Junction, what's your function? Mapping the exon junction complex in Arabidopsis Authors: Neeta Lohani, Meenu Singla-Rastogi Published: 2026 In most eukaryotic genes, the protein-coding exons are interrupted by introns. These introns are cut out of the precursor mRNA by the spliceosome, a large molecular machine that then stitches the flanking exons together. But intron splicing leaves behind more than just properly joined exons. At each exon–exon junction, the spliceosome deposits a protein assembly called the exon junction complex (EJC), creating a molecular mark that stays with the mRNAs throughout their life cycle. In animals, the EJC orchestrates several aspects of RNA metabolism, from nuclear export and translation to nonsense-mediated decay, the surveillance pathway that removes transcripts carrying premature stop codons (He and Jacobson 2015; Woodward et al. 2017). Whether plants use the EJC in a similar manner has remained an open question, largely because technical limitations have prevented genome-wide mapping of EJC deposition in vivo. In their latest study, Kaiwen Sun and colleagues overcome these challenges and provide the first high-resolution map of the Arabidopsis EJC, revealing that plants assemble the complex in a conserved yet functionally distinct manner (Sun et al. 2026). To understand where the plant EJC is deposited, the authors mapped the binding sites of eIF4AIII, the RNA-binding core of the complex, using enhanced crosslinking and immunoprecipitation sequencing (Fig. 1). The resulting map positioned eIF4AIII approximately 25 to 30 nucleotides upstream of exon–exon junctions, confirming earlier indirect predictions from RNA degradome analyses (Lee et al. 2020). Notably, the plant complex sits a little further upstream than in animals, where it binds 20 to 24 nucleotides from the junction (Le Hir et al. 2000). More surprisingly, the authors identified a second binding peak at the 3′ end of introns, a region contacted by spliceosome components during intron removal. This suggests that eIF4AIII first associates with the spliceosome-engaged intron before it is stable at the exon junctions. This observation hinted that EJC assembly might occur progressively during co-transcriptional splicing rather than as a single post-splicing event. EJC deposition during co-transcriptional splicing. Genome-wide distribution of eIF4AIII crosslink sites, showing a peak ∼25 to 30 nucleotides upstream of exon–exon junctions (bottom) and enhanced crosslinking and immunoprecipitation sequencing tracks at 2 representative genes (above). This figure includes panels 1d and 1e of Figure 1 of Sun et al. (2026) and is generated using BioRender. To test this hypothesis, the authors inhibited splicing with herboxidiene (an inhibitor of splicing that stalls spliceosome assembly at an early stage) and repeated the eCLIP analysis. Blocking splicing abolished the junction-associated peak and increased the intronic peaks, demonstrating that eIF4AIII is recruited early during spliceosome assembly and only becomes stably positioned after exon ligation. Interestingly, both EJC abundance and binding stability increased with exon number, indicating that transcript architecture influences not only where the complex is deposited but also how long it remains associated with mature mRNAs. Together, these experiments provide a clear picture of the mechanism of stepwise EJC assembly in plants. Having established how the EJC is deposited, the authors next asked what role it plays in RNA processing. Using an inducible artificial microRNA to reduce eIF4AIII transcript expression, they observed widespread changes in intron retention throughout the transcriptome, meaning many introns were left in the mRNA instead of being cleanly removed. The direct targets of eIF4AIII were among the most affected transcripts, indicating that these splicing defects arise primarily from loss of EJC function rather than secondary changes in gene expression. To distinguish whether the complex promotes splicing or instead marks faulty transcripts for degradation, the authors compared these introns with those that accumulate in the upf3 and hen2 mutants, which are defective in nonsense-mediated decay, and in the nuclear RNA exosome that degrades transcripts in the nucleus, respectively. Only a small proportion overlapped, and introns regulated by eIF4AIII alone remained largely unchanged when translation was blocked, unlike those co-regulated by UPF3. These findings suggest that eIF4AIII functions predominantly in promoting efficient pre-mRNA splicing rather than in directing RNA decay. Like other components of the spliceosome, the plant EJC therefore appears to actively participate in RNA maturation rather than simply marking completed splice junctions. The authors appropriately note that this conclusion has important caveats. Their analysis of premature stop codon–triggered decay focused on 2 CRISPR-edited alleles and used inducible knockdown rather than complete loss of eIF4AIII, and earlier work in Nicotiana benthamiana has shown that an intron downstream of a stop codon can still promote EJC-dependent nonsense-mediated decay (Nyikó et al. 2013). Rather than replacing the canonical model, the study of Kaiwen Sun and colleagues refines it, showing that the EJC's contribution to RNA surveillance in plants depends on transcript context and may fundamentally differ from its role in animals. In doing so, this first genome-wide portrait of EJC deposition in plants lays a foundation for understanding how transcript architecture, EJC deposition, and RNA quality-control pathways together determine mRNA fate. Recent related articles in The Plant Cell: Zhou et al. (2026) demonstrated that the Arabidopsis SF3B2 homolog AtSF3B2 interacts with the splicing factors AtU2AF65B and AtSYF2 and modulates the splicing of its target genes, revealing transcriptional and post-transcriptional roles in the flowering transition. Zhou et al. (2026) demonstrated that the Arabidopsis SF3B2 homolog AtSF3B2 interacts with the splicing factors AtU2AF65B and AtSYF2 and modulates the splicing of its target genes, revealing transcriptional and post-transcriptional roles in the flowering transition. Huang et al. (2025) showed that the histone methylation reader MRG2 interacts with eIF4A3 to enhance its recruitment and to regulate alternative splicing and circadian rhythms, providing a chromatin-level link to EJC loading. Huang et al. (2025) showed that the histone methylation reader MRG2 interacts with eIF4A3 to enhance its recruitment and to regulate alternative splicing and circadian rhythms, providing a chromatin-level link to EJC loading. Szaker et al. (2025) found that the transcription elongation factor TFIIS safeguards splicing accuracy and works alongside nonsense-mediated decay to clear premature stop codon transcripts arising from intron retention and 3′UTR lengthening under heat stress, tying splicing fidelity to transcript surveillance. Szaker et al. (2025) found that the transcription elongation factor TFIIS safeguards splicing accuracy and works alongside nonsense-mediated decay to clear premature stop codon transcripts arising from intron retention and 3′UTR lengthening under heat stress, tying splicing fidelity to transcript surveillance. Cai et al. (2024) identified the m6A reader ECT8 as a stress sensor that accelerates decay of its bound transcripts through the decapping machinery, a reminder that RNA modifications and the EJC both steer cytoplasmic mRNA fate. Cai et al. (2024) identified the m6A reader ECT8 as a stress sensor that accelerates decay of its bound transcripts through the decapping machinery, a reminder that RNA modifications and the EJC both steer cytoplasmic mRNA fate.