Intron architecture predicts chromatin features in Arabidopsis thaliana

Introns are ubiquitous yet enigmatic features of eukaryotic genomes. They are reported to affect gene expression regulation but the mechanisms remain largely unresolved. Here we investigate the connections between intron architecture and chromatin features - histone marks, histone variants, DNA methylation, and gene expression patterns in Arabidopsis thaliana. We found that first intron positions predict active chromatin marks found near transcription start sites while the number of introns explains different chromatin marks enriched in the core of gene bodies in broadly expressed genes. Notably, gene body chromatin marks exhibit a positional gradient distribution across the ordinal position of exons and introns. We tested these relationships by comparing recent gene duplicates with diverged intron architectures, and confirmed that intron number is positively associated with H3K4me1, H3K36me3, H2A.X, meCG, and broad expression across tissues. Our results suggest two distinct mechanisms in which plant intron architecture may affect chromatin states and ultimately gene expression, motivating future experiments: intron positions early in genes may affect the establishment of activating histone marks around transcription start sites, while a greater total number of introns may increase the number of intronic motifs and gene length in general, allowing for the increased accumulation of gene-body-associated chromatin features. Short summaryHere we investigate the association between intron architecture and chromatin features in Arabidopsis thaliana. We found that the position of the first intron is negatively correlated with marks found near the transcription start site and associated with active expression. We also found that the number of introns within genes is correlated with marks found in gene bodies and are associated with actively transcribed genes.