Activation of Cryptic Donor Splice Sites within the UDP-Glucuronosyltransferase (UGT)1A First-Exon Region Generates Variant Transcripts That Encode UGT1A Proteins with Truncated Aglycone-Binding Domains

Abstract

The human UDP-glucuronosyltransferases (UGTs) have crucial roles in metabolizing and clearing numerous small lipophilic compounds. The UGT1A locus generates nine UGT1A mRNAs, 65 spliced transcripts and 34 circular RNAs. In this study, our analysis of published UGT-CaptureSeq datasets identified novel splice junctions that predict 24 variant UGT1A transcripts derived from ligation of exon 2 to unique sequences within the UGT1A first-exon region using cryptic donor splice sites. Of these variants, seven (1A1_n1, 1A3_n3, 1A4_n4, 1A5_n1, 1A8_n2, 1A9_n2v, 1A10_n7) are predicted to encode UGT1A proteins with truncated aglycone-binding domains. We assessed their expression profiles and deregulation in cancer using four RNA-seq datasets of paired normal and cancerous drug-metabolizing tissues from large patient cohorts. Variants were generally co-expressed with their canonical counterparts with a higher relative abundance in tumor than in normal tissues. Variants showed tissuespecific expression with high interindividual variability but overall low abundance. However, 1A8_n2 showed high abundance in normal and cancerous colorectal tissues with levels that approached or surpassed canonical 1A8 mRNA levels in many samples. We cloned 1A8_n2 and showed expression of the predicted protein (1A8_i3) in HEK293T cells. Glucuronidation assays with 4-MU showed that 1A8_i3 had no activity and was unable to inhibit the activity of 1A8_i1 protein. In summary, the activation of cryptic donor splice sites within the UGT1A first-exon region expands the UGT1A transcriptome and proteome. The 1A8_n2 cryptic donor splice site is highly active in colorectal tissues, representing an important cis-regulatory element that negatively regulates the function of the UGT1A8 gene through pre-mRNA splicing.