MicroRNAs (miRNAs) are short regulatory RNAs that are encoded in the genome. They act on messenger RNAs (mRNAs), which they recognize based on sequence complementarity, reducing the rate of protein translation and inducing to some extent the degradation of these target mRNAs. Hundreds of miRNAs have been found in the human genome, and computational models estimate that a miRNA targets on average hundreds of mRNAs. It is thus clear that miRNAs are part of extensive regulatory networks. An interesting question that emerges now is what factors can modulate the effects that miRNAs have on their targets.
Various analyses of miRNA targets suggested that many mRNAs contain more than one binding site for a single miRNA or can be recognized by several different miRNAs that are simultaneously expressed. This observation prompted the speculation that cross-talk between multiple complexes containing miRNAs, assembled on the same mRNA molecule, may increase the robustness of the translation regulatory response. Several studies have also reported that the activity of miRNAs can be modulated by RNA-binding proteins: HuR has been shown to relieve the miR-122-dependent inhibition of the cationic amino acid transporter 1 message under stress, while the dead-end protein (Dnd1) was shown to inhibit the access of miRNAs to their target sites in the primordial cells of zebrafish. In this study we will combine analyses of experimental data with computational modeling in order to determine such modulatory interactions and to understand the nature of the selection pressures that act on mRNAs in evolution.