An overwhelming majority of eukaryotic introns have GT/AG ends, whose identities play a critical role for their recognition and removal by the U2 spliceosome, a well-conserved complex of protein and RNAs. Introns with other splice sites exist at very low frequencies in various genomes, and some of them are processed by the U12 spliceosome. Here, we show that, in the chordate Fritillaria borealis, the majority of old introns have been lost and replaced by introns with highly divergent splice sites. The new introns of F. borealis are exceptionally diverse, though more frequently AG/AC or AG/AT, and features of thousands of them support an origin from transposons. They cannot be processed in human cells, but their splicing is rescued by mutating terminal dinucleotides to GT/AG. With lariat sequencing and splicing inhibitor assays, we show that F. borealis introns are spliced by the U2 spliceosome, which thus evolved to a different selectivity, with neither novel U1 small nuclear RNA (snRNA) types nor major remodeling of its protein and snRNA complements. This genome-wide recolonization by non-canonical introns emphasizes the importance of transposons as a resource of novel introns in a context of massive intron loss. An evolution of the spliceosome may also permit to neutralize harmful transposons through their conversion into introns.
Keywords: intron evolution; non-canonical introns; splicing; transposon.
Copyright © 2019 Elsevier Ltd. All rights reserved.