Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2023 Mar 13;192(2):805–820. doi: 10.1093/plphys/kiad158

Genome-wide profiling of rice Double-stranded RNA-Binding Protein 1–associated RNAs by targeted RNA editing

Shuai Yin 1,2,3,4, Yuedan Chen 5,6, Yache Chen 7,8, Lizhong Xiong 9, Kabin Xie 10,11,12,13,✉,b,c
PMCID: PMC10231461  PMID: 36913253

Abstract

RNA-binding proteins (RBPs) play essential roles in regulating gene expression. However, the RNA ligands of RBPs are poorly understood in plants, not least due to the lack of efficient tools for genome-wide identification of RBP-bound RNAs. An RBP-fused adenosine deaminase acting on RNA (ADAR) can edit RBP-bound RNAs, which allows efficient identification of RNA ligands of RBPs in vivo. Here, we report the RNA editing activities of the ADAR deaminase domain (ADARdd) in plants. Protoplast experiments indicated that RBP-ADARdd fusions efficiently edited adenosines within 41 nucleotides (nt) of their binding sites. We then engineered ADARdd to profile the RNA ligands of rice (Oryza sativa) Double-stranded RNA-Binding Protein 1 (OsDRB1). Overexpressing the OsDRB1-ADARdd fusion protein in rice introduced thousands of A-to-G and T-to-C RNA‒DNA variants (RDVs). We developed a stringent bioinformatic approach to identify A-to-I RNA edits from RDVs, which removed 99.7% to 100% of background single-nucleotide variants in RNA-seq data. This pipeline identified a total of 1,798 high-confidence RNA editing (HiCE) sites, which marked 799 transcripts as OsDRB1-binding RNAs, from the leaf and root samples of OsDRB1-ADARdd–overexpressing plants. These HiCE sites were predominantly located in repetitive elements, 3′-UTRs, and introns. Small RNA sequencing also identified 191 A-to-I RNA edits in miRNAs and other sRNAs, confirming that OsDRB1 is involved in sRNA biogenesis or function. Our study presents a valuable tool for genome-wide profiling of RNA ligands of RBPs in plants and provides a global view of OsDRB1-binding RNAs.

Targeted RNA editing identifies rice DOUBLE-STRANDED RNA BINDING PROTEIN 1-associated RNAs in vivo with high efficiency and accuracy.

Introduction

RNA is a vital component in translating genetic information from DNA to protein and regulates gene activity. As faithful companions of RNA, RNA-binding proteins (RBPs) are required during the whole life cycle of RNAs. Increasing evidence shows that RBPs act as essential regulators to reprogram gene expression during plant growth and stress responses (Cho et al. 2019). RBP–RNA interactions are highly dynamic, and various RBPs are involved in different steps of RNA metabolism, such as RNA processing, localization, turnover, and function (Hentze et al. 2018). In plants, RBPs constitute a large gene family and have diverse RNA targets. In each of two model plant organisms, Arabidopsis (Arabidopsis thaliana) and rice (Oryza sativa), over 200 RBPs were identified from the mRNA-binding proteome (Doroshenk et al. 2012; Reichel et al. 2016). RBPs often contain one or multiple modular RNA-binding domains to recognize RNA sequences, RNA modifications, and/or secondary structures (Lunde et al. 2007). Identification of RNA targets of RBPs is critical to elucidate the function of RBPs during gene expression. However, our knowledge of plant RBP–RNA interaction networks is quite limited, not least due to technical challenges for genome-wide identification of RNA ligands of RBPs in vivo.

RNA immunoprecipitation sequencing (RIP-seq) is the classic method for discovering RNA transcripts associated with RBPs. However, RIP-seq often has a high background. To increase RIP-seq accuracy and specificity, ultraviolet (UV) crosslinking immunoprecipitation sequencing (CLIP-seq) was developed in animal systems to investigate RBP–RNA interactions (Ule et al. 2003). The CLIP-seq method was further optimized as photoactivatable ribonucleoside-enhanced (PAR)-CLIP (Hafner et al. 2010), individual-nucleotide resolution crosslinking immunoprecipitation (iCLIP) (König et al. 2010), and enhanced CLIP (eCLIP) (Van Nostrand et al. 2016). iCLIP-seq has been used to identify RNA ligands of Arabidopsis thaliana glycine-rich RNA-binding protein 7 (AtGRP7, AT2G21660) (Meyer et al. 2017). On the other hand, UV-mediated crosslinking was also used to capture RBPs that were associated with poly(A) RNAs in Arabidopsis (Reichel et al. 2016; Bach-Pages et al. 2020). Successful RIP-seq and CLIP-seq depend on high-quality antibodies for immunoprecipitation and efficient crosslinking conditions, respectively. Darnell (2010) estimated that UV-mediated crosslinking of RBP–RNA occurs on only a minority of contact sites (1% to 5%). In plants, UV irradiation can be absorbed by secondary metabolites (e.g. wax, flavonoids, phenols, nucleic acids, and cell membranes) and pigments, which impairs the RBP–RNA crosslinking efficiencies of UV irradiation (Kostyuk et al. 2008; Koster et al. 2020). In addition, double-stranded RNA (dsRNA) is known to be poorly crosslinked in animal systems (Kim and Kim 2019). Chemical reagents such as formaldehyde crosslink not only RNAs with RBPs but also DNA and proteins, severely compromising the sequencing analysis. Therefore, different methods are needed to study RBP–RNA interactions in plants.

A method named targets of RNA-binding proteins identified by editing (TRIBE) was developed to identify RBP-associated RNAs with high accuracy and sensitivity in animals (McMahon et al. 2016). The TRIBE method reprograms the adenosine-to-inosine (A-to-I) RNA editing enzyme adenosine deaminase acting on RNA (ADAR) to label the RNAs associated with the RBP of interest. Fusing RBP with the ADAR deaminase domain (RBP-ADARdd) can mark its in vivo RNA ligands by A-to-I editing. The edited RNAs could be easily identified as A-to-G conversions using Sanger sequencing and RNA-seq (Fig. 1A). Owing to highly active RNA editing enzymes and the rapid progress of sequencing technology, TRIBE allows fast and efficient characterization of in vivo RBP-targeted RNAs. The TRIBE method was recently validated in Arabidopsis (Arribas-Hernandez et al. 2021; Zhou et al. 2021). However, its editing window, activity, sequence bias, and off-target risk were not evaluated in plants.

Figure 1.

Figure 1.

Open in a new tab

Targeted RNA editing using RBP-ADARdd in rice and Arabidopsis protoplasts. A) Schematic illustration of engineering ADARs to edit RNAs associated with RBPs of interest. The ADAR deaminase domain (ADARdd) is fused with RBP and expressed in plants or protoplasts. The deaminase activity of RBP-ADARdd marks its RNA ligands by converting adenosines to inosines (A-to-I), which are read as A-to-G or T-to-C RDVs in Sanger sequencing and RNA-seq. B–G) Characterization of ADARdd editing activity using MS2-RNA and its binding protein MCP in rice protoplasts. B) Structure of plasmid constructs expressing MS2-RNA with FLAG-tagged MCP, ADARdd, or MCP-ADARdd. U3p, rice snoRNA U3 promoter; UBI10p, rice UBIQUITIN10 promoter; Linker, 3×EAAAK. C) Western blotting showing the expression of MCP-ADARdd, MCP, and ADARdd in rice protoplasts. Coomassie Brilliant Blue (CBB)–stained total protein was used as a loading control. D and F) Six adenosines in MS2-RNA were edited by MCP-ADARdd. The reverse primer (R) used for Sanger sequencing locates at the 3'-end; the 19-nt MS2 stem-loop is colored purple; and the adenosines (bold font) edited by MCP-ADARdd are colored red. E and G) Comparison of editing percentages of 6 sites and their distance to MCP binding region in MS2-RNA. Center line, median; box limits, upper and lower quantiles; whiskers, 1.5 times interquartile range; each point indicates a replicate. The letters in E indicate statistical significance of one-way ANOVA followed by post hoc Tukey's least significant difference (LSD), α = 0.05. Data are presented as the mean ± standard deviation (SD) (n = 10 biological replicates). H–I) Expression of AtGRP7-ADARdd, AtGRP7, and ADARdd with the 35S promoter in Arabidopsis protoplasts. CBB, Coomassie Brilliant Blue–stained total protein. J) Schematic illustration of the A-to-G editing site and percentages in AtGRP7-ADARdd RNA. See also Supplemental Fig. S4 for sequencing chromatograms.

Rice is one of the most important food crops in the world. A few rice RBPs have been characterized as important regulators of agronomic traits, including disease resistance (Zhou et al. 2018), abiotic stress tolerance (Shim et al. 2021), and yield (Tian et al. 2019). Uncovering the regulatory role of RBPs in cellular signaling pathways requires convenient and efficient methods to capture their in vivo RNA ligands. In this study, we adopted TRIBE in rice to profile RNAs associated with RBPs of interest. We evaluated the RNA editing efficiency and the editing window of ADARdd using rice protoplasts. Then, ADARdd was used to profile RNAs associated with rice Double-stranded RNA-Binding Protein 1 (OsDRB1,Os11g0119900). OsDRB1 is one of the rice homologs of Arabidopsis DRB1/Hyponastic Leaves 1 (HYL1), which interacts with DICER-like 1 (DCL1) to process small RNA precursors (Kurihara et al. 2006). However, the function of OsDRB1, particularly its RNA ligands, remains unknown. Using RNA-seq and small RNA-seq (sRNA-seq), we found that OsDRB1-ADARdd generated enormous A-to-I RNA edits in stable transgenic lines. We established a stringent bioinformatic pipeline to identify RNA editing sites with high confidence from RNA-seq data and provide a genome-wide view of OsDRB1-binding RNAs. Overall, this study presents a useful RNA editing toolkit for profiling RNA ligands of RBPs in plants.

