Final Report Summary - WHYMIR (Interplays between miRNAs and transcription factors in the determination and maintenance of cell identity)
The main aim of the ERC Starting Grant WHYMIR was to understand how the small changes that miRNAs induce at the molecular level translate into large phenotypic effects. We proposed to test three concrete hypotheses using specific experimental systems. These were that (1) miRNAs control gene expression noise, (2) the effects of miRNAs can be amplified by their direct targets that are also regulatory molecules, and (3) the RNA interference machinery may have other functions, beyond mRNA destabilization.
Although miRNAs are viewed as a means to buffer stochastic fluctuations (Hornstein and Shomron 2006), few studies have investigated how miRNA targets respond to the induction of miRNA expression at the individual cell level, and almost exclusively using reporter targets with multiple miRNA-complementary sites (Mukherji et al. 2011; Schmiedel et al. 2015). The reduction in protein expression noise has been related directly to the reduction in the average expression level, which is generally limited, except for reporters that carry multiple perfectly-complementary miRNA binding sites in their 3’ UTRs. In this project we characterized the response of individual mRNA targets to miRNA induction. As reported by previous studies, we found that miRNA targets have generally a limited response to the expression of the miRNA and that this is not due to an increased transcription rate masking a large effect on mRNA degradation. Furthermore, we combined single cell mRNA sequencing with computational modeling to develop a robust methodology for inferring the interaction parameters of a miRNA with all its mRNA targets, expressed at physiological levels in the environment of individual cells. We found that the limited response of miRNA targets to the miRNA carries to the level of single cells. Thus, the effect of miRNAs on gene expression noise is likely to be limited for the majority of miRNA targets. However, targets differ over two orders of magnitude in their sensitivity to the miRNA. These results provide the first comprehensive set of parameters of miRNA-target interactions in vivo, which will facilitate the interpretation of miRNA-dependent perturbation experiments. Furthermore, the approach is generalizable to other regulators of gene expression such as RNA-binding proteins.
In a second sub-project we investigate the interplay and "coordination" between miRNAs and transcription factors (TFs) in determining cell identity and differentiation, using miRNA-mediated reprogramming of somatic cells into iPSCs as a model system. Although we have not been able to reproduce somatic cell reprogramming by means of only miRNA over-expression, we have found, as reported before, that embryonically-expressed miRNAs enhance the efficiency of the process by ~3 fold. Analyzing the expression profile of embryonic stem cells that did and did not express miRNAs, we further identified key transcriptional regulators that are immediately downstream of embryonic miRNAs, and probably amplify the miRNA effects on gene expression. These transcription factors are involved in the regulation of the cell cycle, chromatin structure and innate immunity and their effects converge on suppressing the differentiation of embryonic stem cells. Our results are thus consistent with the hypothesis that transcription factors that are direct targets of miRNA propagate and amplify the effects of the miRNAs. Moreover these effects can be substantial, as demonstrated by the increased efficiency of somatic cell reprogramming in the presence of embryonic stem cell-specific miRNAs.
Finally, we have made an unexpected discovery, namely that splicing regulatory proteins, can also modulate the efficiency of pluripotency induction in somatic cells. Consistently, as our study was in progress, work from the Blencowe lab indeed showed that MBNL1/2 act as suppressors of reprogramming, their knock-down increasing the reprogramming efficiency. However, we have been able to demonstrate that the epithelial splicing regulatory proteins 1 and 2 (ESRP1/2) are equally important, their expression increasing the efficiency of somatic cell reprogramming to a level comparable to the miR-302/367 cluster of miRNAs.
Although miRNAs are viewed as a means to buffer stochastic fluctuations (Hornstein and Shomron 2006), few studies have investigated how miRNA targets respond to the induction of miRNA expression at the individual cell level, and almost exclusively using reporter targets with multiple miRNA-complementary sites (Mukherji et al. 2011; Schmiedel et al. 2015). The reduction in protein expression noise has been related directly to the reduction in the average expression level, which is generally limited, except for reporters that carry multiple perfectly-complementary miRNA binding sites in their 3’ UTRs. In this project we characterized the response of individual mRNA targets to miRNA induction. As reported by previous studies, we found that miRNA targets have generally a limited response to the expression of the miRNA and that this is not due to an increased transcription rate masking a large effect on mRNA degradation. Furthermore, we combined single cell mRNA sequencing with computational modeling to develop a robust methodology for inferring the interaction parameters of a miRNA with all its mRNA targets, expressed at physiological levels in the environment of individual cells. We found that the limited response of miRNA targets to the miRNA carries to the level of single cells. Thus, the effect of miRNAs on gene expression noise is likely to be limited for the majority of miRNA targets. However, targets differ over two orders of magnitude in their sensitivity to the miRNA. These results provide the first comprehensive set of parameters of miRNA-target interactions in vivo, which will facilitate the interpretation of miRNA-dependent perturbation experiments. Furthermore, the approach is generalizable to other regulators of gene expression such as RNA-binding proteins.
In a second sub-project we investigate the interplay and "coordination" between miRNAs and transcription factors (TFs) in determining cell identity and differentiation, using miRNA-mediated reprogramming of somatic cells into iPSCs as a model system. Although we have not been able to reproduce somatic cell reprogramming by means of only miRNA over-expression, we have found, as reported before, that embryonically-expressed miRNAs enhance the efficiency of the process by ~3 fold. Analyzing the expression profile of embryonic stem cells that did and did not express miRNAs, we further identified key transcriptional regulators that are immediately downstream of embryonic miRNAs, and probably amplify the miRNA effects on gene expression. These transcription factors are involved in the regulation of the cell cycle, chromatin structure and innate immunity and their effects converge on suppressing the differentiation of embryonic stem cells. Our results are thus consistent with the hypothesis that transcription factors that are direct targets of miRNA propagate and amplify the effects of the miRNAs. Moreover these effects can be substantial, as demonstrated by the increased efficiency of somatic cell reprogramming in the presence of embryonic stem cell-specific miRNAs.
Finally, we have made an unexpected discovery, namely that splicing regulatory proteins, can also modulate the efficiency of pluripotency induction in somatic cells. Consistently, as our study was in progress, work from the Blencowe lab indeed showed that MBNL1/2 act as suppressors of reprogramming, their knock-down increasing the reprogramming efficiency. However, we have been able to demonstrate that the epithelial splicing regulatory proteins 1 and 2 (ESRP1/2) are equally important, their expression increasing the efficiency of somatic cell reprogramming to a level comparable to the miR-302/367 cluster of miRNAs.