A Comprehensive Bioinformatic Analysis of RNA-seq Datasets Reveals a Differential and Variable Expression of Wildtype and Variant UGT1A Transcripts in Human Tissues and Their Deregulation in Cancers

Simple Summary UGT enzymes metabolize and detoxify numerous small molecules that are important in cancer, including carcinogens, endogenous growth regulators, and anticancer drugs. Alternatively spliced UGT transcripts can encode truncated proteins that inhibit canonical UGTs, thus reducing detoxification activity. We assessed the expression of specific variant transcripts, designated as UGT1A_v2 and _v3, in six different cancers using RNA-seq datasets with large cohorts of paired normal and tumor tissues. Our results show high interindividual variation in v2 and v3 transcript abundance, as well as tissue- and tumor-specific expression patterns. These findings suggest that the variants have tissue-specific impacts on glucuronidation and may have a more significant role in tumors than in normal tissues. The high interindividual variability is likely relevant to differing personalized drug metabolisms through the UGT conjugation pathway. Finally, our discovery of novel UGT1A variant transcripts further highlights the diversity of the UGT1A transcriptome and proteome.Abstract The UGT1A locus generates over 60 different alternatively spliced transcripts and 30 circular RNAs. To date, v2 and v3 transcripts are the only variant UGT1A transcripts that have been functionally characterized. Both v2 and v3 transcripts encode the same inactive variant UGT1A proteins (i2s) that can negatively regulate glucuronidation activity and influence cancer cell metabolism. However, the abundance and interindividual variability in the expression of v2 and v3 transcripts in human tissues and their potential deregulation in cancers have not been comprehensively assessed. To address this knowledge gap, we quantified the expression levels of v1, v2, and v3 transcripts using RNA-seq datasets with large cohorts of normal tissues and paired normal and tumor tissues from patients with six different cancer types (liver, kidney, colon, stomach, esophagus, and bladder cancer). We found that v2 and v3 abundance varied significantly between different tissue types, and that interindividual variation was also high within the same tissue type. Moreover, the ratio of v2 to v3 variants varied between tissues, implying their differential regulation. Our results showed higher v2 abundance in gastrointestinal tissues than liver and kidney tissues, suggesting a more significant negative regulation of glucuronidation by i2 proteins in gastrointestinal tissues than in liver and kidney tissues. We further showed differential deregulation of wildtype (v1) and variant transcripts (v2, v3) in cancers that generally increased the v2/v1 and/or v3/v1 expression ratios in tumors compared to normal tissues, indicating a more significant role of the variants in tumors. Finally, we report ten novel UGT1A transcripts with novel 3 ' terminal exons, most of which encode variant proteins with a similar structure to UGT1A_i2 proteins. These findings further emphasize the diversity of the UGT1A transcriptome and proteome.