Post-transcriptional Regulation is the Major Driver of microRNA Expression Variation
MicroRNA (miRNA) expression patterns are highly variable across human tissues and across cancer specimens. The intuitive assumption is that transcription is the main contributor to mature miRNA expression patterns, with post-transcriptional processes further modifying miRNA expression levels. Here we report the surprising model that, on the global level, post-transcriptional regulation dominates over transcriptional regulation in determining mature miRNA expression patterns in both normal
... n both normal tissues and cancer. Taking advantage of large genomic datasets in which the expression of both mature miRNAs and their host genes have been quantified, we establish and validate transcriptional and post-transcriptional metrics, with miRNA host gene expression estimating transcriptional regulation and mature miRNA to host gene ratio estimating post-transcriptional regulation. On average, the post-transcriptional metric contributes 2.8-fold more than the transcriptional metric to the variance of mature miRNA expression. The variation of the balance between the two mature miRNAs (5p and 3p miRNAs) produced from the same precursor hairpin is a non-negligible contributor to miRNA expression, explaining ~27% of the variance of miRNAs' post-transcriptional metric. Data of normal tissues yield similar results as cancer specimens. Additionally, the post-transcriptional metric is superior to the transcriptional metric in classifying cancer types. We further demonstrate that the post-transcriptional metric separates miRNAs into distinct groups, suggesting that there are groups of miRNAs that are co-regulated on the post-transcriptional level. Our data support a model in which the post-transcriptional regulation is the major driver of miRNA expression variation, and paves a way toward better mechanistic understanding of post-transcriptional regulation of mature miRNA expression.