Results

ADAR fused with RBP efficiently edits RNA ligands in rice and Arabidopsis protoplasts

We first tested the editing window and efficiency of ADARdd in rice using bacteriophage MS2 coat protein (MCP) and its RNA ligand (a stem-loop sequence from MS2). MCP and MS2 stem-loop RNA have been used in many RNA tools (Tutucci et al. 2018; Yang et al. 2019) and are ideal for evaluating RNA editing enzymes in vivo. To this end, codon-optimized ADARdd and MCP were synthesized and fused with triple repeats of linker sequence (3×EAAAK), an SV40 nuclear localization signal, and a FLAG tag (see Supplemental Figs. S1 and S2 for ADARdd and MCP sequences). Of note, the ADARdd used in this study contains an E488Q substitution, which increases its RNA editing activity (Hyper ADAR) and specificity (Xu et al. 2018). An artificial noncoding gene of MS2-RNA, which is 322 bp long and contains tandem repeated MS2 stem-loop sequences in the center, was designed as the target of MCP-ADARdd (see Supplemental Fig. S3 for the MS2-RNA sequence). MS2-RNA and MCP-ADARdd were expressed under the rice snoRNA U3 promoter (OsU3p) and UBIQUITIN 10 promoter (UBI10p) in protoplasts (Fig. 1B), respectively. In this experiment, protoplasts expressing MS2-RNA with ADARdd or MCP were included as negative controls (Fig. 1B). Western blotting showed that MCP-ADARdd, MCP, and ARARdd proteins were expressed properly in protoplasts (Fig. 1C). The A-to-I RNA editing in MS2-RNA transcripts was analyzed using reverse transcription polymerase chain reaction (RT-PCR) followed by Sanger sequencing. In protoplasts expressing MS2-RNA and MCP-ADARdd, six A/G and T/C double peaks were found in the MS2-RNA transcript using sense and antisense primer sequencing, respectively. Interestingly, sense primer sequencing chromatograms (A/G double peaks) generated higher editing percentages than antisense primers (T/C double peaks) using the same RT-PCR products (Supplemental Fig. S4A). This phenomenon was previously observed in animal RNA editing, and the ratios of T/C peaks are more accurate and precise than the ratios of A/G peaks for A-to-I RNA editing measurement (Eggington et al. 2011; Rinkevich et al. 2012). Thus, this study used antisense primer sequencing chromatograms to calculate A-to-I editing percentages.

In 10 independent experiments, these 6 A-to-I edits in the MS2-RNA transcript were repeatedly detected in protoplasts expressing the MCP-ADARdd fusion (Fig. 1D). In contrast, no mutation was detected in negative control samples that expressed MCP or ADARdd alone. The editing percentages of these sites ranged from 3.2% to 80% (Fig. 1E), and all editing sites were located within the 4 to 41 nt MCP binding sites, according to the predicted secondary structure of the MS2-RNA transcript (Fig. 1F). We found that the editing percentages were affected by the distance between the MCP binding sites and the edited adenosines (Fig. 1E). The highest editing percentages were observed at A82 and A236 of MS2-RNA, which are located at 20 and 27 nt, ∼20 bp in the stem-loop RNA structure, to the MCP binding site (Fig. 1, F and G). Thus, MCP-ADARdd is efficient for editing adenosines within 41 nt, and its optimal editing window is located at around 20 bp (base pair in stem-loop RNA) from its binding site.

Because the original ADAR prefers dsRNA substrates (Phelps et al. 2015) and the MS2-RNA forms a stable dsRNA structure (Fig. 1F), we asked whether RBP-ADARdd fusions efficiently edit single-stranded RNA targets. To this end, Arabidopsis thaliana Glycine-Rich RNA-binding Protein 7 (AtGRP7, AT2G21660), which binds to its own mRNA (Staiger et al. 2003), was fused with ADARdd. AtGRP7, ADARdd, and AtGRP7-ADARdd fusion were transiently expressed with the cauliflower mosaic virus 35S promoter (35S) in Arabidopsis protoplasts (Fig. 1, H and I), and RNA editing of cognate mRNAs was analyzed. A total of 20 A-to-G edits were identified in the AtGRP7-ADARdd transcript (Fig. 1J). Again, we did not observe any editing sites in protoplasts expressing only AtGRP7 or ADARdd (Supplemental Fig. S4B). Similar to MCP-ADARdd, AtGRP7-ADARdd displayed editing percentages ranging from 2% to 49%. Together, these data indicate that ADARdd-mediated RNA editing is compatible with different RBPs.

Rice plants overexpressing OsDRB1-ADARdd introduce thousands of RNA edits

We then used this system to identify the RNAs associated with rice Double-stranded RNA-Binding Protein 1 (OsDRB1, Os11g0119900) in rice plants. To this end, OsDRB1, ADARdd, and OsDRB1-ADARdd fusion proteins were overexpressed under the Ubiquitin 10 promoter in the rice cultivar Nipponbare (Fig. 2A; referred to as OsDRB1-OE, ADARdd-OE, OsDRB1-ADARdd-OE hereafter). We obtained 20 to 30 transgenic plants per construct, and all plants showed normal growth like wild-type (WT) plants (Supplemental Fig. S5A). Two independent T0 lines with comparable mRNA levels of the transgenes were selected as two biological replicates (rep1 and rep2) for further analysis (Supplemental Fig. S5, B and C). Then, standard RNA-seq was used to analyze the poly(A)+ RNAs of leaves and roots of OsDRB1-ADARdd-OE, OsDRB1-OE, and ADARdd-OE plants. For each sample, ∼20 million clean reads were obtained, and 89% to 96% of them were uniquely mapped to the rice reference genome (Supplemental Fig. S5D; Supplemental Table S1). These RNA-seq reads detected 61% (23790/38978) of annotated rice genes (TPM > 1; Supplemental Table S2), and their expression patterns were highly correlated in these samples (Pearson's correlation coefficient, r > 0.93; Supplemental Fig. S6).

Figure 2.

Figure 2.

Open in a new tab

Identification of OsDRB1-ADARdd editing sites using RNA-seq. A) Plasmid constructs overexpressing OsDRB1-ADARdd, OsDRB1, and ADARdd with a FLAG tag. UBI10p, rice UBIQUITIN10 promoter; Linker, 3×EAAAK. B–C) Number of 12 types of RDVs in OsDRB1-ADARdd-OE, ADARdd-OE, and OsDRB1-OE samples (leaf and root). Rep 1/2, biological replicates. Data are presented as the mean ± SD, and the P values are shown in the plots (Student's t-test; for A > G and T > C, n = 4; other 10 types of RDVs, n = 20). D) Flowchart of the RDV filtering pipeline for HiCE identification. E) Comparisons of RDVs removed in each filtering step. The percentages below the plot are presented as the mean ± SD using leaf and root data sets. For negative controls (ADARdd-OE and OsDRB1-OE), the percentages of all 12 types of RDVs were combined. For OsDRB1-ADARdd-OE, potential RNA editing (A-to-G and T-to-C) and non–RNA editing RDVs are shown separately. See Supplemental Table S3 for the number of each type of RDV that was filtered in each step.

We then analyzed the RNA-DNA variants (RDVs) between RNA-seq reads and reference genome sequences. RNA-seq detected 7,642–9,789 RDVs in leaf and root samples of OsDRB1-ADARdd-OE, which were 1.9 to 3.4 times higher than the RDVs of OsDRB1-OE and ADARdd-OE (Fig. 2B, Supplemental Table S3). We compared the amount of 12 types of RDVs in different samples. In comparison with two negative controls (OsDRB1-OE and ADARdd-OE), the number of A-to-G and T-to-C RDVs, which can be introduced by A-to-I RNA editing, was increased at least 4-fold in OsDRB1-ADARdd-OE data sets, while the amount of the other 10 types of non–RNA editing RDVs showed no significant difference between these transformants (Fig. 2, B and C; Supplemental Table S3). Furthermore, the number of 12 types of RDVs displayed no significant differences between OsDRB1-OE and ADARdd-OE, implying that ADARdd alone did not edit endogenous RNAs in rice and that these RDVs resulted from either background mutations or sequencing artifacts (Fig. 2, B and C; Supplemental Table S3). In summary, we identified thousands of RNA editing events introduced by OsDRB1-ADARdd in rice plants.

