A topologically conserved unstructured region helps positioning the evolutionarily conserved Prp40 WW domains to promote non-canonical intron splicing

Abstract

Newly transcribed introns are immediately identified by early splicing factors that recognize the intron 5 splice site (5SS) and branch site (BS). In the budding yeast, these critical splice-site sequences are generally constrained, whereas degeneracy is the rule in higher eukaryotes. Yet, [~]40% of the yeast introns do diverge, to a certain degree, from the canonical sequences. Exactly how these non-canonical introns are recognized and spliced remains unknown. Here we show that the conserved Prp40 WW domains promote non-canonical intron splicing by enhancing stable U1 snRNP and BBP recruitments. AlphaFold predicts a topologically conserved unstructured region between Prp40 WW and FF domains. Alignment of the AlphaFold Prp40 structure with published U1 snRNP structure positions WW domains adjacent to 5SS and Luc7, which is known to be critical for 5SS recognition. Indeed, deletion of this unstructured region negatively impacts on splicing of the non-canonical 5SS introns. Taken together, our results suggest that the conserved WW domains may have evolved to deal with the highly degenerate 5SS and BS sequences in higher eukaryotes, so as to accommodate increased splicing complexity. HighlightsO_LIThe evolutionarily conserved Prp40 WW domains promote splicing of introns harboring non-canonical 5 splice site or branch site. C_LIO_LIPrp40 WW domains enhance stable U1 snRNP and BBP recruitments to nascent transcripts containing non-canonical splice sites. C_LIO_LIA topologically conserved unstructured region between WW and FF domains helps to position Prp40 WW domains close to the 5 splice site. C_LIO_LIThe N-terminal WW domain sterically hinders conformational rearrangements required for efficient release of a BBP variant during spliceosome assembly. C_LIO_LIA reporter assay identified 13 non-canonical introns whose splicing, under various environmental conditions, depend on Prp40 WW domains. C_LI