1To characterize the rules governing exon recognition during splicing, we analysed dozens of RNA-seq 2 datasets and identified ~3,200 GC-rich exons and ~4,000 AT-rich exons whose inclusion depends on 3 different sets of splicing factors. We show that GC-rich exons have predicted RNA secondary 4 structures at 5′-ss, and are dependent on U1 snRNP-associated proteins. In contrast, AT-rich exons 5 have a large number of branchpoints and SF1-or U2AF2-binding sites and are dependent on U2 6 snRNP-associated proteins. Nucleotide composition bias also influences local chromatin 7 organization, with consequences for exon recognition during splicing. As the GC content of exons 8 correlates with that of their hosting genes, isochores and topologically-associated domains, we 9 propose that regional nucleotide composition bias leaves a local footprint at the exon level and 10 induces constraints during splicing that can be alleviated by local chromatin organization and 11 recruitment of specific splicing factors. Therefore, nucleotide composition bias establishes a direct 12 link between genome organization and local regulatory processes, like alternative splicing. 13 14 Keywords: splicing; genomic; chromatin organization; nucleotide composition bias 15 genes than AT-rich isochores, and genes in GC-rich isochores contain smaller introns than genes in 1 AT-rich isochores [21][22][23] . It has been proposed that splicing of short introns in a GC-rich context may 2 occur through the intron definition model, while the splicing of large introns in an AT-rich context 3 may occur through the exon definition model 10,24 . Collectively, these observations support a model in 4 which the gene architecture (e.g., size of introns) and gene nucleotide composition bias (e.g., GC or 5 AT content) influence local processes at the exon level, such as nucleosome positioning and intron 6 removal. As exon recognition also depends on the binding to the pre-mRNAs of splicing factors that 7 interact with compositionally-biased sequences, one interesting possibility is that the nature of these 8 splicing factors depends at least in part on the gene nucleotide composition bias. In this setting, we 9 have recently reported that exons regulated by different splicing factors have different nucleotide 10 composition bias 25 . 11Here, we have investigated the relationship between the splicing process, gene nucleotide 12 composition bias and chromatin organization at both the local and global levels. We initially 13 identified sets of exons activated by different splicing factors and then demonstrated that analysing 14 the nucleotide composition bias provided a better understanding of the interplay between chromatin 15 organization and splicing-related features, which collectively affect exon recognition. We propose 16 that nucleotide composition bias not only contributes to the 1D and 3D genome organization, but 17 has also local consequences at the exon level during the splicing process. 18 19 20
Results
21Splicing factor-dependent GC-rich and AT-rich ex...