At-RS31 orchestrates hierarchical cross-regulation of splicing factors and integrates alternative splicing with TOR-ABA pathways

Abstract

Alternative splicing is essential for plants, enabling a single gene to produce multiple transcript variants to boost functional diversity and fine-tune responses to environmental and developmental cues. Arabidopsis thaliana At-RS31, a plant-specific splicing factor in the Serine/Arginine-rich (SR) protein family, responds to light and the Target of Rapamycin (TOR) signalling pathway, yet its downstream targets and regulatory impact remain unknown. To identify At-RS31 targets, we applied individual-nucleotide resolution crosslinking and immunoprecipitation (iCLIP) and RNAcompete assays. Transcriptomic analyses of At-RS31 mutant and overexpressing plants further revealed its effects on alternative splicing. iCLIP identified 4034 At-RS31 binding sites across 1421 genes, enriched in CU-rich and CAGA RNA motifs. Comparative iCLIP and RNAcompete data indicate that the arginine/serine (RS) domain of At-RS31 may influence its binding specificity in planta, underscoring the value of combining in vivo and in vitro approaches. Transcriptomic analysis showed that At-RS31 modulates diverse splicing events, particularly intron retention and exitron splicing, and influences other splicing modulators, acting as a hierarchical regulator. By regulating stress response genes and genes in both TOR and abscisic acid signalling pathways, At-RS31 may help integrate these signals, balancing plant growth with environmental adaptability through alternative splicing.

Keywords: ABA; Arabidopsis thaliana; RNAcompete; SR proteins; TOR kinase; alternative splicing; binding site; iCLIP.

Fig. 1

Determination of Arabidopsis thaliana serine/arginine‐rich (SR) protein At‐RS31 in vivo binding sites by individual‐nucleotide resolution crosslinking and immunoprecipitation (iCLIP). (a) Left: autoradiogram of RS31‐GFP and green fluorescent protein (GFP) protein–RNA complexes. After ultraviolet crosslinking, cell lysates were subjected to immunoprecipitation with GFP Trap beads. RNAs were radioactively labelled, and the complexes were separated by denaturing gel electrophoresis. IN, input (lysate). Treatment of the lysate with RNase I (+ RNase) indicates the size of the precipitated proteins. The region above the fusion protein containing the co‐precipitated RNAs used for library preparation is indicated. Right: iCLIP western blot. Immunoblot analysis of the membrane shown in the left panel with anti‐GFP antibody. Bands for GFP and RS31‐GFP are marked accordingly. (b) Functional profiling of genes containing At‐RS31 binding sites. Gene Ontology (GO) term enrichment analysis was performed using the g:GOSt tool from g:Profiler. Numeric and colour‐coded (capped at –log₁₀ (P_adj) ≤ 16) P‐values are shown for the enriched GO terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. BP, biological process; CC, cellular compartment; MF, molecular function. (c) Distribution of the At‐RS31 binding sites within protein‐coding transcripts. Distribution of RS31‐GFP binding sites within transcripts in relation to the total genomic length of the transcript features. 5′ UTR and 3′ UTR – 5′ and 3′ untranslated regions; CDS, coding sequences. (d) Distribution of the At‐RS31 binding sites within 5' UTRs relative to the transcription start sites (TSS). Distance to TSS is shown in nucleotides (nt). (e) Distribution of At‐RS31 binding sites peaking upstream of 5′ splice sites (5′ SS). Only exons at least 50 nt in length were analysed. The red line represents the local density of binding sites. The blue box marks the −35–20‐nt region upstream of 5′ SS where RS31‐GFP binding sites are enriched. (f) Significant STREME binding site motifs. Sequence logos of the most significant (based on their P‐value) RS31‐GFP binding motifs identified by STREME analysis. For the analysis, only sequences from the 9‐nt binding site regions were considered. (g) Hexamer counts and z‐scores. Scatterplot showing hexamer frequencies and counts computed from the 9‐nt binding site sequences of the RS31‐GFP iCLIP sample. Hexamer counts (x‐axis) are compared against hexamer z‐scores (y‐axis). The most enriched hexamers and the ones with highest counts are labelled according to their sequence. The highlighted hexamers (red) contain the subsequence CAGA. RS, arginine/serine.

