ABSTRACT
Alternative splicing is an important regulatory process in eukaryotes. In plants, the major form of alternative splicing is intron retention. Despite its importance, the global impact of AS on the Arabidopsis proteome has not been investigated. In this study, we address this gap by performing a comprehensive integrated analysis of how changes in AS can affect the Arabidopsis proteome using mutants that disrupt ACINUS and PININ, two evolutionarily conserved alternative splicing factors. We used tandem mass tagging (TMT) with real-time search MS3 (RTS-SPS-MS3) coupled with extensive sample fractionations to achieve very high coverage and accurate protein quantification. We then integrated our proteomic data with transcriptomic data to assess how transcript changes and increased intron retention (IIR) affect the proteome. For differentially expressed transcripts, we have observed a weak to moderate correlation between transcript changes and protein changes. Our studies revealed that some IIRs have no effect on either transcript or protein levels, while some IIRs can significantly affect protein levels. Surprisingly, we found that IIRs have a much smaller effect on increasing protein diversity. Notably, the increased intron retention events detected in the double mutant are also detected in the WT under various biotic or abiotic stresses. We further investigated the characteristics of the retained introns. Our extensive proteomic data help to guide the phenotypic analysis and reveal that collective protein changes contribute to the observed phenotypes of the increased anthocyanin, pale green, reduced growth, and short root observed in the acinus pnn double mutant. Overall, our study provides insight into the intricate regulatory mechanism of intron retention and its impact on protein abundance in plants.
ABSTRACT
TurboID-based proximity labeling coupled to mass spectrometry (PL-MS) has emerged as a powerful tool for mapping protein-protein interactions in both plant and animal systems. Despite advances in sensitivity, PL-MS studies can still suffer from false negatives, especially when dealing with low abundance bait proteins and their transient interactors. Protein-level enrichment for biotinylated proteins is well developed and popular, but direct detection of biotinylated proteins by peptide-level enrichment and the difference in results between direct and indirect detection remain underexplored. To address this gap, we compared and improved enrichment and data analysis methods using TurboID fused to SPY, a low-abundance O-fucose transferase, using an AAL-enriched SPY target library for cross-referencing. Our results showed that MyOne and M280 streptavidin beads significantly outperformed antibody beads for peptide-level enrichment, with M280 performing best. In addition, while a biotin concentration ≤ 50 µM is recommended for protein-level enrichment in plants, higher biotin concentrations can be used for peptide-level enrichment, allowing us to improve detection and data quality. FragPipe's MSFragger protein identification and quantification software outperformed Maxquant and Protein Prospector for SPY interactome enrichment due to its superior detection of biotinylated peptides. Our improved washing protocols for protein-level enrichment mitigated bead collapse issues, improving data quality, and reducing experimental time. We found that the two enrichment methods provided complementary results and identified a total of 160 SPY-TurboID-enriched interactors, including 60 previously identified in the AAL-enriched SPY target list and 100 additional novel interactors. SILIA quantitative proteomics comparing WT and spy-4 mutants showed that SPY affects the protein levels of some of the identified interactors, such as nucleoporin proteins. We expect that our improvement will extend beyond TurboID to benefit other PL systems and hold promise for broader applications in biological research.
