A statistically inferred microRNA network identifies breast cancer target miR-940 as an actin cytoskeleton regulator

International audienceMiRNAs are key regulators of gene expression. By binding to many genes, they create a complex network of gene co-regulation. Here, using a network-based approach, we identified miRNA hub groups by their close connections and common targets. In one cluster containing three miRNAs, miR-612, miR-661 and miR-940, the annotated functions of the co-regulated genes suggested a role in small GTPase signalling. Although the three members of this cluster targeted the same subset of predicted genes, we showed that their overexpression impacted cell fates differently. miR-661 demonstrated enhanced phosphorylation of myosin II and an increase in cell invasion, indicating a possible oncogenic miRNA. On the contrary, miR-612 and miR-940 inhibit phosphorylation of myosin II and cell invasion. Finally, expression profiling in human breast tissues showed that miR-940 was consistently downregulated in breast cancer tissues M icroRNAs are a class of endogenous, small (19–25 nucleotides), single-stranded non-coding RNAs that regulate gene expression in all eukaryotic organisms. In metazoans, microRNAs most commonly bind to the 39 untranslated region (39UTR) of their mRNA target transcript and cause translational repression and/or mRNA degradation. Every microRNA is predicted to regulate from a dozen to thousands of genes, including transcription factors. This fine-tuning of protein expression is known to be involved in many physiological processes, such as development, apoptosis, signal transduction and even cancer progression 1,2. More than 2,000 mature human microRNAs are listed in the 20 th release of miRBase: http://www.mirbase.org (2014) (Date of access:19/08/2013), and some authors hypothesise that the majority of human genes are regulated by microRNAs 3. Since their discovery in 1993 4 , a fair understanding of their role in animal development and in the onset and progression of diseases 2 , as well as of their potential use in therapies 5 , has been gathered. However, the cooperative behaviour of microRNAs is still under investigation. A growing body of experimental evidence suggests that microRNAs can regulate genes through complementarity, meaning that microRNAs can act together to regulate individual genes or groups of genes involved in similar processes 6. For example, Hu and co-workers demonstrated that transducing a cocktail of precursor microRNAs (miR-21, miR-24 and miR-221) can result in more effective engraftment of transplanted cardiac progenitor cells 7. Consistent with these discoveries, Zhu et al. demonstrated that miR-21 and miR-221 coregulate 56 gene ontology (GO) processes 8. In the same study, the authors also showed that cotransfection of miR-1 and miR-21 increases H 2 O 2-induced myocardial apoptosis and oxidative stress. These recent findings support the idea of microRNA-mediated cooperative regulation but also argue for the use of systemic approaches, notably based on graph theory, to decipher individual and complementary roles of microRNAs. Some work has been conducted to use recent high-throughput experiment-derived data sets to infer microRNA synergistic relationships 9–12. Herein, we present a microRNA network based on target similarities among microRNAs to infer clusters of microRNAs. Clusters are defined as groups of microRNAs sharing a set of common targets, predicted by either DIANA-microT v3 13 or TargetScan v6.2 14. Some authors have used GO enrichment analysis as a confirmatory tool for their clustering approach 11. In our case, GO enrichment is not used to infer networks but as a way to estimate the probable metabolic pathway(s) a cluster of microRNAs could co-regulate. Moreover, the novelty of our approach is to consider not only clusters of microRNAs but also OPE