Predicting the function of miRNA-mRNA networks

MicroRNAs (miRNAs) are short, non-coding, RNA regulators of gene expression that have been identified in a broad range of eukaryotes. In addition to regulating growth, development, differentiation, some miRNAs have also been classified as tumour suppressors or oncogenes. We have recently developed a systems biology approach to uncover large networks of interacting genes that are directly targeted by miRs and predict their function. This approach was first demonstrated by elucidating the mechanism behind hsa-miR-17-5p transforming potential. We are now expanding our endeavors to explore the mammalian miRNA and mRNA transcriptomes and search for novel functional miRNA-mRNA networks

Left: Functional enrichment for PicTar predicted targets of miR-17-5p and miR-20a from the miR-17-92 cluster. Arrows indicate the mean significance of randomly selected gene sets of equivalent size, and the grey boxes show ± 4 standard deviations. Right: Functional enrichment for 17-5p mRNA targets acting at the G1/S cell cycle boundary. This prediction was done without a priori knowledge of experimentally validated mRNA targets of miR-17-5p, and the mechanism of action was functionally validated.

All of our studies to date have focused on the population as being static despite all our previous transcriptomic analyses telling us it is anything but. In order to improve our ability to monitor genome wide networks more effectively, we need to integrate tissue specific expression and alternative splicing information when prediction miRNA targets.

Currently, there is no exhaustive transcriptome atlas of expression for mammalian tissues. We are currently completing such a body atlas using a combination of whole transcriptome sequencing, small RNA sequencing a CAGE to determine the tissue specific transcriptome content. The combined mRNA and miRNA data from these atlases data will be used to determine which miRNAs-targets are ubiquitous and which ones are tissue specific (ie in which tissues are miRNA and target mRNA both expressed). We will also be able to study the impact of transcriptional complexity on miR activity as it has been shown that alternate 3’ UTR usage can protect some transcripts from miRNA suppression through loss of miR binding sites.

Finally, it has been long established that clusters of mRNAs showing tight co-ordinate expression across multiple states, frequently work to drive a common biological outcome. We are actively investigating if co-ordinately expressed miRNAs also often drive common phenotypes, by targeting multiple mRNAs involved in the same pathway.