A genomic view of alternative splicing: Function, regulation and evolution

High-throughput genomic data such as expressed sequence tags (ESTs) and microarrays have provided a genome-wide picture of alternative splicing in multiple organisms, while whole genome sequences and comparative genomics have made it possible to "look backwards" into the detailed evolutionary history of any specific alternative splice site or exon. We have performed large-scale computational analyses of alternative splicing in multiple mammalian genomes, in order to understand the functional impact, evolutionary history and regulation of alternative splicing.; Evidence of abundant transcript variation (such as alternative splicing) in complex genomes indicates that cataloging the complete set of transcripts from an organism is an important project. One challenge is the fact that most high-throughput experimental methods (such as EST sequencing, or microarray probe hybridization) give highly detailed information about short fragments of transcripts or protein products, instead of a complete characterization of full-length forms. Assembling full-length transcript and protein isoforms from sequence fragments is critical for assessing the functional impact of splice variants. We analyze this "multiassembly problem"-reconstructing the most likely set of full-length isoform sequences from a mixture of EST fragment data-and present a graph-based algorithm for solving it. Using a combined graph traversal and filtering method, we generated the ASP database of high-quality alternatively spliced protein isoforms in human. We demonstrate that this method performs robustly on a wide variety of genes whose protein isoforms have been surveyed experimentally. We illustrate two particularly exciting examples, showing the effect of alternative splicing on domain interactions of the proteome, and on transmembrane anchoring of membrane-bound receptor proteins.; Another interesting topic about alternative splicing is its role in the evolution of genomes. Recently, it was proposed that alternative splicing may act as a mechanism for opening accelerated paths of evolution, by reducing negative selection pressure against large-scale alterations of gene structure such as exon creation or loss. This hypothesis raises a number of questions. First, has this mechanism actually created new exons in mammalian genomes? The original analysis showed that many alternatively spliced exons are species-specific (i.e. present in the human genome but not mouse, or vice versa), but assessing whether this is due to recent exon creation (as opposed to exon loss) requires using outgroups to infer the ancestral genotype. Second, has neutralization of negative selection by alternative splicing produced actual adaptive benefit during mammalian evolution? How can we assess whether alternatively spliced exons show both evidence of "faster evolution" than other exons, and at the same time strong selection pressure that is indicative of biological function? Third, protein reading frame preservation (alternative splicing that adds or removes an exact multiple of 3nt to the transcript, altering the protein sequence in a strictly modular way) has emerged as an important criterion for functional alternative splicing. What is the evolutionary role of such "modular" alternative splicing events, and what distinguishes them from other alternative splicing events? We investigated the evolution of alternative splicing in mammalian genomes. Our data show that alternative splicing can produce a striking acceleration in new exon creation. Intriguingly, conserved alternatively spliced exons show a remarkable reduction in their silent substitution rate, indicating widespread "RNA-level selection pressure" associated with regulation of alternative splicing.