Selection preserves Ubiquitin Specific Protease 4 alternative exon skipping in therian mammals

Ubiquitin specific protease 4 (USP4) is a highly networked deubiquitinating enzyme with reported roles in cancer, innate immunity and RNA splicing. In mammals it has two dominant isoforms arising from inclusion or skipping of exon 7 (E(7)). We evaluated two plausible mechanisms for the generation of these isoforms: (A) E(7) skipping due to a long upstream intron and (B) E(7) skipping due to inefficient 5′ splice sites (5′SS) and/or branchpoint sites (BPS). We then assessed whether E(7) alternative splicing is maintained by selective pressure or arose from genetic drift. Both transcript variants were generated from a USP4-E(7) minigene construct with short flanking introns, an observation consistent with the second mechanism whereby differential splice signal strengths are the basis of E(7) skipping. Optimization of the downstream 5′SS eliminated E(7) skipping. Experimental validation of the correlation between 5′SS identity and exon skipping in vertebrates pinpointed the +6 site as the key splicing determinant. Therian mammals invariably display a 5′SS configuration favouring alternative splicing and the resulting isoforms have distinct subcellular localizations. We conclude that alternative splicing of mammalian USP4 is under selective maintenance and that long and short USP4 isoforms may target substrates in various cellular compartments.