Fig. 2

Identification of Arabidopsis thaliana Serine/Arginine‐rich (SR) protein At‐RS31 RNA‐binding motifs using RNAcompete. (a) Overview of the RNAcompete assay. Glutathione S‐transferase (GST)‐tagged full‐length At‐RS31 protein and its truncated version comprising both RNA recognition motifs (RRMs) were incubated with a 75‐fold molar excess of designed RNA pool. RNA bound to GST‐RS31 and GST‐RS31RRMs fusion proteins during the GST pulldown was eluted, purified, labelled, and hybridized to custom Agilent 244 K microarrays. Microarray data were analysed computationally to identify 7‐mers specifically bound by At‐RS31 and generate RNA‐binding motifs. RRM1 and RRM2, RNA recognition motif domains; RS, region rich in arginines and serines; nt, nucleotides. (b) RNA‐binding motifs of the full‐length At‐RS31 and its truncated version containing RRMs identified in the RNAcompete assay and represented as sequence logos. (c) Individual‐nucleotide resolution crosslinking and immunoprecipitation (iCLIP) binding site distance to RNAcompete motif sites. Distribution of RS31‐GFP binding site peaks in relation to CAGA sites across At‐RS31 target transcripts. The position 1 on the x‐axis denotes the C in the CAGA motif identified by the RNAcompete assay. RS, arginine/serine.

Fig. 3

Impact of At‐RS31 on alternative splicing and gene expression in Arabidopsis thaliana. (a) Distribution of alternative splicing event types differentially regulated in rs31‐1 and 35S::RS31 lines in comparison with wild‐type (WT) A. thaliana. Diagrams on the top illustrate the analysed alternative splicing event types: alternative donor (AD, or alternative 5′ splice site), alternative acceptor (AA, or alternative 3′ splice site), cassette exon (CE), exitron (EI), and retained intron (RI). (b) Proportions of differential EI and RI events with reduced or increased splicing efficiency in rs31‐1 and 35S::RS31 lines compared with WT. Positive and negative percentages indicate the proportion of events with increased or decreased per cent spliced‐in values (ΔPSI), respectively. (c, e) Venn diagram comparisons of genes exhibiting At‐RS31 individual‐nucleotide resolution crosslinking and immunoprecipitation (iCLIP) binding sites, differential alternative splicing (DAS), and differential expression (DE) in rs31‐1 and 35S::RS31 relative to WT. (c) Overlap between DAS genes and genes with At‐RS31 iCLIP binding sites. (e) Overlap between DAS and DE genes. (d, f) Functional profiling of (d) DAS genes containing At‐RS31 binding sites and (f) 35S::RS31 DE genes. The Gene Ontology (GO) term enrichment analysis was performed using the g:GOSt tool of the g:Profiler. Numeric and colour‐coded (capped at –log₁₀(P_adj) ≤ 16) P‐values are shown for the enriched GO terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. RS, arginine/serine.

Fig. 4

Regulation of Arabidopsis serine/arginine‐rich (SR) protein family by At‐RS31. (a) Overview of differential alternative splicing (DAS) and differential gene expression (DE) in Arabidopsis SR genes in rs31‐1 mutant and 35S::RS31 overexpression plants, relative to wild‐type (WT) controls. The table includes information on individual‐nucleotide resolution crosslinking and immunoprecipitation (iCLIP) peaks and the top 7‐mers identified by RNAcompete within SR genes. ‘NA’ indicates data not applicable. (b) Integrated Genomics Viewer tracks depicting iCLIP crosslink sites for green fluorescent protein (GFP) and RS31‐GFP, along with RNA sequencing read coverage tracks for rs31‐1, 35S::RS31, and WT plants. The related tracks have the same scale. Transcript models for SR genes are displayed using the Boxify tool (https://boxify.boku.ac.at/). Protein‐coding regions, spanning from the translational start codon to the stop codon or premature termination codon, are shown in black. Red vertical lines represent the locations of At‐RS31 binding sites identified by iCLIP. Only transcripts relevant to the identified DAS events are displayed, with dashed rectangles marking these alternatively spliced regions. A black line denotes the position of an upstream open reading frame (uORF) in At‐SR34a. Black upward and downward arrows indicate an increase or a decrease, respectively, in the ratio of splice variants encoding full‐length SR proteins. RS, arginine/serine.

