Integrated Transcriptome and Proteome Analyses Reveal Cell Type-Specific Regulation of Alternative Splicing in Arabidopsis

The extent to which alternative splicing produces cell type-specific isoforms in plants, and how these isoforms contribute to proteome diversity is largely unknown. To address these questions, we generated a comprehensive dataset of gene expression from individual cell types of the Arabidopsis root using short-read, paired-end RNA-seq. We complemented this analysis with long-read Pacific BioSciences (PacBio) sequencing and quantitative mass spectrometry. From these three platforms, we identified > 24,000 new splice isoforms and >1000 non-coding RNAs and detected > 12,000 peptides from over >5,000 proteins. Key insights from the analysis of these extensive data are: 1) Alternative splicing between cell types is rarely a binary process, but results in a difference in degree between major and minor isoforms; 2) Alternative splicing tends to remove parts of coding regions from a longer, major isoform, providing evidence for a progressive mechanism of splicing; 3) Intron retention of evolutionarily conserved introns suggests a possible origin for a common form of alternative splicing; 4) Coordinated alternative splicing appears to play a role in regulating differentiation in the root; and 5) For the majority of detected proteins, peptide abundance correlates well with major isoform abundance in different developmental stages. To characterize the complex regulatory process of alternative splicing, we have developed CoSpliceNet, a computational pipeline that integrates co-expression and motif discovery at splicing regulatory regions to infer co-splicing networks. CoSpliceNet can be used for identifying candidate regulators of splicing events and predict RNA-binding motifs in any biological system.