Identification of highly confident RNA editing events of OsDRB1-ADARdd

To identify the high-confidence editing (HiCE) sites from these RDVs, we established a stringent bioinformatic pipeline to filter sequence variants such as genomic single-nucleotide polymorphisms (SNPs), somatic mutations, PCR/sequencing artifacts, and nonspecific transient RNA edits (Fig. 2D). First, sequence variants that occurred in more than 2 negative control samples were removed as genomic SNPs between the reference genome and the genome of Nipponbare stocks used in this study. The remaining A-to-G and T-to-C RDVs were considered as potential RNA edits. Second, RDVs with low coverage (depth < 10) were removed as RDVs with low sequencing reads are difficult to discriminate from PCR/sequencing artifacts (Sims et al. 2014). Third, to exclude false-positive results of transient editing events and random somatic mutations, the remaining RDVs that were detected in only one replicate were removed. The transient edits are caused by the proximal deaminase activity of RBP-ADARdd (Biswas et al. 2020). After these three steps, the remaining A-to-G and T-to-C RDVs were selected as HiCE sites.

To assess the accuracy of this HiCE extraction pipeline, OsDRB1-OE and ADARdd-OE RDVs were analyzed using these filtering steps. On average, three filtering steps removed 90.9%, 8.4%, and 0.7% of RDVs of negative control samples (Fig. 2E). Remarkably, our pipeline recognized 100% of RDVs in negative control data sets, indicating its power for RNA editing event identification. Sanger sequencing was used to verify several A-to-G and T-to-C RDVs that were recognized as genomic SNPs in our pipeline (see example in Supplemental Fig. S7). Of note, ADARdd-OE samples display RDV spectra similar to those of OsDRB1-OE (Fig. 2, B and E), which again implies that ADARdd alone likely has no off-target editing activity in rice. Together, these data suggest that our bioinformatic pipeline could efficiently remove background RDVs (Fig. 2E; Supplemental Table S3).

Then, this pipeline was used to extract HiCE sites from RDVs of OsDRB-ADARdd-OE data sets. We obtained 1,316 and 887 A-to-G and T-to-C RDVs as HiCE sites from leaf and root samples, respectively. Of note, most A-to-G and T-to-C RDVs were removed due to low read depth (51%) and replicate specificity (15%), which is distinct from negative control data sets in which >90% of RDVs are genomic SNPs. It is possible that some true RNA edits might have been removed due to low sequencing depth. For the other 10 types of non–RNA editing RDVs, our pipeline successfully removed 99.7% of them (Fig. 2E; Supplemental Table S3), but 16 false positives remained in OsDRB1-ADARdd-OE root data sets, including 6 C-to-T, 3 T-to-A, and 7 G-to-A/C/T sites (Supplemental Table S4). We amplified and sequenced the genomic DNA and mRNA fragments of these false positives using different transformants and WT siblings. The results indicate that these RDVs were heterozygous SNPs and segregated between different transformants (Supplemental Fig. S8). According to non–RNA edits of RDVs in OsDRB1-ADARdd-OE, we estimated that the rate of false positives was less than 1.8% (16/887) in root-HiCE data and could be recognized by sequencing the genomic DNA fragments. After removing genomic SNPs, low-depth RDVs, and replicate-specific RDVs, ∼19% of A-to-G and T-to-C RDVs in OsDRB1-ADARdd-OE were identified as HiCE sites (Fig. 2E; Supplemental Table S5).

Features of OsDRB1-ADARdd HiCE sites

We then analyzed these HiCE sites of OsDRB1-ADARdd. After merging the leaf and root data, we obtained 1,798 HiCE sites, including 911 (50.7%) and 482 (26.8%) HiCEs that were detected only in leaves and roots, respectively. These data suggest that OsDRB1 interacts with most RNAs in a tissue-specific manner (Fig. 3A; Supplemental Table S5). The editing percentages of HiCE sites ranged from 23.1% to 100% with an average value of 62% (Fig. 3, B and C). Furthermore, the editing percentages of each site were consistent between the two biological replicates (Pearson's correlation coefficient, leaf, r = 0.73; root, r = 0.78). The HiCE sites were distributed on all 12 rice chromosomes and enriched at several loci (Fig. 3D).

Figure 3.

Figure 3.

Open in a new tab

Features of HiCE sites from the OsDRB1-ADARdd transcriptome. A) Venn diagram showing the HiCE sites from OsDRB1-ADARdd-OE leaf and root data sets. B–C) Editing percentages of HiCE sites of two biological replicates. Pearson's correlation coefficients are shown in the plots. D) Genome-wide distribution of leaf- and root-HiCEs in 12 rice chromosomes. The green (Leaf) and brown (Root) curves indicate the density of HiCE sites in chromosomes (window size: 1 × 107 bp). E) Nucleotide frequencies of sequences flanking HiCE sites. The size of the letter indicates the percentages of each nucleotide (A/U/G/C) at 1∼5 nt to the HiCE sites (EDIT).

We noticed that not all adenosines proximal to the MCP-ADARdd binding site were edited in our protoplast experiment (Fig. 1F), raising the question of whether ADARdd prefers a specific sequence motif. We calculated the nucleotide frequencies in ±5 nt sequences flanking 1,798 HiCE sites. Indeed, the 5′ nearest nucleotide (−1 nt position) of the HiCE site shows strong bias as U (44.1%) > A (33.7%) > C (19.2%) >G (3.1%) (Fig. 3E). Additionally, the frequencies of the 3′ nearest nucleotides (+1 nt position) of the HiCEs were G (37.4%) > A (27.2%) > U (21.1%) > C (14.5%) (Fig. 3E), implying a weak bias at the +1 nt position. These data indicate that ADARdd less favors editing A after G. A similar bias was also reported in Drosophila (McMahon et al. 2016). Nevertheless, this editing bias at −1 nt position would not affect the identification of RBP-bound RNAs as ADARdd edited multiple sites in an RNA.

OsDRB1 dominantly binds with noncoding sequences

Then, we analyzed the HiCE-marked transcript units (TUs), which are considered as OsDRB1-binding RNAs. According to rice genome annotation, 75 leaf-HiCE and 69 root-HiCE were located in unannotated regions, with the remaining HiCEs corresponding to 799 annotated TUs, including 305 coding genes, 16 noncoding genes, 25 miRNA genes (MIR), and 453 repetitive elements (Supplemental Table S5). Among repetitive elements, 74.9% (674) and 71.4% (810) of leaf- and root-HiCEs, respectively, were located in TUs from transposons such as MITE, LTR, MuDR, SINE, hAT, En/Spm, and TYPEU (Fig. 4, A and B; Supplemental Table S5). These transposon TUs have 2 to 6 HiCE sites on average (Fig. 4, B and C), consistent with OsDRB1's role in small RNA biogenesis since many sRNAs are processed from transcripts of repetitive elements (Borges and Martienssen 2015). In HiCE-marked protein-coding TUs, ∼50% of them contain only 1 HiCE site (Fig. 4D) dominantly located in noncoding regions. For example, 44.69% (leaf, 37.92%; root, 48.02%) and 46.85% (leaf, 45.85%; root, 38.99%) of HiCE was located in the 3′-UTR and introns, respectively (Fig. 4D), also consistent with dsRNA structures, and sRNAs mostly are formed from noncoding regions of mRNAs (Reich and Bass 2019). Together, these data suggest that OsDRB1 prevalently binds to noncoding sequences of RNAs.

Figure 4.

Figure 4.

Open in a new tab

Annotation of OsDRB1-associated RNAs with HiCE sites. A–C) Annotation of OsDRB1-associated transcripts from repetitive elements. A) Sankey plot showing HiCE-marked TUs belonging to different repetitive elements. B) Bar plot showing the number of TUs. C) Scatter plot showing the average number of HiCE sites in TUs from repetitive elements. D) HiCE-marked protein-coding genes in leaf and root data sets. The bar plot shows the number of TUs with different HiCE sites. The pie charts show the percentages of HiCEs localized in UTR, intron, and coding sequences. See Supplemental Table S5 for annotation of all HiCE sites. E–F) Prediction of overrepresented motifs from sequences flanking HiCE sites. E) A schematic showing the procedure of motif searching with the MEME program. Red color indicates the RNA editing sites. F) Sequence logo plots showing two predicted sequence motifs that pair with each other. See also Supplemental Fig. S9 for all predicted sequence motifs.

We next asked whether any sequence motifs and gene functions were overrepresented in the HiCE-marked region. To this end, 250 bp upstream and downstream sequences of HiCE sites were extracted and analyzed (Fig. 4E). The motif prediction predicted 12 overrepresented motifs flanking all HiCEs. Interestingly, sequence complementarities were found among the 10 predicted motifs (Fig. 4F; Supplemental Fig. S9). Figure 4F shows that two overrepresented sequence motifs perfectly match each other and tend to form dsRNA. We also analyzed the gene ontology (GO) terms associated with HiCE-marked transcripts. Interestingly, two GO terms under the catalog of biological process, including response to biotic stimulus and lipid metabolic process, were significantly enriched (Supplemental Fig. S10, FDR adjusted P < 0.05). Nevertheless, further studies are needed to investigate the biochemical and physiological connections between these HiCE-marked transcripts and OsDRB1.

