The Origins of Spliceosomal Ribozymes

Human introns are active ribozymes that assemble with three small RNA molecules (U2, U6 and U5) at the core of the spliceosome. Our introns appear structurally homologous to prokaryotic mobile ribozymes termed Group II introns. Presented side by side, RNA secondary structures of human and bacterial introns highlight conserved tertiary links essential for ribozyme folding. The RNA splicing mechanism took time to evolve; there are three principal ways preserved in cellular life, of which lariat ribozymes that include Group II and spliceosomal introns appear the winning design. Splicing requires two chemical reactions: the challenge for the ribozyme is to reposition intermediates, which is achieved by a tertiary link between the intron ends. Crucially, splicing is reversible; target sites for reverse splicing are recognised by base-pairing with Group II intron RNA. In prokaryotic genomes, Group II introns are rare self-splicing elements, while the eukaryotic spliceosome evolved as a universal splicing catalyst for the multitude of unique introns in the whole transcriptome. How did this come about? In eukaryogenesis, due to small population size introns escaped selective pressure. Protosplice site repeats CAG|GU, preserved as structural elements of the ribozyme at both 5’ and 3’ss of eukaryotic introns today, indicate that the successful intron progenitor was able to reconstruct its insertion site upon each integration. The proliferation of introns was instrumental to shaping eukaryotic genome and radiation of eukaryotic functions, essential for transition to multicellularity. Conversely, monocellular eukaryotes lost most of their introns and reduced spliceosomal complexity. I focus on the evolution of splice site recognition. Group IIA introns define splice sites exclusively by pairing with their exons. Spliceosomal introns reduced 5’exon pairing and instead use intron recognition sites as intramolecular joints with snRNAs. Spliceosomal and Group IIA introns share RNA recognition loops aligning exons for ligation. These loops are guided by Watson-Crick base-pairing using isosteric pairs to support exon diversity. The exon consensus is lost for a few remaining introns of S. cerevisiae. By contrast, positional dependencies at human splice sites, that feature well-preserved exon consensus, allowed us to uncover U5 snRNA role in exon recognition.