A novel minor spliceosome component required for splicing of AT-AC introns

U12-type introns were originally recognized based on their highly conserved non-consensus AT-AC termini1,2, which are spliced by a separate minor spliceosome3,4. Padgett and Krainer groups later showed that terminal dinucleotides do not differentiate U12-type from U2-type introns, as there are U12-type introns with GT-AG termini and U2-type introns with AT-AC termini5,6. Rather, U12-type introns are recognized by their divergent and highly conserved 5' splice site (5'ss) and branch point sequences, which both differ from the consensus sequences found in U2-type introns. To date, no functional differences have been ascribed to AT-AC or GT-AG subtypes of U12-type introns, nor have RNAseq analyses of minor spliceosome diseases reported any subtype specificity. Here, we describe a novel protein component of the minor spliceosome, encoded by the CENATAC locus, that is required for accurate splicing of AT-AC but not GT-AG type minor introns. CENATAC was initially identified in a subset of Mosaic Variegated Aneuploidy (MVA) patients with mutations in CENATAC, which lead to chromosome congression defects during mitosis. Earlier large-scale proteomic analyses tentatively classified CENATAC as a spliceosome component and phylogenetic analyses showed co-segregation of CENATAC with minor spliceosome components. Targeted depletion of CENATAC in HeLa cells, followed by RNAseq revealed global retention of AT-AC minor subtype introns with more than 60% showing statistically significant (up to 90%) intron retention (IR). Additionally, U12-type introns with 5'-AT, but divergent 3'-terminal dinucleotides also showed significant IR. We also detected cryptic U2-type splice site activation near affected AT-AC introns. In contrast, about 10% of GT-AG subtype introns responded to CENATAC depletion. Co-IP experiments revealed that CENATAC is not a U11/U12 di-snRNP component as expected for a specificity factor, but rather associates with the U4atac/U6atac.U5 tri-snRNP via interaction with PRPF3/4, suggesting a role for minor tri-snRNP in initial 5'ss recognition. CENATAC also interacts with TXNL4B, a paralog of TXNL4A in the major tri-snRNP. Finally, several genes encoding chromosome congression factors harbor U12 AT-AC-type introns that were highly retained in CENATAC depleted cells, potentially explaining the aneuploidy phenotype observed in MVA patients. Overall design: Targeted depletion of CENATAC gene in HeLa cells, followed by RNAseq.