OsDRB1-ADARdd also marked mature miRNAs and other sRNAs

OsDRB1 is one of the homologs of Arabidopsis HYL1, which interacts with DCL1 to process stem-loop precursors (pre-miRNAs) in MIR transcripts (Raghuram et al. 2015). A recent study indicated that miRNA processing couples with transcription (Gonzalo et al. 2022); therefore, we asked whether RNA editing in OsDRB1-ADARdd-OE plants was passed to small RNAs. We sequenced the small RNA fraction of leaf samples of OsDRB1-ADARdd-OE and ADARdd-OE plants. Small RNA-seq (sRNA-seq) obtained 8.2 to 8.8 million clean reads for each replicate, and 92% to 94% of them were mapped to the rice reference genome (Supplemental Table S6). However, only 9.5% to 19.7% of small RNA reads were uniquely mapped to the rice reference genome using alignment conditions that were used to identify sRNA editing in animals (de Hoon et al. 2010) because many small RNAs from repetitive sequences and small RNAs from the same family often share high sequence identities in plants. In this study, we analyzed only RDVs from sRNA reads that uniquely mapped to the reference genome.

The sRNA edits were extracted from RDVs of sRNA-DNA alignments. In comparison with ADARdd-OE, OsDRB1-ADARdd had a comparable amount of background RDVs, except for the A-to-G and T-to-C RDVs, which were 2 to 4 times more abundant than the others (Fig. 5A; one-way ANOVA and Tukey's HSD test, P < 0.05). These A-to-G and T-to-C RDVs were predominantly present in 21-, 22-, and 24-nt sRNAs, which are also the most abundant sRNA fractions in rice (Fig. 5B). After processing these sRNA-RDVs using the pipelines described in Fig. 2D, 12.5% of the 10 types of non–RNA editing sRNA-RDVs remained. Such slightly high backgrounds are likely due to the presence of many single-nucleotide variations (SNVs) between endogenous sRNAs. In contrast, 38.5% (191) of A-to-G and T-to-C RDVs remained in the OsDRB1-ADARdd-OE data set and were considered sRNA edits (Fig. 5C, Supplemental Table S7).

Figure 5.

Figure 5.

Open in a new tab

Characterization of RNA edits in small RNAs of OsDRB1-ARADdd-OE. A) Comparison of RDV numbers from sRNA-seq data sets of OsDRB1-ADARdd-OE and ADARdd-OE. Data are presented as the mean ± SD (n = 2 biological replicates). The letters indicate statistical significance using one-way ANOVA followed by post hoc Tukey's LSD, α = 0.05. B) The number of uniquely mapped sRNA-seq reads with different lengths. The upper plot shows all sRNA-seq reads of OsDRB1-ADARDD-OE that are uniquely mapped to the rice genome. The bottom plot shows the number of mapped sRNA-seq reads containing A-to-G and T-to-C RDVs. The data are shown as mean ± SD. C) Percentages of filtered RDVs during sRNA edit identification. The RDV filtering pipeline is shown in Fig. 2D. D) Venn diagram showing the percentages and number of edited miRNAs, repeat-derived sRNAs, and other sRNAs. E–F) Number of editing sites and editing percentages in miRNAs. Pearson's correlation coefficient is shown in F. G) Heatmap of the amount of sRNA-seq reads mapped to pre-miRNAs. The number in the boxes indicates the total number of sRNA editing sites in pre-miRNAs. H–K) sRNA editing sites in pre-miR159a and pre-miR169a. H and I) Genome browser screen shots of sRNA-seq reads mapped to pre-miR159a and pre-miR169a. The positions of editing sites are shown with red arrows. J and K) Positions of editing sites in pre-miR159a and pre-miR169a stem-loop structures. Red letters (also bold font), edited adenosines; blue letters, mature miRNAs.

After mapping 191 sRNA edits to the rice genome, 47%, 42%, and 11% of them were located in unannotated regions, repetitive elements, and MIR genes, respectively (Fig. 5D). A total of 21 sRNA edits were found in 18 mature miRNAs, including 3 miRNAs containing 2 edits (Fig. 5E). The editing percentages ranged from 6.3% to 100% and showed strong correlations between the two replicates (Pearson's correlation coefficient r = 0.98, Fig. 5F). In addition to mature miRNAs, A-to-G and T-to-C RNA edits were also found in other sRNAs processed from pre-miRNAs. As a result, up to 4 edits (e.g. pre-miR1851) were found in these pre-miRNAs (Fig. 5G). The expression levels of different miRNAs are vastly different (Fig. 5G), implying that OsDRB1-ADARdd has high sensitivity to edit lowly expressed RNAs. We show the sRNA edits in pre-miR159a and pre-miR169a in Fig. 5, H to K. In both examples, the sRNA edits were found outside of the miRNA coding region (within 2 bp to the mature miRNA region), consistent with previous studies showing that Arabidopsis DRB1/HYL1 bind sequences adjacent to mature miRNA regions (Wu et al. 2007). The sRNA edits data further demonstrate that OsDRB1-ADARdd robustly edited its RNA ligands in vivo and that the edited adenosines were passed to mature miRNAs.

Comparisons of RIP-seq and RNA editing data of OsDRB1

We finally tested the RNA immunoprecipitation sequencing (RIP-seq) method to identify OsDRB1-associated RNAs. We first tried RIP-seq using leaf samples of OsDRB1-OE plants; however, in all attempts, we always obtained a high background in negative controls. We therefore performed formaldehyde crosslinking and RIP-seq using protoplasts, which might have better crosslinking efficiency, and the crosslinked OsDRB1–RNA complex allowed rigorous washing to reduce the background of RIP (Fig. 6A). Western blotting showed that the RNA–protein complex in protoplasts was specifically immunoprecipitated using an anti-FLAG antibody (Fig. 6B). Following standard RIP-seq procedures, the negative control, which was transfected with an empty plasmid vector, showed a clean background, as neither RNA nor cDNA amplification was detected (Fig. 6C). After sequencing the RNAs recovered from the OsDRB1–FLAG IP complex, however, 72% of OsDRB1 RIP-seq reads were ribosome RNA (rRNA) (Fig. 6D), likely due to the nonspecifically crosslinking of abundant rRNA with the OsDRB1–RNA complex. Despite such a high background, RIP-seq detected 51 HiCE-marked transcripts (Supplemental Table S8). The RIP-seq signal peaks of these TUs overlapped with HiCE sites (Fig. 6E). Figure 6F and G shows an example of HiCE and RIP-seq signal peaks at Os01g0310550, which forms dsRNA with its antisense transcript OsRLCK31. Sanger sequencing also confirmed that OsDRB1-ADARdd edited multiple sites in this RNA (Fig. 6G). Our data show that conventional RIP-seq has a high background and therefore requires much work to identify the true RNA ligands of RBPs of interest. In contrast, ADARdd-mediated RNA editing provides a simple and sensitive method to characterize RBP–RNA interactions in plants.

Figure 6.

Figure 6.

Open in a new tab

RIP-seq and sanger sequencing validate the TRIBE analysis. A) Schematic of RIP-seq procedures to capture OsDRB1-associated RNAs in protoplasts. B–D) Results of OsDRB1 RIP-seq. B) Western blotting showing the expression of FLAG-tagged OsDRB1 in input and IP samples. C) Comparisons of IP purified RNAs (OsDRB1 vs CK) by agarose gel electrophoresis. CK, protoplast transfected with empty vectors. D) Pie plot showing the percentages of RIP-seq reads mapped to abundant rRNA and annotated transcript units (TUs). A total of 52 TUs were marked by HiCE and RIP-seq signals. E) The density curve showing the distance between RIP-seq peaks and HiCE sites. F) Schematic illustration of RIP-seq reads, RNA-seq reads, and HiCE sites at the Os01g0310550 locus. F and R indicate the primers used for PCR. Sequencing data of replicate 1 of OsDRB1-ADARdd-OE, OsDRB1-OE, and ADARdd-OE lines are shown. G) Sanger sequencing chromatograms showing RNA editing at the Os01g0310550 transcript. Asterisks indicate the edited adenosines.

Discussion

In plants, gene expression is governed by a wide range of RBPs during transcription and posttranscriptional regulation. RBP–RNA interactions are highly dynamic in different tissues, cell types, developmental stages, and environmental conditions (Cho et al. 2019). Therefore, the identification of RNA ligands of RBPs is important to dissect RBP-mediated regulation of gene expression in plants. In this study, we employed ADARdd to edit endogenous RNA ligands of RBPs including MCP, AtGRP7, and OsDRB1 in rice. More importantly, we established a simple RDV processing pipeline to identify HiCE sites from RNA-seq data, which almost completely removed false-positive RNA edits. Our approach sets a standard for ADARdd-mediated RNA editing that provides a simple, robust, and sensitive approach for genome-wide identification of RNA ligands of RBPs in vivo.

