Genomic determinants of tissue-specific splicing in the human genome

Abstract

Cellular differentiation into various tissues is a critical developmental feature of multi- cellular organisms. One of the key gene regulatory processes that play a vital role in tissue differentiation is RNA splicing. Splicing differs between tissues, generating a multitude of mRNA isoforms that encode diverse proteins across tissues. Despite a substantial understanding of RNA splicing, very little is known about the genomic determinants of tissue-specific splicing. Until recently, there have been no systematic ways to analyse splice-sites that are regulated in a tissue-specific manner at a scale that would allow deciphering genome-wide patterns. This is primarily due to methods that focus on mRNA isoforms or splicing events as opposed to analysis at the level of splice sites, whose selection primarily determines the regulation of splicing. This thesis makes an attempt to exploit splice-site level analysis to address determinants of tissue-specific splicing. Here, we quantified the usage of individual splice sites for genes expressed in the human testis from GTEx data, revealing significant variation between individuals. Using splice-site strength as a phenotype, we conducted over 130,000 Genome-Wide Association Studies (GWAS) to identify genetic variation associated with differential splice-site usage. We compared our results from testis with previous analyses carried out in the heart to identify and catalogue splice sites that are utilised in a tissue-specific manner. By motif-enrichment analysis, we reveal several genomic motifs that are enriched among splice sites that are used in a tissue-specific manner both in heart and in testis. This thesis presents these findings in both tissue-specific and evolutionary contexts of genomic determinants of splicing variation. Our studies suggest that tissue-specific splicing is a function of tissue- specific gene expression. We also reveal that genetic variation can affect splicing in a tissue-specific manner, which has potential implications for human disease.