The construction of an annotation and database system for alternative splicing and its application in the study of different modes of splicing
Over half the genes in the human genome are alternatively spliced, with similar observations reported for other eukaryotic organisms. In addition to being implicated in disease and development, alternative splicing also plays a role in increasing proteome diversity. One gene has the potential to generate thousands of transcripts. For example, the Drosophila gene Dscam has the potential to code for ∼32,000 transcripts, which is approximately twice the number of genes in its genome. Because of its biological importance, it is important to understand the evolution and regulatory mechanisms of alternative splicing.
Previous reports observed distinguishing characteristics differentiating alternative and constitutive splicing. Alternative splice sites are weaker than constitutive splice sites, alternative exons and constitutive exons have different length distributions, and orthologous alternative exons are more conserved than orthologous constitutive exons. However less is known about individual types of alternative exons. Skipped exons maintain the reading frame of a transcript at a higher frequency than constitutive exons. Approximately 5% of skipped exons have also been reported to originate from Alu elements within intronic sequences, a mechanism termed “exonization”. Mutually exclusive exons, however, are reported to originate from tandem exon duplications. These findings suggest that distinct modes of alternative splicing may differ from each other.
To better understand distinct modes of alternative exons, I conducted a detailed analysis on different modes of alternatively spliced exons. The first step in this process consisted of constructing the MAASE database, a database system to annotate, store and organize information on different modes of alternative splicing. With the highly curated information in MAASE, I confirmed many previously reported characteristic distinctions between constitutive and alternative splicing such as shorter alternative exon lengths and weaker splicing signals at alternative splice site. Results also revealed novel distinctions between constitutive and alternative splicing and among different modes of alternative splicing. For example, a hierarchical order of splice site strength exists among the different modes, a trait that is consistent in human and mouse. Furthermore, results show that different modes of alternative exons may have evolved from distinct paths and may be regulated by diverse mechanisms.