In comparison with traditional RIP-seq, TRIBE only requires stable transgenic plants expressing RBP-ADARdd fusion proteins, and then, the RNA ligands could be identified from RDVs in RNA-seq data (Fig. 2). The CLIP method, which purifies radioactively labeled RNAs by excising the RBP–RNA complex from polyacrylamide gel separated immunoprecipitation products (McMahon et al. 2016), can greatly reduce the background of RIP-seq but is comparatively challenging and laborious. Previous studies suggested that CLIP and TRIBE results were in high agreement with each other but TRIBE was more sensitive to detect low-abundant transcripts than other methods (McMahon et al. 2016; Arribas-Hernandez et al. 2021). As genetic transformation has been established in many plants, TRIBE could be widely used to study RBP–RNA interactions in different plant species. In addition to stable transformation, transient expression systems such as protoplasts and agroinfiltration could also be used. The transient expression system would be particularly useful for fast verification of interactions between RBPs and candidate RNA ligands, as the MCP and MS2-RNA example we presented here (Fig. 1). In addition, the application of TRIBE in protoplasts could fast screen the RNA ligands of RBPs even though the physiological conditions of protoplasts are distinct for those of normal cells and many genes are not expressed in protoplasts. Our sRNA-seq data indicate that mature sRNAs in sDRB1-ADARdd-OE plants also contain RNA edits (Fig. 5), suggesting that RNA editing events were stable during small RNA processing. Furthermore, the editing window was located 4 to 41 nt adjacent to the RBP-ADARdd binding site (Fig. 1G). This parameter could potentially be used to predict the binding sequence of RBPs.

ADARdd deaminase is highly active and compatible with different RBPs in plants. This study and other reports show that ADARdd is flexible to fuse with different RBPs in plants, including MCP (this study), Arabidopsis ECT2 (Arribas-Hernandez et al. 2021), Arabidopsis GRP7 (this study and (Zhou et al. 2021)), Arabidopsis UBC (Zhou et al. 2021), and rice DRB1 (this study). We did not observe RNA edits in ADARdd-OE rice plants (Fig. 2), suggesting that the deaminase activity of ADARdd requires an RNA-binding domain; therefore, ADARdd alone likely has no (or undetectable) off-target editing in rice. However, the proximal editing activity of RBP-ADARdd may have introduced off-target editing, as we observed that 15% of A-to-C and T-to-G RDVs were replicate specific (Fig. 2E). The editing percentage of most HiCE sites was higher than 40% in this study, which is higher than the editing percentage of most HiCE sites in a previous report (Arribas-Hernandez et al. 2021), likely because we expressed OsDRB1-ADARdd with a strong promoter and used a stringent filtering pipeline to extract HiCEs that tend to exclude the editing sites with low frequencies. We found that RBP-ADARdd fusion was less efficient in editing the GA site (Fig. 3E). However, such sequence bias has minimum impact on RNA ligand identification because multiple adenosines should be presented within the ADARdd editing window. In addition, other RNA editing enzymes (Pecori et al. 2022) with less sequence bias could be used to mitigate this bias in the future. Together, ADARdd-mediated RNA editing is efficient and accurate for studying RBP–RNA interactions in plants.

To engineer ADAR to identify RNA ligands of RBPs, the following factors should be considered. First, negative controls and multiple biological replicates are required to filter genomic SNPs, somatic mutations, and transient RNA edits from RDVs. Genomic SNPs accounted for more than 90% of non–RNA editing RDVs in rice transformants (Fig. 2E). To filter these sequence noises, multiple negative controls, especially plants regenerated from the same transformation experiment, are required to train the RDV analysis pipeline. Second, whole genome sequencing might be included in the future to verify RNA edits. Our results imply that the current analysis pipeline resulted in very few false positives from genomic mutations in root samples (Supplemental Table S4). These genomic mutations could be identified using whole genome sequencing data from the same plants. Third, similar to RNA editing identification in animals (Lee et al. 2013; Sims et al. 2014), sequencing depth is important to identify editing sites for genes with low expression. On average, 51% of A-to-G and T-to-C RDVs in OsDRB1-ADARdd-OE data sets were removed due to low read depth (Fig. 2E) although some true RNA editing events may also be removed. In the future, increasing the sequencing depth and/or depleting abundant transcripts could be used to identify low-expression RNA ligands of RBPs of interest. Fourth, single-nucleotide variations (SNVs) between endogenous RNAs increase the difficulties of RNA editing event identification. As we observed in sRNA-seq data (Fig. 5C), SNVs between endogenous sRNAs are difficult to distinguish from RNA edits. Fifth, RNA editing may impair gene expression and affect plant growth. The A-to-I edits may change the start/stop codon or mRNA splicing site. Such RNA edits disrupt gene function, and overexpressing some RBP-ADARdd may affect plant growth. In such cases, inducible expression of RBP-ADARdd fusion protein could be used to profile their RNA ligands. Last, although ADARdd does contain signal peptide, the subcellular localization of RBP-ADARdd fusions should be examined in some experiments. The same subcellular localization of RBP-ADARdd fusion and RBP is important to capture the spatiotemporal RBP–RNA interactions since RBP may only interact with its RNA ligands in specific organelles (Qin et al. 2021) or membraneless organelles (Wang and Gu 2022).

OsDRB1 is predicted to participate in small RNA biogenesis, but its function has not been characterized. In consistent with this hypothesis, OsDRB1-ADARdd editing sites were found in the 3′-UTR and intron of mRNAs, MIR genes, and repetitive elements (Fig. 4, A to D). These regions are also sRNA hotspots (Ganie et al. 2017; Campo et al. 2021). Additionally, A-to-G RNA edits were identified in miRNAs and other sRNAs (Fig. 5). These data demonstrate that OsDRB1 is involved in sRNA biogenesis or function. Interestingly, we found that OsDRB1-ADARdd editing sites are located within or adjacent to mature miRNA regions in pre-miRNAs (Fig. 5, G and H). Based on the ADARdd editing window we observed, the OsDRB1-binding region in pre-miRNAs is likely within 40 bp of miRNA sequences. These data will help to elucidate the biochemical role of OsDRB1 in sRNA biogenesis. Furthermore, several lines of evidence imply that OsDRB1 is involved in biotic stresses. For example, the transcription of OsDRB1 was induced by rice blast fungus infection and in vitro assays found that OsDRB1 was phosphorylated by a stress-activated mitogen-activated protein kinase (Raghuram et al. 2015). Genes related to biotic stress response were also enriched in OsDRB1 HiCE-marked transcripts (Supplemental Fig. S10). Given that many miRNAs and siRNAs were involved in plant–microbe interactions (Weiberg et al. 2014; Feng et al. 2021), the identification of RNA targets of OsDRB1 would help to understand how stress signals are transduced from upstream protein kinases to sRNA-mediated transcriptome reprogramming.

In summary, this study presents an RNA editing toolkit to identify RNA ligands of RBPs of interest in vivo. Owing to its simplicity and robustness, we anticipate that this tool and the bioinformatic pipeline would be helpful to study the regulatory roles of RBP–RNA interactions.

Materials and methods

Plasmid construction

The codon-optimized ADARdd and MCP were synthesized by GenScript (Supplemental Figs. S1 and S2). The E488Q mutation was introduced into ADARdd using a QuikChange site-directed mutagenesis kit (Agilent, Santa Clara, CA, USA). OsDRB1 (Os11g0109900) and AtGRP7 (AT2G21660.1) were amplified from the rice (Oryza sativa) cultivar Nipponbare and Arabidopsis (Arabidopsis thaliana) Col-0 by RT-PCR. Then, MCP, OsDRB1, and AtGRP7 were fused with ADARdd by overlap extension PCR. The PCR products of MCP, MCP-ADARdd, AtGRP7, AtGRP7-ADARdd, OsDRB1, OsDRB1-ADARdd, and ADARdd were cloned into KpnI and XhoI sites of pENTR11 (Life Technologies, Carlsbad, CA, USA). The MS2 stem-loop fragment (MS2-RNA) was generated by annealing four DNA oligos (MS2–1∼4) and subsequently cloned into pEASY-Blunt (Transgene Tech).

For the RNA editing experiment in protoplasts, MCP, ADARdd, and MCP-ADARdd in pENTR11 were cloned into p1300-32 (Ding et al. 2018) through the LR reaction, resulting in the p32-MCP, p32-ADARdd, and p32-MCP-ADARdd plasmids. MS2-RNA was PCR amplified from the plasmid using primers Bsa-MS2-F and Bsa-MS2-R and then inserted into the BsaI sites of the above plasmids to generate p32-MS2-MCP, p32-MS2-ADARdd, and p32-MS2-MCP-ADARdd (Fig. 1B), respectively. To edit AtGRP7 RNA in Arabidopsis protoplasts, AtGRP7, ADARdd, and AtGRP7-ADARdd in pENTR11 were subcloned into pUGW11 (Nakagawa et al. 2007) through the LR reaction, resulting in pUGW11-AtGRP7, pUGW11-ADARdd, and pUGW11-AtGRP7-ADARdd (Fig. 1H), respectively.

To generate stable transgenic rice plants, pENTR11-OsDRB1, pENTR11-ADARdd, and pENTR11-OsDRB1-ADARdd were subcloned into p1300-33 (Ding et al. 2018) through the LR reaction to build p33-OsDRB1, p33-ADARdd, and p33-OsDRB1-ADARdd constructs for rice transformation (Fig. 2A).

Sequences of DNA oligos used in this study are listed in Supplemental Table S9.

Protoplast transfection and rice transformation

Rice protoplast preparation and transfection were performed as previously reported (Xie and Yang 2013). Briefly, 20 μg of plasmid DNA was used to transfect ∼1 × 107 rice protoplasts. The protoplasts were harvested after incubation in the dark for 24 to 36 h. Rice (Oryza sativa L. ssp.) cultivar Nipponbare was used in this study. Agrobacterium tumefaciens–mediated transformation was performed by Towin (Wuhan, China).