Fig. 5

Arabidopsis thaliana serine/arginine‐rich (SR) protein At‐RS31 regulates alternative splicing of Target of Rapamycin (TOR)‐related genes. (a, c, d) Venn diagram comparisons showing the overlap between genes undergoing differential alternative splicing (DAS) in rs31‐1 or 35S::RS31, genes with At‐RS31 binding sites identified by individual‐nucleotide resolution crosslinking and immunoprecipitation (iCLIP), and TOR‐related genes. (a) Genes undergoing DAS in response to TOR inhibition; (c) genes encoding TOR‐phosphorylated proteins; (d) genes encoding proteins interacting with TORC1 components. (b, e) At‐RS31 modulates alternative splicing of CBL‐INTERACTING PROTEIN KINASE 3 (CIPK3) (b) and SNF1‐RELATED PROTEIN KINASE 2–8 (SnRK2.8) (e). Integrated Genomics Viewer tracks show iCLIP crosslink sites for green fluorescent protein (GFP) and RS31‐GFP, alongside RNA sequencing read coverage for rs31‐1, 35S::RS31, and wild‐type (WT) plants. The related tracks have the same scale. In transcript models, protein‐coding regions (from the translational start codon to the stop codon or premature termination codon) are shown in black. Dashed rectangles highlight regions undergoing DAS. A red vertical line marks the At‐RS31 dominant binding site in CIPK3. A black line indicates the position of an upstream open reading frame (uORF) in SnRK2.8. (b) Lower panel and (e) right panel show representative reverse transcriptase polymerase chain reaction (RT‐PCR) gel images of CIPK3 (b) and SnRK2.8 (e) splicing patterns in rs31‐1, 35S::RS31, and WT. Arrow heads indicate primer positions. CE, cassette exon; ref, reference transcript; RI, retained intron. Bar graphs show splicing index values quantified using imagej. Bars represent mean ± SE (n = 3). Different letters indicate statistically significant differences between genotypes (P < 0.05 by ANOVA followed by Fisher's LSD test for multiple comparisons). RS, arginine/serine.

Fig. 6

Arabidopsis thaliana serine/arginine‐rich (SR) protein At‐RS31 is involved in regulating abscisic acid (ABA) pathway. (a–d) Phenotypic analysis of rs31‐1, 35S::RS31 and wild‐type (WT) seedlings in response to different ABA concentrations. (a, c) Representative images of cotyledon greening (a) and root growth (c). (b) Percentage of seedlings with green and open cotyledons after 9 d. Data represent mean ± SE of three independent biological replicates (n = 25 seedlings per genotype per replicate). Different letters indicate statistically significant differences (P < 0.05 by ANOVA of transformed distribution followed by Tukey's post hoc test). (d) Root length measurements after 9 d. Violin plots show the range of the distribution of data across six biological replicates (n = 8–15 seedlings per replicate). The median is represented by the widest part of each plot. Individual points correspond to the lengths of individual roots. Different letters indicate statistically significant differences (ANOVA followed by Tukey's post hoc test). (e) Venn diagram comparison between genes undergoing differential alternative splicing (DAS) or differential expression (DE) in rs31‐1 or 35S::RS31, genes with At‐RS31 binding sites identified by individual‐nucleotide resolution crosslinking and immunoprecipitation (iCLIP), and ABA‐related genes. (f) At‐RS31 modulates alternative splicing of HAB1 (HYPERSENSITIVE TO ABA1). Integrated Genomics Viewer tracks show iCLIP crosslink sites for green fluorescent protein (GFP) and RS31‐GFP, alongside RNA sequencing read coverage for rs31‐1, 35S::RS31, and WT plants. The related tracks have the same scale. In the transcript models, protein‐coding regions (from the translational start codon to the stop codon or premature termination codon) are shown in black. Only transcripts relevant to the DAS events are shown, with dashed rectangles marking these alternatively spliced regions. Representative reverse transcriptase polymerase chain reaction (RT‐PCR) gels show HAB1 isoforms in the three genotypes. Colour‐coded arrowheads indicate primers. EI, exitron; ref, reference transcript; RI, intron retention. Bar graphs show splicing index values (mean ± SE, n = 3). Different letters indicate statistically significant differences (P < 0.05, ANOVA followed by Fisher's LSD test). RS, arginine/serine.