Abstract

Retrotransposons are self-replicating genetic elements and pervasive in mammalian genomes. Most insertions are passive bystanders of genome function. Yet, individual sequences have been adapted in diverse biological functions, from neurogenesis to placentation.

I will present my work on exonisation of Alu, LINE and LTR retrotransposons. I have previously demonstrated that retrotransposon sequences are poised with binding sites for splice-repressive proteins and that loss of these binding sites is a necessity for formation of novel, lineage-specific exons at retrotransposon-derived sequences. Alus and LIN Es are recognised through multivalent binding sites for splice-repressive proteins, and as a consequence, exonisation during genome evolution is a gradual process. More recently, I have catalogued extensive usage of LINE and LTR -elements as part of novel transcripts across the tumour types represented in the TCGA consortium data. These transcripts are recurrently and exclusively expressed in tumours, and I will present a proof-of-principle study that they generate antigenic peptides. Altogether, I conclude that thousands of past retrotransposon insertions are in a dormant state, and suggest that selective pressure acts on the host proteins and/or retrotransposon sequences to maintain repression in somatic cells.