Genomic DNA extraction, RNA purification, and RT-PCR

The rice genomic DNA was isolated from protoplasts and leaves using the cetyltrimethylammonium bromide method (Ding et al. 2018). 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 then, reverse transcription was performed using M-MLV (Takara Bio) with oligo (dT) primers or sequence-specific primers (RT-MS2-R for MS2-RNA). PCR was performed using Q5 High-Fidelity DNA Polymerase (New England Biolabs, Ipswich, MA, USA) for Sanger sequencing (see Supplemental Table S9 for primer sequences).

Quantification of editing efficiency using Sanger sequencing

After transfecting protoplasts with plasmids, the RNA targets were amplified by RT-PCR, and the PCR products were analyzed by Sanger sequencing. The A-to-I RNA editing was calculated using T/C peak heights in the sequencing chromatogram according to previous studies (Eggington et al. 2011; Rinkevich et al. 2012).

Protein extraction and Western blotting

The total proteins were extracted using extraction buffer containing 50-mM Tris-HCl (pH 7.4), 150-mM NaCl, 1% (v/v) Triton X-100, 1% (v/v) proteinase inhibitor cocktail (Sigma-Aldrich, Burlington, MA, USA), and 10% glycerol. For Western blotting, 2 μg of total protein was separated by 10% (w/v) sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and then transferred to a polyvinylidene fluoride (PVDF) membrane. After blocking with 5% (w/v) nonfat milk in Tris-buffered saline with Tween 20 (TBST), the blot was probed with anti-FLAG antibody (Sigma-Aldrich, 1:1000 dilution). After washing with TBST 3 times, the blot was incubated with horseradish peroxidase–conjugated antimouse secondary antibody (Sigma-Aldrich, 1:10,000 dilution). Finally, the proteins in the blot were detected using a SuperSignal West Femto Maximum Sensitivity Substrate kit (Life Technologies, Carlsbad, CA, USA) and chemiluminescence imager (Tanon-2500).

RNA-seq

Total RNA from leaf and root tissues was extracted from 3-week-old plants. Two independent T0 lines were used as biological replicates for OsDRB1-OE, ADARdd-OE, and OsDRB1-ADARdd-OE (Supplemental Fig. S5). Briefly, RNA-seq libraries were prepared using the NEBNext Ultra RNA Library Prep Kit for Illumina, and 2 × 150 bp paired-end sequencing was performed on a NovaSeq 6000 (Illumina, San Diego, CA, USA).

For small RNA sequencing, 3 μg of total RNA was used to construct a small RNA library with an NEB Next Multiplex Small RNA Library Prep Set. Single-end 50-bp sequencing was performed on a NovaSeq 6000 (Illumina).

RIP-seq

Rice protoplasts transfected with p32-OsDRB1-ADARdd (Fig. 2A) were collected and washed 3 times with phosphate-buffered saline (PBS). Protoplast cells were then resuspended in PBS with 1% (w/v) formaldehyde, and the protoplast was crosslinked for 10 min at room temperature. Then, glycine was added to a final concentration of 100 mM to quench the formaldehyde and terminate the crosslinking reaction. The protoplasts were centrifuged at 300 × g for 3 min at 4 °C, and then, the cell pellets were resuspended in ice-cold lysis buffer containing 50-mM Tris-HCl (pH 7.4), 150-mM NaCl, 1% (v/v) Triton X-100, 1-mM ethylenediaminetetraacetic acid (EDTA), 200-U/ml RNase inhibitor (Takara Bio, San Jose, CA, USA), and 1% (v/v) proteinase inhibitor cocktail (New England Biolabs). Cell lysates were then centrifuged at 13,000 × g at 4 °C for 20 min, and the supernatant was used for IP. The RBP–RNA complex was immunoprecipitated with Anti-FLAG M2 affinity gel (Sigma-Aldrich) by incubating at 4 °C for 4 h. The beads were washed 3 times with 50-mM Tris-HCl (pH 7.4), 150-mM NaCl, 0.1% (v/v) Triton X-100, 20-U/ml RNase inhibitor (Takara Bio), and 1% (v/v) proteinase inhibitor cocktail (New England Biolabs). Then, the IP products were subjected to RNA extraction using TRIzol reagent (Life Technologies). The RNA was dissolved in 15-μL nuclease-free water and used for library construction with a NEBNext Ultra RNA Library Prep Kit. The cDNA was sequenced using paired-end 2 × 150 bp sequencing in NovaSeq 6000 (Illumina).

RNA-seq data analysis

The rice reference genome sequence and annotation (IRGSP 1.0) were downloaded from Ensemble Plants (http://plants.ensembl.org/). The rice repetitive element information was downloaded from the Rice Annotation Project (RAP) (https://rapdb.dna.affrc.go.jp/). Rice microRNA annotation was downloaded from miRBase (https://mirbase.org/).

RNA-seq data were analyzed using a modified bioinformatic pipeline from TRIBE protocols published previously (McMahon et al. 2016). In brief, adaptor sequences were trimmed, and low-quality reads were removed using Trimmomatic (v0.36) (Bolger et al. 2014). The clean reads were then aligned to the rice reference genome using STAR (v2.5.3a) (Dobin et al. 2013) and the following settings: –outFilterMismatchNoverLmax 0.07 –outSAMstrandField intronMotif –outFilterMultimapNmax 1. The PCR duplicates were removed using Picard (v2.23.1), and mapped reads were sorted using SAMtools (v1.7.1). Gene expression profiles were generated by featureCounts (v2.0.1) (Liao et al. 2014) and converted to transcripts per million (TPM) values. The RDVs were extracted from alignment results using SAMtools and bcftools. To extract HiCE sites, the RDVs were filtered as follows using inhouse computer scripts (Fig. 2D): (1) remove genomic SNPs that were detected in negative controls; (2) remove low-depth RDVs (read < 10); and (3) remove replicated-specific RDVs. The editing percentages of HiCE sites are determined as the value of ReadG/(ReadA + ReadG), in which ReadA and ReadG are the number of RNA-seq reads containing A and G at the editing site, respectively. All computer codes used in RNA-seq analysis are available in GitHub (https://github.com/YS-HZAU/OsDRB1_TRIBE).

Small RNA-seq data analysis

Raw sequencing data were trimmed by fastp (v0.20.1) (Chen et al. 2018) using the following parameters: -w 12 –q 25 –u 20 –n 5 -A –L, and then adapters were trimmed using cutadapt (v3.2, parameters: -a AGATCGGAAGA -O 5 -e 0). The remaining reads were further trimmed using Trimmomatic (v0.36) (Bolger et al. 2014). The clean reads sized 18 to 34 nt and without poly(A) tails were mapped to the rice genome using bowtie (v1.3.0) (Langmead et al. 2009). RDVs with a depth > 5 and editing percentages >5% were extracted and filtered as shown in Fig. 2D. All computer codes are available in GitHub (https://github.com/YS-HZAU/OsDRB1_TRIBE).

RIP-seq data analysis

Raw sequencing reads were cleaned as described for RNA-seq data processing. Then, clean reads were merged and compared with the rice rRNA, chloroplast (Pt), and mitochondria (Mt) genomes using bowtie2 (v2.2.3). Unmapped sequences were further aligned to the rice reference genome using bwa-mem (v0.7.17) (Li and Durbin 2009), and PCR duplicates were removed using Picard (v2.23.1). After removing low-coverage reads (<4), the remaining uniquely mapped reads were used as RIP signals.

Motif analysis

Sequences (250 bp each side) flanking HiCE were extracted and merged. Then, these unduplicated sequences were used to predict overrepresented sequence motifs using MEME (v4.11.2) (Bailey et al. 2009) (parameters: -minw 10 -maxw 12 -maxsize 10000000 -dna -nmotifs 12 -maxsites 200).

Supplementary Material

kiad158_Supplementary_Data
Click here for additional data file. (6.8MB, zip)

Acknowledgments

We thank Professor Kenichi Tsuda at Huazhong Agricultural University for his suggestions in writing the manuscript.

Contributor Information

Shuai Yin, National Key Laboratory of Crop Genetic Improvement, Hubei Hongshan Laboratory, Huazhong Agricultural University, Wuhan 430070, China; Hubei Key Laboratory of Plant Pathology, Huazhong Agricultural University, Wuhan 430070, China; Shenzhen Institute of Nutrition and Health, Huazhong Agricultural University, Wuhan 430070, China; Shenzhen Branch, Guangdong Laboratory for Lingnan Modern Agriculture, Genome Analysis Laboratory of the Ministry of Agriculture, Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences, Shenzhen 518120, China.

Yuedan Chen, National Key Laboratory of Crop Genetic Improvement, Hubei Hongshan Laboratory, Huazhong Agricultural University, Wuhan 430070, China; Hubei Key Laboratory of Plant Pathology, Huazhong Agricultural University, Wuhan 430070, China.

Yache Chen, National Key Laboratory of Crop Genetic Improvement, Hubei Hongshan Laboratory, Huazhong Agricultural University, Wuhan 430070, China; Hubei Key Laboratory of Plant Pathology, Huazhong Agricultural University, Wuhan 430070, China.

