnaporthe oryzae. Plant J. 2025;122(6) doi: 10.1111/tpj.70285. [DOI] [PubMed] [Google Scholar]
  • 19.Dobin A., Davis C.A., Schlesinger F., Drenkow J., Zaleski C., Jha S., et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21. doi: 10.1093/bioinformatics/bts635. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Liao Y., Smyth G.K., Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–930. doi: 10.1093/bioinformatics/btt656. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.xlsx (6.8MB, xlsx)
Multimedia component 2
mmc2.docx (1.2MB, docx)

Data Availability Statement

The RNA-seq data described in this study have been deposited in the BIG Data Center (https://bigd.big.ac.cn/) under accession numbers CRA023046for the protoplast data and CRA005307 for the leaf data (under BioProject PRJCA007082). The computer code used for data analysis is available on GitHub (https://github.com/YS-HZAU/protoplast-TRIBE). All other data underlying this study are included in the article and the online supplementary materials.

Articles from aBIOTECH are provided here courtesy of Elsevier

RESOURCES