MicroRNAs Reduce Gene Expression Noise

MicroRNAs regulate a large number of genes in metazoan organisms by accelerating mRNA degradation and inhibiting translation. Although the physiological function of some microRNAs is known in detail, it is not clear why microRNA regulation is so ubiquitous and conserved, since individual microRNAs only weakly repress the vast majority of their target genes and knockouts rarely result in mutant phenotypes. Compelling reasons for this widespread regulation have been suggested, including the ability of microRNAs to provide robustness to gene expression - e.g. by buffering stochastic variations in gene expression. Here we use mathematical modeling and a single cell reporter assay to quantify the effect of microRNA regulation on protein expression variability, or noise, in individual murine embryonic stem cells. We find that microRNA regulation decreases noise for lowly expressed - inherently variable - targets, whereas it increases noise for highly expressed targets. We show that decreased noise results from a reduction in intrinsic noise, which is the dominant noise for lowly expressed genes. However, additional extrinsic noise propagates from the microRNA-pool to the regulated gene and can tip the overall balance towards increased noise at high expression levels. We find that this propagated microRNA-pool noise can be effectively dampened if the target is regulated by multiple independently transcribed microRNAs, and we can verify this for two endogenous microRNA targets. Our findings show how microRNAs confer robustness to gene expression and offer a plausible explanation for the commonly observed combinatorial targeting of endogenous genes by multiple microRNAs.