Lizhong Xiong, National Key Laboratory of Crop Genetic Improvement, Hubei Hongshan Laboratory, Huazhong Agricultural University, Wuhan 430070, China.

Kabin Xie, National Key Laboratory of Crop Genetic Improvement, Hubei Hongshan Laboratory, Huazhong Agricultural University, Wuhan 430070, China; Hubei Key Laboratory of Plant Pathology, Huazhong Agricultural University, Wuhan 430070, China; Shenzhen Institute of Nutrition and Health, Huazhong Agricultural University, Wuhan 430070, China; Shenzhen Branch, Guangdong Laboratory for Lingnan Modern Agriculture, Genome Analysis Laboratory of the Ministry of Agriculture, Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences, Shenzhen 518120, China.

Author contributions

K.X, L.X., and S.Y. conceived the project and designed the experiments. S.Y., Yu.C., and Y.C. performed the experiments. K.X. and S.Y. analyzed the data. K.X. and S.Y. wrote the manuscript with input from others.

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1 . Sequence of the codon-optimized ADARdd.

Supplemental Figure S2 . Sequence of codon-optimized MS2 coat protein (MCP).

Supplemental Figure S3 . Sequence of MS2-RNA.

Supplemental Figure S4 . Sanger sequencing chromatograms showing RNA edits in MS2-RNA and AtGRP7 transcripts.

Supplemental Figure S5 . Morphology and transgene expression of OsDRB1-OE, OsDRB1-ADARdd-OE, and ADARdd-OE plants.

Supplemental Figure S6 . Correlations of global gene expression between OsDRB1-OE, OsDRB1-ADARdd-OE, and ADARdd-OE transcriptomes.

Supplemental Figure S7 . Verification of genomic SNPs by Sanger sequencing.

Supplemental Figure S8 . Examination of two non–RNA editing RDVs in OsDRB1-ADARdd-OE.

Supplemental Figure S9 . Overrepresented sequence motifs flanking the leaf- and root-HiCE sites.

Supplemental Figure S10 . Enriched Gene Ontology (GO) terms of HiCE-marked transcripts.

Supplemental Table S1 . Information on RNA-seq data sets.

Supplemental Table S2 . List of RNA-seq detected genes.

Supplemental Table S3 . Number of RDVs filtered in each step of the HiCE identification pipeline.

Supplemental Table S4 . The remaining non–RNA editing RDVs in OsDRB1-ADARdd-OE root data sets after three filtering steps.

Supplemental Table S5 . The high-confidence editing (HiCE) sites identified in OsDRB1-ADARdd-OE.

Supplemental Table S6 . Summary of small RNA-seq data sets.

Supplemental Table S7 . The sRNA edits extracted from sRNA-seq data sets.

Supplemental Table S8 . The HiCE sites with RIP-seq signals.

Supplemental Table S9 . Plasmid vectors and primer sequences used in this study.

Funding

This work was supported by funding from the National Natural Science Foundation of China (31622047 and 31821005), the Collaborative Fund of Huazhong Agricultural University and Agricultural Genomics Institute at Shenzhen (SZYJY2022004 and SZYJY2021007), and the Fundamental Research Funds for the Central Universities (2021ZKPY002).

Data availability

The RNA-seq data described in this study have been deposited in the BIG Data Center (https://bigd.big.ac.cn/) under the accession number CRA007567. The program codes for data analysis are available at GitHub (https://github.com/YS-HZAU/OsDRB1_TRIBE). Sequence information of the genes used in this study can be found in the Arabidopsis TAIR database (https://www.arabidopsis.org) and the Rice Annotation Project (RAP) (https://rapdb.dna.affrc.go.jp/) under the following accession numbers: OsDRB1 (Os11g0119900) and AtGRP7 (AT2G21660).

References

  1. Arribas-Hernandez L, Rennie S, Koster T, Porcelli C, Lewinski M, Staiger D, Andersson R, Brodersen P. Principles of mRNA targeting via the Arabidopsis m(6)A-binding protein ECT2. Elife. 2021:10:e72375. 10.7554/eLife.72375 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bach-Pages M, Homma F, Kourelis J, Kaschani F, Mohammed S, Kaiser M, van der Hoorn RAL, Castello A, Preston GM. Discovering the RNA-binding proteome of plant leaves with an improved RNA interactome capture method. Biomolecules 2020:10(4): 661. 10.3390/biom10040661 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, Ren J, Li WW, Noble WS. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 2009:37(Web Server):W202–W208. 10.1093/nar/gkp335 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Biswas J, Rahman R, Gupta V, Rosbash M, Singer RH. MS2-TRIBE evaluates both protein–RNA interactions and nuclear organization of transcription by RNA editing. iScience. 2020:23(7):101318. 10.1016/j.isci.2020.101318 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014:30(15):2114–2120. 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Borges F, Martienssen RA. The expanding world of small RNAs in plants. Nat Rev Mol Cell Biol. 2015:16(12):727–741. 10.1038/nrm4085 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Campo S, Sánchez-Sanuy F, Camargo-Ramírez R, Gómez-Ariza J, Baldrich P, Campos-Soriano L, Soto-Suárez M, San Segundo B. A novel transposable element-derived microRNA participates in plant immunity to rice blast disease. Plant Biotechnol J. 2021:19(9):1798–1811. 10.1111/pbi.13592 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen S, Zhou Y, Chen Y, Gu J. Fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018:34(17):i884–i890. 10.1093/bioinformatics/bty560 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cho H, Cho HS, Hwang I. Emerging roles of RNA-binding proteins in plant development. Curr Opin Plant Biol. 2019:51:51–57. 10.1016/j.pbi.2019.03.016 [DOI] [PubMed] [Google Scholar]
  10. Darnell RB. HITS-CLIP: panoramic views of protein–RNA regulation in living cells. Wiley Interdiscip Rev RNA. 2010:1(2):266–286. 10.1002/wrna.31 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. de Hoon MJ, Taft RJ, Hashimoto T, Kanamori-Katayama M, Kawaji H, Kawano M, Kishima M, Lassmann T, Faulkner GJ, Mattick JS, et al. Cross-mapping and the identification of editing sites in mature microRNAs in high-throughput sequencing libraries. Genome Res. 2010:20(2):257–264. 10.1101/gr.095273.109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ding D, Chen K, Chen Y, Li H, Xie K. Engineering introns to express RNA guides for Cas9- and Cpf1-mediated multiplex genome editing. Mol Plant. 2018:11(4):542–552. 10.1016/j.molp.2018.02.005 [DOI] [PubMed] [Google Scholar]
  13. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013:29(1):15–21. 10.1093/bioinformatics/bts635 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Doroshenk KA, Crofts AJ, Morris RT, Wyrick JJ, Okita TW. RiceRBP: a resource for experimentally identified RNA binding proteins in Oryza sativa. Front Plant Sci. 2012:3:90. 10.3389/fpls.2012.00090 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Eggington JM, Greene T, Bass BL. Predicting sites of ADAR editing in double-stranded RNA. Nat Commun. 2011:2(1):319. 10.1038/ncomms1324 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Feng Q, Li Y, Zhao ZX, Wang WM. Contribution of small RNA pathway to interactions of rice with pathogens and insect pests. Rice. 2021:14(1):15. 10.1186/s12284-021-00458-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ganie SA, Debnath AB, Gumi AM, Mondal TK. Comprehensive survey and evolutionary analysis of genome-wide miRNA genes from ten diploid Oryza species. BMC Genomics. 2017:18(1):711. 10.1186/s12864-017-4089-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gonzalo L, Tossolini I, Gulanicz T, Cambiagno DA, Kasprowicz-Maluski A, Smolinski DJ, Mammarella MF, Ariel FD, Marquardt S, Szweykowska-Kulinska Z, et al. R-loops at microRNA encoding loci promote co-transcriptional processing of pri-miRNAs in plants. Nat Plants. 2022:8(4):402–418. 10.1038/s41477-022-01125-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hafner M, Landthaler M, Burger L, Khorshid M, Hausser J, Berninger P, Rothballer A, Ascano M Jr, Jungkamp AC, Munschauer M, et al. Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell. 2010:141(1):129–141. 10.1016/j.cell.2010.03.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hentze MW, Castello A, Schwarzl T, Preiss T. A brave new world of RNA-binding proteins. Nat Rev Mol Cell Biol. 2018:19(5):327–341. 10.1038/nrm.2017.130 [DOI] [PubMed] [Google Scholar]
  21. Kim B, Kim VN. fCLIP-seq for transcriptomic footprinting of dsRNA-binding proteins: lessons from DROSHA. Methods. 2019:152:3–11. 10.1016/j.ymeth.2018.06.004 [DOI] [PubMed] [Google Scholar]
  22. König J, Zarnack K, Rot G, Curk T, Kayikci M, Zupan B, Turner DJ, Luscombe NM, Ule J. iCLIP reveals the function of hnRNP particles in splicing at individual nucleotide resolution. Nat Struct Mol Biol. 2010:17(7):909–915. 10.1038/nsmb.1838 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Koster T, Reichel M, Staiger D. CLIP And RNA interactome studies to unravel genome-wide RNA-protein interactions in vivo in Arabidopsis thaliana. Methods. 2020:178:63–71. 10.1016/j.ymeth.2019.09.005 [DOI] [PubMed] [Google Scholar]
  24. Kostyuk V, Potapovich A, Suhan T, De Luca C, Pressi G, Dal Toso R, Korkina L. Plant polyphenols against UV-C-induced cellular death. Planta Med. 2008:74(5):509–514. 10.1055/s-2008-1074499 [DOI] [PubMed] [Google Scholar]
  25. Kurihara Y, Takashi Y, Watanabe Y. The interaction between DCL1 and HYL1 is important for efficient and precise processing of pri-miRNA in plant microRNA biogenesis. RNA. 2006:12(2):206–212. 10.1261/rna.2146906 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009:10(3):R25. 10.1186/gb-2009-10-3-r25 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lee JH, Ang JK, Xiao X. Analysis and design of RNA sequencing experiments for identifying RNA editing and other single-nucleotide variants. RNA. 2013:19(6):725–732. 10.1261/rna.037903.112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Li H, Durbin R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics. 2009:25(14):1754–1760. 10.1093/bioinformatics/btp324 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014:30(7):923–930. 10.1093/bioinformatics/btt656 [DOI] [PubMed] [Google Scholar]
  30. Lunde BM, Moore C, Varani G. RNA-binding proteins: modular design for efficient function. Nat Rev Mol Cell Biol. 2007:8(6):479–490. 10.1038/nrm2178 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. McMahon AC, Rahman R, Jin H, Shen JL, Fieldsend A, Luo W, Rosbash M. TRIBE: hijacking an RNA-editing enzyme to identify cell-specific targets of RNA-binding proteins. Cell. 2016:165(3):742–753. 10.1016/j.cell.2016.03.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Meyer K, Köster T, Nolte C, Weinholdt C, Lewinski M, Grosse I, Staiger D. Adaptation of iCLIP to plants determines the binding landscape of the clock-regulated RNA-binding protein AtGRP7. Genome Biol. 2017:18(1):204. 10.1186/s13059-017-1332-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nakagawa T, Kurose T, Hino T, Tanaka K, Kawamukai M, Niwa Y, Toyooka K, Matsuoka K, Jinbo T, Kimura T. Development of series of gateway binary vectors, pGWBs, for realizing efficient construction of fusion genes for plant transformation. J Biosci Bioeng. 2007:104(1):34–41. 10.1263/jbb.104.34 [DOI] [PubMed] [Google Scholar]
  34. Pecori R, Di Giorgio S, Paulo Lorenzo J, Nina Papavasiliou F. Functions and consequences of AID/APOBEC-mediated DNA and RNA deamination. Nat Rev Genet. 2022:23(8):505–518. 10.1038/s41576-022-00459-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Phelps KJ, Tran K, Eifler T, Erickson AI, Fisher AJ, Beal PA. Recognition of duplex RNA by the deaminase domain of the RNA editing enzyme ADAR2. Nucleic Acids Res. 2015:43(2):1123–1132. 10.1093/nar/gku1345 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Qin W, Myers SA, Carey DK, Carr SA, Ting AY. Spatiotemporally-resolved mapping of RNA binding proteins via functional proximity labeling reveals a mitochondrial mRNA anchor promoting stress recovery. Nat Commun. 2021:12(1):4980. 10.1038/s41467-021-25259-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Raghuram B, Sheikh AH, Rustagi Y, Sinha AK. MicroRNA biogenesis factor DRB1 is a phosphorylation target of mitogen activated protein kinase MPK3 in both rice and Arabidopsis. FEBS J. 2015:282(3):521–536. 10.1111/febs.13159 [DOI] [PubMed] [Google Scholar]
  38. Reich DP, Bass BL. Mapping the dsRNA world. Cold Spring Harb Perspect Biol. 2019:11(3):a035352. 10.1101/cshperspect.a035352 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Reichel M, Liao Y, Rettel M, Ragan C, Evers M, Alleaume AM, Horos R, Hentze MW, Preiss T, Millar AA. In planta determination of the mRNA-binding proteome of Arabidopsis etiolated seedlings. Plant Cell. 2016:28(10):2435–2452. 10.1105/tpc.16.00562 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Rinkevich FD, Schweitzer PA, Scott JG. Antisense sequencing improves the accuracy and precision of A-to-I editing measurements using the peak height ratio method. BMC Res Notes. 2012:5(1):63. 10.1186/1756-0500-5-63 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shim JS, Park SH, Lee DK, Kim YS, Park SC, Redillas M, Seo JS, Kim JK. The rice GLYCINE-RICH PROTEIN 3 confers drought tolerance by regulating mRNA stability of ROS scavenging-related genes. Rice. 2021:14(1):31. 10.1186/s12284-021-00473-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Sims D, Sudbery I, Ilott NE, Heger A, Ponting CP. Sequencing depth and coverage: key considerations in genomic analyses. Nat Rev Genet. 2014:15(2):121–132. 10.1038/nrg3642 [DOI] [PubMed] [Google Scholar]
  43. Staiger D, Zecca L, Wieczorek Kirk DA, Apel K, Eckstein L. The circadian clock regulated RNA-binding protein AtGRP7 autoregulates its expression by influencing alternative splicing of its own pre-mRNA. Plant J. 2003:33(2):361–371. 10.1046/j.1365-313X.2003.01629.x [DOI] [PubMed] [Google Scholar]
  44. Tian L, Chou HL, Zhang L, Okita TW. Targeted endoplasmic reticulum localization of storage protein mRNAs requires the RNA-binding protein RBP-L. Plant Physiol. 2019:179(3):1111–1131. 10.1104/pp.18.01434 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Tutucci E, Vera M, Biswas J, Garcia J, Parker R, Singer RH. An improved MS2 system for accurate reporting of the mRNA life cycle. Nat Methods. 2018:15(1):81–89. 10.1038/nmeth.4502 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Ule J, Jensen KB, Ruggiu M, Mele A, Ule A, Darnell RB. CLIP identifies Nova-regulated RNA networks in the brain. Science. 2003:302(5648):1212–1215. 10.1126/science.1090095 [DOI] [PubMed] [Google Scholar]
  47. Van Nostrand EL, Pratt GA, Shishkin AA, Gelboin-Burkhart C, Fang MY, Sundararaman B, Blue SM, Nguyen TB, Surka C, Elkins K, et al. Robust transcriptome-wide discovery of RNA-binding protein binding sites with enhanced CLIP (eCLIP). Nat Methods. 2016:13(6):508–514. 10.1038/nmeth.3810 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wang W, Gu Y. The emerging role of biomolecular condensates in plant immunity. Plant Cell. 2022:34(5):1568–1572. 10.1093/plcell/koab240 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Weiberg A, Wang M, Bellinger M, Jin H. Small RNAs: a new paradigm in plant-microbe interactions. Annu Rev Phytopathol. 2014:52(1):495–516. 10.1146/annurev-phyto-102313-045933 [DOI] [PubMed] [Google Scholar]
  50. Wu F, Yu L, Cao W, Mao Y, Liu Z, He Y. The N-terminal double-stranded RNA binding domains of Arabidopsis HYPONASTIC LEAVES1 are sufficient for pre-microRNA processing. Plant Cell. 2007:19(3):914–925. 10.1105/tpc.106.048637 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Xie K, Yang Y. RNA-guided genome editing in plants using a CRISPR-Cas system. Mol Plant. 2013:6(6):1975–1983. 10.1093/mp/sst119 [DOI] [PubMed] [Google Scholar]
  52. Xu W, Rahman R, Rosbash M. Mechanistic implications of enhanced editing by a HyperTRIBE RNA-binding protein. RNA. 2018:24(2):173–182. 10.1261/rna.064691.117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Yang LZ, Wang Y, Li SQ, Yao RW, Luan PF, Wu H, Carmichael GG, Chen LL. Dynamic imaging of RNA in living cells by CRISPR-Cas13 systems. Mol Cell. 2019:76(6):981–997.e7. 10.1016/j.molcel.2019.10.024 [DOI] [PubMed] [Google Scholar]
  54. Zhou X, Liao H, Chern M, Yin J, Chen Y, Wang J, Zhu X, Chen Z, Yuan C, Zhao W, et al. Loss of function of a rice TPR-domain RNA-binding protein confers broad-spectrum disease resistance. Proc Natl Acad Sci USA. 2018:115(12):3174–3179. 10.1073/pnas.1705927115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zhou G, Niu R, Zhou Y, Luo M, Peng Y, Wang H, Wang Z, Xu G. Proximity editing to identify RNAs in phase-separated RNA binding protein condensates. Cell Discov. 2021:7(1):72. 10.1038/s41421-021-00288-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

kiad158_Supplementary_Data
Click here for additional data file. (6.8MB, zip)

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 the accession number CRA007567. The program codes for data analysis are available at GitHub (https://github.com/YS-HZAU/OsDRB1_TRIBE). Sequence information of the genes used in this study can be found in the Arabidopsis TAIR database (https://www.arabidopsis.org) and the Rice Annotation Project (RAP) (https://rapdb.dna.affrc.go.jp/) under the following accession numbers: OsDRB1 (Os11g0119900) and AtGRP7 (AT2G21660).

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES